f- -v^ir Trt^VT^fc S~*\Y~* T T"7* f~** rTpir^ T' IYDROEIJECTRI Hfiitlfj/Mffi DEVELOP iff fli UfltW E. & M E. .DO NOT TAKE FROM ROOM GIFT OF ASSOCIATED ELECTRICAL AND MECHANICAL ENGINEERS MECHANICS DEPARTMENT Library tbe same autbor STEAM-ELECTRIC POWER PLANTS A Practical Treatise on the Design of Central Light and Power Stations and their Economical Construction and Oper- ation. 11x8 Inch. 473 Pages. 500 Figures Net, $5.00 (See Last Page) A.E.&M HYDROELECTRIC L UNIV - of CAI DEVELOPMENTS AND ENGINEERING A PRACTICAL AND THEORETICAL TREATISE ON THE DEVELOPMENT, DESIGN, CONSTRUCTION, EQUIPMENT AND OPERATION OF HYDRO- ELECTRIC TRANSMISSION PLANTS BY FRANK KOESTER )i CONSULTING ENGINEER, ASSOC. MEM. AM. INST. E. E. MEMBER SOCIETY GERMAN ENGINEERS (BERLIN), AUTHOR OF " STEAM -ELECTRIC POWER PLANTS 1 WITH 500 ILLUSTRATIONS SECOND EDITION NEW YORK D. VAN NOSTRAND COMPANY 23 MURRAY AND 27 WARREN STREETS LONDON: CONSTABLE & CO., LTD. 191 1 DEPT. COPYRIGHT, 1909 BY D. VAN NOSTRAND COMPANY #7 Engineering Library ALSO ENTERED AT STATIONER'S HALL COURT, LONDON, ENGLAND All Right i Reserved To MY BROTHERS JOHANNES KOESTER, CONSULTING ENGINEER KLAGENFURT, AUSTRIA FRITZ KOESTER, CHIEF ENGINEER HAMM, W., GERMANY 7492C2 PREFACE. OWING to our supposedly inexhaustible coal supply, interest in hydraulic develop- ment has naturally been of tardy growth until recent years; and it is only lately that the government has taken steps toward the development and commercial use of water power resources and the preservation of the forests. In Europe the limited coal supply early induced the utilization of the water resources; and the various Continental governments encouraged this movement by granting favorable franchises, and in many cases advanced money to finance the undertakings, at the same time protecting the water-sheds by rigid enforcement of forest preservation laws. It is but natural, therefore, that hydraulic developments and electric transmission received early and special attention abroad, and as a result Europe abounds in hydraulic developments, utilizing heads varying from 16.5 inches to 3116 feet. Believing that the progress in hydroelectric engineering is stimulated by the interchange of American and European ideas, and having had considerable prac- tical experience, both here and abroad, the author presents this volume as compre- hending the most advanced American and European practice, and trusts that numer- ous novel features of hydraulic, mechanical, and electrical engineering are made obvious. To point out a few of the new features, the following are cited: Air- shafts and equalizing chambers in connection with pressure tunnels. Seamless- welded, flangeless, telescoping penstocks to facilitate shipment and to eliminate expansion joints. Siphon system, in contradistinction to the inverted siphon - which latter is a misnomer. Impulse wheels with draft tubes and multiple, non- \rater-wasting nozzles. Compound turbine on a single shaft, the discharge of one being the supply of the other. Rapid and complete turbine tests by curtain methods and autographic recording device. Thirty-thousand-volt generators and their efficient protective devices against lightning. Unique combination of single and three-phase high-tension transmission systems from three-phase generators. Wagon- panel switchboard systems. Segregation and decentralization of switchboards. Continuous water-flow grounders and horngaps with micrometric setting. Two- legged transmission towers and line-crossing protection. It is not the object of the engineer as a designer of hydroelectric developments to design any particular machine, such as a turbine, generator, transformer, etc., but to provide, by selection from the different makes, an assemblage of machines and devices, each designed to perform its particular function in the most economical manner, and to have the machines properly combined to form one complete unit for the purpose viii PREFACE. of generating and transmitting electrical current from water power on a satisfactory commercial basis. Being of the opinion that a good illustration may tell more at a glance than a long discussion, numerous cuts are presented to readily show the present standing of the American and European Hydroelectric Engineering. Reports, maps, and charts on rainfall, evaporation, and run-off may be directly acquired from the various governments, and therefore are not herein given. As engineers, students, and others desire suggestions and examples in the same or similar lines of work as executed by engineers of standing, there are given in Part III descriptions of several hydroelectric developments, distinctive in their indi- vidual features. From these the experiences and opinions of various authorities and examples of their works are given; for instance, the Niagara, Lockport and Ontario Power Company's development is an epitome of papers by five authorities. The following eight examples are chosen as representative plants of America and Europe: --The Ontario Power Plant (medium head) and its 60,000- volt trans- mission system; The Great Falls Plant (low head), Charlotte, N. C., with its n,ooo and 44,ooo-volt lines, together with data on the fundamental requirements regarding the development, source, and market for power. The Necaxa Plant, Mexico, which gives excellent examples of a high head development, and a i7o-mile, 60,000- volt (ultimate voltage 80,000) distribution. In the description of the Kykkelsrud-Hafs- lund system the parallel operation of two prominent Norwegian low head plants is presented. The Urfttalsperre plant (medium head), Germany, the most promi- nent of the kind in Europe, furnishes striking examples of how to harness the yearly run-off and to husband natural resources in low mountainous countries. The unique price scale adopted enables the consumer to secure power as low as 0.9 to i cent per K.W. hour. Another German plant, the Uppenborn, embodies many novel features in its low head development and 5o,ooo-volt transmission lines, and also illustrates the effect of high voltage on a telephone system. The Brusio Plant (1300 feet head) and its 5o,ooo-volt Swiss-Italian transmission system, probably surpasses all other hydroelectric undertakings because of its many new features. Some discussion has arisen among American engineers as to the practicability of direct generation of high voltages and the consequent elimination of step-up transformers. For many years Continental Europe has had several high voltage generator plants in operation, and that in Manojlovac, Dalmatia, is the latest and foremost of the kind, having four 6ooo-HP. Francis turbines connected to 30,000- volt generators. The current is transmitted over a 2i-mile aerial line, sufficiently protected against lightning by simple devices. It is hoped that the engineer in general, architect, and student, also the manu- facturer, promoter, and financier will find in the text and illustrations a systematic and comprehensive treatise on hydroelectric plants from their inception to the delivery of power to the substation and consumer. FRANK KOESTER. NEW YORK CITY, April, 1909. ACKNOWLEDGMENTS. THE author is indebted to American Institute of Electrical Engineers for embodied paper, by D. R. Scholes, "Transmission-Line Towers and Economical Spans;" also to those whose works have been consulted as indicated throughout the volume. For cooperation: United States Geological Survey; Ambursen Hydraulic Con- struction Company; National Wood Pipe Company; Excelsior Wooden Pipe Company; Wyckoff Wood Pipe Company; The Pelton Water Wheel Company; J. P. Morris Company; Allis-Chalmers Company; S. Morgan Smith Company; The Dayton Globe Iron Works Company; The James Leffel & Company; Escher Wyss & Co.; J. M. Voith, Maschinenfabrik; The Lombard Governor Company; Rep- logle Engineering Company; General Electric Company; Westinghouse Electric and Manufacturing Company; Siemens-Schuckert Werke; Allgemeine Elektricitats- Gesellschaft; Ganz & Co.; Brown, Boveri & Co.; Maschinenfabrik Oerlikon; Elektrizitats-Gesellschaft Alioth; Archbold-Brady Company; Aermotor Company; The Locke Insulator Manufacturing Company; The R. Thomas & Sons Company. From the technical press: Engineering News; The Engineering Record; Elec- trical World; Electrical Review and Western Electrician; Electrical Railway Review; Power and The Engineer; Cassier's Magazine; Electric Journal; The Electrical Age; Transactions of American Society of Civil Engineers; Transactions of American Institute of Electrical Engineers; Transactions of American Society of Mechanical Engineers; Zeitschrift des Vereines deutscher Ingenieure; Schweizerische Bauzeitung; Elektrische Kraftbetriebe und Bahnen; Elektrotechnik und Maschinenbau; Elektro- technische Zeitschrift; Bulletin technique de la Suisse; Journal le Genie Civil. For collaboration and courtesies extended: V. C. Converse, engineer in charge, The Ontario Power Company, Niagara Falls; R. D. Mershon, consulting engineer, New York; C. A. Mees, engineer in charge, Southern Power Company, Charlotte, N. C.; Dr. F. S. Pearson, consulting engineer, New York; R. F. Hay ward, general manager, The Mexican Light and Power Company, Mexico; C. E. Parsons, chief engineer, Hudson River Electric Power Company, Albany, N. Y.; P. P. Barton, general manager, Niagara Falls Power Company, Niagara Falls, N. Y. ; E. W. Cole- man, general manager, Great Northern Power Company, Duluth, Minn.; F. G. Sykes, general manager, Portland Railway Light and Power Company, Portland, Ore.; J. W. Young, vice president, McCall Ferry Power Company, McCall Ferry, Pa. Also the assistance of C. B. Starbird is acknowledged with due appreciation. TABLE OF CONTENTS. PART I. THE TRANSFORMATION OF WATER POWER INTO ELECTRICAL ENERGY. CHAPTER I. PROPOSITION. PAGE INVESTIGATION 3 FOREST PRESERVATION 4 HYDRAULICS. 5 Laws of Hydraulics Gross Horsepower Miner's Inch Weir Dam Flow of River Profile of River Government Reports. ECONOMY IN DEVELOPMENT 15 Preliminaries Problems Involved Designing Staff Drawings and Speci- fications Field Office. CHAPTER II. DAMS. GRAVITY DAMS 19 Masonry Dams Reinforced Concrete Dams, Coffer Dams Crib Dams Timber Dams Steel Frame Dams Earth Dams. MOVABLE DAMS 32 Stony Gate Dam Butterfly Dam Bear Trap Cylindrical Dam Needle Dam Chanoine Dam Flashboards. FISHWAYS 37 CHAPTER III. HEADRACE. SCHEME 39 CONDUITS 40 Cross Section of Conduits Trenches Masonry Flumes Wooden Flumes Protection of Flumes Tunnels Pressure Tunnels Friction in Tunnels Seepage in Tunnels Construction of Tunnels Siphon System. RACKS AND GATES 47 Racks Screens Wooden Sluice Gates Iron Sluice Gates. COLLECTING BASIN 56 Scheme Sand Traps Spillways Gate Valves. xi xii TABLE OF CONTENTS. CHAPTER IV. PENSTOCKS. PAGE STEEL PENSTOCKS 59 Penstock Run Size of Penstocks Friction Loss of Head Strength of Penstock Construction of Steel Penstocks Flanges Anchors Saddles Expansion Joints Safety Devices Standpipes Protection of Penstocks WOODEN PENSTOCKS 78 Adaptability Spacing of Bands Friction Durability Cost Construction of Wooden Penstocks. REINFORCED CONCRETE PENSTOCKS 86 Adaptability Material Reinforcement Strength Construction. CHAPTER V. POWER PLANT. GENERAL ARRANGEMENT 88 Forebay Low Head Plants Medium Head Plants High Head Plants. EXCAVATION AND FOUNDATION , . . . 102 Selection of Site Test Holes Character of Soil Bearing Power of Soil Weight of Masonry Wooden and Concrete Piling Test of Piles Concrete Mat Construction Foundations Anchor Bolts Grouting. SUPERSTRUCTURE ^-T . 107 Architectural Features Material Walls Floors Roof Doors Win- dows Stairways and Elevators Switchboard Gallery Crane Heating Ventilation Lighting Lavatories Conclusions. STRUCTURAL STEEL 122 Roof Trusses Columns Column Bases Floors Expansion Joints Fiber Stresses Character of Steel Workmanship Inspection Painting Pre- vention of Electrolysis. CHAPTER VI. MECHANICAL EQUIPMENT. TURBINES . . 129 Classification Low Head Turbines Medium Head Turbines High Head Turbines Draft Tubes. REGULATING DEVICES . . 143 Principle of Governors Swiss Governors Lombard Governor Replogle Governor Pelton Nozzle Regulation Accessories Couplings. OILING SYSTEM 1 54 Oil Required Filtering Tanks Oi! Pumps Supply Tanks Oil Piping. TESTING TURBINES 157 European Methods Holyoke Tests. TABLE OF CONTENTS. xiii CHAPTER VII. ELECTRICAL EQUIPMENT. PAGE GENERATORS 167 Classification Induction Generator Revolving Field Generator Revolving Armature Generator Regulation Efficiency Frequencies Voltage Ex- citers Generator Leads High Voltage Generators. SWITCHING ROOM : . . 176 General Arrangement. SWITCHBOARDS 18 1 Object Types Panel Type Pedestal or Column Type Desk or Panel Board Direct Current Board Low Tension A. C. Boards Wagon Panel High Tension A. C. Boards. SWITCHBOARD EQUIPMENT 191 Volt and Ammeters Wattmeters Synchronizing Power Factor Meter Frequency Meter Rheostats Illumination of Switchboards. WIRING DIAGRAMS : 194 System of Wiring Diagrams American and European Systems. Bus BARS . 201 Size of Bus Bars Closed Compartments Open Compartments. OIL SWITCHES 204 General Remarks Types of Oil Switches Circuit Breakers Overload Relays Reverse Current Relays Overload Voltage Relays. PART II. THE TRANSMISSION OF HIGH TENSION ELECTRICAL CURRENT. CHAPTER VIII. . ELECTRICAL TRANSMISSION. / GENERAL REMARKS 215 TRANSMISSION CONDUCTORS 215 Strength of Conductors Elasticity of Conductors Cables as Conductors Spacing of Conductors Characteristics of Conductors Size of Conductors D. C. Conductors D. C. Problem A. C. Conductor A. C. Problem Trans- position Corona Effect. POLE AND TOWER CONSTRUCTION 228 WOODEN AND CONCRETE POLES 228 Wooden Poles Strength of Wooden Poles Kind of Wood Cross Arms Life of Wooden Poles Preservation of Wooden Poles Pole Line Construction' Guys Concreted Wooden Poles Reinforced Concrete Poles Steel Pipe Towers. xiv TABLE OF CONTENTS. PAGE REINFORCED CONCRETE TOWERS 234 STEEL TOWERS 235 Wind Pressure on Structure Wind Pressure on Conductors Sleet Founda- tions Portability Two Leg Towers Three Leg Towers Four Leg Towers Tretzo Tower Syracuse Tower Oneida Tower and Specification New York Central Tower Lucerne Tower Brusio Tower Suspended Insulator Tower Line Stresses Transmission Line Towers and Economical Spans. INSULATORS 265 Pin Insulators Suspension Insulators Strain Insulators Insulator Pins Method of Tying Conductors Section Switches Wall Outlets. CHAPTER IX. SUBSTATIONS. GENERAL ARRANGEMENT 280 Location of Substations Size of Units Arrangement of Substations Ventila- tion Drainage Air-Compressor. TRANSFORMERS 286 Type of Transformers Characteristics of Transformers Regulation of Trans- formers Efficiency of Transformers Connections Delta vs. "Y" Connec- tions Oil-cooled Transformers Forced Oil-cooled Transformers Air-cooled Transformers. CONVERTERS 296 Voltage and Frequency Phases Field Connections Starting of Converters Hunting Induction Regulator Compounding Reactances. MOTOR-GENERATORS 302 FREQUENCY CHANGERS 303 SWITCH GEAR OF SUBSTATIONS 307 CHAPTER X. LINE PROTECTION. LIGHTNING ARRESTERS 309 Purpose Lightning Discharge Principles of Arresters Horn Lightning Arresters Horn-gap Setting Choke Coils Multigap Arresters Action of Multigap Arresters Insulation of Multigap Arresters Fluid Arresters Loca- tion of Lightning Arresters. TABLE OF CONTENTS. X V PART III. (Appendix.} MODERN AMERICAN AND EUROPEAN HYDROELECTRIC DEVELOPMENTS. PAGE POWER PLANT OF THE ONTARIO POWER COMPANY, AND THE TRANSMISSION SYSTEM OF THE NIAGARA, LOCKPORT AND ONTARIO POWER COMPANY 327 POWER PLANT OF THE SOUTHERN POWER COMPANY, NORTH CAROLINA, AND ITS TRANS- MISSION SYSTEM 348 POWER PLANT OF THE MEXICO LIGHT AND POWER COMPANY, NECAXA, AND ITS TRANS- MISSION SYSTEM 369 POWER PLANT OF THE AKTIESELSKABET GLOMMENS TRAESLIBERI, CHRISTIANA, AND ITS TRANSMISSION SYSTEM 382 THE URFTTALSPERRE POWER PLANT AT HEIMBACH OF THE RURTALSPERREN-GESELL- SCHAFT, AACHEN, GERMANY, AND ITS TRANSMISSION SYSTEM 393 THE UPPENBORN PLANT AT MOOSBURG OF THE STADTISCHEN ELEKTRIZITATSWERKE, MUNICH, GERMANY, AND ITS TRANSMISSION SYSTEM . . . 403 POWER PLANT OF THE AKTIENGESELLSCHAFT KRAFTWERKE BRUSIO, SWITZERLAND, AND THE TRANSMISSION SYSTEM OF THE SOCIETA LOMBARDA PER DISTRIBUZIONE DI ENERGIA ELETTRICA, ITALY 417 THIRTY THOUSAND GENERATOR VOLTAGE PLANT, AND TRANSMISSION SYSTEM OF THE SOCIETA PER LA UTILIZZAZIONE DELLE FORZE IDRAULICHE BELLA DALMAZIA, AUSTRO-HUNGARY 435 INDEX 445 LIST OF ILLUSTRATIONS. HYDRAULIC MEASURING DEVICE. PAGE Weirs 8 Current Meter 1 1 Plotting Curves for River Discharge 12 Plotting Curves for River Beds 13 Automatic Graphical Registrator 1 58 Curtain Carriage Tests 159 Flume Tests, Holyoke, Mass 160 DAMS. Design of Dams . 19, 20, 21 Concrete Dams 22 Cyclopean Masonry Dam 22 Behavior of Resultants in Solid Dams 24 Behavoir of Resultants in Concrete Steel Dams 24 Reinforced Concrete Dams 25 Submerged Power House, Patapco 25 Bar Harbor Plant, Ellsworth, Maine 27 Reinforced Concrete Dam, Ellsworth, Maine ^ 28 Coffer Dams 29 Timber Dams 29 Steel Frame Dams, Hauser Lake 30 Earth Dams, Necaxa, Mexico 31 Earth Dam with Concrete Core, Dixville, N. H 32 Stoney Roller Sluice Gate Dam 33, 34 Butterfly Dam, Chicago 35 Fishways , 37 Dam and Penstock, Necaxa 373 Dam, Heimbach 393 , 394 HEADRACE. Typical Headrace 39 Timber Flume 43, 44 Tunnel and Overflow 46 Lake Poschiava Siphon 47 xvii xviii LIST OF ILLUSTRATIONS. PAGE Screen House, Niagara Falls Power Company 48 Racks and Deflector, Hafslund, Norway 49 Screens 50, 5 1 Vent . . ^ 50 Penstock Inlet 50 Wooden Sluice Gate 52 Drum Gate 54, 55 Cylindrical Gate 56 Niagara Power Developments 328, 329 Intake of Forebay, Ontario 330 Screen House, Ontario 330 Gate House, Ontario 331 Map of Power Development, Charlotte, N. C 350 Necaxa Power Development ; 370 Headrace, Kykkelsrud 387 Gate House and Penstock, Brusio 421 PENSTOCKS. Flanges 66, 67, 68, 69, 70, 72 Penstock, Anchor 70 Hinged Penstock Support 71 Expansion Slip Joints 71 Expansion Flange Joint 72 Wedge Shaped Expansion Joint 73 Penstock Run, Loch Leven, Scotland 73 Collecting Basin 74 Penstock Arrangement, Brusio 74 Penstock Vent 74 Automatic Flap for Penstocks 75 Vacuum Relief Valve 77 Automatic Low Water Valve 78 Wooden Stave Penstock 79, 80 Detail of Wooden Stave Penstock 82 Penstock Partly Embedded, Ontario 331 Penstock Intake, Necaxa 371 POWER PLANTS. Chevres, France 54 Lyon, France 56 Georgia, Albany 89 Holyoke, Mass 89 Winnipeg, Manitoba ' 89 Colliersville, Oswego, N. Y 90, 91 Shawinigan, Can 92 LIST OF ILLUSTRATIONS. A.E.&M.E xix 9 2 , 94 McCall Ferry, Pa Niagara Falls Power Company 7T 95, 107, 109 Toronto, Can 96, 97 Kern River, Plant No. i 97, 99 Snoqualmie Falls 100 Kykkelsrud, Norway 102 Wooden and Concrete Piles 105 Stuttgart, Germany no Obermatt, Lucerne, Switzerland 112 Urfttalsperre, Germany 113 Geneva, Switzerland 114 Sillwerke, Tyrol 1 20 Tivoli, Rome, Italy 121 Typical Roof Trusses 122 Typical Columns 123 Crane Column 124 Ontario Power Plant and Distributing Station 333~33^ Charlotte, N. C 351, 356 Necaxa, Mexico 374~377 Kykkelsrud, Norway 383-388 Heimbach, Germany 395, 396 Uppenborn, Germany 404, 405 Brusio Power Plant and Penstock Run, Italy 418, 422, 424 Manojlovac Power Plant, Dalmatia 436-438 MECHANICAL EQUIPMENT OF POWER PLANTS. American Turbine 130 American Turbine, Runner, and Gate 131 Low Head Turbines 132 Vertical Shaft Francis Turbine 133 Double Spiral Francis Turbine* .' 134 Compound Francis Turbine 135 Double Flow Francis Turbine 136 Thrust Bearing 137 Impulse Wheel and Nozzles 138, 139 Francis Spiral Turbine, Horizontal Shaft 140 9700 H.P. Francis Turbine 142 Impulse Wheel Nozzle and Regulator 144 Spoon Wheel Turbine and Governor 145 Turbine Gate 146 Governors 147, 149, 150 Hydraulic Relief Valves 148 Pressure Regulator 150 Pelton Wheel Buckets 151 Automatic Deflecting Nozzle 152 CM xx LIST OF ILLUSTRATIONS. PAGE Relief Valves 153 Flexible Coupling . 153 Oil Filters. ... 154, 155 Oil Filter for Large Capacities 155 6ooo-Horsepower Francis Turbine 439 ELECTRICAL EQUIPMENT OF POWER PLANTS. Generator, Two-Phase, Umbrella Type 168 Induction Generator 168 Flywheel Type Generator 169 Three-Phase Generators 171 Exciters 171 Three-Phase Generator 172 Characteristic Curves of Generator . . 172 Switch and Transformer Room, Obermatt 177 Switch House, Castelnuovo-Valdarno, Italy 178 Shawinigan Falls Plant 179 Puget Sound Plant 180 Switchboard, A. C. and D. C., Necaxa Plant, Mexico 181 Switchboard, Panel and Bench Desk 182 Instrument Columns 183 Switchboard Panel, Rear View, Obermatt 184 Remote Control Switches 185 Instrument Bench Desk 186 Switchboard Panel, D. C 186 Switchboard, Combined Panel, D. C 187 Switchboard for Three-Phase Generator 188 Switchboard, Wagon Panel 189 Oil Switch and Bus Bar Compartment 190 Wattmeter Connections 192 Motor Controlled Rheostat 193 Remote Hand Operated Rheostat 194 Wiring Diagram, Valdarno 195 Wiring Diagram, Ontario 196 Wiring Diagram, Necaxa 197 Wiring Diagram for Exciters 198 Wiring Diagram, Urfttalsperre 199 Wiring Diagram, Obermatt 200 Wiring Diagram for Single Generator and Step-up Transformer 201 Circuit Breaker and Bus Bar Compartments 202 Open Bus Bar Compartments, Lontsch, and Luzerne Plants 203 Solenoid Controlled 30,ooo-volt Oil Switch 205 io,ooo-volt Air Break Switch 205 i i,ooo-volt Oil Switch, Solenoid Controlled 206 6o,ooo-volt Oil Switch Compartment, Ontario 207 3o,ooo-volt, Motor Controlled Oil Switch 208 LIST OF ILLUSTRATIONS. xxi PAGE 1 1,000 and 5o,ooo-volt Oil Switch, Castellanza, Italy 209 88,ooo-volt Circuit Breakers 210 35,ooo-volt Switch with Current Blowouts 210 High Tension Time Relay 210 Control Room, Ontario 337 High Tension Bus Bars, Ontario 340 Oil Pipe System for Transformers, Charlotte, N. C 354 Interior of Generating Room, Charlotte, N. C 357 Wiring Diagram, Charlotte Plant 358 High Tension Switch Room, Charlotte, N. C 359 Solenoid Operated Oil Switch 360 Switch Room, Kykkelsrud 389 Generators and Exciters, Heimbach 397 Bus Bar and Switch Room, Heimbach 400 Switchboard, Heimbach 398 Generating Room, Uppenborn 406 Wagon Panel, Siemens-Schuckert, Uppenborn 414 Switchboard, Individual Generator, Brusio 425 Switchboard, Rear View, Brusio 427 Bus Bars, Brusio 427 Exciter Switchboard, Brusio 427 3o,ooo-volt Generator 439 HIGH TENSION TRANSMISSION. Wooden Poles for 5o,ooo-volt Transmission , 228, 229 " A " Frame Tower, 6o,ooo-volt 230 Cross Arm and Guard Wire, 4o,ooo-volt 231 Reinforced Concrete Tower 234 Two-Legged Steel Tower, 37 feet 237 Types of Poles and Towers, 5o,ooo-volt 238 Four-Legged Twin Steel Tower 239 Steel Tower, 45 feet, Syracuse 240 Steel Towers, Niagara Crossing 241 Dead End Steel Towers, Rochester 242 Steel Tower, Standard, New York Central 245 Tower, Bracket and Guard, 35,ooo-volt, Heimbach 246 Cantilever and Steel Tower, 27,ooo-volt, Obermatt 246 Steel Tower, 5o,ooo-volt, Piattamala 247 Highway Crossing, 5o,ooo-volt, Lecco 248 Steel Towers, i io,ooo-volt, Muskegon 249 Steel Tower, Suspended Insulator Type 250 Steel Towers, Diagrams of Different Types 253, 256, 259 Chart for Economical Width of Tower Base 259 Chart for Economical Span 261 Type of Foundation 263 Chart for Sag of Conductors 264 xxii LIST OF ILLUSTRATIONS. PAGE Insulator, 4o,ooo-volt, Kern River 265 Insulator, 5o,ooo-volt, Taylor's Falls 265 Insulator, 40,000 to 5o,ooo-volt, Paderno 268 Insulators, Methods of Suspending 269 Insulators, Suspended 268, 270 Insulator, Suspended no,ooo-volt, Muskegon 270 Application of Strain Insulators 271 Anchor Insulator 271 Rolling Insulating for Long Spans, Tofwehult 271 Insulator Pin, Porcelain Base 272 Insulator Pin, all Steel 272 Insulator Pin, Detail, 6o,ooo-volt 273 Insulator Tie and Clamp, 6o,ooo-volt, Ontario 273 Insulators, 6o,ooo-volt, Ontario 273 Line Disconnecting Switch 275 Two Break Section Switch 276 Typical Wall Outlets 276-277 Cross Arm Guard Wire, Lightning Rod 323 Transmission Line, 62,ooo-volt, Ontario 341 Niagara Crossing 342 Cantilever, Niagara Crossing 343 Cross Connection and Open Air Fuse, Auburn 343 Four Leg Tower on "Floating" Foundation, Montezuma Swamp 344 Transmission System, Charlotte, N. C 349 Transmission Feeders 361 Main Transmission Line, Charlotte, N. C 363 Insulator, 5o,ooo-volt, Charlotte, N. C, 364 Transmission Line in Cities, Charlotte, N. C 365 Map of Transmission Line, Necaxa 378 Transmission Tower, Necaxa 379 Transmission Line, Necaxa 380 Transmission Line, 5o,ooo-volt, Kykkelsrud to Hafslund 392 Wall Outlet, Uppenborn 408 Insulators, 5o,ooo-volt, Uppenborn 410 Cable Tunnel, Brusio, Piattamala 427 Line Crossing, 5o,ooo-volt, Piattamala 43 1 Insulator, 3O,ooo-volt, Manojlovac 441 Transmission Line Crossing, Manojlovac 442 SUBSTATIONS AND EQUIPMENT. Duluth Substation no Stansstad Substation 115 Substation, Waterbury 281-282 Substation, Typical Three-Phase 283-284 Substation, Steghof, Switch Gear 285 Transformer, Shell Type 286 LIST OF ILLUSTRATIONS. xxiii PAGE Transformer, Core Type 286 Transformer, Water Circulated, Oil Cooled 287 Chart Showing Efficiency of Air Blast Transformer 288, 289 Transformer Connections to Rotary Converters 291 Transformer, Forced Air Circulation 292 Transformer, Air Cooled 293 Transformer, Method of Water Cooling, Molinar, Spain 293 Transformers, Arrangement of Air Blast . . - 294 Transformer, Chart Showing Air Required 295 Rotary Converter, Substation, Albina 297 Automatic Regulators, Transformer 300 Typical Continental Motor Generator, Substation 302 Motor Generator, Substation, Steghof 303 Frequency Changers 304 Typical Switchboard Panels for Substations 305 Wiring Diagram, Waterbury Substation 306 Distributing Station and Power House, Ontario 333 Control Room, Ontario 337 Distributing Station, Ontario 337 Transformer Compartment 338 High Tension Bus Bars, Ontario 340 6o,ooo-volt Open Air Bus Bars, Lockport 346 6o,ooo-volt Circuit Breaker, Lockport 347 Transformer Room, Charlotte, N. C '. . . . . 354 Substation, El Oro 381 Transformer Room, Kykkelsrud \ 390 Substation Interior, Hafslund 391 Transformer Room, Heimbach 399 Transformer House, Typical 402 Transformer and Lightning Arrester House, Hirschau, Uppenborn 411 Transformer Water Flow Grounder, Hirschau, Uppenborn 413 Horn Gaps, Substation, Hirschau, Uppenborn 415 Transformer Station, Piattamala 428 Switch Room, Piattamala 429 Air Cooled Transformer, Open Type, Lomazzo 433 Switch Room, Lomazzo 433 LIGHTNING ARRESTERS. Horn Gaps, Showing Principle of Action 310 Horn Gaps, Application with Oil Rheostat 312 Horn Gaps, Construction 313 Horn Gaps, Setting 313 Lightning Arrester Equipment, Hirschau, Uppenborn 413 Protection of Overhead and Underground Lines 313 Water-Flow Grounder 413 Station Protection, Torchio 314 xxiv LIST OF ILLUSTRATIONS. PAGE Protection, 3ooo-volt Circuit, Gola 314 Multigap Arresters 315,316 Aluminum Arrester 318, 319 Electrolytic Arresters 320-32 1 Horn Gaps, Water-Flow Grounder, Steghof 321 Water-Flow Grounders 321-322 Bank of Horn Gaps, Choke Coils, and Water-Flow Grounders, Vandoise 322 Water-Flow Grounders, 5o,ooo-volt, Piattamala 322 Lightning Rod and Guard Wire 323 Lightning Protecting Device, Heimbach 401 LIST OF TABLES. PAGE DISCHARGE OF WEIRS 9 VELOCITY OF WATER IN CHANNELS 42 PROPERTIES OF TIMBER 53 TESTS OF AMERICAN WOODS 53 AREA OF CIRCLES 60 FRICTIONAL HEAD Loss IN PENSTOCKS 62 CAPACITY DISCHARGE OF PIPES 63 RIVETED PIPES 65 SAFE STRAIN OF PENSTOCK BENDS 79 STRENGTH OF DOUGLASS FLR 81 COST OF RIVETED PENSTOCK 82 COMPARATIVE COST OF PENSTOCKS 83 BEARING POWER OF SOIL 103 WEIGHT OF MASONRY 104 RADIATING SURFACE FOR HEATING 117 COMPARATIVE TESTS OF TURBINE 161 MODULUS OF ELASTICITY OF CONDUCTORS 216 COEFFICIENT OF EXPANSION 216 STRAND FACTOR OF CONDUCTORS 216 COMPARISON OF WIRE GAUGES 218 SOLID COPPER WIRE 219 STRANDED COPPER WIRE 220 REACTANCE OF VOLTS 221 WEIGHT AND STRENGTH OF WIRES 221 AMERICAN AND ENGLISH COPPER CABLE 222 AMERICAN AND ENGLISH COPPER WIRE 223 CORRECTED WIND VELOCITIES 235 COMPARATIVE TESTS OF TRANSMISSION TOWERS . 238 TENSILE STRENGTH IN CONDUCTIVITY OF CONDUCTORS 250 SAG OF WIRES AT DIFFERENT TEMPERATURES 252 RESULTANT FORCE IN TRANSMISSION TOWERS 262 AIR REQUIRED FOR TRANSFORMERS 294 AIR REQUIRED FOR TRANSFORMERS OF 25 AND 60 CYCLES 295 SPACING BETWEEN MULTIGAP ARRESTERS 312 EFFICIENCY OF TURBINES 352 EFFICIENCY OF GENERATORS 355 EFFICIENCY OF EXCITERS 355 EFFICIENCY OF TRANSFORMERS 359 COST OF HYDROELECTRIC INSTALLATION 416 XXV PART I. THE TRANSFORMATION OF WATER POWER INTO ELECTRICAL ENERGY. HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. CHAPTER I. PROPOSITION. Investigation. Before developing a water power much preliminary study is necessary to ascertain whether the proposition will be a paying one. Reliable data must be collected and put in complete and definite form before capitalists can be interested. After the investigations are made showing the amount of energy available, the possible field of consumption must be carefully considered. This may include other central stations, either steam, gas, or even other water-power plants. While the selling price of current is known, it might appear difficult to ascertain what it costs existing companies to produce electrical energy. There are, however, several ways by which this information can be obtained, and with the help of an experienced engineer very close figures can be ascertained. These costs are essential as a guide for the new development, because it may have to compete with or possibly sell current to established stations, and in any event this is the. salient factor in determining whether the proposed plant is an advis- able development. In the case of selling power to established electrical systems the plants are cus- tomarily operated in parallel. In some instances the separate companies have found it expedient to merge their interests and form a corporation. Having arrived at the competitor's figures, the other prospective fields for current consumption must be thoroughly canvassed, to ascertain the load and the price for which the current can be sold. In fixing the selling price different rates are charged according to the amount, duration, and time of load. Conclusions as to the cost of current can only be derived after trial load-curves have been plotted, and the careful balancing of the engineering and commercial items for each particular plant. Plants are economical in first cost and in operation in proportion to the constancy of their load factors. With greatly varying loads much machinery is idle a large part of the time. However, 'in competing successfully with existing central station or private plants, prospective consumers who will require current for only a few hours each day or possibly each week, and those who will need emergency current, must not be overlooked. That these consumers pay a high rate for the service rendered is but natural. 3 4 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. After the foregoing investigations have been made, and the figures show that the installation .is warranted, the possible future growth of the locality, such as the industrial increase, electric railroading, etc., must be well considered; with the growth of the cornm-anity an increase in consumers naturally follows. Hydroelectric plants have often been installed without provision for future extension; the dams are located in such sections of the rivers, and are of such a design, that an increase in storage supply or additional head is impossible. Plants have also frequently been installed where sufficient water to carry the full load cannot be obtained during certain seasons of the year. While the former may be due to lack of foresight, the latter is attributable to negligence in proper investiga- tion. Under such conditions competition would be difficult. In many cases water- power plants do not have to meet competition, as they may be pioneers in the field. Under favorable conditions the hydraulic plant may be reinforced with a steam or a gas engine plant, this auxiliary equipment taking the load during periods of low water and hours of heaviest demand. The equipment and size of the units should be such that they will run at their best efficiency throughout the day, that is, they must not run underloaded or greatly overloaded. Reserve units must be kept in readiness to be thrown on the line when the demand calls for them. Present practice is to install as large units as possible. FOREST PRESERVATION. The relation of forests to water-power development is of the greatest impor- tance. It is a well-known fact that the soil of forests retains the water of precipita- tion more uniformly and releases it gradually, so that during dry seasons a supply of water is assured. Flat non-forested land may hold the water and form swamps, but in most cases the water drains off rapidly, so that streams having denuded watersheds are subject either to floods or droughts. Long observation and costly experiments have proven that forests receive a greater quantity of rain, hail, and snow than land in the same vicinity. Mountain- ous countries, whether bare or covered with forest, receive more rain than flat coun- try; and the forests in mountainous countries receive more rain than bare land at the same elevation. The following data throw some light on the effect forests have on water-power developments. According to the report by the Swiss engineer, Lauterberg, the drainage of the canton Tessin, between 1834 and 1862, was reduced about 28 per cent, due to the removal of forests. He states further that prior to the destruction of forests the valleys were flooded, on the average, every 54 months, while after the forests were destroyed floods occurred every 29 months. Professor Ebermeyer states that, considering the evaporating factor of free land as 100 per cent, the evaporation of the forest land is only 22 per cent, other conditions remaining the same. Dr. van Bebber observed that a forest at an elevation of 1000 meters (3280 feet) has about PROPOSITION. 5 50 per cent more drainage than free land situated on the same elevation. Relative to this, Professor Schreiber, who observed conditions in Saxony, came to the con- clusion that forests on open country receive as much precipitation as free land ele- vated 200 meters (656 feet) higher. Further, Professor Landolt, Zurich, states that for every 100 meters (328 feet) elevation the annual drainage will increase 10 inches. Between the years of 1843-1883 the Ekaterinoslaw government, Russia, culti- vated a forest of 5000 acres, and established two meteorological stations in this sec- tion. The reports show that since the Introduction of the forest, showers are much more frequent, and the previously feared dry seasons are things of the past. The stations report that the average rainfall between 1893-1897 was 18 inches for free land, while for the forests 22.25 inches. The French government spent 14,500,000 francs between the years 1861 and 1877 to forestize 235,000 acres in mountainous localities. The result was so bene- ficial that the government decided to forestize about two million acres additional, which will probably cover 60 to 80 years and consume 150,000,000 francs. Austria has at present very elaborate plans to reestablish forests in denuded sections. The Italian government has set aside $8,000,000 as a beginning towards the reestablish- ment of the forests in the southern Alps. For many years Germany has enforced rigid laws for the preservation of her forests, and in recent years has encouraged and assisted water-power developments. In May, 1908, the governors of several States discussed the preservation of our national resources, particularly those of forests and water supply. This was due chiefly to the increase in the fluctuation of streams, which is a direct result of the destruction of forests. Fluctuations of water supply and danger due to floods have forced water-power developments to additional expenditure to harness water of uncertain quantities. For instance, 1 one of the largest power companies had to build new dams 25 per cent greater in cross section than the older ones on the same stream; other hydraulic plants that previously had abundant water are now forced to supplement with auxil- iary steam plants. HYDRAULICS. Laws of Hydraulics. In 1830, Galileo stated that the laws governing the flow of water were not as well known as those governing the movements of the celestial bodies, and even to-day this is still true. Our experimental data of to-day are far in advance of hydraulic theory, hydraulic engineering being based more on empir- ical facts than on rational mathematical formulas. 2 For power purposes water is usually measured in cubic feet of flow per second. The unit weight of water at ordinary temperature is 62.5 pounds per cubic foot. The present theory of the flow of water is based on a few formulas. The funda- mental laws of falling bodies apply also to the flow of water. Of course the formu- 1 Forestry Hearing, Am. Inst. E. E., May, 1908. 2 Merriman, Treatise on Hydraulics. 6 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. las derived from the laws of gravity cannot be directly applied to hydraulics; they must be changed to suit conditions. However, by the judicious use of the funda- mental principles and common sense all power problems may be easily solved. The principal formulas in hydraulic calculations are: v = \/2 gh = 8.03 V/z. q = av = a ^2 gh = 8.03 X fl V/z. , v velocity of flow in feet per second. q quantity of flow in cubic feet per second. h = head or height through which the water falls. a = area of cross section of falling body of water. Whenever the formula for q is applied to water issuing from an orifice, a coeffi- cient must be introduced. Thus, for water issuing from a circular orifice the quantity coming out is not q \ TT d? V '2 gh, but q = c/4 it d 2 \/2 gh, where c is the coefficient, whose value depends on the sharpness of the edges of the opening. This is true for the flow of water issuing from a square, rectangular, or, in fact, any shaped orifice. The values of the coefficients in any case never equal unity. A special case of a rectangular orifice is what is known as the 'weir.' The weir in general is an opening or rather a notch through which the water flows. The fundamental formula used with weirs is Q - i O ed .c Cubic feet S J2 Cubic feet * .S per minute * 1 per minute *J per minute *| per minute *| per minute *o - passed for "o passed for "o passed for ^ passed for "o - passed for J2 53 each foot of I- j- '53 each foot of x '53 each foot of t. ja '53 each foot of u j- "53 each foot of\ fc * length of S* length of * length of S^ length of Z* length of Qg weir. r Q o weir. Q weir. '-" ^j ^ j ,r ,-' X I L MEASUREMENTS 1 Ki lOf NO. 1 I9O3 NO f T 1904 NO 8 I90S NO 9 n 1906 NO 14 9? g! ^- t 1 y >a it & X 1 i / ? &% I L t % /: \ -'- S / / S 7 v7 g //' 7 */ 7 ? / / 7 ^ 7 s / r /? [/ ! / 8 "<. / / "/ / &> I? ,*A i j>i I- l/e lOClt z y in a feec it. per .seed j na tt so Jrc eo ^ ^L, Ijj ai-Q 00 feat I6< 8 ao iao 10,000 20.000 30.000 4O.OOO 50.000 60.000 70.000 60.00O 90,000" 100.000 "0.000 120,000 130.000 l.000 IK Discharge in second-feet FIG. 3. Method of Plotting Curves for Discharge, Mean Velocity and Area of Rivers. shows a typical rating curve with its corresponding mean velocity and area curves. 1 As the discharge is the product of two factors, the area and the mean velocity, any change in either factor alone will produce a corresponding change in the discharge. The curves are therefore constructed in order to study each independently of the other. The area curve can be definitely determined from accurate soundings extending to the limits of high water. It is always concave toward the horizontal axis or on a straight line unless the banks of the stream are overhanging. p 1 Water-Supply and Irrigation Paper No. 192 of the U. S. Geological Survey. PROPOSITION. 13 The form of the mean velocity curve depends chiefly on the surface slope, the roughness of the bed, and the cross section of the stream. Of these the slope is the principal factor. In accordance with the relative change of these factors the curve may be either a straight line, a curve, convex or concave, or a combination of the three. From study of the conditions at any gauging station the form which the vertical velocity curve will take can be predicted, and it may be extended with reasonable certainty to stages beyond the limits of actual measurements. It is used principally in connection with the area curve in locating errors in discharge measurements and in constructing the rating table. The discharge curve is drawn from the measurements of the discharge. The curve may have certain of its points located between and beyond those given by the actual measurements by means of the curves of area and mean velocity. Under normal conditions the discharge curve is concave toward the horizontal axis and is generally parabolic in form. The chart is readily understood; the term " second-feet" is an abbreviation for cubic feet per second, and is the rate of discharge of water flowing in a stream i foot wide, i foot deep at the rate of i foot per second. Profile of River. To ascertain the slope or fall of a river, elevations of the river level have to be read and plotted, so that the best locations for dams can be seen, see Fig. 4. The abscissas give the distance in miles and the ordinates give the elevations in feet; the latter are preferably read as elevations above the sea level. i,roo 1,250 ~ 1,000 = i 0. Miles 5 FIG. 4. Method of Plotting River Bed. (Alcoy River.) Government Reports. The governments of nearly all countries maintain depart- ments for studying the flow of streams, and official reports on stream measurement are regularly issued. The United States Geological Survey has for more than twenty years been studying the various phases of the water resources of the United States. The results of most of these studies have been published as Water-Supply Papers. Some, however, appear in annual reports and bulletins. These studies 14 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. include measurements of the flow of streams, determination of river profiles, and collection of data in regard to water-power development. These data give the records for a number of years of the rainfall, flow of streams in cubic feet per second for wet and dry seasons, and in some cases give the gross horsepower which could probably be developed, and occasionally offer suggestions to the hydraulic engineer. An excellent example is taken from the introduction of "The Relation of the Southern Appalachian Mountains to the Development of Water Power." l According to estimates made by the United States Geological Survey there is a minimum of about 2,800,000 indicated horsepower developed by the rivers having their head waters in the Southern Appalachian Mountains. Mature consideration of the condition leads the Survey to estimate that at least 50 per cent and probably much more of this indicated power is available for economic development. If auxil- iary power were provided, it would be profitable to develop up to 2.5 times this amount. Full development of storage facilities would increase the minimum from 2 to 30 times. Obviously an estimate of present value based on 50 per cent of the mini- mum indicated horsepower is sure to be extremely conservative. The rental of 1,400,000 horsepower at $20 per horsepower per year would amount to an annual return of $28,000,000. This amount is equal to a gross income of 3 per cent on a capital of about $933,000,000. Some of this power has already been developed, but a very small proportion hardly enough to make any appreciable showing when the enormous resources of the region are taken into account. It has been estimated that in the United States more than 30,000,000 horse- power are available, and under certain assumptions as to storage reservoirs this amount can be increased to 150,000,000 horsepower or possibly more. In an address at the conference on the "Conservation of the Natural Resources," at Washington, D.C., May, 1908, St. Clair Putnam made the following statement on the value of the water powers in the United States: "Using the smaller figure of 30 million horsepower as an illustration; to develop an equal amount of energy in our most modern steam electric power plants would require the burning of nearly 225,000,000 tons of coal per annum, and in the average steam engine plant, as now existing, more than 6,000,000 tons of coal, or 50 per cent in excess of the total coal production of the country in 1906. At the average price of $3.00 per ton, it would require the consumption of coal costing $1,800,000,000 to produce an equivalent power in steam plants of the present general type." Of this immense water power available, only a small percentage is developed, estimated to be about 3,000,000 horsepower. Nearly every state in the Union has large water powers available. It has been estimated that the upper Mississippi and its tributaries have an available water power of about 2,000,000 horsepower; that. of the Southern Appalachian region, about 3,000,000; and that of the State of Washington alone, about 3,000,000 horsepower. 1 Forest Service, Circular 144, U. S. Department of Agriculture. PROPOSITION. 1 5 ECONOMY IN DEVELOPMENT. Preliminaries. The first cost, efficiency, and economy of an hydraulic develop- ment depend primarily on the ability of the designer. This fact, although of great importance, is often overlooked by the investors. When a plant is to be built for a railroad company or other large corporation, the designer is frequently in the employ of the company; sometimes contracts are let to firms of contracting engineers, who may furnish the plans only or both the plans and the entire plant. Contracts may be made between investors and engineers for professional services for specific amounts, or for a percentage plus disbursement, or for fixed sum plus disbursements. Capitalists or corporations, before letting contracts, should make thorough investigations, not only of the financial and business reputation of con- sulting and contracting engineers, but should convince themselves of the ability of the firms and their staffs, particularly of the designer in charge. Frequently during the course of construction or after the completion of a plant, an experienced designer shows where thousands of dollars could have been saved by engaging engineers who are specialists in the design of plants. Reports made on plants after their construction have shown in some instances that in stations of 10,000 K.W. capacity several hundred thousand dollars could easily have been saved; while on plants of 50,000 K.W. capacity reports have been made showing where over a million dollars could have been saved. Problems Involved. The problems involved in the design of hydroelectric plants are those of first cost of construction, equipment, operation, and maintenance. It is the ultimate aim to produce electricity at a minimum of expense. To accom- plish this end, experience is necessary. It is not the province of the engineer as a designer of hydroelectric plants to design any particular machine, such as turbines, generator, oil-switches, etc., but to provide a selection of different makes, each designed to perform its function in the most economical manner; and to have these machines properly combined to form one complete unit for the purpose of gener- ating electricity from water on a satisfactory commercial basis. Since on the original design depends the economical operation of the plant, great care and foresight must be exercised in the selection and arrangement of the devices; for instance, a turbine for low head would not be so efficient if connected to a high head, and vice versa. The location of the power plant building must be chosen so as to obtain the greatest head with the least expenditure for head or tail race. As a general rule, the higher the head the cheaper will be the installation. In designing a plant, and in the selection and arrangement of the equipment, some originality should be exercised. No designer should unreservedly copy the scheme of an existing plant, since what might be economical in one would possibly be the reverse in the other. Any attempt to standardize the design of hydraulic plants is practically impossible. However, in the design of a single station, a system of standardization must be adopted to minimize expenses in design, construc- tion, and operation. i . 1 6 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The work embodied in a complete installation comprises hydraulic, mechanical, structural, and electrical work. They are necessarily closely allied, and it is essential that the design of the entire undertaking be placed under one engineer. If this method is not followed, confusion may possibly result, delaying the work and incur- ring additional expense. If the work is divided among several designers, complete cooperation may not exist, and the various designers will probably conflict with one another; for instance, the same article or work may appear in two or more drawings or specifications, or may be entirely omitted, one designer considering it a part of another's work. Designing Staff. Having secured the necessary data, and fixed the size of the plant, the design must be carried on systematically. As the scope of power plant design is a broad one, it is necessary in large plants to employ assistants, designers, and draughtsmen. For instance, in designing a 50,000 K.W. plant the designer's staff may consist of one assistant, who is familiar with the various branches required in the complete plant; four or five draughtsmen (assistant designers), who are experienced in the various branches previously mentioned. The hydraulic or mechanical designer should arrange the general scheme, such as arrangement of turbines and particularly the headrace and the foundation of the building; the electrical designer, the elec- trical layout, such as wiring and switchboard, etc., and will work in conjunction with the mechanical engineer to establish the size of the building. The structural design depends upon the data supplied by the mechanical and electrical engineers for the skeleton of the building, floor loads, roof trusses, etc. The structural engineer is often called upon to assist in the design of the gates and penstocks, also to design the high tension transmission towers. Architects are seldom employed, as is evidenced from the severely plain power houses, of which there are numerous examples. However, in the last few years occasional plants have been erected indicating that architectural talent has been employed. In order to bring about system and economy in the draughting department, a few tracers may be employed to do less important work, such as tracing and lettering. By shifting the tracers around, as necessity requires, they receive proper training and a general knowledge of the whole power plant construction. While the check- ing of all drawings is necessary, it is not feasible to employ a checker to verify draw- ings of all branches. While he may check the dimensions in conjunction with the designer of the individual features, it is impossible to find a checker to verify the design as is usually done in structural steel branches. Such checkers have to be familiar not only with the general scheme but with the detail of every feature employed in the design of the complete plant. Therefore the designers of the differ- ent branches should check each other's drawings. Drawings and Specifications. For convenience of the draughting department and especially the field, all drawings should be standardized. Drawings larger than 24 X 36 inches are cumbersome and inconvenient for constructors. Multi- plicity of drawings should be avoided. Drawings of the several branches must not PROPOSITION. 17 appear on one sheet, i.e., the structural steel must not be on the same sheet as the foundation work or part of same. It is common practice to begin the work after a few drawings which later undergo revision as construction proceeds. Unless a system of revision numbers is used, subsequent construction is seriously handicapped. Duplicate sets of blue prints should be required from the manufacturers, one set for the office files, the other to be returned with indicated changes or approval as the case may be. All drawings, as well as incoming and outgoing blue prints, should be properly indexed on a two-card filing or other efficient system. Before submitting plans and specifications to contractors for bids, they should be complete in every respect, in fact they should be working drawings. The speci- fications must be drawn up after the plans are finished or practically finished, and should be so drawn as to simplify and explain the plans. Each contractor's specifications should start where that of the previous con- tractor stopped, so that the work will not overlap, or gaps be left. It is not infre- quent practice by plant designers to consult engineering salesmen or manufacturers, from whom it is always advisable to secure specifications, and draw comparisons between the products of the various manufacturers. As stated, specifications and plans have to be complete before they are sub- mitted for bids, in order to minimize in extras. Extras are usually overcharged, since it is to these that some contractors look for profit. For instance, the contract for structural steel may be let from prelimi- nary drawings, on a per pound basis, say from three to four cents; when, however, the plans are worked out in detail, it will be found that there are a number of stair- cases, ladders, railings, etc., not shown in preliminary drawings, and the contractor, on a plea that more workmanship is required with this kind of work, will raise his price to seven or eight cents per pound or even more. For certain apparatus, such as turbines, generators, and overhead cranes, etc., preliminary bids may be asked for from rough drawings to ascertain the approximate cost, of the plant. This may be necessary when the design is limited to a fixed sum. Field Office. As most hydraulic developments, particularly those of large size, are far away from the main engineering organization, it is necessary to have an experienced and capable engineer with a good staff in the field. Cases always arise during construction where, for various reasons, the drawings cannot be strictly followed, and to secure the necessary instruction from the home office consumes much time and may cause confusion. Most of the errors discovered in the process of construction are trivial in themselves, but affect the remainder of the plant. Any modifications of the original drawings made in the field must be undertaken only by an experienced resident engineer, who must have sufficient authority from the head office. All field corrections, however, must be at once reported to the main office, so that corresponding changes can be made on the original design. 1 8 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING BIBLIOGRAPHY. TREATISE ON HYDRAULICS. 1903. Mansfield Merriman. HYDRAULIC AND HYDRAULIC MOTORS. P. J. Weisbach. Translated by A. J. Dubois. A TREATISE ON HYDRAULICS. 1901. Henry T. Bovey. HYDRAULICS. 1907. L. M. Hoskins. RIVER DISCHARGE. 1907. J. C. Hoyt and N. C. Grover. NOTES ON HYDROELECTRIC PLANT ORGANIZATION AND OPERATION. Farley Osgood. Pro. Am. Inst. E. E., April, 1907. CONSERVATION OF POWER RESOURCES. H. St. Clair Putnam. Pro. Am. Inst. E. E., August, 1908. WATER POWER DEVELOPMENT IN THE NATIONAL FORESTS. F. G. Baum. Pro. Am. Inst. E. E., July, 1908. SOME PRELIMINARY STEPS IN HYDROELECTRIC ENGINEERING. Frank Koester. Electrical Review and Western Electrician, Nov. 28, 1908. RAINFALL OF THE UNITED STATES. A. J. Henry. Bulletin D, U. S. Weather Bureau, 1897. RAINFALL AND FLOW OF STREAMS. C. C. Babb. Trans. Am. Soc. C. E., vol. 28, p. 329, 1894. THE INFLUENCE OF FORESTS UPON THE RAINFALL AND UPON THE FLOW OF STREAMS. G. F. Swain. Jour. New Eng. W. Wks. Ass'n. DATA OF STREAM FLOW IN RELATION TO FORESTS. G. W. Rafter. Ass'n C. E. Cornell Univ., vol. 7, p. 22, 1899. THE CONSERVATION OF WATER. John B. Freeman. Proc. Am. Inst. E. E., March 24, 1909. THE WASTES OF OUR NATURAL RESOURCES BY FIRE. Charles Whiting Baker. Proc. Am. Inst. E E., March 24, 1909. ELECTRICITY AND THE CONSERVATION OF ENERGY. Lewis B. Stillwell. Proc. Am. Inst. E. E., March 24, 1909. CHAPTER II. DAMS. Gravity Dams. The fundamental principles on which the stability of a dam is calculated are given in the following formulas: First, it is essential to know the location of the center of gravity of a dam. This may be found for a dam of triangular shape, as it is indicated in Fig. i. Assuming that the water behind the dam is even FIG. i. with the crest, as seen in Fig. 2, the pressure of the water against the dam is calculated in the following way: H = head or height in feet. P = total pressure of water in pounds. H ( = center of pressure. 3 W = total weight of dam acting through center of gravity, P =H X i X 62.5 X 2 The pressure, P, acts perpendicularly to the face, and in turning the dam over uses the lever AD. The overturning moment is P X DA. 19 20 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. To counterbalance the tendency of the water to overturn the dam, the weight of the dam acts through the lever AF. EF is drawn perpendicular to the base through the center of gravity. The resisting moment due to the weight of the dam is W X FA. Wx FA - is the factor of safety. P X DA H w FIG. 3. When the dam is turned around as shown in Fig. 3, the stability is calculated as follows : The water pressure acts perpendicularly to the face CB, and the center of pressure acts on a line at the intersection of the slope and J H. The force, P, tends to overturn the dam about A with a lever arm DA, which is the perpendicular distance between the line of application of P and the point A. From this, it will be noticed, that the flatter the face CB, the less tendency the water has to overturn the dam. W X FA P X DA is the factor of safety. By examining the above fraction it will be seen that the factor of safety is increased. That is, as the dam is made to approach the gravity type, the overturning tendency of the water is diminished. The above calculations are based on a theoretical dam, such as is never built, because it is impractical to build such a sharp crest, owing to the flow of water. In practice, gravity dams are built similar to Fig. 4; the center of gravity is found by laying off the breadth of the base on the slope side of the crest, and the breadth of the crest on the opposite side on the base; then draw a line connecting the points W. DAMS. 21 Draw a line from the middle of the crest to the middle of the base, connecting points XY. The center of gravity is located at the intersections of lines UV and XY. As gravity dams usually have water flowing over (Fig. 5), the following calculations FIG. 4. FIG. 5. represent the conditions to be considered. The center of gravity is found as in Fig. 4. The vertical line EF is drawn through the center of gravity. The pressure P acts at point Z, and is located according to the formula 7 _H/ _h_ ~ 3 V 1 H + 2 h The overturning effect of the water is the same as before, P X AD. The dam counterbalances the overturning effect of the water with a moment, W X FA. W X FA P X AD factor of safety. IR is the resultant of forces P and W, and is found by the application of paral lelogram of forces. IW is drawn proportional to the weight of the masonry, and WR is drawn proportional to the pressure of the water. To have the dam stable, the resultant, IR, must cut the base in some point as K, which must be at a distance greater than one-third the length of the base from the toe. The downstream side or face of the spillway of the dam must be made to conform with the shape of the overflowing water, and in order to prevent erosion, the foot must be provided with a curved apron as seen in the accompanying illustrations (Figs. 6 to 8). This apron must be designed to withstand the effect 'of vacuum produced by the overflowing water. 22 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Earthen and timber dams have long upstream faces, so that the tendency for the water to overturn them is greatly lessened, and are more fully treated under their respective sub-headings. FIG. 6. Cross Section of Dam (St. Louis River). Great Northern Power Company, Duluth, Minnesota. FIG. 7. Dam at Morgan Fall, Georgia. Chattahoohee River. Masonry Dams. Masonry dams are made either in the gravity or arch type. The stability of the straight gravity dam depends upon its own weight, while that of the arch type depends upon the thrust action at the ends, which rest on the mountain slope. All gravity dams must rest upon very solid foundations. Where this condition cannot be obtained, an artificial one must be made, which can be done, for instance, by driving wooden, concrete or iron sheet piles. Where a rock bed is found, the bed must be cleaned of all earth and loose boulders, and provided with trenches to increase the resistance of. friction by sliding. The bottom of the dam should extend some five or six feet below the river bed to prevent the water from leaking under, which might make the dam fail. The majority of masonry dams are made of solid concrete or of cyclopean masonry (Fig. 8). The mixture used in concrete dams is 1:3:6 or i : 4 : 8, depending greatly upon the size of the dam. Frequently a coarser mixture is used, known as rubble concrete. In large hydraulic undertakings, particularly in the West, cyclopean masonry is employed ; the stones vary in size from cobble-stones up to stones weighing one to two tons. The stones are so placed that the smaller ones fill FIG. 8. Dam at Spier Fall, Albany. Hudson River Electric Power Company, New York. up the places between the big ones, the whole being embedded in concrete mortar. Care must be taken in the construction of these, as well as other dams, to make them A.E.&M. DAMS. 23 * H I as water tight as possible to prevent seepage. To accomplish this, the side of the dam must be faced with a rich mixture of cement mortar or a finer course of concrete; sometimes tiles are used for facing. In addition to this, vertical drain- age pipes are embedded in the concrete to carry off seepage. The pipes, usually 4 inches in diameter, are set in vertical sections, with space between, so that the seepage can enter same, and join mains which discharge on the downstream side. Reinforced Concrete Dams. In the last five or six years, the reinforced concrete dam has been much favored for hydroelectric plants. The design is specialized and involves striking features of the adaptability of reinforced concrete. The principle, which these designers of dams endeavor to preserve, is, that the water pressure applied to a dam renders it not less but more stable, that is, the vertical component of the static pressure is made use of to pin the dam to its foundations, whereas, with the previously discussed masonry gravity dams, the pressure of the water is exerted horizontally (provided the upstream side is vertical) to overturn the dam, which must therefore be made sufficiently massive to resist the pressure by its own weight. The pressure exerted on the foundation of a gravity dam varies, theoretically, from zero at the upstream edge to a maximum at the downstream edge. The maximum must never exceed the crushing strength of the material. Usually a factor of safety of 2- or i A is employed. The slope of the "deck" of a reinforced concrete dam may be so related to the weight and width, that the pressure on the foundation is controlled at the will of the designer. Usually the proportions are such that the diagram of pressure is nearly a rectangle; i.e., the pressure is kept substantially uniform over the whole foundation, and with the excess pressure, if any, thrown slightly towards the upstream angle instead of being concentrated at the downstream edge. This arises from the fact that the resultant of the water pressure and weight of the dam can be held at, or a little above, the center of the base, instead of passing down to the lower edge of the middle third. The movements of this resultant and the base pressures dependent thereon may be followed in the diagram, Fig. 9, in which the resultant, as the dam fills, is seen to advance slightly upstream from the center, until the dam is about three-quarters full, returning again nearly to the center, when the dam is under its calculated flood. The angle of the resultant also is always kept within the limit of the angle of friction, so that the dam has no tendency to move on its base. Fig. 10 shows about the simplest form of dam adapted to moderate heads and hard foundations. It consists of a series of buttresses variously spaced from 12 feet to 18 feet apart on centers, and covered with a deck of concrete, reinforced between the different bays as a beam after the usual formula. The factor of safety throughout is said to be never less than 5 in all its relations. But little reinforcement is used in the buttresses, except at the edges and around the openings, which are left for con- venience and to save material. The deck reinforcement, however, is abundant, and is within i| inches of the lower side, leaving from 10 inches to several feet of concrete between the steel and the water. The thickness of the deck necessarily increases from top to bottom with the increase of head. The concrete in the deck is mixed Of CA. 24 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. 9. Behavior of Resultants in Solid Dam. W 5 10 Flood FIG. 10. Behavior of Resultants in a Concrete-Steel Dam. DAMS. 25 1:2:4, usually with fine aggregates, and is poured into the forms in the condition known as " slop concrete." This insures a thorough coating of the steel with cement, and furthermore, insures a density of concrete which seems to be sufficient to forestall porosity altogether. A dam of this design, when on rock, has no continuous base, and therefore cannot be threatened by water pressure finding its way through seams in the rock, and exerting a lifting pressure on the dam. On gravel or other porous foundations, an artificial base is first laid down covering the entire area, but in such cases, this base or floor is pierced with numerous "weep holes," so that upward pressure is again forestalled. Being hollow, reinforced concrete dams not only possess the unusual feature of interior inspection, but the hollow space puts at the disposal of the engineer a valuable space from which to work the various adjuncts, such as flashboards, waste gates, log sluices, movable crests, etc., all of which are handled from the inside of the dam, allowing the whole width of the river to be utilized for rollway, instead of being more or less obstructed by bulkheads. The interior of the dam admits of plenty of space for a passageway, which may vary from an ordinary foot bridge to the equip- ment of a complete power plant as seen in Fig. n. FIG. ii. Patapco Dam, Ilchester, Maryland. The spacing of the piers, which leaves a free waterway, often enables this type of dam, on certain foundations, to be built without the use of a coffer dam, by first carrying up piers in caissons to a uniform grade a few feet above the ordinary water level, and then completing the superstructure while the water is allowed to run freely between the piers. When the dam is completed, these spaces are subsequently and permanently closed with concrete. Referring to Fig. .11, this dam, 200 feet long and 30 feet high, is located near Ilchester, Md., and crosses the Patapsco River. The power house is located inside, 26 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. and is equipped with three 5oo-HP. turbines, provided with draft tubes extending vertically into the tailrace; the generators are direct connected to the turbines; the switchboard and other electrical equipment are also located inside, so that the whole plant is housed inside of the dam. As this plant has been in operation for some time, no trouble has been experienced due to moisture; however, the power house itself is inclosed with four-inch walls of ferro-inclave, entirely separate from the structure of the dam itself. Another power plant, where a reinforced concrete dam has been employed, is given in Fig. 12, showing the general arrangement of dam, forebay and power house, of the Bar Harbor and Union River Power Company, Ellsworth, Me. The dam is 450 feet long and 64 feet high. The conditions were such, that a semi-attached power house was necessary at right angles to the dam, and necessarily supplied from a forebay. Access to the power house and waste gates may be had through the body of the dam, which is entered on the opposite end from the power house. Part of the dam is utilized as a machine shop, storeroom, etc. A sluice gate in the crest is provided, to flush away the accumulated trash which may lodge against the dam. The details of construction of the dam are given in Figs. 13 and 14, and are self- explanatory. This dam, as well as those above described, was constructed by the Ambursen Hydraulic Company, Boston, Mass., to whom the writer is indebted for data on reinforced concrete dams. Coffer Dams. In most cases, when dams are built, coffer dams are necessary to hold back the water, so that the construction of the main dam can be carried on. The coffer dams are built so that a section of the stream or the entire river is deflected. They are temporary constructions, and are removed after the main dam is com- pleted. In shallow and still water, they may be built of gravel and clay, or bags filled with gravel and clay, or bundles of fagots, between which is placed gravel and clay. When the water is deeper and a current exists, this material is apt to be washed away; in such a case, sheet piling is used. Where single sheet piling is not sufficient to withstand the current, two rows of sheet piling are used, the space between being filled with puddle; this construction, of course, ., has to be properly braced, and, as it is composed mostly of wood, it is becoming very expensive; owing to this fact, it has been replaced in recent years by sheet steel piling. This sheet steel piling is made of rolled iron, such as Z-bars, channels, I-beams, and in some cases, specially rolled forms; they are so placed that they are interlocked and kept water tight. This system is very much favored, particularly in large construction work. They are easily driven home, and after the work is finished, they may be used again for other or similar purposes. Crib Dams. Where the bed of the river is rock or near to it, the sheet pile coffer dam cannot be used; in place of it, the crib dam must be substituted. Before a crib is sunk, soundings must be made, to ascertain the contour of the bottom; in some cases divers are sent down. The lower part of the crib is made on shore and floated to the place where it is to be sunk, which is done by filling the same with rocks. As the crib sinks, the remainder of the crib is completed. These cribs are made in DAMS. >> CU c .2 e P PQ < o I I 0) bo O 28 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. : J 2"< 'torizontal Wafer Pre$s..< ~ 2,376,570 lbs.~per 15%. \ \v . weracre Pressure = 165 Ibs. oersa'.in. ' Offset El. ?.5~" EN&. NEWS. \ Top of Dft IZ"< W-^3'l ''n Buttress No.17. Openings shown in Full Lines made in " V Buttresses Jtvfc v" ::: ""'' 1 3*| FIG. 15. Coffer Dam. FIG. 16. Timber Dam. Timber Dams. In a timber gravity dam, the timbers are placed alternately parallel to and crosswise the stream, the spaces between being filled with earth and stone. The bearings of the timbers are either notched, or spiked by iron drift bolts. If the dam is built for retaining water only, the upstream side is built on a slope, while the downstream side may be vertical. If the water overflows the dam, the downstream side must be on a slope, in order to prevent the water washing away the river bed in front of the dam. There are several examples of failures of timber dams due to the erosion of the river bed. Trautwine states: 1 the Jones Dam at Cape Fear 1 Trautwine, Civil Engineers' Pocket Book. 30 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. River had a height of 16 feet, and the usual water fall was 10 feet into a pool 6 feet deep, and in a few years wore out the soft shale rock, undermining the dam to such an extent that it gave way. The timber dam at Holyoke is another example of the erosion of the river bed due to the falling of the water, so that the dam had to be reinforced by a downstream apron. The apron of the downstream side of the dam should be on an angle of about 30 degrees, so that the water has an easy overflow and protects the river bed. There are other forms of timber dams, , the frame types. They are built in framework of various forms, which vary according to conditions. The timbers are framed and strongly braced, upon the top of which is placed sheet planking. As they are lighter than the rock-filled timber dam, the upstream side must be on a longer slope, so that advantage may be taken of the weight of the water to secure greater stability. These dams are easily built, and in certain sections of the country they are the cheapest form for hydraulic power developments. However, with the diminution of the forests and the cheapness of masonry and iron, the latter is more profitable to use. FIG. 17. Hauser Lake Steel Frame Dam. Steel Frame Dams. Similar in construction to the timber framework dam is the steel frame dam, as seen in Fig. 17. This dam has been erected by the Helena Power Transmission Company, across the Missouri River at Helena, Mont. 1 During a period of the year, about two to four months, it has to stand high floods and act as an overflow dam. It is about 75 feet high and 630 feet long. The lower section of the upstream side is of concrete, behind which is rubble masonry; the upper section is made of structural steel trusses 9 feet 9 inches apart. The entire upstream side is faced with steel plates; those on the bottom extend into concrete and fasten to sheet piling beneath the river bed, so as to prevent the water from washing beneath 1 Zeitschrift des Vereines deutscher Ingenieure, April 18, 1908. DAMS. 31 the dam. These, as well as those on the upper section, are flat plates, five-sixteenths of an inch thick, while the middle are concave and three-eighths inch thick. Only the upper section of the downstream side is faced with steel plates; the lower section is made of timber and faced with planking. On the top of the dam is a flashboard structure faced with steel plates at both ends, and in the middle is an opening 50 feet long to let through floods, and can be closed after the flood has passed. This dam failed on April 14, 1908, and at the time of the accident, the matter of acceptance and final settlement was in the hands of the attorneys of the power company. The following is an abstract report given by the Electrical Review. 1 ''The initial break occurred at bent 39, about 400 feet from the east or power- house end of the dam. The anchorage at this point apparently gave way, breaking the seal, and allowing the water to pass under the rubble masonry fill. The water rapidly cut away the gravel, permitting settlement of this upstream masonry, and carrying down with it the lower end of the girder, forming the upper member of the steel bent. The expansion joint in this girder, and in the plates of the dam just above the top of the masonry, gave way, leaving the bents and plates unsupported in such a manner that the water pressure pushed over this section. About six minutes elapsed from the time the water first came through under the rubble masonry until the expansion joint failed and the first bent toppled over, carrying out a section about 30 feet in width. The tremendous rush of water rapidly widened the breach, the foundations on each side were undermined, and the posts and steelwork buckled at right angles to the direction of the flow of the river. The bents continued to give way and fall until the breach widened to nearly 300 feet." Earth Dams. The earth dam is the oldest type of dam known, and is still used in hydraulic developments. They are made of loam, clay, and rock, or a combination of same. These dams are used principally in still water; however, if they are intended to be used as an overflow dam, they must be properly faced so that no erosion can take place. The upstream side usually has a slope of 2 : i, while the downstream i *&d&&&&* ; - ?,i: l'-tl; fli^-fyssj i ; ! ';&<<' ' ! FIG. 20. General Arrangement of Stoney Roller Sluice Gate, at Beznau Plant, Switzerland. 34 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. that a gate under a pressure of 300 or 400 tons can be easily moved by hand. A type of this gate is shown in Fig. 20, seven of which have been installed in connection with the Beznau plant, Switzerland. These gates are 49 feet wide and 20 feet high; they are made of structural steel, and provided on the bottom with a square timber, resting on a cast-iron shoe embedded in the concrete of the dam. The sides of the gates are made water tight by steel ropes which are held against the joint by the pressure of the water (see Fig. 21). FIG. 21. Details of Stoney Roller Sluice Gate. Each gate may be operated by hand, two men being necessary, or for quick operation, there is a portable 8-HP. motor. Plans are at present in preparation for the installa- tion of an hydraulic turbine at the dam, for automatically operating these gates in case of emergency. The discharge of the water from this type of dam takes place from underneath as the dam is hoisted; thus the foreign material which collects on the bottom is easily discharged. Butterfly Dam. Another type of movable dam is the butterfly, an example of which is given in Fig. 22. Two of this kind have been installed in connection with the Chicago Drainage Canal, 1 one being 12 feet and the other 48 feet wide. The two movable crest dams are practically alike in details of construction and operation. Each movable crest is built of structural steel shapes and steel plates, and is practically a 45-degree sector of a cylinder with a 26-foot radius. Each sector is hinged horizontally along the axis of the cylinder of which it would form a part, to 1 Movable Crest Dams at the Water Power Development of the Chicago Drainage Canal. The Engineering Record, Aug. 24, 1907. DAMS. A.E.&M.i UNlV. OF CA The radial '"~" the top of a back wall on which it is mounted on the downstream side, deck plane and the curved upstream front of the sector are made water tight with steel plates, the deck being provided with steel angles for ice skids. The deck plane, the lower radial plane and the curved face' are heavily reinforced by intermediate steel frames. When the crest formed by the intersection of the curved face and the radial deck plane is at the maximum operating height, the lower radial plane of the sector is horizontal. As the crest is lowered, the sector rotates on its axis and moves FIG. 22. Butterfly Dam, Chicago. Drainage Canal Power Plant. FIG. 23. Detail of Back Hinge of Butterfly Dam. into a space in the concrete base, which is also approximately a sector of a cylinder, of about the same radius as that of the crest, the radial deck being horizontal when the crest is at its lowest position. The crests of both dams have a vertical range of 18 feet, from 2 feet above to 16 feet below Chicago datum, the water surface above the dams being 4 to 6 feet below that level under normal conditions of flow. Bear Traps. The old bear-traps consisted of two leaves, hinged to the foundations. The upstream leaf overlaps the downstream leaf when the gate is lowered. A culvert leaves from the river upstream, to the space under the leaves, and a second culvert from this space to the river downstream, and are provided with valves. When the first culvert is opened with the second closed, the hydraulic pressure under the leaves causes them to rise, provided the head from the upstream culvert is sufficient. Reversing the process, the leaves will fall. The interior angle formed by the leaves in the raised position must not be less than 90 degrees, since, if it were, the trap when once up, would not fall under the action of the hydraulic forces. The angle should be about 100 to 105 degrees, and if the angle is too great, the width of the base will be excessive in proportion to the height of the crest above the foundations. The principal defects of the old bear-trap, as given by P. S. Bond, 1 are as follows: I. Sliding friction between the leaves. II. Width of base too great for height attained. III. The overlap of upper upon lower leaf. IV. Inability to raise and fall uniformly (tendency to warp). 1 The Permanent Improvement of the Ohio River. The Engineering Record, Jan. 16, 1909. 36 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. V. Necessity for initial head in raising. VI. Difficulty of stopping without shock in raising. VII. Difficulty of operation in wide passes, and division into several sections by piers. VIII. Leakage at time of raising. IX. Liability of binding on debris along side walls, and driftwood lodged in the exterior angle between the leaves. X. Improper proportioning of leaves (unscientific design). XL Cost. Many improvements have been made in late years, which relate more or less to the mechanical construction. Cylindrical Dams. On the continent of Europe, several cylindrical rolling dams have been installed. Two prominent ones are located in the rivers Main and Sau at Schweinfurt, Bavaria; 1 the former is 13.5 feet in diameter and has a clear width of about 60 feet; the other one has a diameter of 6.5 feet and a length of 115 feet. It is practical to make them 39 feet in diameter and 150 feet span. These dams are nothing more than two concentric shells, the space between being air tight. The inner shell is open at the ends, so that when the dam is lowered, the water flows through, thus reducing the buoyancy effect. The dam is raised and lowered by a chain or cable wound around one end; both ends roll on cogwheel tracks. The principal claims for this type of dam are, the elimination of piers in the center of a river, simplicity of construction in dam as well as machinery, and ease of opera- tion. The dam may be easily raised above the river level, thus giving a free and unobstructed passage. Needle Dams. A system for the temporary impending of water is the needle dam, consisting of a row of squared timber or heavy planking set upright against a trestle. In case of excessive flood, a number are removed and the water released. As these needle dams are usually built across the entire width of the river, the trestle remains as an obstruction to floating material when any of the needles are removed. Chanoine Dam. The objection in the needle dam mentioned is overcome in the Chanoine Dam, which can be lowered, thus giving free passage; it will tip auto- matically when the water rises to a certain height overflowing the crest, similar to permanent flashboards. The movable parts of the Chanoine Dam consist of a row of wickets hinged on horses, and held in place by props. A detailed description of this system will be found in The Engineering Record. 