HflMffimmfifflaa LIBRARY OF THE UNIVERSITY OF CALIFORNIA. Class DOMESTIC SANITARY ENGINEERING AND PLUMBING DOMESTIC SANITARY ENGINEERING AND PLUMBING DEALING WITH DOMESTIC WATER SUPPLIES, PUMP & HYDRAULIC RAM WORK, HYDRAULICS, SANITARY WORK, HEATING BY LOW PRESSURE, HOT WATER, & EXTERNAL PLUMBING WORK BY F. W. RAYNES, R.P. MEDALLIST, CITY AND GUILDS OF LONDON INSTITUTE HEAD OF THE PLUMBING DEPARTMENT, AND ASSISTANT LECTURER IN THE DEPARTMENT OF CIVIL ENGINEERING THE GLASGOW AND WEST OF SCOTLAND TECHNICAL COLLEGE, GLASGOW WITH 277 ILLUSTRATIONS LONGMANS; GREEN, AND co. 39 PATERNOSTER ROW, LONDON NEW YORK, BOMBAY, AND CALCUTTA 1909 All rights reserved o PREFACE IN order to cover the subject without making the book unwieldy, and expensive to procure, it has been necessary to omit a great deal of elementary and general matter, and instead of devoting space to the Municipal side of Sanitary Engineering the scope of the work has been limited to the title of the book. Many formulae have been introduced as an aid in the design of work, and in most cases these have been given in as simple a form as possible consistent with accuracy, whilst numerous examples have been worked to show their application. Although the book will be found valuable for Students of Domestic Sanitary Engineering and Plumbing for Examination purposes, the writer hopes that it will have a still greater value for those who are entrusted with the design, the supervision, and the execution of this branch of engineering work. Much time has been entailed in the preparation of suitable drawings, and where a catalogue illustration has been used, it is not intended to convey that a certain manufacturer's goods are superior to those of another firm, but to illustrate some principle or point under discussion. For valuable aid in the preparation of the illustrations the writer's thanks are due to his drawing assistant, Mr. John Burnside. F. W. RAYNES. THE GLASGOW AND WEST OF SCOTLAND TECHNICAL COLLEGE, GLASGOW. 195099 CONTENTS CHAPTER I MATERIALS AND THEIR PROPERTIES : MODE OF MANUFACTURE riai Metals Physical and chemical properties Lead ores Reduction of ores Lead compounds Lead pipes Tinned lead pipes Tin-lined lead pipe Lead traps and bends Cast sheet lead Milled lead Iron ores Wrought iron Steel Malleable cast iron Iron pipes Coatings for pipes Copper Coatings for copper Sheet copper Copper tubes Tin Block tin tube Zinc Alloys Properties of alloys Composition Sanitary pottery Fireclay Earthenware drain pipes Concrete tubes ..... 1-29 CHAPTER II ROOF WORK Metal Coverings Lead flats or platforms Solid rolls Roll ends Soldered dots Hollow rolls Intersecting rolls Lead gutters Drips Drip-boxes or cesspools Soakers Cover flashings Gutter flash- ings Step flashings Dormers Glass roof and skylights Cornices Stone copings Hips and ridges Methods of securing leadwork Ornamental ridging Torus rolls Turret roofs Shape of bays Domes Finials Strengths of lead .... 30-80 CHAPTER III PIPE FIXING AND PIPE BENDING Methods of supporting pipes Wood grounds Plain tacks Ornamental lugs or tacks Ornamental fixings Flange supports Pipe hangers Special wall clips Roller fixings Bending lead pipes Springs Weights Dummies Working drawings Development of elbow pipes Bending copper pipes Bending machines . . 81-96 vii Vlll CONTENTS CHAPTER IV PIPE JOINTS PAGE Joints for lead pipes Preparation of joints Gauges Methods of support- ing joints Joints for tin-lined lead pipes Burnt joints Joints for copper pipes Iron pipe joints Lead wool Expansion joints Pack- ing rings Expansion bends Joints for w.c.'s Joints for drains Patent joints Elastic cement ..... 97-123 CHAPTER V SOLDERS, FLUXES, AND LEAD BURNING Soft solders Composition and fusing points Properties of solder Treat- ment of poisoned solder Hard or brazing solders Composition of hard solders Fluxes, and their use Lead burning Hydrogen generator Method of charging generator Tank for supply- ing atmospheric air Compressed gases in cylinders Cost of oxygen ........ 124-135 CHAPTER VI SANITARY FITTINGS AND ACCESSORIES Principles governing construction Water-closets Wash-out type Wash- down type Flush pipes Combination closets Valve closets, merits and defects Siphonic closets, types, their action and limitations Trough closets Siphonic latrines Ranges of closets Connec- tions of w.c.'s Flushing cisterns, types in use Waste water preventers, mechanical and pneumatic types Lavatories, their merits and defects Baths, treatment of overflows Sinks Wash- tubs Slop sinks Urinals, forms they take, their merits and defects . V . . . . .A 136-166 CHAPTER VII SOIL AND WASTE PIPES Materials Iron soil pipes Copper soil pipes Thickness of soil pipes Arrangement of soil pipes Sizes of soil pipes Sizes of antisiphon- age pipes Effect of arrangement of pipes on sizes of antisiphonage pipes Unsealing of traps, cause and prevention Waste pipes Sizes of waste pipes Arrangement of waste pipes Rust pockets Traps for waste pipes ...... 167-192 CONTENTS IX CHAPTER VIII DRAINAGE OF HOUSES AND OTHER BUILDINGS PAGE General design Merits and drawbacks of earthenware and of iron drains Foundations for drains Connections with drains Junctions and bends Chambers and openings for access Sizes of chambers Manhole covers Gully traps Disconnecting traps Grease traps Tidal traps Drainage of basements and sewage lifts Drainage plans Stable drainage Connections of drains with sewers Venti- lating and flushing of drains Automatic flush tanks Methods of laying drains Boning rods and sight rails Timbering trenches Drain testing Testing appliances Discharging capacity of drains Hydraulic mean depth Velocity formula . . . 193-242 CHAPTER IX DISPOSAL AND TREATMENT OF SEWAGE FROM MANSIONS AND HOUSES IN COUNTRY DISTRICTS Methods of treatment System of subsoil irrigation, sizes of tanks, amount of land required Sewage filters Difference between contact beds and percolating filters ..... 243-250 CHAPTER X WATER SUPPLY Water pollution Sources of supply Collecting area Special collecting surfaces Conditions affecting yield by a surface Rainfall Volume of water available Capacity of storage tanks Water consumpt Rain-water separators Sand filters Surface springs Deep springs Wells as a source of supply Boreholes Hardness of water Soften- ing water Water services, constant and intermittent supplies- Arrangement of service pipes Connections of service pipes with street mains Storage cisterns Cistern overflows and washouts Sizes and capacities of tanks Domestic filters Ball-cocks Screw- down and plug cocks Spring or semi-automatic taps Water- hammer in pipes, cause and remedy .... 