2 Flashboards. In order to take care of surplus water during flooding periods, or minimum use of water, flashboards are employed to impend water for dry season, or maximum demand. They are designed to withstand a certain amount of water, which would otherwise discharge over the dam; should the pressure exceed the designed limit, the supports give way and the boards are washed downstream; 1 Wegmann, The Design and Construction of Dams. * Permanent Improvement of the Ohio River P. S. Bond. The Engineering Record, Jan. 9, 1909. DAMS. 37 this of course in most instances would mean the loss of planking. However, in many instances, heavy floods are anticipated, and the flashboards are removed before the flood reaches its height. Permanent flashboards are so arranged that part or the whole are removed, should the water rise above the limit. They may be built of structural steel; placed and removed from a footpath carried above the dam. Another type of flashboard is of structural steel held in an upright position by rods hinged below the center of the board. When the pressure of the water above the hinge exceeds that exerted against the flashboard below the hinge, the flashboard drops over automatically; it is not washed away, but held by the anchor rods. Fishways. Fishways are frequently required in connection with dams, in order to provide a passage for fish which return upstream to their breeding places during certain seasons of the year. In most countries, it is specified by the government whether they have to be installed or not; the size of these fishways depends upon the kind of fish and their habits; data on this subject can be obtained from the govern- ments as well'as local authorities. FIGS. 24 and 25. Types of Fishways. These fishways are always located on one side of the dam, the outlets being at the bottom of the dam, because the fish usually gather there. The principle of a fishway consists in retarding the velocity of the water in an inclined trough provided with obstructions, so that the mean velocity will be no more than 6 or 8 feet per second, with resting places made by the nature of the obstructions. Such a passage, in most cases, is nothing more than a series of steps forming cascades, between which are pools of water as indicated in Fig. 25. Another form of fishway is seen in Fig. 26; it is nothing more than a chute which reduces the velocity of the water by the friction of a zigzag course. These fishways are made either of wood or masonry. 38 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. BIBLIOGRAPHY. THE DESIGN AND CONSTRUCTION OF DAMS. Edward Wegmann. 1899. John Wiley & Sons. New York. RESERVOIRS FOR IRRIGATION, WATER POWER AND DOMESTIC WATER SUPPLY. James Dix Schuyler. 1901. John Wiley & Sons. New York. MASONRY CONSTRUCTION. I. O. Baker. 1903. John Wiley & Sons. New York. NOTES ON STRESSES IN MASONRY DAMS. Max aus Ende. Engineering, Dec. 8, 1905. INTERNAL STRESSES IN MASONRY DAMS. S. D. Bleich. School of Mines Quarterly, November, 1905. RECENT PRACTICE IN HYDRAULIC-FILL DAM CONSTRUCTION. J. D. Schuyler. A. S. C. E., October, 1906. MASONRY DAMS. Th. G. Bocking. Engineering, Sept. 27, 1907. MASONRY DAM FORMULAS. Orin L. Brodie. School of Mines Quarterly, April, 1908. THE DESIGN OF BUTTRESSED DAMS OF REINFORCED CONCRETE. R. C. Beardsley. Engineering News, April 23, 1908. A FORMULA FOR CALCULATING FLASHBOARDS FOR DAMS. Richard Muller. Engineering Record, Aug. 22, 1908. THE DESIGN OF RETAINING WALLS. Harold A. Petterson. Engineering Record, June 13, 1908. A PROPOSED NEW TYPE OF MASONRY DAM. Geo. L. Dillman. Trans. A. S. C. E., vol. 49, p. 94, 1902. THE CORRECT DESIGN AND STABILITY OF HIGH MASONRY DAMS. Geo. Y. Wisner. Engineering News, Oct. i, 1903. ON THE STABILITY OF MASONRY DAMS. Karl Pearson. Aug. n, 1905. ON THE DISTRIBUTION OF SHEARING STRESSES IN" MASONRY DAMS. W. C. Unwin. Engineering, June 30, 1905. THE STRESSES ON MASONRY DAMS. Engineering, September, 1907. DESIGN OF AMERICAN DAMS. Engineering Record, Feb. 20, 1902. THE LIMITING HEIGHTS OF EARTH DAMS. Engineering Record, Dec. 7, 1901. EARTH DAMS WITH CONCRETE CORE WALLS. Engineering News, Sept. 7, 1905. RECENT PRACTICE IN HYDRAULIC- FILL DAM CONSTRUCTION. Proc. Am. Soc. C. E., October, 1906. A COLLAPSIBLE STEEL DAM CREST. J. C. Wheeler. Engineering News, Oct. 3, 1907. A NEW AUTOMATIC MOVABLE DAM. Engineering Record, March 8, 1902. MOVABLE DAMS. B. F. Thomas. A. S. C. E., March, 1898. AMERICAN TYPES OF MOVABLE DAMS. Hiram M. Chittenden. Engineering News, Feb. 7, 1895. MOVABLE DAMS, SLUICE AND LOCK GATES OF THE BEAR-TRAP TYPE. Archibald O. Powell. Jour- nal Association Engineering Societies, June, 1896. CHAPTER III. HEADRACE. Scheme. The arrangement of headrace is governed entirely by natural conditions. While some plants require a dam only, for securing water supply, others require in addition, miles of headrace, including expensive tunneling and installation of high- pressure penstocks. These are the extremes between high and low head plants. The ultimate aim in both cases is, to secure the greatest amount of energy with least expenditure both in first cost and cost of operation. Therefore, the building containing the turbines should be located so as to utilize the most efficient head. In some cases, tailrace water is discharged directly into the headrace of a plant located below. FIG. i. Typical General Arrangement of a Hydroelectric Development. Fig. i illustrates a way of harnessing a stream and conducting the water to the power house, or, in short, a complete hydraulic installation. In connection with same, the tailrace of a previously installed plant is utilized. This particular plant has been selected because it contains most of the features to be met in harnessing 39 40 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. water for hydroelectric developments. It utilizes the water of the river Sill, and supplies the capital of Tyrol, Innsbruck, with light and power. In the illustration, the letter C designates the tailrace of a 6ooo-HP. plant, a, which joins the headrace at the sluice gates h; g are sluice gates to control the water in the river; e is a spillway dam;/ and /represent racks; one prevents foreign material from entering the entrance basin from the river, and the other, a finer one, prevents material from entering from the headrace. At the lower end of the entrance basin is a sandtrap designated by k; there is also a sandtrap k at the side of the sluice gate /, before the water enters the tunnel; i is a spillway in the flume to discharge surplus water. The letters m designate shafts leading to the headrace tunnel, which is 4.7 miles long. There are seven shafts in all. The tunnel has a slope of i foot in 1000. It takes one hour for the water to travel from the entrance basin to the collecting reservoir. The letter n is an overflow wall in the reservoir, while p is a sluice for emptying same. The water is discharged down the mountain slope, in cascade to break the fall, and joins the tailrace n. As the plant is located in a district of frequent and heavy snowfalls, a snow sluice o is provided. The water, before entering the penstock r, must pass through a fine screen/ then through a sluice g. While the velocity of the water in the tunnel is 7.4 feet, the velocity in front of the screens, before entering the penstocks, is only i foot per second. At the bottom of the screen, which sets on the skew, is another sandtrap which discharges into the cascade. It will be noticed that there is only one penstock r in place, to supply three units; it is about 11,000 feet long, and laid on the mountain at an angle of 33 degrees. The head is 602 feet to the center of the turbine shaft. As the friction loss is 5.5 feet, the effective head is 596.5 feet. The penstock is made up of steel plates in sections, and the diameter is 4.1 feet, the upper section being of five-sixteenths material, and the lower section thirteen- sixteenths inch. The turbines are of the impulse type, mounted in pairs on one shaft; when running at 350 R.P.M. and with an efficiency of 80 per cent they develop 2500 H.P., and consume 45.4 cubic feet of water per second. CONDUITS. Water may be conducted by the following methods: 1. Canals. 2. Tunnels. 3. Penstocks. The canals may be subdivided into trenches and flumes. The tunnels may be non-pressure and pressure. The penstocks always operate under pressure, and may be built of steel, wood or reinforced concrete. Cross Section of Canals. The cross-section area of a canal, may it be a trench or flume, should be such that the water will rise to about three-quarters to seven- HEADRACE. eighths of the height, but should never be higher than the latter. The rectangular form is the most common one in use, and is so proportioned that the depth of the water is about half of the width. The slope of the canal depends on the degree of smoothness of the bottom, and varies from one-half to one foot in a thousand; the latter is more common. Trenches. The most common form of canal is an open trench dug in the soil, and the sides sloped according to the firmness of the soil, usually i : i. If loose soil is encountered, the form of the canal must be such that the velocity is about 2 to 3 feet per second. If a higher velocity is taken, the sides and bed will be disturbed. If good loam is found, the velocity may be taken as 4 feet per second. This may be increased to 4.5 to 5 feet by lining the sides and bottom with paving stone and gravel. According to Bazin's formula, the bottom and mean surface velocity may be found as follows: v max. v 25.4 \/rs\ v = v b + 10.87 \/rs\ .'. v b = v 10.87 ^^s. v = mean velocity in feet per second. max. v = maximum surface velocity in feet per second. v b = bottom velocity in feet per second. r = hydraulic mean depth in feet = area of cross section in square feet divided by wetted perimeter in feet. s sine of slope. The lowest velocity is found in the wetted perimeter. The different velocities, according to Rankine, are in the ratio 2 : 3 : 4 in low-velocity canals, and 3 : 4 : 5 in high-velocity canals. The greatest velocity is found in the middle slightly below the surface. Ganguillet and Kutter give the following table I, of safe bottom and mean veloci- ties in channels^ calculated from the formula, v = 10.87 The results obtained by using this formula are very low, as admitted by the above authorities. TABLE I WATER VELOCITY IN CHANNELS. Material of channel. Safe bottom velocity (vt) in feet per second. Mean velocity (v) in feet per second. Soft brown earth ... o. 249 o. 328 Soft loam O 4OO o. 6s6 Sand I OOO I . 312 Gravel . I 908 2.621; Pebbles 2. OOO 3.918 Broken stone, flint 4 . OO'3 c. cyo Conglomerate, soft slate 4.988 6. ^64 Stratified rock 6.006 8. 204 Hard rock . 10. 009 IV 127 4 2 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The following figures are selected from a diagram by W. A. Burr * showing the resistance of various soils to erosion by flowing water. TABLE II. MAXIMUM WATER VELOCITY IN CHANNELS. Material. Velocity in feet per second. Pure sand I. I Sandy soil, 15 per cent clay I. 2 Sandy loam, 40 per cent clav 1.8 Loamy soil, 65 per cent clav 3. o Clay loam 85 per cent clay. . . 4.8 Agricultural clay 05 per cent clav .... 6 2 Clav 7. 1$ Masonry Flumes. Flumes are made either of masonry or of planking. The former resembles the trench, but has vertical walls. They may be made of brick, concrete or reinforced concrete. Those made of concrete are the most common ones in use, and usually follow the contour of the ground. If concrete flumes have to cross valleys, either filling has to be made or else solid pier construction has to be used, in a way similar to the old Roman aqueducts, in which case reinforced concrete may be successfully employed. The wetted perimeter must be smooth, so as to allow a velocity of 7 to 8 feet per second. The bottom must have a slope of one-half to one foot in a thousand. As an example of modern concrete flume construction, the one of the Kern River Power Plant is cited: The whole structure is carried on 1 5-inch I-beams set 8 feet 10 inches apart, supported by concrete piers. These longitudinal girders carry 9-inch steel I-beams laid transversely 4 feet center to center, and on them is erected a framework of structural steel for the sides and bottom of the flume. Two layers of expanded metal of 1.5 and 3-inch mesh are used in connection with this framework, and, being embedded in concrete, form the sides inclosing the frame. This concrete construction is also reinforced on the floor by twisted half-inch rods. The outside and inside of the flumes are plastered, making the thickness of the reinforced concrete sides and bottom 4 inches. This type of flume or conduit, while it costs more than a wooden flume, has the advantage of being as permanent as tunnels themselves. Wooden Flumes. Wooden flumes, which are used mostly in the West and Pacific Coast, are constructed of California fir, redwood and Oregon pine. They are best carried on trestlework or concrete piers, and are usually of the open type, and built on a slope of one-half to one foot in a thousand. The planking must be laid so that the pieces break joints. The wetted perimeter must be smooth to allow a velocity of 7 to 8 feet per second. They must be water tight, which may be done as given in an example below. 2 1 Engineering News, Feb. 8, 1894. 3 Kern River Power Plant No. i, by C. W. Whitney. The Engineering Record, Aug. 10, 1907. HEADRACE. 43 With the installation of this plant, there are five wooden flumes/ the; longest being 1030 feet, the shortest being 50 feet. They are placed on concrete 'foundations, and are designed with a factor of safety sufficient to make their life from 30 to 40 years. The framework for supporting the flume box is of Oregon pine, being so designed and distributed that no part of the timber comes in contact with the earth, or is exposed to the drip should the flume at any point spring a leak. The flume box is built up of 3 by 12-inch redwood planks obtained from butt-ends of Sequoia Semper Virens, grown in swamp lands west of the coast range of northern California. The grain of this lumber is perfectly clear, and its quality is such that its life should not be less than forty years. The edges of all planks were beveled so as to give a one-quarter inch opening on the inside of the joints, which is calked with FIG. 2. Detail of a Timber Flume, 32-foot Span. ship-chandler's oakum. The bottom seams were covered with hot asphaltum and i by 6-inch redwood battens nailed down over them. On the sides of these flumes a specially designed batten is used. This batten is of i by 6-inch redwood, the upper half being cut away on a curve, permitting asphaltum to be poured between the batten and the side of the flume. At the corners of the flumes a quarter-round strip is nailed (Fig. 2). The design of the flume above described has been thoroughly tested, and even if it should stand dry for months in the hottest weather, the designers stated that it may be again filled with water without having any perceptible leakage. In some of the flumes where streams are crossed that are apt to carry considerable 44 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. AM ixv'i'h. winter, 'span flumes are constructed. One of these span flumes has a length of 32 feet, built 'with a 10 by 12-inch timber frame, resting on 12 by i2-inch beams. In connecting the wooden flume with the portal of a tunnel, a construction of a special nature was used, which offers two points of contact between the wood and the con- crete, and a well between the two, from which the water may be pumped out, and any leaks repaired should these ever occur between the wood and the concrete. ,Wiro nails FIG. 3. Timber Flume, on Mountain Slope. American River Electric Company. Protection of Flumes. Frequently the flumes run on the sides of mountain slopes, and are endangered by loose boulders. Where flumes pass through such sections, they should be provided with some means of protection. One way of protecting the flumes is to build a retaining wall of sufficient height on the mountain side of the flume, to deflect the boulders across same. Another way is to cover the flume with concrete slabs, preferably reinforced with rods. Where loose boulders or land-slides are severe, the cover should have an arch form, and in any case, should be covered with at least two feet of earth to act as a cushion. HEADRACE. 45 Tunnels. As stated, tunnels for hydraulic developments are classified as non- pressure and pressure tunnels. In the former case, the tunnel is only partly filled, and in the latter, it is completely filled and under pressure due to the head of the water. The cross section of the tunnel may be semi-egg-shaped, or rectangular with an arched roof. Where the tunnels run through loose soil, they must be lined, the thickness of which varies with the character of the soil and the pressure required to retain same. The lining may be made of brick, but in later years, concrete has been used exclusively. Where there is a possibility of a cave-in, they must preferably be reinforced with rods. The cross section should be uniform throughout, except near the collecting reservoir, so that it may serve as an additional storage. They are usually built with a slope of i to 2 feet in 1000, and have a velocity of 7 to 8 feet per second. Where the tunnel is several miles long, it is advisable to provide at every mile, access to same; this is usually done by vertical shafts. In cases where overflow side-tunnels are provided, these shafts may be eliminated. Pressure Tunnels. In the last few years several plants have been installed with pressure tunnels, particularly in Europe. These tunnels are thoroughly and strongly lined, and as water tight as possible, because they are under pressure. Such tunnels are provided with a vertical shaft, the upper end of which is enlarged to serve as an air chamber and absorb fluctuations. One of the most notable examples making use of this system, is that of the Urfttalsperre installation, Germany. This plant is operating under a head varying from 230 to 360 feet. At the end of a 885o-foot tunnel is located the vertical or equilizer shaft. The top of this shaft, sunk through the mountain, is higher than the high-water level in the reservoir, so as to prevent, in case of sudden shut-down of the plant, a waste of water. It will be seen from this, that the equilizer shaft acts as a standpipe, similar to those installed on penstocks. These shafts are more economical than standpipes because they do not waste the water. With exceptionally long pressure tunnels, it is advisable to install vent pipes along the line, to let out air which might collect, and prevent same from getting into the penstock. Friction in Tunnels. In order to minimize friction in tunnels, the perimeter must be smooth, which is accomplished by a cement coating. In case of non-pressure tunnels, the coating should extend some 6 inches above the highest water level; while in pressure tunnels, the coating must extend over the whole. of the interior surface. Where the tunnel is cut through hard rock, and a smooth surface is easily obtainable, the lining may be omitted and only a coating provided. This coating has to fill up small recesses in the rock, and to face the concrete lining. It is made of a mixture of one part sand and two of cement, and applied about a quarter of an inch thick. Another way to reduce friction, is to have the course of the tunnel as straight as possible; short radius curves must be avoided. Where the side walls join the bottom, the junction must be a smooth curve. Seepage in Tunnels. Where tunnels run through mountains and seepage water is encountered, provision must be made to take care of same. In high-pressure 4 6 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. tunnels, this trouble amounts to little or nothing, especially as these tunnels have to be constructed as water tight as possible. In non-pressure tunnels overflow devices may overcome the difficulty. These overflow devices are nothing more than overflow weirs, discharging into side tunnels taking the shortest cut to the mountain slope. Fig. 4 shows such a device, as installed in the Brusio Power Plant, Switzerland. The tunnel of this installation is 1756 feet long, built on a slope of 2 feet in 1000, and FIG. 4. Overflow Device in Headrace Tunnel, Brusio Plant, Switzerland. is provided with u overflow devices. It will be noticed that the main tunnel is only partly lined. In connection with same, the horsepower carrying capacity is indicated; of course this capacity applies only to this particular plant. Construction of Tunnels. In the construction of long tunnels, which in some cases represent the greatest expenditure in hydroelectric engineering, temporary power plants are installed, particularly when plants are remote from power supply. Tunnels are usually begun at both ends, arid in some cases at intermediate places, where shafts have been driven. Where intermediate junctions are to be made, and the character of soil is known, and varies from rock to soft earth, the junctions are best made at such places. For cutting rock, pneumatic or electric drills are employed; two or more drills are mounted on a truck running on rails. Where the tunnel is driven through loam, special cutting machines which run on tracks may be employed. Cases sometimes arise, particularly in Switzerland, where the tunnels have to be cut under pressure; when such is the case, it is customary to heavily line the tunnel as fast 1 as the work proceeds. Siphon System. In some installations, in order to utilize the water of a mountain lake, the bottom of same has to be tapped, for which purpose sheet piling is driven. A.E.&M.E HEADRACE. 47 The tapping of a lake on the bottom is very troublesome, particularly "when the soil is soft, and might cause failure in construction. Swiss engineers, in such a case, have adopted the siphoning system, which consists of sinking a vertical shaft, a safe distance from the shore of the lake, to which the headrace tunnel is connected. The water is siphoned from the lake into the shaft. One leg of the siphon is submerged in the shaft, and the other is carried on a trestle out into the lake, and extends down into same as far as possible, to take advantage of all the water available. To the author's knowledge, the first siphon system installed was that of the Kubel plant, and the largest one, in connection with the Brusio plant, both in Switzerland. TT "77 FIG. 5. Siphon System at Lake Poschiavo. An illustration of the latter is given in Fig. 5. The siphon tube is 6.5 feet in diameter and is made of /g-inch material. As it is submerged below the normal water-bed of the lake, it will start automati- cally. The horizontal part of the siphon is placed on a slope of i : 1000, and has, at the highest point, two pipe connections, one 3.5-inch for air pump and the other an 8-inch centrifugal pump connection. If for any reason the siphon should stop operating and the water level is low, either of the pumps may reestablish the siphoning action. Both ends of the siphon are provided with controlling valves. In addition to this, the suction leg has a one-inch mesh screen, which may be cleaned by breaking the siphon, and allowing the water to flow back, or the centrifugal pump may be applied to same. RACKS AND GATES. Racks. There are two different kinds of racks, one in which the bars are spaced far apart, and the other where they are near together. The former is known as rack, the latter as screen. The racks must be placed at the entrance to the forebay or the entrance to headrace, in order to keep the heavy floating material out. In small plants, these racks are made in sections, so that they may be easily removed and cleaned. In large plants, the racks are stationary and are made of heavy bars, 4 8 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. or sometimes light I-beams, all depending upon the force with which the floating material acts on the racks. No round bars or pipes should be used for this purpose, although the latter is sometimes used. Racks, made of round material, will collect more foreign matter than flat or square bars, because the latter deflect better than the former. Further, it is more troublesome to clean racks with round bars than with flat ones, as the wedge action of floating material is greater with round than with flat bars. The space between the bars should not be less than 1.5 inches, and should never exceed 4 inches. This FIG. 6. Screen House, Plant No. 2, Niagara Falls Power Company. is particularly true when the racks are placed at the entrance of a long inclosed head- race, otherwise deflectors must be installed, to deflect material which passes the outside racks. If this is not done, there is a possibility of the headrace becoming clogged. Racks may be set vertically or at an angle, as long as the facility for clean- ing is not sacrificed. Bars 3.5 by 0.5 inches, bolted together with separators, give a suitable construction for average conditions. In plants where heavy material has to be deflected from the headrace, an arrange- ment-similar to Figs. 7 and 8 might be adopted. It is a combined regulating device, rack and deflector, and has been installed in the headrace of the Hafslund Power Plant, Norway. As this is made of heavy I-beams, much clearance is allowed. Any material which passes the openings is prevented from entering the penstocks by two other racks, one rough and one fine. The former is in front of the forebay and provided with a sandtrap; the latter is directly in front of the penstocks. HEADRACE. 49 As is seen in Fig. 8, there are two water levels, which are controlled by hoisting or lowering the I-beam rack. Even if the gate is open, the elevations will differ, owing to the deflector extending some 19 feet into the water. In certain seasons of the year, the rack is lowered so that anchor ice will either be broken up on the rack, or forced upward and deflected by the wall. Because the width of the headrace is 32.8 feet, it is necessary to split the rack into sections, to facilitate handling. By ">v^ ^^\ FIGS. 7 and 8. Deflector and Rack in Headrace, Hafslund Plant, Norway. means of a windlass, which travels on a bridge, the different sections are hoisted and fastened to the latter. This serves a twofold purpose: first, reducing the weight to be lifted; second, the regulation of the water is better accomplished. Screens. Screens are frequently made of 2 by f-inch bars. They are spaced from f to | inch apart. These screens, in almost all cases, are made in sections, with the exception of those at small-sized plants; even then, two screens are provided, so that when one is being cleaned the other is lowered. Such is seen in Figs. 3 and 4, which gives the general arrangement as well as construction. It will be noticed that there are two screens hoisted and lowered by a windlass. There is a walkway supported on brackets in front of the upper screen, so that the attendants may have free access for cleaning. The lower end of the bars of the rear screen are bent, so that when the screen is hoisted, the floating material goes with it. 50 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. It is advisable to place screens on an angle from 45 to 60 degrees; this not only brings submerged floating material to the surface, but adds greatly to the passage area of the screen, and reduces the velocity of water. Such screens are arranged in one row, so as to cover a number of penstock inlets; they are divided into sections, horizontally and vertically, and slide in vertical guides made of cast iron or rolled steel channels. When they are divided into vertical sections, they overlap each other FIG. 9. Arrangement of Screen and Vent of Penstock Inlet. for several inches; they are arranged so that one or more sections may be hoisted independently of the others. By extending the abutments, the partition walls of the turbine chambers provide a very suitable support; otherwise they will have to be braced from the rear with structural steel; this of course depends on the size of the screen. Theory shows, that when the bars of the screen occupy one-quarter of the area of the water inlet to the turbine chamber, the drop in head due to resistance is from 2.5 to 3.5 per cent. 1 Wooden Sluice Gates. For controlling the water supply in the headrace, usually vertical moving sluice gates are employed. The small sizes up to 100 square feet 1 Gelpke, Turbinen und Turbinenanlagen, p. 142. HEADRACE. are made of wood, while the larger ones are made of structural steel. If conditions favor the use of large wooden sluice gates, they must be cut up into sections, other- wise they will be too bulky. The sections may be split up either vertically or hori- ;!:.nt I FIG. 10. Detail of Screen. zontally. In the former case, more guide? are necessary, and as a single sluice gate extends through the entire depth of the water, the water flows underneath the gate. In the latter case, fewer guides are necessary, and the waste may flow through the bottom, middle or upper sluice gate. Besides this, the individual stages of such 52 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. types do not have to withstand the total head. Furthermore, this type of sluice gate is advantageously used in connection with sandtraps and overflows, particularly for ice. They may also be used as an adjustable weir, but in such cases they are usually made of structural steel. FIG. ii. Wooden Sluice Gate. Hand and Motor Operated. For calculating stresses in wooden sluice gates, the table on the following page is submitted. It is compiled from a large number of tests; the maximum and minimum values are given. All of the test specimens had a sectional area of i. 575 by 1.575 square inches. The transverse specimens were 39.37 inches between supports, and the compressive- test specimens were 12.60 inches long. The modulus of rupture is calculated from , pi the formula R = - 2 bd 2 P = load in pounds at the middle. / = length in inches. b = breadth in inches. d = depth in inches. HEADRACE. 53 TABLE I. PROPERTIES OF TIMBER. Description. Weight per cubic foot in pounds. Weight per foot B.M. in pounds, average. Tensile strength per square inch, in pounds. Crushing strength per square inch, in pounds. Relative strength for cross breaking white pine = 100. Shearing strength with the grain, pounds per sq. in. Pressure in pounds per square inch to indent 0.05*. Ash 43-55 - 8 43-53-4 50-56.8 4.1 3-9 4-5 11,00017,207 11,500-18,000 10,300-11,400 4400-9363 5800-9363 5600-6000 130-180 100-144 55-63 130 96-123 96 888-95 132-227 I22-22O 130-177 I55-I 8 9 TOO 98-170 86-110 458-700 1800-1850 Beech Cedar Cherry Chestnut 33 34-3 6 - 7 2-75 2-9 10,500 13,400-13,489 8700 20,500-24,800 10,500-10,584 10,253-19,500 535-5 6 6831-10,331 57 911-5-11,700 '8150 4684-9509 6850 5000-5650 5400-9500 5050-7850 Elm Hemlock Locust 44 49 45-45-5 7 3 28-8-33 3-7 4-i 4-i 5-8 2-5 2.6 Maple 367-647 752-966 1700-1900 23-355 Oak, white Oak, live Pine, white 10,000-12,000 12,600-19,200 10,000-19,500 225-423 296-415 253-374 875-1160 1900 875-1025 Pine, yellow Spruce TABLE II. TESTS OF AMERICAN WOODS. Name of Wood. Transverse tests, modulus of rupture. Compression parallel to grain, pounds per square inch. Max. Min. Max. Min. Yellow poplar, white wood 6,55 6,720 8,610 12,200 8,310 5,95 10,220 8,250 14,870 7,010 9,760 7,900 5,95 13.85 6,310 5.64 5,610 3,78 9,220 9,900 7,59 8,220 10,080 ",75 6 ii,53 13.45 2i,73 16,800 15,800 i3,95 2 i5,7 20,710 18,360 18,37 18,420 12,870 18,840 9,53 15,100 ",53 10,980 21,060 11,650 14,680 17,920 16,770 4,15 3,810 6,O I O 8,33 5,83 4,520 6,980 4,960 7,65 5,8io 4,960 4,54 3,680 5,77 2,660 4,400 3,75 2,580 4,010 4,i5 4,5 4,880 6,810 5,79 6,480 7,5 11,940 9,120 8,830 8,790 8,040 10,280 9,070 8,970 8,55 6,650 7,840 5,810 7,040 5,600 4,680 10,600 5>3 7,420 9,800 10,700 \Vhite wood, basswood . . j__ . Red maple . .... Locust Wild cherry White ash Slippery elm. . . . . . . White elm . ..... Shellbark hickory White oak Red oak Black oak Chestnut ..... . . Beech White cedar ... Red cedar White pine Spruce pine Long-leaved pine, southern pine White spruce Hemlock . . . Red fir, yellow fir Tamarack 54 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. In connection with wooden sluice gates, particularly with less expensive or tem- porary installations, wooden guides are sometimes used. Tables I and II, giving the properties of timber and tests of woods, may be used in connection with same. Wooden sluice gates, as they are usually of small size, may be operated by hand. Where a series of small gates are opened simultaneously, they may be operated by a single motor. Such a device is seen in Fig. n. 1 It will be noticed that two wooden sluice gates, side by side, may be operated separately, by hand or by a motor; when desired, both gates may be operated simultaneously by a motor; friction clutches are located on both sides of the latter. Iron Sluice Gates. Large head gates are usually made of structural steel, and on account of their heavy weight, are counterweighted. Provision must be made, so FIG. 12. Application of Drum Gate, Chevres Plant, France. that they may be operated by hand, in addition to the regular motor operation of same. Low head plants usually employ several large-sized gates; they must be located side by side, and interconnected by an operating bridge. Where the turbine is placed in an open chamber, the water to the chamber is controlled by a separate gate. The type of gate varies greatly with the setting of the turbine. This may be done by a vertically operated sluice, drum or cylindrical gate. Sluice gates are the most common, and are frequently made of structural steel, provided with an auxiliary gate to facilitate operation; large gates usually are operated by a motor or hoist. 1 Gelpke, Turbinen und Turbinenanlagen. HEADRACE. 55 Where quick action is necessary, they are operated by air or hydraulic pressure. A gate of the latter type has been installed in the Beznau plant, Switzerland. They are 21.6 feet wide and 10.5 feet high, and provided with rollers, as they have to with- stand a pressure of 18.7 tons. The lifting or lowering of these gates is accomplished by oil pressure, supplied by the same pumps furnishing oil to the step-bearings. As the sluice gates are in the open air, the cylinders are jacketed to protect them from frost. To protect the steel piston rod from rust, it is fitted with a brass sleeve. Drum gates are made to swing either vertically or horizontally. A type of the latter has been installed in a plant at Rheinfelden, which has been in operation for many years. To the writer's knowledge, it has never been duplicated. A more FIG. 13. Detail of Drum Gate. favored one is the vertical swinging type. They have been installed at Chevres, France, and in many other plants. As the water presses against the drum (the gate being located in the turbine chamber; see Fig. 12), care must be taken to keep it tight, because the water is pressing it away from the seat. This may be accomplished by a rope or steel cable on top of the gate as seen in Fig. 13, and on the bottom by wooden 56 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. blocks. For operating the gate, a chain is fastened on the bottom of the drum and guided by sheaves. It may be raised by a separate hoist, or by the overhead crane in the generating room. The application of a cylindrical gate is seen in Fig. 14. The whole gate consists of a cylinder with vertical slots, which are covered and uncovered, by moving a disk up and down by means of a windlass. COLLECTING BASIN. Scheme. The junction between the headrace and the penstock must be made by a collecting basin, provided it is not a pressure tunnel. This collecting basin should be large enough to overcome slight fluctuations. To increase the capacity of such FIG. 14. Application of Cylinder Gate, Lyon Plant, France. a basin, the area of the headrace near the basin may be enlarged for a certain length. This may be well adopted if the headrace is a tunnel. Under ordinary conditions such a collecting basin is large enough when the velocity is only one foot per second. This velocity is sufficient to allow all heavy foreign material to settle, thus preventing it from entering the penstock. In most cases, the entrance to the penstock must HEADRACE. 57 be protected with a fine screen, to prevent foreign material from entering the same. The screens are preferably placed on the skew, so that the water will push the foreign material to the gate of the sandtrap. Sandtraps. A sandtrap is nothing more than a recess in the bottom of the collecting basin or headrace. The approach to the sandtrap must be gradual, so as to decrease the velocity of the water, and facilitate the settling of sand and gravel. At the deepest point, the trap is about two or three feet deep. The sandtrap must run on a skew, so that the foreign material will roll to the gate at the end. In many instances, there is no gate, only an opening, and the sand and gravel discharge all the time into the spillway. Spillways. The collecting basin must be provided with a spillway, so that in case the water to the penstock is cut off, and the water in the headrace is not cut off, no damage is done to the collecting basin or penstock run. The spillway is usually nothing more than a partition wall in the collecting basin with a lower elevation, so that the surface water can overflow. The spillway must be provided with a sluice gate, so that the collecting basin and headrace may be emptied through same. To provide for protection against floating material, particularly ice, a sluice gate to be lowered is installed, contrary to the one which is raised. These two gates may be combined into one, as will be seen under Sluice Gates. The discharge of the spillway is best done in cascades when the collecting basin is on a steep mountain slope, and discharged either into a river, or into the tailrace of the plant. The cascades break the fall of the water. Attention is here called to the spillway of the Ontario Power Company, Niagara Falls. The water is forced over an adjustable weir and discharged into a vertical, helical shaft, which opens into the gorge below the falls. The helical course was chosen to prevent ice formation. Gate Valves. All penstocks must be provided with cut-off valves at the collecting basin. They are usually of the gate valve type; in very rare instances butterfly valves have been used. The gate valves, also classed as sluice gates, are usually of simple form; the head presses the disk against the seat and keeps it tight. They are usually made of cast iron, and have babbit or bronze seats ( so as to keep them from rusting. When large valves are used, by-pass provision should be made, so as to properly fill the penstocks before the main valve is opened. The valves may be either hand or motor operated; they are sometimes of the remote control type, so that they can be operated from the power house. To protect the operating mechanism of the gate valves, and other devices connected to the inlet of the penstock, a housing should be provided. In medium head plants, screens and gates may be of such size that they are difficult to handle by hand, in which case a hand-operated traveling crane must be used. 58 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. BIBLIOGRAPHY. HYDRAULICS AND HYDRAULIC MOTORS. P. J. Weisbach. Translated by A. Jay Dubois. 1891. John Wiley & Sons. New York. A TREATISE ON HYDRAULICS. Henry T. Bovey. 1901. John Wiley & Sons. New York. TREATISE ON HYDRAULICS. Mansfield Merriman. 1903. John Wiley & Sons. New York. HYDRAULICS. L. M. Hoskins. 1907. Henry Holt & Co. New York. RIVER DISCHARGE. J. C. Hoyt and N. C. Grover. 1907. John Wiley & Sons. New York. ICE TROUBLES IN HYDRAULIC POWER WORKS AND METHODS OF OVERCOMING THEM. John Murphy, Consulting Engineer. May i, 1908. THE FLOW OF WATER OVER DAMS AND SPILLWAYS. Engineering Record, June 2, 1900. FLOW OF WATER OVER ROUNDED CREST. N. Werenskiold. Engineering News, Jan. 31, 1895. THE FLOW OF WATER OVER DAMS. Geo. M. Rafter. Proc. A.S.C. E., March, 1900. EXPERIMENTS ON THE MEASUREMENT OF WATER OVER WEIRS. A. E. Victor Dery. Proc. I. C. E., vol. 114, p. 333. 1893. EXPERIMENTS ON THE FLOW OF WATER THROUGH LARGE GATES AND OVER A WIDE CREST. Chas. E. Kaberstroh. Journal Association Engineering Societies, January, 1890. THE ESTIMATION OF DAMAGES TO POWER PLANTS FROM BACK WATER. Engineering Record, April 26, 1902. INSTRUCTIONS FOR INSTALLING WEIRS, MEASURING FLUMES AND WATER REGISTERS. Johnson. Engineering News, Aug. 29, 1901. NEW FORMULA FOR CALCULATING THE FLOW OF WATER IN PIPES AND CHANNELS. W. E. Foss. Journal Association Engineering Societies, vol. 13, p. 295. 1894. MEASUREMENT AND DIVISION OF WATER. L. G. Carpenter. Bulletin 27, Colorado Agricultural Experiment Station, Fort Collins, Colorado. 1894. DISCHARGE MEASUREMENTS OF STREAMS. F. H. Newell. Proceedings Engineering Club, Philadelphia, vol. 12, No. 2, p. 125. 1895. COEFFICIENTS IN HYDRAULIC FORMULAS. W. J. Keating. Journal Western Society Engineering, vol. i, p. 190. 1896. EXPERIMENTAL DATA FOR FLOW OVER BROAD CREST DAM. F. T. Johnson and E. L. Cooley. Journal Western Society Engineering, vol. i, p. 30. 1896. STEAM GAUGINGS. Clarence T. Johnson. Proceedings Purdue Society Civil Engineers. 1897. METHODS OF STEAM MEASUREMENT. Water Supply and Irrigation Paper, No. 56. 1901. ACCURACY OF STEAM MEASUREMENT. E. C. Murphy. Water Supply and Irrigation Paper, No. 64. 1901. THE LAWS OF RIVER FLOW. C. H. Tutton. Journal Association Engineering Societies, January, 1902. METHODS OF MEASURING THE FLOW OF STREAMS. John C. Hoyt. Engineering News, Jan. 14, 1904. A METHOD OF COMPUTING FLOOD DISCHARGE AND CROSS SECTION AREA OF STREAMS. E. C. Murphy. Engineering News, April 6, 1905. CHAPTER IV. PENSTOCKS. STEEL PENSTOCKS. Penstock Run. Penstocks must be laid in the shortest course to the power house, and in such a way as to avoid sharp turns. This may be done by running the penstocks through tunnels or trenches, or crossing valleys on trestles or piers. Conforming to modern practice, instead of one large penstock, several small ones are installed. They should be arranged side by side, and the bed selected for same should be as even as possible. If the natural ground cannot carry the penstocks, it must be reinforced to avoid any possibility of a semi-landslide, which in some cases have put plants out of commission. Where a number of such penstocks are installed, it is well to build as part of the penstock bed, a permanent cable road, by which means sections of the penstocks are hoisted to place, as well as to facilitate the building of the penstock run; later on, the cableway is used for inspection and repair purposes. Where multiple penstocks are laid, they must be interconnected, preferably at the power house, and properly equipped with valves, so that in case of emergency, the water of one penstock may feed other turbines. Size of Penstocks. The size of the penstocks depends upon the amount of water to be carried and head available. Other conditions remaining the same, the velocity of water in a penstock under a high head must be greater than under a low head. However, plants have been installed, where the velocity in the penstocks under low and high heads is practically the same. For instance, with the horizontal section of the i8-foot penstock of the Ontario Power Company, Niagara Falls, under a head of about 20 feet, the velocity is 15 feet per second; the velocity in the 3O-inch penstock of the Necaxa plant, under a head of 1300 feet, is 15 feet under normal full load, while under extremely high load, it is 18 feet per second. As stated, the smaller the penstocks, the higher the velocity; in some plants with penstocks i to 2 feet in diameter, the water has a velocity of 20 to 30 feet per second. In most high head plants running under a head of 1000 feet and higher, and penstocks 2 to 3 feet in diameter, the velocity chosen is between 10 to 16 feet per second. The velocity of the water in the penstock depends greatly on the velocity in the turbine gates, or, in other words, the speed for which the water wheel has been designed. Therefore, the lower sections or branches-must be designed to suit conditions. The main penstock must not have any sudden enlargements of diameter. The use of a drum at the end of a penstock, from which branches run to several turbines, must be avoided, because of the sudden change of speed in the water. 59 >?.- i At 60 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. TABLE I. AREA OF CIRCLES. 1 i 79 1.77 18 i 254.47 268.80 35 J 962. ii 989.80 52 2123.72 2164. 76 69 3739-29 3793-68 86 \ 5808.82 5876.56 2 1 3-14 4.90 19 i 283-53 298.65 36 1017.87 1046.34 53 1 2206. 19 2248. 01 70 3848.46 3903-63 87 5944-69 6013. 22 3 i 7.06 9. 62 20 314.16 330.06 37 1175.21 1104.46 54 2290. 23 2332-83 71 3958.20 4015.16 88 i 6082. 14 6I5I-45 4 12.56 15.90 21 346.36 363-05 38 1134.11 i i 64 . i 6 55 \ 2375-83 2419.23 72 407I-5 1 4128. 26 89 6221 15 6291. 25 6 i 19.63 23-75 22 380.13 397-61 39 1194.59 1225.42 56 2463.01 2507.19 73 4185.40 4242.93 90 6361 74 6432. 62 6 28.27 33.18 23 i 415.48 433-74 40 i 1256.64 1288.25 57 2551-76 2596.73 74 4300.85 4359.I7 91 6503.90 6575-56 7 38.48 44-17 24 452.39 471-44 41 1320. 25 1352-65 58 i 2642.09 2687.84 75 4417-87 4476.98 92 6647 65 6720 08 8 50.26 S6-74 25 i 490.88 510.71 42 1385-45 1418.63 59 \ 2733-98 2780.51 76 \ 4536.47 4596-36 93 6792.92 6866. i 6 9 i 63.61 70.88 26 530.93 551-55 43 1452.20 1486. 17 60 2827.44 2874.76 77 4656. 64 4117. 31 94 \ 6939-79 7013.82 10 i 78.54 86.59 27 593-95 44 1520.53 I555.29 61 2922.47 2970.58 78 4778.37 4839-83 95 i 7088.23 7163.04 11 95.03 103.87 28 6i5-75 637.94 45 i 1590.43 1625.97 62 3019.08 3067.97 79 4901. 68 4963.92 96 i 7238.25 73I3-84 12 113.10 122. 72 29 660.52 683.49 46 i 1661.91 1698.23 63 3166.93 80 5026.56 5089.59 97 7389-83 7466. 21 13 I43.I3 30 706. 86 730.62 47 1734-95 1772.06 64 i 3217.00 3267.46 81 5216.82 98 7542.98 7620. 15 14 153-94 165-13 31 i 754-76 779-31 48 1809.56 1847.46 65 3318.31 3369-56 82 5281.03 99 i 7697.71 7775.66 15 16 176.72 188.69 2OI.o6 32 33 804.25 855. 30 49 50 1924.43 1067. so 66 67 3421.20 3473-24 3525. 66 83 84 5410.62 5476.01 554I-78 100 7854.00 217. 87 881.41 i i y^ j o 2OO2. Q7 708.48 C6O7. (K 17 O J 226. 98 34 OO7. Q2 51 71 2O42. 83 68 OJf ~ 3631 . 69 85 *J / 7 J 1:674.. | | o a tj a 5 a o a a ' 1 ft 8 a o i V a tj c3 O OS O 6 1 8 19.58 24-5 78.32 12. 2 176.24 8.16 314.12 6. 12 489.68 4-90 705.64 4.0? 8 Si 20. 80 27.4 83-23 13-7 187.25 9. 15 333-75 6.86 520. 61 5-49 749.01 4-57 8* 9 22.03 30.5 88.11 15.2 198. 27 10. I 352-26 7.64 550-89 6. ii 793-72 5-o? 9 9i 23-25 33-8 93-0 16.9 209. 24 II. 2 371.90 8.46 581.25 6-77 837.08 5.61 IO 24.48 37-3 97.90 18.6 220. 30 12.4 391.40 9-33 612. 10 7.46 88i.8c 6. 21 IO ioj 25.70 40.9 102.80 20. 4 231-3! 13.6 411.05 IO. 2 642.43 8.19 925.20 6.82 105 ii 26. 92 44-7 107. 69 22.3 242.33 14.9 430-54 II. I 673-31 8-95 969.88 7-45 II nj 28.15 48.7 112.58 24-3 253-34 16. 2 450. 20 12. I 703.62 9-74 1013.3 8. ii Hi 12 29-37 52.8 117.48 26.4 264.36 17.6 470.68 13.2 734.52 10.5 I0 57-9 8.8c 12 13 31.82 61.5 127.27 30.7 286.39 20.5 509.82 15-3 795-73 12.3 1145.0 10. 2 13 14 34.27 71.0 137.06 35-5 308.42 23-7 548.96 17.7 856.94 14.2 1233.1 ii. 8 14 15 .36.72 81.0 146.85 40.5 330-45 2. 70 587.10 20.3 918.15 16. 2 1321.2 13-5 15 Dia 7-inch. 8- inch. i o- inch. 1 2-inch. i s-inch. 1 8-inch. Dia. Pipe. Pipe. ^ d >, d d >, d >, d d .0 _g _g 1 1 ft I *j a '| m a _o a u I ft 1 Q a O 6 6 u fc !> 8 (KQ. 44 3- 49 1253. 4 3.06 iot;8. 4 2. 45 2820. i 2.08 8 7j7 "^ IOI9.4 3-9 2 o 3-43 * y j~^ . *f 2080.8 2.74 2996.3 2-33 4688. I 1.82 6741.9 1.52 9 1079.4 4-36 1410. i 3.82 2203. 2 3-05 3172.7 2. 60 4957-7 2.04 7138.1 1.70 9 9* II39-4 4-83 1488.0 4-23 2325.6 3-38 3348.9 2.88 5232-1 2.25 7534-8 1.88 9i 10 H99-3 5-33 1566.8 4.66 2448.0 3-73 3525 2 3-17 2-5 793 1 - 2.07 10 loj 1259-3 5-84 1645.8 5- 22 2570.8 4.09 370I.4 3.48 5783-4 2-73 8328.8 2.27 io| ii I3I9.2 6-39 1723-5 5-59 2692.8 4-47 3877.7 3.80 6058. 2 2.98 8724.9 2.48 ii nj 1379.2 6 95 1801.5 6.08 2815.2 4.87 4053.8 4.14 6334.6 3-25 9121.7 2. 70 ni 12 1439-2 7-54 1880.2 6 60 2937.6 5.28 4230.2 4-49 6609 . 9 3-52 95I7.8 2-93 12 13 I559-I 8.79 2036 8 7.90 3182.4 6.15 4582.8 5-23 7 i 60 . 6 4. 10 10310. 3-42 13 14 1679.0 10. I 2193-5 8.87 3427.2 7.10 4935-4 6.03 7711.4 4.73 11104. 3-93 14 15 1799.0 ii. 6 2350.2 10. I 3672.0 8.10 5287.8 6.89 8262.1 5-40 11897. 4-5 15 Dia. Pipe 20-inch. 24-inch. 30- inch. 36-inch. 42-inch. 48-inch. Dia. Pipe. J d d d >, d >, d d i ti "o g 'o .9 o _g o .0 '3 _g 3 ti JD a a a a ^o a ft tj a 'o a o ,0 "3 01 O 6 d O a O 6 6 81 8323-^ i-37 11985. 1.14 18725. 915 26967. .760 36704. 653 47942. 571 8* 9 88l2. i-53 12690. 1.27 19827. I. 01 28554. .847 38863. .728 50762. 636 9 9* 9302. (. i. 69 13395- 1.40 20928. I. 12 30140. -938 41022. .806 53582. .694 9$ 10 9792. 1.87 14100. i-55 22030. 1.24 31726. 1.03 43181. .888 56403. .778 IO loj I028l. 2.05 14805. 1.70 23I3I- 1.36 33313- !-!3 45340. 975 59223- -851 ioj II 10771. 2. 24 15510- 1.86 1-49 34899. 1.16 47499- i. 06 62043. 93 ii II* 11258. 2-43 16215. 2.03 25338^ i. 62 36485. i-35 49658. 1.16 64863. I.OO "I 12 II750. 2. 64 16920. 2. 2O 26436. i. 76 38072. 1.46 51817. i. 26 67683. I. 10 12 13 12729. 3.08 18330. 2.56 28639. 2.05 41244. i. 70 1.46 73324- 1.28 13 14 13708. 3-55 19740. 2-95 30842. 2-37 44417- i-97 60453- i. 69 78964. 1.48 14 15 14688. 4-05 21150. 3-37 33045- 2. 70 47590. 2. 24 64771. i-93 84604. 1.69 15 64 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Strength. In calculating the thickness of the penstock shell, the following formula may be used : t = -^. T t = thickness of shell in inches. P = pressure in pounds per square inch. d = internal diameter in inches. T = tensile strength. s = factor of safety. Tensile strength of mild steel may be taken as 60,000 pounds per square inch. Wrought iron 50,000 pounds per square inch. For calculating the circumferential pressure in a penstock, the following formula may be used: . pr i = t i = intensity of strain. p = pressure head. r = radius of pipe. / = thickness of shell. For stresses in riveted pipes, the following formula may serve: i = elk. i = intensity of strain. e = modulus of elasticity. t = change of temperature. k = coefficient of expansion. In low head plants, where large-sized penstocks are used, the plates are made heavier, so that when the penstocks are emptied, they will retain their form. Another way to accomplish the same result is to reinforce the top section. Construction. Steel penstocks are made either riveted or welded. The former is made up of sheet steel or iron plates, which are rolled in the shop or field. The latter is resorted to, only in the case of very large pipes, in order to save freight rates. Lap joints are made single, double or triple riveted according to the head used. Under high heads the number of rivets in the lower sections of the penstocks increase; this means additional friction or loss of head. To overcome these difficulties, many plants use welded pipes in such places; among the prominent ones on the American Continent is the Necaxa plant in Mexico. Of the penstocks installed at this plant, there are 6 thirty-inch penstocks, each having a length of 2460 feet. The lowest sections are subjected to a static head of 1200 to 1300 feet. They are seamless welded pipes, and have a thickness varying from 0.4 to 0.95 of an inch, and were shipped from Germany in 29.5-foot lengths. A system of welding steel, particularly penstocks, pipes, etc., has been in use in Germany for a number of years. It is strange to note that American firms are very slow to adopt this system; because of this a large number of penstocks, amounting to many miles, are purchased abroad and shipped to the American Continent. PENSTOCKS. TABLE IV. RIVETED HYDRAULIC PIPES. 1 II. O 3 1 CO Ii *j d 3 CO c S -o |l ~s > s a 1 1 S p M _co g rt "* > 0} [3 SI | W I "<* ^ .s| c CO _o 3 c- c d ^5 11 I ^O " _> $ a ~3 cd ^3 fl .5 Q Pi W * 'I 5 w E* 3 18 O^ 810 2.25 18 12 . 109 295 25-25 4 18 O^ 607 3.00 18 II 125 337 29.00 4 16 .062 760 3-75 18 10 .14 378 32-5 5 18 5 485 3-75 18 8 .171 460 40.00 5 16 .062 605 4-5 20 16 .062 151 16.00 5 14 .078 757 5-75 20 14 .078 189 19-75 6 18 .05 405 4-25 20 12 . 109 265 27.50 6 16 .062 505 5-25 20 II .125 34 3!-5o 6 14 .078 630 6.50 2O 10 .14 340 35-oo 7 18 .05 346 4-75 2O 8 .171 415 45-5 7 16 .062 433 6.00 22 16 .062 138 17-75 7 U .078 540 7-5 22 14 .078 172 22.00 8 16 .062 378 7.00 22 12 . 109 240 30-5 8 14 .078 472 8-75 22 II 125 276 "34-50 8 12 . 109 660 12.00 22 10 .14 309 39.00 9 16 .062 336 7-50 22 8 .171 376 50.00 9 14 .078 420 9-25 24 14 .078 158 23-75 9 12 . 109 587 12-75 24 12 . 109 220 32.00 10 16 .062 37 8-25 24 II 125 253 37-50 IO 14 .078 378 10.25 24 10 .14 283 42.00 IO 12 . 109 530 14.25 24 8 .171 346 50.00 10 II 125 607 16.25 24 6 . 20 405 59-oo 10 10 .14 680 18.25 26 14 .078 145 25-5 ii 16 .062 275 9.OO 26 12 . 109 203 35-50 ii 14 .078 344 11.00 26 II .125 233 39-5 ii 12 . 109 480 15-25 26 IO .14 261 44-25 ii II 125 553 17-50 26 8 .171 319 54-oo ii 10 .14 617 19-50 26 6 . 20 373 64.00 12 16 .062 252 IO.OO 28 14 .078 135 27.25 12 14 .078 316 12. 25 28 12 . 109 1 88 38.00 12 12 . 109 442 17.00 28 II 125 216 42.25 12 II 125 506 19.50 28 10 .14 242 47-50 12 10 .14 567 21. 75 28 8 .171 295 58.00 13 16 .062 233 10. 50 28 6 . 20 346 69.00 13 14 .078 291 13.00 30 12 . IO9 176 39-50 13 12 . 109 407 18. oo 3 II 125 202 45-0 13 II .125 467 20.50 3 10 14 226 50-5 13 10 .14 522 23.00 30 8 .171 276 61. 75 14 16 .062 216 11.25 30 6 . 2O 323 73-0 14 14 .078 271 14.00 30 1 25 404 90.00 14 12 . 109 378 19.50 36 ii 125 1 68 54-oo 14 II 125 433 22.25 36 10 .14 189 60.50 14 IO .14 485 25.00 36 A .l8 7 252 81.00 15 16 .062 202 ii. 75 36 1 25 337 109.00 15 14 .078 252 14-75 36 A .312 420 135-0 15 12 . 109 352 20. 50 40 10 .14 170 67.50 15 II .125 405 23. 25 40 A .187 226 90.00 15 10 -14 453 26.OO 40 i 25 33 I2O.OO 16 16 .062 190 13.00 40 A .312 378 I5O.OO 16 14 .078 237 16.00 40 | 375 455 iSo.OO 16 12 . 109 332 22.25 42 IO 162 7I.OO 16 II 125 379 24.50 42 ^ :ls 7 216 94-5 16 10 -14 425 28.50 42 i 25 289 I26.0O 18 16 .062 1 68 14-75 42 A .312 360 158.00 -18 14 .078 2IO 18.50 42 1 375 435 I9O.OO 66 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. This process is known as "autogenous welding." It embraces all methods of welding in which the parts are joined in a homogeneous manner, the joints being made by the intermingling contacts of the fibers of the material to be welded, so that the finished piece is one of uniform quality and properties throughout. The process is carried on by several methods, viz., the oxy-acetylene, the purely electrical, and the oxy-hydrogen method. The latter is the most extensively used, and has been discussed by the author in the technical press. 1 By using welded penstocks, it will be readily observed that additional head is obtained because of the absence of rivets. No fixed rules can be laid down for penstock construction, as it all depends upon the nature of the conditions under which the steel penstock is to be constructed. To give an example of a prominent steel penstock construction on the American Continent, that of the Kern River Power Plant 2 of California is cited. It has many notable features, and, contrary to the customary practice on mountain slopes, it is run through tunnels. The steel penstock serves as a lining to the rock-cut tunnel. The penstock consists of a tunnel approximately 1700 feet long driven through the mountains on an incline and lined with steel, varying in thickness from ^ to i inches. This tunnel begins at the bottom of the forebay, and is carried down at an angle of approximately 45 degrees, and, turning into the horizontal section, emerges at the lower end on a level with the floor of the power station. There are three vertical curves in the tunnel. The upper one forms an angle of 7 degrees, 260 feet from the forebay floor, and turns the pipe from a grade of 130.32 per cent to a grade of 101.35 per cent. The second curve 32.5 feet lower down has an angle of 5 degrees, and turns the pipe into a grade of 4.93 per cent on which it is carried 994.24 feet FIG. i. Type of Penstock Flange used in recent Swiss Practice. to the last vertical curve. The latter has an angle of 40 degrees, and from its lower end the pipe is carried along horizontally to the power house, the total length of the main being 1697 feet. The penstock is finished to give it an inside diameter of 7 feet 6 inches. At the top, a taper, 20 feet long and 10 feet in diameter at the forebay entrance, terminates in the 1 Oxy-hydrogen Welding. Electrical World, May 9, 1908. 2 Kern River Power Plant, by C. W. Whitney. The Engineering Record, Aug. 10, 1907. PENSTOCKS. 67 regular 7^ feet diameter of the completed tunnel tube. This diameter of 7^ feet is maintained throughout the inclined tunnel, and on the horizontal beyond vertical curve No. 3, for a distance of 167.39 feet. At this point, 1454.44 feet from the forebay, the penstock emerges from the solid rock and is carried to the portal, a distance of 243 feet, through a detrital deposit, lying between the mountain and the power-house site. Where the tunnel emerges from the solid rock, a 2o-foot taper was installed, reducing the diameter of the main from 7? to 5^ feet, at which diameter the pipe is carried to the branch piping at the power house. so -*j>o~ soJ FIG. 2. Type of Penstock Flange used in recent Swiss Practice. The inclined part of the pressure main, and the portion of the horizontal section that is carried through solid rock, were finished by installing a steel lining built up of plates three-sixteenths inch thick for the incline, and three-eighths inch thick for the horizontal section, riveted together to form a cylindrical pipe, 7^ feet in internal diameter. The tunnel itself was driven in approximately circular form and 9 feet in diameter. The steel pipe was centered in the tunnel, being installed in lo-foot sections, and the space between the outside of the steel lining and the bed rock was thoroughly filled with a mixture of i : 3 : 3 concrete. The work of installing this lining was begun at the lower end in the horizontal section, where the pipe is tapered down to a diameter of 5^ feet. At this point, the 2o-foot taper, already mentioned, was placed. It consisted of if -inch steel plates riveted together with butt straps. The taper was placed back in the solid rock, and around it was constructed a heavy bulkhead of concrete, which was anchored into the bed rock by means of steel rods driven into the sides. From this point, the installation of the light steel lining with concrete back-fill, as already stated, progressed from the bottom to the top of the tunnel, terminating at the reinforced concrete taper that connects with the floor of the forebay. The rock formation, through which the penstock tunnel was driven, is not of the best kind, being very much fractured and broken. It was necessary to timber the greater part of the shaft or incline when it was excavated, and these timbers had to be removed before the steel lining was installed. The timbers were removed ahead of the steelwork, the bed cleaned off, and the concrete tamped into place without difficulty. At a point about 120 feet below the top, the men in charge removed some 68 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. timbers without bracing the bents above. This precipitated a cave-in of the shaft, and several men lost their lives, one man being imprisoned for two weeks, after which time he was rescued in good condition. In retimbering the caved portion, octagon steel sets of y-inch 15 -pound I-beams were used, these sets being left in place when the concrete was filled in behind the steel lining. The lower end of the pressure main, below the taper reducing the diameter to 5^ feet, was made of if-inch steel plates sufficiently heavy to withstand the static pressure without any external support. No concrete was placed around this pipe, and the tunnel was merely left in its original condition with the timber set to support the ground overhead. FIG. 3. Flange used at the Brusio Plant (i30o-foot. Head). At a point 215 feet above the power house, a manhole was placed in the inclined tunnel for convenience on inspecting, and for use in case any repair work was neces- sary. The regular T \-inch steel lining was replaced at this point by a section of ij-inch pipe 30 feet long. The steel pipe was shipped to Camp No. i at the power house from San Fran- cisco, in 5-foot lengths, 5 sections being nested together for shipment. The outside section was riveted complete on its two longitudinal seams, but the four inner sections were riveted on one seam only, so as to allow for the nesting. At the camp, the pipe was riveted into lo-foot lengths, and hoisted by means of an aerial tram to the forebay site at the upper end of the pressure tunnel. There the sections were secured to a dolly car, and lowered by means of a hoist to the point where they were to be riveted together. This car consisted of a truck at each end of the pipe sections, the latter being hung from two timbers that passed through the pipe, and rested on the axles of the truck. All the piping in the pressure tunnels, which is constructed of steel plates of one-half-inch thickness and under, is made up with standard lap joints, double riveted PENSTOCKS. 69 on the longitudinal seams and single riveted on round seams. All pipe on the work over one-half inch in thickness, is made up of butt strapped joints throughout, with triple riveting on each side of the longitudinal seams, and double riveting on each side of the round seams. After the steel lining was completed, an inspection revealed the fact, that there were several places along the bottom of the pipe where voids had been formed in the concrete backing. FIG. 4. Type of Penstock Flanges for Necaxa Plant, Mexico. These voids, which were revealed by tapping, were caused mainly by the difficulty experienced in tamping the concrete thoroughly around the steel lining. These steel sections were 10 feet in length, and in a few places where irregular rock excavation occurred at the bottom of a section, with but a Q-inch space at the top for handling the tamping bars, some voids were naturally formed because of the insufficient tamping. Whenever these voids occurred, the pipe was tapped and liquid cement forced in until the hole was filled. The apparatus designed on the spot to accomplish this work was an ingenious one. A section of 3 -inch steel tube 20 inches long was fitted at the bottom with a cap that would fit the hole drilled in the steel lining. Liquid HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. cement was poured into the void by means of this pipe, which had a capacity of about an ordinary pail. When no more cement would run in, there was fitted in the pipe, a screw with a plunger at the lower end and a crank at the outer end. By means of this device, the cement was forced into the void under pres- sure until it would hold no more. The pump was then removed and the hole in the lining was stopped up by an ordinary flush pipe plug. There were 116 of these voids tapped and filled through the lining; only three of them were of any size. A number of the voids required only a pint of the liquid cement, the quantity used varying up to the largest, for which ten buckets of the slush were necessary. The slush used was a liquid mixture of Portland cement and sand. The work was carried on from a dolly car fitted with beveled wheels, lowered down from the top by a steel cable. About 15 days were necessary to complete this special work. After all the voids were filled, the entire pipe was painted with asphaltum, the same dolly car being used for the purpose. FIG. 5. Type of Penstock Flange for Sillwerke, Innsbruck. FIG. 6. Method of Anchoring Penstock, Jajce Power Plant, Bosnia. In the rear of the power house, a number of branch penstocks lead to the turbines. They are built tapered, and vary in thickness from if to f of an inch. At the end of the last section of the penstock is a 28-inch gate valve for emptying the entire system when necessary. All branches are provided both outside and inside of the building with a gate valve. PENSTOCKS. Flanges. In connection with high-pressure penstocks, particular attention must be paid to the flanges uniting the different sections. In Fig. 4 are shown three different designs of flanges, as used with the pre- viously mentioned Necaxa penstocks, for different pressures. It will be noticed that these flanges act as a lever upon the flared ends of the penstock. Another type of flange, by the manufacturer of the above, will be found under subheading "Slip-joints.'' A simpler, yet efficient, flange joint is seen in Fig. 5. It has been used in con- nection with the Sillwerke plant, near Innsbruck, Tyrol. Its efficiency lies chiefly in the wedge- shaped packing. Anchors. In order to prevent penstocks from sliding on mountain slopes, they must be anchored. The simplest way is to embed them in concrete blocks. Another way is to rivet on iron saddles, and bolt the same to concrete piers; in addition, anchor rods are sometimes used. Such a method of anchoring is seen in UU FIG. 7. Hinged Penstock Support, Kaiserwerke, Tyrol. t. FIGS. 8 and 9. Expansion Slip-Joints for Penstocks. Fig. 6. By studying this illustration it will be readily seen that embedding the penstock in a single massive block would have accomplished the same purpose. 72, HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Special precautions must be taken, where the penstocks run on turns, either horizontally or vertically, by establishing fixed anchors. Between two anchors, the penstocks rest on supports, and an expansion joint must be provided. Saddles. Penstocks must be properly supported between anchors on saddles, to allow the penstock, in case of expansion, to slide. As the movement of the penstock is slight, it is not essential to provide the saddles with rollers, as has been done in some cases. The saddles are made of cast iron or semisteel, and so designed, that the penstock rests on two carrying surfaces, in order to give a better support and minimize friction. The saddles must be rigidly anchored to concrete or other masonry piers. The spacing of the saddles is closer at the bottom of the mountain slope than at the top, or they may be constructed heavier (similar to the penstocks) as the pressure exerted upon them is greater. When the penstock leaves the ground for short distances, a method for supporting may be adopted as is given in Fig. 6. It has been used in connection with a power plant in North Tyrol. It consists of a steel frame- work hinged to a concrete pier; the upper end forms a saddle, which is clamped to the penstock. This arrangement allows a free movement of the penstock due to expansion. FIG. 10. Wedge-Shaped Expansion Joint, Jajce Plant, Bosnia. FIG. ii. Ferrum Slip- Joint. Flange for High Pressure Penstocks. Expansion Joints. The expansion joints must be located at the upper end of each section between anchors, when descending a mountain slope. This is done to relieve the expansion joint from the weight of the penstock, so that when expansion takes place, the section of the penstock slides up hill. These expansion joints are, in most cases, of the slip-joint type, as seen in Figs. 8 and 9, both of which have been used in connection with recent Swiss power plants. Another type, seen in Fig. 10, has been used in connection with the Jajce plant in Bosnia. It consists of a wedge-shaped drum to take up the expansion of the penstock, which is laid on an angle. The diameter of drum depends upon the amount of expansion to be taken up, and the sides are preferably of copper. PENSTOCKS. A.E.&M.E. 73 UNIV. OF CAL. 2 stuffing-hoy A way to avoid special expansion joints and sliding saddles, is to us flanges, which have been used especially in high head installations. A detail of this flange joint is seen in Fig. n. The manufacturers (Aktiengesellschaft Ferrum Kattowitz, Germany) claim that it is not essential to lay the penstock sections exactly on centers. It does away with slip joints, as each section takes up its own expansion. The joints can be packed without taking the sections apart. Because the flanges are removable, the sections are made easier for shipment, and in some cases they may be telescoped. Fig. u shows six penstock lines with this type of FIG. 12. Penstocks with Slip- Joint Flanges, Loch Leven Plant, Scotland. flange as installed for the Loch Leven Water and Electric Company, Scotland. These penstocks have a diameter of 40 inches, and a shell' thickness varying from three-eighths to seven-eighths inch. The lower sections are designed for a pressure of 425 pounds per square inch. Each of the lines is 6230 feet long. The sections are made in 19. 7-foot lengths. The concrete anchor blocks are placed about 175 feet apart. The constant applied for calculating the expansion for wrought iron and steel is 0.0000067 f an mc h P er inch, or approximately .00008 of an inch per foot for one degree Fahr. The coefficient for cast iron is .0000059 per unit of length per degree Fahr. As cast iron or cast steel fittings amount to but little, the coefficient of wrought iron or steel is best employed. 74 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. PENSTOCKS. 75 Safety Devices. For the protection of penstocks, they must be provided at the upper end with relief devices. If there is no air vent at the top, the penstocks are apt to collapse when the head gates are closed, because of the formation of a partial vacuum. This vent pipe is nothing more than a pipe connection on top and directly behind the head gate, and must extend above high-water level. Care must be taken that the vent pipes do not freeze. In many plants they are simply ducts in the wall of the collecting basin or dam. An example of such a device is given in Fig. 12, in which case there is a separate collecting basin and gate house. The vent pipes extend through the roof of the former and above the high-water level of the collecting FIG. 14. Automatic Flap. Inlet to Penstock, Brusio Plant, Switzerland. basin. Another safety device in the head works of this plant is, the mouth of the penstock in the collecting basin is provided with a flap valve. In case a penstock should fail, this valve will close automatically, and can be operated from the collecting basin or from the power house by remote control. The flap is counterbalanced by a weight. The penstocks can be filled through a by-pass after the flap is closed (see Fig. 14). In addition to this, the collecting basin is provided with a mechanical- 76 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. electric water-level indicator, which indicates in the power house the water level in the collecting basin. Another device for admitting air into a penstock to prevent the formation of a vacuum, is shown in Fig. 15. It has been installed in a recent Swiss power plant, which operates under a head of 767.5 feet. When the penstock is under pressure, the valve is closed; when air is admitted, the disk is held in place by a spring. In plants where the penstocks have a long horizontal run, and suddenly go down a steep slope, a vacuum will be created in the latter if the turbine valve is suddenly opened wide. This is due to the fact that the velocity will be greater in the vertical section than in the long horizontal line. To overcome this, the above-mentioned valve is placed near the junction of the vertical and horizontal sections. Fig. 15 is another safety device, which will act automatically on low-water level, or in case the generator should drop its load. If for any reason the penstock should fail, the valve will shut off the water automatically. The lower end of the penstocks must be provided with relief valves or blow-out plates, so that in case of a sudden shut down of the turbines, the water in the penstock will be released through the safety devices, and discharged into the tailrace. The safety devices must be so located, that when they operate, no damage will result. Manholes must be provided on the lower sections of penstocks. Standpipes. Another safety device, to relieve penstocks of excess pressure, is the standpipe. These standpipes are usually connected to the end of the penstock, and must extend vertically several feet above the head, so that in case of a sudden increase of pressure, the water has a chance to rise before overflowing. Usually these stand- pipes are wasteful, and to overcome this defect, a standpipe system similar to that in the Urfttalsperre, Germany, may be used. This standpipe, however, is not located at the end of the penstock, but lies at the junction of the pressure tunnel and the penstock. Another method is to provide the. lower end of the penstock with an air cushion, which is nothing more than a closed chamber, and acts in the same way as an air chamber on a reciprocating pump. This air chamber must be as air tight as possible, otherwise the air will escape and render the device useless. In connection with Jajce Power Plant, Bosnia, 1899, the author experienced similar trouble, and to remedy the difficulty, an air pump had to be installed to maintain the air cushion. All standpipes must be protected from freezing. Protection. In exceptionally cold regions, penstocks must be protected from frost. They may be embedded in the ground below the frost line, or protected by a wooden box covering. To cite an example what frost might do, at Grand Mere, Quebec, 1 a penstock of 14 feet diameter was left unprotected during the first winter. It was found to have its interior surface covered with solid crystal ice of from 12 to 18 inches in thickness. If the penstock is buried in the ground, the trench must be left open at least several months after operation has commenced, to ascertain whether additional 1 Thurso, Modern Turbine Practice, p. 176. PENSTOCKS. 77 O. S a I o -o S HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. calking is necessary. Before the trench is filled, the penstock must be properly painted. Buried penstocks do not require any device for taking up expansion. All penstocks must be coated with hot asphalt both inside and outside. WOODEN PENSTOCKS. Adaptability. Wooden penstocks are largely used in the western states, particu- larly on the Pacific coast, where it is necessary to conduct water over long runs. They are practically installed for heads up to 300 feet. Wherever they are installed under higher heads, they are always placed in the top section of the conduit. As they have many advantages over metal or reinforced concrete pen- stocks, they are much in favor. Some of the advantages are, that they are smooth and have a greater carrying capacity, ranging from 10 to 20 per cent more than any other. Another import- ant factor is, that they are very much cheaper and do not deteriorate as rapidly as those of metal. Further, they are unaffected by frost. In the construction of some pen- stock lines, it would be difficult to transport heavy steel penstocks over the country where there are no roads. In such cases, the wooden penstock is used. They range in diameter from 10 inches upward, and vary in thickness according to the diameter. The staves are milled from clear, well-seasoned or kiln-dried yellow pine, redwood or fir. The ends of the staves are connected by a tongue, which prevents butt joint leakage. The staves are held in place by steel rods and a cast-iron shoe. FIG. 16. Automatic Low Water Device for Protecting Penstocks. Spacing of Bands. The spacing of the iron bands is determined by a formula given by James D. Schuyler. 1 N = 1200 DP 28 N = number of bands per 100 feet. D = diameter of pipe in inches. P = pressure in pounds per square foot. S = safe working strain in pounds per square inch for bands when threaded for use, determined by regular tests at the mills where they are made. 1 Trans, of Am. Soc. of C. E., vol. xxxi. PENSTOCKS. FIG. 17. 84-inch Wooden Stave Penstock, 3700 feet long, joined to two 60- inch Riveted Steel Penstocks, each 800 feet long. Trenton Falls Water Power Plant, Utica, New York. The following values of 5 give a factor of safety of about five in each case, or about one-fourth of the elastic limit : TABLE I. SAFE WORKING STRAIN OF PENSTOCK BANDS. f-inch bands plain S = 1000 pounds. f-inch bands upset .S = 1 200 pounds. ^-inch bands . . plain 5 2000 pounds. -inch bands f-inch bands f-inch bands. .... upset plain upset 5= 2500 pounds. 5= 3000 pounds. 5= 3500 pounds. The formula given below was used by C. P. Allen, in the construction of a 6^-mile penstock along the Little Conemaugh River near Johnstown, Penn. This penstock has a diameter of 36 to 44 inches. The pressure in this penstock is slight; the regular slope is i to 2 feet in looo. 1 1 Some Applications of Wooden Stave Pipe, by John Birkinbine, in a paper before the Engineers' Club, Philadelphia, Penn. 8o HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Number of bands per 100 feet = 600 DPHF AB D = diameter of pipe in inches. P = pressure due to i foot. (0.44 pound.) H = head in feet. F = factor of safety. A = area of bands in square inches. B = breaking strain of bands per square inch. Thus, for a 44-inch penstock, one-half inch bands, and a 5o-foot head, the number of bands per 100 feet is 600 X 44 X 0.44 X 50 X 4 - = 197. 0.19635 X 60000 Friction. As already indicated, the friction in wooden penstocks is very low, as the staves are planed. The resistance of steel penstocks increases each year, while that of wooden penstocks decreases. In Kutter's formula, the factor of resistance is generally taken as o.oio for wooden stave pipe, while for steel as 0.013. Some- times the factor for wooden stave penstocks goes as low as 0.007. FIG. 18. Three 7-foot Wooden Stave Penstocks, each 4000 feet long and connected to Riveted Steel Penstocks, 1000 feet long. Great Northern Power Company, Duluth, Minnesota. Durability. The durability of wooden stave penstocks if kept continually wet, is yet undetermined. The following data are of interest: in 1898 some of the original PENSTOCKS. 81 wooden pipe laid in the London waterworks in 1802, were taken out sound and free from rot. Some of these wooden mains were in actual use as late as 1865, after having been in the ground for 63 years. Some of the wooden pipes first laid in Philadelphia, after being in use 21 years, were removed and relaid in Burlington where they were in use for 28 years. In a series of tests carried on at the Puget Sound Navy Yard in 1901, comparing Douglas fir and yellow pine for pipe staves, Frank W. Hibbs, naval constructor of the United States Navy, arrived at the following conclusions: In strength, Douglas fir is generally equal to yellow pine, and superior to it in some essential particulars. Douglas fir is decidedly more elastic than yellow pine. Douglas fir is far superior to yellow pine as regards toughness. Yellow pine is superior to Douglas fir in wearing qualities, especially when moisture is present. Yellow pine is superior to Douglas fir in lasting qualities, on account of the greater amount of pitch it contains. Douglas fir is 14 per cent lighter than yellow pine. Following are the average general characteristics of strength of Douglas fir: For well-seasoned, fine-grained, hard, clear stock: TABLE II. STRENGTH OF DOUGLAS FIR. Characteristics. Pounds per square inch. Tensile strength 13,000 Tensile strength across grain 3CO Tensile strength for bending 10,000 Elastic limit for bending 6,000 Modulus of elasticity for bending 1,500,000 Strength for compression across the grain without destructive deformation 1,200 Modulus of elasticity for compression across the grain. Crushing strength for compression, "end on" to grain Modulus of elasticity for "end on" compression Modulus of elasticity for torsion 4,000 9,000 70,000 27,000 Shearing strength with the grain 15,000 Crushing strength for columns whose proportions are such as to resist bending 6,000 ^Veight per cubic foot, pounds . . 7C Cost. The cost of wooden stave penstocks relative to steel penstocks, depends on pressure, size, location, and character of country, through which they are laid. Mr. L. A. Adams states that the details of cost of an i8-inch penstock at Astoria, Ore., 7^ miles long, are as follows: " Steel in bands, $0.048 per pound; lumber, feet board measure in staves measured before milling, $35.40 per thousand. The cost to the city, including all appurte- nances, was $0.903 per foot; and $0.76 excluding such appurtenances. The whole amount of the contract was $36,100, and the total extra work cost, $29.35. 82 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The distribution of the cost was as follows: "Building and spacing bands, 55 per cent; back-cinching, 26 per cent; repainting ironwork, 3 per cent; back-filling to a depth of 6 inches over the pipe, 8.75 per cent; placing specials, 3.5 per cent; placing air valve, 0.75 per cent; unclassified labor, 3 per cent." f FIG. 19. Cross Section and Partial Plan of a Wooden Stave Penstock. Adams also gives the cost for the riveted steel pipe in the same line as follows: Size. Gauge of steel. Price. 14-inch 1 6-inch 1 6-inch No. 12 No. 12 No. 10 $1. IO 1.18 1.38 The manufacturing cost of the riveted steel pipe was about 0.45 of a cent per pound for labor only, including the cost of dipping. FIG. 20. Cast-Iron Saddle for Connection to a Smaller Pipe. Comparative costs on the construction of steel, cast-iron and wooden penstocks are given in Table III, as compiled by Mr. A. L. Adams for Chicago. These figures are supposed to include only the principal items, with no profit to the contractor, or for incidentals, and are therefore for comparison only. PENSTOCKS. TABLE III. COMPARATIVE COST OF PIPE AT CHICAGO, INCLUDING LAYING, BUT OMITTING HAUL. Wooden Stave Pipe. Diameter. 2 5 -foot head. So-foot head. loo-foot head. 2oo-foot head. Inches . 