251-307 CHAPTER XI APPLIANCES FOR RAISING WATER Lift pump Suction pipes Deep well pumps Lift and force pumps Air vessels Double barrelled pumps Formulae for lift pumps Formulae for lift and force pumps Geared pumps Formula for geared pumps Hydraulic rams Long and short drive pipes Air vessels for rams Duty of rams Formulae for rams Ram pumps and their action 308-339 X CONTENTS CHAPTER XII HYDROSTATICS AND HYDRAULICS PAGE Pressure due to head of water Total pressure Resistance to the flow of water Vena contracla Flow of water through orifices and short tubes Flow of water through long pipes Hydraulic gradient Sizes of pipes Head absorbed by friction Thickness and strength of pipes ........ 340-365 CHAPTER XIII DOMESTIC HOT WATER SUPPLY Movement of heat Circulation of water Tank system Cylinder system Details of systems Secondary circuits Cylinder-tank systems, merits and defects Sizes of tanks Range and independent boilers Duty of range boilers Formula} for range and dome top independent boilers Steam apparatus for heating water Properties of steam Boiling point Automatic steam supply valves Steam traps Heat transmitted by steam coils Indirect hot-water systems for preventing incrustation difficulties Forms of indirect heaters Collapse of cylinders, and its prevention Noises in boilers Boiler explosions Safety valves . . . . . . . 366-426 CHAPTER XIV LOW PRESSURE HOT- WATER HEATING APPARATUS Systems of piping Pitch of pipes Circulating head Sizes of pipes Heating surfaces Comparative value of heating surfaces Radiator valves Air valves Feed cisterns Calculations of heating surface Discharge of air through flues Heat to warm air Heat absorbed by wall and glass surfaces Heat emitted by pipe surfaces Drying rooms Boilers for low pressure heating Value of direct and indirect surfaces Boiler draught regulator Sizes of boilers Calorific value of fuels Size of chimneys ..... 427-466 APPENDIX Hydraulic memoranda Weight of water at different temperatures Weight of metals Weight of cast-iron pipes Wire and plate gauges ........ 462-466 CONTENTS xi WORKED EXAMPLES. HOUSE DRAINAGE NO. PAGE 1. Hydraulic mean depth of a 6-inch pipe when water is flowing J the depth of pipe . . . . . . . 238 2. Gradients which produce velocities of 3 feet per second in a 6-inch pipe when the latter is flowing and full .... 240 3. Discharging capacity of a 6 -inch drain when flowing f full, and when laid with a gradient of 1 in 50 . . . . . 242 WATER SUPPLY 4. Volume of rainfall available from a given surface . . .258 5. Collecting area required to yield a given volume of water . . 258 6. To determine width of a storage tank ..... 260 7. Depth of circular storage tank required when other particulars are given ........ 260 8. Diameter of tank required for given conditions . . . 260 9. Capacity of irregular shaped tank ..... 291 10. Depth of irregular shaped tank when other particulars are given . 291 APPLIANCES FOR RAISING WATER 11. Volume of water raised by a 3J-inch diameter lift pump in a given time .......'. 320 12. Diameter of lever pump to raise a given volume of water in a given time ......... 321 13. Force required to be exerted at the end of a lever to overcome a given load ........ . 322 14. Effort necessary to raise water by means of a 4-inch diameter lift pump ........ 323 15. To determine diameter of a lift and force pump when worked with limited power ....... 325 16. Effort necessary to raise water through a given height by means of a lift and force pump ....... 325 17. Volume of water raised by a lift and force pump . . . 325 18. Effort necessary to raise water through a given height by a wheel- handle lift and force pump ...... 328 19. To determine diameter of a geared wheel-handle pump for working with limited power ....... 328 20. Volume of water raised by a geared double barrelled pump . . 329 21. Volume of water raised by a hydraulic ram .... 336 22. Water supplied to ram in any given time .... 336 23. To determine efficiency of a hydraulic ram .... 336 Xll CONTENTS HYDROSTATICS AND HYDRAULICS NO. PAGE 24. Total pressure, and the average pressure per sq. inch on a vertical surface of a cylindrical tank . . . . . . 342 25. Total pressure acting upon a circular stopper .... 343 26. Total pressure acting upon one side of a rectangular cistern . . 343 27. Head of a column of water equivalent to a given load on a safety valve ........ 343 28. Discharge of water by a short tube ..... 346 29. Head absorbed by friction in discharging a given volume of water through a short tube . . . . . .347 30. Discharging capacity of a cast-iron pipe .... 349 31. Diameter of pipe necessary to discharge a given volume with given head . . .-' . . . . . .350 32. Sizes of pipes for a compound main ..... 352 33. Volume discharged by a tap when subjected to a given head, and where length of pipe is comparatively short . '.,' . 355 34. Discharging capacity of a given arrangement and size of pipe . . 357 35. Diameter of pipe necessary for a given discharge . . . 358 36. Sizes of main draw-off pipe and branches, when the latter are dis- charging simultaneously given volumes of water . . . 360 37. Maximum safe working pressure for a lead pipe . . . 364 38. Thickness of lead pipe . . . . . . .365 39. Thickness of a cast-iron pipe required when subjected to a given water pressure .... .... -. . 365 DOMESTIC HOT WATER SUPPLY 40. Volume of water heated by a given type of range boiler . . 395 41. Volume of water heated by a given type of range boiler . . 395 42. Volume of water heated by a given type of range boiler . . 395 43. Time required to heat a given volume of water by a range boiler . 396 44. Heating capacity of a dome top independent boiler . . . 397 45. Size of independent boiler necessary to do a specific amount of work . 397 46. Volume of hot water essential to produce a larger volume at a given lower temperature and capacity of hot-water tank . .- ., . 399 47. Capacity of a cylindrical tank . . ., . . 400 48. Height of a cylindrical tank when capacity and diameter are given . 400 49. Diameter of a cylindrical tank when its capacity and height are given . 400 50. Length of steam heated coil essential to raise water through a given temperature . . . . . . . .411 51. Gallons of water raised by a steam heater in a given time . .412 52. Time required for a steam heater to do a specific amount of work . 412 Low PRESSURE HOT- WATER HEATIJSLG APRARATUS 53. Discharging capacity of an air duct per hour .... 444 54. Heating surface necessary to warm a room to a given temperature . 448 CONTENTS Xlll NO. PAGE 55. Heating surface necessary to warm a room to a given temperature . 449 56. Heating surface for drying room ..... 450 57. Size of cast-iron sectional boiler for a given amount of work . . 460 58. Heating capacity of a given size and type of boiler . . . 460 59. Size of chimney required for heating boiler .... 461 TABLES I. Strengths of lead for roof work . . . . .80 II. Capacity of tanks and size of siphons for flushing drains . . 223 III. Data for obtaining hydraulic mean depth and the sectional area of flow in circular drains, when water is running at different depths . . " . . . . .239 IV. Values of c for Kutter's formula ..... 240 V. Approximate gradients for drains ..... 242 VI. Efficiency values for hydraulic rams .... 336 VII. Sizes of drive and delivery pipes for hydraulic rams . . 337 VIII. Coefficients for hydraulic formula ..... 348 IX. 5th power of pipe diameters ..... 348 X. Coefficients for formula in connection with short pipes and fittings 354 XL Average tensile strength of metals .... 363 XII. Constants or coefficients for different forms of range boilers . 394 XIII. Properties of steam ...... 403 XIV. Heat transmitted by short steam-heated coils . . . 411 XV. Boiling point of water when subjected to different pressure heads 416 XVI. Approximate heating surface supplied by various sizes of pipes (low pressure hot-water heating) . 435 XVII. Heat lost through brick walls in British units . . . 445 XVIII. Heat lost through stone walls in British units . . . 446 XIX. Heat lost by glass and other surfaces in British units . . 446 XX. Heat transmitted by horizontal pipe surfaces . . . 447 XXL Approximate space warmed by a square foot of heating surface . 451 XXII. Calorific value of fuels . . . . . .458 XXIII. Coefficients for various forms of hot water heaters and different rates of firing . . . . . . . 459 DOMESTIC SANITARY ENGINEERING AND PLUMBING CHAPTEE I MATERIALS AND THEIR PROPERTIES: MODE OF MANUFACTURE A GOOD knowledge of the properties of materials is essential for executing work of importance where durability and sound workmanship are required. From time to time much work has resulted in failure owing to ignorance of the properties of the materials employed. Metals. The properties of metals are both physical and chemical. Physical properties are malleability, fusibility, ductility, tenacity, flow of metals, lustre, elasticity, electric conductivity, and heat conductivity. Malleability is that property which permits of a metal being rolled into sheets, and worked into various shapes, without the metal being broken or torn. Fusibility, as the term indicates, is the conversion of a solid into the liquid state by the application of sufficient heat. All metals, with the exception of mercury, are solid at normal temperatures. Ductility is the property which permits a metal to be drawn into the form of wire ; the thinner the wire can be drawn the more ductile is the metal. Density is the relative weight of a body when compared with an equal volume of water. All the metals used in Plumbers' Work are heavier than water. Lead is the heaviest, having a density or specific gravity of 1 1 '4 ; whilst aluminium 2 DOMESTIC SANITARY ENGINEERING AND PLUMBING is the lightest, having a density of 2 '6 7. In other words, volume for volume lead is 11*4 times heavier than water, and aluminium 2*6*7 times heavier. Tenacity indicates that property which resists the particles or molecules of which a body is composed from being torn asunder. The tenacity of metals is greatly re- duced when they are subjected to high temperatures. Flow of Metals. Although not visible to the naked eye, the molecules composing metals are, to a more or less extent, in a state of motion due to variations of temperature. Flow of metals is more pronounced when they are subjected to pressure ; thus a piece of sheet lead can be thinned in one place and thickened in another, owing to the manner in which the molecules can be displaced by the application of the bossing stick or mallet. In other words, the metal is said to flow. It is on account of this property that a block of lead can be rolled into thin sheets, or forced through a die to take the form of a pipe. Lustre. All metals when clean or when polished reflect light, and therefore have a high lustre. Elasticity is that property which permits of a metal regaining after distortion its original shape. Thus if a steel bar is only bent or elongated by the application of force, the bar will regain its normal state when the force is removed, provided the elastic limit of the metal has not been exceeded. When the elastic limit has been passed a permanent set is made. Conductivity. Metals are good conductors of both heat and electricity, although some are much better than others. Copper ranks as the best conductor of the common metals for either heat or electricity, whilst lead is about the worst. Chemical Properties. When in the molten state the common metals have a great affinity for oxygen, with the result that oxides of these metals are rapidly produced. Dry air does not affect metals at normal temperatures to any considerable extent, but when the air is moist, and carbonic acid is present, the surfaces of metals are readily attacked and covered with a film of oxide. Acids, such as Nitric, Hydrochloric, and Sulphuric, tend MATERIALS AND THEIR PROPERTIES 3 to dissolve the common metals to a more or less extent. Sulphuric acid when cold only slowly affects lead, owing to sulphate of lead forming on the surfaces and acting as a protective covering for the metal beneath. Lead and its Ores. The ores which are capable of yielding considerable quantities of metallic lead are Galena and Cerusite. The former is a dark-coloured, metallic looking substance, and is the most widely distributed. Cerusite occurs as a carbonate of lead in the form of a white or dark earthy substance, and intermixed with clay and limestone, etc. ; galena is also often present in its admixture. Galena (PbS) is a compound consisting principally of lead and sulphur, and when this ore is placed in the smelting furnace it contains from 70 to 85 per cent, of lead. Cerusite (PbC0 3 ), or white lead ore, is a compound consisting chiefly of lead, oxygen, and carbonic acid gas. The pure ore will yield as much as 77J per cent, of lead, whilst the crude ore contains about 30 per cent, of lead. Although galena is very widely distributed, it does not occur in many places in sufficient quantity to pay for working it. The principal British localities where lead is obtained are North Wales, Derbyshire, Cornwall, Northum- berland, Lanarkshire, and Laxey, Isle of Man. Large quantities of lead are imported to Great Britain from Spain. Cerusite is found in large deposits in Nevada and in Colorado, U.S., and it also occurs in Scotland, principally at the Lead Hills, Lanarkshire. The lead of commerce, however, is obtained usually from galena. Reduction of Ores. Lead is extracted from the ore in smelting furnaces, which differ in operation and construction in different localities, and according to the nature of the impurities in the ore. The impurities lead contains are silver, iron, zinc, antimony, and copper, etc., and these render the lead very hard. All the impurities, with the exception of silver, can be removed by raising the lead to a high temperature in the presence of air, when oxidation of the impurities takes place at the surface and they can be taken