12 18 24 3 36 42 48 54 60 66 72 $0.42 o. 69 0.79 o. 96 1.19 1.40 i-55 2.23 2.85 3.21 3-65 $0.49 0.8o 0.91 I. 12 1.40 1.68 1.85 21 62 3-35 3.81 4-38 $0.63 1.02 I.I4 1.44 1.82 2.23 2. 46 3-43 4-37 5.00 5-83 $0. i. i. 2. 2. 3- 3- 5- 6. 7- 8. 85 46 61 06 65 33 67 02 40 38 73 Riveted Steel Pipe. Diameter. No. 14. No. 12. No. 10. No. 8. No. 6. J-inch. T^-inch. f-inch. Inches. 12 18 24 30 36 42 48 54 60 66 72 $0.32 $0.38 o-57 $0.44 0.65 0.85 $0.78 1.04 1.27 i-55 1.61 $0.98 1.28 i-59 1-93 2.18 2.48 2.80 $i-55 i-93 2.30 2.66 3-3 3-4i 3-79 4-35 4-52 $1.99 2.46 2.92 3-37 3-83 4.29 4-75 5-21 5.66 $3.04 3-58 4.12 4.66 5.21 5-74 6. 29 6.83 Cast-Iron Pipe. Diameter. 2 5 -foot head. 5o-foot head. ioo-foot head. zoo-foot head. Inches. 12 18 24 3 36 42 48 54 60 66 72 $0.73 1.29 1.91 2. 67 3-47 4.42 5-5 6.65 8.04 9-Si 11.32 $0. I. 2. 2. 3- 4- 5- 7- 8. 10. 12. 77 35 00 80 67 69 84 10 63 16 oo $0.84 1.46 2.18 3-7 4.06 5-22 6-53 8.00 9.80 11-55 13.26 tOf r 51 . I. 2 3- 4- 6. 7- 9- 12. 14. 16. 00 70 55 61 85 28 92 78 13 5 oo 84 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. \ Construction. A very good example of wooden stave pipe construction is that built in connection with the Bishop Creek Power Plant. 1 The penstock is about 12,000 feet long, consisting of 6700 feet of 42-inch wood-stave pipe, 2150 feet of 30-inch wood-stave pipe, and 3150 feet of 24-inch steel penstock; all diameters being inside measurements. The 42-inch penstock lies on a nearly level grade, the static head at the lower end being about 30 feet. At this point, are placed two 3O-inch gate valves, one opening into the 30-inch penstock, and the other provided for a future line. The 30-inch penstock descends the hill to a point that gives a static head of 265 feet. Here it joins the 24-inch steel penstock, which descends a steep hill to the power house; the total static head is 1068 feet. For about 500 feet, this steel penstock is laid on an angle of 38 degrees from the horizontal. The 42-inch penstock is made up of twenty-five staves, and the 3O-inch penstock of nineteen staves, milled from a 2 by 6-inch piece. The lumber is red fir from Southern Oregon, and the staves were milled there to the proper circle. The bands are mild steel, one-half inch in diameter, with ends upset to five-eighths inch diameter. They were shipped straight, and bent to form at the work. The lugs are of cast iron, of a form that allows the ends of the bands to pass each other, and be tightened with nuts on each end. When the bands are tightened there is a slight bending of the rods, but this is not believed to be injurious. The ends of the staves were slotted, and a three-sixteenths by one-half inch compressed paper dowel inserted. When wet, this dowel swells, and proves very effective in closing leaks. Bands on the 42-inch penstock were spaced on 6-inch centers, although the pressure did not demand this close spacing. On the 3O-inch pipe, which is subject to a considerable pressure, the designer used the pressure due to swelling of wood given by A. L. Adams, 100 pounds per square inch; in addition to this, an allowance was made for initial tension in the rods due to the stress necessary to bring the staves into form. Red fir is a very stiff wood, and by observation it was determined necessary to use 2100 pounds to bring the staves into position; this was calculated as being distributed along the lineal foot of pipe, and distributed among whatever number of bands occurred in that length under different heads. Bands were spaced under these assumptions with a safety factor of four. The question of initial tension on the rods is believed to be a vital one, as, in the case of the stiff lumber used, it required much cinching to make the staves come together. As the wood is hard, the bands crushed into it but little under pressure, and hence there is little relief to the stress. By test it was determined, that, with the wrenches used, an initial tension of 8000 pounds per square inch could easily be obtained. The wood pipe is laid directly on the ground, a few sills of culled material being placed at intervals. At points, it was covered with earth as a protection from possible rocks rolling down the steep hillsides. The penstock was laid in easy curves, but in one case a curve of 100 feet radius was made through nearly 90 degrees. For nearly two months the daily amount laid averaged no feet. 1 Bishop Creek, Cal., Hydro-electric Power Plant, by J. D. Galloway. Electrical World, June 30, 1906. PENSTOCKS. 85 The wood pipe is provided with two 6-inch and one 3o-inch standpipes, the former being of casing and the latter of 3O-inch wood pipe. In addition, 6-inch air valves are supplied in such number, that there is a 6-inch opening every noo feet. On the steel pipe, three 6-inch air valves were placed at the upper end. The material for the wood pipe was mostly hauled to the site on wagons, a minor portion being hauled up from the power house on the tramway used in laying the steel pipe. Another very interesting, and in some respects difficult, wooden stave penstock construction was laid in the American Fork Canyon, Utah. 1 Where conditions justified, the continuous wood-stave penstock was laid on a grade as near the hydraulic grade as was thought advisable, when considering the necessity of keeping the pipe filled. It was found necessary, however, to construct three inverted siphons, the first and longest one being at the upper end of the line, where the pipe follows down the bottom of the canyon for some distance; the other two are about midway between the point of diversion and the power house. The maximum head on this pipe, at one of the siphons, is 175 feet. The remaining portion of this line was laid on a table, cut on a grade contour along the mountain side. As the canyon is rough and broken, this alignment necessitated the introduction of many curves, some of which were quite sharp, and the driving of many tunnels through solid quartzite ledges, the length of the tunnels varying from 25 to 1 60 feet. The tunnels were rectangular in shape, the dimensions being 4 by 5 feet, and some of them were located on curves. The pipe is 36 inches in diameter, and a section at right angles to its axis shows 22 staves. These staves were sawed from well-seasoned Oregon fir, with their faces dressed to true segments of circles, and the edges to true radial lines. They are if inches thick, but vary in length from 8 to 15 feet. In making the joints, a special malleable casting patented by Frank C. Kelsey was used. A section through this casting is similar in shape to that of an I-beam, except that there is a shorter flange projecting from the middle of the web on either side. The distance between the outer edges of the main flanges corresponds with the thickness of the staves, but one of them has a batter of one-thirty-second inch. This means, that when in place, the outer flanges project over the ends of the stave, compressing it one-thirty-second inch, while the middle triangular flanges are driven into them. The flanges of this casting are longer than the web, so they not only project over the two ends of the staves, but also over the side adjoining. The bands are one-half inch in diameter and have square heads, with upset threaded ends. The foreman on the work was provided with a sheet showing the required band spacing for the various portions of the pipe, the spacing having first been calculated for the given pressure. The coupling has a curved seat, which sets on the outside of the stave, and two lugs so designed that they will hold both the head and nut of the band. The staves were hauled as near to the work as possible in wagons. From this point, in the bottom of the canyon, a narrow T-rail track was laid up the mountain 1 An Hydro-electric Development in American Fork Canyon, Utah, by A. P. Merrill. The Engineer- ing Record, May 9, 1908. 86 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. side to the grade. The cars containing the bands, staves, and so forth, were drawn up the track by a horse. All castings and steel bands were dipped in a special paint before being used. This was done in the bottom of the canyon near the lower end of the track. Construction was quite difficult in many places, especially at the two lower siphons, and in the tunnels located on the curves. In building the pipe around curves, short straight lengths from 50 to 75 feet long were first constructed, only enough bands being used to hold the staves in place. It was then shifted into the proper place with jacks. In doing this, some of the staves tended to slide longitudinally, which condi- tion required driving them to place, from the end of the pipe, with heavy mallets. The rest of the required number of bands were then placed. When the bapds were first placed around the dry staves, they were made just tight enough to hold the pipe together. After the water had been turned in, and the staves had become fairly well saturated, all bands were tightened. But at no time were they drawn so tight that the fiber of the wood was cut or crushed. The second tightening of the bands stopped practically all leaks, there being none in the two miles of pipe of any importance, and only a small number of minor ones where the pressure was highest. Air vents or standpipes were placed at all summits, and washout valves at the lowest point in the siphons. Much of the sediment that succeeds in passing the head- works will settle in the siphons, where it can be washed out, and its deteriorating effects on the machinery avoided. REINFORCED CONCRETE PENSTOCKS. The reinforced concrete penstock has not been used to a great extent, although, in some French plants, they have been used for many years under low heads, and are made in one continuous piece by hand. A newer process (System Siegwart) has been developed in Switzerland, whereby penstocks of reinforced concrete can be made by machinery, and capable of carrying pressures up to 300 pounds and higher, if desired. The thickness of the shell is a matter of requirements to suit the conditions at hand. Being manufactured by automatic machinery of very compact design, the pen- stocks can be readily made in the field. They are made in sections, the length of which depends on the size of the machine. The ends are provided with special joints, which, after in place, are filled with asphalt. To insure further tightness, an external band is slipped over the joint, and sealed by the asphalt. After coming from the machine, the penstock sections are coated inside with a layer of asphalt. This treatment renders the penstock absolutely water tight, and, in addition, reduces the skin friction, which means an increase in head above the ordinary concrete penstock. PENSTOCKS. BIBLIOGRAPHY. A.E.&M.E. 87 UNIV. OF C EXPERIMENTAL STUDY OF THE RESISTANCE OF THE FLOW OF WATER IN PIPES. A. V. Saph and E. W. Schoder. Proc. A. S. C. E., May, 1903. FLOW OF WATER IN WOODEN PIPES. T. A. Noble. Trans. A. S. C. E., vol. 49, 1902. FLOW OF WATER IN PIPES. C. H. Fulton. Journal Association Engineering Societies, October, 1899. FRICTION COEFFICIENT OF RIVETED STEEL PIPES. A. McL. Hawks. Proc. A.S.C. E., August, 1899. A TRAVELING MOLD FOR MAKING REINFORCED CONCRETE PIPES. F. Teichman. Engineering News, Feb. 20, 1908. PIPE LINES FOR HYDRAULIC PLANTS. Engineering Record, Dec. 21, 1907. A DIAGRAM FOR CALCULATING PENSTOCKS. Richard Muller. Engineering Record, Nov. 14, 1908. PENSTOCKS FOR WATER POWER PLANTS. Frank Koester. Engineering Record, Feb. 20, 1909. WOODEN STAVE vs. RIVETED PIPE. Journal Association Engineering Societies, p. 239. 1898. STAVE-PIPE. A. L. Adams. Trans. A. S. C. ., p. 676. 1898. CHAPTER V. POWER PLANT. GENERAL ARRANGEMENT. HYDRAULIC power plants have no standard arrangement, as there are so many types of turbines which are fed under various conditions; low heads may be utilized by horizontal or vertical turbines, requiring an entirely different proposition in the layout of the plant. The same is true for average as well as high head turbines; even in the latter case, which usually requires horizontal impulse wheels, vertical impulse wheels are sometimes used. Whatever arrangement is chosen, care should be exercised to locate the turbines so as to secure the highest possible head. Many tur- bines are dependent upon draft tubes to give additional heads. The turbines and regulators should be set in straight lines, and not scattered about the generating room. This also applies to high head turbines which are sometimes set on 45 degrees, which is done to give a more easy penstock connection. Such arrangements can be easily overcome by exercising a little judgment, and with little or no expenditure of money. This must be done for the sake of the appearance of the plant, and, of still greater importance, ease of operation. Whatever arrangement is decided upon, the flow of water to the turbines should be as free and easy as possible, to avoid friction. Heads utilized vary very greatly. At Genoa, Switzerland, 1 is a plant utilizing a head of 16.5 inches, while at Vouvry on Lake Geneva, Switzerland, there is a 2o,ooo-HP. plant operating under a head of 3116 feet. 2 From the foregoing figures, it will be seen that it is impossible to delineate the different designs of plants operating under various heads and conditions. In the following pages, only a few typical arrangements of low, medium and high head plants are discussed. In low and medium plants, the power house frequently forms a part of the dam, or adjoins the dam on the downstream side, or, in the case of a hollow concrete dam, is located in the body of the dam. In any case, the walls must be made waterproof to prevent seepage leaking into the power house. Forebays. Forebays must be located so as to deflect all foreign material as much as possible. This is best done by placing the deflecting wall at an angle of 30 to 45 degrees to the flow of the stream. This wall extends 2 or 3 feet into the water. Below the water level and fastened to the deflecting wall are rough screens, which in large plants consist of heavy bars. The forebay itself should be provided with a spillway, icerun and a sandtrap. The latter is best located just before the water enters the penstock. The bottom of the forebay must slope towards the sandtrap, 1 Thurso, Modern Turbine Practice, p. 13. 2 Wagenbach, Turbinenanlagen, p. 3. POWER PLANT. 89 FIG. i. Winnipeg, Manitoba, Plant. FIG. 2. Cross Section of Albany, Georgia, Plant. FIG. 3. Cross Section of Low Head Plant, Holyoke Water Power Company, Holyoke, Massachusetts. 9 o HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. to decrease the velocity of approach, and to give the sand and gravel an opportunity to settle. There must be an opening on the spillway side, to carry off the stuff which has collected in the trap. The penstocks must be provided with fine screens located above the sandtrap and in front of the penstock openings. In large plants, the screens and gates to the penstocks are usually placed in a house by themselves, which is provided with a crane to remove and lower screens. In temperate zones, it is advisable to install a heating system in the screen and gatehouse, for the prevention and thawing of ice. The design and arrangement of the forebay, in connection with the power house, depend upon nature and character of stream, particularly on the nature of the floating material; also the formation of ice is an important factor in this consideration. FIG. 4. General Plan of Colliersville Plant, Oswego County, New York. Low Head Plants. Low head turbines are usually located in open flumes, with vertical or horizontal shafts. With the vertical shaft, frequently several turbines are connected, by gearing, to one horizontal shaft. With this arrangement, special precaution must be taken to exclude moisture from the generating room. Fig. i shows the power plant of the Winnipeg Electric Railway Company, which utilizes water from the Winnipeg River, some 65 miles from the city of Winnipeg, and transmits the power at 60,000 volts. 1 A channel had to be cut to the Upper River near the Otter Falls, 120 feet wide, with a clear depth of 8 feet at normal low water; the channel is 8 miles long with a drop of 5 feet to the mile, thus giving a head of 40 feet. The units are McCormick turbines coupled to looo-K.W. generators, making 200 R.P.M., and are equipped with Lombard governors. It will be noticed that two pairs of turbines with two draft tubes are located in one casing, which is an extension of the penstock. The gates to the penstock are 1 Winnipeg, Manitoba, 6o,ooo-VoIt Hydro-electric Plant, by V. D. Moody. Electrical World, June 33, 1906. POWER PLANT. 91 provided with by-pass or relief valves, to facilitate the operation of the main gate. Fig. 2 gives the power house cross section of the Albany Power and Manufacturing Company, located on Big Shoals on the Muckafoonee River, about one mile below the city of Albany, Ga. 1 For utilizing the water, a dam 360 feet long and 20 feet high, and a spillway 150 feet long, have been erected. The turbines are of the hori- zontal, radial, inward flow type, made by the S. Morgan Smith Company. Each unit comprises four wheels, 33 inches in diameter, mounted in a cast-iron housing. fv H-*- ^,^%,^-r^,^-.-. "Tr^rt^rr'^r'j'-VJ'jrrrrJL^'^ FIG. 5. Cross Section of Colliersville Plant, Oswego County, New York. The discharge through' each pair of turbines passes through a draft tube 7 feet 9 inches in diameter, made of one-quarter-inch steel plate. The head on the turbines is 23 feet, and each unit is capable of developing 900 HP. at full gate. They are controlled by Lombard governors, and connected to 5OO-K.W. generators. Medium Head Plants. Medium head plants are usually equipped with Francis turbines of the horizontal or vertical type. When the horizontal type is chosen, there is usually only one wheel mounted on the shaft coupled to the generator, and supplied by a closed penstock. The unit must be set above ground water. With the adoption 1 Hydro-electric Plant at Albany, Ga., by R. W. Hutchinson, Jr. Electrical World, June 16, 1906. 9 2 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. 6. Map of the Shawinigan Power Plant, showing Arrangement of Headrace and Power House. FIG. 7. Cross Section of McCall Ferry Power House. POWER PLANT. 93 of the vertical shaft turbine, in almost all cases, there is more than one wheel mounted on the shaft, and they are located in an inclosed wheel pit, or else in a steel or iron casing. The intake and draft tubes to and from the wheel pit must be smooth, particularly when made of concrete. All intakes must be provided with vent pipes. Fig. 7 shows the general arrangement of the turbine, generating and transformer rooms of the McCall Ferry Power Company's plant. The plant is located on the Susquehanna River, about 25 miles from Chesapeake Bay, and is designed for a normal capacity of 100,000 HP., half of which at present is being installed. The dam is 75 feet high and 2500 feet long. The turbines, furnished by the I. P. Morris Company, are of the inward and flow Francis type, mounted in pairs on a single shaft. At a speed of 94 R.P.M., a head of 53 feet, and with a gate opening of 80 per cent, they are capable of developing 13,500 HP. The turbines are connected to 7500-K.W., 25-cycle generators. The location of the power house in relation to the forebay and dam is shown in Fig. 8. The building stands at an angle of 42 degrees with the face of the main dam. 1 It comprises a screen and gatehouse, generating and transformer room. At one end is a chute for ice and other floating material which may collect in the forebay. The whole building is built of concrete. The intake conduit for one main unit is comprised of three openings, 6 feet wide and 16 feet high. Eight feet back from the gates, they merge into one, which is 15 feet wide and for a short distance 13 feet high, and expanding, as the conduit forms the turbine chamber, to a height of 33 feet. There are two draft tubes, one leading from each wheel of the unit, and are separated by a vertical wall; the discharge outlet into the tailrace of each unit is composed of two passages, each 13 feet wide and 15 feet high. This arrangement of the draft tubes, since they are constructed of solid concrete, necessitated very complicated form work, especially as it was necessary to have easily curving surfaces, which would offer little or no resistance to the flow of water. The gates closing the intake conduits are 16 feet high and 6 feet wide, and are raised and lowered by a 1 5-ton crane. To facilitate the operation of the gates, they are provided with auxiliary gates which are operated by the crane. In front of the gates are screens, which are built up in panels, 10 feet wide, n feet high, and 4 tiers to each unit. They are handled by the overhead crane. The draft tubes are provided with grooves for stop-logs. The Niagara Falls power plants are medium head plants, but of an entirely different design from the McCall Ferry Company's plant. Four of these plants have adopted the vertical shaft turbine located in a pit. The arrangements vary somewhat, especially in the tailrace end. Power House No. i of the Niagara Falls Power Com- pany, equipped in 1895 w i tn 5ooo-HP. units, had no draft tubes; in the later plants, both on the American and Canadian sides, the turbines are equipped with draft tubes of different design, as is shown in Figs. 9 and 10. By eliminating the draft tube in the first plant, 700 HP. was lost for each of the ten units. It will be noticed in Fig. n, representing the arrangement of the turbines of plant No. 2 of the Niagara Falls Power Company, that the draft tube is divided, and the 1 The Engineering Record, Sept. 21, 1907. 94 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. A a. N n MHO POWER PLANT. 95 FIG. 9. Arrangement of Turbines, Niagara Falls Power Company, Plant 2. 9 6 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. r . _ tfK,' _ 1. - IS'IO- - f- - ^' 2 ' "5 FIG. 10. Cross Section of Power Plant of the Toronto and Niagara Power Company. POWER PLANT. 97 FIG. ii. Plan of Headworks of the Toronto and Niagara Power Company. FIG. 12. Cross Section of Kern River Plant No. i. 98 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. branches run down the sides of the tailrace tunnel. In Fig. 12, showing the arrange- ment of the turbines of the Toronto and Niagara Power Company, it will be noticed that there are two tailrace tunnels, and the draft tubes discharge into the tailrace from underneath. Half of the turbines discharge into the right-hand tunnel, and half into the left-hand. The turbines operate, under normal conditions, under a head of 145 feet. As the generators are located some feet above the headrace or forebay, it will be readily seen that long shafts are necessary. These shafts are divided into three sections carried on one adjustable thrust bearing, and guided by two side bearings. In the American plant, the bearings are supported on structural steel galleries, while in the Toronto and Niagara Company, they are supported on heavy concrete arched floors. The vertical penstocks leading to the turbines are well anchored at the upper and lower end. To allow for expansion, slip joints are provided at the upper end. In digging the turbine pits at Niagara Falls, many difficulties were experienced; some of them have a depth of 175 feet, a width of only 17 to 22 feet, and run the entire length of the power houses, which are 400 to 500 feet long. Much of the work was done in rock of different formation, which was exposed as the work progressed. All the plants are well provided with screens and ice racks, in separate houses. To prevent floating material from entering the forebay and screen rooms, the curtain walls are extended some several feet into the water. High Head Plants. High head plants usually have a simple arrangement of turbines, but, on the other hand, they have a more complicated arrangement of penstock and regulating devices. It is but natural that high head plants are located away from center of current distribution, therefore a large electrical equipment is connected to the plant. As the ground is cheap in such localities, it would be an unwise policy to crowd the generating room. Ample space must be allowed, particu- larly for the regulating devices. One of the most prominent high head plants in the United States is the Kern River Plant No. I, of the Edison Company, Los Angeles. It utilizes the water of the Kern River, and has a rated capacity of 20,000 K.W. The current at 60,000 volts is transmitted 117 miles to Los Angeles and other towns. To harness the water of the Kern River, a dam 203 feet long, 20 feet above the river level, was constructed. 1 The water is first led through 19 tunnels, aggregating a length of 42,910 feet; then through timber flumes 1520 feet long, then through a reinforced concrete conduit 503 feet long, where it enters a forebay. From the forebay leads a penstock 1697 feet long, to the power house. Contrary to the usual practice of laying the conduit on the mountain slope, this penstock is run through a tunnel. Throughout the course, there are several horizontal curves and vertical bends, amounting to 40 and 45 degrees. The penstock in the tunnel has a diameter of 7.5 feet, and near the power house the diameter is reduced to 5.25 feet. 2 1 The Kern River Power No. i, by C. W. Whitney. The Engineering Record, Aug. 10, 17, 24, 31, 1907. 2 For details on construction, see chapter on Penstocks. POWER PLANT. 99 As seen in the plan (Figs. 12 and 13), the branches from the main penstocks lead beneath the switching room to the impulse wheels. Two Allis-Chalmers impulse wheels drive one 5ooo-K.W. generator, and are mounted on the overhanging shaft of the generator. Each wheel is 9 feet 8 inches in diameter, and has 18 buckets. The guaranteed output of one pair of wheels is 10,750 H.P. at 150 R.P.M. The nozzle is adjustable, so that the stream can be deflected from the buckets. It is provided at the stationary end with a ball and socket joint heavily bolted down to the con- crete foundation. This swiveling head has to take up the full pressure of 375,000 pounds. The regulation of the wheels is effected by a governor, which deflects the jets of the two nozzles. The needles are adjusted by hand, and are usually set so that maximum size of jet which will be sufficient to develop the maximum peak FIG. 13. Plan of Kern River Plant No. i. loads expected for that period of the needle setting; in other words, there is always a maximum amount of water leaving the nozzles. The governor adjusts the deflect- ing nozzles in such a way that only as much water is directed upon the buckets as is needed for the load for the time being. The balance discharges below the buckets into the tailrace. Each jet has a maximum diameter of 7! inches and leaves the nozzle tip at a velocity exceeding 225 feet per second. It was necessary to provide means for receiving this tremendous power and deflecting the jet into the tailrace in such a way that its impact would not be detrimental to the structure against which it is directed. The arrangement designed, consists of a pair of heavy deflector plates by which the jet is diverted. These plates are curved, their design being such as to turn the water through two right angles before it is allowed to pass into the tailrace, thus reduc- ing the force of the water so that it can do no damage. The upper of these plates IOO HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. MM* dfff 1 'IronScnen ' -TimberScrtm consists of a channel, heavily ribbed and bolted to the concrete foundation. The channel at its upper end is slightly more inclined than the deflected jet. Thus, the jet strikes the bottom of the channel at a small angle, and therefore tends to spread and fill the section. The channel gradually widens, and the jet is consequently offered a larger resistance area. The lower part of the channel is curved, and at its end the jet discharges almost perpen- dicularly downward. The bottom plate is S-shaped, its upper end being flush with the bottom of the wheel pit, the lower end being practically level. The jet strikes the bottom plate almost in the turn of the S and under a small angle. Thus the jet is again forced to spread and follow the base of the bottom plate. The deflectors are lined with movable steel plates wherever the surfaces are exposed to the flow of the deflected jet, and held in position by lag screws. The plates are 7 feet wide, and the lower one projects out into the tailrace 8 feet. The wheel races are lined with steel on both sides, and fitted with steel back plates just back of the nozzle tips, to keep the splash water out of the shaft alley. The tailrace is 29 feet wide and ex- tends the length of the power house. It is fitted with two 2 5 -foot steel plate weirs; the lower weir at the end of the tailrace being 4 feet below the level of the upper, which has its crest 13 feet 6 inches below the line of the nozzles. The penstock branches enter the power house at the south side, and after passing across the transformer rooms, and before joining the nozzle bases, connect to 28-inch These valves are of a special design, and are separately SwHcH "Board FIG. 14. Power Plant of Snoqualmie Falls. Power Development. cast-steel gate valves, operated from the control switchboard by a I.2-H.P., i2o-volt Allis-Chalmers motor. These motors are mounted vertically and operate at 460 R.P.M. It requires 7$ minutes to open or close a valve by means of the motor. All of the gate valves are equipped with 4-inch by-passes. In the machine room of the power house is installed a Dibble reservoir gate equipped with an indicating dial and a registering chart, for measuring the water in the forebay. A very unique arrangement of a high head power plant layout is that of the Snoqualmie Falls and White River Power Transmission Plant in Washington (see 102 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Fig. 14). A shaft of rectangular section, 27 feet long and 10 feet wide, was sunk in the river bed about 300 feet above the Snoqualmie Falls. This shaft reaches a depth of 270 feet, or the level of the river below the falls, and connects with a tunnel 24 feet high and 12 feet wide, having a slope of two feet, utilized as a tailrace. 1 The underground power station begins at the bottom of this shaft and is 30 feet high, 40 feet wide, and about 200 feet long. The tunnel, power house and shaft are lighted by incandescent lamps; the natural draft through the tailrace and up the shaft provides good ventilation. The temperature remains constant at 55 F., and the generating room is perfectly dry. There are three vertical shafts leading to the power house; one for elevator, and cables conducting the current to the distributing station on the shore; the others for penstocks. The elevator shaft, 10 feet long and 8 feet wide, is lined with steel casing backed with concrete. The penstock has a diameter of 7.5 feet, and is built of steel plate sections, having a thickness of 0.5 inch on the top and one inch on the bottom. Here it is connected to a horizontal chamber, 10 feet in diameter at the penstock junction and 8 feet in diameter at the opposite end. From this chamber, four 4-foot branches lead to the turbines, which are of the Doble impulse wheel type, having a capacity of 2500 HP., making 300 R.P.M., and connected to I5OO-K.W., icoo-volt, 3-phase Westing- house generators. Each unit is composed of 6 impulse wheels mounted on a common shaft, each wheel having two jets. EXCAVATION AND FOUNDATION. Selection of Site. In connection with the head and tail race, and the selection of the site for the power house, the character of the soil must be considered. It frequently happens that forced choice for the site of the power house of a plant is on unsuitable soil. To overcome this difficulty there are two ways: either change the location of the site, or strengthen the soil. It is therefore essential, before drawing up plans of the general arrangement, that accurate information is obtained regarding the bearing power of the soil. This is secured by sinking test holes. If the soil is of an unknown character, test loads must be applied. Test Holes. Test holes must be sunk in alluvial soil, or made land, to secure accurate knowledge of the underlying strata. Frequently, in the sinking of test holes, rocks are encountered; this indication must not always be taken for strata of rock. To ascertain that rock does not exist, holes, short distances apart, must be sunk, so that an accurate plot of the soil is obtained. These holes are usually 25 to 50 feet apart. When the magnitude of the project does not warrant the use of well or core drilling machines, test holes can be put down by driving iron pipe; a small and large one can be used, the large one acting as a shell for the smaller, to prevent the hole from caving in. A core can be secured by leaving the lower end of the small pipe open, and working without the use of water. An easy method is to force a 1 Electrical Review, June 18, 1904. POWER PLANT. stream of water down the small pipe, which, returning to the surface through the large one, brings up specimens of the soil. With a scheme of this kind, holes may be driven 50 feet or more, depending on the character of the soil. Character of the Soil. When rock is present within moderate depth, the founda- tion must be carried down to same. The surface of the rock must be leveled and cleaned in order to give a good bearing. The bearing value of rock varies within wide limits, from 10 to 200 tons per square foot and even higher. A clean sharp sand makes an ideal bearing soil, and is easy to excavate; in addition, it may be utilized in the making of mortar and concrete, which means considerable in saving and expense. Quicksand, either wet or dry, if in thin layers, should be removed entirely; where it is underlaid with a firm strata, it can be confined by means of a concrete coffer dam, and the foundations can then be floated on same. The footing or mat covers the entire area within the coffer dam. In soft or alluvial soil, piling is necessary for heavy foundations. As a general rule, the firmness of the soil increases with the depth; there are exceptions, however. In Chicago, the firm upper layer of soil, from 10 to 20 feet in thickness, is underlaid by a soft clay stratum about 70 feet thick, under which is a stratum of firm clay. Similar conditions have been noticed in different parts of the country. Clay varies greatly in consistency, varying from fluid to hard shale; the latter, when exposed, will disintegrate. It varies greatly according to the opportunity for absorbing or losing water, and because of this, it is very troublesome. To make a good foundation, clay, sand and stone are spread on it, and then well rammed down. The stones should be small enough to permit their being handled by one man. As a general rule, hydraulic plants are located in rocky, mountainous districts; and as the plants are frequently located on the mountain slope, a simple and efficient way is to blast the rock in steps. Bearing Power of Soil. For the bearing power of soils, the values in the following table are from Baker's " Treatise on Masonry Construction": TABLE I. SAFE BEARING POWER OF SOILS. Kind of material. Safe bearing power in tons per square foot. Minimum. Maximum. Rock the hardest in thick layers in native bed.. Rock, equal to best ashlar masonry 200 25 IS 5 4 2 8 4 2 o-S 3 20 10 6 4 10 6 4 I Rock, equal to best brick masonry Rock, equal to poor brick masonry Clay, in thick beds, always dry Clay, in thick beds, moderately dry Gravel and coarse sand, well cemented Sand, compact and well cemented Sand, clean, dry Ouicksand, alluvial soils, etc 104 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Weight of Masonry. For the total bearing stress of a foundation on the soil, the weight of the foundation itself must be included. For use in this connection, Table II, taken from the same authority, gives the weights of various types of masonry : TABLE II. WEIGHT OF MASONRY. Kind of masonry. Weight in pounds per cubic foot. Brickwork, pressed brick, thin joints ... . 1AC Brickwork, ordinary quality T2C Brickwork, soft brick, thick joints l *5 Concrete, i cement, 3 sand, and 6 broken stone I/1O Granite, 6 per cent more than the corresponding limestone Limestone, ashlar, largest blocks and thinnest joints. Limestone, ashlar, 12 to zo-inch courses and f to }-inch joints . . ... 160 TCC Limestone, squared stone 148 Limestone, rubble, best 14.2 Limestone, rubble, rough 1*6 Mortar, i Rosendale cement and 2 sand 116 Mortar, common lime, dried IOO Sandstone, 14 per cent less than the corresponding limestone ... Piling. There are different kinds of piles, such as wooden, sand and concrete. Wooden piles are used in two ways: they are driven down in soft soil to compact it, in which case the bearing power depends entirely upon friction. In other cases, the piles are driven to rock or solid strata, in which case they act as a column. Wooden piles must be cut off below the permanent ground water line to prevent the caps from decay, and on top of wooden piles is spread a concrete cap. Sand piles are used for strengthening the soil, by driving wooden piles or hollow sheet steel tubes down, then withdrawing them; the hole is then rilled with sand. These piles are usually spaced close together, and do not act like wooden piles; the whole soil is made compact. In the last few years, much use has been made of concrete piles, both plain and reinforced with steel. Two general methods are employed: in one, the piles are molded, then driven; in the other, the mold is driven with a removable core, the concrete being placed after the core has been removed. There are several other schemes of concrete piling, but the principles do not differ from those mentioned. The many advantages of concrete piling are obvious, such as, they cannot decay, and for this reason they can be left as high as desired; and more economical foundations secured, owing to the fact that the ground water line does not signify the point for the cap, as it does with wooden piles. The friction and bearing power is higher than that of wood. In addition, the diameter of a concrete pile can be varied at will, while the diameter of a wooden pile rarely exceeds 14 inches. For this reason fewer concrete piles are necessary for supporting a given load. They are, in a way, down- ward projections of the monolithic mass of the foundations. POWER PLANT. 105 Corrvqafecf Reinforced Concrete Piling Wood FIG. 16. Comparison of Foundations using Wooden and Concrete Piles. Test of Piles. The difference in bearing power between a conical and a cylin- drical pile was shown by an experiment, tried on some work at the United States Naval Academy at Annapolis, Md. A Raymond pile core, tapering from 6 inches at the point to 20 inches at the butt, was driven 19 feet, until the penetration, under two blows from a 2100- pound hammer falling 20 feet, was seven-eighths of an inch. A wooden pile 9^ inches at the point and n inches at the butt, and of the same 106 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. length as the conical pile, had a penetration of 5 finches under two blows of the same hammer falling 20 feet. A ly^-foot test pile having the same dimensions as the concrete pile above mentioned, and having a penetration of i inch under twenty blows of a steam hammer, was loaded with 41 tons. Levels were taken during the loading, and at intervals for one month. At the end of the month the total settle- ment was 0.007 foot, or three-thirty-seconds inch. Concrete Mat Construction. Concrete mat construction is frequently used with earth filling, also on soft ground where pile driving has been done. To guard against unequal settlement, it is preferable to extend the mat Under the entire building. The thickness of the mat varies with the load to be applied, and may be kept down by reinforcing same, preferably with old rails. The mixture for the concrete in the latter case is i : 3 : 6; if a more expensive mixture is desired, use i : 2\ : 5. When plain concrete slabs are used, rubble concrete may be employed up to a few inches of the floor line. Foundations. In determining the size of foundations, the weights of the machinery must be secured from the manufacturer. In most cases, the sizes are indicated on the blue prints; this, however, is not sufficient, as one case cannot serve for all; as all depends on the character of the soil. In the case of turbines, the weight of water must be figured in with the weight of the machines. Foundations must always be made of concrete, i : 2\ : 5 for the smaller type, and i : 3 : 6 for larger foundations. As will be seen in the chapter on buildings, the substructure of an hydraulic plant is usually a monolithic mass of concrete. The forms for the foundation should be so designed that they can be used over and over again, where there are a number of isolated foundations. This is also true in the case of core forms in wheel pits, draft tubes, etc., provided there will be no serious interruption of the work. The forms must not be removed until the concrete is thoroughly set, otherwise the concrete will assume a different shape. For locating anchor bolts, templates must be constructed. They are made of planking and thoroughly braced with diagonal bracing, otherwise the template will warp out of shape, and throw out the location of the anchor bolts. The drawings should not only contain the elevations of foundations, bolts, etc., but also dimensions to simplify construction of same. Anchor Bolts. The anchor bolts for machinery are preferably made removable, particularly with large machinery. Under ordinary conditions, they need not project into the foundations more than 18 or 24 inches. They are provided on the bottom end with a cast-iron washer, 6 to 12 inches square. The 6-inch washer is sufficient for bolts i to i^ inches in diameter. They are inclosed, between the washer and foundation (with no grouting), in a pipe, with a diameter about one inch larger than the bolt. The bolts are threaded at both ends to permit adjustment. All bolts, washers and pipes should preferably be of standard size, to minimize expense in draughting department, shop and field. Grouting. After the machinery has been properly set in place, and anchored down, grout must be poured in to establish a final setting for the bed plate. There must be an allowance in the foundation from one inch to two inches; and even in POWER PLANT. 107 small foundations such as for pumps, etc., it must not be less than three-fourths of an inch thick. The grouting itself is a thin, rich mixture of cement mortar with little or no sand, in order to fill up all spaces between the bed plate and foundation and around the anchor bolts. SUPERSTRUCTURE. Architectural Features. Rapid progress has been made in the last few years in the design of hydraulic plants and their substations. The designs show a more harmo- nious agreement between engineer and architect; this, however, varies with the different countries; some lay much stress upon the artistic appearance, while others confine their attention solely to utilitarian objects, disregarding entirely the architectural features. Necessity requires only a building of sufficient support, to shelter and FIG. i. Entrance Hall of Plant No. 2, of the Niagara Falls Power Company. protect the machinery and those who operate it, and must be of durable construction. An ornamental building will not increase the efficiency of the machinery; it increases the fixed charges. But, at the same time, it is required from an aesthetic point of view, and will, no doubt, have certain effect upon the moral of the operating force, ' whose efficiency will be increased thereby. The ultimate aim in the design of an hydraulic plant is, to generate electricity 108 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. upon a commercial and economical basis. As a rule, hydraulic plants are located away from centers of population, consequently the architectural features are neglected, as is evidenced by many of the soap-box-like structures in America and England. It seems strange that the hydraulic and electrical engineers of these countries pay little or no attention to the architectural features of buildings. There are, however, a few examples which show the excellent harmony between the engineer and architect. Some of them are found at Niagara Falls, for example the upper works, or headrace at Dufferin Islands, of the Ontario Power Company, whose power house is located in the Gorge. It might be of interest to give an extract from a report of the park commissioners, when the franchises were granted to the various power companies on the Canadian side of Niagara Falls. "All of the works and structures connected with the electrical power projects have been designed with the object, not only of doing the least possible injury to scenic conditions, but the commissioners are confident in the belief, that when the several works are completed, the consensus of opinion, by the vastly increased number of visitors that are expected to visit the park, will abundantly sustain them in their contention, that the park as a whole, with its wealth of electrical machinery, will then be of tenfold greater interest to the great majority visiting it." A building for power-house purposes should not be too ornate, as is frequently found in Europe. Simplicity of design and harmonious agreement with its sur- roundings are of prime importance. The machinery must be well arranged, sufficient ventilation and an abundance of light provided. A plant, one of the foremost in America, not only regarding the equipment and capacity, but also from the architectural point of view, both exterior and interior, is that of the Niagara Falls Power Co., Figs. 2 and 3. The superstructure is of rough faced granite blocks with a slate roof. In front of the main structure is the screen house. Both structures are well provided with windows for light and ventilation. The building is tasteful in design and is typical for its purpose, namely, that of a power plant. The interior design of the generating room is in keeping with the exterior, while the entrance hall has been more elaborately treated (see Fig. i). The electric illumination of the entrance is in perfect harmony with the architectural design. What has been done in main hydraulic plants has been typified in sub- stations, as seen in Fig. 4. In Europe, the architectural features, from an American point of view, are exaggerated in the extreme. Fig. 5 represents the hydraulic plant of the city of Stuttgart, Germany. The style is that of the fourteenth century, and is much favored in Continental power plant practice. The approach leading from the street to the power house harmonizes with the main structure. This plant is equipped with four 3OO-HP. hydraulic units, and a small storage battery of 300 ampere-hour capacity. In America this plant is considered small, and in all probability the architectural features would be neglected. It will be observed that much money is spent for architectural purposes, in fancy cornices and off-sets, in the above building. However, it is not necessary to secure pleasing architecture in such a manner; contrast the above plant with that of Obermatt, near the city of Lucerne, Switzerland POWER PLANT. 109 FIGS. 2 and 3. Exterior and Interior of Plant No. 2 of the Niagara Falls Power Company. FIG. 4. Substation of Great Northern Power Company, Duluth, Minnesota. (no) FIG. 5. Municipal Plant, Stuttgart. Germany. POWER PLANT. ill (Fig. 6). Attention is called to the novel design of the windows. The interior of this power house is given in Fig. 7. A German plant designed entirely on modern " Secession " style is that of the power plant of the Urfttalsperre at Heimbach, shown in Figs. 8 and 9. The entire design, such as the arrangement and design of the pilasters, roof trusses, windows and switchboards, is patterned on the same line. Special attention is called to the design of the windows and doors, the rear-end wall and side-wing towers. The latter may be taken for ornamental bases for smokestacks. While individual features, such as switchboards, have appeared in Continental practice for years, the design, as a whole, is a bold one in power plant practice. Material. In the construction of electric plants, it is essential to have the buildings as fireproof as possible. This can be secured by using concrete, brick, terra-cotta, or steel. The material adopted depends greatly upon the locality and on the labor supply. In some countries or sections of countries, skilled labor is easily obtainable; in many places material and skilled labor have to be carried to the site. For buildings in such localities concrete and steel are the best; with a few skilled foremen and pick- up labor, such buildings can be easily erected. This is particularly true in tropical countries where labor is difficult to secure. In countries subject to earthquakes, a framework of steel covered with corrugated iron serves admirably well; the corrugated sheets lapping over each other five inches, and from one and a half to two inches on the side. In some cases painted corrugated sheets are used, owing to their cheapness of cost. The material should preferably be galvanized, in which condition it should not receive a coat of paint until it is exposed to the weather one or two years, and the surface has become slightly oxidized. Walls. The interior of the generating and switching room should be kept as light as possible. It is advisable to apply a smooth surface of cement plaster, and whitewash same. A more pleasing effect will be secured by facing the walls with enameled tile and a wainscoting of contrasting color. Pilasters may be used to break up the monotony of a smooth surface, and conceal steel columns of the crane runway. The tiling of the walls should preferably be of cream color, while the wainscoting and ornamental panels of olive-green. However, the selection of the latter color is governed by other conditions, particularly that of the floor. The switching room should be separated from the generating room by a parti- tion wall of fireproof material. Large openings should be left in this wall, particu- larly in the control switchroom, so that the operators can have an unobstructed view of the generating room; glass partition walls serve the same purpose, and have the advantage of excluding all dust. Floors. In the construction of floors, non-combustible material must be employed. As the substructure is, in most cases, built of concrete, it is but natural that the floors should be of concrete; in this case, they must have a granolithic finish of dark color, to render drips of oil inconspicuous. Some authorities dislike concrete floors, for the reason that such floors produce grit by wear, which is stirred up by walking and sweeping, thereby getting into the bearings and other parts of the 112 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIGS. 6 and 7. Exterior and Interior of Obermatt Plant, Lucerne, Switzerland. POWER PLANT. FIGS. 8 and 9. Exterior and Interior of Urfttalsperre Plant at Heimbach, Germany. 114 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. machinery. Another reason is, that a person on a concrete floor coming in contact with any high-tension wiring, would instantly be killed, while a wooden floor would minimize the risk. The use of wood in power plant construction, and particularly in floors, is obsolete; the principal reason being, that around machinery, there is more or less dripping of oil, which soaks into such a floor and shortly gets into a very inflammable condition. In fact, the whole trouble of some power plant fires has been due to an insignificant blaze of the wooden flooring. Probably, the best floor finish for a generating room is tile or mosaic; being smooth, it is easy to keep clean, and has a very handsome appearance. Penstock connections and generator leads are frequently laid in trenches, curbed and covered by plates; for the sake of appearance, if possible, they should run lengthwise or transverse to the building. It is poor engineering to have the FIG. 10. Municipal Plant, Geneva, Switzerland. branches of the penstock embedded in the concrete floor, and it is still worse to have flanges of same project above the floor. Roof. The cheapest non-fireproof roof construction is boards covered with roofing felt, on which is laid a pitch and gravel roof. This type of roof is suitable for slopes ranging from two inches per foot up to 45 degrees, but is preferably con- fined to the flatter slopes; steep inclines increase the expense materially. Slag and gravel roof is often applied to reinforced concrete slabs or arches. In Continental Europe, pumice stone is occasionally used in concrete for roof purposes; while in America, cinder concrete is often used instead of gravel. Both of these concretes are much lighter than the ordinary gravel concrete. POWER PLANT 115 In constructing a gravel roof, the concrete is first covered with a layer of hot pitch over which is laid the tarred roofing felt, the sheets lapping over each other about half the width of the roll, and each sheet being mopped with pitch as it is laid. Over the entire surface an even layer of pitch is then spread, in which, while still hot, slag or gravel is embedded. Another well-designed roof requires a preliminary preparation in regard to the steelwork in the shape of T-irons laid over the roof purlins. Between these, book tiles are laid, covered with Spanish roll tile. The advantage of the concrete roof construction, and the two latter methods mentioned, is that they are entirely fireproof. Steep inclines are necessary for any tile or metallic roof, and the height should be at least one-third of the span. Where flat roofs are used, surrounded by parapet walls, metal flashing should be provided. FIG. ii. Substation, Stansstad, Switzerland. For tropical countries, the pitch and gravel roofs, so frequently used in the temperate zones, are not suitable, a special material being prepared for use in such climates. One of the troubles with corrugated iron roofing arises from its making an oven out of the building which it covers, unless an air space is provided to insulate the room directly below the roof from the heat, which may be done by applying sheathing on the bottom chords of the roof trusses. This sheathing reduces the height of the room and increases somewhat the difficulty of properly ventilating it. Another trouble with corrugated iron roofs arises from the condensation of moisture upon their surface, when the roof for any reason becomes cooler than the air. This moisture occasionally causes trouble with electrical machinery. Il6 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Leaders. One square inch of leader area must be provided for each 100 to 150 square feet of roof. The leaders must not be smaller than four inches in diameter. In ordinary buildings, galvanized iron leaders are used, while in more pretentious plants, copper of rectangular shape is sometimes employed. All leaders must be provided on their upper ends with removable guards or strainers. Doors. The door through which material is received should .be large enough to admit a railroad car; a door 12 feet wide and 16 feet high will suffice. In some cases, Dutch doors are used, of which the upper half can be used for ventilating purposes, the lower half remaining closed. When it is desirable to open the whole door at once, a folding gate is provided to keep out the curious. Swinging doors are for many reasons inconvenient. There are on the market various designs of doors for economizing room, such as vertical and horizontal sliding doors, sectional folding and rolling doors. The doors should be ornamental and massive. Oak, well paneled, makes a very hand- some door, particularly when trimmed with bronze. In many cases, entire metallic doors of ornamental design are used, as it is desired to avoid the dull appearance of the usual fireproof shutters. Windows. The generating and switching room must have abundant light, and large windows must be provided. It must be borne in mind, however, that fireproof windows cost from $0.80 to $1.00 per square foot; as walls are always calculated by the builder as solid, the cost for doors and windows is an additional expense. It is common practice to have the windows of ribbed wire glass, because they keep out intense rays of the sun, and do not shatter when broken. It is desirable in some localities, to protect the lower windows of a building with a wire mesh or bars. The window sashes should be metallic or covered with metal. The windows, together with the crane pilasters, must be symmetrical with regard to the arrangement of turbines. Arched windows are preferable for power plants and are handsome in appearance. If the windows are of large design, care must be exercised to properly panel them to harmonize with the design of the building. Too frequently, the design, as well as the arrangement of windows, spoils the appearance of an otherwise well-designed building. Stairways and Elevators. Ample stairway provision must be made, because easy access to all points is essential. The stairways should be about 4 feet wide, have easy steps, and be free from turns. Where the floors are more than 12 feet apart, the stairway should be broken by a landing. Stairs should be built of steel framing, with treads of checkered steel, slate, or covered with other anti-slip material. Where elevators are installed, to eliminate the service of an attendant, they must be of the self-starting control type. Switchboard Gallery. The switchboard galleries must be designed to give plenty of room for all ducts and passages necessary for wiring. In some plants, part of the flooring is made up of slate or soapstone slabs, which can be removed should the necessity arise. The reason for employing this material is, that such stones contain very few metallic elements and are first-class insulators. As a matter of precaution POWER PLANT. 117 in other cases, rubber mats are sometimes placed on the floors where attendants have to stand, while operating or making inspections. The switchboard itself should be of artistic design, harmonizing with the costly instruments. It should be made up of an ornamental iron structure faced with white marble or enameled slate. In central and substations, white marble panels are more in vogue abroad, while the enameled slate is favored in America. All instru- ments must be well grouped. When instrument pedestals are used, they must be well arranged, and at the same time, convenience of operation must not be sacrificed. Crane. The crane is not an architectural feature, but even this unpromising subject may yield to proper treatment. It should be designed in a way to conform with the roof trusses. In this connection, the latticed type crane has a better appear- ance than the unsightly, fish-bellied, box girder, so prominent in use. Heating. If the plant be located in a temperate latitude, it will be necessary to supply means for heating. The most common means are by steam or hot water. The latter is, however, in many cases, very inconvenient, due to the large amount of radiating surface necessary; there is also the danger of freezing exposed pipes. Steam heating is far better for a large building, being more economically installed and more easily handled. There are two systems of steam heating which, may, l?e used that will produce satisfactory results, viz., direct radiating system and hot blast. With the direct radiating system, the heating may be done either by pipe coils or radiators or a combination of both. The coils may be located either on the ceiling or under the windows; the latter method is the more efficient, the neces- sary radiating surface being about 10 per cent less than that of ceiling coils. A table for calculating the necessary amount of radiating surface to heat a room of given dimensions is given in Table I. TABLE I. FACTORS FOR RADIATING SURFACES. iladiatoi S. Coils. Outside temperature. o o o O Inside temperature. 45 54 63 72 Si 45 54 63 7* 81 4-inch brick wall .112 137 I 7 I .209 .156 .095 .121 I 5 I .182 .226 8-inch brick wall .076 OQ3 . 106 143 174 .064 .082 . 102 I2 5 145 12-incb brick wall .O?2 .06? .081 .099 . 121 .045 .057 .072 .088 . 107 1 6-inch brick wall O43 O^ .066 .081 .098 .036 .046 .058 .072 .087 2o-inch brick wall .078 .046 .058 .072 .087 .032 .041 .052 .063 .077 24-inch brick wall O33 .041 .OC.I .063 .076 .028 .036 45 .055 .067 28-inch brick wall .O2Q .036 .04? .054 .064 .025 .031 .028 .048 .056 Window, single I 99 .241; . 306 377 .458 . 169 .214 .270 .331 403 Window, double - 004 . IOC. *44 .178 .215 79 . 102 . 127 .156 . 190 Skylight, single .18? . 22? .282 .348 .421 .156 .199 .249 .305 .371 Skylight, double . IO2 .12? J 57 J 93 .234 .087 .III .138 . 169 . 206 Floor, wood .OI4 .Ol8 .021 .027 .032 .012 .015 .018 .023 .028 Ceiling, wood .Ol8 .O2I .027 .033 .041 .015 .OI9 . O24 . 029 .035 Floor, fireproof .O2O .026 .032 .039 .047 .018 .022 .029 .035 .042 Ceiling, fireproof .024 .029 .037 .045 .055 .O2O .027 .033 .041 .048 Door .068 .084 . 10? . 129 .156 .058 .074 . 092 II 3 .138 Cubics .OO4 .OO? .006 .007 .008 .003 .004 .OO5 .006 .007 Ii8 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The method of using this table is as follows: Calculate the amount of exposed wall and glass surface in square feet, and figure the cubic contents of the room. Multiply these results by the various amounts shown under the headings for the temperature desired; the sum of the results will be the number of square feet of heating surface necessary. It will be noticed that the table calls for an outside temperature of oF.; it is general practice to figure this temperature as the minimum. Hot blast heating is done by blowing air, by means of fans, over coil surface and transmitting the heated air to various points of discharge by means of galvanized iron ducts. There are two methods of doing this, one by recirculat- ing the air, that is, using the air in the room over and over again. The other method costs more to operate, but insures better ventilation, as it supplies heated outside air. The style of boiler to be used depends entirely upon the size of the plant; up to 25 HP. a cast-iron sectional boiler will produce good results; up to 100 HP. a ;hprizpntal tubular or a locomotive boiler is as good as can be used; anything over this; ' a .water tube boiler will be found most economical. It is advisable to locate the btojler in a separate building, in order to remove any possibility of dust and ashes accumulating on the machines. For calculating the size of the boiler, 100 square feet of radiating surface to one boiler horsepower (30 pounds of steam per hour) for direct radiation is the accepted practice. For blast coil work, a rough rule is 30 square feet of radiating surface per boiler horsepower. This latter, however, is inaccurate, as there are numerous other conditions, such as shape of heaters, number of air changes per hour, etc., which must be taken into consideration. The steam mains supplying the heating system may be calculated on the following velocities: 5000 to 6000 feet for the main distributing lines; 3000 to 3500 feet for vertical rising lines, and 1200 to 1500 feet for individual branch mains. No branch main should be less than i-inch pipe. These velocities are for low-pressure steam. The return mains are generally one- third to one-half the size of the supply mains. Careful provision must be made that all piping pitches in the direction of the flow of steam; this pitch must be at least three-fourths inch in every 10 feet. It is important to cover all supply and return mains with a good non-conductor, not only for economy's sake, but also to minimize the fall in pressure in the steam main which often produces snapping and cracking in the pipe. Ventilation. For ventilating generating room, louvres with swinging windows should be avoided in the roof, above or near current-carrying apparatus such as generators, motors or switchboards. The windows in the wall on the switchboard side must be fastened, no provision being made to open same, or else a locking system provided. This precaution must be taken to prevent short-circuits caused by rain and dust particles, blown in from the outside. Where air blast transformers and storage batteries are installed, forced draft must be used; the discharge gases of the latter must be carried through special ducts up through the roof. 119 POWER PLANT. Lighting. To provide for an emergency, in case of a complete shut ( generating and switching room must be provided with a multiple systemltrf wiring:' This is essential, for instance, in the alternating current plant with motor-driven exciters; a complete shut-down would seriously handicap the locating of the trouble. In modern plants, the switch gear is operated by motors, supplied by a storage battery which may furnish light also. As much as possible, wires must be concealed and run in ducts of approved design. Lavatories. For sanitary reasons, well-equipped lavatories must be installed in all plants. The plumbing must be of good substantial material, enameled basins, bowls or sinks. Bowls are preferable to sinks; bath and toilet floors should be tiled; the partitions of white enameled slate, or, if a more expensive construction is desired, of marble. The advantage of white finish is, that it enforces cleanliness by making dirt conspicuous. The drain must run to avoid all ducts and wiring, and be properly provided with traps and vent pipes; the latter must extend above the roof. Preferably at the side of lavatories, lockers should be installed to enable the men to change their clothes and clean up. The lockers must be large enough to contain a complete change of clothing, permitting a man in winter to hang up an overcoat. Sufficient room must be given in the aisles to allow the men to make necessary changes. Conclusions. Many have biased opinions, that by the erection of power plants for the utilization of water power, the scenery of the country will be destroyed. It is an entirely mistaken idea. It is only a matter of ability, on the part of the engi- neer and architect, to design a plant to harmonize with the surroundings. In fact, plants have been installed, greatly enhancing the beauty of the scenery of the country. To bear out this statement, illustrations are given in Figs. 12 and 13. The latter is the head race of the i5,ooo-HP. plant utilizing the water of the Sill, near Innsbruck, Tyrol. The valley without the canal, spillway, and building for attendants, would undoubtedly be monotonous, particularly in the mountainous country. Considering Fig. 14, the power plant at Tivoli on the Tiber, it is perhaps the most picturesque power plant ever designed. The building itself with its few arched openings is simple; the head-race is designed after the style of the old Roman aqueducts, and carries more water than the power plant needs. Without question, the scenic value of the country has been increased. This is more remarkable, because the plant was not erected for the immediate locality, but to serve the city of Rome. It cannot be expected that all plants should be architecturally treated in the same manner as some of the above cited. However, it must always be remembered that a pleasing appearance can be secured at little or no additional expense. In fact, many prominent plants which are masterpieces of ugliness, have cost more per unit capacity than those which are noted for their fine appearance; a proper knowledge of architecture being requisite to secure good results. A.E.&M.E FIG. 12 Bird's-Eye View of "Sillwerke," Tyrol. (120) FIG. 13. Headrace of the "Sillwerke," Tyrol. POWER PLANT. 121 By courtesy of the Proprietor of Gassier' s Magazine. FIG. 14. Tivoli Plant, Rome, Italy. 122 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. STRUCTURAL STEEL. Roof Trusses. As it is essential that structures be fireproof, first of all, structural steel roof trusses are required. In small plants, the roof trusses are preferably carried on the side walls; while in large plants, the trusses are carried on columns, which support also the crane runway. The outline of the roof trusses depends upon the kind of roof to be supported and the pitch, which depends also upon architectural conditions. When slate and shingles are the only roofing materials available, steep slopes are necessary, in order to cause the water to run off rapidly, and prevent its working up under the roof and causing leaks. With modern methods of waterproofing, a slope of 2 inches per foot is sufficient to supply the requisite drainage. Such roofs have many advantages: they require less material than those of steeper pitch; also, they are easier to build, and the waterproofing is readily applied. Steeper roofs, however, are often used, owing to the fact that they are considered more economical in steel, but this advantage is more than offset by applying the roofing. -e. ^uo.1 Panels. - FIGS, i to 4. Typical Roof Trusses. The accompanying sketches, Figs, i to 4, illustrate some of the usual forms employed in roof construction. An inspection of the various cross sections of plants given in different parts of this volume, show a number of other forms of roof trusses in actual use, some of which are more or less ornamental. In the design of roof trusses, it is necessary to know the span and the load as well as the spacing of the trusses. In power plant work, the location of columns is largely determined by the equipment. For the sake of rigidity, the trusses must be directly connected to the columns; it will therefore be seen that the span and distance between trusses is POWER PLANT. 123 fixed, and that this distance may or may not permit the most economical design of truss. The loading depends upon the locality of the plant, and the style of roof to be used. In New York City, the live load of a roof having a pitch less than 20 per cent, is 50 pounds per square foot, and for a pitch exceeding 20 per cent, is 30 pounds per square foot. This live load (snow and wind) is the vertical component on the projected area. In localities subject to severe wind storms, the roofs must be properly anchored, particularly when they rest on walls. The top chords of the trusses are tied together by purlins, which support the roof; on deep trusses, the lower chords have longitudinal bracing. In addition, at the end panels, and in long buildings at intermediate panels, angles or chords are used for diagonal bracing in the plane of the upper and lower chords. An overhead crane, operated by power or hand, is essential in the generating room, also in the transformer room of larger power houses. In brick buildings, the crane runways may be supported on pilasters of the wall designed for this purpose. This type of construction is adopted only in small and comparatively low buildings, or in localities where masonry is cheaper than structural steel. In localities where building materials are expensive, it is better to erect a corru- gated sheet steel iron frame structure. This is particularly true in tropical countries where the labor and building materials have to be imported. For this type of building, galvanized corrugated iron is used, Nos. 18 and 24 gauge, and for the roof, Nos. 20 to 26 (United States Standard). The siding is in all cases two gauges lighter than that of the roofing. Black or painted sheets are occasionally used, but as they are not so durable as the galvanized sheets, they cannot be recommended. The best grade of this material is called "Muck Bar" corrugated sheeting, and is much more durable when exposed to moist air. A corrugated iron building can hardly be classified as a permanent structure, and cannot be recommended for modern power plant practice. There are several methods in use for the design of structural steel buildings. In one, the frame is self-supporting, and the light curtain walls are partly supported by the steelwork; while on the other hand, the structural steel, as well as the walls, is entirely self-supporting. Columns. The building columns should be of the open type as much as possible (see Figs. 5 to 7). The use of box girders and columns should be avoided, because ~r T r i FIGS. 5 to 7. Typical Columns. they are usually built up of channels, I-beams and plates, and are more apt to corrode inside than out, as they cannot be painted inside; to overcome corrosion, they may be filled with concrete. Column Bases. The columns are preferably designed with a base of sufficient area to permit of their being set directly upon the foundation, but cast-iron base 124 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. plates are sometimes interposed at this point, as they can be leveled up before the column is erected. Grillages are undesirable, but cannot be avoided with heavy column loads. Floors. In hydraulic plants, little use is made of floorbeams as far as the generating room is concerned; the whole substructure is made of solid concrete. However, in the design of the floors in the switching and transformer rooms, con- siderable structural steel is used. In many cases, the floors are subjected to concen- trated local loads at various points, which require special treatment. In other cases a railroad spur extends into the building, to facilitate the handling of heavy material by an overhead crane. The load for which the floors must be designed, is the weight of the heaviest pieces of machinery placed on them. It must be borne in mind, that ' ' , ', ' - ///' -'\/ ''/** i.'*.'.*' / / '///s'LSsL- ////////. FIG. 8. Crane Column. during construction, large quantities of material are apt to be piled on the floor, consequently precautions must be taken. As the weights of the various pieces of machinery, such as generators, transformers, etc., change so very greatly with their functions, this data must be obtained from the manufacturer. In countries where only the lowest grade of labor can be obtained, and conditions do not warrant the sending of an erecting force, but only a foreman, the steel sections may be bolted together and filled instead of being riveted. Experience has shown in some cases, that abutting pieces had to be provided with dowel pins to facilitate erection. The use of floor arches causes a lateral thrust against all of the beams composing the floor system; for this reason it is necessary to introduce tie rods, suitably spaced, to take care of this stress. These tie rods should always be placed high enough so that they will be hidden by the floor arches, as this adds greatly to the appearance of the ceilings. In some plants this detail has been neglected, and the result, to say the least, is unsightly. Another small point, is the provision of curb angles around all hatches and other openings in the floors; these angles should project from 2 to 3 inches above the finished floor level, their purpose being, to prevent wash water, sweepings, etc., going down to the floor below. The value of these curbs is more apparent in those cases where machinery is located on the lower floors, or under the galleries, which would be liable to damage from anything dropping on to them. Expansion Joints. In very long buildings, the expansion due to changes of tem- perature must be taken care of during erection, but such precautions are not required POWER PLANT. 125 in small buildings. In buildings under 300 feet in length, temperature variations do not cause much trouble, and no special precautions are required to care for them. In some cases, where expansion joints are used, it is specified, that after the building has been walled in, the joints shall be blocked with lead to prevent any motion of the steelwork cracking the concrete flooring, etc. The necessity of these joints is only during the erection period, when longitudinal expansion is very apt to make it difficult to erect portions of the steelwork. Fiber Stresses. Steel structures are proportioned, in regard to the sections used, by a limit set on the fiber stress in tension, which is reduced, for compression members, usually by Gordon's formula. In many localities, the limiting unit stresses are specified in the building laws, these in some cases being limited in their application, to some particular city; in other cases, they apply to a state or nation; the legal requirements differ greatly in different localities, hence it is advisable to investigate the subject unless the requirements are well known. In practice, the fiber or unit stresses for steel in tension vary from 13,500 to 20,000 pounds per square inch; for most of the important structures, the working stresses have been kept between 15,000 and 16,000 pounds per square inch. Character of Steel. A large portion of the structural steel manufactured in the United States is made under the "Manufacturers' Standard Specifications" as revised to Feb. 6, 1903, which permit the use of either open-hearth (Siemens- Martin) or Bessemer steel (the Bessemer steel produced in the United States is made by the acid process, no basic Bessemer steel being produced). The practice of specifying open-hearth steel exclusively, for most structures, is growing, owing to the fact that it is more homogeneous in its physical properties. Bessemer steel, on the contrary, is liable to fail in service, in an irregular and inexplicable manner, and for this reason it is not desirable for structural work. All steel made by the open-hearth process must be of uniform quality, tough and ductile. The phosphorus must not exceed 0.08 per cent. Rivet steel must have an ultimate tensile strength of from 45,000 to 55,000 pounds per square inch. Structural steel must have an ultimate tensile strength of from 55,000 to 65,000 pounds per square inch. The elastic limit must not be less than one-half of the ultimate tensile strength. The percentage of elongation must be equal to: 1,400,000 Ultimate strength in pounds per square inch Rivet steel, before or after heating to a light yellow heat and quenching in cold water, must stand bending 180 degrees flat on itself, without signs of fracture. Structural steel, before or after heating to a light cherry-red heat and quenching in cold water, must stand bending 180 degrees, to a curve whose diameter does not exceed the thickness of the sample, without signs of fracture. The finished bar plate and shapes must be free from all cracks, flaws, seams, blisters and all other defects; it must have a smooth surface and be well straightened at the mill before shipment. The tensile strength, limit of elasticity and ductility must be determined from 126 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. standard test pieces, of at least one-half square inch sectional area, cut from the finished material; two opposite sides of the test piece must be the rolled surface, the other two opposite surfaces to be milled or planed parallel; rivet rounds, however, must be tested of the full size, as rolled. All test pieces must show a fracture of a uniform fine-grained, silky appearance, of a bluish gray or "dove'' color, and must be entirely free from granular, brilliant and black specks of a fiery luster. Every finished piece must be clearly stamped with the melt numbers. The inspection of the steel, to insure its compliance with the specifications, necessarily takes place at the mill. It is common to introduce a clause in the specifications, by which if any material accepted at the mill, when under the punches or shears, shows that it is not of uniform quality, it may be rejected at the shops. In some cases a drifting test is called for, by which a hole punched in a plate or piece, the thickness of the material in some cases being specified, can be drifted to a larger diameter, without cracking either the edges of the hole or the external edge of the piece, the increase in the diameter of the hole ranging from one-third to one-half the original. The distance from the center of the hole to the edge of the piece may be specified. Workmanship. The following, in reference to workmanship, is based on the standard practice of some of the leading concerns. All material must be punched one-sixteenth of an inch larger than the nominal size of the rivets, except that material five-eighths of an inch thick and over, must be drilled or subpunched and reamed one-eighth of an inch larger in diameter, so as to remove all sheared or burred edges. (In some cases subpunching is insisted upon when more than one cover plate is used on columns or girders, in which case the reaming must be done after the parts are assembled and clamped together.) All work must match so accurately, that after assembling, the rivets can be entered without drifting. Whenever possible, all rivets should be machine driven by direct acting machines, operated by compressed air, steam or hydraulic pressure, which should be capable of retaining the applied pressure after the upsetting has been completed. Field riveting should be done, preferably, by long-stroke pneumatic riveters. Hand riveting should not be permitted for rivets over seven-eighths of an inch in diameter. The details must be designed to avoid riveting in difficult or inaccessible places. No bolts should be used, except by permission; they must be turned to a driving fit, and the bolt holes drilled and reamed after the parts are assembled and clamped together. In many cases, however, the roof purlins are bolted with ordinary black bolts, all other connections being riveted. The abutting surfaces of compression members must be truly faced to an even bearing. (In some cases this clause is extended to cover the tops of column bed plates in a specific manner, and in some rare instances it is specified that the abutting ends of tension members must be faced.) All rivets, when heated and ready for driving, must be clean. When driven, they must completely fill the hole and have round concentric heads of uniform size, thoroughly pinching the connected pieces. POWER PLANT. 127 Inspection. All facilities for the inspection, testing of material and workmanship, must be furnished by the contractor to duly appointed inspectors, but the inspection for the raw materials must be made at the mills or foundries where the steel is rolled or the castings made. The inspectors must be allowed free access to all portions of the plant in which any portion of the material is made. Painting. In regard to painting, there are a number of differing requirements, such as, raw and boiled linseed oil, iron ore or iron oxide paint, red lead paint, graphite paint, etc., and there are a number of proprietary mixtures on the market of more or less value. The proportion of the materials to be used in preparing the paint, and the kind of brushes to be used in applying it, are sometimes enumerated. The proportion of red lead used, varies from 16 to 40 pounds per gallon of oil, depending upon the quality; a paint containing 25 pounds of red lead per gallon of oil makes a very satisfactory coating for steel, the following formula being a very good mixture: 25 pounds of pure red lead, i gallon of pure raw linseed oil, | pint of japan, free from benzine. Iron ore or oxide paints possess the merit of being cheap, and for this reason are much used. They are not reliable, and should be avoided in good practice. Boiled linseed oil without a pigment makes a good coating for iron or steel. The pigment addition acts as a filler for the pores in the oil, and retards its drying or oxidization, and for this reason driers are used, japan being one of the best materials for this purpose, provided it is free from benzine. The use of benzine, sometimes called gasolene or naphtha, must not be permitted in any paint which is to be applied to ironwork, for the rapid evaporation of the benzine will cool the material to a point where the surface to be painted will be covered with a dew or moisture. At least 48 hours must elapse between the application of each coat of paint. Painting should not be permitted during freezing or wet weather. The writer would be inclined to specify that painting should only be permitted on clear days, when the temperature was above 40 F. In riveted work, all surfaces coming in contact must be painted, before assembling, with one coat of paint on each surface. Occasionally it is specified that all portions of the work to be embedded in concrete or brickwork must receive one or two coats of asphaltum varnish. All the work must receive at least one coat of paint before it is shipped; and after erection, all places where the paint has been rubbed off, and the heads of the field rivets, must be painted, after which the entire structure must receive two coats of paint. There is very little agreement in regard to the best coating for any particular case, probably because so much depends upon the preparation of the surface to receive the paint, the care with which it is applied, and the exposure conditions. All dust and loose scale must be removed before the paint is applied, and the painter should follow immediately after the cleaner. Prevention of Electrolysis. At various times it has been proposed to insulate the steel frames of power houses, with the idea of preventing electrolytic action. The 128 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. complete insulation of the frame is impractical, owing to the fact that a number of pipes must be supported by hangers bolted, or otherwise secured, to the framing; some of these pipes being in connection, electrically, with the ground water, an attempt at insulation is extremely liable to localize the electrolytic action at a few points, which would be worse than the troubles arising from the entire omission of insulation. At the site of erection, or adjacent thereto, it is usually necessary to store portions of the structural material, after it is unloaded and until it is required for erection. This material must be laid on skids, so that it does not come in contact with the ground and must be kept clean. CHAPTER VI. MECHANICAL EQUIPMENT. TURBINES. Classification. As regards the behavior of the water, turbines may be divided into two general types, the reaction and impulse. In the former, the flow of water must be continuous in all parts of the turbine, that is, the entire runner is under water; in the impulse type, the water impinges on parts of the wheel, and in nearly all cases the atmospheric air has free access to the remainder of the runner. Turbines may be further divided as regards their construction, into radial, axial or parallel flow and combined or mixed flow. In the radial type, the water passes through the wheel, either inward or outward at right angles to the axis of rotation. In the axial turbine, the general direction is parallel to the axis of rotation. In the mixed flow turbine, the water enters radially and discharges axially, or vice versa. The different types of reaction turbines are commonly known by the names of their inventors, as the Fourneyron, which is a radial outward flow; the Francis, a radial inward flow; and the Jonval, a parallel flow. A combination of the Jonval and Francis is known as the American type, and is to-day the most common one used in America, where it had its origin. Of the impulse type, the Girad and the Zuppinger in Europe and the Pelton in America are the most familiar. In regard to the origin of the impulse wheel, it might be of interest to state, that Zuppinger in 1846 installed his first tangential wheel at Weiler's Mills near Friedrichshafen on Lake Constance. 1 The same engineer, who was at the time connected with the Escher Wyss Company, built in 1868-69 for the Haemmerle Cotton Mill in Dornbirn, Voral Mountains, Germany, a tangential turbine of 220 H.P., making 300 R.P.M. under a head of 550 feet. 2 This tangential impulse wheel was 5 feet in diameter, 30 inches wide, and mounted upon a vertical shaft and provided with two diametri- cally opposite jets or nozzles; it was designed for a water supply varying from one to six cubic feet per second, and as the designer up to that time had not constructed wheels to operate with over 380 feet head, he thought it advisable not to give a guarantee of more than 65 per cent. However, during actual operation, with a water consumption of 1.5 to 2 cubic feet per second, the efficiency was 70 to 75 per cent; with a water consumption of about 5 cubic feet per second, was 65 to 70 per cent. Turbines are also classified as follows: I. Low head up to 30 feet. II. Medium head from 30 to 200 feet. III. High head above 200 feet. 1 Letter by Prof. Escher, Zeitschrift des Vereines deutscher Ingenieure, Feb. 18, 1905. 3 Grosse moderns Turbinenanlagen. L. Zodel, Schweitzerische-Bauzeitung, June 13, 1908. . 129 130 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. This classification is not strictly adhered to, as many manufacturers and plant designers are ignorant of the fact that a low head turbine is less efficient when applied to a high head, and vice versa. Reputable manufacturers with competent engineering staffs will advise the use of such a turbine as is most suitable for the condition at hand. Slight variations of the above are sometimes made, when other conditions favor the same. A rigid classification by European engineers is as follows: 1 Low head turbine, up to about 10 feet. (With open flumes, vertical shafts with bevel gearing. When above ground water, horizontal shaft with belt or rope drive.) First intermediate head turbine, from 10 feet to about 35 feet. (Open penstock, which is possible up to 35 feet; vertical, or when advisable, horizontal shaft.) Second intermediate head turbine, from 35 to about 165 feet. (Closed penstock, spiral casing, horizontal shaft. Of course for special conditions vertical shafts may be used.) High head turbine, above 165 feet. (Closed penstock, spiral casing, horizontal shaft, as long as reaction wheels are considered; otherwise, impulse wheels with horizontal or vertical shaft.) As this close classification is seldom applied to American practice, a more liberal classification must be made. FIG. i. American Turbine as designed by the Dayton Globe Iron Works. Low Head Turbine. In America, the low head turbine, a combination of Francis and Jonval type, is manufactured in the horizontal and vertical type, and is known as the American turbine. Frequently a number of runners are mounted on a single shaft, as seen in Fig. i. In many cases they are placed in an open flume. The runner 1 Hiitte. Vol. I, p. 802. Edition, 18. MECHANICAL EQUIPMENT. 131 of this type of turbine was previously made of steel buckets riveted or bolted to the frames. To-day, most manufacturers make the runner in one solid casting. Fig. 2 shows such a runner. The regulating mechanism of the American turbine, as manufactured by the Dayton Globe Iron Works, is shown in Fig. 3. The ring C which actuates the guides D controlling the water supply is governed through sector E. Other American low head turbines and application of same will be found throughout the text. FIG. 2. Runner of "Amer- FIG. 3. Crown Plate and Gate ican" Turbine. of "American" Turbine. Medium Head Turbines. As the line of demarcation between low and medium, and medium and high head, is not distinctly drawn, one will find under this head a great variation of turbines, including high and low head types. The majority of turbines used under medium head are of the Francis type. It is not the purpose of this book to go into details of the design of turbines; only the typical features will be given. The Francis turbines are built in either the horizontal or vertical type, with one or more runners mounted on a single shaft. These turbines are placed, either in an open chamber of the power house, or inclosed chamber made of cast iron or structural steel. Figs. 5 and 6 show a vertical Francis turbine as built by J. M. Voith, Heidenheim, Germany, for the Kykkelsrud plant, Norway. It operates under a head of 52.5 feet to 62.5 feet, with a water consumption of 670 to 530 cubic feet per second, and with 150 R.P.M., develops 3000- HP. It will be noticed that the turbine casing is spiral in plan and rectangular in elevation; it is made of structural steel. The water enters the turbine casing with a velocity of 9 feet per second, and gradually increases to 20 feet, and discharges with a velocity of 3.9 feet per second. The runner has a diameter of 5.9 feet, and is mounted on a 1 2-inch vertical shaft, 25 feet long, on the end of which is coupled the shaft of the generator. Between the turbine and generator is a thrust bearing, supplied with oil at 220 pounds pressure per square inch, to take up the weight of the revolving element which is 32 tons. The water supply in the turbine is controlled by clam-shell gates. A turbine of same make and pattern as the above will be found in the plant of the Ontario Power Company, with the exception that two turbines are mounted upon a horizontal shaft (see Fig. 7). 132 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. -a c 'rt CJ 5 r Ban JU..\.KeptoyU FIG. 7. Replogle Governor. FIG. 8. Pressure Regulator (Bell & Co.). MECHANICAL EQUIPMENT. A.E.&M, UN'IV. OF CA position during the act of opening the valves. Therefore the speed is in reality- lower after the new power is added than it was before the change in load. It is now clear that there is a continuous dropping in the speed while the valves are opening. In practice this permanent drop is enough to insure the correct division of load. It is also enough to permit of successful government where adequate power storage exists in the unit to be governed. In this governor there is no special provision for temporary relay. Such provision is unnecessary except where the momentum effects are small. (In the governor shown the permanent drop can be varied by the pitch of the thread used at X.) In ordinary practice it is about 2 per cent." Pelton Nozzle Regulation. The usual method of controlling the speed of the Pelton wheel is by means of a deflecting nozzle, needle nozzle, or a combination of both. Which one of these types is most suitable depends on the condition of the head, power and character of load. The deflecting nozzle is a cast-iron nozzle provided with a ball and socket joint, permitting of its being raised or lowered, thus throwing the stream on or off the buckets. The power of the wheel is consequently increased or diminished, according to the change of load, and a constant speed is maintained. A steel deflecting plate, which deflects the stream itself, the nozzle remaining stationary, is sometimes used to accomplish the same results when the design will not admit of a deflecting nozzle. FIG. 9. Instantaneous photograph of Tangential Wheel fitted with Pelton Buckets when running at high efficiency, showing the discharge from the sides of the buckets parallel with the entering jet; the photograph also shows clearly that the front of the Pelton Bucket enters the stream without shock or disturbance of any kind and that all of the energy is removed from the water by the bucket. The needle nozzle consists of a nozzle body in which is inserted a concentric tapered needle. A change of position of this needle produces a corresponding change of discharge area of the nozzle. The amount of water used is thus varied and the power of the wheel influenced proportionately. The needle and deflecting nozzle is a most valuable combination*, consisting of a deflecting nozzle, with which is incorporated a needle nozzle with means for operating 152 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. either the needle or deflecting nozzle simultaneously or separately. The deflecting nozzle in itself is a most sensitive means of regulation when actuated by an auto- matic governor, but does not save water. On the other hand, the needle nozzle, while it is extremely economical in the use of water, is difficult to control quickly by means of the governor. The operation of the combination is as follows: Assuming the full load to be on the water wheel, and the nozzle in position of greatest efficiency, a decrease of load will cause the nozzle to be suddenly deflected by the automatic governor. Simultaneously the needle portion of the nozzle will be actuated by hand, or by another automatic device, tending to gradually close the needle and decrease the flow. The governor then raises the nozzle to accommodate the decreased flow of water (and consequent decrease of power), and the nozzle is then brought back to the position of greatest efficiency, having, at the same time, FIG. 10. Automatic Needle and Deflecting Nozzle, Pelton Impulse Wheel. controlled the speed within the required limits. Such a device is essential where water is valuable and where economy is necessary to carry over the peak load. The needle portion need not necessarily be operated by an automatic device, but may be controlled by hand, and the same results obtained, although necessarily in a longer period of time. An installation of this combination is given in Fig. 10. The lower end of penstocks, particularly of high-head plants, must be provided with reb'ef valves, as already discussed under penstocks. Fig. n shows a battery of relief valves as employed by the Pelton Water Wheel Company in connection with their impulse wheels. They may be installed either singly or in a battery, which depends on the size of the penstock and the working head. These valves are set to operate at a pressure slightly greater than the normal, and in the event of the water flow being suddenly checked by the closing of the gate or operation of the governors the safety valves momentarily open and relieve the pressure, thus guarding the penstock against the possibility of water hammer. Accessories. For operating hydraulic governors either by water or oil pressure, additional auxiliaries such as oil pumps, pressure accumulators, and water filters MECHANICAL EQUIPMENT. 153 are necessary. Particular care must be taken, if the penstock water is used in hydraulic governors, to clean same, which is done by sending the water through a screen chamber. There must be at least two screens, so that one may be in use when the other is being cleaned. The pressure oil for the relays or pilot valves of the governors is usually supplied by motor-driven plunger pumps. As this oil is also used in the step and thrust bearings and frequently must be under high pressure, the pressure to the governor FIG. ii. Battery of Relief Valves. FIG. 12. Combination of Fly- Wheel and Flexible Leather Link Coupling. must be lowered by reducing valves. To insure continuity of operation two or more pumps must be installed. In connection with these, accumulators are installed to take up fluctuations in the pressure. Couplings. The turbines may be rigidly or flexibly coupled to the generators. The rigid coupling is used where there is little fluctuation on either the hydraulic or the electrical end of the plant. The flexible couplings serve two purposes: first, to take up light speed variations; second, in most cases it insulates the turbine from the generator, as the actual connection between the turbine and generator is done by means of leather or rubber. A coupling very much used in Switzerland is the Zodel. It consists of two concentric cylindrical flanges provided with slots, through which a belt is wound in and out. Frequently these couplings are so designed as to act as a fly wheel to balance fluctuations of load. A coupling designed on this principle is shown in Fig. 12. 154 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. OILING SYSTEM. Oil Required. It is of vital importance to install an oiling system in all power plants, large as well as small. A complete oiling system collects the oil from the bearings, filters it and returns it to the machine, all of which is done automatically. From 50 to 70 per cent of oil used in power plants is wasted if means are not provided to collect same. Filtering Tanks. The filtering tank must be so located that the oil will flow to it by gravity. The tanks must be installed in a fireproof compartment. This AUTOMATIC WATER SEPARATING! APPARATUS TO SEWER FIG. i. Burt Oil Filter. compartment may also contain the oil pumps as well as the waste cleaner and drier. The door must be so arranged that it will shut automatically. If the room is large, it is better to install two doors, one as a means of easy escape for the attendant. The floor must be provided with proper drainage, as it is frequently necessary to clean the tanks and filters. Many of the larger power plants have filtering tanks of special design, but common practice is to install some regularly manufactured article. The tanks must be in duplicate, or so arranged in compartments that one may be cleaned at MECHANICAL EQUIPMENT. 155 a time without putting the entire tank out of service. Large tanks may be con- structed of many compartments. The oil, entering through cheese cloth or light canvas filters, passes through the compartments at a low velocity, precipitating any foreign substances. Section. 1 {Action 2 Section 3 Section 4 FIG. 2. Turner Oil Filter. 4 Suction Bj Pal. OOOOOOO OOOOOOO ooooooo OOOOOOO OOOOOOO oooo&oo OOOOX3OO OOOOOOO ooooooo 0006000 ooo@ooo OOOOOOO Cf&O O O O O ilTO O O O O on'ooooo irooooooo fuuuuuuu uuuuuuu uuuuuuin FIG. 3. Oil Filtering Tank for Large Capacities. The filtering tanks may have a heating coil to heat the oil, thereby increasing the speed of filtration and causing more rapid precipitation. When, however, high speed turbines are used, and the temperature of the oil returned to the filters is high, the use of the coil may be dispensed with. 156 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Very frequently the oil returned contains a certain amount of water. It is important to abstract this water. Fig. i shows a typical oil filter of this type as manufactured by the Burt Manufacturing Company. The oil entering at the top passes through the waste contained in the center chamber, from where it passes downward through the pipe C, is heated by the coil and flows upward through the water contained in the lower portion of the tank. This water forces the oil through the waste F into the pure oil compartment, from which it is drawn off and reused. The water is dis- charged to the sewer through the automatic water separator, shown on the left-hand side of the cut. Another very efficient oil filter is shown in Fig. 2, representing the Turner system. As will be seen, this tank is divided up into four sections. The oil passes through the filtering material of each section, having its temperature raised by coils in the first two sections. A very efficient oil filtering tank is shown in Fig. 3. The tank is divided into chambers by partition walls extending alternately to the top and bottom of the tank, giving the oil an up and down flow, thus increasing precipi- tation, which will be greater the lower the velocity. The oil before entering the tank passes through Canton flannel bags, arranged in trays as shown in the illus- tration. These bags are removable, and when dirty, may be replaced by clean ones. The pipe connections are such that any chamber may be separately cleaned without shutting down the entire filter. Oil Pumps. The pumps required for an oiling system are either high or low pressure. The latter are used with a central oiling system. Duplicate pumps must be installed in order to keep one in reserve. With certain turbines high-pressure pumps are required to pump the oil into the step bearing. It is better practice to install several small-size pumps than one or two large ones, as the possibility of shut-down is thereby lessened. With the vertical turbine in some instances water is used for the step bearing, with practically the same results as those obtained with the use of oil. The entire equipment, with the exception of the filtering tanks, is the same as the oiling system. Supply Tanks. Frequently it is necessary to install one or two elevated supply tanks, from which the oil is fed by gravity to the various bearings. These tanks must be properly vented, and where more than one tank is employed, they must be interconnected. In order to avoid complicated and long pipe mains, these tanks are preferably placed somewhere in the center of the plant. As the oil is used over and over again, and its temperature is increased each time it is used (especially with high-speed turbines), it might be necessary to cool the oil by means of water coils placed in the supply tank, before it returns to the bearings. Oil Piping. The return pipes leading the oil from the various bearings or collecting pans to the filtering tank may be of either wrought or cast iron. The former is preferable, however, for small pipes. If wrought iron is employed screw fittings may be used. In order to secure a good gravity flow for the oil the pipes should be pitched at least one inch in every ten feet. Where many returns are connected to one common header, provision has to be made for the removal of air. This is accomplished by placing one-half inch or MECHANICAL EQUIPMENT. 157 three-quarter inch vent pipes on the header. These vents must extend above the highest point in the return piping, so that, if the pipe discharging to the filter becomes plugged, the oil will not escape through the vents. To facilitate cleaning the pipe, it is good practice to install crosses instead of tees in the header, one leg being plugged. The supply pipes from the filter to the elevated tank, and also the pipe from the tank to the machines, must preferably be made of brass or copper. This is absolutely necessary, as steel, wrought-iron, or cast-iron pipe contains a scale which oil loosens, and if this scale gets into the bearings it is liable to cause considerable damage. Galvanized iron pipe has been tried for supply piping, but experience has shown that the galvanizing will wear off and the pipe will scale as badly as a black iron pipe. It is essential to keep the pressure constant in high-pressure oiling systems. This may be accomplished by accumulators. TESTING TURBINES. European Methods. It is difficult to keep the load and revolutions of a turbine steady for long periods, to secure data for figuring the exact water consumption. It is therefore essential to devise a system whereby the flow of water is indicated simultaneously with the load and revolution of the turbine. This is best accomplished by automatic graphical methods, registering the load, revolutions, water levels in head and tail race, water discharged, and time. A device of this kind (Reichel and Fuess system) is seen in Fig. i. It has been used for some years in Germany, and consists of a vertically revolving drum, with six different recording indicators for the different readings. The drum, by means of worms and gears, is actuated by a 220 V. \f HP. motor, making 2750 R.P.M., and the speed of the drum can be varied at will between 0.6 mm. and 15 mm. per second. The drum itself can be set in four different positions on the vertical shaft, so that four complete tests can be recorded on the same sheet. It will be observed that there is a clock connected with the recording mechanism, cutting in and out the four relays for the four lower indicators. The two upper indicators are attached to the wires running to the floats, one for the headrace and the other for the tailrace. The load on the turbine is measured by a Prony brake, and indicated on the recording device. The discharge of the tailrace, measured by current meters, is also recorded. It is usually difficult to measure the exact discharge of the tailrace, as it varies greatly according to the proximity of the channel walls, and, as an exact average flow throughout the channel can hardly be ascertained, because a constant turbine load is only of short duration, therefore it is well to install a number of current meters. This may be done by having three or five meters, according to the depth of the tail- race channel, on a vertical shaft secured to a carriage, which is moved across the channel to four or six points, depending on the width of same. The carriage must move easily, the rollers resting on an I-beam or channel iron, 158 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. so that successive measurements can be taken very rapidly by running the carriage along to different points. Practice shows that by having three current meters on one shaft, and moving the carriage in five different positions, fifteen readings of the tailrace cross section can be made in five minutes. For measuring the discharge of the turbine in a simpler and perhaps the most accurate way, a method has been long in vogue in Norway and Sweden and recently FIG. i. Automatic, Graphical Registrator for Testing Turbines. introduced into Germany. It was developed by Prof. Erik Anderson, Stockholm. To make use of this method, the tailrace must be some 30 feet to 40 feet long and have a uniform cross section with smooth surfaces. A carriage, preferably made of aluminum or light steel bicycle tubes, rests on a smooth track, preferably on the planed legs of angle iron. To the carriage is hinged a light framework of wood or steel, of the width of the tailrace, giving on each side a clearance of about a quarter MECHANICAL EQUIPMENT. 159 to three-eighths of an inch. This frame is covered with oiled cloth or other water- proof canvas. The total weight of those as illustrated in Fig. 2 is about 80 pounds, and takes about 0.8 pound to move same. Fig. 3 shows the general arrangement of a testing plant at Heidenheim. At point / the curtain is lowered and soon assumes a vertical position before entering the area of measurement. Point /// shows the carriage in a position, with the FIG. 2. Carriage with Curtain for Testing the Water Discharge of Turbines. FIG. 3. Arrangement for Measuring Tailrace Water in Testing Turbine by the Curtain Method. curtain released from the vertical position by means of a trip device; the carriage is then drawn back for another run. Practice shows that every four minutes a com- plete test can be recorded. The speed of movement depends on the exact uniform water velocity throughout the tailrace channel. It must here be stated that with a water velocity of less than 0.5 foot per second the measurements become inaccurate. The position and time of travel are recorded by electrical contacts placed some three to five feet apart. i6o HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The difference in the water level at both ends of the tailrace varies between one and two millimeters; the average is taken as final. Such tests are not made on windy days, because the outside water is swept into the tailrace, and the force of same is oftentimes sufficient to reverse a current meter. FIGS. 4 and 5. Plan and Section of Holyoke Testing Flume, showing Turbine in Position for Test. Holyoke Tests. Most of the low-head turbines manufactured in America are tested at the flume of the Holyoke Water Power Company, Holyoke, Mass. As the head on this plant is only 18 feet, and is seldom constant, due to great fluctuations, all the readings therefore have to be reduced to a uniform head. Further, as the MECHANICAL EQUIPMENT. 161 conditions of the flume and the setting of the wheel are different from those at the plant where the wheel is to be installed, the value of the Holyoke tests may be judged from the following comparison. Due to the high efficiency claims of some American manufacturer, a German concern intended to build turbines after the American type, for which purpose it bought a i6-inch turbine, duly tested at Holyoke, then tested in Germany by Professor Pfarr, one of the highest German authorities on turbines. The results of these tests are published in the ZeUschrift des Vereines deutcher Ingenieure, June 7, 1902. The comparison of the efficiencies is as follows: Discharge I. O o. o 0.8 O. 7 0.6 o. t; O 4. O 1 Holyoke test 0.81 0. 70? o. 76? o. 72? o. 67 German test 0.718 0.703 o 693 0.658 0.591 0.491 0-358 O. 121 The discharge is the actual discharge and is not figured on the proportional gate opening. During the Holyoke tests the total weight on the step bearing was 25 per cent greater than that in the German tests. For these reasons, guarantees of such tests must not be accepted by the power plant designer; he should accept only such guarantees as are made in the power plant itself. However, as the Holyoke tests are used in many respects as a standard in American practice, a brief description of the Holyoke Test Flume method of testing and deduction as given by the Dayton Globe Iron Works Company is given below. For the purpose of making the necessary experiments on the wheels, the Holyoke Water Power Company built a permanent testing flume, in which the wheels are tested both for power and for amount of water discharged. They are usually tested at five or six different openings of the gate, ranging from full open to the opening at which the discharge is one-half that at full opening, and at six or eight different velocities of revolution at each gate opening, and making some thirty to fifty experiments on each wheel. The final result is that for all practical purposes the water wheel is converted into a water meter, and its discharge may be known under any of the conditions under which it will have to run. Besides this, its efficiency or value as a motor is also known. The essential portion of the testing flume consists, in the main, of the trunk or penstock M, bring- ing the water into the wheelpit D and the tailrace E. In the passageway M are placed two sets of racks, or baffleboards, to stop eddies and oscillations in the flowing water. Baffleboards are also placed in the tailrace for the same purpose Flume wheels are set in the center of the floor of D, and D is filled with water. They discharge through the floor of D and out of the three culverts N, N, N into the tailrace E. At the downstream end of this tailrace is the measuring weir O, the crest being formed of a piece of planed wrought iron. It can be used with or without end contractions. The depth of water on the weir is measured by a hook gauge, in a cylinder P, set in a recess Q fashioned into the sides of the tailrace. These recesses are water tight, and the observer is thus enabled to stand with the water level about breast high, or at a convenient height for accurate observation. The methods of measuring water over this weir are those described in Lowell Hydraulic Experiments, by James B. Francis. A platform R surrounds the tailrace, and is suspended from the iron beams that roof it in. The wheels to be tested are lifted from the wagon or cars by a traveling windlass, and run into the building 1 62 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. and lowered into the wheel pit D Winding stairs S lead into a passageway that leads in turn to the platform R. In the well hole of these stairs is set up the glass tube X, which measures the head of water upon the wheel. It is connected with the pit D by means of pipe running through a cast-iron pipe T, built into the masonry dam which forms the downstream end of the wheel pit D. The power is weighed by a Prony brake, consisting of a cast-iron pulley surrounded by a wood-lined jacket, cooled and lubricated by water from the city mains, or, if necessary, with the addition of a small stream of soap-water. The pull of the jacket is weighed by a bent lever and weights, the friction being regulated by an attendant at the temper screw, so that the weights are kept balanced. To enable the observer at the brake wheel, the one at the head gauge and the one at the measuring weir, to take simultaneous observations at intervals of one minute, an electric clock is set up, which rings three bells simultaneously at intervals of one minute, or of half a minute if desired. The whole structure is built in a durable and efficient manner. The pits and tailrace are all lined with brick laid in cement. The stone masonry was intended, by careful work and grouting, to be water tight without the brick lining, and the brick lining was then carefully laid up with joints full of mortar, as an extra precaution. As a consequence, the front of the wall forming the downstream side of the pit D is built so tight that an exact measure- ment of the leakage of the wheel gate could be made if desired. An approximate estimate is readily made by filling the pit before the tailrace is allowed to fill up and apportioning the total measured leakage of the wheelgate and that of the flume. W shows a waste pipe. Another not shown serves to draw the water out of and through the floor of the pit D. To close or open these waste pipes they are fitted with cases of small water wheels, which thus form convenient valves for the purpose indicated. The pipe W leads into a sewer on the other side of the second level canal and thence into the river. It enables the tailrace to be emptied of water down to within some three inches of the bottom plank- ing. After the wheel to be tested is placed in the flume, and the dynamometer placed on the shaft, the lever is adjusted, care being taken that it is horizontal and tangent to the circumference of the brake at point of application of the pull. It is balanced by placing a small weight first on one side and then on the other of the fulcrum, and at equal distances from it, and noting the time necessary to move the long arm a certain distance above and below the center, and then adjustment of the counterpoise until the times become equal. A dashpot is always used with the lever to steady its oscillations. An indicator is attached to the turbine gate or to the mechanism controlling it, so that the position of the gate is always known. The hook gauge is set by an engineer's level so that point of hook is level with crest of weir when scale on the gauge reads zero. The length of the weir is adjusted to the proper length for the quantity of water to be measured. The floating gauge, by which head on wheel is measured, is adjusted so that the zero of its scale is at the level of tail water. In the system followed there are three observers, each taking a reading of his gauge every minute and keeping a separate set of notes. The notes from which the theoretical or gross power of the water is computed are kept by the men at the head gauge and hook gauge, and consist of a simple measure- ment of the head, or vertical distance from surface of water in flume over the wheel to surface of tail water, the length of the measuring weir, the number of its end contractions, depth of water flowing over the weir, and temperature of water. All data for the effective power of the wheel are taken by the third observer, and consist of the circumference of the brake at point of application of the weight, ratio of the lever arms, number of pounds on the lever, revolutions of the wheel per minute, and setting of the wheelgate. Of the data, all excepting setting of the gate, weight on lever, revolutions, head and depth on the weir, are generally constant throughout one wheel test. The variable data are compared with each other, and for any one experiment consecutive readings are selected where everything goes to show that revolutions, head, and depth on the weir are steady and consistent with each other. These readings in each notebook are then averaged, and these averages compose the variable data for the experiments, a full set for each change of weight on the lever. MECHANICAL EQUIPMENT. 163 The quantity of water passing the weir is computed by the Francis formula: Q = 3.33 (L . i nh) h\, in which Q = quantity in cubic feet per second. L length of weir in feet. n = number of end contractions. h = depth on the weir. If the volume of water renders it necessary, Q is corrected for velocity of approach. Q is then diminished by the leak of flume floor, and result is the net quantity of water passing the wheel. Theoretical power of water in horsepower is q X H X 60 X Wt HP. (water) = - , 33,000 in which q = cubic feet per second passing the wheel. H = head on wheel, in feet. Wt = weight in pounds of one cubic foot of water, according to temperature. Effective power of wheel is W X R X I X c HP. (wheel) = - , 33,000 in which W = weight in pounds on lever-arm. R - - revolutions per minute. I = ratio of lever-arms. c = circumference of brake. HP. (wheel) Efficiency of wheel = HP. (water) On account of fluctuations in height of the canal from which water is drawn for testing, and on account of varying depth of water over the weir, the head on wheel is not constant throughout the test, so that the discharges at various gates and speeds cannot be directly compared, but must first be reduced to a uniform head, H', by the rule Jir = q \H' The discharge at full gate and at the speed of revolution giving the maximum efficiency, is taken as the unit discharge, and the discharge ( CHAPTER VII. ELECTRICAL EQUIPMENT. GENERATORS. Classification. In modern high-tension transmission systems, alternating current generators are practically exclusively used. They are wound either for single, two or three phase, and connected either in star or delta. The choice of any of the three systems depends on the character of the transmission system. The generators are classified as inductor, revolving armature and revolving field type. Inductor Generator. This type of generator derives its name from the projecting ends of the rotating element which are termed inductors. The chief advantage of this generator is that there is no rotating winding, as both field and armature wires are stationary. The rotating element is nothing but a mass of iron, consisting of a cast spider made in two or more parts, depending on the size. The rim of the spider is provided with lugs, to which the laminated pole pieces are fastened. Due to the simple construction of the revolving element, high peripheral speed may be attained without setting up excessive stresses. As seen in Fig. i, the field coil is stationary and clamped in the middle of the machine. Its winding consists of copper wire or strips, properly insulated and well ventilated. As it is not necessary to surround each individual pole piece on all four sides with copper winding, the amount of copper used in winding these field coils is less than that in revolving field generators. The armature winding usually employed is what is called the "concentrated winding," that is, there is only one slot per phase per pole. The percentage of space taken up by the insulation in this style of winding is less than that in the "distrib- uted winding," consequently more space is left for copper conductors, a factor which is of special importance in high-voltage machines. They are wound for 2300 or 6600 volts, and usually are of the two or three phase type. The pressure wave of an induction alternator can be made to closely approach a pure sine curve, a factor of great importance in long-distance lines, and also in connection with arc lamps. The efficiency for both full and partial load is high, which is partly due to the fact that the magnetization of the iron is never reversed, but merely increases from zero to maximum value and then decreasing to zero; if the iron is worked at ordinary densities, the iron losses are small. As a rule, the regulation of inductor alternators is not as close as that of the revolving field type. 167 168 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. i. Sections of an Induction Generator, General Electric Company. FIG. 2. 5000-HP., azo-volt, Umbrella type, 2-phase Alternator with Internal Stationary Armature, Niagara Falls Power Company, Plant No. 2. ELECTRICAL EQUIPMENT 169 Revolving Armature Generator. The armature in this type of alternator consists of laminated steel rings, mounted on the cast-iron rim of a wheel. The armature ring is built up of thin sheet steel punched in such a way, that when assembled, the completed armature core is pierced with slots for the reception of the winding; ventilating spaces are provided at intervals, as the armature is assembled. FIG. 3. 225-K.W., 4oo-volt, 5o-cycle, 3-phase, Flywheel Alternator, with Internal Stationary Armature, and Exciter. According to the amount of current to be carried, the winding consists of wire, straps or bars. For high-voltage alternators of small current capacity, wire wind- ing, in machine wound coils, is used. For low voltage and large current capacity strap wound windings are employed. Copper bars are used where the current in the armature is very high. The fields of large alternators are made in two or more pieces, the division being vertical or horizontal, so that the frame may be removed and the armature winding is easily accessible. The field poles are made up of thin annealed steel plates and are bolted into the field yoke. I/O HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. In small alternators, the field coils are made of wire instead of straps, used in large alternators. When strap winding is used, the strap is wound on edge. This type of alternator is not used to any extent for high-tension transmission; its field is confined to isolated or similar power plants. Revolving Field Alternator. As classified by the name, in this type of alternator, the field revolves while the armature is stationary. This method of construction facilitates the insulation of the armature winding and requires that field current instead of the armature current shall pass through the collector rings and brushes. Due to this, alternators of this type are specially adapted for high voltages for large current output. The revolving element or field consists of a wheel, upon the rim of which are mounted laminated plates, bolted together; at intervals are air spaces for ventilation. As these pole pieces are built on the circumference of the rim, the revolving element frequently serves the purpose of a flywheel, particularly in connection with low-head turbine plants where the speed is low. The field coils, according to the size of generator, are either of wire or copper strap wound on edge. When placed in position on the frame, they are securely held by wedges. The armature winding is stationary and usually external to the field (armature internal to the field and stationary is seen in Fig. 2) and carried in the frame of the machine. The winding is similar to that of the revolving armature type. The stationary parts for small machines are made in one piece, and so arranged that the whole frame can be shifted for inspection. In large machines the frame is split up into sections for inspection and repair purposes. Regulation. According to the standardization committee of the American Insti- tute of Electrical Engineers, the regulation of an apparatus intended for the genera- tion of a potential, current, speed, etc., varying in a definite manner between full and no load, is to be measured by the maximum variation of potential, current, speed and so forth, from the satisfied condition under such constant conditions of operation as give the required full load values. The regulation of an alternator is the percentage rise in voltage obtained by throwing off the entire non-inductive full load. The speed and excitation of course are constant. Good machines have a regulation of 6 per cent on non-inductive loads and 8 per cent on inductive loads with a power factor of 0.85. Where synchronous machines are connected to the transmission line, close regu- lation of the generators is very essential. As these machines run in synchronism with the generator, any sudden variation in generator voltage is transmitted to the synchronous apparatus, which cannot respond owing to the inertia effect of the rotating element. If the changes are very sudden, the synchronous machines will fall out of step and eventually stop. For slight changes, the speed of the synchronous machine will try to keep step with the generator. This action is known as " hunting." Where the alternator is subject to much fluctuation in voltage, an automatic regulator facilitates the regulation of the machine, that is, it automatically controls ELECTRICAL EQUIPMENT. 171 the exciter current to give nearly constant voltage at the generator terminals. It has this advantage: it is independent of the inherent regulation of the machine itself and gives superior results. As an alternator with high regulation is expensive, it is often cheaper to install an automatic regulator on a machine with inferior regulation. The Tirrell Regulator is used on constant-potential circuits and the Thury on con- stant-current. FIG. 4. Eleven i2oo-HP., Sooo-volt, 5o-cycle, 66-R.P.M., 3-phase, Brown, Boveri Alternators. In Front two 4OO-HP. Exciters, Beznau Plant, Switzerland. Efficiency. The efficiency of a generator is the ratio of the power output to the power input and is expressed in per cent. Some of the best machines have an efficiency as high as 98 per cent. However, average efficiencies are in the neighborhood of 96 per cent. As the efficiency of a generator depends exclusively on the design and workmanship, it is the best policy for the manufacturer to produce a machine of highest efficiency, and the power plant designer must not hesitate to use same. The efficiencies must be high, not only on a full load, but correspondingly high on frac- tional loads. It is the practice, particularly in plants supplying power for railroad purposes, to operate the generators at 50 per cent overload. The highest efficiency is usually attained at 25 per cent overload. This indicates that the generator is designed for greater capacity than actually rated by the manufacturer. European manufacturers HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. HO aaaa MOD. $0 10 v 2190 FIG. 5. Characteristic Curves of 275O-K.V.A. Generator. 9000-10,500 volts, 150 amperes, 42 cycles. E = No-load characteristic. 4 = Armature copper loss at cos < = 0.75. Jc = Short circuit characteristic. 5 = Exciter loss at cos = 0.75. n = Efficiency. 6 = Armature copper loss at cos = i.oo. 2 = Iron loss. 7 = Exciter loss at cos = i.oo. 3 = Friction and Windage loss. FIG. 6. 2750-K.V.A., 9ooo-io,5oo-volts, 42-cycle, 3I5-R.P.M., 3-phase Alternator (Oerlikon) connected to an Impulse Wheel, Caffaro Plant, Italy. ELECTRICAL EQUIPMENT. 173 rate their machines according to the capacity at highest efficiencies, and the over- load capacity is usually 25 per cent. When these concerns sell their machines to foreign countries, where 50 per cent overload capacity is required, they follow the practice in America, that is, they underrate the generator. Fig. 5 shows the charac- tistic curves of a 2750-]$.. V.A. generator, which is given in Fig. 6. When generators continuously run for 24 hours, the temperature rise of any part of the machine must not exceed from 40 to 45 C. for normal load with a power factor of 0.90 to i.oo. With the same power factor the rise in temperature on 25 per cent overload must not exceed 50 C., and with 50 per cent overload for one hour the rise must not exceed 60 C. above that of the surrounding temperature. Frequencies. The most common frequencies used are 25 and 60, and depend chiefly on the character of service; 25 is used for power purposes and 60 for lighting. However, there are exceptions where the reverse is true. The lower frequency is chosen because the iron losses in the generators are less and consequently the machine is cheaper. The higher frequency is used for lighting, as it does away with fluctuation. Synchronous machines, such as rotary converters, give much trouble on 6o-cycle lines, and in their stead motor-generator sets are substituted, as will be seen in chapter on Substations. In the last few years 15 cycles have been used for railroading, particularly in connection with single phase, and it is still being discussed in the technical press whether or not it should be adopted as a standard. 1 The principal arguments in favor of 15 cycles are given as : 1. An increase of from 30 to 40 per cent in the output of a motor of a given size, and a consequent reduction in the total number of motors required to operate a rail- way, and in the cost of equipment. 2. Better performance of the i5~cycle motors, including higher efficiency, higher power factor, and better commutation. 3. Less dead weight to be carried on cars and locomotives. 4. Lower line losses. In other countries the choice of frequencies varies greatly; for instance, one will find frequencies of 15, 25, 32.5, 42, 60, etc. Voltage. For low-tension distribution, voltages of no, 220, and 440 may be con- sidered as standard. For higher generator voltages, noo, 2200, 3300, 6600, n,ooo, and 12,000 are most frequently used in American practice. The choice of voltage depends chiefly on the system of distribution, particularly for long-distance trans- mission. When the bulk of the power is used in the vicinity, the voltage of the gen- erator must be chosen to suit the most economical distribution, that is, to reduce the use of transformers to a minimum. When the power is transmitted over a long distance, one of the above generator voltages is used and stepped up to a suitable transmission voltage. Exciters. The exciters are driven either by a separate turbine or from the shaft of the main unit. In the latter case each generator has its own exciter and is 1 Twenty-five versus Fifteen Cycles for Heavy Railways, by N. W. Storer. Am. Inst. E. E., July, 1907. 174 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. 7. Interior of Submerged Power Plant of the Patapsco Electric and Manufacturing Company. 3OO-K.W., 3-phase, 6o-cycle, n,ooo-volt Allis-Chalmers Alternators, running 240 R.P.M. FIG. 8. Interior of Sill Plant, Insbruck, Tyrol. Each Unit consists of two Impulse Wheels, Zodel Coupling, and a 2ooo-K.W., io,ooo-volt Generator with Overhanging Exciter. ELECTRICAL EQUIPMENT. 175 seldom belt-driven, but is mounted on the overhanging shaft of the generator; it is therefore dependent on the operation of the main unit, the disadvantage being, in case the speed of the main unit should drop the excitation diminishes, thus neces- sitating the installation of an automatic regulator. With a turbine-driven exciter the excitation of the generator is independent of its speed. The exciter has the same type of turbine as the main unit. Lighting of the station is frequently supplied by the exciter units. The voltage of the exciters depends much on the voltage used for lighting purposes, also for operating plant auxiliaries. In connection with the exciters, storage batteries are installed which float on the exciter busses to take care of fluctuations and peak loads. This is particularly true where the exciters are mounted on the main generator shaft. Current may be drawn from the storage batteries for operating the high-tension oil switches. When the turbine-driven exciter is employed there must be more than one, to take care of emergency cases. In small or average-size plants the exciter is of sufficient capacity to excite all of the generators at once, while the second unit is kept in reserve. In plants above say 50,000 K.W. capacity it is good policy to install several exciter units instead of two. These are so connected that they feed one common bus from which the main units are excited. The size of the exciters is from one-half to one per cent of the output of the plant, and as reserve must be provided, the combined capacity is about two per cent of that of the plant. The voltage usually employed for American or British practice is either 125 or 250 volts. Generator Leads. The generator leads to the switchboards must run as inconspicuously as possible. They are laid in the floor either in tile, lori- cated, iron, or other approved ducts. A more convenient way, particularly in large-sized plants, is to run the leads in trenches or tunnels, on insulators; and must be so arranged that the cables may be easily inspected; and free from any possible chance of short circuit. The size of the leads is always specified by the manufacturer. High Voltage Generators. The advantages of high voltage generator plants lie in lower first cost, low operating expenses, simplicity in station wiring and better line regulation. With the employment of high voltage generators, transformers and their attendant troubles are eliminated. A 3o,ooo-volt generator is, of course, more expensive than a 6ooo-volt or other low voltage generator. However, taking into con- sideration the step-up transformers necessary with low voltage generators, the cost for high voltage generators is lower; further, with the latter equipment, the power house is smaller and therefore less expensive. The simplicity in the station wiring is at once evident; there are no low tension bus systems, with switches, transformers and measuring instruments, therefore no special low tension compartments are necessary. However, better and more expen- sive instrument transformers are required. The operating performance of the station is simplified, there being no step-up transformers to look after when generators are put into service; when generators are to be run in parallel and thrown on the line, all that is necessary is to bring the incoming machines to synchronism and 176 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. throw them onto the bus. The absence of transformers relieves the system of surges, attendant when throwing unloaded transformers onto the busses. That there is sufficient reliability in operating high voltage generators is proven by the number of such generator plants on the continent of Europe; for instance, the 2o,ooo-volt, 3-phase, i5-cycle generator plant at Ponte de Desco, operating the Valtellina Railroad, Italy. Further, a 15,000- volt, 42-cycle, 2-phase gene- rator plant at Jaruga, Dalmatia, which has been in operation since 1903, feeding a six mile aerial transmission system. Owing to the successful operation of this latter plant, in 1906, another plant (Manojlovac, Dalmatia) was put into operation by the same company for the same purpose. This plant possesses four 6ooo-HP. Francis turbines, directly connected to 42-cycle, 3-phase generators, making 420 R.P.M. The 30,000 generator- voltage is transmitted over a twenty-one mile aerial line. SWITCHING ROOMS. General Arrangement. The switching rooms are located either in the power house itself or in a separate building. The latter is exclusively done in connection with large power plants; in such cases only the controlling apparatus are located in the power plant, while the entire switchgear for the outgoing feeders is in a separate building, some distance away. Examples of this arrangement are that of the Ontario Power Company, and the Canadian-Niagara Power Company. However, in smaller and average sized plants the whole switchgear is embodied in the gene- rating plant, and if transformers are necessary they are located in the same building or in an annex. The majority of hydroelectric power plants are of the latter type. It is decidedly bad practice to have the whole switchgear located in one single room. It must be separated regarding high and low tension, transformers, etc., either on separate floors or separated by partition walls. An example of this kind of arrangement is given in Fig. i, representing the relation of the various divisions of the Obermatt, Lucerne, power plant. It is considered one. of the best Swiss plants. It will be observed that the switching apparatus is located on three floors. The two lower are each separated by two longitudinal division walls; the upper is located in the tower-like structure from whence the long distance lines leave the building. The 6000 generator- voltage is stepped up to 27,000 volts through transformers located in an annex, longitudinal to the switching room. Good examples of American switching room practice are given in Figs. 3 and 4. Studying the cuts of the Shawinigan Falls Power Plant, 1 it will be observed that the 2200 generator- voltage is led to a separate switchboard, controlling the 25,000 and 5o,ooo-volt transmission groups. Beneath the switchboard gallery in the gen- erating room are the oil switches, while the two groups of lightning arresters are 1 Electric Power from Shawinigan Falls, Canada, by W. C. Johnson, Gassier 's Magazine, June, 1904. ELECTRICAL EQUIPMENT. 177 kept in separate compartments. Another very interesting arrangement of switching rooms is that of the Puyallup River plant of the Puget Sound Power Company, near Tacoma, Washington. Owing to the fact that the plant is located on the FIG. i. Cross Section of Obermatt Transformer and Switch Rooms. MA = Circuit Breakers. MO = Overload Switches. T = Transformers. / == Induction Coils. B = Horn Gaps. WW = Water Rheostats. WS = Water Flow Grounder's. hillside, the whole switching gear is located in two adjoining buildings as seen in Fig- 5- 1 The transformer rooms are at the same level as the generator room, but isolated from the latter by rolling steel doors. On floor No. 2 are the low tension discon- necting switches, the generator and transformer cables going to the sets of discon- necting switches on either side of the middle partition; the disconnecting switches are 1 Electrical World and Engineer, October 8, 1904. 178 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. . _ ' / \ L V \ _ _ _._ _ -J - ( . . ..^ .- FIG. 2. Castelnuovo-Valdarno Plant. Cross Section of Switch House. A-Generator leads; B-Circuit breaker 6000 V.; C-Potential transformer; D-Series transformer; E-Selector switch; F-Generator rheostat; G Transformers; H-Blowers; J-Circuit breaker, 33,000 V ; K-Series transformer, 33,000 V.; L- and N-Hook switches; M-Busbars, 33,000 V.; O-Circuit breaker, P-Series transformer; Q-Choke coils; R-Outgoing line; S-Ground-detector; T-Horn gaps; U-Water rheostats; V-Instrument column; W Switchboard. ELECTRICAL EQUIPMENT. 179 an aa aa ARRESTER) RAISING TRANSFORMERS 2200 TO 25,000 VOLTS 2200 TO 50,000 VOLTS GQOO OOQOD ^^^ ^ ^ ^**^-^ v_ s& b^^" ^--^ ^ -^ **^*^ ^ -A gag _ FIG. 3. Plan cf Shawinigan Falls Plant. FIG. 4. Shawinigan Falls Power Plant. i8o HYDROELECTRIC DEVELOPMENTS AND EXGEVEEREN'G. installed between the oil switches and the bus, being on the outer walls and imme- diately below the bus bar compartments, which are above on floor No. 3. In the center of floor No. 3 are the low tension oil switches, the two oil switches correspond- ing to a generator or a transformer bank being arranged back to back and facing their corresponding set of bus bars. The bus bars are of the laminated type, consisting of flat copper bars with expansion joints, and supported on marble slabs set on edge, which in turn rest on concrete slabs, forming barriers between adjacent bus bars. The compartments formed by the concrete slabs are covered by insulating fireproof doors. FIG. 5. Section through Generator Room and Switchhouse, Puget Sound Plant, Puyallup River, Washington. The oil switches are installed in brick cells with soapstone bottom and top slabs and doors. Each pole of a switch is separated from the others by brick barriers. The same general scheme is used for both the high and low tension disconnecting and oil switches, except that only one set of high tension bus bars is at present installed, provision being made for later installation of the second set. The high tension disconnecting switches and current transformers are on floor No. 5, while the high tension oil switches are on floor No. 6. Above floor No. 6 are the two outgoing high tension line towers, in the north end of which are the high tension lightning arresters, each pole being separated from its adjacent pole by brick barriers extending the full length of the arrester. The lines emerge from the wire tower centrally through an extra heavy 3o-inch sewer tile covered by a glass plate. ELECTRICAL EQUIPMENT. 181 SWITCHBOARDS. Object. The object of the switchboard is to collect the generated current for the purpose of controlling, measuring and distribution of same. The structure and apparatus mounted on same should be fireproof, and so arranged that easy access may be had to all parts to facilitate inspection and repair. The arrangement of the apparatus as well as the whole switching gear must be simple and symmetrical FIG. i. 5o-cycle, A.C. and D.C. Switchboard, Necaxa Power Plant, Mexico. s to prevent as much as possible the making of wrong connections; the number of instruments and protecting devices must be sufficient to secure a flexible and con- tinuous operation. All live parts, especially those of high potential, must be eliminated from the front of the switchboard. When installing a switchboard, provision must be made for further extension. 182 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. In laying out a switchboard, either for direct or alternating current, each gene- rator or machine must have its own panel; and the various panels for the same type of machines should be in a separate group; according to American practice, for instance, all alternator panels should be together, so that they can be operated from a central panel when working in synchronism. European practice is virtually the same as the above; however, in a very recent installation at Brusio, Switzerland, each generator has its own switchboard directly opposite the generator. When the various generators are working in parallel they are controlled from a single instrument column in front of the exciter switchboard o o o o 6- o o o o <> eo o ' c ! FIG. 2. Westinghouse Type of Panel and Bench Desk Switchboard. FIG. 3. General Electric Type of Panel and Bench Desk Switchboard. in the middle of the generating room. This system was adopted owing to the great number of generators installed. The generator attendants of the various machines look after the switchboards, while the outgoing feeders, of which there are few, are controlled from the above central instrument column. The leads from generators come to the switchboard from beneath, and the out- going feeders usually leave from the top. The latter particularly must be well arranged and inconspicuously placed. Types. Switchboards are either direct or remote controlled. For voltages, both alternating and direct current, the switches under 600 volts are direct control, while above this they are remote control, either by mechanical devices, such as bell cranks, rods, and gears, or electrically by solenoids or motors. There are a few installations ELECTRICAL EQUIPMENT. 183 where the remote control system is operated by compressed air, but such a system is not favored in present practice. The switchboards installed are for direct and alternating current. The direct current board is used principally for controlling excitation and the alternating for controlling the output of the main generators. In isolated plants for small industrial purposes, having no long transmission lines, a common switchboard is usually employed. FIG. 4. Generator Instrument Columns, Obermatt Plant, Luzerne, Switzerland. Oerlikon Co. Panel Type. Under switchboards one finds different types, such as panel, pedestal or column, and desk or benchboard. The panel type usually has mounted on it the entire switchboard equipment. In many instances however, in recent high tension practice, the switchboard has only on the front the meters and other indicating instruments, while the controlling switches are placed on a desk or bench in front of the panel or instrument board. These boards are made up of structural steel or pipe frames faced with white or blue marble or slate slabs. The marble presents a much finer appearance, but it readily shows oil stains and scratches, and if the board is extended at any future time it is difficult to match the panels. Marble 184 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. panels are chiefly used in isolated plants, and in central stations in Europe, in which case the panelboard is made very ornate. In most American central plants and substations slate with dull black or oil finish is used. It has the advantage of having a uniform shade, while scratches and oil spots are readily eradicated, also the instruments stand out in a bolder relief. The back of the switchboard for low tension wiring must be at least from 3 to 4 feet away from the wall and thoroughly braced at the foot and top. The sizes of the FiG. 5. Apparatus in back of aooo-K.W., 6ooo-volt Generator Switchboard Panel, Obermatt Plant, Luzerne, Switzerland. panels are practically standard. For instance, the General Electric Company's panel consists of two slabs, the lower one 28 inches high and the upper one 62 inches high, the width being 24 inches. The power section of the Westinghouse panel is 25 and the upper 65 inches. The Westinghouse panel is sometimes made up in three sections, the lower being 25, the middle 45, and the top 20 inches, the upper being primarily made for the mounting of a circuit breaker to enable easy removal in case of repair and substitution. ELECTRICAL EQUIPMENT. 185 Pedestal or Column Type. For controlling a single generator, a pedestal or column with all the necessary switches, instruments, etc., mounted on same is employed. In most cases, they are arranged in front of the feeder panel board, with the back of the column toward the generating room, so that the operator faces the instruments and generating room. A novel arrangement of pedestal and columns has been adopted by the Ontario Power Company, where they are arranged in a semi-circle, and easily overlooked from the desk of the chief operator. FIG. 6. Cross Section of Generating Station Switchboard Arrangement, with Oil Switches for Remote Control by means of Switchboard Lever. Desk or Panel Board. The desk or bench board are chiefly used for controlling the main oil switches, both of the generator and outgoing feeders. They are equipped with pilot switches and lamps, for operating the main circuit breakers, field switches, field rheostats, governor, motors, etc. These benches are usually placed in front of the instrument board, and must be so arranged that the panel of one circuit is directly behind the section of the same circuit on the controlling bench. To further facilitate operation, the panel and control bench are provided with card holders or name plates to classify the groups. In addition, dummy bus bars are mounted on the bench. 1 86 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. 7. Instrument and Controlling Bench Lontsch Hydroelectric Plant, Switzerland, Brown, Boveri & Co. FIG. 8. Typical Exciter or D.C. Generator Panel Arrangement (Walker Electric Co.). ELECTRICAL EQUIPMENT. 187 Where it is desirable that the switchboard operator should command a view of the bench, panel board and the generating room at the same time, the bench and board are placed back to the generating room with the board elevated on posts, with a space of 3 to 4 feet between the top of the bench and the bottom of the panel board. In some of the European plants, the designers entirely dispose of the instrument board by mounting the instruments and the controlling devices on a common bench. It is common practice to place all switchboards, columns and benches for controlling generators and outgoing feeders on galleries or mezzanine floors. Wtmeterf Ammeter - " Around Detec'b/ s 2a/nj& fffteostat Handn'/tee/' feeder^ (Switches Generator Switch Card Holder Feeder <5w/tc/i&$ FIG. 9. Typical D.C. Combination Panel Switchboard. Direct Current Board. The direct current board is usually made up of one panel for each exciter unit, containing a voltmeter, ammeter, main and field switch, circuit breaker and field rheostat, field discharge resistance, also equilizer switch for parallel operation. In some cases the latter switch is placed on a stand near the machine or mounted on the machine itself. In modern practice this exciter switchboard is placed on the main operating floor near the exciters, although in some cases they are placed on the operating gallery with the rest of the control apparatus. The circuit breakers on exciter panels must be non-automatic, while on D.C. generator panels it is essential that they are automatic. The reason for this is that the exciter current must not be interrupted except when the main unit supplied by it is shut down. For ordinary direct current distribution, the switchboard is divided into two parts, the generator panels and the feeder panels. The above-mentioned instruments are mounted on the generator end, while on the feeder end each feeder panel has an ammeter, integrating wattmeter, circuit breaker and single throw knife switch. In many cases there is a totalizing watt meter connected to the feeder busses. Where the switchboard is provided with two sets of busses, the feeder panels are provided i88 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. with double throw knife switches, so as to connect onto either bus. Fig. 8 shows a typical arrangement of a direct current generator panel with only one set of busses and an equilizer bus. Another type of D.C. switchboard equipment is seen in Fig. 9. It will be observed that it contains the necessary instruments for the generator and switches for feeder circuits, as the feeder circuits do not contain any instruments it is a typical switchboard for an isolated plant. Low Tension A.C. Boards. In isolated plants supplying light and power for manufacturing, low tension three phase and two phase is usually employed, or some modification, for instance, a 4-wire, 3-phase or a 3-wire 2-phase. The voltage varies from 200 to 600. Fig. 10 shows a typical layout of a low tension, 3-phase ;#)- Wattmeter Voltmeter Potent/at Receptacle CtciterKheostat-*- Synchronizing Receptacle Field Switcn Oil Switch H 24 H FIG. 10. 480 and 6oo-volt, 3-phase Generator Power with Three Main Ammeters, General Electric Company. generator panels as designed by the General Electric Company. These switch- boards are equipped with either three or a single ammeter. When three are used, there is one for each phase, while when one is used, it is assumed that the phases are balanced so that one meter is sufficient, being continuously connected to one phase, or by means of a receptacle and plug, connected to any of the three phases. All alternating current boards always have an oil switch for a main switch. Where there is a number of such panels, the synchronizing voltmeter and lamps are placed on a swinging bracket at the end of the board, while each panel is provided only with a synchronizing receptacle. The synchronizing voltmeter is sometimes replaced by a synchroscope or synchronism indicator. Wagon Panel. A novel feature in switchboard design, in use only a few years on the Continent of Europe, is the Wagon Panel System. Fig. n shows a carriage as constructed by the Allgemeine Elektricitats-Gesellschaft, Berlin. It consists of a ELECTRICAL EQUIPMENT. UNIV. OF CA ,O po Jj i ^ I i.i FIG. ii. Wagon Panel Switchboard of the Allgemeine Elektricitats-Gesellschaft. FIG. 12. Siemens-Schuckert Wagon Panel Switchboard. 190 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. carriage running on small wheels upon the structural steel frame of the switchboard and carries the panel with all the instruments. When a panel is to be removed, a portable wagon is backed up to the panel, and the latter is pulled out (each panel is provided with two handles) onto the wagon and removed. It will be observed that the electrical connections do not have to be disturbed as they are similar to a knife FIG. 13. Typical Arrangement of i3,2oo-volt Bus Bars, Electrically Operated Oil Switches, and Disconnecting Switches in a Three-Phase Station, General Electric Company. blade switch, that is, by means of heavy clips, which make and break the circuit when the carriage is rolled in or out. In the Siemens-Schuckert System, the entire panel and its equipment is built on a carriage which rolls on tracks in the floor. The electrical connections are made in a way similar to the above. The wagon of each system is so provided with ELECTRICAL EQUIPMENT. 191 locking switches, that it cannot be withdrawn while the panel is in operation, which is particularly essential for high tension switchboards. In the Siemens-Schuckert system the entire panel and its equipment is built on a carriage which rolls on tracks in the floor. The electrical connections are made in a way similar to the above. The wagon of each system is so provided with lock- ing switches that it cannot be withdrawn while the panel is in operation, which is particularly essential for high tension switchboards. The principal advantage of this " Wagon-panel System " is that a panel can be withdrawn for inspection and repairs and that a reserve panel can eventually replace an old one without disturbing the operation of the remaining units. Thus the danger otherwise encountered in making repairs on the switchboard is eliminated. High Tension Alternating Current Boards. High tension switches are of the remote control type, that is, the switches are located at a distance from the switchboard, so that the switchboard contains only low tension current apparatus used for operating the high tension switches, thus eliminating the danger of high tension apparatus from the operator. The oil switches are frequently mounted in masonry cells and operated either by motor or solenoid; and as they have no mechani- cal connections with the switchboard, they may be located at any convenient place. The motors or solenoids for operating the switches are mounted on top of the oil compartments and usually operated by no or 220 volt direct current. In some cases the current for operating the switches is taken from the exciter busses, while in others a storage battery is maintained for the purpose. SWITCHBOARD EQUIPMENT. Volt and Ammeters. The voltmeters and ammeters are either of the round, sector dial or of the edgewise type. The latter are frequently installed in such a way that they may be readily removed by lifting them out of the swtichboard slab. In some recent European plants the instruments are set so that the faces are flush with the panel. Instruments are always placed on the generator side of the line transformers. Where the voltage is higher than 150, potential transformers are required in connection with the voltmeter. For ammeters greater than 5 amperes series transformers are necessary, and for capacities greater than 800 amperes the series transformers are so arranged, to slip over the bus bar or cable. Wattmeter. Wattmeters are used to indicate the output of a generator. They are made both indicating and recording. The former gives only the momentary output, while the latter gives a continuous record. With Westinghouse wattmeters, except those for 5 amperes and not over 400 volts, series transformers are required. Where the alternations exceed 3000 and the voltage 200, shunt or potential trans- formers are required. For the General Electric Company's Thomson Induction Wattmeter a series and potential transformer are necessary only when the amperes exceed 150 and the volts 1150. Fig. i shows the wattmeter connections with and without a current transformer. 192 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Synchronizing. To facilitate the synchronizing of generators suitable instru- ments must be installed. They consist usually of two lamps and a bell, so that when the alternators are out of step this condition is indicated. The lamps are of dif- ferent colors, green indicating that the machine is running too slow and red that it is too fast. An instrument which indicates this condition more readily, that is, indicat- ing how much the generator is out of step, has been put on the market in recent years under the name of synchroscope and synchronism indicator. The attendant, when synchronizing a generator, operates a pilot switch con- trolling the motor on the water wheel governor. Thus when the synchroscope indi- cates that the incoming machine is too slow, he turns the pilot switch so that the motor will allow the governor to increase the speed of the turbine until synchronism is reached; and the generator is thrown on the bus. To minimize the time lost in synchronizing, an automatic synchronizer has been placed on the market. When using this instrument the attendant has only to put it in operation and adjust the speed of the incoming generator. At the first instance of synchronism the instrument will automatically throw the machine on the busses. FIG. i. Connections of Two Wire Induction Wattmeters. Power Factor Meter. In most alternating installations the power factor meter is of great value. It is built for single and polyphase circuits and for 3000 and 7200 alternations and in standard sizes up to 2000 volts and 2000 amperes. It indicates directly and accurately results which otherwise can only be reached by computation from readings which if not taken simultaneously may lead to error. The dial of the power factor meter is divided into four quadrants, each being marked from o to 100. These figures represent percentages which would be obtained by dividing the " true watts " in the circuit (as indicated by a wattmeter) by the " apparent watts " (the product of volts and amperes). The angular position of the pointer at any moment also indicates the angular difference in phase between current and voltage. The upper half of the dial indicates the power factor for lagging or leading currents when power is being delivered in one direction, and the lower half gives similar indications for power delivered in the opposite direction. In this way the power factor meter may serve to show a reversal of the direction of power on the line. In the operation of rotary converters this instrument finds important uses; it simplifies the adjustment of field strength, either for minimum armature current or ELECTRICAL EQUIPMENT. 193 to produce some desired effect on the system as a whole. The poor power factors resulting from heavy inductive loads may often be much improved or entirely neutralized by proper field adjustment of rotary converters and other synchronous apparatus. With generators running in parallel the proper distribution of the load can be checked, and in many cases the number of ammeters required may be considerably reduced by the use of the power factor meter. FIG. 2. Motor Control Rheostat for Field Excitation of Main Generator, Westinghouse Electric Manufacturing Company. Frequency Meter. To eliminate calculations necessary to ascertain the frequency of a generator, and incidentally the revolutions, a frequency meter may be employed. It may be mounted on the switchboard and occupy the same space as any other indicating instrument. It is built for any frequency with a + or -- variation of 25 per cent of the normal. When used on circuits exceeding 100 volts a shunt transformer is required. Rheostats. Rheostats in generating stations are used for controlling the excitation of generator fields. The rheostats for the exciters, being small, are mounted on the back of the exciter switchboard and in most cases are hand controlled, while those for the main generators are of large size, and have to be placed in compart- ments which must be well ventilated. These large rheostats are always remote controlled, either by shaft and gearing, much used in Europe, or by motors, common American practice. The pilot switches for controlling same are mounted on the control bench. 194 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Illumination. The switchboard proper must be suitably illuminated so that the condition of switches can at all times be readily seen, also, illuminate the faces of the. different meters. Some of the meters have opal glass scales and are illumi- nated from the rear, making it possible to read the indication from a distance or in otherwise insufficient light. The lamps are inclosed in a compartment separate FIG. 3. Remote Control Hand Operated Rheostats for Field Excitation of the Main Generator, Oerlikon Company. from the working part of the meter. Where meters with illuminated dials are not installed, a system of incandescent lamps is mounted on the switchboard proper or suspended in front of same, and for either alternating or direct current, but preferably of both to meet emergency. WIRING DIAGRAM. Systems. All power plants, whether small or large, depending on continuous operation, must have a double set of bus bars, or the equivalent. The leads from the generators to the bus bars must be so provided with switches that current can be thrown on to either of the systems. For small plants, where there is a light and power load, one bus may be kept separately for light and the other for power. To increase the flexibility of the system, the feeders to and from the bus bars must be ELECTRICAL EQUIPMENT. 