off in the form of a scum. To extract the silver, the lead is subjected to either the Pattinson or Parkes 4 DOMESTIC SANITARY ENGINEERING AND PLUMBING process. The object aimed at in each process, is to con- centrate the silver into a small quantity of lead, when the rich argentiferous lead is afterwards finally freed from the silver by cupellation. A little impurity in lead, however, is not always objectionable. In the case of sheet lead for roof work, and where it requires to be worked into various shapes, the metal should be as free from impurity as possible. Soil pipes which require bends made in them should also be of soft lead. Lead water pipes, on the other hand, and which deliver water under more or less considerable pressure, are better made of lead which contains a little impurity. In this case the harder pipes better resist the erosive action of water when flowing through them at high velocities. Thick sheet lead and plate lead, which is used for lining vitriol tanks, is said to better resist the action of the acid when it contains a small percentage of antimony. Physical Properties of Lead. In colour lead is bluish- gray, and when newly cut has a bright metallic lustre, but is rapidly oxidised in the presence of moist air. Lead is not very ductile, so it cannot be drawn into very thin wire ; it has a low tenacity, and is useless where strength or toughness is required. The specific gravity of lead is about 11*4, and a cubic foot weighs approximately 712 Ib. It is not a perfectly elastic metal, and the rate of expansion slightly exceeds that of contraction. The latter property is noticeable in many lead-lined sinks, where the metal has formed itself into ridges or buckles owing to its size being intermittently increased. Many lead pipes are either distorted or fractured by increase of length due to alternate heating and cooling. Lead is a poor conductor of heat and electricity, but it is very malleable and soft, and can be readily worked into various shapes without the application of heat. Chemical Properties of Lead. To a more or less extent lead is acted upon by all acids, and also by moist air. After lead has been newly laid in gutters, or fixed in connection with other roof work, it is generally found that on the follow- ing day the surfaces of the lead are covered with a thin film of basic carbonate of lead ; this film is due to the moisture and carbonic acid gas in the atmosphere acting MATERIALS AND THEIR PROPERTIES 5 upon the lead. The action readily takes place after sunset when dew is deposited, or in the daytime when the atmo- sphere is in a saturated state. Sheet lead work always has a better appearance a day or two after it has been done, owing to the carbonate film producing a dull surface and obscuring tool marks. Water, when pure, is said to have no action upon lead, but when it contains free oxygen the lead is attacked, forming an oxide which is soluble in water. If carbonic acid gas is also present in the water, the dissolved oxide is precipitated as basic carbonate, and the surfaces of the lead are again laid bare to the action of the water. Impurities in water act differently ; some tend to prevent water acting upon lead, whilst other impurities tend to accelerate the action. Sulphuric acid, when dilute, has no action upon lead ; strong solutions of the acid at ordinary temperatures act slowly upon it, but the action is accelerated by the concen- tration of the acid and with rise of temperature. Boiling sulphuric acid readily converts the lead into a sulphate with the evolution of sulphurous acid. Lead is readily dis- solved by dilute nitric acid, and it is also acted upon by hydrochloric acid. Lead Compounds. The principal lead compounds, so far as plumbers are directly concerned, are red lead and white lead ; the former is an oxide of lead, whilst the latter is a carbonate of lead. These two compounds are largely used by plumbers for jointing materials. Red Lead. (Pb 3 4 ) is made by exposing molten lead in a furnace to the action of the air ; as oxidation takes place the metallic surfaces are repeatedly renewed by pushing the oxide towards the back of the furnace, this operation being continued until apparently the whole of the metallic lead is converted into an oxide. It is then removed from the furnace to the grinding mill, where it is ground in water with heavy, revolving stone rollers. The grinding process, besides reducing the oxide into a fine state of division, separates the oxidised from any metallic lead which may be present. After leaving the grinding mill the oxidised lead is put into a furnace which is called the colouring oven, and which has 6 DOMESTIC SANITARY ENGINEERING AND PLUMBING a temperature of about 600 F. In this oven the oxide is exposed to the action of the air for about twenty-four hours, in order that it may take up a further amount of oxygen, which is absorbed at a gradually decreasing rate until oxida- tion is complete, and the oxide has assumed a bright red colour. The red lead is now removed from the colouring oven and reground in water, and afterwards dried. Litharge (PbO) is made precisely in the same manner as red lead, excepting that the furnaces are raised to a higher temperature. Litharge is yellow in colour. The difference in colour of red lead and litharge is due to the different amounts of oxygen which have entered into their compositions. White Lead (PbC0 3 ) is made in different ways, but the Dutch process is the one best known. In the Dutch process either specially cast grids of lead, or sheet lead loosely formed into coils, are placed in jars which contain acetic, acid. The lead is fixed clear of the acid, and the jars are arranged in rows and built up in stacks. At the bottom of the stack a thick layer of fermenting material, such as tan, is placed, and upon this the jars are arranged side by side, and surrounded with the same material. Over the mouths of the jars thin lead plates are fixed, and then another floor is formed by laying boards on the top of the jars. On this floor another layer of tan is placed, and other jars are arranged as above described. The jars are built up in tiers in this way until the stack is sufficiently high. When complete, the stack is left for about three months, during which time fermentation takes place, and the whole mass becomes thoroughly heated. The heat generated in the stack vaporises the acetic acid, and as the vapour combines with the carbonic acid gas given off by the decomposing tan, the metallic lead is attacked and converted into a basic carbonate of lead. In clue time the corroded lead is removed from the jars, and freed from any metallic lead by passing it between corrugated rollers. The grinding process is the next operation, and if the lead is required in the form of paste it is reground in linseed oil. Lead Pipes. For purposes of comparison