195 196 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. such that either or both light and power can be drawn from either of the busses. The leads between generator and bus bars must be provided with a main switch, so that each individual unit may be cut out without affecting the operation of the remainder of the plant. An additional safeguard for continuous operation is to place sectionalizing switches in the bus bars. In high tension plants where oil switches are used, section or hook switches must be placed on either side of same, so that they may be isolated for inspection and High Voltage Feeders ill ill Low Voltage Feeders rtt ttt tt ttt ttt ttt III 111 III til 111 Ui II 1 TTT TTT TTT TTT &,&&&&&, Generators laoooVoffs FIG. i. Wiring Diagram of Power Circuits, Ontario Power Company. repair. Where transformers are installed either for stepping up or stepping down, particularly where there are a number, it is advisable to have a bus on the outgoing feeder side. In American practice the double bus bar system for outgoing feeders is seldom used, but will be found to a great extent in Swiss practice, in the form of a ring system. Fig. i shows the general arrangement of the wiring diagram of the Ontario Power Company's plant. 1 It will be observed that the wiring system is simple, yet as flexible as possible. 1 The Electrical Plant of the Ontario Power Company, by V. G. Converse. Canadian Electrical Asso- ciation, Niagara Falls, June, 1906. ELECTRICAL EQUIPMENT. 197 The generators can be thrown on either of the two low tension (12,000 volt) busses, and the transformers can draw from either of same. It will also be noticed that both high (62,000 volt) and low tension busses are well provided with section switches. The local distribution (12,000 volts) may draw current from either of the two low tension busses. A somewhat more simplified wiring system is that of the Necaxa power plant, Mexico (Fig. 2). There is only a single low tension (4000 volt) bus bar system. 40 000 Vollt outgoing Li: DIAGRAM OF CONNECTION OF NECAXA POWER PLANT FOR MEXICAN LIGKT ANO POWER COMPANY 000 Volts Fui in I! Switch \ V \ \ CO 000 Volts Di&om. Switch _ i^N^i J^ i ^^- "^ <" MO v "< VW|W^ VW^VVA VWWWW VWAWtfWV /WW1/VWW y Q /D .U U i.OOOV^Fo^.HS.itoh acdioulizing Svitcho FIG. 2. Wiring Diagram of the Necaxa Power Plant, Mexico. The current from the generators may be thrown upon same or directly on the transformers. This would mean that the transformers can draw from the bus bars irrespective of the operation of their own generator. There is but one high tension (60,000 volt) bus bar system. A wiring diagram in which the generator or low tension bus has been eliminated, although the plant is of large capacity, is that of the Urfttalsperre, Germany. Each 5ooo-volt generator feeds through a fuse directly to its transformer. The reason for this arrangement is that if a transformer is out of commission the particular 198 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. ELECTRICAL EQUIPMENT. 199 generator is shut down and vice versa. The high tension side is tied together with a ring bus bar system, the ends of which are connected through choke coils (see Fig. 4). Between the transformer and bus bar are automatic oil switches. Each outgoing feeder is provided with an automatic oil switch, choke coil, and three-pole tie switch. The outgoing feeders are doubly interconnected with tie switches, so that in fact two ring systems are secured on the high tension side. The whole is amply protected by lightning arresters. WSt FIG. 4. Wiring Diagram of the Urfttalsperre Plant, Heimbach, Germany. G = Generators. T = Transformers. t = Section Switches. WSt = Waterflow Grounders. WW= Water Rheostats. OW = Oil Rheostats. E = Earth. S = Circuit Breaker. s = Fuses. D = Choke Coils. Considering the protection and particularly the flexibility of the high tension side, it seems strange /that the low tension side should be so rigidly connected. It will be readily seen that this plant can be seriously handicapped when a transformer, a generator, and a turbine of different groups are out of commission. Swiss power plants employ either the double bus bar or ring system. Either one in itself is very intricate. A good example of this kind is given in Fig. 5, and is that of the Obermatt plant, Lucerne. It supplies light and power for various pur- poses, for which four 2ooo-HP. generators serve; a fifth generator used exclusively for street railway purposes (located in the left hand of the wiring diagram) is inde- pendent of the rest of the plant. As these four generators supply three phase for power or single phase for lighting a certain section, there are two ring bus bar 1-4-1 -"* FIG. 5. Wiring Diagram of the Obermatt Plant, Switzerland. ED - GB - DW- T RT - M AB - R AU KA - MA OA Exciter. MO Railway Generator U Three-phase Alternators. VU Transformers. D Reserve Transformers. TS Measuring Transformers. S Storage Battery. A Regulator. ST Cut out Switch V Carbon Cut-out Switch. DV Overload Switch SV SecUonalizing Switch. GV Overload Oil Switch. W '-- Double Throw Switch. L -- Voltmeter Switch. SL -- Double-cell Switch. WW '- Disconnecting Switch. B Fuse. F Ammeter. / = Series Transformer. WA Voltmeter. E Double Voltmeter Z = Static Voltmeter. ZO : Bus Bar Voltmeter ZR Wattmeter. Phase Lamp. Signal Lamp. Water Rheostats. Lightning Arrester. Lightning Arrester. Choke Coils Water Flow Grounders. Earth Plate Three phase Ammeter Overload Oil Circuit Breaker Time Relays 200 ELECTRICAL EQUIPMENT. 201 Direct-Current Bui Ban FIG. 6. Complete Wiring Diagram of a Single Generator and Step-up Transformer. systems on the low (6000 volt) tension as well as the high (27,000 volt) tension side. The two bus bar systems on either side consist of a single phase and three phase group. BUS BARS. Bus bars are made up of cables or flat bars of copper or aluminum. Flat bars, for mechanical reasons, are preferable to round ones, as connections are readily made. Where the bus bars are of short length they may be made of uniform section throughout. However, where the bus bars extend over the entire length of the generating room, as is frequently the case for sake of economy, the bus bars are flat, each of across section area necessary for one generator; thus where a generator lead joins it an additional bus bar section is added. Size of Bus Bars. The size of the bus bar is determined by the number of amperes it has to carry. From 700 to 800 amperes per square inch of copper is usually chosen as a safe value. For electrical reasons flat bars are better than round ones, 2O2 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. because they present a greater radiating surface thus keeping the resistance from increasing due to heating. This is also a reason in favor of open bus bar com- partments. If aluminum is used in place of copper, the bus bars must have a cross sectional area of 1.66 times that of copper, for equivalent electrical conductivity. Closed Compartments. Similar to the arrangement for high tension oil switches, the bus bars and disconnecting switches are also placed in masonry compartments. The compartments are made either of brick or concrete, and are entirely closed FIG. i. Three-Deck Oil Circuit Breaker and Bus Bar Structure. Two Sets of Bus Bars. %%%^#3^^ FIG. 2. Typical American Arrangement of Oil Circuit Breakers and Bus Bars for 15,000 volts or less. (having access through openings) or else open altogether. With high tension busses, where space is limited, the former arrangement is preferable. The open- ings to the compartments are about 15 inches square and are staggered. As a means of protection for inspectors and repair men they are best closed with a sheet steel door. Open Compartments. Where space is ptenty, which is usually the case with hydraulic plants, the entire front of the bus bar compartment may be left open. The insulators for carrying the busses are mounted either on the back wall or shelves. ELECTRICAL EQUIPMENT. 203 FIG. 3. 6ooo-volt Bus Bar Room, Obermatt Plant, Luzerne, Switzerland. FIG. 4. 8ooo-volt Oppen Bus Bar Chamber, Lontsch Plant, Switzerland. 204 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. In order to eliminate posts or partitions for carrying the shelves, the latter are best made of reinforced concrete, giving an unobstructed view of the busses and also facilitating construction, inspection, and repair. Fig. 4 gives an illustration of open bus bar construction at the Lontsch plant, Switzerland. The busses carry 80,000 volts. Another view of Swiss practice is seen in Fig. 3 where no shelves between individual phases are used. There is only one shelf which separates the single phase from the three phase busses (6000 volts.) OIL SWITCHES. General Remarks. In modern switchboard engineering one will find the most contradictory practice. Oil switches for 600 volts are designed and installed on the back of the switchboard, operated by a single lever, while on the other hand they are remote controlled. Again, switches for 10,000 volts are mounted directly on the back of the switchboard. Also, with 6ooo-volt switches, the oil chambers for the indi- vidual phases are placed in large and very expensive masonry cells, the access to which is well protected by specially designed fireproof doors; while on the other hand, with 5o,ooo-volt oil switches, all phases are placed in a single sheet metal tank, unprotected and exposed to view. In addition to this, the former is operated by motor or solenoid, the latter by lever and rods. The difference between America and Europe in this practice is clearly indicated in accompanying illustrations. To go still further in citing the difference existing in American practice, there are 60,000 and 80,000 volt oil switches in operation which have the phases in separate sheet steel tanks exposed to view, being entirely unprotected by masonry construc- tion, as is done in the 6ooo-volt switches. Even 120,000- volt oil switches of the former type are advocated. From the foregoing contradictory practice it will be observed that arguments about the danger to the operating force from exposed high tension apparatus are without reason, particularly if one bears in mind that practice has proven that 220 volts may kill as readily as 30,000 volts and higher. It is evident from the above that much could be saved on first cost, maintenance, and floor space in the design of modern oil switches. Types. Oil switches up to 5000 volts are, according to American practice, of the self contained oil type; they are made up of a sheet steel tank with partitions to separate the phases, and lined with insulating material. The lining and partitions are frequently made of wood. The contacts of the switch are in oil, so that the make and break are made submerged. The phases are usually equipped with multiple break contacts, thus securing a great current breaking capacity. The practice regarding the operation of high voltage switches varies greatly. In some cases 2300-volt switches are operated by remote control levers, rods, and bell cranks, or by motors or solenoids. In Swiss practice, io,ooo-volt switches, in some instances, are mounted on the back of the switchboard and operated by hand levers as seen in Fig. i. Frequently in American practice switches larger than 5000 kilowatts capacity, the phases are submerged in individual oil tanks and have multiple break ELECTRICAL EQUIPMENT. 205 FIG. i. Oerlikon io,ooo-volt Air Break Switch. FIG. 2. Oerlikon 3o,ooo-volt, 3oo-ampere, Solenoid Controlled Circuit Breaker. 2O6 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING, contacts. Each pole of these oil switches is inclosed in a separate fireproof structure made of brick or concrete. The use of soapstone is not essential and it must be borne in mind that it readily absorbs oil. The doors to the tank compartments are either asbestos fiber, slate slabs, or wire glass, and are held in place by clamps, or hung from the top of the compartment. It is common practice on the continent of Europe to have the poles of all phases for any voltage in one sheet metal tank. In some cases, however, phases are placed in small separate oil tanks and not separated by partition walls. The FIG. 3. Westinghouse n,ooo-volt, 6oo-ampere, Remote Solenoid Controlled, 3-pole Oil Switch. oil tanks are grouped and mounted on iron frames in compartments of reinforced concrete and always exposed to view. Switches or circuit breakers of this construc- tion have been installed up to 50,000 volts normal capacity (see Fig. 2). American switches for 60,000, 80,000, 100,000 volts, and even higher are designed on the same principle as those for 30,000 volts. They are either top or bottom connected. The bottom connected switch is arranged with two pots forming one pole of the switch mounted on a common horizontal platform, and is usually operated by a motor. This type of switch requires a comparatively small amount of oil, and has a further advantage, that the circuit is opened in two independent receptacles ELECTRICAL EQUIPMENT. per phase. According to Hayes/ the exposed metal parts of this switc tank and bare terminals below, necessitate the inclosing of the switch in a masonry structure for the protection of the attendant. Doors are provided for each compart- ment of the structure, to permit the ready inspection of the tanks, etc., but the removal or breaking of a door leaves these live metal parts a source of danger. Such switches FIG. 4. High Tension Oil Switch Compartment, Ontario Power Company. have been installed in the 6o,ooo-volt circuit of the Electrical Development Corn- pan)' of Toronto at Niagara Falls. The top connected switches are usually solenoid operated, and the oil tanks are of sheet metal. The two stationary contacts forming each pole are located near the top of the oil, where sediment cannot settle. The contacts are separated by barriers, Switchboard Practice for Voltages of 60,000 and upwards, by S. Q. Hayes Am. Inst. E. E., June, 1907. 208 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. as though each contact were in a separate tank. This type is usually installed without being encased in masonry compartments; however, the tanks and all the mechanism must be properly grounded for the protection of attendants. In the distributing FIG. 5. 3o,ooo-volt, 5o-ampere, Remote Motor Control, 3-pole Oil Switch, General Electric Company. station of the Ontario Power Company types of this switch are installed for 62,000 volts. Both switches, of course, can be installed as circuit breakers, as practically all high tension switches have this provision, and can be arranged to work with a scheme of inclosed or open wiring. The bottom connected switch is essentially ELECTRICAL EQUIPMENT. 209 designed for plants where the wiring, bus bars, etc., are placed in separate com- partments, while the top connected breaker is designed for plants where the wiring is overhead. Circuit Breakers. A high tension oil circuit breaker is nothing more than an oil switch provided with an automatic opening device. The purpose of the circuit FIG. 6. 11,000 and 5o,ooo-volt Switches and Substation, Castellanza, Italy. breakers is to protect the generators and transformers from overloads, reversal of line current, and excess voltage, for which purpose (a) overload relays, (&) reverse current relays, (c) over-voltage relays, are installed. (a) Overload Relays. Overloads in most cases are caused by short circuits on the line. The common practice is to maintain the short circuit and burn it out. There are cases, however, where the short circuit cannot be burned out, and to maintain it would damage the line. To prevent long and excessive shorts from. 210 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. 7. 88,ooo-volt Top-connected Circuit Breaker with Sheet-Metal Tanks. FIG. 8. Siemens-Schuckert 35,ooo-volt Remote Control FIG. 9. Siemens-Schuckert High Switch with two Current Blowouts. Oil Tank removed. Tension Time Relay. ELECTRICAL EQUIPMENT. 211 damaging the line, the circuit breakers are provided with an overload relay, which will cause them to open in a certain set time after the short is established. The time varies usually from one to five seconds, depending upon the setting of the time limit of the relay. (b) Reverse Current Relays. The reversals of current are caused chiefly by synchronous apparatus connected on the line, such as synchronous motors and rotary converters. Cases sometimes arise that these machines instead of absorbing power will pump it back, and when this happens there is usually trouble, especially if there are underground conduits in the system. To prevent any damage from this source, the power house circuit breakers are provided with reverse current relays, which cause the breaker to open upon reversal of line current. (c) Overload Voltage Relays. The over-voltage relay causes the breaker to open upon excess of normal voltage, due to over-excitation or poor line regulation. The former difficulty is under the control of the operator, while the latter is under the control of the designing engineer and may be eliminated. These relays are not extensively used on high tension systems. BIBLIOGRAPHY. THE THEORY AND CALCULATION OF ALTERNATING CURRENT PHENOMENA. Charles P. Steinmetz, 1908. GENERAL LECTURES ON ELECTRICAL ENGINEERING. Charles P. Steinmetz. 1908. ALTERNATING CURRENT ENGINEERING PRACTICALLY TREATED. E. B. Raymond. 1905. ALTERNATING CURRENT MACHINERY. William Esty. 1907. ALTERNATING CURRENT MACHINES. Samuel Sheldon and H. Mason. 1908. EXPERIMENTS WITH ALTERNATING CURRENT OF HIGH POTENTIAL AND HIGH FREQUENCY. N. Tesla. 1904. HIGH SPEED ELECTRICAL MACHINERY. H. M. Hobart and A. G. Ellis. 1908. DYNAMOS, MOTORS, ALTERNATORS, AND ROTARY CONVERTERS. G. Knapp. 1902. POLYPHASE ELECTRIC CURRENTS AND ALTERNATE-CURRENT MOTORS. S. P. Thompson. 1903. THE INSULATION OF ELECTRIC MACHINERY. H. Turner and H. M. Hobart. 1907. PRACTICAL DYNAMO AND MOTOR CONSTRUCTION. A. W. Marshall. 1907. ELEMENTARY PRINCIPLES OF CONTINUOUS-CURRENT DYNAMO DESIGN. H. M. Hobart and H. F. Parshall. 1908. THE MANAGEMENT OF ELECTRICAL MACHINERY. F. B. Crocker and S. S. Wheeler. 1907. DYNAMO-ELECTRIC MACHINERY. F. B. Crocker. 1907. TROUBLES OF CENTRAL STATION SWITCHING APPARATUS AND METHODS OF HANDLING THEM. C. F. Conrad. Electrical World, Aug. i, 1908. EXTRA-HIGH-PRESSURE IRONCLAD SwiTCHGEAR. Electrical Review, London, July 24, 1908. METER AND RELAY CONNECTIONS. H. W. Brown. Electric Journal, May, 1908. SWITCHING APPARATUS AND ITS PRACTICAL OPERATION IN LARGE HYDRO-ELECTRIC STATION. Frank E. Conrad. Electrical World, July 25, 1908. OPERATION OF LARGE HYDRO-ELECTRIC STATION SWITCHING APPARATUS. F. E. Conrad. Electrical World, May 30, 1908. SOME FEATURES OF EUROPEAN HIGH-TENSION PRACTICE. Frank Koester. Electrical Age, December, 1008. SWITCHBOARD PRACTICE FOR VOLTAGES OF 60,000 AND UPWARD. Stephen Q. Hayes. Pro. Am. Inst. E. E., June, 1907. \ PART II. THE TRANSMISSION OF HIGH TENSION ELECTRICAL CURRENT. PART II. THE TRANSMISSION OF HIGH TENSION ELECTRICAL CURRENT. CHAPTER VIII. ELECTRICAL TRANSMISSION. A TRANSMISSION line should run at low levels and near highways, to facilitate erection, inspection and repair; it must further be borne in mind that the line must be as straight and short as possible, to minimize first cost, maintenance, line loss, and the expenditure for securing the right of way. Where the line runs through moun- tainous countries, high peaks must be avoided, because the temperature range between peak and valley is great and atmospheric electrical discharges are frequent. It is therefore better policy to detour the line than be troubled with atmospheric discharges. For high-tension lines, two separate circuits must be run either on a common or two separate towers, so that one is always in reserve. Such lines must be divided up into sections, provided with section switches and by-pass connections, so that a continuity of service is assured. The sections may be about 20 miles long, and at the end of each section must be a repair shop and accommodations for the patrolman, whose duty is to inspect the section once or twice a day. All poles and towers must be properly numbered to facilitate the location of trouble. Telephone connections must be established at the patrolman's quarters and at frequent intervals. For further convenience, portable telephones may be used. The telephone line, in duplicate, is best run on separate poles, and is also used for telegraph purposes. These lines must be for the exclusive use of the transmission company. TRANSMISSION CONDUCTORS. Strength of Conductors. Aerial lines transmit the electrical energy from hydraulic plants to the center of distribution. The material used for conductors is copper, hard or soft drawn, aluminum and steel; all three are used in cable form, while copper is sometimes used as solid conductor. For transmission purposes, hard drawn copper is almost always used, as it has an ultimate tensile strength of 60,000 pounds per square inch, while that of soft drawn copper is only 30,000 pounds. The resistance of the former is 5 per cent greater than that of the latter. Aluminum has a tensile strength of about 28,000 pounds; while steel varies greatly, it averages 100,000 pounds. 215 216 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Elasticity of Conductors. The elasticity of a cable is much greater than that of a sojid wire, therefore cables are preferable for long-span transmission lines. TABLE I. MODULUS OF ELASTICITY. Copper wire, hard drawn Copper, hard drawn cable strand 1 6 300 ooo Aluminum, hard drawn Steel wire TABLE II. COEFFICIENT OF EXPANSION PER DEGREE FAHRENHEIT. Copper . . o 0000006 Aluminum Steel From the tables it will be seen that steel cables compare very favorably for long spans, but the disadvantage is the low conductivity, being only 12, while that of copper is 100. Cables as Conductors. The strands in a cable must be continuous, to give it uniform strength and conducting area. The lengths of cable obtainable are longer than those of a solid conductor, for the reason that a cable is made up of a number of small strands, each of which is made from the same-sized ingot as a larger con- ductor. The weights of cables are calculated about one per cent heavier than a solid wire of the same circular millage, while the resistance is calculated for a solid con- ductor. The number () of strands in a cable of given circular millage (C.M.), composed of wires of diameter (d), is found by the following formula: C.M. = d 2 X n. C.M. The diameter of a cable is found by multiplying the diameter of one wire by the factors given in Table III, according to the number of strands composing the cable. Another convenient table on this subject is found in the appendix. TABLE III. STRAND FACTOR. Number of strands Factor. 3 2.25 7 3.00 12 4-25 19 27 5.00 6.25 ELECTRICAL TRANSMISSION. 217 Spacing of Conductors. The spacing of conductors depends on the voltage and on the length of the spans, and varies from 24 to 96 inches. The increase in spacing increases the inductive drop, and also the line loss. There are no fixed rules estab- lished for the spacing of conductors. The following are approximate distances: for voltages from 5000 to 10,000, 24 to 36 inches; for 10,000 to 30,000, 36 to 60 inches; for 30,000 to 60,000, 60 to 84 inches; for 60,000 to 100,000 and higher, from 84 to 96 inches. Characteristics of Conductors. Comparing the specific gravities of aluminum and copper, the latter is about 3.3 times greater than the former, so that for cables of equal length and resistance, the copper cable is twice as heavy as the aluminum. As the tensile strength is only about 28,000 pounds per square inch, aluminum wires must be used only in the form of a cable. For equivalent conditions, the diameter of an aluminum cable is 1.28 times that of copper, while the cross-sectional area is about 1.65 times larger, because the conductivity of aluminum is only 63 per cent that of copper. Cables present a larger surface to the wind, and also, for the forma- tion of ice. The advantages gained in the use of aluminum are, cheaper than copper and light weight, which in turn reduces the cost of pole line construction. Owing to the high coefficient of expansion, aluminum wires must be strung either in the spring or fall, preferably the latter. Heretofore, much difficulty was experienced in splicing cables of aluminum; this has been overcome in recent practice. According to the laws of transformation of alternating currents, the higher the voltage, the smaller the current for a given amount of power, and vice versa. So far as the transmission line itself is concerned, the highest voltage which can commer- cially be produced, is the best voltage. A small current on the line means that the size of wire can be reduced until the mechanical strength of the wire predominates. The use of high voltages reduces the line drop, losses in transmission, and gives better regulation than low voltages. Size of Conductors. The essential factors necessary for calculating the size of line conductors are: the load to be transmitted, the voltage desired at the receiving end, the permissible loss of energy in transmission, the frequency, spacing of the wires on the cross-arms, length of transmission line. The size of conductors is, for standard sizes, usually designated by a number; for sizes larger, they are designated by their cross-sectional area in circular mils. The largest standard size is oooo and is 0.46 of an inch in diameter. The circular mil is the thousandth part of an inch, and is chosen as the unit of measurement for electrical conductors. Thus the diameter of a one-inch cable is 1000 mils, and its area is a million circular mils or (iooo) 2 circular mils. 218 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. In calculating the size of transmission conductors it is poor policy to use a larger conductor than is absolutely necessary. This fact is well illustrated in a law developed by Lord Kelvin, and known by that name. The usual way of stating it is: " The most economical area of conductor will be that for which the annual interest on the capital outlay equals the annual cost of the energy wasted." A more precise statement of the same thing is given by Gisbert Kapp, 1 " The most economical area of conductor is that for which the annual cost of the energy wasted is equal to the interest on that portion of the capital outlay which can be considered to be proportional to the weight of the metal used." Direct Current Conductors. For calculating the size of conductors for direct current distribution the following formulas suffice: _K.W. X looo E 2 X D XI X ii .. = size of conductor in circular mils. E X p K.W. = load in kilowatts. / = current. E = voltage. D = distance one way in feet. ii = resistance of copper per mil-foot. p = percentage drop in voltage. Direct Current Problem. For example, it is desired to transmit 500 K.W. at 500 volts for a distance of two miles with a line drop of 10 per cent. 2 miles = 5280 X 2 = 10,560 ft. 500 X 1000 Then / = - = 1000 amp. 500 2 X 10,560 X looo X ii C.M. = - = 4,646,400. 500 X o.io Upon consulting the wire tables it will be seen that this is not of standard size, in fact it is a little less than 22 wires of No. oooo, the largest standard size. Suppose 22 wires were used. The resistance of the whole circuit is one twenty- second of one circuit of No. oooo. The resistance of No. oooo is about 0.05 ohm per 1000 feet. The resistance for the whole is 2 X 10.56 X .05 -* 22 = 0.048 ohm. To check up results on this assumption, Voltage drop = 0.048 X 1000 = 48 volts. - X 100 = 9.6 per cent, line drop. 500 (icoo) 2 X 0.048 -5- 1000 = 48 K.W., line loss. 48 -H 500 X 100 = 9.6 per cent, energy loss in transmission. 1 Economical Conductor Section, by Frank G. Baum. Electrical World, May 25, 1907. ELECTRICAL TRANSMISSION. If in the above calculations the voltage be doubled, the size of the wire will be only one-quarter as great; that is, the amount of copper varies inversely as the squt of the voltage. Alternating Current Conductor. In the calculations of alternating currents new factors have to be taken into consideration, and their value depends upon the fre- quency and the distance the wires are apart, etc. In direct current transmission the losses can be calculated either E X I or PR, but in alternating current only PR gives the real loss; E X I gives the apparent loss. Alternating Current Problem. For single phase transmission the following exam- ple gives approximate results. Suppose a load 1000 K.W. is to be received at a distance of ten miles at 10,000 volts, frequency 25 cycles, power factor 0.85, line loss 10 per cent, wires spaced 36 inches apart. 1000 K.W. = actual energy. 1000 -4- 0.85 = 1176.4 K.W., apparent energy. 1176.4 X 1000 -T- 10,000 = 117.64 amperes. PR = (117.64)^ = line loss. Also 1000 K.W. X 10 per cent = ioo K.W., line loss. Then (n 7.64) 2 X R = ioo X 1000. ioo X 1000 R = = 7.23 ohms (total). (ii7.6 4 ) 2 7.23 X 1000 Resistance per 1000 feet = - - = 0.0684 ohm. 2 X 10 X 5280 This corresponds nearly to a No. oo wire, which has a resistance of 0.076 ohms per 1000 feet. The resistance of the circuit with No. oo is 5.280 X 2 X 10 X 0.076 = 8.02 ohms. 117.6 X 8.02 = 941 volts; resistance volts. To calculate the drop due to reactance, recourse to Table VII is necessary. It represents the reactance volts per ampere per 1000 feet of line (2000 feet of wire) at 60 cycles. For distances not given in the tables interpolations are made directly. From the table the constant for No. oo placed at 36 inches is 0.254. This is the drop per ampere at 60 cycles. As this value varies directly with the frequency, for 25 cycles the reactance volts are 105.6 (thousands of feet) X 117.6 X 0.254 X = 1314- 60 The line drop is not current X resistance, but \/(94i) 2 + (1314)* = 1597- 1597 H- (10,000 + 1597) X ioo = 13.77 P er cent > drop in generator station voltage. (117. 6) 2 X 8.02 -T- 1000 = 17.24 K.W., line loss. 17.24 -5- (1000 + 17.24) X ioo = 16.9 per cent, loss of energy in transmission. 22O HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. TABLE IV. COMPARISON OF WIRE GAUGES. No. A. W. G. Brown & Sharpe. (B. & S. G.) Stubs. Birmingham. (B. W. G.) (N. B. S.) English Standard. (S. W. G.) A. S. & W. Co. Washburn & M. Roebling's. Diam. Mils. Cir. Mils. Diam. Mils. Cir. Mils. Diam. Mils. Cir. Mils. Diam. Mils. Cir. Mils. oooooo 580. oo 516.50 460.00 409 . 64 364.80 324.86 289.30 257-63 229.42 204.31 181.94 162.02 144.28 128.49 H4-43 101 . 89 90.742 80.808 71.964 64.084 57.068 50.820 45-257 40.303 35-890 31.961 28.462 25-347 22.571 20. IOO 336,400 266,772 211,600 167,805 133.079 105,534 83,694 66,373 52,633 4i,742 33,i2 26,250 20,817 16,509 i3>094 10,381 8,234. i 6,529.9 5,178-4 4,106.8 3.256-8 2,582.7 2,048.2 1,624.3 1,288.1 1,021.5 810.08 642.47 509-45 404.01 464 432 400 372 348 324 300 276 252 232 212 192 I 7 6 160 144 128 116 104 92 80 72 64 56 48 40 36 11 24 22 215,296 186,624 1 60,000 138,384 121,104 104,976 90,000 76,176 63,504 53,824 44,944 36,864 30,976 25,600 20,736 16,384 13.456 10,816 8,464 6,400 5.184 4,096 3.136 2,304 i, 600 1,296 1,024 784 576 484 460 43 394 363 33i 307 283 263 244 225 207 192 177 162 148 135 121 106 92 80 72 63 54 48 4i 35 32 29 26 2 3 211,600 184,900 154,764 131,406 109,561 94,249 80,089 69,169 59,536 50,625 42,849 36,864 31.329 26,244 21,904 18,225 14,520 11,130 8,464 6,400 5.184 3.969 2,916 2,256 i, 68 1 1.225 1,024 818 666 529 1 ooooo oooo 454 425 380 340 300 284 259 238 220 203 180 165 148 134 1 20 109 95 83 72 65 58 49 42 35 32 28 25 22 206,116 180,625 144,400 115,600 90,000 80,656 67,081 56,644 48,400 41,209 32,400 27,225 21,904 17,956 14,400 11,881 9,025 6,889 5.184 4,225 3,364 2,401 1,764 1,225 1,024 784 625 484 ooo oo o I 2 A. . e. . 6 7 8 10 ; ii 12 H 14 1C . . 16 17 . . 18 10. . 20 21 22 27 . . 24 N. B. S. W. G. has No. ooooooo; diameter, 500 mils; area, 350,000 cir. mils. ELECTRICAL TRANSMISSION. 221 TABLE V. SOLID COPPER WIRE. BARE AND INSULATED. Diam. Mils. Weight, pounds. Resistance. Carrying ca- pacity. Amp. No. Area. T. B. Int. Ohms, 20 C. (68 F.)-Matt. Std. B. &S. T. B. Cir. Mils. Bare. Weatherproof. i6F. rise, 3 2F. Bare. Weather con- r s , proof. 1000'. Mile. 1000'. Mile. 1000'. Mile. cealed. open. oooo 460 780 211,600 640.5 3.38i 754 3,98o . 04893 -2583 2IO 312 000 409. 6 700 167,805 508.0 2,682 614 3,240 .06170 3258 177 262 00 364.8 635 I33. 79 402.8 2,127 486 2,570 .07780 .4108 ISO 22O o 324.9 59 105.534 3I9-5 1,687 388 2,050 .09811 .5180 127 I8 5 I 289.3 55 83,694 253-3 i,337 312 1,650 .12370 6531 107 156 2 257.6 5i5 66,373 200. 9 1,062 254 1,340 .1560 .8237 90 131 3 229.4 45 52.633 159-3 841.1 201 i, 060 .1967 1.0386 76 110 4 204.3 43 4i,742 126.4 667.4 I6 3 860 . 2480 1.3094 65 9 2 5 181.9 400 33, 102 100. 2 529.0 134 710 .3128 1.6516 54 77 6 162.0 360 26,250 79.46 4I9-5 112 59 3944 2.0824 46 65 7 J 44-3 335 20,817 63.02 332-7 8 9 47 4973 2.6257 39 55 8 128.5 280 16,509 49.98 263.9 73-8 39 .6271 3-3III 33 46 9 114.4 255 i3, 94 39-63 209. 2 60.6 320 .7908 4-1754 28 38 10 101.9 220 10,381 31-43 166.0 50.2 265 .9972 5-2652 24 3 2 ii 90.74 205 8,234 24-93 131.6 42. 6 225 1-257 6. 6370 20 27 12 80. 81 I8 5 6,530 19.77 104.4 35-o 185 1.586 8-374 17 2 3 13 71.96 170 5,i78 15-68 82.79 30-3 160 1.999 10. 560 14 19 14 64.08 160 4,107 12.43 65-63 25-9 137 2.521 ^sn 12 16 15 S7-7 *55 3,257 9.858 52-05 22. 7 I2O 3-179 16.785 9 12 16 50.82 15 2,583 7.M 41.28 19.9 105 4. 009 21. i. 68 6 8 i? 45.26 147 2,048 6. 200 32-74 18.0 95 5-55 26. 690 4-5 6-5 18 40.30 i45 1,624 4.917 25.96 16. i 85 6-374 33-655 3 5 19 35.89 140 1,288 3-899 20.59 14.2 75 8.038 42.440 2-3 4 20 31.96 i35 1,022 3.092 16.33 12.3 65 10. 14 53-540 i-5 3 A.I.E.E. s.u.c.c. A. I. E. E. Average. 1 A. I. E. E. Nat. Elec. Code. 1 Std. Und. Cable Co., J. A. Roebling's Co., Amer. Elec. Works, Gen. Elec. Co., Amer. St. & W. Co., Hazard Mfg. Co. 222 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. TABLE VI. STRANDED COPPER WIRE. BARE AND INSULATED. Diam. Mils. Weight, pounds. Resistance. Carrying Int. Ohms, 20 C. capacity. No. and (68 F.) Matt. Std. Amperes. Diam. of T. B. Strands Area. Bare. Weatherproof. or Size T. B. Cir. Mils. Bare. Weather- 16 F. O p B. & S. proof. 1000'. Mile. rise, rise, 1000'. Mile. i ooo'. Mile. con- cealed. open. 9i/. 128 1,408 1.875 1.500,000 4,575 24,156 5,335 28,169 . 006902 .03644 850 1,360 9i/. 117 1,287 i,775 1,250,000 3,813 20,132 4,45 23,496 .008282 04373 750 1,185 6i/. 128 ,I5 2 1,6.35 t ,000,000 3.050 16,104 3,610 19,061 010353 .05466 650 I, OOO 6i/. 125 ,125 i, 600 950,000 2,898 I5>299 3,450 18,216 . 010900 05755 625 960 6l/. 121 ,092 1.525 900,000 2,745 14,494 3,268 17,255 .01150 .06072 600 920 6i/. 118 ,062 1,500 850,000 2,593 13,688 3,083 16,278 .01218 .06431 575 880 6i/. 115 ,035 1,475 800,000 2,440 12,883 2 ,905 15,338 .01294 . 06832 550 840 6i/. in 999 1,425 750,000 2,288 12,078 2,730 14,415 .01380 .07286 525 Soo 6i/. 107 963 1,405 700,000 2,i35 ",273 2,569 13.565 .01479 . 07809 5 760 6r/. 103 927 i,375 650,000 1,983 10,468 2,393 12,635 01593 .08411 475 720 6i/.o99 891 600,000 1,830 9,662 2,215 11,695 .01725 .09108 45 680 617.095 855 1,300 550,000 1,678 8,857 2,040 10,771 .01882 09937 420 635 6i/.ogi 819 1,250 500,000 1,525 8,052 1,870 9,875 .02070 10930 390 590 37/. 110 770 1,200 450,000 1,373 7,247 1,694 8,945 . 02300 .12144 360 545 37/-I04 728 I,19O 400,000 'i, 220 6,442 1,529 8,075 .02588 13664 33 500 37/-097 679 1,125 350,000 i, 068 5,636 1,320 6,970 .02958 .15618 300 45 C 377.090 630 955 300,000 9i5 4,831 i,i33 5,982 03451 .18221 270 400 37/-o8 3 59 920 250,000 762 4,026 940 4,963 .04141 .21864 235 35 C oooo B &S 530 805 211,600 641 3,381 779 4,115 04893 2583 210 312 ooo " 47 75o 167,805 508 2,682 650 3>432 .06170 3258 177 262 00 " 420 660 133,079 403 2,127 527 2,785 .07780 .4108 150 220 o " 375 615 105,534 320 1,687 43 2,270 .09811 .5180 127 185 I " 330 575 83,694 253 i>337 326 1,722 .12370 6531 . 1C? 156 2 " 291 535 ' 66,373 2OI 1,062 270 1,426 .. 15600 .8237 90 131 3 " 261 475 52,633 159 841 218 1,15 . 19670 1.0386 76 no 4 231 4i,742 126 667 176 93 . 2480 I . 3094 65 92 Haz. M.C. J.AR Haz. M.C. A.I.E.E. J. A. Roebling's Haz. Mfg. Co. A. I E. E. Nat Elec. Code ELECTRICAL TRANSMISSION. TABLE VII. REACTANCE VOLTS. 223 Distance apart of conductors in inches. Size of Conductor. i2-inch. 1 8- inch. a 4- inch. 30- inch. 36-inch. 48-inch. 96-inch OOOO 193 . 212 .225 235 .244 251 .283 ooo .199 .217 230 .241 .249 255 .287 oo . 204 . 222 236 .246 254 . 262 .294 . 209 .228 .241 251 2 59 . 267 .299 I .214 233 . 246 .256 265 273 35 2 .220 .238 .252 . 262 .270 277 309 3 .225 244 257 . 267 275 .282 3M 4 .230 .249 . 262 .272 .281 .294 .326 5 .236 254 .268 .278 .286 .299 33 1 6 .241 . 260 .272 .283 . 291 35 337 7 .246 .265 .278 .288 . 296 .310 342 8 .252 . 270 .284 293 .302 3i5 347 TABLE VIII. WEIGHT AND STRENGTH OF ELECTRICAL WIRES. Weight. Pounds per 1000 feet (bare). Breaking weight. Pounds. No. B. &S. Copper & Phono- Alum- Wrought Copper, Copper, Phono- Alum- Charcoal Crucible Elec. inum. Iron. Steel. Hard. Soft. Electric. inum. Iron, Steel. Bright Ordinary 0000 640.5 192.86 553-97 565-50 8,310 5,650 11,460 4,15 13,000 16,250 ooo 508.0 I52-94 439-33 448.45 6,580 4,480 9,140 3, 2 90 10,400 13,000 oo 402. 8 121.28 348.40 355-65 5,226 3,553 7,400 2,620 8,350 10,430 o 3I9-5 96.18 276.30 282.02 4,55 8 2,818 6,300 2,070 6,650 8,300 I 253-3 76.29 219. ii 223. 68 3,746 2,234 5,250 1,640 5,45 6,810 2 200.9 60.50 I73-7 8 I77-38 3-129 1,772 4,180 1,300 4,3o 5,370 3 159-3 47-97 137.80 140. 67 2,480 1,405 3>36o 1,030 3,550 4,43 4 126.4 38-05 109. 28 ni-57 1,967 1,114 2,700 819 2,850 3,56o 5 100. 2 30.17 86.68 88.46 i,559 883 2,080 650 2,300 2,875 6 79-5 23-93 68.73 70.15 1,237 700 1 ,680 515 1,850 2,310 7 63.0 18.98 54-43 55-56 980 555 1,350 408 1,500 1-875 8 50.0 I5-05 43-23 44.12 778 440 I.07S 324 1,200 1,500 9 39-6 "93 34.28 34-99 617 349 850 255 970 1,210 10 3i-4 9.46 27.18 27.74 489 277 68 S 204 765 955 ii 24.9 7-5i 21.56 22. OI 388 219 545 162 610 762 12 19.8 5-95 17. 10 17.46 37 174 420 128 49 612 13 iS-7 4-72 I3-56 I 3 .84 245 138 340 103 395 493 14 12.4 3-74 10-75 10.98 *93 109 270 84 3i5 394 15 9-9 2-97 8-53 8.70 153 87 220 67 255 319 16 7.8 2-35 6.76 6.90 133 69 1 80 52 205 256 17 6.2 1.87 5-36 5-47 97 55 i35 42 170 212 18 4-9 1.48 4-25 4-34 77 43 107 34 *35 169 *9 3-9 1.17 3-37 3-44 61 34 86 27 no 137 20 3-i 93 2. 67 2 -73 48 27 70 21 90 112 Weight. Pounds per cubic inch. Tensile strength. Pounds per square inch. 50,000 69,000 25,000 78,000 97,000 .322 .0967 .278 .284 6o,000 34,000 85,000 39,000 112,000 140,000 Authority B. B. Co. Pittsburgh Reduction Co. Bridgeport Brass Co. P. R. Co. Trenton Iron Co. 224 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. TABLE IX. AMERICAN AND ENGLISH STANDARD COPPER CABLE. STRANDED, BARE. American. B. &S. English. No. and Diameter of Single Strands. Diam. Cable. Area. Weight. Resistance. Int. Ohms. Inch. mm. Sq. In Sq. mm. Cir. Mil. Lb. loooft Kilog. Kilom Matt. Std.-2oC. 60 F. IOOO Yd. 1000 Ft. Kilom. 4 3 3 I O 00 coo oooo 250,000 cm. 300,000 " 400,000 " 500,000 " 600,000 " 700,000 " 800,000 " i ,000,000 ' ' 7/.o68 7/-095 . 204 2 3i . 261 .285 .291 33 .360 375 .410 .420 .460 .470 55 53 574 59 .630 .707 .728 .819 .828 .891 .909 9 6 3 .990 i-35 1.078 1.144 I. I5 2 5 i,59i 1,830 1,921 2,i35 2,279 2,440 2,698 3-038 3-05 R-B 149 189 238 287 32 380 453 479 586 605 738 766 886 960 1,144 i, 1 35 1,362 i,735 1,815 2,268 2 .372 2,725 2,860 3,180 3,390 3-630 4,015 4,524 4-540 C .320 . 2480 .1967 . 1640 .1560 1237 .1051 .09811 .08100 .07780 . 06440 .06170 05340 04893 .04170 .04141 0345 1 .02748 .02588 .02070 . 02008 .01725 .01666 01479 .01403 .01294 .01185 .01052 01035 A 1.050 .814 6452 5380 5 JI 9 .4058 3446 .3220 2655 2 552 . 2112 . 2O24 1752 . 1606 .1367 .1360 .1132 .O90I .0849 .0679 .0662 .0566 .05468 .04850 .04602 .04250 . 03888 03450 .03396 C .962 493 197.072 197.082 197.092 i9/. 101 317 .244 .194 .161 37/.o8 2 125 37/-ioi .0827 617.092 .0605 6i/. 101 6i/. no 917.098 9i/. 104 .0502 .0423 0357 W B R-C B AUTHORITIES. A Amer. Inst. Elec. Eng.; B Standard of Cable Makers' Ass., Feb. 5, 1903, England; C Calculated from Strand or by Conversion; R J. A. Roebling's Sons Co.; D Heavy figures denote nominal English sizes, but check only approximately with other quantities. ELECTRICAL CURRENT. TABLE X. AMERICAN AND ENGLISH SOLID COPPER WIRE. BARE, WITH ENGLISH AND METRIC MEASURES. 225 American . B. &S. English. S. W. G. Diameter. Area. Weight. Resistance- 20 C. Int. Ohms. (68F.) Matthiessen's St'd. Inch. mm. Sq. In. Sq. mm. Cir. Mils. Pounds. 1000'. Kilog. Kilom. 1000'. Kilom. 7-0 6-0 .5000 .4640 .4600 .4320 . 4096 .4000 .3720 .3648 .3480 3249 .3240 .3000 .2893 . 2760 .2576 2 52O .2320 .2294 . 2120 .2043 . I92O . 1819 . 1760 . l62O . l6oO 1443 .1285 . 1160 1 1'44 . 1040 . 1019 .0920 .0907 .0808 .0720 .0641 057i .0560 .0508 .0480 453 .0403 359 .0320 .0285 . 0280 0254 .0240 .0226 .0201 I 2 . 700 11.785 11.683 10.972 10.404 10. 160 9.448 9. 266 8.839 8.251 8. 229 7. 620 7-348 7.010 6-544 6.401 5-893 5.827 5.385 5.190 4-877 4.621 4.470 4-115 4.064 3-665 3-263 2.947 2. 906 2.641 2.588 2-337 2-305 2.052 1.828 1.628 1.449 1.422 i. 291 i. 219 1.150 1.024 .9116 .8118 . 7229 .7112 .6438 .6096 5733 5 I0 5 .1963 . 1691 . 1662 . 1466 .I3I 8 1257 .1087 1045 .0951 . 0829 .0825 .0707 .0657 .0598 .0521 .0499 .0423 0413 0353 .0328 .0290 .0260 .0243 .0206 .0201 .0164 0130 .0100 .0103 . 00850 .00815 . 00665 . 00646 00513 . 00407 .00323 .00256 . 00246 . 00203 .00181 .00161 .00128 .00101 . 00080 . 000638 .000615 . 000507 .000452 .000401 .000317 126. 670 109.090 107. 100 94.560 85 . ooo 81.070 7O. 120 67 . 40O 61.360 53-400 53- 19 45.600 42.360 38.600 33.600 32.176 27.272 26.650 22.772 21.150 18.678 16.770 I5-659 13-35 I3-035 10.440 8-367 6.818 6.580 5-48o 5-250 4.288 4.168 3-38 2. 626 2.083 1.652 1.590 I.3IO I. 167 1.039 .823 653 517 .412 397 327 .291 .258 .204 250,000 215,296 211,600 186,624 167,805 1 60,000 I38,3 8 4 133.079 121,104 105,534 104,976 90,000 83,694 76,176 66,373 63,504 53,824 52,633 44,944 41,742 36,864 33,io2 30,976 26,250 25,600 20,817 16,509 13,456 13,094 10,816 10,381 8,464 8,234 6,53 5,178 4,107 3,257 3,i36 2,583 2,304 2,048 1,624 1,288 1,022 810 784 643 576 5i 404 756 651 641 565 508 484 419 403 366 320 317 272 253 230 2OI 192 l6 3 159 136 126 III IOO 93-7 79-5 77-5 63.0 50.0 40.7 39-6 32-8 3r-4 25.6 24.9 19.8 15-7 12.4 9.86 9-49 7.82 6-97 6. 20 4.92 3.90 3.09 2.45 2.40 i-95 1.70 1-54 1.22 1126 968 955 840 756 721 623 600 545 476 472 405 377 343 299 286 243 237 203 1 88 165 149 139 118 "5 93-9 74-4 60.6 59- 48.8 46.7 38.1 37-i 29-5 23-3 18.5 14.7 14.1 ii. 6 10.4 9. 22 7-3 5-79 4.60 3-65 3-57 2.89 2-53 2. 29 1.82 .04141 04805 .04893 05550 .06170 .06472 .07481 .07780 .08550 .09811 . 09870 .1151 .1237 .1361 .1560 .1631 I 9 2 5 .1967 .2308 . 2480 .2812 .3128 3346 3944 4050 4973 .6271 .7700 .7908 9570 9972 i. 226 1-257 1.586 1.999 2.521 3-179 3-305 4.009 4.496 5-055 6-374 8.038 10. 14 12.78 13.21 16. 12 18.00 20. 32 25-63 .1360 1576 1605 .1820 2025 .2123 2453 2552 . 2804 .3220 .3240 3773 4059 4465 .5116 5348 6313 6450 757 813 .922 1.026 1.098 1.294 1.329 2.632 2-058 2-525 2.592 3-i4o 3.270 4.020 4.124 5-200 6.560 8.260 10.42 10.84 13-15 14.74 16.59 20.90 26.36 33.25 41.90 43.30 52.8 59-0 66.7 84.0 oooo 5- ooo 0000 ooo oo o oo o I I 2 3 3 4 3 5 4 6 5 7 6 8 9 * 10* ii 7 8 9 12 10 13 ii 12 13 14 15 14* 15* 1 6* 17 16 18 17 18 19 20 21 19* 20* 21* 22 22 23 24* 25* 23 24 * Approximate English Equivalents to American sizes. 226 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The size of wire chosen shows that the line loss is greater than the regulation, which is impractical. As this is a cut and try method, such conditions appear only in checking assumptions. This is corrected by using No. ooo, instead of No. oo, which was tried in this calculation. If the transmission is to be two phase, the calculations are similar except that each phase carries half of the power. For three phase, calculate for a single phase to carry half of the power, and each wire is the size determined. Transposition. Excessive inductive effects can be counteracted by transposing the conductors. In the final transposition the phases must occupy the same relative position as at the beginning. The number of transpositions is arbitrary; for instance, the 52,ooo-volt line, Gaucin-Seville, Spain, has 35 transpositions throughout its length of 75 miles, while the 5o,ooo-volt line from the Uppenborn plant to the city of Munich, Germany, has but three transpositions in its run of 33 miles. Corona Effect. When two conductors in the neighborhood of each other are charged with a very high potential, and after a certain value of electrical pressure has been reached, a bluish glow surrounds the conductors; this glow is distinctly visible in the dark. Coincidently with the appearance of this glow, loss of power begins. A further increase in electromotive force produces a brush discharge, which takes place, not from surface of the conductor, but from the external limits of the luminous envelope surrounding the conductor. This brush discharge results in further augmentation of electrical losses, and is usually accompanied by a hissing or crackling noise. It is intermittent in character and is reddish violet in color. The name "corona" has been given to the combined luminous manifestations of initial glow and subsequent brush discharge. If the electromotive force were carried still higher, the current would jump through the air from one conductor to the adjacent one; but in the case of wide separations, such as are usual, this would require an electromotive force greater than anything either contemplated or necessary. In the case of arcing, the line is short- circuited, and continued arcing would mean a cessation of power supply. Since the appearance of the corona and the brush discharge represent power losses which may be of considerable magnitude, it is desirable that the line be so proportioned and operated at such potential as to avoid them, and this is particularly necessary in the case of transmission systems which supply only a small amount of power, because the energy loss due to the corona effect is independent of the energy transmitted over the line, being fixed by the voltage and not by the energy. There- fore, while the losses might represent a small percentage of a large amount of power, they would be a large percentage of a small amount. A proper investigation of this subject requires first a consideration of electro- static phenomena in general. In a paper l by Lamar Lyndon, who has collated existing data on the subject by authorities, Mershon, Ryan and Steinmetz, the following conclusions are enumerated : 1 The Corona Effect and its Influence on the Design of High Tension Transmission Lines. Philadelphia Section, Am. Inst. E. E., Nov. 9, 1908. ELECTRICAL TRANSMISSION. 227 1. The critical voltage is dependent on the diameter of the conductors, their distance apart and atmospheric conditions, increasing with both diameter and separation of the conductors. 2. After the critical voltage is reached the losses increase very rapidly with increase in voltage. 3. The critical voltage and the magnitude of the losses after it is obtained are affected by atmospheric conditions, and therefore varies with the locality and season of the year. 4. In the same section of country a voltage which is normally below the crit- ical point may be at times above the critical voltage with changes in weather conditions. 5. Smoke, fog, moisture or floating particles increase the losses, while the effect of rain is appreciable. 6. With increase in separation of conductors the regulation and power factor are diminished. 7. A separation of ten feet between wires is as great as is commercial or desirable. 8. The same law applies to cables as to solid wires, the diameter of the cable being effective diameter of the conductor. 9. The losses appear to be independent of the material of which the conductors are composed. 10. The losses and the critical voltage appear to decrease slightly with the frequency. 11. A transmission line should be designed for the atmospheric conditions that may obtain in the locality through which it passes. 12. All corona formation and losses depend on the maximum value of the volt- age waves. Therefore the ratio of the maximum to the mean value should be definitely known to properly design a transmission system. 13. The limiting voltage (effective) which may be applied to a circuit of No. o wires, seven feet apart, with a maximum vapor product of 0.4, and keep down the line loss due to the corona within three kilowatts per mile, is approximately 110,000 volts. 14. High-tension transmission systems working under potentials exceeding 150,000 volts must have the wires covered with some insulating material having a greater dielectric strength than air, or use conductors of abnormally great diameter. The paper shows that under usual atmospheric conditions, which prevail through- out the United States, the following formula is applicable: E = 148,000 X (r X .07) X Iog 10 > in which r E = effective voltage at which the corona will form and loss begin. r = radius of conductors in inches. D = distance apart of conductors in inches. Obviously the voltage applied should be less than that at which the corona is formed. 228 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. From the formula, it is evident that after a separation of 100 times the radius of the conductors has been reached, any further separation is practically negligible in its effect, and with very high potentials the only remedy against corona losses is the increase in the diameter of the wire. A practical example shows, that for a potential of 250,000 volts and a conductor separation of 90 inches, the diameter of the conductors must be 1.5 inches. Such a conductor would contain far too much metal to be easily supported in the air or for the necessary conductivity. Therefore it is believed, that a large jute or hemp core overlaid with a thin sheath of stranded copper or aluminum is the proper conductor to use on high-tension lines; the metal sheath being of such a thickness as will give the requisite cross section to transmit the energy of the system. POLE AND TOWER CONSTRUCTION. For carrying the conductors of a transmission system, the following pole and tower construction is used: i. Wooden poles. Concreted wooden poles. Reinforced concrete poles and towers. Steel pipe poles and towers. Structural steel towers. Of this wide variation wooden pole and structural steel towers are chiefly used; however, the different types will be successively treated. _t_ a i sa- P : ! 1 j r i <"n* 1 WW 1 * i - FIG. i. Types of poles used at the 50,000 V. Line of Taylor's Falls, Minneapolis Power Transmission. WOODEN AND CONCRETE POLES. Wooden Poles. Where the transmission line runs through a section or in the vicinity of a forest district, where poles may be cut, the wooden pole is more apt to be chosen because of its cheapness, little or no transportation, and ease of erection. Another advantage is, that they offer better protection for the community, because they are insulators. The disadvantages in the use of wooden poles are, that they ELECTRICAL TRANSMISSION. 229 decay very rapidly; more insulators are required owing to the short spacing; they are readily destroyed by storms, lightning and fire. Taking the given disadvantages into consideration, which in many instances outstrip the advantages, for instance the first cost, one will find to-day in thickly wooded sections, steel towers carrying the transmission line conductors. . '*- _I_U FIG. 2. Typical Three Phase Circuit Poles. The ordinary type of line construction is a single pole with cross arms as seen in Figs, i and 2. Other types are the " A " frame and " H " frame, both of which require two poles. The latter types must be properly braced and securely bolted, to prevent any deformation due to excessive stresses. These structures are, accord- ing to A. C. Wade 1 who made exhaustive tests on the various types of wooden poles and frames, 3 to 4.5 times as strong as a single pole. Strength of Wooden Poles. For calculating the stresses in wooden poles the fol- lowing formula may be used. M= M = bending moment. R = radius of section at ground level. 5 = strength of wood per square inch. H = height above ground of force applied. This formula applies only to a pole of uniform cross section. In the usual case, the pole is tapered and will break about half-way between the ground and cross arm. Kind of Wood. The kind of wood depends upon the locality through which the line runs, the cost and factor of safety desired. In western transmission lines, spruce and fir are much used, while in California, redwood is prevalent. Cedar and chest- nut are also used; the former is expensive and has great durability. Pine is the most extensively used in pole line construction, owing to its cheapness, but has the dis- advantage of a short life and inferior strength. 1 The Use of Wooden Poles for Overhead Power Transmission, by A. C. Wade. Inst. oj Elec. Engin- eers of Great Britain, May 2, 1907. 230 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. 2A. Wooden Towers of the "A" Frame Type, with Disconnecting Switches. ELECTRICAL TRANSMISSION. 231 Poles come in lengths varying from 30 to 60 feet; the butt diameters vary from 10 to 12 inches, and the top 7 to 9 inches. They are set with one-sixth to one-seventh of their length in the ground, and sometimes as high as one-eighth. Cross Arms. The cross arms are made of first-class wood, such as chestnut, white oak, cedar, redwood, red or yellow pine. Arms 8 feet long are 4 by 5 inches in cross section, 3-foot arms about 3 by 4 inches. For long-span lines, the cross arms are sometimes made 6 by 6 inches. The cross arm must be properly braced to stand stresses applied on same. They are fastened to the pole by means of lag screws or bolts usually f to f inch in diameter. As a protection against splitting, the cross ^ arms are through bolted on both sides of the pin. Experience has proven that certain cross arms split at a stress of 1200 pounds, while when through bolts were used, the cross arm failed to yield at 2000 pounds. For long stretches and on corners, the cross arms must be doubled, to stand the stresses. For medium spans where single cross arms are used, they must face the same direction on alternate poles, while the intermediate cross arms must face the opposite direction. Life of Wooden Poles. The life of a pole depends on the nature of the wood, chemical treatment, and climatic conditions, also character of soil. Redwood and cedar poles under favorable conditions may last 20 years, while the life of chestnut is about . TIG. 3. Arrangement of Cross Arm and 15 years, and that of pine and white Guard Wire 40>000 v Line cedar, 10 years. When chemical treat- ment is applied, the life, of course, will be materially increased. Where poles are set in marshy ground, or ground which is alternately wet and dry, the life of the pole is correspondingly decreased. Preservation of Wooden Poles. To lengthen the life of a wooden pole it must be properly preserved; it must be treated with some chemical compound. The simplest and least expensive way is to paint the top and butt, at least two feet above the ground level, with tar or creosote. More thorough methods of treating poles are done by special concerns, who treat the entire pole. The treatments are more or less iden- tical, that is, the poles are placed en masse in inclosed cylinders and subjected to intense heat (usually steam), varying from 200 to 250 F.; then they are shifted to vacuum cylinders to remove the sap, after which they receive chemical treatment in a cylinder under pressure. The cross arms should be subjected to the same treatment. All cutting and trimming of poles must be done before they are sub- 232 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. jected to chemical treatment. Where drilling or cutting has to be done in the field, after treatment, these places must again be treated with chemicals, which are usually tar or creosote. The top of the pole must be beveled or a cast iron cap provided. Pole Line Construction. To secure correct alignment, the location of the poles must be made with the aid of a transit, and the pole itself must be lined up with a plumb bob. The poles must be correctly distributed according to length. The cross arms may be mounted before the pole is set, and frequently the insulators are mounted at the same time. Poles up to 40 feet in length are usually erected by about 6 men with pikes, while poles above this, are preferably erected by means of a portable derrick and a team of horses, otherwise the services of 10 to 12 men are required. The poles are set either in concrete blocks or directly in the ground; in the latter case, they are frequently provided with a cross member to resist uplift. Where poles have to be set in marshy or swampy ground, it is frequently impossible to set them without very heavy footbracing, consisting of bracing with a semi-crib con- struction filled with ballast. Guys. Where the line is dead-ended, and on sharp turns, poles are guyed where permissible. This practice is not to be recommended with steel towers, because, as the expenditure is made for a structural steel tower, the structure should be made stable enough to resist any strain or stress applied to same; otherwise wooden poles may be erected. The methods commonly employed in guying are to bury a "dead man" in the ground or use some of the patent guy anchors now on the market. The "dead man" is a pole of short length buried in the ground some distance from the pole, and so placed that it lies normal to the direction of the guy wire fastened to it. The patent anchors are so made that little or no excavation is necessary to bury them. Other advantages of patent anchors are the ease of transportation, erection and removing same; several anchors can be readily applied to one guy. Concreted Wooden Poles. The concreted wooden pole has not been used to any great extent. It has only been used, to the writer's knowledge, in Switzerland. It consists of an ordinary pole covered with a layer about one inch thick of concrete mortar. As this coating covers the entire pole, its life is made practically indefinite and the strength of the poles is materially increased, and so fewer poles and insulators are needed. The concrete block setting frequently required is eliminated. Reinforced Concrete Poles. The ordinary reinforced concrete pole is of similar construction as most types of concrete piles. They are made solid or hollow, in cross, square or circular cross sections, and are reinforced by a number of iron or steel rods according to the strength of pole desired. As these poles may be made for any practical strength and length, it is a very convenient type of pole, particularly as they are readily made in the field. A type of reinforced concrete pole, developed and used to some extent in Germany and recently introduced from Switzerland into England, 1 are hollow and tapering, in lengths up to about 40 feet. The machine is capable of making poles of any size and lengths within the limits of 40 feet long and 2 feet in diameter. 1 Electrician, London, July 31, 1908. ELECTRICAL TRANSMISSION. 233 In the process of manufacture, a long sheet iron core is mounted on two trestles, run- ning on rails, so as to be capable of rotational and longitudinal movements. Upon this core, small longitudinal steel rods are fixed. The core is drawn through the machine, which is stationary. Concrete made of clean screened grit and Portland cement is mixed dry in a mechanical mixer and discharged through a chute into a hopper or drum in which rotating paddle wheels regularly discharge the concrete upon a bandage of coarse webbing laid on a conveyor belt, that takes one lap around the core. This con- tinuous traveling conveyor belt is stretched so that the concrete is wrapped about the core under great pressure. As the core issues beyond the conveyor belt, wire is fed spirally around it so as to press into the concrete wrapping, and small rollers then apply great pressure by working on the webbing, the slack of which, caused by the reduction in diameter resulting from this pressure, is taken up by another device. . The core as it issues from the machine is wrapped about spirally with a bandage of cloth. The machine pulls the trestles forward with the suspended core as the concrete is wrapped on, and when the core has passed completely through the machine it is lifted by an overhead crane and laid to one side to harden. It is kept constantly damp so as to secure the maximum hardness. In about twelve hours the interior sheet metal core is reduced in diameter by means of a screw attachment inside and withdrawn. After hardening six days the bandage of webbing is removed, and the whole is then complete for setting. The poles are estimated to have a life of fifty years, and during that time will cost nothing for maintenance. On this basis the total cost of a pole at the end of fifty years is estimated to be $20.00 for the concrete pole, $50.00 for an iron pole and $53.00 for a wooden pole, all including maintenance, repairs and renewals. This is for a 29-foot pole. For a 36-foot pole for transmission service and for the same period, the corresponding figures are: for the concrete pole, $26.00; for the iron pole, $60.00; and for the wooden pole, $68.50. Any desired amount of ornamentation may be given to the poles. Some tests on a pole of this type, 32 feet 9 inches long, showed a deflection of 2f inches with a tensile strain of 15,000 pounds. Tests on a Siegwart, lo-meter pole (39.3 feet) show the following results: with a pull of 880 pounds, the deflection was i.i inch, which increased to 3.5 inches with a pull of 1540 pounds; the permanent deflection being 2.5 inches, owing to the fact that the strain was slowly released. In another test, with a pull of 1540 pounds, a deflection of 2.75 inches was produced; when suddenly released, the top of the pole assumed its original position. A third test, with the application of a pull of 2200 pounds and 6.1 inches deflection, showed a permanent set of 4.72 inches with a gradual release of the pull. In all of these tests no signs of fracture or cracks appeared. Steel Pipe Towers. The simplest form of a steel pipe pole is that of a single pipe with cross arms. For more rigid and higher constructions, three-legged poles have been constructed. Owing to the length, each leg is made in sections and coupled by nipples, the legs being cross-braced by angle irons and rods. Such towers have been constructed for the Ontario Power Company, but this type is now obsolete; it has been superseded by structural steel towers, which are more economical. 234 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. [L 151.75 C.Urt ft 3 *Cntt -Ana Clint of t"Cmi-Arm_ ][il EU4 l 4 {' !' *! I! K 8 f/3J5 Section A-B. ENC.NTN5. FIG. i. Reinforced Concrete Tower and Foundation. REINFORCED CONCRETE TOWERS. Where special long and high spans are required, high towers are necessary. A few of such towers have recently been built of reinforced concrete; for instance those at Brownsville, Penn., 1 erected by the West Pennsylvania Railway Company, for carrying a transmission line across the Monongahela River. The main tower rises 150 feet above its foundation; the second tower is but 55 feet high and located 1 Reinforced Concrete Towers for High Potential Transmission Lines, by F. W. Scheidenhelm. Engineer- ing News, May 2, 1907. ELECTRICAL TRANSMISSION. 230 feet behind the first, and acts as an anchorage, taking the dwct, strain of 'the main span which is 1014 feet. They are reinforced as follows: For reinforcement old T-rails were used. All of the rails forming the vertical reinforcement were of 60 pounds section. The safe unit stresses were cut down to allow for the wear which many of the rails showed. Incidentally, the use of rails solved the problem of the end-to-end connection in the case of the vertical reinforce- ment, for the ordinary spliced-plate joint thus became possible. On the other hand, the large cross section of each rail was a disadvantage. In certain sections of the towers, for instance, it was necessary, under the circumstances, to insert the full cross section of a rail, even though only a fraction of it was required by the stress to be carried. The base section of the main tower contains twelve 6o-pound rails, three being placed at each corner, while the base section of the anchorage tower contains ten rails in the tension side and two in compression. Thus the main tower base con- tains 1.73 per cent of steel, and the anchorage tower base 1.25 per cent. In addition to the vertical reinforcement of rails, a spiral winding of three-eighths cable was used. Two spirals were wound, i foot apart, thus giving a 2-foot pitch. Tie wires secured the spiral winding to the vertical reinforcement, the concrete being i : 2.5 : 5 for the footings and 1:2.5:4 for the tower. STEEL TOWERS. With the introduction of high-tension transmission, wooden poles are fast being substituted by structural steel towers. The majority of transmission lines now in use employ this type of tower. They are made up of angles, channels and lattice construction, and in two, three and four-legged type. All towers for carrying transmission lines have to be calculated to withstand the following general conditions: They must be self-supporting, strong enough to carry the line conductors and to resist the wind pressure on the conductors and tower itself. To this must be added the load due to sleet, and the effects of temperature changes, as well as a factor of safety to guard against accident, such as the breaking of one or more conductors. Wind Pressure on Structures. The records of the United States Weather Bureau are available as an aid in estimating the maximum velocity to be expected in a given locality. These published velocities are not accurate, but must be corrected by a correction table, which may be obtained from the Weather Bureau and is as follows : TABLE I. CORRECTED WIND VELOCITIES. Indicated Actual Indicated Actual velocity. velocity. velocity. velocity. 10 9.6 60 48 20 I 7 .8 7 55- 2 3 25-7 80 62. 2 40 33-3 90 69. 2 5 40.8 IOO 76.2 236 , HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. 'The relation between wind velocity and the pressure produced by the wind on a plane surface normal to the direction of the wind is given by Scholes 1 in the following: M = KV\ M = pressure in pounds per square foot. V = wind velocity in miles per hour. K = constant. Experiments in general, indicate that this form of equation is correct, but differ as to the proper value of K. According to tests by the Weather Bureau, K 0.004, which is probably the most reliable figure. Experiments indicate in general, higher pressures are to be expected at the top of a tower than near the ground, but little is known as to how the pressure is dis- tributed. There is considerable doubt as to what should properly be considered the exposed area of a structure; it is certain, however, that both faces are not, in general, subject to the same pressure. It is usually considered that a reduction factor of 0.5 should be used in figuring the wind pressure per square foot, of projected area of cylindrical surfaces. The wide use which has been given this factor is its principal recommendation. It appears, therefore, that it would be good practice in transmission-line con- struction to specify that the poles or towers should, in addition to their other prop- erties, have strength to resist loads on their members due to a wind pressure of 40 pounds per square foot, with a factor of safety from 1.5 to 2, based on actual test. Such a structure would be suitable for locations where the winds are high; in other locations these figures would be reduced by judgment, aided by a consultation of the weather reports and other such data. Wind Pressure on Conductors. It is not necessary to allow as high a pressure on the conductors of long spans as on the tower itself; however, there are little definite data available for such calculation, but a value of 30 pounds per square foot is usually chosen for localities where the wind velocities are high. In order to keep the stresses of a conductor within its elastic limit, a factor of safety of at least 2 should be chosen. Sleet. Where the transmission line runs in temperate zones, the weight of sleet must be considered. (Specific Gravity of ice is 0.92, or 57.4 pounds per cubic foot.) Although sleet collects in the middle of the span with a greater thickness than near the towers, the usual practice is to allow 0.5 inch thickness, so that the diameter of the cable is increased one inch. As the sleet is apt to remain several days, during which time high wind storms may occur, it is necessary, therefore, that when calcu- lating the wind pressure, the increased diameter must be considered. Foundations. The foundations of towers are made of concrete, or cross members are buried in the ground under heavy ballast. The former is the most common method. They must be heavy enough, or so designed to resist the uplift, equal to the weight of the foundation plus the weight of earth taken at the angle of repose. 1 Fundamental Considerations Governing the Design of Transmission Line Structures, by D. R. Scholes. Am. Inst. of E. E., Atlantic City, N. J., June 30, 1908. ELECTRICAL TRANSMISSION A.E.&M, In designing towers, tests of the soil should be made to determine its holding power and carrying capacity. Many of the foundations for recent installations arc of reinforced concrete. They have the form of an inverted T. The horizontal cross arm gives additional anchorage to resist uplift. Another advantage of the reinforced concrete foundations is, as they are comparatively light they may be made on one or more sections of the line and transported to place. Portability. It is but natural that most transmission lines run through sections of country where transportation facilities are seriously handicapped. Besides this, the towers for modern transmission lines are of such large size that it is difficult to ship them by rail or water, therefore it is necessary to design them so that they can be transported in pieces, known as "knocked down." The members of the towers can be readily transported on the backs of burros or mules. Two-Legged Towers. As stated, towers are designed with two, three or four legs. The two-leg type is made of two channels or I-beams cither in H or A form, and cross-braced. With this type of tower carrying three conductors, the possibility is, that if two or all conductors break, the adjacent towers may be deflected, and like- wise the next towers may be somewhat affected. In order to overcome the possibility of several towers being affected, every fourth or fifth tower may be rigidly guyed. This type of tower for three conductors is sufficiently strong in the transverse direction, and for short spans in general; however, when long spans come into consideration, they are weak in the longitudinal direction. When, how- ever, the tower carries six or more conductors, as, for instance, in the later described Italian tower at Tretzo, where twelve conductors are carried on a tower, the breakage of a few con ductors amounts to but a comparatively small per cent of the total, and the tower is little or none affected. As stated, this tower has been erected to carry spans up to 600 feet. The standard towers of the 75-mile, 52,000- volt Gaucin-Seville, Spain, transmission system, are made in H-frame of two channels with diagonal bracing of flat bars, and accommodate two 3-phase circuits. Another two-legged tower transmission system is that for Moosburg to Munich, Germany, a distance of 32 miles. These towers are of the A-frame type, and carry FIG. i. Two-Legged 37-Foot Transmission Tower, Used in Switzerland and Italy. 238 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. a single 3-phase circuit and two telephone lines. One conductor is carried on the peak of the frame, and the other two on a common cross arm. Before the contracts for the transmission structures were let, tests were con- ducted on (i) wooden A-frame structure; (2) steel tube poles; (3) Mannsmann tube poles; (4) latticed tower of angle iron; (5) I-beam A-frame. i. FIG. 2. Types of Poles and Towers tested before Contract was let for 5o,ooo-volt Transmission System, Moosburg, Munich, Germany. The following table gives a comparison of the tests on the above structures, together with the prices in marks. The structures are tabulated successively as above numbered, and are expressed in the metric system and serve for ready comparison. .*v i o 3 4 5 Res. -Mom. in line direction in cm 3 ooo ^2. 2 Ji. o 2C2 C2 Res. -Mom. in transverse direction in cm 3 Safe load in kilograms 5440 IOO 32.2 l8oO 33-9 2 2OO 353 870 5 o 1800 Cost in marks including two cross arms 7C. 2O 4O. ?? 4C. 2O 8O. 7O 40. oo It will be noticed that the wooden structure was not favorable, especially as the line passes through marshes, and the life of a wooden structure is short. The I-beam structure outstripped the others regarding safe load and price, which is the reason why this structure was adopted. The standard tower is 23 feet to the lower insulator, and is embedded 5 feet deep in a concrete block, and carries three solid copper conductors; the standard spacing is 165 feet. The towers are made up of two I-beams braced at three points; as the cross sec- tions of the conductors vary from 70 to 16 square millimeters, the size of the I-beams varies from 5.5 to 3.5 inches correspondingly. None of them are provided with ELECTRICAL TRANSMISSION. 239 FIG. 3. Aermotor Four-Legged Twin Tower with Ground Wire Pin as used by the Southern Power Co., Charlotte, N. C. This Type is a Combination of Two Three-Legged Towers. 240 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. guy wires. There is a total of 2260 towers, which were erected by two gan The conductors will be of No. o stranded copper supported upon structural steel towers with concrete foundations. 244 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. TOWERS. The towers will be of structural steel, as per blue prints herewith, and will be designed to sustain the assumed loads as follows: The side pressure of the wind will be taken at ij pounds per lineal foot of cable based on a wind pressure of 30 pounds per square foot on a flat surface or 15 pounds per foot on a round surface, acting upon a cable covered with a thickness of sleet equal to its own diameter. The heights of towers will be arranged to provide a minimum clearance of 10 feet over buildings and such wires as may be crossed by this line. Towers at angles will be designed to provide for side strains due to the tension of the cable itself and for the pressure of the wind on the cables and on the tower. Enough insulators will be provided at the angle towers so that the cable will not make an angle of over 7^ degrees on any one insulator. The wind pressure on the tower itself will be assumed at 60 pounds per panel on each half of tower. UNIT STRAINS. The section of members in the tower will be calculated for a unit strain of 24,000 pounds per square inch due to the combination of loads stated above. The strains in compression will be based on the formula, 24,000 96 . DESIGN. The design of towers in general will be as per drawing of 45-foot tower herewith. In towers of greater heights, sections of upright members in lower panels will be increased. The cross- arms of all towers will be designed to resist torsional strains due to the pull of the cable on the tops of insulators. The maximum pull allowable on a single cross-arm tower will be 1000 pounds for each cable, the ties being designed to break at this tension. The cross-arms of towers at dead-ends will carry three insulators for each cable and will be designed to resist the maximum calculated pull due to the assumed conditions of load and sag. Bolted joints in the main members of towers will be designed on the basis of 10,000 pounds shearing per square inch and 20,000 pounds bearing per square inch. FOUNDATIONS. Foundations of all towers will be as per drawing herewith. For towers below 57 feet high, the legs of tower will be extended into the ground to a concrete footing about 3 by 7 feet reinforced at top and bottom as per drawing. The metal below the surface of the ground will be pro- tected by a 6-inch concrete sleeve run in sheet-metal form and extended slightly above the ground sur- face. In wet places this protection will be carried above the surface of the ground, and the portion of corner posts below the splice angle will be lengthened in proportion. Foundations of towers 57 feet high and over will be concrete piers as per drawing. Each leg of the tower will be anchored by two one-inch bolts running to footings. The concrete will be proportioned one part cement, three parts sand and five parts broken stone, or may be one to six cement and clean sharp gravel. The footing under each pier will be 5 feet by 3 feet by 10 inches thick reinforced at top and bottom. Where the ground is firm, no forms will be used around footings. The concrete for piers will be placed in forms of planed lumber. All piers must be carefully leveled up with neat cement at the top, this cement to be put on before the forms are stripped. The foundation bolts will be set with wooden templates and to elevation shown on working drawings. STEPS. Steps about 20-inch centers will be provided for each tower extending from first panel to cross-arm. CONSTRUCTION. The towers will be riveted in the shop as far as practicable. The field con- nections may be bolted. The bolts in the two lower panels will be upset. PINS. The pins will be of malleable iron 18 inches high above cross-arm and designed to with- stand a strain of 2000 pounds in any direction applied to the insulator. They will be attached to the tower with four f -inch bolts. DEAD-END AND SPECIAL STRUCTURES. Where the line is dead-ended or where special structures are required at the sub-stations a separate agreement will be made, these structures not being included in this contract. ,- HANDLING AND STRINGING CABLES. The cable will be delivered to the contractor at convenient freight stations along the line. The contractor will string same upon the towers, using great care that the cable is not kinked or damaged during the operation. The cable will be strung in general with a ELECTRICAL TRANSMISSION. 245 ,;- Malleable Iron ;6x5"YeJtoHrPine &'x5' 'Yellow Pine.. Dressed 6'*y 'Timb to 5%'x i2-foot sag on 48o-foot span at 32 F,, which corresponds to a normal tension of 300 pounds in the cable. Allowance will be made for temperature so that the cable will have a sag of 12 feet at 32 F. Where the spans vary, the sag will be proportioned so that the normal tension in the cable will remain practically the same. The cable will be tied in with wire furnished by the company, the form of tie to be agreed upon later, but these ties will have, as near as practicable, a breaking strength of 1000 pounds. On the double cross-arm equalizing saddles will be provided to insure equal strains being brought upon the insulators. At the dead-end towers, clamps will be arranged for holding the ends of the cable securely. Saddles will be designed to facilitate removing of a defective insulator. CHARACTER OF WORK. All work in shop and field must be carefully and accurately done, and the struc- tures left complete and finished according to the best practice in this class of work. PAINT. All work to have one shop coat of red lead and oil and one coat in field of graphite paint of approved manufacture. All work to be done accord- ing to the directions of the Engineer of the Oneida Railway Company. New York Central Tower. A steel tower of lattice construction for carrying a number of conductors on wooden cross-arms, as in- stalled in connection with the New York Cen- tral and Hudson River Railroad, is seen in Fig. y. 1 "The component parts of the tower consist of the following: Four L's 3 inches by 3 inches by five-sixteenths inch; lacing, one L 2\ inches by i^ inches by three-six- teenths inch (single) ; connecting L's 2\ inches by z\ by one-fourth inch; cap plate of malle- able iron; rivets three-fourths inch in dia- meter. The estimated quantities of material for one pole are: steel, 1340 pounds; concrete, 6.5 cubic yards; timber, 71 feet board measure. The general conditions in installing the lines were as follows: Distance from the center to center of poles on tangents is 150 feet, sag 30 inches; distance on i-degree . Section A-A l*4 Concrete *%** . Top of Foundation 6'ottove Base of Rail in Cut \ lop of Foundation 6' I below Base of Rail on nil. -I' | Concrete I i'4-?r - No.3 Annealed Solid Wire Connection to Elevation.. FIG. 7. Standard Steel Tower of the New York Central Railroad. curve is 141 feet, sag 27 inches; on 2-degree curve, 133 feet, sag 24 inches; on 3-degree curve, 125 feet, sag 21 inches; on 4-degree curve, 118 feet, sag 18^ inches; on 5-degree curve, 112 feet, sag 16^ inches; on 6-degree curve, 107 feet, sag 15 inches. The sag of wires for all spans is computed at 70 F. with no wind. Load on poles: Six-wire circuit No. i, each 0.728 inch diameter, area 400,000 C.M., weight 1 Two Forms of Transmission Towers in New York State. Street Railway Journal, July 14, 1906. ELECTRICAL TRANSMISSION. A.E.&M 2 4 7 UNIV. OF C -yo 'M i ! f i i>! ; i ii li i Iji i|| i if . li . T 1 1 ||' if! i !' ' A t 1 i 1 J II u ft 1 1, 4-1 I i i i \ i K 9 .,--- ' FIG. 1 1 . Type of Tower used in 5o,ooo-volt Italian Transmission System. 1.22 pounds per linear foot; four-wire circuit No. 2, one-fifth inch diameter, area 1,000,000 cm., weight 3.55 pounds per linear foot; three wires, circuit No. 3, each 0.165 inch diameter, area 27,225 cm., weight .074 pound per linear foot, together with one-half inch coating of ice on all wires. The wind pressure is 30 pounds per square foot on the surface of the pole, and on all wires covered with one-half inch coating of ice. Unit stresses: Tension, 30,000 pounds per square inch net section; 30,000 pounds L 2 I = 125 r 3 the compression is per square inch cross section; shear on rivets 22,500 per square inch; bearing on rivets, 45,000 pounds per square inch; maximum bending moment on pole, 2,910,000 inch-pounds; maximum overturning moment of pole, 3,340,000 inch- 248 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. pounds. The painting is one coat of New York Central standard red lead paint on each surface in contact before assembling, and one coat on the entire pole before leaving the shop. Before erection two heavy coats of New York Central asphaltum varnish are added." Luzerne Tower. A type of recent Swiss transmission line construction is given in Fig. 9, and is used in connection with the 27,ooo-volt line at Luzerne, Switzerland. 1 The towers are about 45 feet high measured from the ground up to the lower insulator. As will be seen in detail, Fig. 10, the insulators are mounted on vertical oak members carried by transverse channels and are 3.3 feet apart on the leg. The insulators are fastened to galvanized iron pins by hemp, linseed oil, and shellac. To prevent a line from dropping to the ground, guard angles are provided. Owing to adverse con- ditions, several of the towers had to be placed on a cantilever construction overhanging the lake as seen in Fig. 9. For this purpose a cantilever structure had to be embedded in a heavy concrete block, in order to protect the cantilever and the tower from boulders coming down the mountain slope; heavy masonry abutments were placed on top of the concrete block; a passageway is provided to reach the tower. The total length of the canti- lever is about 30 feet. The spacing of the cantilever poles is 400 feet, while the normal spacing for the land towers is about 200 feet. FIG. 12. Highway and Telephone Crossing for 5o,ooo-volt Line Near Lecco, Italy; also Section Switch House. Brusio Tower. The latest and most prominent transmission line is that of the Brusio Plant, Switzerland, transmitting 50,000 volts for some 88.5 miles in the 1 The 27,000- Volt Transmission System of the Obermatt Power Plant, Switzerland. Electrical Review, June 13, 1908. ELECTRICAL TRANSMISSION. 249 northern part of Italy. Duplicate parallel lines (13 to 16.5 feet apart) run the entire length of the line. As they run over mountains and valleys of great variation in altitude, and owing to great difference in temperatures, frequent storms and atmos- pheric discharges, the peaks of the moun- tains were avoided. The standard tower (see Fig. n) is of angle iron lattice construction, and has brackets to accommodate two 3-phase circuits. Each cable consists of nineteen 2.6 mm. copper wires, the total diameter being 14 mm. (about 0.5 inch). The insulators are of the two-petticoat type, and are, according to Swiss practice, mounted on wood, carried on steel brackets. The tower is about 40 feet high from the ground to the lowest insulator, and spaced on the average, 393 feet; the longest span being 1280 feet, for which special towers were employed. All towers are set in concrete, and designed for. a wind pressure of 70 miles an hour, allowing a stress of 17,000 pounds; the stress of the copper cables is 8500 pounds per square inch, accommo- dating a temperature change of 120 F. Of the 3100 towers erected, there are only four different types employed, weigh- ing from 1250 to 2500 pounds, and cost on the average, $80 each, including founda- tion and erection. The insulators cost $2.60 each, including mounting and the wooden blocks. At present, each line carries only one circuit, amounting to 900 FIG. 13. Aermotor Towers. At the Right a gross tons of copper; the laying of same Medium Weight Tower for Suspended Insu- ... . i lators for the no,ooo-volt Transmission cost $28 per mile of transmission. System of the Grand R ap id s -Muskegon Suspended Insulator Towers. All the Power Co., Mich, above-discussed towers are for pin insula- tors; the suspended type of insulators requires a change in the brackets of trans- mission towers. 1 Types of such towers have been constructed by the Aermotor Company for the Grand Rapids-Muskegon Power Company; 2 they are from 40 to 1 Some New Methods in High Tension Line Construction, by H. W. Buck. Am. Inst. of E. E., June, 1907. 2 The 100,000- Volt Steel Tower Line of the Grand Rapids-Muskegon Power Company. Electrical World, Nov. 2, 1907. 250 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. 60 feet high and spaced about 500 feet apart. Each leg is anchored to a 3-inch angle 7 feet 10 inches long, set in concrete to prevent corrosion, except for about 10 inches at the bottom, which is left bare to pro- vide an effective ground. Stranded copper cables with hemp centers, and having a con- ductivity equal to No. 2 solid wire, are sus- pended from brackets by means of 5 disk link insulators. The steel angles, to which the links of the towers are anchored, were set in concrete at a mixing plant at one end of the line and afterwards transported to the points needed. Each complete anchor weighs about 275 pounds. The concrete envelope is elliptical in section, the axes of the ellipse being 4.5 inches and 6 inches respectively. One 3-inch 4-pound steel channel and several short reinforcing rods were fastened horizon- tally near the bottom of each main angle as anchors. These channels and rods also were set in concrete disks, sheet-iron molds being used for the purpose. Economical Spans. In laying out a trans- mission line it is of foremost importance first, to find out the most economical span, that is, after the size of the conductors has been calculated, not omitting the line loss; the next step is to ascertain the proper spac- ing. Having determined from the foregoing chapter the necessary cross section of copper conductor, the choice of material, whether FIG. 14. Type of Tower for Suspended Insulator, Southern Power Co., Char- lotte, N. C. copper, aluminum, or steel conductors should be used, must be decided. TABLE I. TENSILE STRENGTH AND CONDUCTIVITY OF CONDUCTORS. Material. Tensile strength, pounds per square inch. Conductivity. Copper Aluminum 55. 28,000 100 62 Steel . . IOO,OOO 12 In the foregoing table the conductivities and tensile strengths of conductors for high tension transmission lines are given; in connection with same the market price of the materials must be considered, especially those of copper and aluminum which vary greatly. From this and in conjunction with the design of the tower the most economical span can only be determined by making comparative estimates. ELECTRICAL TRANSMISSION. 251 Line Stresses. The following example, by B. Wiley, illustrates a method for calculating the stresses on steel towers. 1 This problem was worked out for a span at the Homestead Steel Works, Pennsylvania, to cross the Monongahela River. The dimensions and other data are given in the illustration and calculations. The conditions that form the basis of the calculations are as follows: Line voltage, 250; load to be carried, 800 amperes; drop of voltage permissible, 40 volts; necessary size of copper conductor, 1,000,000 circular mils; necessary size of alumi- num conductor, 1,600,000 circular mils (duplicate lines of 800,000 circular mil cable were used for convenience of construction) ; maximum sag allowable at 212 F., 35 feet; maximum wind probable pressure, 40 pounds per square foot; minimum temperature, 20 F.; probable ice coating, one-fourth inch thick. The tensile strength of hard drawn aluminum wire is 35,000 pounds per square inch; its conduc- tivity, 63, as compared with copper at 100; and the coefficient of expansion, .0000231 per degree Fahrenheit. When a wire is suspended between two supports it takes a curve known tech- nically as the catenary. In the case at hand the catenary comes very close to the parabola, which gives the following relations: T = . Bd where T = tension in cable at ends, L = length of span in feet, w = weight per foot of wire, d = the central deflection in feet. Obviously T will be a maximum when w is at its maximum and d at its mini- mum. The wire will have its greatest weight per foot when coated with ice and is withstanding a heavy wind pressure; and the deflection will vary directly as the temperature. The weight of i foot of 8oo,ooo-centimeter Ai cable = .736 pound The weight of one-fourth inch ice coating per foot = .389 pound Total weight per foot = 1.125 pounds Taking the wind pressure at 40 pounds per square foot and as acting on the cross- section of ice covered wire, the pressure per foot is 4.166 pounds. As this force acts at right angles to the weight, the resultant force = Vi.i25 2 + 4-i66 2 = 4.31 pounds, which may be considered the maximum for w. Sd 2 For the catenary curve, L' = L -\ , where L' = actual length of cable, L = length of span, d = central deflection. t / 3 L (L' - L) Transposing, d = y - 1 A Long Span Transmission Line, by B. Wiley. Electrical World and Engineer, April 16, 1904. 252 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. From these two formulae d can be figured for any temperature, the initial sag being 35 feet at 212 F. The following table gives the sag for temperatures between 212 F. and minus 20 F.: TABLE II. SAG AT DIFFERENT TEMPERATURES. Temperature, degrees F. Deflection, feet. Temperature, degrees F. Deflection, feet. 212 35 -o 90 27.1 2OO 34-7 80 26. 4 IQO 33-8 70 25.6 180 33- 2 60 24.9 170 32 5 5 24.1 160 31.8 40 23-3 15 31.2 3 22. 5 140 3-5 20 21-7 130 29.9 10 2O-9 1 20 29. 2 O 2O. I no 28 5 2G 19. 2 100 27.8 2O 183 Substituting in equation (i) the values w =4.31 pounds (the maximum weight), d = 18.3 feet (minimum deflection), L = iooo feet, T (the maximum tension) = iooo 2 X 4.31 8 X 18.3 = 29,400 pounds. The sectional area of 8oo,ooo-C.M. cable is .8 square inch, giving a tensile strength of .8 X 35,000 = 28,000 pounds per cable. Comparing this result with the maximum tension, 29,400 pounds, it is seen that the line will not stand the severe conditions as set down. To relieve the strain, the line should be lengthened in the fall and, to prevent excessive sag, taken up again in the spring. Suppose a range of temperature from 60 F. to 20 F. be taken for the winter. Then the line could be allowed a drop of 35 feet at this maximum temperature, which, by reference to the table, would make the equivalent sag at 20 F. 28.4 feet. Substituting in formula (i), T = iooo 2 X 4.31 8 X 28.4 = 19.000 pounds, or the one setting would give a safe tension on the cable for the conditions noted, though for severe conditions it would be well to give the maximum drop of 35 feet, as the adjustment requires only a few minutes' work. As an example, the maximum strain per cable is 19,000 pounds, or per tower, 4 X 19,000 = 76,000 pounds. The horizontal component due to the wind pressure is transmitted to the foundations and the direct pull to the steel brace rods behind. ELECTRICAL TRANSMISSION. 253 TRANSMISSION LINE TOWERS AND ECONOMICAL SPANS. 1 For any given transmission line there is a certain length of span which is most economical. A determination of what the economical span is, in any case, can only be made by obtaining data showing the variation of each item of cost which changes with the length of span. In a steel-tower line the cost of the tower is probably the most important among those items which vary with the length of span. As the span is made longer, the towers must be made higher and stronger. The purpose of this paper is to describe a method by which the relation between the height, strength, and cost of a tower of given form may be expressed. The application of this method to the problem of fixing the economical span will also be shown. -J FIG. i. FIG. 2. A transmission tower has, in general, three duties to perform: 1. It must have strength to resist wind pressure on its various members. 2. It must have strength to withstand certain external loads due to cables, guys, etc. 3. It must have strength to sustain its own weight. The weight of a given transmission tower may therefore be considered to be made up of three components, each component corresponding to one of these sources of stress. The following equation may then be written for the weight of the structure shown in diagram in Fig. i, W = W w + W, + W 5 (i) in which W = total weight. 1 A paper by D. R. Scholes. Am. Inst. of E. E., Niagara Falls, N.Y., June 26, 1907. 254 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. TF W = weight necessary to provide strength against wind pressure. W L = weight necessary to provide strength against external loads. W s = weight necessary to enable the structure to sustain its own weight. Assume that the structure shown in Fig. i has been designed for a certain wind pressure, and for certain external loads of given amount and manner of application. Each member in the structure may be considered to involve three components of thickness, each component corresponding to one of the three general sources of stress. In determining the value of W w , the stress in each member resulting from wind pressure alone would first be computed; with this as a basis, the component of thickness of each member necessary to sustain the stress due to wind pressure alone would then be calculated. Having determined the component of thickness of each member corresponding to the stated wind pressure, the value of W w would follow directly. A similar method would be used in finding W SL and W s . This method will, perhaps, be made more clear by referring to Fig. 2, which shows in cross section one of the members of the tower of Fig. i. In the figure, t = total thickness, / w = thickness corresponding to wind pressure, / L = thickness corresponding to external loads, t s = thickness corresponding to weight of structure, / sw = thickness corresponding to component W w of the weight of the structure, / SL = thickness corresponding to component W \ of weight of load. It is seen that t = t w + t L + t s (2) and / = / w + / t + / sw + /. (3) since / s = / sw + / SL . The thickness of any other member of the tower may be considered to be divided up into parts in the same manner. Since / s is divided into the parts / sw and SL , a corresponding division may be made in the term W & of equation (i) which gives W = W w + W, + W sw + W^ (4) where TF WS = weight necessary to provide strength to sustain TF W and W aw , and W SL = weight necessary to provide strength to sustain W L and W SL . The structure shown in diagram in Fig. i involves members of three general kinds; namely, beams, struts, and tension members. The bending moment produced in a given beam by a given load W may be expressed by the equation M = CWl, (5) M = maximum bending moment, / = distance between supports, C constant, dependent on the manner in which the load is distributed. The relation between the bending moment and the stress in the most remote fiber of the beam is given by the equation p/?p Jf- , (6) ELECTRICAL TRANSMISSION. 255 M = bending moment, P = stress per unit area in most remote fiber of beam, F = cross-sectional area of beam, k = radius of gyration of beam section, e = distance of most remote fiber from neutral axis. Combining these two expressions, the equation Cle is obtained, which gives the load which the beam will carry, P' being the ultimate strength of the material in the beam. Now if k is the radius of gyration of a given figure, the radius of gyration of a second figure similar to the first but of different size is equal to nk, n being the ratio between corresponding linear dimensions of the two figures. If, therefore, a second beam be considered, exactly similar to the first, but of different size and length, n being the ratio between corresponding linear dimensions of the two beams, the load which this second beam will carry is Pn 2 Fn 2 k 2 2 PFk W 2 W~ = - = n 2 , and - = n 2 . (8) Cnlne Cle W Expressed in words, this relation may be stated as follows: The load which a beam of given form will carry varies as the square of its linear dimensions. The strength of a strut against compressive stress is given by Rankine's formula: P' F W = - (9) , I 2 * ' . W = ultimate strength of strut. P' = ultimate compressive strength of material. F = cross-sectional area. I = length. k = radius of gyration. C = constant, depending on kind of material. And the strength of another strut, exactly similar to the first but of different size and length, n being the ratio between corresponding linear dimensions of the two struts, is P'n 2 F P'F W, W. = - - = , - also - = n 2 . (10) , n 2 l 2 P W I+C ; I+C- n 2 k 2 k 2 Expressed in words, this relation may be stated as follows: The load which a strut of given form will carry* varies as the square of its linear dimensions. 256 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The strength of a tension member is directly proportional to its cross-sectional area; that is, it varies as the square of its linear dimensions. An investigation of the action of a member subjected to torsional loads, similar to those just made for beams, struts, and tension members, would show a like rela- tion; that is, the load which a member of given form subjected to torsion will carry, varies as the square of its linear dimensions. This investigation is not undertaken here, however, because members of this character are little used in transmission towers. Returning to the structure shown roughly by Fig. i. It is usually assumed that the actual pressure on any part of such a structure, produced by a wind of given velocity, is directly proportional to the exposed area of that part. Now the exposed area of any part is, in general, dependent on its length and breadth, but not upon its thickness. It therefore follows that if the structure shown in Fig. 3 is geometrically FIG. 3. FIG. 4. similar to that of Fig. i, in every respect except the thickness of its parts, and is of different size, the ratio between corresponding linear dimensions being n, the load produced on any part of the second structure by a wind of given velocity is equal to n 2 times the load produced on the corresponding part of the first structure by the same wind. It also follows that the stress in any member of the second structure under these conditions, due to wind pressure, is equal to n 2 times that in the corresponding member of the first structure. For the structure of Fig. 3, W = +W' and (12) ELECTRICAL TRANSMISSION. 257 From the foregoing discussion of the relation between the size and strength of beams, struts, etc., of given form, it is evident that and W' w = nW w (14) both structures being calculated for the same wind pressure. Again referring to the equation for beams, P'Fk 2 WCle w =aT> or F = ^' It is evident that if k and e can be kept constant, the sectional area which a given beam must have to sustain a load distributed in a given manner varies directly as the load and directly as the length of the beam. The sections commonly employed as beams are angles, channels, and I-sections. By reference to any handbook of such sections it will be seen that for any of these sections of a given nominal size the area of the beam may vary considerably without producing more than a negligible change in the value of k or e. Hence, if after the nominal size of a beam has been determined, it is desired to vary either the load or the length of the beam, the sectional area should be made to vary directly as the load and directly as the length of the beam. From the formula for columns, P'F it is seen that, if the ratio l/k is kept constant, the strength of the column is directly proportional to its cross-sectional area. From the nature of a tension member, its strength is proportional to its sectional area. Again refer to Fig. i. It is assumed that this structure is subjected to the loads Gp G 2 , G 3 , etc., these loads being placed upon it through cables, guys, or the like. The application of each of these loads will, in general, produce certain stresses in each of the members of the structure. The stress in a given member produced by a given load will be directly proportional to the load, and the magnitude of the stress will depend on the particular position which the member occupies. If a certain system of loads, as G,, G 2 , G 3 , and G 4 , is applied to the structure, the resultant stress in any given part may be considered to be made up of the components .4Gp BG 2 , CG 3 , and DG 4 A, B, C, and D being constants. Also, if each of the loads is multi- plied by a factor r, the resultant stress in any member will also be multiplied by that factor. Moreover, if a system of loads as G p G 2 , G 3 , etc., be similarly applied to another structure geometrically similar to that of Fig. i, but of different size, the stress pro- 258 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. duced in a given member of the second structure by these loads will be equal to that produced by them in the corresponding member of the first structure. In Other words, the stress in any member is dependent only upon the geometrical form of the structure and the amount and manner of application of the loads producing it; and is not affected by the actual size of the structure. Let the structure indicated in Fig. 4 be geometrically similar to that of Fig. i in all respects except the thickness of its members. Let the system of loads, rG } , rG 2 , rG 3 , and rG 4 , applied to this structure be similar to that applied to the structure of Fig. i, but of different magnitude, the ratio between corresponding loads being r. Also let the structure of Fig. 4 be designed for a different wind pressure from that of Fig. i, the ratio between the wind pressures per unit area in the two cases being p. For the structure of Fig. 4, W" = W\ + W'\ + W" sw + W" SL (17) v = <" w + 1\ + <" sw + r SL (18) In view of the relations pointed out between the length, sectional area, and strength of the various kinds of members involved in the structures, it follows that (19) (20) + (21) JP' SL =nW aL +. ... (22) To make equations (21) and (22) strictly accurate, terms must be added to represent the weight added to provide for the strength necessary to take care of each individual increment of weight. This will involve a convergent infinite series in each case. All terms of these series, except the first, are, however, relatively unimpor- tant and will therefore be neglected. Substituting in equation (17) W" = n*pW w + nrW^ + n*pW svl + W SL . (23) This is a general equation, and, given the values of TF W , W L , W svf and PF SL for the structure of Fig. i, it makes it possible to calculate the weight of the structure of Fig. 4 without going through the routine of calculating the stresses in each member and the sizes and weights of the parts necessary to carry these stresses. The application of this formula to the problem of fixing the economical span for a given transmission line is obvious. A tower for a given length of span would be designed to furnish the strengths necessary for that span. The design would be made in accordance with the manufacturing facilities available for producing the structures. The stresses in each member would be carefully calculated and the val- ues of TF W , W L , W sw and TF SL found for the structure. Having found these values, the weight of any similar structure for any length of span could be determined by substitution in equation (23). It is to be observed that this method of treating the case assumes that both wind ELECTRICAL TRANSMISSION. 259 loads and external loads are to be applied to the structure simultaneously. This is usually the case. In other cases, however, the method to be pursued wpuld be similar, but modified to suit the peculiarities of the case. t 3 \ \ I 4 \ x x. *> L^ ~ -. 9 i 3. ) 4 r 6 7 9 10 II it 13 14 4 WIDTH or DAH /// /r/A FIG. 6. It is also to be borne in mind that formula (23) contemplates that variations in the cross section of any member will be made in such manner that the radius of gyration of the section will be kept proportional to n in every case, and also that no appreciable variation from geometric similarity will occur. These assumptions do not involve any appreciable inaccuracy within the range of ordinary practice. 260 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Before the problem of providing steel towers for supporting the cables of a given transmission line can be considered, the general features of the line, its voltage, size of conductor, etc., must be fixed. To show the application of the formula just developed, the following set of general assumptions has been selected as a working basis, and it is believed that they are in accord with average high-grade practice. General Assumptions. System: three-phase alternating current. Conductor: 400,000 circular mils stranded copper. Cross-sectional area 0.3145 square inch. Outside diameter 0.73 inch. Weight per foot 1.22 pounds. Spacing: 7-foot delta, for 5oo-foot span. Minimum clearance: 30 feet between ground and lowest conductor at center of span. Temperature range: 40 F. to uoF. Sleet: 0.5 inch all around cables. Diameter of conductor with sleet 1.73 inches. Weight per foot with sleet 1.98 pounds. Wind pressure: 30 pounds per square foot normal to plane surfaces. Test factor of safety: 2. It is further assumed that at occasional intervals along the line, the structures will be stayed by guy cables in the direction of the line, and that the cost of such staying will not vary with the length of span. To provide in all structures a certain amount of strength against loads on the insulators, in the direction of the line, it is assumed that in the tower for the 5oo-foot span, an unbalanced test load of 2000 pounds will be applied to the top of each insulator pin in a horizontal direction parallel to the line. In explanation of the term "test factor of safety," it may be said that it has become usual for purchasers, in issuing specifications for towers, to require that the structures must show, under actual test, their ability to withstand the loads due to the assumed wind pressures, weights, etc., with a certain factor of safety. In calculating the load to be applied to the top of an insulator pin, for instance, to test it for strength against wind pressure on cables, the effective area of the cable with sleet would be multiplied by the stated wind pressure and by the factor 2. The load thus obtained would then be actually applied to the structure, and its acceptance would depend upon its ability to withstand such tests. In order that the structure may have a certain margin of strength over and above that actually required to withstand tests based on a test factor of safety of 2, the sizes of the members will be calculated with reference to a factor of safety of 2.5 based on ultimate strength. In determining the sag corresponding to each length of span, reference has been had to the curves given in Fig. 9, calculated by Mr. Ralph D. Mershon, and here reproduced through his courtesy. These curves indicate in each case the sag for maximum temperature, this sag being so determined that, when under minimum temperature and maximum wind and sleet loads, the conductor will not be stressed beyond its elastic limit. With the foregoing set of conditions at hand, computations have been made of the cost of each of a series of structures for a 5oo-foot span, these structures being of ELECTRICAL TRANSMISSION. 261 varying width of base but uniform in height. The purpose of these computations is to show the relation between the width of base and cost for such structures, and to obtain an indication as to what ratio between height and width of base is most economical. This series of structures is shown in diagram in Fig. 5. A curve is given in Fig. 6 showing the relation between the width of base and the cost per structure. The cost of each structure has been figured on a basis of $4.50 per 100 pounds delivered in the field. The construction involves standard angle and flat steel sections, standard butt-weld pipe, and some simple forgings. It has been 300 4OO 306 tOO goo FIG. 7. assumed that all parts would be properly galvanized, so no limitation has been made as to the minimum thickness of material, it being simply required that the members be of sufficient strength to meet the conditions laid down. The construc- tion admits of shipment knocked down and bundled, and it is believed that the figure $4.50 per 100 pounds for structures of this class delivered in the field, is quite safe. It will be seen, by reference to the curve in Fig. 6, that the cost of the structure alone is least when the ratio of width of base to height is about i to 4. This con- clusion has reference, of course, only to the span of 500 feet and to the conditions and type of construction adopted. 262 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The width of base of the structure has an important bearing on the cost of the line, aside from its effect on the cost of the tower structure itself, since it affects the cost of foundations, the cost of right of way, and the cost of assembling and raising the structure in the field. Now it is a difficult and uncertain matter to estimate the variation of cost of these items for a general case. Hence a determination of the economical width of base for certain assumed conditions would be of but little interest in the present connection. Application of the Formula. The structure having a width of base equal to one-fourth its height has been selected as a basis for calculations of the weights of towers for longer spans. An investigation of this structure has been made to deter- mine the values of W w) W L , W sw , and W SL , and the following values arrived at: w = 383 . = 813 W sw - 34 W., =60 The table given below gives the results obtained by means of the formula for a series of towers similar to No. 4 in Fig. 5, but for spans up to 1000 feet. Since all towers in the series are to be for the same wind pressure, p is equal to unity in each case. Also, r is proportional to the length of span, since the external loads are due to wind pressure on the cables and the weight of the cables. TABLE I. Span. Sag. Height. H r? 4 P r W 3 />^ W r.rv L n'pu'svt n*ru'sL w" 200 2.0 32.0 o. 780 0.608 Q-475 0.370 i 0.4 182 254 13 15 464 300 4-5 34-5 0.832 0.692 0.576 0.479 i 0.6 221 406 16 25 668 400 7-5 37-5 0.915 0.837 o. 766 0.701 i 0.8 294 596 24 40 954 500 II. 41.0 I. 00 I. 00 I.OO I.OO i I.O 383 813 34 60 1290 600 15-5 45-5 I. II 1.23 i-37 i-5i i I. 2 5 2 3 1082 5 2 89 1746 700 20.5 5-5 1.23 1-51 1.86 2.28 i 1.4 73 1405 78 128 23M 800 26.0 56.0 1.366 1.86 2 -55 3-46 i 1.6 97 8 1778 118 179 3053 900 33-o 63.0 1-537 2.36 3.62 5-57 i 1.8 I 3 86 2250 190 255 4081 IOOO 40.5 7-5 1.72 2.96 5-9 8.76 i 2.O 195 2800 298 356 5404 These results are shown graphically in Fig. 7 by the curve which gives the relation between the length of span and the cost of towers per thousand feet of line. By properly representing to this same scale the cost of insulators, foundations, right of way, etc., per thousand feet of line, corresponding to the various lengths of span, and adding the corresponding ordinates of all these curves, a resultant curve will be obtained. This resultant curve will show the relation between the length of span and cost of those items which vary with the length of span, and it will therefore indicate the economical span for the assumed conditions. ELECTRICAL TRANSMISSION. 263 A curve showing the cost of insulators per 1000 feet of line is given in Fig. 7, the insulators having been figured at $5.00 each, erected on the tower and with the con- ductor secured to them. The curve in Fig. 7, showing the cost of foundations per 1000 feet of line, has reference to the type of foundation shown in Fig. 8, and to the following method of calculation. It is a usual assumption that the strength of a foundation against a force tending to pull it out of the ground is directly proportional to the weight of the foundation plus the weight of earth contained in the figure ABCD. If the foundation in Fig. 8 has strength to resist a resultant force P, a second foundation, exactly similar to it but of different size, would have strength to resist the force n 3 , pn being the ratio between corresponding linear dimensions of the two foundations. Now it seems fair to assume that the cost of such a foundation would vary directly as its volume. The cost of the foundation would therefore vary directly as the resultant force which it is capable of resisting. Referring to some experiments made at Chicago on a foundation similar to that of Fig. 8, and to the records showing the actual cost of the foundation in the field, ready to receive the structure, the following basis for calculation was obtained: Resultant force sustained by foundation 24,000 Ib. Cost of foundation $15-25 By calculating the resultant force which would come upon the foundation from each of the structures given in Table I, and making the cost of foundation for each structure proportional to that force, on the basis of the data above given, the curve showing the foundation cost per' 1000 feet of line given in Fig. 7 was obtained. It is to be observed that this curve is quite flat, indicating that the foundation cost does not vary to any great extent as the length of the span is varied. The curve of combined cost of towers, foundations, and insulators was obtained by adding the respective ordinates of the curves giving the separate costs of these items. This curve indicates that, for the assumed conditions, a span of about 425 feet would be most economical. It is to be observed that in the foregoing solution the determining factors are the tower cost and the insulator cost. As the price per insulator is increased, the econom- 264 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. ical length of span would be increased, and vice versa; in other words, the higher the voltage the longer the span should be. For a low-voltage line the economical span would be somewhere between 300 and 400 feet, as far as the methods of calculation here employed can determine. Each structure in this case would, however, be a very light affair. It is probable that in the average case a somewhat longer span would be decided upon in order to give each structure greater individual strength and thus make it safer against damage due to external causes. too' ,,o' FIG. 9. In case it is desirable to impose limitations of this sort, the formula must be modi- fied accordingly, by subdividing the component of weight into parts; as, for instance, by letting _ ^ + + + where W Gl , W G2 , W G3 and W G4 are components of weight corresponding to the loads G l) G 2 , G 3 and G 4 respectively. These loads may then be made to vary at different rates, or some may be kept constant and the others varied in such manner as may be desired. Suppose, for example, it is assumed that each structure should have strength to resist the loads due to the breakage of any two conductors. These loads would be the same regard- less of the length of span, whereas the loads due to wind pressure on the cables would vary according to the length of span. These assumptions will, in general, tend to make the economical span longer. ELECTRICAL TRANSMISSION. 265 INSULATORS. Pin Insulators. With high-tension transmission systems multi-petticoat porce- lain insulators are extensively used. However, recently a new type, known as the " link " insulator, has been developed. The petticoat types are made in several sections cemented together, and with exceptionally large sizes they are frequently cemented in the field. When this is done, care must be exercised to prevent the cement from being chilled while setting. The cement mixture must be a fine rich mortar free from impurities. Porcelain for electrical purposes 1 is a mixture of ground flint or silicon dioxide and feldspar (KjO.A^Og.SiO^, potassium aluminum silicate, raised to the vitrifying temperature, that is, to a temperature sufficiently high to melt the feldspar and per- mit it to unite the particles of flint into a perfectly homogeneous body of uniform electrical and mechanical strength. FIG. i. Insulator for 6o,ooo-volt used on the Kern River Transmission System. FIG. 2. 5o,ooo-volt Insulator used on the Taylor's Falls Transmission System. Besides the electrical stresses, the insulators must be made strong enough to with- stand mechanical stresses imposed on them by the span. Mechanically, insulators can be designed for any load by the proper disposition of material. Good electrical porcelain has a crushing strength in excess of 15,000 pounds, and tensile strength ranging between 1500 and 2000 pounds per square inch. Fig. i shows an insulator as installed at the Kern River Plant. 2 The specification called for a guarantee of a ioo,ooo-volt test from the groove to the pin for half an hour, under a precipitation of i inch in 5 minutes, at an angle of 30 degrees from the vertical. The assembled insulator was required to withstand under a wet test a potential of 150,000 volts for 30 seconds, and the separate parts are guaranteed to withstand a voltage of 25 per 1 "High Voltage Insulator Manufacture," by Walter T. Goddard. Canadian Society of Civil Engineers, Dec. 19, 1907. 2 Kern River Plant No. i. Electrical World, Aug. 31, 1907. 266 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. cent in excess of the normal proportion of over-voltage test. The insulators are guaranteed to withstand a side strain of 4000 pounds, and actually fail at approxi- mately 9000 pounds. A cross section of the insulator used on the Taylor's Falls 5o,ooo-volt transmission system 1 is seen in Fig. 2. It is known as the S.W. No. i, made by the Locke Insulator Manufacturing Company. It consists of four parts held together with neat cement. These insulators are shipped in crates assembled, but without pins. The crates were provided with holes just the size to take the pin. The cement- ing in of pins was done before the insulators were uncrated, the crate thus serving the purpose of a template to hold the pins in position while the cement dried. The insulator, as seen by the drawing, is 12^ inches high by 14 inches in diameter over all. The four parts were tested before assembling with a 6o-cycle, 2oo-kilovolt-ampere testing set. The top piece withstood a test pressure of 60,000 volts; the second shell 40,000 volts; the third shell, 50,000 volts; and the fourth inner shell or center, 50,000 volts. The assembled insulator without cement was tested at 120,000 volts. The experience of noted Swiss and Italian engineers in the development of high- tension insulators is summed up in an article entitled, "Present Status of European Practice in Transmission Line Work," 2 of which the following is an extract: " The experiences of Mr. Charles Brown, one of the earliest workers in this field, and a pioneer of many of the more recent developments, gives a word of warning to those engaged in insulator testing, claiming that most insulators have a very pro- nounced fatigue effect. Though an insulator may stand a given tension for 15 min- utes, it may possibly break down at this tension if it is maintained for two hours. He recalls the great difficulty always experienced in attempts to deduce reliable results from any testing, except actual use on the transmission line. He further states that very little trouble is experienced from heavy rainstorms or climatic conditions causing insulator breakdowns, the trouble being almost entirely mechanical and due to lightning. Such mechanical defects as have been experienced are thought to be due largely to the use of cement for connecting the petticoats of the insulators together, since great difficulty is experienced in obtaining a cement which does not swell with increase of temperature and thus fracture the insulator. In Switzerland, use is made of sulphur cement (when sulphur is used, the pins must be galvanized) and plaster of Paris, both of which have given satisfaction. The latter, however, being some- what porous, must be varnished with shellac wherever it is exposed to the air at the outside of joints, etc. " For fixing the pins of the insulators, tow or hemp is used, which is twisted around the end of the pin, the whole being then dipped in asphalt or shellac and screwed into the insulator. Mr. Brown states that no splitting or fracturing of insulators occurs with this method of fixing the pins. " It is further stated that no good results have been obtained with the Fox cement, which is thought to have too high a coefficient of expansion, producing splitting 1 The 50,000- Volt Line of the Taylor's Falls, Minneapolis, Power Transmission. Electrical World, Sept. 7, 1907. 2 Electrical World, Dec. 22, 1906. .itu uiimiui f\ A I INIY OF CA ter country ELECTRICAL TRANSMISSION. troubles. In this connection, however, it should be noted that somewhat thinner ' * m m M. f * *" insulators are used in Switzerland than in Italy, and the engineers of the li have found very little trouble due to this cause. " Other Swiss experts have referred to the difficulty in the manufacture of perfect insulators, pointing out that minute holes in the enamel in the surface which cannot be seen by the eye, may pass test, and then cause breakdown after some months' installation. " Some again claim to have overcome the difficulty due to insulator splitting, by using no cement at all, the insulator being made in two or more pieces which are tested independently and then screwed together and the whole rebaked. " They have also paid considerable attention to the exact shaping of the edges of the insulator petticoats, a rounded edge being considered very bad, since in a heavy rainstorm it will cause the water to run under and drop on the surface of the lower petticoats. They at present very much favor a petticoat slightly turned up near the edge to check the velocity of the running water and then dropping to a sharp point on the extreme edge, which seems to prevent this running under. By this means it is considered that the effect of rainstorms may be considerably reduced. For all transmission lines for electromotive forces of 40,000 volts and above, iron poles are preferred, and if the insulators have not more than two petticoats, wooden cross-beams are used; if three or more, then they are placed directly on the iron poles. " In Italy, Mr. Guido Semenza, whose name is associated with the well-known Paderno transmission and numerous others throughout the country, referring to his early experiences, stated that on the Paderno line, after some experiment, his conical type of insulator was chosen and a triple petticoat was used; the line being, however, finally completed with two of these cemented together as one six-petticoat insulator, The dimensions of this were, height over all, 7 inches; diameter of petticoat, 6f inches. In other plants the Paderno type has been superseded by a much lighter insulator, but it has lately returned to favor and is in general use. The original Paderno insulator is shown in Fig. 3." This type of insulator has also been installed on the 5o,ooo-volt system of the Brusio plant. It is fastened to the pins by hemp and shellac as above described. The pins are mounted on wooden blocks, secured to the steel cross-arm. On many European high-tension tranmissions systems, the insulators are made in one piece to eliminate cementing. Such insulators are employed on the 35,000- volt transmission system of the Urfttalsperre plant, Germany. Suspension Insulators. A new type of insulator successfully used in recent practice is the suspension type. The advantages of this type over the pin insulator are given by Mr. Goddard as follows: 1 "The reason for using suspended insulators is largely a matter of cost, since it is entirely possible to build porcelain insulators of the conventional type of sufficient size to successfully operate at any voltage, but the extreme height and diameter of 1 High Voltage Insulator Manufacture, by Walter T. Goddard. Canadian Society oj Civil Engineers, Dec. 19, 1907. 268 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. a pin-type insulator for 100,000 or 150,000 volts makes the cost prohibitive. A sus- pended type of insulator has several advantages which it is well to understand before going into details of design. Of paramount importance is the unit formation making it possible to increase the effective insulation whenever it is desired to raise the line voltage or wherever it seems desirable to present extra leakage surface because of FIG. 3. Paderno Type Insulator for 40,000 to 50,000 Volts. FIG. 4. Suspended Insulator, showing the Progressive Breaking Down be- tween the Beginning of Actual Leak- age and the Maximum Arc. salt fogs or smoke from railways and factories. Many lines start operation at much lower potential than designed for, because the initial load is light, and the potential need be increased only when regulation demands it. With the pin type of insulator there is no alternative but to invest at the start in the largest insulators which the line will ever need, whereas in the suspended form, additional units may be intro- duced whenever the growth of power business warrants an increase in potential. In the pin type of insulator the nearness of line wire and pin must always prove a weak point for lightning assault as well as an aggravator of line-charging current difficulties. The suspended type gets away from both difficulties by a wide separa- tion of line conductor and supporting structure. Incidentally the position of the ELECTRICAL TRANSMISSION. 269 conductor below the cross-arm permits the supporting structure to act as a lightning rod and so to relieve the line of much lightning stress. " Mechanically, provision must be made to prevent the swinging conductor from coming too near the tower structure, but the extra length of cross-arm necessitated by this feature is more than compensated for in cost by the fact that there are no twisting strains upon the arm. Insulator unit formation presents another very posi- tive advantage in the matter of breakage. When a shell of a pin-type insulator FIG. 5. Method of Suspending Insulators. FIG. 6. Detail of a Dead Ending. becomes cracked or broken the whole device is rendered worthless, as it is utterly impossible to break the cement joint forming the bond between shells. Further, the cracking of a shell, especially an inner shell, may cause immediate shut-down, or at least shut-down during the first severe rainstorm. On the contrary, the break- ing or cracking of one of the shells of a suspended-unit type insulator takes away but that one unit from the series; thus, in the case of a five-unit, 100,000- volt insulator, a broken unit reduces the total strength but twenty per cent. 270 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. " The underhung system of insulation works out with pleasing directness and simplicity, and its comparative cheapness argues for its wide adoption for the higher voltages. The cost of such insulators, as at present manufactured, ranges from $1.60 to $2.00 per unit, depending on the nature of the fittings. "At least two 14-inch units would be required for 60,000 volts, and as good 6o,ooo-volt insula- tors can be secured for prices ranging from $1.70 to $2.30 each, the question of the use of sus- pended units for voltages below 75,000 to 80,000 is largely one of safety factor and investment. " The foregoing has given little which could be used in the determination of the proper insulator to use for any particular voltage, and it is quite in point to add here, that every case is special. Insulators well suited to one locality are out of reason for use elsewhere. A single transmission line of less than 100 miles in FIG. 7. Suspended Insulators as used length may easily pass from high, clear in the no.ooo-volt System of the , . r ,. Grand Rapids Muskegon Power Co., mountam air to f ggy> sm ky surroundings General Electric Co. which are a constant menace to continuity of service. Again, the cost of complete immunity may well be balanced against cost of possible shut-downs." FIG. 8. Dead-End Insulator of the Suspended Type. General Electric Co. FIG. 9. Single Desk of Suspended In- sulator. A detail of the Locke Suspended Insulator is given in Fig. 4, while two other illustrations give the method of application of same. The dead-ending scheme as proposed for all towers has the advantage that in case of a breakdown of a conductor ELECTRICAL TRANSMISSION. 271 FIGS. 10 and n. Application of Cooke Strain Insulators. FIG. 12. Diagram of Anchor Insulators. FIG. 13. Insulating and Rolling Support for Long Spans, Tofwehult-Westerwik Trans- mission System, Sweden. 2/2 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. adjacent sections do not have to be slacked down in order to repair the break. The disadvantage is that the line conductor has to be cut up into sections, the current being by-passed by a loop or " jumper " as seen in the illustration. A suspended type of insulator is used on the 1 10,000- volt transmission system of the Muskegon Grand Rapids Power Company. They are of the type described by Mr. Hewlett in his paper presented at the last convention of the American Institute of Electrical Engineers, at Niagara Falls, June 26, 1907. Figs. 7 and 8 show the construction of the members of this insulator. Five of these insulators are suspended in series to insulate the line. The diameter of each porcelain link is 10 inches, and the rated voltage that each link will withstand is 25,000, although the links arc over where wet at approximately 60,000 volts each. Fig. 9, while showing the interior FIG. 14. Porcelain Base Insulator Pins. FIG. 15. All Steel Insulator Pins. construction, also shows the form of petticoat on the insulator used in a horizontal position as a strain insulator at curves and at intervals to anchor the line. The spans of this line are on the average about 150 feet. The conductors consist of stranded copper cables with hemp centers, having a conductivity equal to No. 2 solid wire. This line was designed for 100,000 volts, but recently the voltage has been raised to 1 10,000.' Strain Insulators. Strain insulators must be placed at the beginning and end of lines, and at all sharp turns to take up the pull of the spans which ordinary insulators cannot stand. Such insulators as seen in Figs. 10 and n are usually held at top and bottom. Two or more ordinary insulators, when used in connection with an anchoring 1 Editorial, Engineering Record, Aug. 15, 1908. ELECTRICAL TRANSMISSION. 273 FIG. 17. Insulator with Single Tie. 6o,ooo-volt Insulator of the Ontario Power Co. FIG. 16. Detail of Iron Insulator Pin, used for Insulators seen in Figs. 17 and 18. FIG. 18. Insulator with Clamp. 6o,ooo-volt Insulators of the Ontario Power Co. Upper and Lower Insu- lators are of Uniform Size. 274 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. device, may take the place of a regular strain insulator. An arrangement of such insulators has been installed in the transmission system of the Pennsylvania Rail- road Company. 1 The insulators are placed one behind the other, each couple forming an anchor insulator and practically eliminating all danger of wires breaking at the insulators, as shown in Fig. 12. The two cross-arms provide accommodation for 1 8 wires, but at the present time only 10 are in use. The height of the first row of insulators from the floor of the platform is 7 feet 6 inches, and of the top row, 10 feet 6 inches. This gives an abundance of room for linemen to work with safety. The insulators are placed longitudinally 3 feet between centers, with the exception of the third insulator from each end, which is placed 3 feet 6 inches from the second. This variation permits of symmetrical spacing with reference to the upright supports of the cross-arms. They measure 6| inches across the umbrella and are 6 inches high, each insulator having two petticoats. They were designed for a voltage of 25,000, but the present service pressure is only 13,000 volts. Insulator Pins. Wooden pins have been extensively used on transmission systems up to 30,000 volts. However, in localities subject to salt storms, heavy sea fogs and near chemical manufactories, there has been more or less pin burning without regard to the type of insulator used, or the voltage of the system. It has been reported 2 that certain plants using only 440 volts, have at times great trouble from the burning of pins, although io,ooo-volt insulators are used. To overcome such difficulties, pins are provided with porcelain bases. Nearly all wooden pins are made of locust, oak or eucalyptus, and are chemically treated for preservation. All standard wooden pins are i inch in diameter and have 4 threads per inch. A modification of the wooden pin to-day more commonly used, is an iron bolt with a wooden top which screws into the insulator, and is provided with a porcelain base. Such a pin is illustrated in Fig. 14. A still more satisfactory type for high-tension transmission is an all-steel pin as seen in Fig. 15. It is made in various modifications and used on wooden cross-arms as well as steel. Sometimes the steel pin is made in a single piece, either forged or cast, a type of which is illustrated in connection with the insulators of the Ontario Power Company (see Fig. 16). Method of Tying Conductors. The tying of the line conductor to the insulator is done in different ways; such as the patent Clark system or as illustrated in the accompanying illustrations. Figs. 17 and 18 show the methods adopted by the Ontario Power Company. One shows aluminum tie wires; and in the other, the conductor is held in place by a clamp on a cast iron cap cemented to the insulator. These insulators are 14 inches in diameter and are designed for the 6o,ooo-volt trans- mission system. They are about 27 inches high including the steel cast pin, and weigh about 80 pounds. Section Switches. Section switches are located where duplicate lines run parallel and near each other, so that, in emergency cases, defective sections may be easily cut out and by-passed. They are also located at places where, in the near future, 1 Steel Transmission Towers on the Jersey Meadows. Electrical World, Dec. 14, 1907. 2 Burning of Wooden Pins on High Tension Transmission Lines, by C. C. Chesney. Am. Inst. E. E., March, 1903. ELECTRICAL TRANSMISSION. 275 TV 7ZX r FIG. 19. Line Disconnecting Switch. Attach Grounding Cabll of (Jpermtlng Poll. FIG. 20. Open Air Section Switch. I fel FIG. 21. Outdoor Two Break Section Switch used on the Pacific Coast. *NLs Standard^ Cpnorclion Bolt "* 4*H*rd Wood 276 FIGS. 22 and 23. Typical Wall Outlets. ELECTRICAL TRANSMISSION. 277 taps will have to be made. The common section switch is nothing more than a disconnecting switch such as used in the generating station, but usually larger and heavier, and mounted on line insulators. They are usually placed directly in the line, similar to that shown in Fig. 20, which has been installed in a transmission system FIGS. 24 and 25. Typical Wall Outlets. Locke Insulator Company. FIG. 26. Provo, Permanent Wall Outlet. Three Concentric Tubes of Fibre Conduit. in Auburn, N.Y. Where section houses are located in long transmission lines, the section switches are preferably placed in the houses. Another type of section switch as used on the Pacific coast, is seen in Fig. 21. It will be noticed that the blades of the switch revolve and can be operated from the ground. 278 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Wall Outlets. Where high-tension wires leave or enter a building, the outlet must be protected against the weather. This is accomplished in American practice, by building hoods over the wall opening as seen in some of the accompanying illus- trations. Other methods are, by inserting insulating bushings in the wall. Common Continental practice is to lead the conductor through a hole of a glass panel; the hole is from one-fourth to three-eighths of an inch larger than the line conductor. Insulators are placed on both sides of the panel so that the section of the conductor going through the wall is always straight. There are no hoods or other protection provided, and it is a simple and inexpensive yet efficient arrangement. Many Western plants and those on the Pacific coast have, for a wall outlet, a tile pipe, 18 to 24 inches, provided with a plate glass cover. Tile pipes and bushings used for wall outlets must be set on the slant, so that collected moisture can drain off outdoors. BIBLIOGRAPHY. HIGH TENSION POWER TRANSMISSION. 1906. A Series of Papers and Discussions at the Meetings of the A. I. E. E. under the Auspices of Committee on High Tension Power Transmission. Vol. I, 1906. THE ELECTRICAL TRANSMISSION OF ENERGY. A. V. Abbott. 1907. LINE CONSTANTS AND ABNORMAL VOLTAGES AND CURRENTS IN HIGH-POTENTIAL TRANSMISSIONS. E. J. Berg. Proc. Am. Inst. E. E., September, 1907. THE GROUNDED NEUTRAL, WITH AND WITHOUT SERIES RESISTANCE IN HIGH -TENSION SYSTEMS. P. M. Lincoln. Proc. Am. Inst. E. E., September, 1907. LINE CONSTRUCTION FOR OVERHEAD LIGHT AND POWER SERVICE. Paul Spencer. Canadian Elec- trical News, September, 1906. A NEW TYPE OF INSULATOR FOR HIGH TENSION TRANSMISSION LINES. E. M. Hewlett, Proc. Am. Inst. E. E., June, 1907. SQME NEW METHODS IN HIGH-TENSION LINE CONSTRUCTION. H. W. Buck. Proc. Am. Inst. E. E., June, 1907. HIGH-TENSION INSULATORS, FROM AN ENGINEERING AND COMMERCIAL STANDPOINT. C. E. Delafield. Electrical Review, N.Y., Sept. 28, 1907. THE CORONA EFFECT AND ITS INFLUENCE ON THE DESIGN OF HIGH TENSION TRANSMISSION LINES. Lamar Lyndon. Am. Inst. E. E., Philadelphia Section, Nov. 9, 1909. TRANSMISSION LINE CROSSINGS OVER RAILROADS. Ralph D. Mershon. Railroad Gazette, Feb. 7, 1908. THE CENTRAL STATION DISTRIBUTING SYSTEM. H. B. Gear. Electrical Age, January, 1908. AMPERES IN ALTERNATING-CURRENT CIRCUITS. A. D. Williams, Jr. Electrical World, Aug. 8, 1908. GROUND DETECTORS AND THEIR CONNECTIONS. James T. Coe. American Electrician, December, THE DISTRIBUTION OF PRESSURE AND CURRENT OVER ALTERNATING-CURRENT CIRCUITS. A. E. Kennely. Harvard Engineering Journal, November, 1905. SIMPLE DIAGRAMS FOR THREE-PHASE POWER CALCULATIONS. Alfred Still. Power, March, 1906. PRESENT STATUS OF EUROPEAN PRACTICE IN TRANSMISSION LINE WORK. Electrical World, Dec. 22, 1906. SOME POWER TRANSMISSION ECONOMICS. Frank G. Baum. Proc. Am. Inst. E. E., May, 1907. HIGH-PRESSURE DIRECT CURRENT TRANSMISSION. Electrical Review, London, June 7, 1907. THE TRANSMISSION OF ELECTRICAL ENERGY BY DIRECT CURRENT. J. S. Highfield. Inst. E. E. March 7, 1907. ELECTRICAL TRANSMISSION. 279 POTENTIAL STRESSES AS AFFECTED BY OVERHEAD GROUNDED CONDUCTORS. R. P. Jackson. Proc. Am. Inst. E. E., April, 1907. EARTHING THE NEUTRAL, WITH AND WITHOUT SERIES RESISTANCE IN HIGH TENSION SYSTEMS. Paul M. Lincoln. Proc. Am. Inst. E. E., September, 1907. THE GROUNDED NEUTRAL. F. G. Clark. Proc. Am. I. E. E., September, 1907. RECENT PRACTICE IN ELECTRICAL TRANSMISSION OF POWER. W. B. Esson. Engineer, London, Dec. 14, 1906. PRESSURE RISE ON HIGH TENSION TRANSMISSION LINES. E. Hudson. Electrical Engineer, London, April 26, 1907. EXPERIMENTS WITH HIGH POTENTIALS. Electrical World, Jan. 26, 1907. LINE CONSTANTS AND ABNORMAL VOLTAGES AND CURRENTS IN HIGH-POTENTIAL TRANSMISSIONS. Ernst J. Berg. Proc. Am. Inst. E. E., September, 1907. THE LOCALIZATION OF EARTH LEAKAGES ON A THREE-WIRE NETWORK. Electrical Engineer, London, April 12, 1907. PROTECTIVE DEVICES FOR HIGH TENSION TRANSMISSION CIRCUITS. J. S. Peck. Institute of Electrical Engineers, March, 1908. DETERMINING THE SIZES OF ALTERNATING CURRENT LINE WIRES. N. T. Carl. Power, July 28, 1908. LONG DISTANCE ELECTRIC TRANSMISSION OF POWER. L. S. Bruner. Proc. Engr's Club of Phila., April, 1908. A TRANSMISSION LINE CONSIDERED AS A MECHANICAL STRUCTURE. W. T. Ryan. Electrical World, Feb. 29, 1908. COMPENSATION OF PRESSURE VARIATIONS ON ALTERNATING CURRENT NETWORK SUPPLYING MOTORS. A. Heyland. Electrician, London, April 24, 1908. THE TANGENTIAL SYSTEM OF SUSPENDING OVERHEAD TROLLEY AND TRANSMISSION WIRES. Robert N. Tweedy. Electrician, London, May 15, 1908. SOME FEATURES OF EUROPEAN HIGH TENSION PRACTICE. Frank Koester. Electrical Age, December, 1908. SOME POWER TRANSMISSION ECONOMICS. F. G. Baum. Proc. Am. Inst. E. E., May, 1907. ONE-PHASE HIGH -TENSION POWER TRANSMISSION. E. J. Young. Proc. Am. Inst. E. E., May, 1907. INDUCTIVE DISTURBANCE IN TELEPHONE LINES. Louis Cohen. Proc. Am. Inst. E. E., May, 1907. TRANSMISSION-LINE TOWERS AND ECONOMICAL SPANS. D. R. Scholes. Proc. Am. Inst. E. E., May, 1907. POTENTIAL STRESSES AS AFFECTED BY OVERHEAD GROUNDED CONDUCTORS. R. P. Jackson. Proc. Am. Inst. E. E., April, 1907. HlGH-VOLTAGE DIRECT-CURRENT AND ALTERNATING-CURRENT SYSTEMS FOR INTERURBAN RAILWAYS. W. J. Davis, Jr. Proc. Am. Inst. E. E., August, 1907. CHAPTER IX. SUBSTATIONS. GENERAL ARRANGEMENT. Location of Substations. Substations or receiving stations are designed to act as distributing centers for light and power. Where a source of direct current is desired, the substation houses, rotary converters, or motor generators set. The substations as a rule are located as near as possible to the center of gravity of their systems of distribution. This cannot always be done, as the demands on the station vary in certain sections during the different seasons, particularly in street railroad work. In many cases, to help out in the latter instance, portable sub- stations are run to the centers of increased demand, and remain until the load on the line can be taken care of by the substation proper. Size of Units. The size of units, may they be generators, transformers, converters or motor generator sets, depends upon the capacity of the plant and upon the load factor. Care must be taken to have one or two units in reserve, depending upon the size of the plants. American practice is to overload the units 50 per cent, while European, only 20 to 30 per cent. Fixed rules as regards the size of the individual capacities cannot be laid down, as they all depend on the nature of loads at various times. Each case has to be individually treated, which is best done by plotting load curves for the day, week, and possibly for the month; in many instances it is necessary to plot curves for the whole year, particularly for suburban railways, and heavy lighting loads, where great fluctuations occur during certain seasons of the year. Arrangement of Substation. In transformer substations, transformers are usually located in fireproof compartments, provided with iron rolling shutters. To facili- tate inspection and repairs, a track is run in front of the compartments, so that the transformers may be readily removed on a small truck and then shifted to the repair room. Such transformers are provided with wheels and ratchet, resting on a rack. The truck is also provided with a rack so that the transformers are easily shifted to the truck without the aid of an overhead crane. Where converters are used, the transformers are not housed in compartments, but are set opposite the converter on the main floor, where they are handled by an overhead crane. Figs, i and 2 show typical arrangements of transformers, switchboards, con- verters, etc. It will be observed that the substation of the Connecticut Railway and Lighting Company is provided with a large storage battery; for such auxiliaries separate apartments are required. 280 SUBSTATIONS. 281 282 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. A very novel arrangement of a substation is that at Piattamala, Italy. The transformers are arranged in two banks each, to accommodate twelve I250-K.W. 7ooo/5o,ooo-volt single-phase, oil-cooled transformers. The current is received at one end of the building; the two sets of low-voltage busses are located on a FIG. 2. Cross Section of Waterbury Substation. mezzanine floor above the passage between the banks of transformers. The current leaves at the other end of the building. Ventilation. In laying out a substation, it must be borne in mind that even the normal operation of the transformers and converters will considerably increase the temperature, therefore provision must be made for good ventilation. This is par- ticularly important where oil-cooled transformers are used. It is unnecessary to provide any auxiliary means for heating in compact substations which carry a station load factor equal to average practice, and run 24 hours per day. Drainage. Where air-blast transformers are used, the air chambers must be waterproofed and the ducts located at such an elevation that water will not stand in the bottom. If this is not done, the transformer may be damaged by the warm air from the blowers picking up moisture and depositing it in the transformers not in service. Where any cable comes into the station, underground, the entering conduit must be sealed, and suitable drainage provided, so that water cannot leak through these open- ings. Where oil-cooled transformers are installed, it is good practice to provide a pit of sufficient capacity to hold the oil from several transformers, and also drainage- piping from the oil drain cocks on the transformers to the pit. These pipes must be of ample size, so that the oil can be drained off very quickly in case of emergency. Air Compressor. An air compressor is an item which must never be overlooked in a substation as well as in a power house of any considerable size, as the life of all electrical apparatus depends to a very great extent upon cleanliness. SUBSTATIONS. ^ m^^s^^^^^ 284 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. 4. Plan and Sectional Elevation of Small Substation with Single-phase Oil-insulated Self-cooling Transformers and Hand-operated Oil Switches, 11,000 or i3,2Oo-volt, Overhead High Tension Lines. SUBSTATIONS. 285 * H O bO CO .o "S 4- cfl J2 CO I i o 286 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Frequently, a portable motor-compressor with a small storage tank is used. Where stationary compressors are used, the air must be piped from the tank to various points in the station, where cocks must be provided for the attachment of a rubber hose. An air-pump governor is a convenient means for keeping the air in the storage tank at a constant pressure. TRANSFORMERS. Types of Transformers. Transformers are made either single phase or three phase, and in shell or core type. The core type is more extensively used abroad and made three phase. In America, besides the core type, the shell type is widely used, but chiefly in single-phase design. The advantage of a three-phase transformer is its greater compactness and lighter weight, resulting in a considerable saving in first cost of transformer itself, and a saving in floor space of about 30 per cent. The connections of the transformers are simpler and fewer than in three single-phase units. The method of winding and insulating is practically the same as in single-phase transformers. FIG. i . Shell Type Transformer in Process of Construction, General Electric Co. FIG. 2. Core Type Transformer in Pro- cess of Construction. The difference between a shell and core-type transformer is best illustrated in Fig. i. In the shell type, it will be noticed that the coils are almost entirely sur- rounded by the sheet steel laminations, and are known as " pancake " coils. To secure mechanical strength, the conductors must be rectangular in cross section SUBSTATIONS. 287 and of sufficient width. As the " pancake " coil is difficult to wind for a small transformer, the core type is preferable for small sizes. In the core type, the core is made of sheet steel laminations and almost entirely surrounded by the winding, giving it great stability and mechanical strength, for which reason it is used for small as well as large size transformers. The coils are made of flat copper strips wound on edge. The secondary or low potential wind- ings of these transformers are usually divided into two or more coils connected in series, on each of the vertical legs of the core. The coils in the secondary windings of pole transformers have their leads run to a common terminal block. By interconnecting the terminals with jumpers, a limited FIG. 3. 2000-K.W., 4ooo/6o,ooo-volt Gen- eral Electric Co.'s Shell Type Trans- former. FIG. 4. i6oo-K.V.A., 4o,ooo/5oo-volt 3- phase Water Cooled Oil Transformer, Ocrlikon Co. range of voltages may be impressed on the service mains. The high tension wind- ings are arranged for series or multiple connection with other transformers. Characteristics of Transformers. Aside from the reliability and safety of operation of a transformer, the most important electrical features are the efficiency and the regulation. Although good regulation and good efficiency are always to be desired, the relative importance of the two is determined by the local conditions under which the transformer is to operate. 288 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Where power is expensive or is used at only short intervals, the efficiency of the transformer, especially at light loads, is of great importance, but where the power is cheap, the efficiency as a rule is not so important a feature. On account of the low cost of the power and the double transformation of the potential necessary, the important feature of transformers designed for use on high voltage circuits, is the regulation and not the efficiency, especially not the efficiency at light loads. Regulation of Transformers. In large transformers for long-distance transmission, close regulation is of even greater importance than in the ordinary small transformer for lighting circuits, as the drop in the line is often of considerable magnitude; and WESTINGHOUSE AIR BLAST TRANSFORMER. 550 K.W. 10500 VOLTS. 3000 ALTERNATIONS. 50 75 JOO PEK CENT LOAD CHART I. 125 150 with raising and lowering transformers, the transformer drop occurs twice between the generator and the load. This drop is generally increased when the power factor of the load falls below unity, as is usual in power work. It is therefore particularly necessary that close regulation be obtained in the transformers designed for trans- mission work, especially if they are to be used for supplying inductive loads. Transformers for such service are usually designed to have good regulation for loads of any power factor. The second set of curves, Chart I, illustrates the operating characteristics of a transformer designed for transmission work where the power factor of the circuit is low. SUBSTATIONS. 289 The regulation of a transformer depends largely upon the resistance drop, and the inductive drop within it. The former is fixed by the amount of copper loss at full load, the latter by the number of turns in the winding and the relative position of the coils and the space between them. In a transformer designed for good regulation, it is therefore essential to have the two windings as close together as possible, a result obtainable only by using the best insulating materials to separate them, and to have low copper loss at full load. Some stations supply a service where the transformers are connected to the supply mains continuously, and current is taken from the secondary for only a few hours during the day. In such a case, the iron losses are incessant and the copper losses intermittent. The transformer must be of such a design that the iron losses are the lowest possible, otherwise the total work received during the day will greatly exceed the work given out. The ratio of the work given out to the work received during the day is called the all-day efficiency. o 3 D u OC *. 1.8 1.- 1.4 1.2 1.0 .8 .6 .4 .2 z ui u u. u. u w. 100 90 80 TO 60 50 40 30 20 10 WESTINGHOUSE AIR BLAST TRANSFORMER. 100 K.W. 12000 VOLTS. 3000 ALTERNATIONS. x L EFF ICIEtv CY / 7 ,^- ' ouu HON_ ^ *T-FO / X Ji^4 . *--^ 1 ^^ ^ at FIG. 6. Forced Oil Circulation for cooling Oil-Insulated Transformers. consequently the use of oil maintains a uniform core loss and a superior insulation. The oil in the transformer is cooled, either by its natural gravity circulation or by means of submerged coils through which water is circulated. The amount of water necessary for cooling the oil depends on the temperature of the incoming and outgoing water. Theoretically, each kilowatt loss will give up 57 B.t.u. per minute, or, in other words, 57 pounds of water are raised i F. In practice, however, the amount of water required varies with the design, and the amount of water necessary can be obtained from the manufacturer. 1 See paper, Y or A Connections of Transformers, by F. O. Blackwell, presented at aoth Annual Con- vention Am. Inst. E. E., Niagara Falls, N.Y., July i, 1903. SUBSTATIONS. 293 Another design of transformer, instead of using water coils, the upper part of the transformer is provided with submerged radiating ribs cooled by circulating water. Transformers of this design, having a capacity of 1250 K.V.A., 7700/50,000 volts, have been installed at the Italian substation at Piattamala. FIG. 7. Method of Cooling Circulating Water for 6750-K.V.A. 6600/66, ooo-volt 3-Phase Siemens-Schuckert Transformer, Molinar Plant, Spain. In order to keep the temperature rise of this transformer below 45 C., 5 gallons of water per minute at a temperature of 15 C. are required. For a 25 per cent over- load for 6 hours, 10 gallons are required; for 2 hours at same overload and using 5 gallons, the permissible temperature rise is 60 C. Forced Oil-Cooled Transformer. Another method of cool- ing the transformer oil, is by forced circulation, and has the advantage of doing away with the cooling coils. Instead of the oil being cooled in the transformer, it is cooled outside in a cooling device which works on the same principle as a cooling pond or a surface condenser. A very elaborate system of this kind is given in Fig. 6. 1 It will be observed that besides water pumps, a set of oil pumps is necessary, while with the water-cooled system, only water pumps were required. Where sufficient head is obtain- able, the water-pumps may, of course, be eliminated. With a forced-oil circulation, the transformers are small and less expensive, due to the elimination of cooling coils; however, an extra cooling system is necessary, the cost of which in small plants will outstrip the reduced cost in trans- transformer plants over 4000 K.W., the forced-oil system seems FIG. 8. Air-Cooled Transformer formers. In preferable. Air-Cooled Transformers. In the early type of transformers, the cooling was done by natural air-draft, or forced draft. The latter is still very much in use. The cores 1 Forced-Oil and Forced- Water Circulation for Cooling Oil Insulated Transformers, by C. C. Chesney. Am. Inst. E. E., April, 1907. 294 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. of the transformer are incased, through which air is forced by means of a blower. Where there are a number of transformers, they are preferably set over a common duct and supplied with air from a blower at either end, one being kept in reserve. They are most conveniently operated by motors. The volume of air required for air- blast transformers depends on the outside temperature, as well as the entering and 1 1 1 V I " PPiTg 1 / L j L / x / z ~ J '$ / / 2 >7//? CHAM&ER / FIG. 9. Arrangement of Air Blast Transformers. discharge temperature, and to great extent on the design. Under normal load and continuous operation, the temperature rise must not exceed 35 to 40 C; at 25 per cent overload, 50 C. ; at 50 per cent overload, 60 C. The temperature rise is taken by thermometers or calculated from the increase in resistance. The pressure furnished by the blowers depends on their size and the length of the ducts. High- voltage transformers usually require higher air pressure. Fig. 10 gives approximately the air pressure required for different capacity transformers. A more complete table on this subject is found in Table I. TABLE I. AIR REQUIRED FOR TRANSFORMERS. Horse Total kilowatt trans. Size of units kilowatt. Cubic feet air required per transformer per minute. Cubic feet air required for all transformers Cubic feet air furnished by standard blower set. Oz. Press. Freq. Mot'r . Size blower, inches . Speed blower. power to drive blower full vol. and per min. pressure. 900 IOO 45 4,050 6,OOO } 2 5 5 75 2-5 40 5 800 ^ 60 40 900 1800 200 90x5 8,100 8,000 1 25 55 75o 4 40 55 800 60 5 900 2700 300 1125 10,125 10,000 1 25 55 75o 5 40 55 800 60 55 720 4500 500 1625 14,625 14,000 ' 1 25 75 500 5-5 40 70 600 60 70 720 6750 75 1875 16,875 2O,OOO \ 25 90 500 12 40 80 600 60 80 600 7500 1250 2800 16,800 20,OOO I 25 90 500 12 40 80 600 60 80 600 SUBSTATIONS. 60 CYCLES. 295 Volls. 2200 66OO 1 1 000 16500 22OOO 33000 K\v. 100 2OO 2 5 300 375 500 1000 1500 2OOO 2500 3000 * Oz. Pres sure sure a Oz. Pres Oz. Pres sure i 25 CYCLE TABLE Volts. 22OO 6600 IIOOO 16500 22000 33000 Kw. IOO 125 \ Oz. Pres sure 15 2OO 1 250 1 300 1 375 500 3 4 Oz. Pres sure 75 1 JOOO 1 1500 2OOO I Oz. Pres sure 2500 3000 FIG. 10. Air Required for Transformers. 296 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. With air-cooled transformers, more or less dirt is carried along with the air, and deposited along the various air passages. Much of the dirt may be obviated by keeping the air passages to and from, the transformer closed when not in use. A frequent cleaning of the transformer windings with a blast of compressed air will improve conditions, and at the same time remove a possible fire risk. Oil-insulated transformers should have the oil drained off once in a while, and all evidences of sediment removed. The emergency drains must be cleaned out at the same time. In a recent Italian plant at Lomazzo, I25O-K.V.A., 4 2, 000/11,000- volt trans- formers without casing, are placed in masonry compartments, through which the air is forced from ducts beneath. Each compartment, reaching to the ceiling, is provided with a ventilator. The reason given, is, that ready inspection can be made without removing the core, although provision is made for doing so in case of extensive repairs. The guaranteed and test efficiencies of these transformers is as follows: Regulation at cos< = i.oo full load i per cent Regulation at cos = 0.80 full load 3 per cent Regulation at short circuit 3 per cent Efficiency full load .97 per cent Efficiency half load 9.65 per cent The operation of the blowers is included in the above-named efficiencies. CONVERTERS. Rotary converters are installed for transforming alternating current into direct current; however, they may be otherwise used. They may be supplied with direct current and deliver alternating. They may be connected to alternating mains and operate as simple synchronous motors, or connected to direct current mains and operated as simple direct current motors. There are a number of other electrical and mechanical connections which can be applied, but the main purpose is to serve as a means of conversion from alternating to direct current. In general appearance and construction, the rotary converter resembles a direct current generator to which a set of collecting rings has been added. The field is composed of a cast iron yoke with inwardly projecting poles of laminated steel. It may be either shunt or compound wound. The armature consists of a slotted, laminated core with embedded coils, with the addition of taps or leads to the collector rings. The method of cross-connecting the armature windings, which has been a means of securing superior performance in direct current generators, is applied to rotary converters with equal success. This is an effective way of preventing sparking at the commutator, as it insures uniform field strength under all the poles. Voltage and Frequency. The ratio between the voltages at the alternating and direct current ends of a given rotary converter, is approximately constant, and cannot SUBSTATIONS. 297 be changed by altering the speed or by using a rheostat. Therefore, any alteration in one voltage will proportionately alter the other, and vice versa. In most rotary converters, the voltage on the collector rings of a two-phase machine is about seven- tenths of that at the commutator, and the voltage on the collector rings of a three- phase rotary converter is about six-tenths of that on the commutator. FIG. i. Rotary Converters in Sub-Station " d," Albina, Portland Railway Light and Power Company. Thus, a two-phase converter receiving alternating current at approximately 385 volts will deliver direct current at 550 volts, and a three-phase converter receiving alternating current at approximately 330 volts alternating current will deliver at 550 volts direct current. In installations supplying three-wire lighting systems, or where it is necessary to obtain two voltages, for the operation of variable speed, direct current motors, a special neutral-wire connection is required for use in conjunction with the positive and negative leads on the direct current side of the rotary. If a conductor be con- nected to the middle points of the secondary windings of the transformers, which supply the alternating current for a two-phase rotary, it will be found that the E.M.F. between this conductor and either of the direct current terminals is equal to one-half of the E.M.F. between those terminals. In this way no volts can be secured from a 22o-volt machine. A similar arrangement for three-phase rotaries is secured by employing the interconnected star system of connections for the secondary winding of the transformers, the neutral lead being connected at the common junction point of the secondary windings. The same relation exists between speed, number of poles and frequency that is 298 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. found in alternating current generators. The product of the number of poles by the speed, in revolutions per minute, is equal to the number of alternations per minute. Rotary converters can be had for any frequency up to 60 cycles per second. The standard frequencies are 25 and 60 cycles, the former being generally used for rail- way service, and the latter when a combined railway and lighting service is operated, or where power is obtained from existing 6o-cycle transmission plants. Phases. The alternating current may be applied to the collector rings of the rotary converter in the form of either single, two, three or six phase currents. How- ever, single-phase currents are seldom used, and the majority of machines are wound for either three or six phases. It is now general practice to wind 25 cycle units for railway work for three phases when under 500 K.W. capacity, and for six phases when of 500 K.W. or above. In 6o-cycle machines, those under 300 K.W. are wound three-phase, and 300 K.W. and above, six-phase. Six-phase winding in the larger machines is highly desirable, because it reduces the heating, and increases the effi- ciency and stability of operation. Field Connections. For highly fluctuating loads, such as those on interurban or small city railway systems, rotary converters with compound-wound fields are prefer- able. The compounding of such machines differs from that of direct current gen- erators; the direct current voltage is dependent upon the alternating current voltage, and is therefore affected by the line drop, generator voltage, etc. The standard com- pound winding is designed to give 600 volts at both no load and full load, that is, a flat compounding, with 5 per cent drop between the generating station and the rotary converter substation. This flat compounding is obtained with the assistance of reactive coils connected between the stepdown transformers and the rotary con- verter. The compound-wound field coils are of the ventilated type, that is, the winding is in two layers separated by a space through which air is blown by the centrifugal action of the armature, thus greatly increasing the radiating surface of the field spools and reducing the temperature rise. The utility of reactance will readily be understood when it is borne in mind that a rotary converter is simply a transforming device, and the ratio of the alternating voltage impressed to the direct voltage delivered, is approximately a fixed quantity and independent of the field strength. Therefore, any increase in the direct current voltage, or overcompounding, must be secured by a proportional increase of pressure at the collector rings, and it is the presence of a reactance in circuit which brings about this desired result. Or, in other words, by inserting this reactance, the line itself has been compounded, and thus made self-regulating. In order to make use of single-pole switchboards and to do away with pedestals, each compound-wound rotary converter is fitted with a panel mounted on the machine frame, and carries the equalizer switch, as well as a switch used in connection with the shunt provided for the adjustment of the series winding. In order to make the rotary converter panels of the switchboard of the same polarity as the direct current feeder panels (positive or trolley polarity), the series fields are connected on the negative or ground side of the circuit between the armature and the rail returns; this makes all switches on the machine frame panel of the negative SUBSTATIONS. 