lead pipes may be divided into four classes, viz : First, ordinary lead pipes ; MATERIALS AND THEIR PROPERTIES 7 second, lead pipes which have one or both surfaces tinned ; third, tin-lined lead pipes where the tin and lead are in contact ; and fourth, tin-lined lead pipes where the tin and lead are separated by a covering of asbestos or similar material. Lead pipes are made by the aid of the hydraulic pipe press, hand-made pipes being now a thing of the past. The manufacture of lead pipes is apparently a simple process, but great care is required when setting the die and core of a machine, in order that a pipe will be turned out as nearly true in section as possible. In Fig. 1 a lead pipe making machine is given. Its principal parts are the ram P, container C, the core or mandril M, and the die D. Molten lead is run into the container C until the latter is filled and, as a rule, holds just sufficient lead to make two bundles of half-inch pipe. The lead is allowed to solidify, but whilst still hot hydraulic pressure is brought to bear upon the ram P, and to raise the piston. From the rising container the lead only has one point of escape, and that is through the annular space between the mandril and the die. Through this space the lead issues in the form of a pipe. The die D, it will be observed, forms the external diameter of the pipe, whilst the mandril M makes the bore. It is often thought by people who have had no opportunity of seeing pipes produced, that the lead is forced from the container whilst in a molten state ; this, however, is not the case, as the lead would be simply squirted into the air. Considerable pressure is required to form a pipe, the intensity of the pressure varying with the size of pipe and the amount of impurity in the lead. Small pipes, other things being equal, require a greater hydraulic pressure than larger sizes to produce. When water pipes are made, they are wound round drums to form bundles as they issue from the machine. For making soil and waste pipes in straight lengths, a cord is passed over a pulley which is fixed high over the machine ; one end of the cord is attached to the pipe, and the other is kept taut by a man as the pipe is being drawn ; another person measures the pipe and cuts it off to the required length. 8 DOMESTIC SANITARY ENGINEERING AND PLUMBING Lead pipe making machines differ considerably in structural details, but their general mode of operation is practically the same in each case. By changing the die and mandril one machine can be used for making the various sizes and thick- nesses of pipes in general use. FIG. 1. Machine for making lead pipe. Tinned Lead Pipe is sometimes confused with tin-lined lead pipe, but in the former the surface of a pipe is only covered with a thin film of tin, whilst in the latter the tin lining may form a substantial part of the pipe. The tinning process is very simple, and is effected as the pipe issues from the machine. When the inner surface of a pipe requires to be MATERIALS AND THEIR PROPERTIES 9 tinned, all that is necessary is to pour a little molten tin inside the pipe ; the hot mandril and lead keep the tin in the molten state, and as the pipe is being formed its inner surface is covered with a film of the molten tin. When this class of pipe was first produced it was thought that the covering of tin would tend to prevent the corrosive action which some waters have upon lead. Instead, however, of the tinning preventing such action, it was soon discovered that such superficial treatment often tended to accelerate it. At their best, internally tinned lead pipes are little or no better than those which are not tinned ; the tin alloys to a more or less extent with the lead, but the interior surfaces of the pipes are not evenly covered. To tin the outer surface of a lead pipe, the die of the machine is frequently formed with a hollow, or pocket, into which molten tin is poured. Some- times the tin is melted in the upper part of the machine by heating it with gas, or by other means. In the hollow of the die block the molten tin surrounds the pipe, and as the latter passes through it the external surface of the pipe receives its film of tin. To remove any superfluous metal, and to give the pipe a smooth appearance, cotton waste or similar material is pressed against the pipe as it issues from the machine. Lead gas-pipe which has its external surface tinned is frequently called composition pipe, but the term " com- position " is rather misleading in this case. Tin-Lined Lead Pipe is also made by the ordinary pipe press, Fig. 1 ; but in this case a double process of charging the container C is involved. Instead of the container being fully charged with lead, an annular space next the mandril M is left, and is afterwards filled with molten tin. Thus whilst in the container the lead and tin really take the form of a very thick, tin-lined lead pipe. The tin is added after the lead, and when set hydraulic pressure is brought to bear upon the under side of the container, and as the latter rises the two metals are forced together through the die, and issue in the form of a compound pipe, with the tin and lead in the correct proportion. The thickness of the tin lining varies from about IT* inch to T V inch, according to the size and quality of the pipe. Tin-lined lead pipes have been largely used to minimise 10 DOMESTIC SANITARY ENGINEERING AND PLUMBING the risk of lead poisoning when portable waters dissolve the latter metal. Although these pipes are superior to lead ones for conveying waters which attack lead, still they are far from being satisfactory, as traces of the latter metal will generally be found in water which has been lying stagnant in them for a short time. It is quite possible for imperfections to occur during the manufacture of tin-lined lead pipe, and doubtless a certain amount of alloying takes place when the molten tin comes in contact with the lead during the charging of the machine. Another drawback generally arises through the un- satisfactory method of jointing these pipes, owing to the lead being laid bare by some parts of the tin lining being destroyed. It is therefore obvious that the use of tin-lined lead pipe is no guarantee that a soft and acid water which passes through it will not contain traces of lead. On the other hand, if a water has no effect upon lead, tin-lined lead pipes serve no very special purpose excepting that they are stronger than lead pipes. Insulated Tin-Lined Lead Pipes. Some time ago an attempt was made to overcome the defects of ordinary tin- lined lead pipe by producing a pipe in which the tin and lead are kept separate by an asbestos covering. It is not possible, in this case, for the tin and lead to alloy when making the pipe, as the tin and lead tubes are separately made. The tin lining is afterwards covered with the