299 or ground potential. A double-throw field break-up switch is also included in the machine equipment, by means of which the polarity of the machine may be reversed if necessary, in starting. Where the load is practically constant, such as in heavy services, shunt-wound rotary converters are more generally used. In such cases, the fluctuations of the load are so low that they can be followed by hand control of the field rheostats. Starting of Converters. There are different ways to throw converters on the line starting from rest. One way is to supply the converter with a separate starting motor or one mounted on the shaft, which brings the machine up to the desired speed, and with the application of an automatic synchronizer, the converter is thrown on the line at the first instant of synchronism, and the starting motor cut off. Another method is by supplying alternating current directly to the slip rings, and is impressed upon the windings at a lower voltage than is used after the machine is run up to speed, and is in synchronism with the source of supply. This low-voltage alternating current is obtained from the stepdown transformer by means of switches, which connect the armature to low-voltage taps on the transformers at starting, and establish the connections to the full-voltage taps when the machine has reached synchronous speed. The most common method used in street railroad work, is to start the converter as a direct current motor, supplied with current from the trolley mains. By adjust- ing the strength of the field windings, synchronous speed is reached and the machine thrown on the line. This method does not require any special starting apparatus, such as starting motors, or transformers with special taps. Its disadvantage lies in the fact that it is dependent on the direct current supply of the system. Hunting. Rotary converters, to give the best service, must run in exact synchro- nism with the supply current, but it frequently occurs that the speed of the generator is not exactly uniform; and in such cases, the rotary will tend to follow the fluctua- tions of the generator speed, resulting in a surging action of the rotary armature, alternately above and below synchronism. This is commonly known as "hunting," and often assumes disastrous proportions where no provisions are made for counter- acting this effect. Since hunting is an oscillation in the relative positions of the converter ajiead of the generator at one instant and behind the next the correc- tive currents in the circuit due to the oscillations are first in one direction and then in the opposite. The effect of this varying current is to strengthen the leading pole tip when flowing in one direction and to strengthen the lagging pole tip when flowing in the other direction, thus constantly changing the distribution of magnetism over the pole face, and, in effect, causing the magnetic flux to continually shift back and forth across the pole face. . The fact that hunting is always accompanied by a shifting field makes possiLI,- an effective method of reducing it. Most rotaries are provided with heavy copper grids, that surround each pole face and extend across it, embedded in one or more slots. The function of the copper grids is to act as dampers, preventing the relative position of the converter armature being changed, by the corrective currents, more than the initial change in the generator. The action is essentially a damping one, 300 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. and is the same as that of the copper magnet damper used in galvanometers, and is analogous to the action of a dashpot on an engine governor. To consider the action of the damper in detail, assume that the generator speed momentarily increases. This causes a difference in phase of the generator and converter E.M.F.'s. The difference in the instantaneous E.M.F.'s due to the difference in phase will cause a corrective current to flow in the circuit, that will FIG. 2. Automatic Regulators. distort the field and accelerate the converter armature. The shifting flux cuts the copper grid and generates in it eddy currents, which retard the converter armature. The retarding action of the eddy currents occurs only while the relative positions of the converter and generator armatures are changing, i.e., while the magnetic flux is moving across the pole face. The eddy currents, therefore, do not act as a constant opposing force to the corrective currents, but as a true damping force, becoming zero whenever the generator and converter armatures revolve exactly in synchronism. SUBSTATIONS. 301 Induction Regulator. The induction regulator consists of a polyphase transformer with primary movable with respect to the secondary. The construction of core and winding resembles that of an induction motor. The primary is connected across and the secondary in series with the line. By shifting the position of the primary winding of the regulator, the secondary voltage delivered to the alternating side of the converter may be raised or lowered without opening any part of the circuit, and the voltage on the direct current side thus varied. When of sufficient size, the regu- lator may be operated by a small motor. This method of regulation may be employed to overcome small and infrequent fluctuations in the line voltage in lighting, electro- lytic and similar service. Compounding. It is generally understood that there is a certain adjustment of field strength which gives a minimum alternating input for a given direct current output, and that an overexcited field sets up a leading current in the line, while an underexcited field causes the line current to lag. As change in the field strength alone cannot appreciably affect the direct current voltage, the ratio between the two E.M.F.'s remaining practically fixed, the only way to vary the direct potential is to vary the alternating potential at the collector rings. It is, however, possible by a proper proportion of series excitation and the provision of sufficient inductance in the supply line, to produce a change in the voltage at the collector rings, resulting in a corresponding effect at the direct current terminals. The conditions for rotary converter compounding are, therefore, a series winding on the field connected to assist the shunt, and inductance in the line between the generator and converter. The series winding of a rotary converter does not directly increase the direct current voltage, as in a direct current generator, but acts indirectly with the aid of inductance in the supply circuit. Rotary converters which are compounded to give a constant or increasing voltage with increasing load, maintain a practically uniform voltage at the generator ter- minals, and therefore do not produce the drop in voltage which usually occurs when the generator load increases. This enables a practically constant voltage to be maintained on other circuits supplied by the same generator, independent of the variations in load upon the rotary converter. Both lighting and railway loads may thus be supplied simultaneously from the same bus bars, provided the proper compensation is effected, and the fluctuations in load do not cause an appreciable variation in the speed of the generators. In some systems, alternating current is supplied to rotary converters 'at a distance from the power house, while other converters, located in the power house, are sup- plied with current from the same generators. If the converters in the power house are to be compounded to give a rising voltage with increase of load, it is necessary to provide self-induction either in transformers or in choke coils, placed between the bus bars and the converter. Reactances. To enable the direct current voltage to be altered by the field rheo- stat or automatically by compounding, which calls for a corresponding change of the alternating current voltage, a phase reactance-coil is provided between the low-ten- sion windings of the transformer and the converter. Without such a reactance, the 302 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. maintenance of the same voltage at full load as at no load, involves excessive leading and lagging currents, and consequently excessive heating in the converter armature, unless the resistance drop from the source of constant potential is small, or the natural reactance of the circuit is unusually high. If the armature field is weakened, a lagging current is set up, which causes a drop in the reactive coil. If the field is strengthened, a leading current is set up which gives a rise of voltage in the reactive coil. Under heavy load, the series field of a compound converter tends to produce leading currents, which tendency is practically balanced by the reactance, improving the power factor of transformers, lines and generators when loaded. FIG. 3. Typical Continental Motor Generator Substation. Vienna Railway System. Motor Generators. Motor generator sets may be used in place of rotary con- verters, if the line voltage is not too high; the motor of the set may be directly con- nected to the line, thus eliminating transformers. The advantages of using a motor generator are, that no synchronizing apparatus is necessary, the voltage of the gen- erator bears no relation to that of the supply, same as in a converter, and it may be adjusted through a wide variation. The disturbance known as " hunting " is unknown in the motor generator, when an induction motor is used as the driver, and no skilled attendants are necessary. The disadvantages are, that the efficiency is from 4 to 7 per cent less than that of a transformer-converter set, and they cost SUBSTATIONS. 303 more. With what voltage a motor generator can be used without the use of a step- down transformer, depends entirely upon the design of the motor. The accompanying illustrations give an idea of motor generator sets as used in Europe where they are most extensively employed. Fig. 4 shows the interior of a substation at Steghof, Switzerland; each motor generator consists of a 34O-K.W., 265o-volt, alternating current motor, coupled to a 30O-K.W., 575-volt, direct current generator, running at 490 R.P.M. Frequency Changers. A frequency changer differs from a motor-generator set in the following respects: The driving motor must be a synchronous motor and the FIG. 4. Motor Generator Sets. Substation " Steghof," Switzerland. generator, an alternating current machine. The generator has more or less pairs of poles than the motor, depending upon the frequency desired. Sometimes an induc- tion motor is substituted in place of the generator and made to rotate above or below its rated speed. The alternating current line is connected to the rotor of the motor, and if the rotor operates above its normal speed, the frequency is increased; if below, the frequency is decreased. Frequency changers are not very much used; however, when they are employed, they are used in plants which run in parallel with others of different frequencies. A notable example in the use of frequency changers is in Montreal, Quebec. 1 1 Frequency Changers at Montreal, by B. A. Behrend, Electrical World and Engineer, Feb. 13, 1904. 304 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. " The City of Montreal, Quebec, obtains electric energy for power and lighting from three plants, which have three different frequencies. The Chambly power plant supplies alternating current at 66 cycles, the Lachine Hydraulic Land & Power Company generates alternating current at 60 cycles, and the Shawinigan Water & Power Company generates alternating current at 30 cycles. Since the consolida- tion of these three plants, a compromise frequency of 63 cycles has been adopted. " At Shawinigan Falls there are installed two 3750-]$.. W. generators operating at 180 revolutions and 30 cycles, and generating 230x3 volts two-phase. By means of transformers, the two-phase current is changed to three-phase 55,000 volts, which is transmitted from Shawinigan to Maisonneuve, a distance of 85 miles. In the sub- station at Miasonneuve, a suburb of Montreal, the 3O-cycle, three-phase current is stepped down from 44,000 to 2300 volts. The long distance line between Shawini- gan Falls and Maisonneuve is operated at the potential of 55,000 volts at the gener- ating end, and 44,000 volts at the receiving end. The three-phase, high potential current is reduced by three transformers from 44,000 to 2300 volts. " The five groups of frequency changers change the current from 2300 volts three-phase 30 cycles to 2300 volts three-phase 60 cycles. Fig. 5 is the 3O-cycle motor; FIG. 5. Outline of io65~KW. Frequency Changers. the machine to the right is the oo-cycle generator; the exciter shown on the right hand side of the set serves as a starting motor and also excites the two alternators. The rating of each set is 1068 K.W. at 2300 volts 60 cycles, 100 per cent power-factor, or 800 K.W. at 75 per cent power-factor. The speed of the frequency changers is 450 revolutions, the motor being an 8-pole machine, the generator a i6-pole machine. "The frequency changers are started from the exciters, which are good for 75 K.W. at 120 volts. Although the excitation of each machine does not exceed 18 K.W. under any condition of load, it was deemed advisable to use large exciters in order to facilitate the starting of these sets, as at the moment of starting the current taken is quite considerable. A 3o-cycle induction motor direct-connected to an 8o-K.W. direct-current generator is used for the starting of the frequency changers. "The operation in multiple of frequency changers is of considerable interest. Imagine a frequency changer to be in operation and that a second frequency changer SUBSTATIONS. 305 is to be connected in parallel with the first. Imagine that the first set is carrying full load and that the second set is to divide the load with it. "The motor can be synchronized in the usual manner by adjusting the field current, so that the potential difference between the bus bars and the synchronous motor vanishes. If the generator is synchronized in the same way it is not possible to put a load on the machine. If the field current of the generator is diminished or increased the load of the frequency changer remains unaltered and the effect of changing the excitation results only in an increase of the cross currents between the two sets. "Now then, in order to make the second frequency changer divide the load with the first, it becomes essential to abandon the usual way of paralleling. Let it be assumed that both sets are in operation and are dividing the load equally. The T T I v- T J/' I * i # i FIG. 6. Typical Substation Switch Board Panels, i. Incoming Line or A.C. Converter Panel. 2. Outgoing Line Panel. 3. D.C. Converter Panel. 4. D.C. Single Circuit Feeder Panel. saturation curves of the machines being the same, it is clear that the exciting currents of the machines must also be the same if the load be distributed uniformly between them. As juggling the field currents after the machine has been thrown in parallel has no other effect than to increase the cross currents, it is evident that the field currents have to be adjusted properly before the machines are thrown in parallel. Hence, assume the first set in operation with 125 amperes excitation on the fields of the generator. To throw the second set in parallel with the first set, first synchronize the motor of the second set and then make the excitation of the generator of the second set, 125 amperes. The bus bar voltage on which the first set is operating is 2300 volts; the second set has the same excitation and the terminal voltage of its 306 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. o fc a o C/3 S a ~ SP bC c !S - SUBSTATIONS. 307 generator is, therefore, greater than the bus bar voltage on which the machine is to operate. Assume the drop of the machine at its load to be 12 per cent; then the generator of the second set at 125 amperes excitation on its fields will generate 2580 volts. The switches must be closed between the two machines at these unequal voltages and the two sets will pull each other in parallel with the load distributed equally between them." Switch Gear. The switch gear is similar to that of the main generating station. Each incoming feeder circuit has its own panel, and the equipment depends some- what on the form of switch adopted, whether hand or electric operated. Fig. 6 shows typical substation panels one and two for alternating current; the former, for incoming high tension lines, having a lever for remote control automatically tripped oil switch, and one ammeter; the latter, for low tension distribution, having a three- pole, automatically tripped switch and three ammeters. No. 3 is the main con- verter panel, having a circuit-breaker with overload and low voltage release, one ammeTer, one field rheostat, a potential receptacle, single-pole main switch, a double throw station-lighting switch, and the bottom panel contains a recording wattmeter. The last panel is a direct current feeder panel equipped with an overload circuit- breaker, ammeter, main switch, lightning arrester and choke coil, one potential receptacle by which the feeder voltage may be determined with the circuit-breaker open. This is used to advantage when the converters are started up on the direct current side. The high and low tension alternating current bus bars must be separated if such are installed. In small stations the transformers are directly connected without the use of a bus. In large stations, high and low tension bus bars are placed on either side of the transformers, so that a converter may be operated without its own transformer when necessary. Separate converters must be kept for lighting and railroad work, which means two direct current bus bar systems. They must, however, be interconnected that the converters can supply either systems. What previously has been said under switchboards regarding flexibility, etc., applies also to substation equipment. BIBLIOGRAPHY. TRANSFORMERS FOR SINGLE AND MULTIPHASE CURRENTS. G. Kapp. 1906. THE ALTERNATING CURRENT TRANSFORMER. F. G. Baum. 1903. THE ALTERNATING CURRENT TRANSFORMER IN THEORY AND PRACTICE. J. A. Fleming. 1900. PRINCIPLES OF THE TRANSFORMER. F. Bedell. 1908. THE SERIES TRANSFORMER. E. S. Harrar. Electrical World, May 16, 1908. THE CHOICE OF TRANSFORMERS FOR CENTRAL STATIONS. L. A. Sterrett. Electrical World, May 2, 1908. THE CENTRAL STATION DISTRIBUTING SYSTEM. H. B. Gear and P. F. Williams. Electrical Age, February, 1908. SOME FEATURES OF EUROPEAN HIGH TENSION PRACTICE. Frank Koester. Electrical Age, December, 1908. 308 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. THE DETERMINATION OF THE ECONOMIC LOCATION OF SUB-STATIONS IN ELECTRIC RAILWAYS. Gerard B. Werner. Proc. Am. Inst. E. E., May, 1908. RECEIVING STATION OPERATED FROM HIGH-TENSION TRANSMISSION LINE. Electrical Age, July, 1908. INSTRUCTIONS TO OPERATORS IN RAILWAY CONVERTER SUB-STATIONS. J. E. Woodbridge. Electric Railway Journal, June 13, 1908. PROTECTION OF THE INTERNAL INSULATION OF A STATIC TRANSFORMER AGAINST HIGH FREQUENCY STRAINS. W. S. Moody. Proc. Am. Inst. E. E., May, 1907. RELATIVE MERITS OF THREE-PHASE AND ONE-PHASE TRANSFORMERS. H. W. Tobey. Proc. Am. Inst. E. E., April, 1907. RELATIVE ADVANTAGES OF ONE-PHASE AND THREE-PHASE TRANSFORMERS. J. S. Peck. Proc. Am. Inst. E. E., April, 1907. FORCED OIL AND FORCED WATER CIRCULATION FOR COOLING OIL INSULATED TRANSFORMERS. C. C. Chesney. Proc. Am. Inst. E. E., April, 1907. A.E.&M UNIV. OF C CHAPTER X. LINE PROTECTION. LIGHTNING ARRESTERS. Purpose. To guard against interruption of service of the generating plant or substation, the electrical apparatus of same must be protected, particularly against atmospheric discharges. This is done by providing the transmission system with lightning arresters or some form of grounding device. The function of same is to act as a relief vent. Various sources of disturbances (particularly where the transmission line runs through sections of country of different altitudes), the chief of which is lightning, causing surges and oscillations in the circuit of such frequency and high potential as will endanger the apparatus in the generating plant, substation or probably both. Lightning Discharges. Lightning, as commonly understood, means the electric discharges from cloud to ground, or from cloud to cloud, but the word " lightning " as applied to electric circuits, means much more than this. It includes, besides the lightning referred to, disturbances due to static unbalancing of the circuit and surges, that is, disturbances in the flow of generated power, brought about by various causes and depending for their energy on the power of the generating system. A very small per cent of these electrical disturbances results from direct strokes, the far greater number resulting from induction by charged clouds suddenly discharging or per- haps from the static charges collected from rain, snow or fog drifting across the line. Regardless of their source all static disturbances on transmission lines are charac- terized by abnormal potentials and abnormal frequencies. Principle of Arresters. The lightning arrester must permit sufficient freedom of escape of the charge from transmission lines so as to limit their potential to a safe value. To do this, a vent is required which will permit a very large flow of current when the potential is above a certain value, but which will suppress this flow of cur- rent quickly, quietly, and completely as soon as the potential has resumed a normal value. In other words, the arrester must permit the escape of the abnormal surge, but should preferably take no current whatever and consequently cause no additional disturbance or drop in voltage. In general, a lightning arrester is made up of three elements, as follows: An air gap, a current limiting element and an arc suppressing device. Two of these ele- ments are always present, and usually the other is combined in some form with the other two. The air-gap holds the voltage ordinarily, but is broken over by any great excess potential, thus permitting current to flow. The current limiting element 39 310 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. i. Siemens-Schuckert Horn Lightning Arrester. Showing Principle of Action. LINE PROTECTION. usually appears in the form of a series resistance of some kind which limits the power current to a reasonable value. The arc-suppressing device is provided by some modification of the air-gap, and may consist of a magnetic blow-out, a mechanical arrangement by which the length of the gap is increased until the arc breaks (horn type), or non-arc-metal gaps consisting of a number of small cylinders with the proper spacing between them. Horn Lightning Arresters. In Continental Europe, where it originated, the horn type lightning arrester has been extensively used since the early stages of electric transmission. It is based on the principle that a short-circuited arc once started at the narrow gap between the horns, the heat of the arc will cause it to travel upwards along the members of the horn and break by reason of its attenuation. In some recent practice, auxiliary apparatus is used in connection with it, such as water flow grounders, oil resistances, choke coils, relays, condensers, etc. Fig. 2 shows a Siemens-Schuckert Relay Horn Light- ning Arrester with condensers, Tesla transformer, rheostat, and automatic blow-out, etc. The horns are placed 3 to 4 mm. apart, which is the lowest practical setting, because dust or other particles may collect and cause it to discharge when set lower. The gap of 3 to 4 mm. will cause the arrester to discharge under ordi- nary operating conditions at 8000 volts, but with the use of the auxiliary apparatus, it will discharge at 3000 volts and lower without changing the setting of the horns. This is accomplished by the discharge of an auxiliary gap set off by two condensers; the auxiliary discharge causes high frequencies to be set up in the rr, i f ,. , ,, . T, ,, . Tesla transformer which starts the mam gap. By this means the main gap can be set to several times the 3. 2. Siemens-Schuckert Horn Gaps with Micro- metric Setting. opening otherwise required for breakdown at 2000 or 3000 volts. Fig. 3 shows the arrangement of six such arresters connected to an oil resistor. Three of the arresters are connected in " Y " to relieve the line of lightning discharges and three are connected in delta between phases to relieve one another of unbalancing effects. The American type of horn lightning arresters is usually built on a large scale and preferably installed out of doors. Fig. 4 shows such an arrester as installed by the American River Electric Company, California. They are installed on the 40,000- volt transmission line, and are made of galvanized iron gas-pipe mounted on insu- lators on a pole construction; the gap is 2.25 inches. The horns are grounded through a 25-gallon water tank with a film of oil on top to keep down evaporation. Expe- rience has proved that pure water in the tank gives better satisfaction than water with salt. The company reports: " In one instance they discharged several times in suc- cession, the arc traveling halfway up before breaking. Every discharge had the same 312 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. 3. Application of Siemens-Schuckert Lightning Arresters with Micrometric Setting, and Oil Rheostat. LINE PROTECTION. 313 effect as a temporary short-circuit, causing the voltmeter to swing entirely across the scale, and the lights to dim to perhaps half candle power. We have had no trouble from these arresters, no damage done by lightning, and consider the arrester as satis- factory for high voltage as any now in use." To Line FIG. 4. Construction of a Horn Type Arrester, FIG. 5. Curve Showing Setting of Horn American River Electric Company. Gaps. Horn-Gap Setting. In Fig. 5 is shown a curve which gives the proper gap-lengths for horn-gaps when used on certain voltages. According to American ideas, horn- gaps should not be used for potentials lower than 13,500 volts, since the gap is so small FIG. 6. Protection of a Combined Overhead FIG. 7. Protection of a Combined Overhead and Underground Line, Using Oil- and Underground Transmission Line, Immersed Choke Coils. *Using Horn Gaps and Oil-Immersed Resistances. that the arc will not rise properly and break. Some latitude is allowable in the setting of the horn-gaps. The gap must be so set that small arcs will not strike back and rise again repeatedly. 314 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. 8. Scheme of Station Protection by FIG. 9. Arrangement of Protecting Apparatus Torchio. of 3ooo-volt Circuit as Proposed by Gola. FIG. 10. Flat Choke Coil. FIG. ii. Hour Glass Type of Choke Coil. LINE PROTECTION. 315 Choke Coils. Choke coils are installed to take care of surges in the line and are always used in connection with the various kinds of lightning protective devices. The advantages of using a choke coil are, as there is normally no voltage between the turns, and there is no tendency to hold a short-circuit in case of a surface momen- tary discharge, it pc'rmits of a cheapor transformer construction. They are made in various forms, such as flat copper strips wound in spiral, copper wire wound spirally in the form of an hour-glass, or in cylindrical spiral form. In most forms the turns are insulated from one another. Multigap Arresters. Of the various makes of multigap lightning arresters, the differences between them amount to but little. They are built to operate on the same principles, which are as follows: The greater the value of the dynamic current, the greater the number of gaps required to extinguish the arc. Any arc is unstable GOOOO Vtf(9 JOOOO Volts FIGS. 12 and 13. Three Phase Multigap Lightning Arresters, General Electric Company. and can be extinguished by placing a properly proportioned resistance in parallel with it. Further, the higher the frequency of the lightning oscillations, the more readily will the multigap respond to the potential. Being made up of units, the multigap arrester can be built for all commercial voltages. Those used on circuits below 6000 volts are classified as low tension, and those above, as high tension arresters. Action of Multigap Arrester. The essential elements of this arrester are a number of cylinders spaced with a small air-gap between them, and placed between line and ground, and between line and line. In operation, the multigap arrester discharges at a much lower voltage than would a single gap having a length equal to the sum of the small gaps. In explaining the action of multigaps, there are three things to take into considera- tion; the transmission of the static stress along the line of the cylinders; the sparking of the gaps; the action and duration of the dynamic current which follows the spark, HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. and the extinguishment of the arc. A spark may be defined as conduction of elec- tricity by the air, and an arc as conduction of electricity by vapor of the electrode. The cylinders of the multigap arrester act like plates of condensers in series. This condenser function is the essential feature of its operation. When a static stress is applied to a series of cylinders between line and ground, the stress is instantly carried from end to end. If the top cylinder is positive it will attract a negative charge on the face of the adjacent cylinder and repel an equal positive charge to the opposite face, and so on down the entire row. The second cylinder has a definite capacity relative to the third and also to the ground; con- sequently the charge induced on the third cylinder will be less than on the second, due to the fact that only part of the positive charge on the second cylinder induces negative electricity on the third, while the rest of the charge induces negative electricity to the ground. Each succes- sive cylinder, counting from the top of the arrester, w r ill have a slightly less charge of electricity than the pre- ceding one. This condition has been expressed as "a steeper potential gradient near the line." The quantity of electricity induced on the second cylinder is greater than on any lower cylinder, and its gap has a greater potential strain across it. When the potential across the first gap is sufficient to spark, the second cylinder is charged to line poten- tial and the second gap receives the static stress and breaks down. The successive action is similar to overturning a row of nine-pins by pushing the first pin against the second. This phenomenon explains why a given length of air-gap concentrated in one gap requires more potential to spark across it than the same total length made up of a row of multigaps. As the spark crosses each successive gap, the potential gradient along the remainder readjusts itself. FIG. 14. Graded Shunt Resistance, Multigap Lightning Arrester, General Electric Company. LINE PROTECTION. 317 When the sparks extend across all the gaps, the dynamic current will follow if, at that instant, the dynamic potential is sufficient. On account of the relatively greater current of the dynamic flow, the distribution of potential along the gaps becomes equal, and has the value necessary to maintain the dynamic current arc on a gap. The dynamic current continues to flow until the potential of the generator passes through zero to the next half cycle, when the arc-extinguishing quality of the metal cylinders comes into action. The alloy contains a metal of low boiling point which prevents the reversal of the dynamic current. It is a rectifying effect, and before the potential again reverses, the arc vapor in the gaps has cooled to a non- conducting state. Installation of Multigap Arresters. The multigap arresters may be installed on delta connected and also on "Y" connected circuits, with the neutral grounded or ungrounded. The difference lies in the use of a fourth arrester leg between the multiplex connection and ground on underground systems. The reason for introducing the fourth leg is evident, for if one leg becomes accidentally grounded, the full line potential would be thrown across one leg if the fourth or ground leg were not present. On a " Y" system with a grounded neutral, the accidentally grounded phase causes a short-circuit of the phase and the arrester is relieved of the stress by the tipping of the circuit breaker. Briefly stated, the fourth or grounded leg of the arrester is used when, for any reason, the system could be operated even for a short time, with one phase grounded. In, protecting 2-phase 4-wire circuits, two single phase, multiplex connected arresters are used; when protecting 2-phase 3-wire circuits, two single phase arresters are connected in between the outside leg and the common leg, no multiplex cross connection being between the outside legs. As much wall space as possible must be provided, and plenty of room in front must be left for the operator. The following minimum separation distances, recommended by the General Electric Company for the past few years, have proved entirely satisfactory. TABLE I. GIVING PROPER SPACE BETWEEN ARRESTERS. Volts. Distance between live parts of adjacent phases. Minimum distance between centers. 1 Inches. Inches. 6,600 8 28 IO,OOO 8 28 12,500 8 33 15,000 10 35 2O,ooo 12 37 25,000 IS 48 30,000 22 52 35' 26 56 40,000 28 62 45,000 32 67 50,000 36 72 60,000 40 78 1 If barriers are used, the width of barriers should be added to distances given. HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. It is advisable to install arresters in a dry place, and before assembling them, the wooden supports, insulators, etc., must be thoroughly dried of all moisture which may have collected. Fluid Arresters. There are two different kinds of fluid arresters in American practice. In one, the components are submerged in oil in a steel tank, and known as Aluminum Arresters (General Electric Company) : in the other, the components are incased in an empty porcelain jar, and known as the Electrolytic Arrester (Westinghouse Electric and Manufacturing Company). The principle of both is practically the same. It consists of a series of concentric aluminum pans, placed one above the other, separated by an electrolyte, usually a borax solution. Experiments have been made for a number of years with a film, which may be formed on aluminum plates, when treated with certain electrolytes. This film being very thin, that is, comparable in thickness with a wave length of light, its electrostatic capacity as a condenser is very great. If the electromotive force is constant, only leakage current passes through, but if it is alter- nating, there is a leakage and a charging or condenser current superimposed. It was discovered that this film has a very desirable characteristic for lightning arrester purposes, in that it has an apparent resistance of a very high value when moderate voltages are impressed upon it. When the voltage or pressure reaches a certain value, however, this film breaks down in myriads of minute punctures making almost a short circuit for these higher voltages. As soon, however, as the voltage is reduced again, the minute punctures seal up at once, and original high resistance reasserts itself. It may be seen that for elec- trical pressure, this action is exactly the same as that of a safety valve on a boiler. In the aluminum arrester, each cell is designed to operate normally at 300 volts with a very small leakage current, and with a perma- nent critical value of 420 volts, that is, the voltage at which the film opens and allows a free and heavy discharge is 420, and the permanent critical value is thus 40 per cent above the normal operating voltage. If the potential rises to any value greater than 300 and less than 420 volts, a tem- porary critical value is reached and the film allows the arrester to discharge for a short time. A thicker film is soon formed and the leakage current is decreased to a small amount. When the line potential again becomes normal, this extra thickness of the film gradually dissolves. If the voltage continues to rise, this process of form- ing a temporary critical film continues until the permanent critical value, 420 volts, is reached, when the cells discharge freely, allowing a heavy rush of current. The FIG. 15. General Electric Company's Aluminum Lightning Arrester. LINE PROTECTION. 319 700 .600 soo Ifpo VOLT-AMPERE: CHARACTER/STIC FIG. 16. Discharge Rate Above Permanent Critical Value. /MO [360 $230 eo 140 i Uf Vot-T-AMPEffc CHARA ALUMJNUM UGHTN/N CTER/5T/C GAffRSTE> _ > ,O/ .OS ^7J ^* .OS .06 -<77 ,.O0 ^> ^O .// Amperes Fig. 17. Characteristic Curve at Permanent Critical Value of General Electric Aluminum Arrester. For Delta or Ungrounded F System. For Grounded Neutral System. FIGS. 18 and 19. Arrangement of General Electric Company's Arresters. In the Installation the Bases of the Horn Gaps are, of course, Horizontal. HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. plate area of the cells is sufficient to discharge a quantity of electricity many times greater than that which would be liberated by an ordinary induced lightning stroke. The upper cut in Fig. 16 is a volt-ampere curve showing the characteristics of a film which has been formed up to its permanent critical value ; the lower (drawn to a FIG. 20. Outside Installation of Westinghouse Electrolytic Lightning Arresters. different scale) shows the discharge rate above the permanent critical value. Suffi- cient cells are placed in series on circuits of any given voltage to allow a normal volt- age of 300 volts per cell. The arresters are connected permanently between line and ground. A multi- gap or horn-gap, set at a suitable value above line potential, is inserted in series, and LINE PROTECTIONS. 321 FIG. 21. Westinghouse Electrolytic FIG. 22. Horn Lightning Arresters with Water Lightning Arrester. Flow Grounder at Substation Steghof, Switzerland. FIG. 23. Oerlikon Water Flow Grounder. 322 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FlG. 24. 5o,ooo-volt Alioth Waterflow Grounder at Step-up Station, Piattamala, Italy. FIG. 25. Bank of Horn Gaps, Choke Coils, and Water Flow Grounders installed at the Vandoise Motor Power Company's Plant at Lakes of Joux and Orbe, Switzerland. LINE PROTECTIONS. 323 xawoovo TO TI-U9 ^O//VT prevents the arrester from being subjected continuously to the line voltage. In this way leakage is prevented during normal operation, and a longer life is assured. The accompanying illustrations show the application of these types of lightning arresters. Frequently the horn arrester is connected to water flow grounders. Fig. 22 shows such an arrester as installed by the Oerlikon Company, in connection with a 27,000- volt transmission line. The grounding device consists of a pair of glass tubes through which water is continuously flowing. Another arrangement of a water flow grounder as installed by The Alioth Company, in connection with the 50,000- volt Swiss-Italian transmission system, is seen in Fig. 24. It has been installed in addition to horn-lightning arresters and choke coils, to take care of light surges in the line and to maintain uniform line pressure. This apparatus consists of a nozzle or jet of water (from a spring), playing against a baffle plate connected to the line. The stream of water is three- eighths of an inch in diameter, 28 inches high, and allows a leakage of o.i ampere. Ammeters are inserted in the line connec- tion to detect failures in grounding. Water-flow grounders, in different forms, have been used successfully for a number of years on the Continent of Europe. However, in America its use has not been advocated, for the reason that the assumption of the failure of water supply points out that the apparatus is inefficient. This argument is not quite justifiable, as the water may be drawn from the same supply as the turbines, and in substations, usually located in or near cities, water from the city mains can be used. It is the practice in Euro- pean countries to make use of the water which circulates through the cooling coils in the oil transformers. Further, the water from nearby springs is oftentimes available. Location of Arresters. The main generating station and all substations must be equipped with lightning arresters. Practice of recent years shows that it is good policy to install more than one form of arrester, for instance, a combination of multi- gap and horn type for direct lightning strokes; choke coils and fluid arresters to take care of slight atmospheric discharges and surges. Some power plants are equipped with all four of the above-mentioned forms, for example, the Ontario Power Company, which has the electrolytic form of fluid arrester, and gaps of the different horns set for various voltages. In a recently installed 7000/50,000- volt transformer station at Piattamala, Italy, the station protection is as FIG. 26. Lightning Rod. 324 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. follows: Flat choke coils are placed on both sides of the transformers in connection with horn lightning arresters provided with water rheostats; for taking up lighter static and atmospheric discharges, cylindrical choke coils with non-inductive resist- ances are provided. Finally, as all surges will create more or less variation in pres- sure, water jet grounders are installed to maintain uniform pressure. The location of lightning protection devices is a matter of opinion. In American practice, the choke coils and multigap arresters are located inside the station, while the fluid and horn-gap arresters are outdoors. In Europe, the practice is to locate all the lightning protection apparatus inside of the stations, with exception of those on the line. Even these are sometimes placed in section houses. The transmission line itself must be protected against lightning either by horn- lightning arresters at frequent intervals (about 2 or 3 miles), or by the overhead guard wire. The latter is more frequently used on wooded pole line construction. Where no guard wire is used on wooden pole lines, the individual poles must be provided with a lightning rod extending some ten to twelve inches above the top, sometimes fastened to an iron pole-cap. Guard wires, and all lightning arresters, must be well grounded by a copper wire having a short and straight run to ground. The end may be wound in a coil or connected to a copper plate buried in the ground. Flat copper strip is sometimes used in place of copper wire. The efficiency of the grounding wire is increased if the earth plate is buried in moist ground. BIBLIOGRAPHY. LlGHTNING-RODS AND GROUNDED CABLES AS A MEANS OF PROTECTING TRANSMISSION LlNES AGAINST LIGHTNING. Norman Rowe. Proc. Am. Inst. E. E., May, 1907. PRACTICAL TESTING OF COMMERCIAL LIGHTNING ARRESTERS. P. H. Thomas. Proc. Am. Inst. E. E., June, 1907. A PROPOSED LIGHTNING ARRESTER TEST. N. J. Neall. Proc. Ant. Inst. E. E , June, 1907. PROTECTIVE APPARATUS ENGINEERING. E. E. F. Creighton. Proc. Am. Inst. E. E., June, 1907. PROTECTION AGAINST LIGHTNING, AND THE MULTIGAP LIGHTNING ARRESTER. D. B. Rushmore and D. Dubois. Proc. Am. Inst. E. E., April, 1907. NEW PRINCIPLES IN THE DESIGN OF LIGHTNING ARRESTERS. E. E. F. Creighton. Proc. Am. Inst. E E., 1907. SOME FEATURES OF EUROPEAN HIGH-TENSION PRACTICE. Frank Koester. Electrical Age, December, 1908. PART III. (APPENDIX.) MODERN AMERICAN AND EUROPEAN HYDROELECTRIC DEVELOPMENTS. A.E.& UNIV. o APPENDIX. TYPICAL HYDROELECTRIC PLANTS. THE POWER PLANT AND TRANSMISSION SYSTEM OF THE ONTARIO POWER COMPANY. ACCORDING to Zoelly, in a paper before the Engineering and Architectural Society of Zurich, Switzerland, 1 there is no accurate data on the flow of water over Niagara Falls; it is estimated that the flow is one hundred million cubic meters per minute (three thousand five hundred and thirty million cubic feet). This is sufficient to develop 16,800,000 HP., or, figuring on an efficiency of 75 per cent, 12,600,000 HP. 2 Although this enormous amount of power is available, and in spite of the number of large plants already erected, only a small percentage of the water is utilized. Since 1890, when the International Power Commission met to decide upon the utilization of the water of Niagara, great progress has been made in water-power development. On April 4, 1895, the first 5ooo-HP. turbine of the Niagara Falls Power Company was set in motion. The many plants now located around Niagara Falls give ample proof of the success of this first installation, especially as three other Niagara plants have been built on the same lines. It will be noticed by studying the accompanying drawing, that on the American side are located, besides several small, four large installations; two above the rapids, and two below the falls in the gorge. The law of the New York State Reservation, 1885, stipulated that the big power developments had to be one mile away from the Falls. On the Canadian side, conditions were different; the whole of Victoria Park was thrown open wide to the development of power from the Horse-Shoe Falls. Starting on the American side, above the Falls, are located power houses Nos. i and 2 of the Niagara Falls Power Company. These plants are located on either side of an indented forebay about a mile and a quarter above the Falls. The turbines in station No. i are of the 5ooo-HP., vertical type, located at the bottom of a pit, and operate under a head of about 135 feet. There are ten units installed. Power house No. 2 is designed on the same principle and contains ten 55OO-HP. units. The tailrace of both plants empties into a tunnel 1000 feet long and discharges into the gorge at the side of the pillar of the steel arch bridge. It might be of interest to state, that after thirteen years of continuous operation, the lining, of ordinary brick, has been in no way damaged. Beneath the Falls at the water's edge on the American side, are power houses Nos. 2 and 3 of the Niagara Falls Power and Manufacturing Company. On the Canadian side, the intake of the Ontario Power Company plant will be 1 Neuere Turbinenanlagen. Zeitschrift des Vereines deutscher Ingenieure, 1901, p. 1239. 2 Prof. W. C.'Unwin made a rough estimate of 7,000,000 HP. 3 2 7 328 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. =5 I Q. o o .j- TYPICAL HYDROELECTRIC PLANTS. 329 observed to be the farthest above the Horse-Shoe Falls. Below this is the plant of the Electric Development Company. It is equipped with vertical turbines similar to those of the Niagara Falls Power Company. The tailrace discharges into the Gorge at the base and behind the Horse-Shoe Fall. Below this power plant is that of the Canadian-Niagara Power Company, which is allied with the Niagara Falls Power FIG. 2. Map of Niagara Falls, showing Location of Power Developments. Company. It is designed for vertical turbines with the general arrangement as the three above mentioned. The water from this plant is discharged into the Gorge at the foot of Table Rock Cliff. The four plants above the Falls are identical in many respects. Ontario Power Plant. 1 The largest and most prominent power plant contem- plated is that of the Ontario Power Company, located on the Canadian side of the Falls. There is no installation in the world which exceeds it in capacity. The power house is located in the Gorge near the Table Rock Cliff, and draws its water above the Falls among the Dufferin Islands. The ultimate capacity of the plant will exceed 200,000 HP. This power is controlled and distributed, at 60,000 volts, from an isolated distributing station, situated on the cliff some 600 feet away and 260 feet above the generating station. Forebay. The forebay or intake is about 600 feet long, stretched across the inlet of Dufferin Islands and practically parallel to the main stream. The deflecting 1 Abstract from a paper by P. N. Nunn, The Development of the Ontario Power Company. Inst. E. ., Ashville, N. C., June 19-23, 1905. Am. 330 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. curtain wall of the outer forebay is made of reinforced concrete faced with wooden planking (see Fig. 3). The water here is 15 feet deep; as the curtain wall extends 9 feet into the water, only deep water is admitted to the forebay, and all floating material is deflected. The water, after entering the outer forebay, passes through a rack into the inner forebay. The rack structure is 320 feet long and lies across the entrance of the inner fore- bay, practically parallel with the flow in the outer forebay. All finer floating material which passes the outer deflecting wall is deflected by the main screen house and carried over the spillway (see Fig. 4). At the foot of the rack is a trench or sand- FIG. 3. Section through Intake of Forebay, Ontario Power Company. FIG. 4. Section through Screen House, Ontario Power Company. trap to carry off sand, gravel, etc. The water at the screen house is 20 feet deep, while at the gate house it is 30 feet deep. The gate house is provided to accommodate three penstocks, with motor-operated head gates. In front of the head gates are wide mesh screens and a curtain wall, extending about 3 feet into the water, to prevent foreign material from entering the penstock. The screen and gate houses are well provided with steam for heating and thawing, also electrically operated cranes to facilitate the changing of screens. As the buildings are located in the reservation, special attention has been paid to the architectural features of same. Penstocks. The main penstocks are three, of which two are installed; are 18 feet in diameter and laid in the top of the lower cliff. This penstock is made of o.5-inch material, reinforced on the upper half with bulb tees and covered with concrete (see Fig. 6). It is 650x3 feet long and calculated for a velocity of 15 feet per second. TYPICAL HYDROELECTRIC PLANTS. 331 FIG. 5. Section through Gate House, Ontario Power Company. FIG. 6. i8-foot Penstock, partly embedded in Concrete, Ontario Power Company. 332 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. At the end of the penstock above the power house, cut in the rock, is the valve chamber, from whence seven branches, 9 feet in diameter, lead down to the turbine room, supplying water at a velocity of 10 feet per second. The main penstock also has two 3o-inch branches for supplying the exciter turbines. These branches are provided with electrically operated gate valves controlled from the generating room, and run vertically, then horizontally, to the turbines. At the bend they are securely anchored, being embedded in concrete. Each branch is provided with two expansion joints. At the end of the main penstock is a spillway, with a helical discharge. The function of the spillway is to act as a relief valve, in case a generator dropped its load. The characteristic features of the spillway are, the adjustable weir and helical discharge, which preserves a smooth, unbroken water column, with highest velocity and least expenditure of energy. This scheme has been adopted to pre- vent erosion, restricted flow and excessive air suction, the latter on account of the formation of ice from spray under forced circulation of air. Power House. The power house is located at the bottom of the cliff, and is 76 feet wide with an ultimate length of 1000 feet. The main turbines are arranged in a single row, and the exciter turbines are set in recesses (see plan and cross section). The cross section is taken through the extreme width, including the recesses. The whole building, including the roof, is made of concrete and reinforced concrete. It is of handsome design, both exterior and interior. The walls of the latter are faced with white enameled brick. The turbine room is served by a 5 I I I c 6 10 6 TYPICAL HYDROELECTRIC PLANTS. 355 The runners of the Holyoke turbines are 48 inches in diameter. The runners of the Allis-Chalmers turbines are 53 inches in diameter, and the draft connection ii feet in diameter, which gradually increases to 18 feet 3 inches, by n feet 2 inches. The intake flumes are 18.5 feet by 16 feet, and taper down to 15 feet in diameter at the turbine casing. The runners of the exciter turbines are 24.5 inches in diameter. The intake of the same is 9 feet high with semicircular ends of 3-foot radius, tapering down to 6 feet in diameter at the turbine casing. The draft tubes are 5 feet 6 inches in diameter at the casing, and flare to a width of 9 feet 10 inches, with semicircular ends of 2 feet 9 inches radius. As the turbines are located in the body of the dam, a tunnel in the latter is provided, so that access may be had to the outside bearings. The turbo-generator sets are controlled in pairs by Lombard governors; there is one type "N" governor for two sets of main turbines, while the two exciter units are controlled by a single type "P" governor. The "N" type, developing 31,000 foot-pounds, are guaranteed to completely open and close the gates in 1.5 seconds; while the "P" type, developing 6700 foot-pounds, in 4 seconds. The former are electric controlled from the switchboard. There are 4 by 6-inch triplex pumps operated by belts from the turbine shafts; these, and the pressure tanks are located in the above mentioned tunnel. Power House. The power house is 250 feet long and 37 feet wide. Adjoining same is a two-story switch and transformer house 85 feet long and 75 feet wide. The generating room is well provided with 20-inch roof ventilators, and is served by a 25-ton hand crane, with a 5-ton auxiliary trolley. On the main floor, in the switch and transformer house, are located the low tension oil-switches and transformers, while on the upper floors is the high tension apparatus. Generators. The main generators are of 3000 K.W. capacity, 60 cycles, 2300 volts. The exciters are 400 K.W. capacity, 250 volts. The guarantee of the main generators is as follows: Load Full I } \ Efficiency per cent 06 QC. C 04 oo The temperature is guaranteed not to exceed 35 C., after 24 hours run at normal load, and 50 C., at 37.5 per cent overload for the same duration. The guaranteed efficiencies of the exciters are as follows: i | Full ij Efficiency per cent so 88 01 02 02 356 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Switchboard. As will be noticed in the accompanying illustration, the switchboard is located several feet above the floor on a raised platform. It is made of blue Ver- mont marble and contains two transformer panels, two double-circuit feeder panels, one station and two blank panels. In front of the switchboard, arranged in a semicircle, are eight instrument posts and eight pedestals for controlling the main generators. Switchboard, instrument columns, and control pedestals are well equipped according to modern practice. FIG. 6. Power House, seen from Tailrace, Southern Power Company. The generator leads are lead-covered cables, and run through tile ducts laid in the floor. Wiring Diagram. There are two sets of exciter bus bars and one main generator bus. The generators feed the latter through non- automatic oil circuit breakers; between the transformers and the low tension bus are located overload time limit oil circuit breakers. Between the transformers and the high tension bus (44,000 volts) are reverse current circuit breakers. Both bus bars are divided up into two sections by sectionalizing switches, on both sides of which are disconnecting switches. The outgoing feeders are provided with overload time limit circuit breakers. The whole wiring diagram is such that two generators can feed, through one transformer, a single transmission circuit, or they may feed any of the transformers or outgoing lines. Again, the transformers may feed directly the outgoing feeders by by-passing the high tension bus bar. The tow tension bus bar is made up of five strips of 3 by one-fourth inch copper, TYPICAL HYDROELECTRIC PLANTS. . OF O, a o U -i .s 386 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. 330 feet wide. The penstocks leave the forebay at right angles, and are provided with fine screens and sluice gates. The gates are operated from the top of the forebay wall, where also the cleaning of the racks is accomplished. As the forebay wall is only 45 feet away from the rear wall of the power house, the penstocks have a short run, being only 100 feet long. Power House. The power house was originally designed to accommodate four main turbine units and two exciter units. The general arrangement consists of a generating room 150 feet long and 50 feet wide, set directly over the turbine pits, which are 25 feet wide and 15 feet deep. Behind the generating room is the transformer and switching room; the transformer room floor is about 5 feet below the main generating room floor. The switchboard gallery is 16.5 feet above the main generating room floor. FIG. 4. Power Plant, Kykkelsrud, Norway. In the middle of the rear, adjoining the transformer room and beneath the street level, is the pump room. The arrangement of the windows, pilasters and location of generating units is symmetrical. There are five bays, the middle one containing two 28o-HP. exciter units and the controlling switchboards. On each side of the middle section are two generators. The entire interior is finished off in light color, the floor finished with diamond-shaped tile. The generating room is provided with abundant light and ventilation. Turbines. Owing to the great fluctuation in the water level (the head varies from 40 to 64 feet), it was decided to use inclosed Francis turbines. Of the first installa- tion, one main and two exciter turbines were furnished by Voith, Heidemheim, while the other main unit was furnished by Escher Wyss, Zurich. The Voith 3OOO-HP. TYPICAL HYDROELECTRIC PLANTS. 387 turbine was designed for a head varying from 52 to 62 feet and consuming from 670 to 530 cubic feet per second, and running with a speed of 150 R.P.M. The water is fed to the turbine with a velocity of 9 feet per second through a 9.8-foot penstock embedded in concrete. Where the penstock joins the turbine casing, there is an 8.5-foot hand-operated geared butterfly valve. The turbine casing is rec- tangular and built of structural steel. The inlet of the spiral casing is 6.5 by 3.4 feet, thus giving a velocity to the water of 9 feet per second. The velocity of the water at discharge is 3.9 feet per second. FIG. 5. Headrace, Kykkelsrud Plant, Norway. The vertical shaft of the turbine is 12 inches in diameter and about 25 feet long; on top of this is coupled the shaft of the generator. The weight of the revolving part is 32 tons and is taken up in a step-bearing running under an oil pressure of 220 pounds per square inch. The regulation of the turbines is accomplished by an hydraulically operated governor and works in conjunction with the oil pressure in the step bearing. The principal difference between the Escher Wyss and the Voith turbine is that the former has a cylindrical casing and vertical moving ring-gates, while the latter has a spiral turbine casing with a so-called clam-shell gate. The penstock connec- tions and butterfly valves are the same. The cylindrical gate is operated by a three- piston arrangement, worked by oil pressure controlled by the hydraulic governor. This governor is similar to that of the Voith turbine and is operated by gearing from 388 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. the main shaft. The guarantee of the main turbines, from their manufacturers, under a head of 52.5 feet, is 75 per c^nt at normal speed, 150 R.P.M. Keeping the speed constant, and a head variation of 6.5 feet up or down, the guarantee efficiency is 72 per cent. The oil pressure for the turbines is supplied by two motor-driven pumps pro- vided with air chambers to maintain uniform pressure. Each pump has a capacity FIG. 6. Interior of Kykkelsrud Plant, Norway. of 90 gallons per minute; under ordinary conditions, only one pump is in operation. If one pump is out of commission, the other starts up automatically. The two recently installed turbines are of the same type as the above described, but have a capacity of 3750 HP. each. Exciter Units. The two exciter turbines are located in one wheel pit and are supplied by one 6.5-foot penstock. The branches to the turbines are 4.1 feet in diameter and are fitted with butterfly valves. The guarantee of these turbines under a head of 52.5 feet, is 76 per cent running at a speed of 325 R.P.M., the water consumption being 60.8 cubic feet per minute. These turbines are not supplied with oil pressure step bearings, because the oil pressure pumps are driven by current from TYPICAL HYDROELECTRIC PLANTS. 389 the exciters. The weight of the revolving part of the turbines is taken up by relief disks in the turbine casing. The generator shaft is coupled directly to the turbine shaft. The exciter is wound for 115 volts, 1580 amperes, giving 181.7 K.W. They also supply the station with light, besides running the oil pressure pumps. FIG. 7. View in Rear of Switchboard, Kykkelsrud Plant, Norway. Generators. The main generators of the first equipment are 5ooo-volt, 3-phase, 5o-cycle, 4o-pole revolving field type and have a 2ooo-K.W. capacity. A full load test was run continuously for 48 hours, and no part showed a temperature greater than 26 C. With unity power factor, the efficiency was 96 per cent, and with power factor 0.80 the efficiency was 94.8. The copper loss was 31 K.W.; the iron loss with unity power factor was 16 K.W.; with power factor 0.80, was 21 K.W. The friction and windage loss was 59 K.W.; this included all friction losses in turbine and shaft. The excitation at full load is 290 amperes. The two recently installed generators are of the same type and make (Siemens Schuckert Werke) as the above, and have a capacity of 2500 K.W. 390 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Switchboard Room. The switchboard is provided for four generator panels, and two panels for outgoing feeders located on a gallery 16.5 feet above the main floor; the exciter switchboard is located on the main floor directly below this gallery. Each generator panel has a voltmeter, ammeter, synchroscope and phase lamps. The lower front part of the switchboard has hand wheels for controlling the exciter current. The panel for the 2o,ooo-volt outgoing feeders has instruments and levers for oil switches controlling the current. As there are two complete bus-bar systems, FIG. 8. Transformer Room, Kykkelsrud Plant, Norway. current may be thrown on or drawn from either. These bus bars as well as all 5000- volt apparatus axe mounted upon a structural steel frame, located five feet in the rear of the switchboard, thus giving a passage for inspection and repairs. The oil switches are mounted upon the top of this framework and are operated by levers (see Fig. 7). The current from the generators passes through fuses placed in marble compart- ments. There are three transformers for each generator of the first equipment and they are located in the basement of the switchroom. They are arranged in two banks on either side of a track used for the removal of the transformers, which are of the air- TYPICAL HYDROELECTRIC PLANTS. 391 cooled oil type, 950 K. W. each, and step up the voltage from 5000 to 20,000. The transformer tanks are surrounded by a corrugated iron casing. The air for cooling comes up from the basement beneath, under a pressure of three-fourths of an inch. The transformer losses for full load at unity power factor are 19.75 K.W. com- posed of 8.25 K.W. iron and 11.5 K.W. copper loss. With power factor unity from half to full load, the efficiency is constant, 98 per cent. For continuous full load, the temperature of the transformer oil never exceeds 40 C. FIG. 9. Interior of Substation at Hafslund, Norway. Leads from the transformer go to the high-tension busses located in a structural steel frame on the same floor as the 5ooo-volt busses. With the extension of the plant, four 3-phase transformers of 225O-K.V.A. capacity each, have been installed. They are of the water-cooled oil type and wound for 5 000/50,000- volt transformation, and serve exclusively the transmission line to Hafslund. Transmission Line. From the Kykkelsrud power house, lead 2o,ooo-volt, 3-phase transmission lines toward Christiana, then around Christiana Fjord to Slemmestad, where the last substation is located. Each cable has a cross section 392 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. of 50 square mm. for a distance of 38 miles, then it is reduced to 35 square mm. which runs for 15 miles; the total length of the line is 53 miles. There are seven substations along the line, stepping down the line voltage to 5000 for local distribution. The lines are carried on wooden poles except at railroad crossings and turns, where structural steel poles are used. This line was put into operation in 1903. In 1907 with additional plant equipment, a 5o,oco-volt line was run 25 miles southward to assist the plant at Hafslund. This line is also run on wooden poles except at railroad crossings and turns where structural steel towers are used. As seen in Fig. 10, the steel towers have a close spacing, and in addition, the tops of same are provided with triangular steel frames, which ground the line in case of a break. These precautions were required by the Public Service Commission. The wooden poles are about 40 feet long and stick FIG. 10. 5o,ooo-volt Transmission Line, Kykkelsrud to Hafslund, Norway. about 6 feet in the ground. They are spaced about 100 feet apart. The cables have a cross section of 64 square mm. and are carried on porcelain insulators arranged in triangular form, 5.5 feet on a leg; the cross arm is of steel. The insulators are fastened to the pin by hemp and shellac. At present there is only one 5o,ooo-volt line; a duplicate one is projected to run parallel about 35 feet from it. Substation. The substation at Hafslund is equipped with four aooo-K.V.A. transformers of the water-cooled oil type, and are designed for 45,ooo/5ooo-volt step-down transformation. This station is used for a distributing center for the power from Kykkelsrud as well as the power from the Hafslund power house. The 5o,ooo-volt line is protected in this station by water flow grounders, choke coils placed in layers, and a series of horn lightning arresters; there are also used in con- nection with these several oil resistances. In addition to this, the line on both sides is protected by horn lightning arresters with exceptionally large gaps. TYPICAL HYDROELECTRIC PLANTS. 393 From the power plant at Hafslund, lead four circuits of 5000 volts into the sub- station. From here, the power from Kykkelsrud and Hafslund may be distributed separately, or, as in common practice, in parallel. The early equipment of the plants and transmission system was furnished by the Schuckert Company, Nuremburg, and the later equipment, by the Siemens-Schuckert Werke, Berlin. HYDROELECTRIC PLANT, URFTTALSPERRE, GERMANY. 1 Forced to husband natural resources, particularly coal, advantage has been taken of all kinds of water resources. The continent of Europe, for a number of years, has harnessed the yearly supply of the drainage area of low mountainous or hilly countries. FIG. i. Urfttalsperre Dam, showing Valve Chamber Shafts and Spillway. In order to provide for a steady water supply for the whole year, more particularly for the dry season, large dams have been built across valleys. Having once stored such large bodies of water, the generation of electricity and the transmission by high- tension lines is the next step. Such water resources are usually located away from centers of industry, and it is but natural that advantage will be taken of modern 1 Reprint of author's article, " The Urfttal Hydro-electric Development in Germany," The Engineering Record, Sept. 19, 1908. Based on Data submitted by the Designing and Constructing Engineers. 394 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. high-tension distribution. The following is a brief description of the most prominent one of this kind in Europe. It has a storage capacity of sixteen hundred million cubic feet and a transmission system of 35,000 volts. This plant is known as the "Urfttalsperre," l and is situated on the river Urft in the Eifel Mountains, Germany. The capital of $2,000,000 was subscribed by seven cities and districts, four of which supplied one-fifth each; the remaining fifth was supplied by the other towns. The former are entitled to draw power, and the other three are not. The plant is capable of developing twenty-two million K.W. hours per year, of which sixteen million has already been contracted, giving a yearly income of $165,000. Dam. About 2.5 miles above the junction of the Urft and Rur, is located a dam, establishing a drainage area of 145 square miles. This dam is 190 feet high, having a width at the bottom of 165 feet, and at the top, 17 feet. The dam was built in arch form, with a radius of 650 feet, giving the crown a total length of about 1000 feet, about 300 feet of which is used for spillway, the water flowing over in cascade. The dam itself is made up of cyclo- pean masonry, the largest of the stones being of a size which required to be handled by two men. Both sides of the dam are faced with rough-faced cut stone. The dam was built in one con- tinuous mass, for which purpose, three timber towers were erected for elevating the material. From these towers, tracks ran across the dam, and by means of turntables the cars were placed on longitudinal tracks. The mortar is composed of lime, sand and trass. Cement was not used, because it was feared the cement would harden too quickly and unequally throughout the mass, thus producing unequal stresses. Trass has of late years been much used in German dam con- struction, as it forms, when mixed with lime and sand, a hard and impervious substance which dries slowly and equally. The cross section of the dam will be seen in Fig. 2. For the purpose of draining the storage basin, two discharge pipes are led through tunnels, through the bottom of the dam. The gates are located on the upstream side in valve chambers at the bottom of a shaft extending above the high-water level. The tops of the shafts and crown of dam are connected by bridges to facilitate the operation of the gates. The upstream side of the dam is plastered with "Siderosthen," a waterproof material, then faced with tile. To drain off the seepage, the dam is provided with vertical seepage drains. FIG. 2. Section through Dam and Valve Chamber, Urfttalsperre Plant, Germany. TYPICAL HYDROELECTRIC PLANTS. 395 Headrace. The power plant itself is located at Heimbach on the Rur, 1.7 miles away from the dam, so that at low water it has a head of 230 feet, and at high water, 360 feet. A tunnel 8850 feet long, having an area of 60 square feet, is cut through the mountains, thus connecting the collecting basin with the penstocks. On the basin side of the tunnel is located a sluice gate operated through a vertical shaft about 150 feet high. On the other side of the mountain, at the junc- tion of penstock and tunnel, is located an equalizing shaft which has on the top a reservoir that absorbs all fluctuations in the water flow. This chamber per- forms the same duty as a standpipe on a penstock, but in this case no water is wasted. From this shaft are also operated the sluice gates controlling the water supply in each of two pen- stocks. The velocity of the water in the tunnel is six and a half feet per second. From the bottom of the equalizer shaft, run horizon- tally two penstocks parallel to the slope of the mountain, from whence they run to the power house. The upper portions of the penstocks run through tunnels and are partly embedded in con- crete, and covered with rilling to protect the penstock from loose boulders. c rt S OJ O .rf I | ffi PL, u = i. Cos 4> = 0.7. Per cent. Per cent. Per cent. 0.25 93-5 92.0 o-75 I. 00 1.25 95- 96.0 96-5 93-5 94-5 95- The four exciters are of the 6-pole, ii5-volt, shunt-wound type. They develop 150 K.W. at 450 R.P.M. Each exciter serves four generators, with twenty-five per cent overload. 424 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Switchgear. Contrary to the usual practice of centralizing the switchgear, because it was thought best for the convenience of operation and a material decrease in first cost and simplification of the wiring system, each generator has its own switchboard. As will be seen in Fig. 4, these switchboards are located against the wall next to the switchroom, and directly opposite each generator. Thus the station is divided into complete unit systems. However, to control all .switchboards from one central point an instrument column has been installed. FIG. 4. Interior of Brusio Power Plant. The switchboards are of ornamental design and faced with white marble slabs. All high tension parts of the switchgear are located on the opposite side of the wall in masonry compartments fitted with corrugated iron rolling shutters. Each generator switchboard is equipped with the following instruments: two voltmeters; one syn- chroscope; with phase lamps; three ammeters, one for each phase; one three-pole oil-switch, which may be operated by hand or automatically. There are, further, an ammeter on the central column, a main current rheostat for excitation, and a field discharge resistance. Owing to the non-centralization of the switchgear system, it was not considered necessary to install a double bus-bar or ring system, so common in Swiss practice. There is one main and one exciter bus; both systems are divided in the middle by TYPICAL HYDROELECTRIC PLANTS. 425 sectionalizing switches. The arrangement is such that a group of three generators may also be independently excited and thrown upon separate bus-bars. The current from these three generators is intended for the valley of Brusio and for the operation of the Bernian Railway. FIG. 5. Individual Generator Switchboard. The outgoing feeders, with the exception of those just mentioned, are connected at the middle of the bus-bars, which are made up of copper strips, two by three- sixteenths inch being sufficient for one generator. Thus, where each generator connection joins the bus-bar an additional layer has been added. The bus-bars run the entire length of the switchroom, above the aisle and close to the ceiling. They are carried on petticoat insulators fastened to I-beams, and are securely 426 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. anchored in the middle and at the ends, so that in case of a severe short circuit the different phases will not be thrown together. The exciter switchboard (Fig. 8) is located upon a platform in the middle of the generating room, opposite the exciters. It is provided with four white marble panels, one for each exciter, and upon each are mounted a voltmeter, ammeter, knife-switch, shunt rheostat, and a reverse current circuit-breaker. In front of the exciter switchboard is the above mentioned central instrument column, upon which are mounted the following instruments: an ammeter, with multiple throw switch, to read the current of each generator; one voltmeter, with plugs, for each phase; two ammeters, one for each of the outgoing feeder systems of the Societa Lombarda, and one hand wheel, operating a shaft to which are connected the shunt rheostats of the four exciters. From this column one attendant may control the operation of the entire plant. Current Supply. As previously stated, much of the current generated is trans- mitted across the boundary line into Italy, and it was deemed advisable to run duplicate circuits to the substation at Piattamala. Since, however, the valley is quite narrow and atmospheric discharges are of great frequency, a tunnel was built for the purpose of carrying these wires to the station. The conductors leave the basement of the switchroom and cross the River Poschiavino through a covered bridge (Fig. i), where they then enter the tunnel mentioned. This tunnel, which runs to a substation, is 1650 feet long. It is 8.2 feet wide and 9.8 feet high, the top being arched. Owing to the customs regulations between the two countries, the tunnel cannot be entered from the power house end. Entrance is obtained, however, through a door visible from the street; at the boundary line, the tunnel is closed off by an iron door separating the Italian and Swiss sections. The accompanying cross section, Fig. 9, illustrates the scheme of arranging the conductors in the tunnel. They consist of copper bars 0.25 square inch in section, which are carried on petticoat insulators supported on channel irons pro- jecting from the side walls of the tunnel. These channels are spaced longitudinally for 4.9 feet, with reinforced concrete slabs spanning them, forming partitions between the conductors. The outgoing yooo-volt feeders tap the middle of the bus-bar system, then are carried on either side of the tunnel to the substation. For the protection of the customs officials, the circuits are fenced off by removable wire netting. Step-up Station, Piattamala. This station is built in the shape of a T, 180.5 ^ eet long, 68.8 feet wide, and 28.2 feet high, the cross wing being 92 feet long and 42.6 feet high. It is designed to accommodate 24 single phase transformers having a capacity of 1250 K.W. each. At present there are thirteen installed, with a total normal capacity of 16,250 K.W. At one end of the transformer room is the meter room, where the current is checked by the two companies. The transformers are arranged in two rows, between which are two tracks leading into the inspection and repair room. This is in the middle of the cross arm of the T, in which there is a lo-ton traveling crane. The TYPICAL HYDROELECTRIC PLANTS. 427 FIG. 6. Back of Individual Generator Switchboard, Brusio Plant, Switzerland. FIG. 7. Rear of Switchboards, and Gen- erator Busses. FIG. 8. Exciter Switchboard and Control Pedestal. FIG. 9. Cross Section of Cable Tunnel leading across Boundary, between Power House, Brusio, Switzerland, and Step-up Station, Piattamala, Italy. 428 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. transformer switchboard rooms are directly behind each row of transformers. The substation is divided into two distinct sections. The outgoing feeders leave the building from the third story of the cross wing. FIG. 10. Step-up Transformer Station, Piattamala, Italy. The feeder lines from the power station enter the substation from the tunnel on the ground floor, as the station is built into the hillside. As two companies are concerned in the amount of current used, the Brusio Company supplying and the Societa Lombarda receiving the current for distribution, this room, on the ground floor, is thoroughly equipped with measuring instruments, some of which are kilowatt meters of different makes, and are switched in series in order to check each other. TYPICAL HYDROELECTRIC PLANTS. 