asbestos, and the lead pipe requires to be of sufficient size to receive them. The composite pipe is then formed by inserting the insulated lining into the lead casing, and by afterwards passing the whole through a special machine the three materials are compressed together. This form of tin-lined lead pipe overcomes some of the failings of the first form, but it is not free from defects, and trouble has been caused by the tin lining col- lapsing and the water passages becoming stopped. A great amount of care is also necessary in making the joints, in order to prevent water coming into contact with lead at these points. It also has the drawback of a high initial cost, and, like the ordinary form of tin-lined pipe, is more costly to fix than lead pipe, on account of the special fittings and extra time involved in making the joints. MATERIALS AND THEIR PROPERTIES 11 Where water is known to act upon lead, it is desirable not to use lead pipes of any form for distributing water which .is intended for dietetic purposes. The best form of pipe at present in use for the conveyance of such water is the tin- lined iron pipe with right- and left-hand screwed and socketed joints. Lead Traps and Bends. Lead traps, which are used at the present time, are the seamless, hydraulically drawn productions, and the cast types ; hand-made lead traps with soldered seams have had their day. The seamless traps possess the advantages of cheapness and smoothness, whilst those which are cast have the merit of being stronger than the former. Opinions still differ amongst many good plumbers with regard to the merits and demerits of the hand-made and machine-made trap. It is contended by some that the modern drawn seamless trap will not withstand the strain due to changes of temperature to the same extent as the hand-made trap witli soldered seams ; that the former has often failed in a comparatively short time, whilst hand-made traps fixed under similar conditions have been much more durable. This presents a problem worthy of consideration, because if a certain fitting fails in a com- paratively short time when compared with a similar one which is only made in a different way, the cause of such failure should not be difficult to ascertain. Very often when comparisons are made some of the most important factors are overlooked, and thus the conclusion arrived at may be only partly true. It is quite possible, however, for a hand- made trap to be much more durable than a drawn one under certain conditions, but the writer sees no reason why a seamless trap under ordinary circumstances should not be just as durable as a hand-made one, provided it is properly fixed, made of good lead, and of sufficient strength. Thickness for thickness, hand-made traps are stronger than those which have been drawn, owing to the soldered seams, which impart a fair amount of rigidity. Should a trap with soldered seams be fixed in connection with a length of light lead waste pipe, and the pipe arranged that movement due to expansion and contraction can freely take place, the trap under such conditions would have a long life, and especially 1 2 DOMESTIC SANITARY ENGINEERING AND PLUMBING if the waste water passed through it was not of a very high temperature. On the other hand, if the same trap had been fixed to a length of strong waste pipe, through which very hot water was discharged, and no provision were allowed for the pipe to expand, the life of the trap under these circum- stances would be comparatively a short one. In the latter case the whole of the strain which accompanies expansion would be concentrated upon the trap, and the latter would naturally be distorted by the gradual increasing length of pipe. If, now, we assume that drawn lead traps had been used for the cases already described, the results would be similar in each. Of course it is quite possible for special cases to occur where greater distortion can take place with drawn lead traps than with those with soldered seams, but where a strong- form of lead trap is desirable a cast one could be used. Lead is not, however, an ideal material for traps which receive alter- nately hot and cold discharges of water ; and where durability is an important factor, traps of hard metal, such as iron and brass, should be used. For general work the cheapness of drawn lead traps, and the many forms in which they are made, are important advantages. Their life may also be considerably increased if the waste pipes are arranged to prevent the pull and thrust which accompany contraction and expansion being concen- trated upon the traps. Many of the defective, old, seamed lead traps, which have been taken out from time to time, instead of being damaged by different rates of expansion and contraction between the two different materials of which they were constructed, were chiefly the result of corrosion owing to lack of ventilation. Drawn lead traps should not be thinner in substance than 6-lb. sheet lead, and a greater thickness as a rule does not proportionally increase the life of traps which receive alternate discharges of hot and of cold water. Much, of course, depends whether the metal is pure or not. For example, it is quite possible for a trap which is equal to 8-lb. sheet lead to be less durable than one whose thickness is only equal to 6-lb. lead, provided the lead of the former contained a higher percentage of impurity. MATERIALS AND THEIR PROPERTIES 13 The manufacture of drawn lead traps and bends is similar to that of ordinary lead pipe, but the machines differ in construction. In a trap making machine the lead container is arranged with a piston at each end, whilst the lead issues from the centre of the machine. After the container is charged with lead, hydraulic pressure is brought to bear upon the pistons, and by manipulating the pressure on each piston the issuing pipe can be curved or bent in the direction desired. When equal pressures are applied to the pistons the lead issues from the machine in the form of a straight pipe, and when differential pressure is acting on the pistons the pipe curves in the direction of the greater force. Sheet Lead. This may either be cast or milled, but the former is very seldom required at the present time, and it is very doubtful if \ per cent, of the present day plumbers will ever be called upon to cast and lay it. Cast sheet lead when required can be made on a suitable casting frame in the workshop, or other suitable place ; or it may be procured from a lead merchant. Cast Sheet Lead. For roof work cast sheet lead is con- sidered by some to be superior to milled sheet lead, but on the whole its drawbacks outweigh any merits it may possess. Cast lead can only be made in comparatively small sheets, it requires to be thicker than milled lead, it is not as a rule of uniform thickness, and sometimes it is porous. It is also dearer than milled lead, and neither does it permit of the same neatness of finish when it requires to be worked into various shapes. The chief advantages of cast lead are its architectural appearance, and it will expand and contract with less risk of breaking than milled sheet lead. The latter is owing to the molecules of cast lead being in normal positions, whilst those of milled lead are squeezed into un- natural places, and may be more or less in a state of internal stress. Milled Lead is made in sheets about 33 feet long and from 7 ft. 6 in. to 8 feet in width. Some few rolling mills make sheets up to about 40 feet in length and "9 feet in width. Narrower sheets than those usually made can be 14 DOMESTIC SANITARY ENGINEERING AND PLUMBING obtained if desired, and special widths are often an advan- tage as much scrap can in many cases be avoided. Milled sheet lead is made by first casting a block of lead approximately 8 feet square and 6 inches in thickness. After this has cooled sufficiently it is hoisted on the rolling machine, and passed forwards and backwards between two heavy chilled steel or cast-iron rollers, which are located in the centre of the frame-work of the machine. The rolling process is continued until the block of lead has been reduced to about three-quarters of an inch thick. The flow of the metal, due to the enormous pressure which is brought to bear upon it, is practically all in a longitudinal direction ; as regards the width, that is not much affected, the edges only taking a ragged or irregular form. When the thickness of f inch has been reached, the lengthened plate of lead is cut cross- ways into suitable lengths, which vary with the size of the finished sheet and strength of lead required. Very often two sheets are rolled at one operation ; this is done by rolling one of the pieces, into which the whole plate has been divided, until its thickness is equal to about that of 10 -Ib. sheet lead; at this point the sheet is doubled, when by further rolling the desired thickness is obtained. The ragged edges are now straightened and the lead rolled up. When lead thicker than 6 Ib. per sq. foot is required the sheets are rolled separately. Fig. 2 gives drawings of a milling machine. It will be observed that numerous wood rollers are arranged from end to end of the machine, and these carry the lead, and of course impart easy motion to the lead when being rolled. Pressure is applied to the centre rollers by the aid of a wheel and suitable gearing, and the top roller is raised and lowered by the same means. The cutting roller, which is shown on the right side of Fig. 2, is for trimming the edge of a sheet. One is fixed on each side of the machine, and occasionally a guillotine arrangement is also provided for shearing the sheets cross- wise. Iron. The principal ores from which iron is obtained are Ked Hematite, Brown Hematite, Magnetic Iron Ore, 16 DOMESTIC SANITARY ENGINEERING AND PLUMBING and Siderite. The first three occur as oxides of iron, whilst the latter occurs as a carbonate of iron. Iron takes three principal forms, viz. : Cast Iron, Wrought Iron, and Steel. The difference in their physical and chemical properties is due chiefly to the difference MATERIALS AND THEIR PROPERTIES 17 18 DOMESTIC SANITARY ENGINEERING AND PLUMBING in the amount of carbon which enters into their com- position. Cast iron contains over 1J per cent, of carbon, and a large amount of other impurities. Wrought iron is the purest form of iron, and only contains about -J per cent., and less, of total impurities ; the amount of carbon should not exceed -J- per cent. Mild steel contains less than J per cent, of carbon, whilst hard steel contains from J per cent, to about 2 per cent, of carbon. Both forms of steel are nearly free from other impurities. The principal impurities in iron are carbon, manganese, phosphorus, silicon, and sulphur. Cast iron at the present time is very largely used in the manufacture of sanitary fittings, such as baths, lavatories, etc. ; for drainage work, soil and waste pipes, and for boilers, pipes, and fittings in connection with low pressure hot-water and steam-heating work. Properties of Iron. Cast iron is hard and brittle, and varies in colour when fractured from a silver white to a dark gray. On account of its fusibility it can be run into moulds so as to take various forms. The specific gravity of cast iron varies from 7 to 7'6, and a cubic foot weighs from 437 to 474 Ib. Its tensile strength is comparatively low, whilst its resistance to a crushing force is high ; on an average the tensile strength of cast iron is about 7 tons per sq. inch, whilst the average crushing load is about 48 tons per sq. inch. Cast iron is less affected by oxidation than either wrought iron or steel. Wrought iron is of a bluish -white or bluish -gray colour, and it is readily oxidised with moist air. It is very malleable and ductile, and at high temperatures can be forged, rolled, and hammered into various shapes. The average specific gravity of wrought iron is 7 '7 8, and it weighs 485 Ib. per cubic foot. Its average tensile strength is 22 tons per sq. inch, whilst the average crushing load is only about 17 tons per sq. inch. It requires a very high temperature to effect its fusion and, unlike cast iron, cannot be run into moulds. Steel may be hard or soft according to the amount of MATERIALS AND THEIR PROPERTIES 19 carbon it contains. Like wrought iron, it can be forged and welded ; it is malleable, ductile, and very tenacious. Its colour is bluish-gray, and it is readily oxidised upon exposure to moist air. Ordinary steel can be tempered to take different degrees of hardness by cooling in liquids such as oil, water, etc. The tensile strength of ordinary steel is about 50 tons per sq. inch, and its crushing load is about 150 tons per sq. inch. The strengths of mild steel vary considerably, and whilst they are much less than those of ordinary steel they exceed those of wrought iron. The tensile strength of mild steel varies from 25 to 35 tons per sq. inch. Malleable Cast Iron is largely used for fittings in connection with wrought iron tubes, for pipe hangers, clips, and brackets, and for many other small fittings. The property of malleability is imparted to cast iron by decarbonising it. To effect de- carbonisation the cast fittings may be embedded in hematite or oxide of manganese, and subjected to a red heat for a period which varies from a few days to several weeks, according to the size of the casting that is being treated. This annealing process is also said to increase the tensile strength of the iron to about 1*6 times that of cast iron. Iron Pipes are made of either wrought or cast iron, accord- ing to the purposes they are intended to serve and the pressure they are required to withstand. Wrought-iron pipes are of two kinds : the first are those which have