429 The switches are so arranged that the current may be thrown onto either row of transformers, from either of the two feeder lines, or the current from both feeders may be thrown on one row of transformers only. The oil switches, in the meter room, are of the remote-control, hand-operated type. It was not deemed advisable to install automatic switches, because a sudden cutting out the whole load, which might amount to 20,000 K.W., might seriously interfere with the operation of the plant, particularly the hydraulic end. Above the aisle, between the two rows of transformers, and extending the full length of the room, is a mezzanine floor carrying the feeders in two vertical rows, FIG. ii. Switch Room and Step-up Transformer Station, Piattamala, Italy. one on either side of the transformers. The phases of the bus-bar system are sepa- rated by concrete shelves, the front remaining open. The high-tension, or 50,000- volt bus-bars run on the mezzanine floor above the transformer switchboard or operating rooms. These bus-bars are arranged in horizontal rows separated by concrete partitions, but not covered. The transformers are of the Alioth water-cooled oil type, a system of water circu- lation from a spring, under a head of 26 feet, being provided. The efficiency of the transformers under actual test at full load was 97.5 per cent; at half load, 96.5 per cent. The drop in voltage between no load and full load, with a power factor 430 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. cos = i, is one per cent. With cos < = 0.8, is 2.2 per cent. The greatest drop is 2.8 per cent. Each transformer is contained in a well ventilated concrete compartment, the front being provided with a corrugated iron rolling shutter. The transformers are provided with pinion wheels, resting on pairs of racks, secured to the floor, the transfer table also being provided with such racks. This device greatly facilitates the handling of the transformers, a ratchet being used for moving them onto the transfer table, by which they are transported on the track to the inspection and repair room, where the cores are easily taken out by the overhead crane. Each transformer is provided on the low-tension (7000 volts) side, with a three- pole oil switch, while on the high-tension side (50,000 volts), three oil switches, one for each phase, are provided. These switches, interconnected, are remote controlled, and may be operated either by hand or automatically. Access to the yooo-volt switches, which are protected by doors, can only be had when the current is off. The 5o,ooo-volt switches are similarly protected. All these switches are accessible from the aisles of the operating rooms. Between each group of transformers, sec- tionalizing switches and choke coils are provided for protection against variations in load caused by throwing the switches. Protecting Devices. On account of the high tension and long transmission line, the great variation in altitude and consequent difference in temperatures, and par- ticularly on account of the frequent storms and atmospheric discharges, various devices were installed for protection against surges. For this purpose, the choke coils above mentioned are placed on each side of the transformers, and horn lightning arresters are placed on the outgoing feeders. The latter have a gap of two and three-eighths inches and are connected in series with waterflow resistances. The choke coils consist of two spools, having a brass core, upon which is tightly wound a copper band of sixty turns, separated by insulating material, forming a solid, tightly wound spool, which sudden surges will not distort. For taking up lighter static and atmospheric discharges, the more sensitive role lightning arresters were installed and connected in series with waterflow resistances. Finally, as all surges will create more or less variation in pressure, waterflow grounders are installed for each phase, to maintain a uniform pressure. This apparatus consists of a nozzle for forcing a jet of water, under a head of 26 feet (supplied from above- mentioned spring), against a baffle plate connected to the line. The stream of water is three-eighths inch diameter and 28 inches high, and allows a leakage of one-tenth ampere. Ammeters are inserted in the wire connection to this apparatus, in order to detect failures in the grounding. All lightning arresters, as well as the outgoing lines, are provided with disconnect- ing switches. All metallic features of the installation are interconnected and well grounded. Transmission Lines. The transmission line (50,000 volts) may be considered the most important in Europe. It consists of two independent lines, each 88.5 miles long. As the line runs over mountains and valleys, peaks were avoided as much as possible, to escape the unavoidable difficulties due to atmospheric discharges. These TYPICAL HYDROELECTRIC PLANTS. 431 lines cross three provinces and 94 townships, and required the right of way through 6000 properties, the cost of which averaged about $800 per mile. The lines cross ten railways, one tramway, ten state roads and 120 county roads. From the main substation at Piattamala, the line runs westward through the Adda Valley to Colico, thence along the shore of Lake Como to Bellano, from which point it runs in a southeasterly direction over the Valsasina Plateau. Palasco, the highest point of the line, is 2130 feet above sea level. From Valsasina the line runs in the mountains of Lecco in a southwesterly direction, and cross the Adda Valley with a span of 720 feet, this being the lowest point of the line (640 feet above sea level). From here, until the first step-down station, at Lomazzo, is reached, 88.5 miles distant from the step-up station at Piattamala, the run is practically straight. Eight and one-half miles beyond Lomazzo, at Castellanza, is another step-down station. FIG. 12. Brusio 5o,ooo-volt Line, Crossing Railway. The average span is 393 feet. In 87 cases, however, the span exceeded the average, the longest span being 1280 feet, across the Gravina Valley at Colico. The transmission line consists of two parallel rows of towers, from 13 to 16.5 feet apart, of latticed-girder construction embedded in concrete. Each tower is provided with six brackets, three for present use and three for future extension, so that there will be eventually four separate three-phase circuits. The porcelain insulators are supported on pins, fastened to oak and chestnut blocks secured to steel brackets. Each cable consists of nineteen wires, 2.6 mm. in diameter, the total diameter of the cable being 14 mm. (105 square mm. area). 432 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The towers are calculated for a wind pressure of 70 miles per hour, allowing a stress in the copper of 8500 pounds per square inch, and on the tower of 17,000 pounds per square inch. Allowance is made for a temperature difference of i2oF. On account of the difference in the spans and frequent changes in direction of the lines, four different types of towers are employed, weighing from 1250 to 2500 pounds each. There is a total of 3100 towers, averaging in price $80 each, including foundation and erection. The two existing lines represent 900 gross tons of copper and 10,000 insulators at $2.60 each, including mounting and wooden blocks. The laying of the cables cost $128 per mile of transmission. The transmission system is divided into six sections, varying from 8.5 to 25.5 miles, and is provided with section switches so arranged that in case of a break in a section of one line the current may be by-passed over the other. There is a small station at each section, for housing the sectionalizing switches, measuring apparatus, lightning arresters, some of which are of the horn type, some of the coil type, and some are also provided with water-flow grounders as described previously. At a distance of 65 feet, and parallel with the high tension lines, a telephone and telegraph line is carried the entire length of the transmission system, for the exclusive use of the plant. There are two wires carried on wooden poles; and 30 stations costing $30 each, while the line costs about $380 per mile. Transformer Station, Lomazzo. This substation is located centrally in the low tension distributing district. It is built in the form of an I. The wing at one end, containing the apparatus for the incoming feeders, is 85 by 30 feet, and 48 feet high. The wing at the opposite end is of the same dimensions, and contains the apparatus for the outgoing feeders. The middle member of the building, containing the trans- formers, is 55 feet wide, 60 feet long, and 33 feet high. The over-all dimensions are 85 by 1 20 feet. The two 5o,ooo-volt circuits enter the second floor of one of the wings in a way similar to the outgoing feeders leaving the step-up station at Piattamala. They are similarly protected against electrical discharges, except that the water-flow lightning arresters are supplied with water by a centrifugal pump and tank under a head of 40 feet instead of a natural head from the mountain stream. The transformers (1250 K.W. 50,000-11,000 volts) are arranged in two rows, similar to those at Piattamala, with tracks in front of the compartments, of which there are six on each side. There are also six three-phase transformers of 5000 K.W. each (11,000-20,000 volts). There are at present installed only three single-phase and three three-phase transformers. While the transformers at Piattamala are of the oil-cooled, water- circulating type, those at this station (Lomazzo) are of the forced air-cooled type, for which two blowers are at present installed. The final equipment demands four blowers, of which two will be kept in reserve. The blowers are motor-driven and discharge through air ducts located beneath the two rows of transformers. The cores of the transformers are not encased. The fronts of the transformer compartments are provided with rolling shutters; ventilators are placed in the roof. Good results were obtained with these trans- TYPICAL HYDROELECTRIC PLANTS. 433 formers, an advantage being that the cores can be easily inspected. The primary winding is provided with taps, so that the voltage may be reduced to 35,000. This was done so that easy regulation might be secured. The tests show that the efficiency at full load is 97 per cent, and at half load 96.5 per cent. The pressure loss at full load with power factor of cos = i is one per cent, and with a power factor 0.8 it is 3 per cent. The temperature rise is 40 C. The high and low tension sides, respectively, were tested to 65,000 and 17,000 volts, 10 minutes duration. The trans- formers are capable of standing an overload of 25 per cent with a total temperature FIG. 13. 5000-K.V.A. Open Type Air- Cooled Transformer at Lomazzo, Italy. FIG. 14 5o,ooo-volt. Switch Room at Sub- station, Lomazzo, Italy. rise of 60 C. The operation of the blowers is included in the aboved-named efficiencies. The 11,000-20,000 volt, 500 K.W., three-phase transformers have an efficiency of 97 per cent at full load with a power factor of cos = i, while with cos (f> = 0.8 it is 96 per cent, and three-quarters load 96 per cent and 95 per cent, while at half load it is 95.5 and 94.5 per cent. The drop in pressure is 1.5 per cent with a power factor of cos< = i, and 3 percent with a power factor of 0.8. The temperature rise is 50 C., and the overload capacity is 20 per cent for two hours. Distribution. The wiring diagram is made so that under normal operating conditions the line "A" will distribute ii,ooo-volt current in the district about Lomazzo, and "B" and "C" will supply Castellanza. The arrangement is such that one bus-bar system may feed either of the outgoing lines, or that the line "A" to Lomazzo may be fed from the line "C." Through the line "C" ii,ooo-volt current ma^ be drawn from the steam-power plant at Castellanza of the Societa 434 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. Lombarda, which is a reserve for the hydraulic plants at Turbigo and Vizzola. It will be seen that with this auxiliary source of supply, in case of emergency, current may be sent through this station (Lomazzo) and through the station at Piattamala to the hydraulic plant at Brusio. A fourth line of 20,0x30 volts leads northward to Como, for which purpose the three-phase, 1 1 ,000-20,000- volt transformers were installed. The feeders from the 50,000-11,000- volt transformers lead to the three-pole oil switches on the mezzanine floor above the aisle, between the two rows of trans- formers. The feeders to and from the transformers are provided with cutout switches. The 50,000, 11,000, and 20,000 volt bus-bars are arranged, according to the space available, in horizontal or vertical rows, and the phases separated by concrete shelves or partitions. These bus-bar compartments remain uncovered. The 2o,ooo-volt outgoing feeders are protected like those at the step-up station at Piattamala. Transformer Station, Castellanza. As previously stated, the Societa Lombarda possesses a steam-power plant at Castellanza, having an equipment of two 2500-!!?. engines and two 5ooo-HP. steam turbines, which work in parallel with the above described hydroelectric plants at Brusio, Turbigo, and Vizzola. A temporary trans- former station has been erected in the engine room of this power house, and contains six single-phase, I250-K.W. transformers arranged in groups of three. The whole apparatus, owing to the small space available, has been located on three floors. The transformers, which are of the oil, water-cooled type, are designed similarly to those at Piattamala, except for a voltage of 11,000-40,000. Taps are provided, so that some coils may be cut out, to secure a voltage of 35,000. The efficiency of the transformers at full load is 98 per cent, and at half load 97 per cent. The drop in pressure at full load with a power factor of cos = i is i per cent, while with cos = 0.8 it is 2 per cent. The rise in temperature is 45 C., using five gallons of water in twenty minutes at 15 C. They are capable of standing an overload of 25 per cent, maintaining the temperature of 45 C., and using ten gallons of water, or with a rise of temperature of 60 degrees, using five gallons of water. The transformers were tested at 65,000 volts for a duration of ten minutes. As the capacity of the steam-power plant is expected to be increased in the near future, an isolated transformer station is now being erected alongside of this power house, which will accommodate eighteen transformers. The entire installation was put in operation within 2.5 years after the organization of the company, and is giving most satisfactory results, the expectation being that the maximum output will be reached during this year. TYPICAL HYDROELECTRIC PLANTS. 435 THIRTY THOUSAND GENERATOR VOLTAGE TRANSMISSION SYSTEM. DALMATIA, AUSTRO-HUNGARY. 1 The manufacture of carbide has been carried on extensively, for a number of years, in certain sections of the Austro-Hungarian empire, particularly in Dalmatia and Bosnia. In order to produce carbide on an economical scale, the question of obtaining low-rate electric current was an essential one. This resulted, for a section of Dalmatia, in utilizing the Kerka river to such an extent that this undertaking is one of the foremost hydroelectric developments of Austro-Hungary. Of the many novel and unique features embodied in the hydraulic and electrical end, the adoption of high-voltage generators, feeding directly a twenty-one mile aerial transmission system, at a potential of 30,000, and its simple, yet highly efficient protecting devices against atmospheric discharges, stand out most prominently. This is another Continental step in the practicability and simplicity of generating current at high voltage, for long transmission systems, without the aid of step-up transformers. The river Kerka rises at the foot of Dinaria Mountains, forming the boundary between Bosnia and Dalmatia, and flows southwesterly, emptying into the Adriatic Sea, in the bay of Sebenico, below the town Scardona. The Kerka, although com- paratively short, has, throughout its length, many scenic falls, varying in height from 25 to 147 feet; the latter, named after the river Kerka and owing to their grandeur, are well known to Dalmatian travelers. The first hydroelectric plant on this river, and to-day still in operation, was installed at the Kerka Falls in 1894; a 3OO-HP. Girad turbine, operating under a head of 33 feet, is bevel-geared to a 22O-volt, 42-cycle, single-phase generator. The voltage is stepped-up to 3000 volts and transmitted a distance of six miles, to Sebenico, for light and power. With the commercial success of carbide manufacture by electric current in 1898 a second 5oo-HP. unit was added for experimental purposes in connection with two carbide furnaces. The present owners of the water rights, Societa per la utilizzazione della forze idrauliche della Dalmazia" of Trieste, started up a new plant at Jaruga in 1903, with two 350O-HP. double Francis turbines, operating under a head of 80 feet. They are directly connected to 3ooo-K.V.A., 42-cycle, two-phase alternators, making 315 R.P.M. The 15,000 generator voltage is directly transmitted, over 9 mm. conductors, to the carbide works, some 6 miles away, not far from the town Sebenico where the voltage is stepped-dewn to forty-eight by oil-cooled, water-circulated, single-phase transformers. The step-down station adjoins the carbide furnaces, so that the transmission line for 15,000 amperes is very short. The current from this plant is consumed in eight carbide furnaces, requiring, on the average, 5000 HP per hour throughout the year. With the increased demand for carbide, the factory has been recently extended to accommodate thirty-two furnaces, consuming, on the average, 32,000 HP. per hour throughout the year. For this 1 Author's article. Electrical Review, Jan. 9, 1909. Based on Data Submitted by the Designing and Constructing Engineers. 436 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. FIG. i. Manojlovac Plant, Dalmatia, Austro-Hungary. purpose, a new hydroelectric plant, of 24,ooo-HP. capacity, has been installed at Manojlovac Falls, near Kistanje, some 21 miles upstream, above the Sebenico carbide works. This plant, together with the above-mentioned earlier plants, was designed and installed by Ganz & Co., Budapest, who also supplied all the hydraulic, mechanical and electrical equipments of all these plants. Near the Manojlovac Falls, the river forms an S, and in the course of 1.2 miles has a drop of 360 feet. The flow varies greatly; in spring, due to snow thaws, TYPICAL HYDROELECTRIC PLANTS. 437 amounting to 1700 cubic feet per second, and in exceptionally dry summer season, to but 350 cubic feet per second. Manojlovac Plant. Just above the mentioned S, the river forms a natural lake, with an outlet over a natural dam, which is tapped 6.5 feet below the crest, where the inlet to the headrace is provided with three sluice gates. It will be seen that it was unnecessary to build a dam, yet sufficient water is impended for dry season. The headrace is 5250 feet long, and has a slope of 2 feet in 1000. It has a cross-section area of 117 square feet, cut through the solid rock of the mountain. To reduce skin friction, it is cement-coated up to the water level. FIG. 2. Plan of Manojlovac Plant, Dalmatia, Austro-Hungary. In order to save excavation, two separate collecting basins, joining each other, have been installed. As there are four penstocks, and due to the arrangement of the turbines in the generating room (a right and left hand turbine facing one another), there are two penstock beds. At the junction of the headrace and collecting basin are fine screens; each inlet to the penstocks is provided with a vertical swinging sector-gate, which is hydraulic- operated, the pressure being supplied by gravity from a reservoir situated on the mountain slope, some 165 feet above the collecting basin. The water for the reservoir is supplied by a small piston pump in the power house, driven by a Pelton wheel, under a head of 328 feet. By this arrangement, the piston pump has to supply water against a head of about 300 feet. Should the supply water fail, provision is made to operate the sector gates manually by worms and gears. Adjoining the collecting basin is a filtering system of three gravel filters, to supply the hydraulic governors of the turbines. The water is conveyed to same by means of cast iron bell and spigot pipes. The water to the filter system is supplied by a small channel, branching off from the main headrace. Thus the filtering is done by gravity, instead of under pressure, as is the case in many European power plants, where the 438 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. connections to the filters are made at the foot of the main penstocks. Of course, with the latter arrangement, a different kind of filtering system is adopted. The penstocks leave the collecting basin by bellmouthed connections; just outside of the wall are vents, so that, should the sector gates close before the turbines are cut off, the penstocks will not collapse. Each penstock is 558 feet long, 63 inches in diameter, having a shell thickness at the top of one-fourth inch and at the lower end, of nine-sixteenths inch. They were shipped in sections 19.7 feet long, and contrary to the usual practice of bolting same by means of flanges, the sections are riveted together. The penstocks rest on concrete piers; the lower ends are well anchored, while the upper ends are provided with expansion joints. Generator Room. The turbines are of the Francis spiral type, provided with two draft tubes, and operate under a head of 328 feet, and, with a water consumption of FIG. 3. Interior of Manojlovac Plant, Dalmatia, showing Four 3O,ooo-volt Generator Units. 212 cubic feet per second, develop, at 420 R.P.M., 66,000 HP. each. Owing to the large units, the double flow was adopted to obviate the side thrust, which in the single flow type is usually overcome by special thrust bearing. The counterbalancing effect is adjusted by regulating the guide vanes. The regulation of each turbine is accomplished by an hydraulic-actuated governor, which, when the revolutions exceed 10 per cent above the normal, operates a trip lever, which cuts off the supply. As the load is entirely for the manufacture of carbide, a very regular one, the governors come into play practically only when the turbines run away. It requires three seconds to cut off the supply from full to no gate. The generators of the Ganz & Co. type are rigidly coupled to the shafts of the TYPICAL HYDROELECTRIC PLANTS. turbines; they are of a very unique design. In order to eliminate step-up transform mers, the generators, which are the three-phase, 42-cycle type, are designed for 30,000 volts, and at 420 R.P.M., with a power factor 0.8, deliver 5200 K.V.A. each. The efficiencies at full and half load, with power factor 0.8, are 94 and 91 per cent, respectively. When running with full load and a power factor 0.8, at constant speed and excitation, a sudden dropping of the load will cause the voltage to rise 18 per cent. With maximum excitation, the windings will stand a short circuit for two minutes. The revolving field consists of a cast-steel ring shrunk upon a spider wheel; twelve cast-steel pole cores are fitted into dove-tailed slots and secured by conical bolts. . OF C/ FIG. 4. 6000 HP. Unit, Manojlovac Plant, Dalmatia, Austro-Hungary. The field windings consist of flat copper strips, wound on edge, and held in place by the pole shoe, which is part of the pole itself. The insulation of the coils consists of paper sheets, and the whole is incased in paper casing, formed to suit the coil. The whole revolving field, with shaft, weighs 26 long tons. The armature frame consists of halves, which again are split perpendicular to the axis, and when bolted together form a perfect circular ring. It will be observed in the illustration that the feet for the frame are removable; this was provided for the following purpose: the liability of a breakdown in a high-tension generator feeding directly into an overhead transmission line, is greater, owing to atmospheric dis- 440 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. charges, than one feeding an underground cable system, or that of a lower-tension generator, feeding transmission lines through step-up transformers. In the pit, the generator frame rests on two pairs of rollers, by means of which, after the feet have been removed, the whole frame can be revolved, and the lower section be brought on top and removed by the overhead crane, should it be necessary to inspect the coils in the lower half of the armature. By this arrangement it is not necessary to remove the revolving element of the generator. The coils are machine form-wound in five different shapes. Each is composed of a rectangular copper conductor, wound for twenty-six per slot per phase. The convolutions are insulated by several layers of Micanite, over which are wound several layers of insulating tape. Connections to the coils are made through brass terminals, soldered to the ends of the winding. Each generator has its own exciter mounted on the overhang of the shaft. The most striking feature of this arrangement is the method by which the exciting current is led to the revolving field. On the extensions of the carbon brush holders are the copper brushes bearing on the collector rings (one of which is insulated), mounted on the shaft, adjoining the commutator. The generator shaft is bored up to the field; through this hole the exciter current is supplied by an insulated cable. The return is through the shaft itself. The generator bearings are 37^ inches long and icf inches in diameter, lined with white metal, and are water-cooled. Switch Room. Parallel to the generating room in the middle and sunk in the opposite wall, is the switchboard. There are four generator panels, one collector or totalizing and three outgoing panels. Upon each generator panel are mounted, a rheostat for the exciter field; lever for the generator switch; a volt and ammeter, also voltmeter for excitation; phase lamps, synchronism indicator and double throw switch for parallel operation. The totalizing panel contains three ammeters and a totalizing recording wattmeter. Further, there are three automatic switch devices, which open the field circuits of all exciters, in case of an excess of generator voltage or current overload, or a diminu- tion of the generator pressure, and by means of the automatic turbine regulator, the water supply to the turbines is cut off. Each feeder panel has an ammeter and a pilot switch for the overload oil circuit breaker. Behind the switchboard is the switchroom, the low building above the tailrace; the tower at the end is for the outgoing lines. All the switches and measuring trans- formers for each machine are placed in concrete cells; wherever possible, the apparatus for each phase is in a separate cell. The generator switches are of the single-pole, oil type, actuated from the switch- board by means of cable and sheaves. The moving element of each phase of a switch is connected to a common operating shaft. Adjacent to the oil switch cells are those for the series and potential transformers, and so continue for the four generator units. On the roof of the cells are hook switches, also placed in cells. On top of these, is a single set of bus-bars; the different phases are separated by low partitions. TYPICAL HYDROELECTRIC PLANTS. 441 At one end of the bus-bars, after the fourth unit connections, are three double cells containing the general station-protecting devices, consisting, for each phase of the busses, of a condenser submerged in oil; a horn-gap provided with auxiliary gap and a multigap arrester shunted by a resistance placed in oil. From here the busses branch out into two feeders per phase for the two aerial circuits. The phases of each circuit are provided with overload circuit breakers, potential and series transformers for the ammeters and relays, and hook switches. From here the lines pass to the upper floor of the tower, and just before leaving the building, each is provided with the following combination of lightning protecting devices; a choke coil with capacity cylinder; a horn-gap with auxiliary gap, by means of which the main gap can be adjusted to a lower breakdown setting than the usual. The horns are shunted by graphite resistance-rods; further, a multigap arrester with shunted resistance; finally, the ground connection is made through continuous water- flow grounders, which lead off light static discharges. The conductors leave the building through porcelain bushings. It will be noticed that the protecting equipment is simple, yet very complete; this precaution had to be taken because the generators feed directly an aerial transmission line, which leads through a section of country passing over plateaus, canyons and valleys, very frequently visited by violent thunder storms and other atmospheric electrical JT IG c Insulator used on discharges. 30,000- volt Transmission Transmission System. Both circuits led to the carbide System, Manojlovac, .. 01- M T i j Dalmatia, Austro-Hun- factory near Sebenico, some 21 miles distant, and carried _., gary. practically the entire length on wooden poles spaced nor- mally 108 feet apart; the lowest conductor is some 19 or 23 feet above the ground. The conductors are 9 mm. copper wire, carried on three-piece, two-petticoat porce- lain insulators, the head diameter being 7 inches, the total height being 8.5 inches. The pin and first petticoat are held together by a glaze of lead and glycerine. The head or second petticoat rests on the first, with an air space between, formed by the ribs on the inside of the head. The insulators were tested at 80,000 volts, and during operation none have broken down. Where the transmission line crosses small country roads, the poles are placed on either side of the road, where they are also provided with grounded guard arms, so that in case of breakage, the line is grounded; as the spacing is so close, a broken conductor cannot touch the ground. To take up side stresses on turns, the circuits are carried on A-frames. To protect the wooden poles against light- ning, each has a pointed castiron cap with a ground wire. Fig. 6 shows a latticed construction used in crossing the highway and telegraph lines. The bottom and sides of this steel construction are provided with a wire netting. 442 HYDROELECTRIC DEVELOPMENTS AND ENGINEERING. The transmission lines enter the carbide plant with a similar lightning protection equipment as in leaving the power-house tower, with the exception of the water-flow grounders, owing to the lack of fresh water. After passing the lightning arresters, connections of the two circuits are made to a common bus-bar system by automatic oil circuit breakers, and series transformers with their recording instruments. There are installed 12 single-phase oil-cooled water-circulated transformers of 1500 K.V.A. each, stepping-down the line voltage, which is here 26,000, to 48 volts, FIG. 6. Transmission Line crossing Highway and Telegraph. In Rear Carbide Plant. Manojlovac System, Dalmatia, Austro-Hungary. used in the carbide furnaces. The wiring system is so arranged that from the con- trol panel at each furnace, the transformer feeding same can be thrown on to any phase in order to balance the load of the circuit. Again, the division of load between the two furnaces of one transformer is indicated on a differential meter expressing .the division of lead in per cent. There are further two I50-K.V.A. 26,ooo/33O-volt three phase transformers to operate auxiliary apparatus, such as pumps supplying salt water for cooling the transformers; crushers and conveyors for limestone, coal and carbide; repair shops and driving the ventilators of the furnaces, etc. The Manojlovac plant has been continuously in operation since the earlier part of 1907, and has given entire satisfaction; no trouble has been experienced with the transmission line, or the high tension generators, although the country was frequented by heavy storms and electrical discharges. INDEX INDEX. A. Action of horn gap lightning arrester, 310. Adaptability of wooden penstocks, 78. Air compressors, 281. Aluminum lightning arrester, 318. American and European hydroelectric develop- ments, 325. American Fork plant, penstocks, 85. turbine, 130. Ammeter, 191. Anchors, 71. Anchor bolts, 106. insulator, 271. Appendix, 325. Application of lightning arresters, 312. Architectural features, 107. Area of circles, 60. Arrangement of substations, 280. B. Bar Harbor, Ellsworth, power plant, 27. Basin, collecting, 56. Bear traps, 35. Bearing power of soil, 103. Beznau roller sluice gate, 33, 34. Bibliography on mechanical equipment of power plants, 165. on dams, 38. on electrical equipment of power plants, 211. on headraces, 58. on high tension transmission, 278. on hydraulic developments, 18. on line protection, 324. on substations, 307. on penstocks, 87. Bishop Creek plant, penstock, 84. Bolts, anchor, 106. Buildings, 101. bearing power of soil, 103. character of soil, 103. concrete mat construction, 106. excavation, 101. foundations, 101, 106. piling, 104. site for building, 101. Buijdings Continued test holes, 101. test of piles, 105. weight of masonry, 104. Bus bars, 201. compartments, 202. room, Lontsch plant, 203. room, Lucerne plant, 203. Ontario, 340. outdoor, 346. Butterfly dam, 34. Brusio plant, Swiss-Italian, 417. collecting basin, 420. current supply, 426. distribution, 433. gate house, 421. generators, 423. headrace, 419. headrace tunnel, 46. insulators, 267. open air cooled transformer, 433. penstocks, 420. penstock flange, 68. penstock flap, 75. power house, 423. protecting device, 430. railway crossing, 431. secondary water supply, 419. siphon system, 419. step-up station, Piattamala, 426. substation, Castellanza, 434. substation, Lomazzo, 432. switch gear, 424. tower, 247. transmission lines, 430. transmission tunnel, 427. turbines, 423. C. Canals, 40. Cantilever, Niagara Crossing, Ontario, 343. tower, Obermatt, 246. Capacity and discharge of penstock, 63. Castellanza substation, oil switches, 209. Castelnuovo-Valdarno, switch house, 178. Channels, water velocity, 41-42. 445 446 INDEX. Chanoine dam, 36. Chevres plant, France, 54. Chicago drainage canal, 34. Choke coils, 315. Circles, area, 60. Circuit breakers, 209. Clamps and ties for insulators, 265, 273. Concrete, dams, 23. mat construction, 106 penstocks, 86. piling, 105. Concreted wooden poles, 232. Conductors, method of tying, 274. wind pressure, 236. Conduits, 40. Converter, field connections, 298. frequencies, 296. phase, 298. starting, 299. Converters, 296. compounding, 301. hunting, 299. reactances, 301. Cost of current, 3. of developments, 3. of wooden penstocks, 81, 83. Costs, Uppenborn, 416. Couplings, 153. Coffer dam, 26, 29. Cooling of circulating water for transformers, 293- Collecting basin, 56. basin, Brusio, 74, 420. Colliersville plant, 90-91. Columns, 123. Compounding, rotary converters, 301. Compound Francis turbine, 135 Conductors, 215. alternating current conductor, 219. alternating current, problem, 219. cables, copper, 224. cables as, 216. characteristics, 217. direct current, 218. elasticity of, 216. size, 217. spacing, 217. strength of, 215. corona effect, 226. direct current problem, 218. reactance volts, table, 223. transposition of conductors, 226. wire gauges, comparison, 220. wire, solid copper, 221, 225. stranded copper, 222. weight and strength of, 223. Crane, 117. Crib dam, 26. Cross arms, 231. Curtain method, testing turbines, 159. Cylindrical dams, 36. gate, 56. D Dams, 19. behavior of resultants, 24. bibliography, 38. butterfly, 34 Chanoine, 36. Charlotte, 352. coffer, 26, 29. concrete, 23. crib, 26. cylindrical, 36. earth construction, 32. earth, Necaxa, 31. Ellsworth, 28. gravity, 19. Heimbach, 394. masonry, 22. Necaxa, 372. needle, 36. Patapco, 25. reinforced concrete core, 32. steel frame, 30. submerged power plant, 25. timber, 29. Dixville, 32. Deflector in headrace, 49. Designing staff, 16. Detail of drum gate, 55. Development, economy, 15. first costs, 15. Developments, investigation, 3. Direct current switchboards, 187. Doors, 1 1 6. Draft tubes, 141. Drawings, charge of extra work, 16. checking, 16. and specifications, 16. Drum gate, 54. detail, 55. Duluth substation, no. E Earth dam, Necaxa, 31. Economy in development, 15. Economical spans, 250. Electrolysis prevention, 127. Electrolytic arrester, 320. Electrical equipment of power plants, 167. bibliography, 211. INDEX. 447 Elevators, 116. Ellsworth dam, 28. El Oro, substation, 381. Equalizing chambers, 76. European methods of testing turbine, 157. Exciters, 173. wiring diagram, 198. Expansion slip joints for penstocks, 71-72. Expansion joints of structural steel, 124. F. Field office, 17. Financial aspect, 362, 403. Fishways, 37. Flanges, 66-75. Flashboards, 36. Floors, in, 124. Floating foundation for transmission tower, 344- Flow of river, 10. Fluid arresters, 318. Flumes, masonry, 42. timber, 42, 44. Flywheels, 153. Flywheel alternator, 169. with internal stationary armature, 169. Foundations, 106. for steel towers, 236, 263. Four-legged towers, 240, 344. twin tower, 239. Forest preservation, 4. Forebays, 88, 329. Francis turbines, 131. Frequencies, 173, 364. Frequency changers, 303. meter, 193. Friction in steel penstocks, 61. in tunnels, 45. in wooden penstocks, 80. Gate house, 74, 331, 420. valves, 57. Gates and racks, 47. Generators, 167. auxiliaries, 332. 30,000 volt, 175. leads, i74,33 2 - umbrella type, 168. Brusio, 423. Charlotte, 355. Heimbach, 397. Kykkelsrud, 389, Manojlovac, 438. Necaxa, 376. Generators Continued Ontario, 332. Uppenborn, 407. Geneva plant, 114. Georgia plant, 89. Gola lightning protection, 314. Governors, 143. Bell, 145- Escher Wyss, 144. Lombard, 146. Glocker-White, 147, 149. Replogle, 148, 150. Governmental reports, 13. Gross horsepower of falling water, 7. Gradient, hydraulic, 7. Gravity dams, 19. Great Falls plant, Southern Power Company, Charlotte, N. C., 348. auxiliary power, 367. dam, 352. financial aspect, 362. frequency, 364. generators, 355. high tension room, 359. insulator, 364. lightning protection, 360. oil switches, 360. penstocks, 80. power development, map, 350. power house, 355. secondary power, 367. spillway, 348. towers, 362. transformers, 358. transmission feeder circuit, 361. lines, 360, 366. system, map, 349. turbines, 352. voltage, 365. wiring diagram, 356. Guard wire, 231. Guys, 232. H. Hafslund, deflector and rack, 49. Hamilton cataract turbine, 140. Hauser Lake, dam, 30. Heimbach plant. See also Urfttalsperre, 113. Heimbach plant, wiring diagram, 199. tower, 246. Head, loss, 6. Headrace, arrangement, 39-40. bibliography, 58. Brusio, 419. Heimbach, 395. Kykkelsrud, 382. scheme, 39. 448 INDEX. Headrace Continued Sillwerke, 120. tunnel, Brusio, 46. Uppenborn, 405. Heating, 117. factors of radiating surface, 117. High head plant, 98. Holyoke, plant, 89. tests, 160. test flume, 161. Horn gaps, 311. Horn-gap construction, 313. setting, 313. Horsepower, gross, of falling water, 7. Hunting of converters, 299. Hydraulics, 5. fundamental formulae, 6. principal formulae, 6. Hydraulic gradient, 7. pipes, riveted, 65. relief valve, 148. I. Induction generator, 167. regulator, 301. Innsbruck plant, 174. Installation of multigap arresters, 317. Instrument pedestal, 185. Insulators, 265. anchor, 271. Charlotte, 364. Manojlovac, 441. Necaxa, 380. Paderno and Brusio, 267. strain, 271. suspended, 268. Swiss and Italian, 266. Uppenborn, 410. Insulator pins, 272-274. Insulator tie and clamp, 265, 273. Insulating and rolling support, 271. Investigation of developments, 3. Iron sluice gates, 54. Italian insulators, 266. steel tower, 247. J Jajce plant, penstock anchor, 70. wedge shaped expansion joint, 72. K. Kaiserwerke, penstock support, 71. Kern River plant, 97-99. insulators, 265. Kern River Plant Continued penstock, 66. Kykkelsrud plant, Norway, 102, 382. exciter units, 388. generators, 389. headrace, 382. power house, 386. substation, Hafslund, 392. switchboard room, 390. transformer room, 390. transmission line, 391. turbines, 386. L. Lavatories, 119. Laws of hydraulics, 5. Leaders, roof, 116. Length of economical spans, 261. Lighting, 119. Lightning arresters, 309. fluid, 318. horn type, 311. location, 323. multigap, 315. principle, 309. Lightning discharges, 309. Line protection, 309-360. bibliography, 324. disconnecting switches, 275. stresses, 251. Location of arresters, 323. of substations, 280. Loch Leven plant, penstock, run, 73. Lockport substation, Ontario, 345. Lontsch plant, bus bar room, 203. Loss of head in penstock, 6, 61, 62. Low head plants, 90. Lucerne, cantilever tower, 246-248. plant, 112. bus bar room, 203. turbine, 144. wiring diagram, 200. Lyon plant, France, 56. M. Manojlovac plant, Dalmatia, 435. building, 437. drainage area, 436. exciters, 440. generators, 30,000- volt, 175, 438. highway crossing, 442. insulators, 441. lightning arresters, 442. line protection, 441. INDEX. 449 Manojlovac plant Continued switch room, 440. transformers, 442. transmission poles, 441. transmission system, 441. turbines, 438. Masonry, dams, 22. flumes, 42. weight, 104. Material, building, in. McCall Ferry plant, 92-94. Mechanical equipment of power plants, 129. bibliography, 165. Medium head plant, 91, 131. Meter, Venturi, u. water current, n. Method of plotting discharge of rivers, 12. of plotting river bed, 13. Medium head plants, 91. Miner's inch, 8. Molinar plant, Spain, method of cooling circu- lating water, 293. Montreal substation frequency changers, 304. Morgan Fall, Georgia, dam, 22. Mountain lakes, siphoning, 47. Motor generators, 302. Motor generating station, Vienna, 302. Multigap arresters, 315. Muskegon no,ooo-volt insulator, 270. tower, no,ooo-volt, 249. N. Necaxa plant, Mexico, 369. building, 374. conductors, 381. dams, 31, 372. development, map, 370. drainage area, 272. generators, 376. insulators, 380. oil switches, 380. penstocks, 372. penstock flange, 69. transmission system, 380. substation, El Oro, 381. switching room, 377. switchboard, 181, 380. towers, 380. transformers, 377. turbines, 376. wiring diagram, 197. wiring system, 377. Needle dam, 36. New York Central Towers, 245. Niagara crossing, 342. Niagara crossing tower, 240. Niagara, Lockport and Ontario development, 327- Niagara Falls, 327. Niagara Falls Power Company plant, 48, 93, 95, 97, 107, 109. O. Obermatt, cantilever tower, 246. plant, 112. wiring diagram, 200. switch room, 176-177. Oil filtering tanks, 154. piping, 156. pumps, 156. required, 154. rheostat, 312. switches, 204. Oiling system, 154. Ontario plant, bus bars, 340. bus bars, outdoor, 346. cantilever, Niagara, crossing, 343. circuit breaker, 6o,ooo-volt, 347. control room, 340. distributing station, 338. exciters, 332. floating foundation for transmission tower, 344- forebay, 329. four-legged tower, 344. gate house, 331. generator auxiliaries, 332. generators, 332. generators, leads, 332. high tension room, 339. insulator and pin, 273. low tension room, 339. Niagara crossing, 342. oil switch compartment, 207. Oneida tower, 243. open air fuses, 343. outdoor bus bars, 346. penstocks, 330. power house, 332. screen house, 330. substation, Lockport, 345. transformer room, 339. transmission line, 341. three-legged towers, 341. turbines, 332. wiring diagram, 196. wiring system, 339. Outdoor bus bars, Ontario, 346. disconnecting switch, 276. 450 INDEX. Open air fuses, Ontario, 343. Overload relays, 209. voltage relay, 211. P. Paderno insulator, 268. Painting of structural steel, 127. Part I, i. Part II, 215. Part III, 323. Patapco dam, 25. plant, 174. Penstocks, 59. anchor, 70-71. air cushion, 76. American Fork plant, 85. bibliography, 87. Brusio, 74, 420. flanges, 66, 67, 68, 69, 71, 75. flap, 75. head loss, 61,62. Kern River plant, 66. Loch Leven, 73. Necaxa, 372. Ontario, 330. protection, 76. run, 59. reinforced concrete, 86. shell, strength, 64. size, 59. slip joints, 71, 72. steel construction, 64. support, hinged, 71. vents, 50. wooden, Bishop Creek, 84. Piles, tests, 105. Piling, 104. Pin insulators, 265. Piping, oil, 156. Poles and towers, tests, 238. Poles, see wooden poles. Pole and tower construction, 228. Porcelain base insulator pin, 272. Portability of steel towers, 237. Poschiavo siphon system, 47. Power factor meter, 192. Power plants, 88. Bar Harbor plant, Ellsworth, 27. Brusio, Swiss-Italian plant, 417. Colliersville plant, 90, 91. Georgia plant, 89. Great Falls plant, 348. Holyoke plant, 89. Kern River plant, 97,99. Kykkelsrud-Hafslund plant, Norway, 382. Lyon plant, France. Power plants Continued Manojlovac plant, Dalmatia, 435. McCall Ferry plant, 92, 94. Necaxa plant, Mexico, 369. Niagara Falls Power Company's plant, 93, 95, 97- Ontario plant, 309. Puget Sound plant, 180. Schaffhausen Low Head plant, 132. Shawinigan plant, 92. Sill plant, Tirol, 120. Snoqualmic plant, 100. Stuttgart plant, no. Toronto plant, 96. Uppenborn plant, Munich, Germany, 403. Urfttalsperre plant, Heimbach, Germany, 393- Winnipeg plant, 89. Preservation of wooden poles, 231. Preservation of forest, 4. Pressure tunnels, 45. Principle of arresters, 309. Profile of rivers, 13. Properties of timber, 53. Protection of penstocks, 76. Pumps, oil, 156. R. Racks and gates, 47. Railway crossing, Brusio, 431. Reactance, converters, 301. Regulating devices, 143. Regulation of generators, 170. Reinforced concrete, dams, 23. penstocks, 86. poles, 232. tower, 234. Relief valve, 153. Remote control switchboards, 185. Reports, governmental, 13. Reservoirs, 56. Reverse current relay, 211. Revolving armature generator, 169. field alternator, 170. Rheostats, 193. River bed, method of plotting, 13. River, flow, 10. Rivers, method of plotting discharge, 12. profile, 13. velocity of flow, 7. Riveted hydraulic pipes, 65. Rochester dead end tower, 242. Rolling support for long spans, 271. Roof truss, 122. Rotary converter connections, 291. INDEX. 451 S. Saddles for penstocks, 72. Sag at different temperatures, 252. Sandtraps, 57. Screens, 49. Screen, detail, 51. Screen house, 48, 330. Scholes, D. R., paper on transmission line towers and economical spans, 253. Secondary power, Charlotte, 367. Second-feet, 13. Section house, Uppenborn, 414. Section switches, 274. Seepage in tunnels, 45. Setting of horn gaps, 313. Shawinigan plant, 92, 176, 179. Siegwart concrete poles, 233. Siphoning lakes, 47. Siphon system, 46, 419. Site for buildings, 101. Size of bus bars, 201. of units, 280. Sleet on conductors, 236. Sluice gate, Stoney, 33-34. Sluice gates, iron, 54. wooden, 50-52. Snoqualmie Falls plant, 100. Soil, bearing power, 103. character of, 103. Spacing of bands on wooden penstocks, 78. Spans economical, 250. economical length, 261. Spier Fall, dam, 22. Specifications and drawings, 16. Specification of steel towers, 243. Spillways, 57. Spillway, Charlotte, 348. Spoon wheel turbine, 144. Stairways, 116. Standpipes, 76. Stansstad substation, 115. Starting of converters, 299. Steel, frame dam, 30. insulator pin, 272. penstocks, 59. penstock, friction, 61. pipe towers, 233. Steel tower, cantilever, 246. foundations, 236, 263. Heimbach, 246. New York Central, 245. Oneida, 243. specifications, 243. Syracuse, 240, 241. Steel towers and poles, tests, 238. for suspended insulators, 249, 250. Steel towers and poles Continued two-legged, 237. wind pressure, 235. Steel transmission towers, 235. Steghof substation, general arrangement, 285. Stoney roller sluice gate, 33, 34. Strain insulators, application, 271. Strength of Douglas fir, 81. of penstock shell, 64. Structural steel, 122. character of steel, 125. column bases, 123. crane column, 124. expansion joints, 124. fiber stresses, 125. floors, 124. inspection, 127. painting, 127. prevention of electrolysis, 127. typical columns, 123. typical roof trusses, 122. workmanship, 126. Substations, 280. Substation, arrangements, 280. bibliography, 307. Castellanza, 434. drainage, 281. Duluth, no. Heimbach, 401. Hirschau, 411, 415. Hafslund, Kykkelsrud, 390. location, 280. Lomazzo, 432. Piattamala, 426. Stansstad, 115. Steghof, 285. Steghof, motor generator, 303. switchboard panel, 305. switch gear, 307. typical arrangement, 283. ventilation, 281. Waterbury, 281. Superstructure, 107. Survey, Geological, United States, 13. Suspended, insulator, 268. insulator tower, 249-250. Swiss-Italian steel tower, 247. Swiss insulators, 266. penstock flanges, 66, 67. switching room arrangement, 183, 184, 186. typical bus bar and wiring system, 195. wiring practice, 195. Switchboards, 176. bus bars chambers, 190. combined panel, 187. D. C. board, 187. desk, type, 185. 452 INDEX. Switchboards Continued exciter or D. C. panel, 186. high tension A. C. boards, 191. instrument bench, 186. low tension A. C. board, 188. oil switch arrangement, 190. panel switchboard, 183. pedestal, or column type, 185. types of switchboards, 182. wagon panel, 188, 189, 414. Switchboard, equipment, 191. gallery, 116. Switch gear, Brusio, 424. for substations, 307. Necaxa, 380 Ontario, 356. panels for substations, 305. Uppenborn, 407. Switching room, 176. Heimbach, 399. Kykkelsrud, 390. Manojlovac, 440. Necaxa, 377. Puget Sound, 180. Synchronizing, 192. Syracuse 45-foot tower, 240-241. Systems of wiring, diagram, 194. T. Tailrace measurement, 159. Tanks, supply, 156. Taylor's Fall, insulators, 265. Telephone, Uppenborn, 416. Telluride plant, 139. Test of American woods, 53. Test holes, 101. Tests of transmission poles and towers, 238. of reinforced concrete poles, 233. Testing turbines, 157. Three-legged towers, 240-341. Thrust bearing, 137. Ties and clamps for insulators, 265, 273. Timber dam, 29. flumes, 42, 44. properties, 53. Tivoli plant, 121. Tofwehult-Westerwik rolling support, 271. Torchio lightning protection, 314. Toronto plant, 96. Tower, dead end, Rochester, 242. reinforced concrete, 234. Towers, Charlotte, 362. Necaxa, 380. steel pipe, 233. Transmission, feeder circuit, Charlotte, 361. Transmission Continued high tension, bibliography, 279. lines, Brusio, 430. Charlotte, 360-366. Heimbach, 401. Kykkelsrud, 391. Ontario, 341. stresses, 251. towers and economical spans, 253. poles, Manojlovac, 441. see wooden poles, system, Manojlovac, 441. Necaxa, 380. Uppenborn, 410. towers, Uppenborn, 412. see steel towers, transformers, Manojlovac, 442. tunnel, Brusio, 427. Transformation of water power, i. Transformers, 286. air cooled, 293. air required for, 294. arrangement of air blast, 294. Charlotte, 358. characteristics of, 287. connections of, 290. core type, 286. delta vs. Y connection, 290. efficiency of, 289. forced oil-cooled transformers, 293. method of connecting transformers to rotary converters, 291. oil circulation for cooling, 292. oil cooled, 292. regulation of, 288. shell type, 286. transformer connections, 290. characteristics, 287. efficiency, 289. Necaxa, 377. oil circulation, 292. open air cooled Brusio, 433. regulation, 288. room, Kykkelsrud, 390. room, Ontario, 339. types of, 286. Uppenborn, 407. Traps, bear, 35. Trenches, 41. Trenton Water Fall plant, 79. Tretzo tower, 241. Tunnels, friction, 45. pressure, 45. seepage, 45. Turbines, 129. accessories, 152. INDEX. Turbines Continued Brusio, 423. Charlotte, 352. curtain carriage testing, 159. European methods of testing, 157. graphical registrator, 158. high head turbines, 138. Holyoke testing flume, 160. Heimbach, 397. Kykkelsrud, 386. low head, 130. medium head, 131. Manojlovac, 438. Necaxa, 376. Ontario, 332. testing, 157. Uppenborn, 406. Two-legged tower, Gaucin-Seville, Spain, 237. tower, Moosburg, 237. towers, type used in Switzerland and Italy, 237- Typical arrangement of headrace, 39, 40. of substations, 283. Types of oil switches, 204. steel towers, 259. switchboards, 182. U. Umbrella type generator, 168. United States Geological Survey, 13. Uppenborn plant, Germany, 403. costs, 416. generators, 407. generating plant, 406. headrace, 404. horn-gaps, 415. insulators, 410. lightning arrester station, 411. lightning protection, 409. section house, 414. sluice gates, 405. substation, Hirschau, 411-415. switch gear, 407. telephone, 416. transmission system, 410. transmission towers, 412. transformers, 407. turbines, 406. wall outlet, 408. water flow grounders, 413. Urfttalsperre, Heimbach plant, Germany, 393. dam, 394. financial aspecr, 40;, generators, 397. headrace, 395. Urfttalsperre, Heimbach plant C power house, 396. substations, 401. switching room, 399. transmission lines, 401. turbines, 397. V. Vacuum relief valve, 77. Vandoise penstock relief valve, 77. Velocity of flow in rivers, 7. Ventilating of power plants, 118. Ventilation of substations, 281. Venturi meter, u. Vents for penstocks, 50. Vienna motor generator station, 302. Voltage, 173. of converter and frequency, 296. Voltmeter, 191. W. Walls, in. Wall outlets, typical, 276-278. outlet, Uppenborn, 408. Water, current meter, n. flow grounders, Uppenborn, 413. flow grounders, 321. power transformation, i. velocity in channels, 41-42. Waterbury substation, 281. Wattmeter, 191. Weight of masonry, 104. Weir dam, 8. Weirs, construction, 9. quantity of water passing over, 9. Windows, 116. Wind pressure on conductors, 236. on steel towers, 235. Winnipeg plant, 89. Wiring diagrams, 194. Charlotte, 356. for exciters, 198. Heimbach plant, 199. Necaxa plant, 197. Obermatt plant, 200. Ontario plant, 196. of a single generator and step-up trans- former, 201. Waterbury, substation, 306. wiring system, Ontario, 339. Wooden flumes, 42. piling, 105. penstocks, 78. comparative costs, 83. wiv. OF a 454 INDEX. Wooden penstocks Continued connection to smaller pipes, 82. construction, 84. cost, 81. durability, 80. friction, 80. spacing of bands, 78. poles, 228. "A" frame tower, 230. cross arms, 231. for three-phase circuit, 229. kind of wood, 229. Wooden Poles Continued life of, 231. preservation, 231. strength, 229. 50,000 volts, Taylor's Falls, 228. pole line construction, 232. pole guys, 232. pole, concreted, 232. sluice gates, 50, 52. Woods, test of American, 53. Working strain of penstock bands, 70. Workmanship on structural steel, 126. Uniform with "Hydroelectric Developments and Engineering." 8x11 inches. 473 Pages. 500 Illustrations. 80 Tables. $5.00 Net. STEAM-ELECTRIC POWER PLANTS BY FRANK KOESTER CONTENTS: Problems, Efficiency and Cost of Plants. Location. General Layout. Coal Storage. Condenser Water Supply. Excavation and Foundations. Steel Work. Archi- tectural Features. Boilers. Mechanical Stokers and Grates. Coal. Combustion. Draught. Smoke Flues. Chimneys. Boiler Feed Water. Feed Water Heater. Superheaters. Super- heated Steam. High Pressure Piping. Low Pressure Piping. Prime Movers. Reciprocating Engines. Turbines. Condensers. Pumping Machinery. Oiling System. Electrical Equipment. The Design of Small Light and Power Plants. Testing Power Plants. Descriptive Discussion of Five Typical American and European Light and Power Plants. Principal Dimensions and Data of Recently Constructed Light and Power Plants. Tables per K. W. Capacity. Series of Tables. Recommended by the technical journals, to experts and advanced engineers. Adapted as text by leading universities. SOME OPINIONS " The book represents a lot of diligent work, much research, and investigation ; it is fully up to the high standard of the author, who has an international reputa- tion in the engineering profession." Electrical Review, " This book will undoubtedly take a high place among classical works of the industry ; it is evidently the result of an exceptional experience, such as falls to the lot of very few engineers." The Electrician. " It would be difficult to mention any detail that is not touched upon." Power and The Engineer. " A very considerable amount of information. The Engineering Digest. " Probably no other volume contains so much information." The Engineering Record. "To a great extent a classic on the subject." Engineering Times. "Considerable value to any student of power engineering." Sibley Journal. " It is the work of a very experienced engineer ; everybody interested in power plants should have this book." Zeitschrift des Vereines deutscher Ingenieure. "A valuable guide to modern practice." Engineering. " A comprehensive and useful volume." Engineering Review. " It is a conscientious guinea's worth." Electrical Times. D. VAN NOSTRAND COMPANY PUBLISHERS AND BOOKSELLERS 23 MURRAY AND 27 WARREN STREETS = = NEW YORK ' LIST OF WORKS ON ELECTRICAL SCIENCE PUBLISHED AND FOR SALE BY D. VAN NOSTRAND COMPANY, 23 Murray and 27 Warren Streets, New York. * WRITE FOR COMPLETE 96 PAGE CATALOG, GRATIS ABBOTT, A. V. The Electrical Transmission of Energy. A Manual for the Design of Electrical Circuits. Fifth Edition, enlarged and rewritten. With many Diagrams, Engravings and Folding Plates. 8vo., cloth, 675 pp Net, $5.00 ADDYMAN, F. T. Practical X-Ray Work. Illustrated. 8vo., cloth, 200 pp Net, $4.00 ALEXANDER, J. H. Elementary Electrical Engineering in Theory and Prac- tice. A class-book for junior and senior students and working electricians. Illustrated. 12mo., cloth, 208 pp $2.00 ANDERSON, GEO. L. Handbook for the Use of Electricians in the operation and care of Electrical Machinery and Apparatus of the United States Seacoast Defenses. Prepared under the direction of Lieut. -General Commanding the Army. Illustrated. 8vo., cloth, 161 pp $3 .00 ARNOLD, E. Armature Windings of Direct-Current Dynamos. Extension and Application of a general Winding Rule. Translated from the original Ger- man by Francis B. DeGress. Illustrated. 8vo., cloth, 124 pp $2.00 ASHE, SYDNEY W. Electricity Experimentally and Practically Applied. 422 illustrated. 12mo., cloth, 375pp. Net, $2.00 ASHE, S. W., and KEILEY, J. D. Electric Railways Theoretically and Practi- cally Treated. Illustrated. 12mo., cloth. ^ Vol. I. Rolling Stock. Second Edition. 285 pp Net, $2 . 50 Vol. II. Substations and Distributing Systems. 296 pp Net. $2.50 2 LIST OF WORKS ON ELECTRICAL SCIENCE. ATKINSON, A. A. Electric and Magnetic Circulations. For the use of Electrical Engineers and others interested in the Theory and Application of Electricity and Magnetism. Third Edition, revised. Illustrated. 8vo., cloth, 310 pp. Net, $1 . 50 ATKINSON, PHILIP. The Elements of Dynamic Electricity and Magnetism. Fourth Edition. Illustrated. 12mo., cloth, 405 pp $2 .00 Elements of Electric Lighting, including Electric Generation, Measurement, Storage, and Distribution. Tenth Edition, fully revised and new matter added. Illustrated. 12mo., cloth, 280 pp $1 .50 Power Transmitted by Electricity and Applied by the Electric Motor, including Electric Railway Construction. Illustrated. Fourth Edition, fully revised and new matter added. ]2mo., cloth, 241 pp $2 .00 AYRTON, HERTHA, The Electric Arc. Illus. 8vo., cloth, 479 pp.. . .Net, $5.00 W. E. Practical Electricity. A Laboratory and Lecture Course. Illus- trated. 12mo., cloth, 643 pp $2.00 BAKER, J. T. The Telegraphic Transmission of Photographs. 63 Illustrations. 12mo., cloth, 155 pp Net, $1.25 BEDELL, FREDERICK. Direct and Alternating Current Manual. Assisted by C. A. Pierce. Second Edition, enlarged. Illustrated. 8vo., cloth, 365 pp. In Press BEDELL, F. & CREHORE, ALBERT C. Alternating Currents. An analytical and graphical treatment for students and engineers. Fifth Edition. 112 illus- trations. 8vo., cloth, 325 pp Net, $2.50 ELAINE, ROBERT, G. The calculus and Its Applications. A practical treatise for beginners especially engineering students. 79 illustrated. 12mo., cloth, 330 pp Net, $1.50 BIGGS, C. H. W. First Principles of Electricity and Megnetism. Illustrated. 12mo., cloth, 495 pp $2.00 BONNY, G. E. The Electro-Plater's Hand Book. rX A Manual for Amateurs and Young Students on Electro-Metallurgy. Fourth Edition, enlarged. 61 illustrations. 12mo., cloth, 208 pp $1 .20 BOTTONE, S. R. Magnetos For Automobilists How Made and How Used. A handbook of practical instruction on the manufacture and adaptation of the magneto to the needs of the motorist. Second Edition, enlarged. 52 Illustrations. 12mo., cloth, 188 pp Net, $1 .00 Electric Motors, How Made aud How Used. Illustrated. 12mo., cloth. 168 pp 75 cents BOWKER, WM. R. Dynamo, Motor, and Switchboard Circuits for Electrical Engineers : A practical book dealing with subject of direct, alternating, and polyphase circuits. Second Edition, greatly enlarged. 130 illustrations. 8vo., cloth, 180 pp Net, $2 .00 BROADFOOT, S. K. Motors, Secondary Batteries and Accessory Apparatus. Illus- trated. 16mo., cloth. (Installation Manuals Series) In Press LIST OF WORKS ON ELECTRICAL SCIENCE. 3 CARTER, E. T. Motive Power and Gearing for Electrical Machinery; a treat- ise on the theory and practice of the mechanical equipment of power stations for electric supply and for electric traction. Second Edition, revised. Illustrated. 8vo., cloth, 700 pp Net, $5.00 CHILD, CHAS. T. The How and Why of Electricity: a book of information for non-technical readers, treating of the properties of Electricity, and how it is generated, handled, controlled, measured, and set to work. Also explaining the operation of Electrical Apparatus. Illustrated. 8vo., cloth, 140 pp $1 00 CLARK, D. K. Tramways, Their Construction and Working. Second Edition. Illustrated. 8vo., cloth, 758 pp $9.00 COOPER, W. R. Primary Batteries: their Theory, Construction, and Use. 131 Illustrations. 8vo., cloth, 324 pp Net, $4.00 The Electrician Primers. Being a series of helpful primers on electrical subjects, for the use of students, pupils, artisans, and general readers. Second Edition. Illustrated. Three volumes in one. 8 vo., cloth, Net, $5 00 Vol. I. Theory Net, $2 .00 Vol. II. Electric Traction, Lighting and Power Net, $3.00 Vol. III. Telegraphy, Telephony, etc Net, $2.00 CROCKER, F. B. Electric Lighting. A. Practical Exposition of the Art for the use of Electricians, Students, and others interested in the Installation or Operation of Electric-Lighting Plants. Vol. I. The Generating Plant. Sixth Edition, entirely revised. Illustrated. 8vo., cloth, 482 pp $3.00 Vol. II. Distributing System and Lamps. Sixth Edition. Illustrated. 8vo., cloth, 505 pp $3 .00 and ARENDT, M. Electric Motors: Their Action, Control, and Application. 160 Illustrations. 8vo., cloth, 296 pp Net, 2.50 and WHEELER, S. S. The Management of Electrical Machinery. Being a thoroughly revised and rewritten edition of the authors' " Practical Manage- ment of Dynamos and Motors." Eighth Edition. Illustrated. 16mo., cloth, 232 pp Net, $1 .00 GUSHING, H. C., Jr. Standard Wiring for Electric Light and Power. Illustrated. 16mo., leather, 156 pp $1 .00 DAVIES, F. H. Electric Power and Traction. Illustrated. 8vo., cloth, 293 pp (Van Nostrand's Westminster Series.) Net, $2 . 00 DAWSON, PHILIP. Electric Traction on Railways. 610 Illustrations. 8vo., half leather, 891 pp Net, $9.00 DEL MAR, W. A. Electric Power Conductors. 69 illustrations. 8vo., cloth, 330 pp. Net, $2.00 DEVEY, R. G. Mill and Factory Wiring. Illustrated. 16mo., cloth. (Installa- Manuals Series) In Press DIHGER, Lieut. H. C. Handbook for the Care and Operation of Naval Machinery. Second Edition. Illustrated. 16mo., cloth, 302 pp Net, $2.00 4 LIST OF WORKS ON ELECTRICAL SCIENCE DYNAMIC ELECTRICITY; Its Modern Use and Measurement, chiefly in its appli- cation to Electric Lighting and Telegraphy, including: 1. Some Points in Elect. ic Lighting, by Dr. John Hopkinson. 2. On the Treatment of Elec- tricity for Commercial Purposes, by J. N. Schoolbred. 3. Electric-Light Arithmetic, by R. E. Day, M.E. Fourth Edition. Illustrated. 16mo., boards, 166 pp. (No. 71 Van Nostrand's Science Series.) 50 cents EDGCUMBE, K. Industrial Electrical Measuring Instruments. Illustrated. Svo., cloth, 227 pp Net, $2.50 ERSKINE-MURRAY, J. A Handbook of Wireless Telegraphy : Its Theory and Practice. For the use of electrical engineers, students, and operators. Third Edition, revised and enlarged. 180 illustrations. 8vo., cloth, 388 pp Net, $3 .50 Wireless Telephones and How they Work. Illustrated. 16mo., cloth, 75 pp. $1.00 EWING, J. A. Magnetic Induction in Iron and other Metals. Third Edition, revised. Illustrated. 8vo.' cloth, 393 pp Net, $4.00 FISHER, H. K. C., and DARBY, W. C. Students' Guide to Submarine Cable Test- ing. Third Edition, enlarged. Illus. 8vo., cloth, 326 pp Net, $3.50 FLEMING, J. A. The Alternate-Current Transformer in Theory and Practice. Vol. I.: The Induction of Electric Currents. Fifth Issue. Illustrated. 8vo., cloth, 641 pp Net, $5 .00 Vol.11.: The Utilization of Induced Currents. Third Issue. Illustrated. 8vo., cloth, 587 pp Net, $5.00 Propagation of Electric Currents in Telephone and Telegraph Circuits. Illustrated. 12mo., cloth In Press Handbook for the Electrical Laboratory and Testing Room. Two volumes. Illustrated. 8vo., cloth, 1160 pp. Each vol Net, $5.00 FOSTER, H. A. With the Collaboration of Eminent Specialists. Electrical Engi- neers' -Pocket Book. A handbook of useful data for Electricians and Electrical Engineers. With innumerable Tables, Diagrams, and Figures. The most complete book of its kind ever published, treating of the latest and best Practise in Electrical Engineering. Sixth Edition, completely revised and enlarged. Fully Illustrated. Pocket Size. Leather. Thumb Indexed. 1636 pp $5.00 FOWLE, F. F. The Protection of Railroads from Overhead Nransmission Line Crossings. 35 illustrations. 12mo., cloth, 76 rp Net, $1.50 FRENDEMACHER, P. W. Electrical Mining Installations. 16mo., cloth. (In- stallation Manuals Series) In Press GANT, L W. Elements of Electric Traction for Motormen and Others. Illustrated with Diagrams. 8vo., cloth. 217 pp Net, $2.50 GERHARDI, C. H. W. Electicity Meters; their Construction and Management. A practical manual for central station engineers, distribution engineers and students. Illustrated. 8vo., cloth, 337 pp Net, $4.00 LIST OF WORKS ON ELECTRICAL SCIENCE. 5 GEAR, H. B. and WILLIAMS, P. F. Electric Central Station Distribution Systems. Their Design and Construction. 139 illustration. 12mo., cloth. 352 pp Net, $3.00 GORE, GEORGE. The Art of Electrolytic Separation of Metals (Theoretical and Practical). Illustrated. 8vo., cloth, 295 pp Net, $3 . 50 GRAY, J. Electrical Influence Machines: Their Historical Development and Modern Forms. With Instructions for making them. Second Edition, revised and enlarged. With 105 Figures and Diagrams. 12mo., cloth, 296 pp $2 . 00 GROTH, L. A. Welding and Cutting Metals by Aid of Gases or Electricity. 124 illustrations. 8vo., cloth, 280 pp Net, $3.00 HAMMER, W. J. Radium, and Other Radio Active Substances; Polonium, Actin- ium, and Thorium. With a consideration of Phosphorescent and Fluo- rescent Substances, the properties and applications of Selenium, and the treatment of disease by the Ultra-Violet Light. With Engravings and Plates. 8vo., cloth, 72 pp $1 .00 HARRISON, N. Electric Wiring Diagrams and Switchboards. Illustrated. 12mo., cloth, 272 pp $1 .50 HASKINS, C. H. The Galvanometer and its Uses. A Manual for Electricians and Students. Fifth Edition, revised. Illus. 16mo., morocco, 75 pp. . .$1 .50 HAY, ALFRED. Principles of Alternate-Current Working. Second Edition. Illustrated. 12mo., cloth, 390 pp $2.00 Alternating Currents; their theory, generation, and transformation. Second Edition. 1 178 Illustrations. 8vo., cloth, 319 pp Net, $2 . 50 An Introductory Course of Continuous-Current Engineering. Illustrated. 8vo., cloth, 327 pp Net, $2.50 HEAVISIDE, 0. Electromagnetic Theory. Two Volumes with Many Diagrams. 8vo., cloth, 1006 pp. Each Vol Net, $5.00 HEDGES, K. Modern Lightning Conductors. An illustrated Supplement to the Report of the Research Committee of 1905, with notes as to methods of protection and specifications. Illustrated. 8vo., cloth, 119 pp. .Net, $3.00 HOBART, H. M. Heavy Electrical Engineering. Illustrated. 8vo., cloth, 338 pp Net, $4.50 Electricity. A text-book designed in particular for engineering students. 115 illustrations. 43 tables. 8vo., cloth, 266 pp., Net, $2.00 Design of Static Transformers. 12mo In Press HOBBS, W. R. P. The Arithmetic of Electrical Measurements. Withn umerous examples, fully worked. Twelfth Edition. 12mo., cloth, 126 pp.. 50 cents HOMANS, J. E. A B C of the Telephone. With 269 Illustrations. 12mo., cloth, 352 pp $1 .00 HOPKINS, N. M. Experimental Electrochemistry, Theoretically and Practically Treated. Profusely illustrated with 130 new drawings, diagrams, and photographs, accompanied by a Bibliography. Illustrated. 8vo , cloth, 29S pp Net, $3 .00 6 LIST OF WORKS ON ELECTRICAL SCIENCE. HOUSTON, EDWIN J. A Dictionary of Electrical Words, Terms, and Phrases. Fourth Edition, rewritten and greatly enlarged. 582 Illustrations. 4to.. cloth Net, $7.00 A Pocket Dictionary of Electrical Words, Terms, and Phrases. 12mo., cloth, 950 pp .Net, $2.50 HUTCHINSON, R. W., Jr. Long-Distance Electric Power Transmission: Being a Treatise on the Hydro-Electric Generation of Energy; Its Transformation, Transmission, and Distribution. Second Edition. Illustrated. 12mo., cloth, 350 pp Net, $3 .00 and IHLSENG, M. C. Electricity in Mining. Being a theoretical and prac- tical treatise on the construction, operation, and maintenance of electrical mining machinery. Illustrated. 12mo., cloth In Press INCANDESCENT ELECTRIC LIGHTING. A Practical Description of the Edison System, by H. Latimer. To which is added: The Design and Operation of Incandescent Stations, by C. J. Field; A Description of the Edison Electro- lyte Meter, by A. E. Kennelly; and a Paper on the Maximum Efficiency of Incandescent Lamps, by T. W. Ho well. Fifth Edition. Illustrated. 16mo., cloth, 140 pp. (No. 57 Van Nostrand's Science Series.). . . .50 cents INDUCTION COILS: How Made and How Used. Eleventh Edition. Illustrated. 16mo., cloth, 123 pp. (No. 53 Van Nostrand's Science Series.). . .50 cents JEHL, FRANCIS. The Manufacture of Carbons for Electric Lighting and Other Purposes. Illustrated. 8vo., cloth, 232 pp Net, $4 . 00 JOHNSON, J. H. Arc Lamps and Accessory Apparatus. Illustrated. 16mo., cloth. (Installation Manuals Series) In Press JONES, HARRY C. The Electrical Nature of Matter and Radioactivity. Second Edition, completely revised. 12mo., cloth, 212 pp $2 .00 KAPP, GISBERT. Electric Transmission of Energy and its Transformation, Subdivision, and Distribution. A Practical Handbook. Fourth Edition, thoroughly revised. Illustrated. 12mo., cloth, 445 pp $3.50 KAPP, GISBERT. Alternate-Current Machinery. Illustrated. 16mo., cloth, 190 pp. (No. 96 Van Nostrand's Science Series.) 50 cents Dynamos, Alternators, and Transformers. Illustrated. 8vo., cloth, 507 pp $4.00 KELSEY, W. R. Continuous-Current Dynamos and Motors, and their Control; being a series of articles reprinted from the " Practical Engineer," and com- pleted by W. R. Kelsey, B.Sc. With Tables, Figures, and Diagrams. 8vo., cloth, 439 pp $2.50 KEMPE, H. R. A Handbook of Electrical Testing. Seventh Edition, revised and enlarged. 285 Illustrations. 8vo., cloth, 706 pp. . . .Net, $6.00 KENNEDY, R. Modern Engines and Power Generators. Illustrated. 4to., Electrical Installations of Electric Light, Power, and Traction Machinery. Illustrated. 8 vo. , cloth, 5 vols. The Set, $15. 00 Each, $3 . 50 LIST OF WORKS ON ELECTRICAL SCIENCE. 7 KENNELLY, A. E. Theoretical Elements of Electro-Dynamic Machinery. Vol I. Illustrated. Svo., cloth, 90 pp $1 .50 KERSHAW, J. B. C. The Electric Furnace in Iron and Steel Production. Illus- trated. 8vo., cloth, 74 pp Net, $1 . 50 Electrometallurgy. Illustrated. 8vo., cloth, 303 pp. (Van Nostrand's West- minster Series.) Net, $2 . 00 KINZBRUNNER, C. Continuous-Current Armatures; their Winding and Con- struction. 79 Illustrations. 8vo., cloth, 80 pp Net, $1 .50 Alternate-Current Windings; their Theory and Construction. 89 Illustrations. 8vo., cloth, 80 pp Net, $1 .50 KOESTER, F. Hydroelectric Developments and Engineering. A practical and theoretical treatise on the development, design, construction, equipment and operation of hydroelectric transmission plants. 500 illustrations. 4to., cloth, 475 pp Net, $5.00 Steam-Electric Power Plants. A practical treatise on the design of central light and power stations and their economical construction and operation. Fully Illustrated. 4to., cloth, 455 pp Net, $5 . 00 LARNER, E. T. The Principles of Alternating Currents for Students of Electrical Engineering. Illustrated with Diagrams. 12mo., cloth, 144 pp. Net, $1.50 LEMSTROM, S. Electricity in Agriculture and Horticulture. Illustrated. 8vo., cloth Net, $1 .50 LIVERMORE, V. P., and WILLIAMS, J. How to Become a Competent Motorman: Being a practical treatise on the proper method of operating a street-railway motor-car; also giving details how to overcome certain defects. Second Edition. Illustrated. 16mo., cloth, 247 pp Net, $1 .00 LOCKWOOD, T. D. Electricity, Magnetism, and Electro-Telegraphy. A Prac- tical Guide and Handbook of General Information for Electrical Students, Operators, and Inspectors. Fourth Edition. Illustrated. 8vo., cloth, 374 pp $2 . 50 LODGE, OLIVER J. Signalling Across Space Without Wires: Being a description of the work of Hertz and his successors. Fourth Edition. Illustrated. 8vo., cloth, 156 pp Net, $2 .00 LORING, A. E. A Handbook of the Electro-Magnetic Telegraph. Fourth Edition, revised. Illustrated. 16mo., cloth, 116 pp. (No. 39 Van Nostrand's Science Series.) 50 cents LUPTON, A. PARR, G. D. A., and PERKIN, H. Electricity Applied to Mining. Second Edition. With Tables, Diagrams, and Folding Plates. 8vo., cloth, 320 pp Net, $4.50 MAILLOUX, C. O. Electric Traction Machinery. Illustrated. 8vo., cloth. In Press 8 LIST OF WORKS ON ELECTRICAL SCIENCE. MANSFIELD, A. N. Electromagnets: Their Design and Construction. Second Edition. Illustrated. 16mo., cloth, 155 pp. (Van Nostrand's Science Series No. 64) 50 cents MASSIE, W. W., and UNDERBILL, C. R. Wireless Telegraphy and Telephony Popularly Explained. Illustrated. 12mo., cloth, 82 pp Net, $1 .00 MAURICE, W. Electrical Blasting Apparatus and Explosives, with special reference to colliery practice. Illustrated. 8vo., cloth, 167 pp. .Net, $3.50 MORECROFT, J. H. and HEHRE, F. W. A Short Course in Testing of Electrical Machinery. Illustrated. 8vo., cloth In Press MORGAN, ALFRED P. Wireless Telegraph Construction for Amateurs. 153 illus- trations. 12mo., cloth, 220 pp Net, $1.50 NIPHER, FRANCIS E. Theory of Magnetic Measurements. With an Appendix on the Method of Least Squares. Illustrated. 12mo., cloth, 94 pp. $1 .00 NOLL, AUGUSTUS. How to Wire Buildings. A Manual of the Art of Interior Wiring. Fourth Edition. Illustrated. 12mo., cloth, 165 pp $1.50 OHM, G. S. The Galvanic Circuit Investigated Mathematically. Berlin, 1827. Translated by William Francis. With Preface and Notes by the Editor, Thos. D. Lockwood. Second Edition. Illustrated. 16mo., cloth, 269 pp. (No. 102 Van Nostrand's Science Series.) 50 cents OLSSON, ANDREW. Motor Control as used in Connection with Turret Turning and Gun Elevating. (The Ward Leonard System.) 13 illustrations. 12mo., paper, 27 pp. (U. S. Navy Electrical Series No. 1.) Net, .50 OUDIN, MAURICE A. Standard Polyphase Apparatus and Systems. Illustrated with many Photo-reproductions, Diagrams, and Tables. Fifth Edition, revised. 8vo., cloth, 369 pp Net, $3.00 PALAZ, A. Treatise on Industrial Photometry. Specially applied to Electric Lighting. Translated from the French by G. W. Patterson, Jr., Assistant Professor of Physics in the University of Michigan, and M. R. Patterson, B.A. Second Edition. Fully Illustrated. 8vo., cloth, 324 pp $4.00 PARR, G. D. A. Electrical Engineering Measuring Instruments for Commercial and Laboratory Purposes. W r ith 370 Diagrams and Engravings. 8vo., cloth, 328 pp Net, $3 .50 PARSHALL, H. F., and HOBART, H. M. Armature Windings of Electric Machines. Third Edition. With 140 full-page Plates, 65 Tables, and 165 pages of descriptive letter-press. 4to., cloth, 300 pp $7 .50 Electric Railway Engineering. With 437 Figures and Diagrams and many Tables. 4to., cloth, 475 pp Net, $10.00 Electric Machine Design. Being a revised and enlarged edition of "Electric Generators." 648 Illustrations. 4to., half morocco, 601 pp. ..Net, $12.50 LIST OF WORKS ON ELECTRICAL SCIENCE. 9 PERRINE, F. A. C. Conductors for Electrical Distribution : Their Manufacture and Materials, the Calculation of Circuits, Pole-Line Construction, Under- ground Working, and other Uses. Second Edition. Illustrated. 8vo., cloth, 287 pp Net, $3.50 POPE, F. L. Modern Practice of the Electric Telegraph. A Handbook for Elec- tricians and Operators. Seventeenth Edition. Illustrated. 8vo., cloth, 234 pp $1 .50 RAPHAEL, F. C. Localization of Faults in Electric Light Mains. Second Edition, revised. Illustrated. Svo., cloth, 205 pp Net, $3 . 00 RAYMOND, E. B. Alternating-Current Engineering, Practically Treated. Third Edition, revised. With many Figures and Diagrams. 8vo., cloth, 244 pp Net, $2.50 RICHARDSON, S. S. Magnetism and Electricity and the Principles of Electrical Measurement. 254 Illustrations. 12mo., cloth, 596 pp Net, $2 .00 ROBERTS, J. Laboratory Work in Electrical Engineering Preliminary Grade. A series of laboratory experiments for first and second-year students in electrical engineering. Illustrated with many Diagrams. 8vo., cloth, 218 pp Net, $2 .00 RUHMER, ERNST. Wireless Telephony in Theory and Practice. Translated from the German by James Erskine-Murray. Illustrated. 8vo., cloth, 224 pp Net, $3.50 RUSSELL, A. The Theory of Electric Cables and Networks. 71 Illustrations. 8vo., cloth, 275 pp Net, $3.00 SALOMONS, DAVID. Electric-Light Installations. A Practical Handbook. Illus- trated. 12mo., cloth. Vol. I.: Management of Accumulators. Ninth Edition. 178 pp $2.50 Vol. II.: Apparatus. Seventh Edition. 318 pp $2.25 Vol. III.: Application. Seventh Edition. 234 pp $1 .50 SEVER, G. F. Electrical Engineering Experiments and Tests on Direct-Current Machinery. Second Edition, enlarged. With Diagrams and Figures. 8vo., pamphlet, 75 pp Net, $1 . 00 and TOWNSEND, F. Laboratory and Factory Tests in Electrical Engineering. Second Edition. Illustrated. 8vo., cloth, 269 pp Net, $2 . 50 SEWALL, C. H. Wireless Telegraphy. With Diagrams and Figures. Second Edition, corrected. Illustrated . 8vo., cloth, 229 pp Net, $2 . 00 Lessons in Telegraphy. Illustrated. 12mo., cloth, 104 pp Net, $1 .00 T. Elements of Electrical Engineering. Third Edition, revised. Illustrated. 8vo., cloth, 444 pp $3.00 The Construction of Dynamos (Alternating and Direct Current). A Text- book for students, engineering contractors, and electricians-in-charge. Illustrated. 8vo., cloth, 316 pp $3 .00 10 LIST OF WORKS ON ELECTRICAL SCIENCE. SHELDON, S., and HAUSMANN, E. Dynamo-Electric Machinery : Its Construction, Design, and Operation. Vol. I.: Direct- Current Machines. Eighth Edition, completely rewritten. Illustrated. 12mo., cloth, 281 pp Net, $2 . 50 Vol. II.: Alternating-Current Machines: Eighth Edition, rewritten. 12mo., cloth, 353 pp Net, $2.50 Electric Traction and Transmission Engineering. 127 illustration. 12mo., cloth. 317 pp Net, $2 .50 SLOANE, T. O'CONOR. Standard Electrical Dictionary. 300 Illustrations. 12mo., cloth, 682 pp $3.00 Elementary Electrical Calculations. A Manual of Simple Engineering Mathematics, covering the whole field of Direct Current Calculations, the basis of Alternating Current Mathematics, Networks, and typical cases of Circuits, with Appendices on special subject. 8vo., cloth. Illustrated. 304 pp Net, $2 .00 SNELL, ALBION T. Electric Motive Power. The Transmission and Distribution of Electric Power by Continuous and Alternating Currents. With a Section on the Applications of Electricity to Mining Work. Second Edition. Illustrated. 8vo., cloth, 411 pp Net, $4.00 SODDY, F. Radio-Activity ; an Elementary Treatise from the Standpoint of the Disintegration Theory. Fully Illustrated. 8vo., cloth, 214 pp. .Net, $3.00 SOLOMON, MAURICE. Electric Lamps. Illustrated. 8vo., cloth. (Van Nos- trand's Westminster Series.) Net, $2 .00 STEWART, A. Modern Polyphase Machinery. Illustrated. 12mo., cloth, 296 pp Net, $2 .00 SWINBURNE, JAS., and WORDINGHAM, C. H. The Measurement of Electric Currents. Electrical Measuring Instruments. Meters for Electrical Energy. Edited, with Preface, by T. Commerford Martin. Folding Plate and Numer- ous Illustrations. 16mo., cloth, 241 pp. (No. 109 Van Nostrand's Science Series.) 50 cents SWOOPE, C. WALTON. Lessons in Practical Electricity: Principles, Experi- ments, and Arithmetical Problems. An Elementary Text-book. Eleventh Edition, enlarged with a chapter on alternating currents. 404 illustrations. 12mo., cloth, 507 pp Net, $2 . 00 THIESS, J. B. and JOY, G. A. Toll Telephone Practice. 268 illustrations. 8vo. cloth, about 400 pp In Press THOM, C., and JONES, W. H. Telegraphic Connections, embracing recent methods in Quadruplex Telegraphy. 20 Colored Plates. 8vo., cloth, 59 pp. .$1.50 THOMPSON, S. P., Prof. Dynamo-Electric Machinery. With an Introduction and Notes by Frank L. Pope and H. R. Butler. Fully Illustrated. 16mo., cloth, 214 pp. (No. 66 Van Nostrand's Science Series.) 50 cents Recent Progress in Dynamo-Electric Machines. Being a Supplement to "Dynamo-Electric Machinery." Illustrated. 16mo., cloth, 113 pp. (No. 75 Van Nostrand's Science Series.) 50 cents LIST OF WORKS ON ELECTRICAL SCIENCE. 11 TOWNSEND, FITZHUGH. Alternating Current Engineering. Illustrated. 8vo., paper, 32 pp Net, 75 cents UNDERBILL, C. R. Solenoids, Electromagnets and Electromagnetic Windings. 218 Illustrations. 12mo., cloth, 345 pp Net, $2 .00 URQUHART, J. W. Dynamo Construction. A Practical Handbook for the use of Engineer Constructors and Electricians in Charge. Illustrated. 12mo., cloth $3.00 Electric Ship-Lighting. A Handbook on the Practical Fitting and Running of Ship's Electrical Plant, for the use of Ship Owners and Builders, Marine .Electricians, and Sea-going Engineers in Charge. 88 Illustrations. 12mo., cloth, 303 pp $3.00 Electric-Light Fitting. A Handbook for Working Electrical Engineers, em- bodying Practical Notes on Installation Management. Second Edition, with numerous Illustrations. 12mo., cloth $2.00 Electroplating. Fifth Edition. Illustrated. 12mo., cloth, 230 pp $2.00 Electrotyping. Illustrated. 12mo., cloth, 228 pp $2.00 WADE, E. J. Secondary Batteries: Their Theory, Construction, and Use. Second Edition, corrected. 265 Illustrations. 8vo., cloth, 501 pp Net, $4.00 WADS WORTH, C. Electric Battery Ignition. 20 Illustrations. 16mo. paper. In Press WALKER, FREDERICK. Practical Dynamo-Building for Amateurs. How to Wind for any Output. Third Edition. Illustrated. 16mo., cloth, 104 pp. (No. 98 Van Nostrand's Science Series.) 50 cents SYDNEY F. Electricity in Homes and Workshops. A Practical Treatise on Auxiliary Electrical Apparatus. Fourth Edition. Illustrated. 12mo., cloth, 358 pp $2 .00 Electricity in Mining. Illustrated. 8vo., cloth, 385 pp $3.50 WALLING, B. T., Lieut.-Com. U.S.N., and MARTIN, JULIUS. Electrical Installa- tions of the United States Navy. With many Diagrams and Engravings. 8vo., cloth, 648 pp $6.00 WATT, ALEXANDER. Electroplating and Refining of Metals. New Edition, rewritten by Arnold Philip. Illustrated. 8vo., cloth, 704 pp. .Net, $4.50 Electro -metallurgy. Fifteenth Edition. Illustrated. 12mo., cloth, 225 pp. .$1 .00 WEBB, H. L. A Practical Guide to the Testing of Insulated Wires and Cables. Fifth Edition. Illustrated. 12mo., cloth., 118 pp $1 . 00 WEEKS, R. W. The Design of Alternate-Current Transformer. New Edition in Press 12 LIST OF WORKS ON ELECTRICAL SCIENCE. WEYMOUTH, F. MARTEN. Drum Armatures and Commutators. (Theory and Practice.) A complete treatise on the theory and construction of drum- winding, and of commutators for closed-coil armatures, together with a full resume of some of the principal points involved in their design, and an exposition of armature reactions and sparking. Illustrated. 8vo., cloth, 295 pp Net, $3.00 WILKINSON, H. D. Submarine Cable-Laying, Repairing, and Testing. New Edition. Illustrated. 8vo., cloth In Press YOUNG, J. ELTON. Electrical Testing for Telegraph Engineers. Illustrated. 8vo., cloth, 264 pp Net, $4.00 ZEIDLER, J. and LUSTGARTEN, J. Electric Arc Lamps. Their principles, con- struction and working. 160 illustrations. 8vo., cloth, 200 pp Net, $2.00 96=page Catalog of Books on Electricity, classified by subjects, will be furnished gratis, postage prepaid, on application. RETURN TO the circulation desk of any University of California Library or to the NORTHERN REGIONAL LIBRARY FACILITY Bldg. 400, Richmond Field Station University of California Richmond, CA 94804-4698 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS 2-month loans may be renewed by callina (510)642-6753 1-year loans may be recharged by brinqinq books to NRLF Renewals and recharges may be made 4 days prior to due date DUE AS STAMPED BELOW DD20 6M 9-03 749202 ? UNIVERSITY OF CALIFORNIA LIBRARY