plain welded joints, and the second are those which have lap welded joints. The former are used for general work, such as for conveying water, gas, or steam ; whilst the latter are for withstanding high pressures, as in hydraulic work, and for small bore hot-water heating ap- paratus. Cast - iron pipes may be classified under three heads. Firstly, those which are cast in a horizontal position ; secondly, those cast on an inclined plane : and thirdly, those which are cast when vertically arranged. Pipes which come under the first head are chiefly those of a light character, such as rain-water pipes ; frequently such pipes have thick and thin sides, owing to the cores being buoyed or bent 20 DOMESTIC SANITARY ENGINEERING AND PLUMBING upwards when the metal is flowing into the moulds. Under the second heading fall the heavy section cast-iron soil, waste, and drain pipes ; many water pipes are also cast in this position. When pipes are cast on an inclined plane the socket end is at the higher point, and the pipes produced in this way are superior to those cast in a horizontal position. Water pipes which are required to withstand high pressure should be vertically cast, and the best and most reliable cast pipes are produced in this way. Vertically cast pipes should have their socket ends downwards, their cores should be well and truly formed, and they should also be cast at least one foot longer than their finished lengths. The extra length is to ensure compactness of grain at the spiggot end of the pipes, and it should be cut off afterwards in a lathe. The pipes should be true in section, and be free from defects and flaws of all descriptions. It is very important that pipes for conveying drinking water be coated with some preservative immediately after casting, and before their surfaces are covered with a thin film of rust. Pipes should be tested before leaving the foundry to not less than twice the pressure to which they will be subjected when laid or fixed in position. Coatings for Iron Pipes. By arresting corrosion the life of iron pipes may be prolonged for a more or less considerable time. For this purpose different substances are applied to the surfaces of pipes, or their surfaces are treated in some special way. Protective materials take the form of oil paints, enamels, bituminous compositions and glazes, or the surfaces of the pipes may be subjected to a barffing or galvanising process. Oil Paints are frequently used for treating the surfaces of iron rain-water pipes, gas pipes, etc., but painting only has a limited life and requires periodical renewal. Corrosion is generally very active on the inside surfaces of pipes, and the repainting of these is generally omitted after once they are fixed in position. The inner surfaces of pipes can be easily painted in the workshop or other suitable place, but if the coating is to be successful the surfaces will require to be properly prepared and all loose scales and sand removed. A MATERIALS AND THEIR PROPERTIES 21 wire brush or similar tool is very suitable for cleansing the inner surfaces of pipes. A Bituminous Composition, such as Dr. Angus Smith's, is fairly durable, provided the pipes have been properly cleansed and no rust is present when the coating is applied. It is essential, however, that dilute acids are not brought in direct contact with the coating, or the latter will be readily destroyed. To apply the composition, it is raised to a temperature of about 350 F., when the pipes are dipped into the hot solution, and remain submerged until they acquire the same temperature as the solution itself. The pipes are afterwards withdrawn and allowed to drain. Dr. Smith's preparation consists of a mixture of coal-tar and pitch, with a small added quantity of mineral oil. It is largely used for coating cast-iron soil pipes, waste pipes, water pipes, and drains. Srtiall ivrought-iron pipes when laid in the ground can be protected by putting them in small wooden channels and surrounding the pipes with pitch. Glazes are also frequently used for coating the inner surfaces of cast-iron soil pipes and drain pipes, and although these may ensure smooth surfaces when the pipes are new, it is very doubtful if glazing is worth the price it costs. The glazes which are applied to iron pipes are easily destroyed by dilute acids, and the cutting of pipes also tends to damage them. Barffing. This process may consist of heating the articles to be treated to redness, and by subjecting them to the action of superheated steam. The steam is decomposed, and a thin adherent film of magnetic oxide of iron is formed on the surfaces, and this, for a time, prevents further oxidation taking place. Barffing is suitable for boilers and hot-water pipes. Galvanising is largely resorted to for the protection of wrought-iron cisterns, cylindrical tanks and wrought-iron pipes. It is very effective in many cases, but certain waters readily attack and destroy it. The galvanising of wrought-iron pipes often tends to make them brittle, and great care is necessary when bending the pipes cold. If, however, these pipes are heated to redness, so that bends can be the more 22 DOMESTIC SANITARY ENGINEERING AND PLUMBING readily made, the zinc coating is destroyed, and the pipes require to be regalvanised. Galvanising is effected by first removing any scales or dirt from the articles to be treated, afterwards they are thoroughly cleaned by submerging in an acid bath, and finally by dipping them into a bath of molten zinc. Copper. This metal occurs in the native state in certain places, but only in few in sufficient quantity to work it. The principal locality in which native copper is found is in the district south of Lake Superior. The ores of copper are chiefly found in the form of oxides, carbonates, and sulphides, the latter being the most important. Sulphides are Copper Glance, Copper Pyrites, Erubescite, and Fahl Ore. In this country Cornwall is the only county where large deposits of the ore are found. Properties of Copper. It is a tough, very malleable, and ductile metal, and its specific gravity varies from 8*6 to 8 '9, its highest value being when in the form of wire. The weight of copper per cubic foot varies from 537 to 555 Ib. It is one of the best conductors of heat and electricity. The tensile strength of cast copper is about 8 J tons per sq. inch, and when in the form of wire 2 6 tons per sq. inch ; in the sheet form its tensile strength is approximately 14 tons per sq. inch. It can be forged when either hot or cold, and is softened when heated to redness and suddenly submerged in cold water. Copper often contains impurities such as traces of lead, iron, zinc, tin, and of other metals. When in the presence of moisture copper is covered with a film of basic carbonate, and when heated to redness it is covered with a film of oxide. It is dissolved by cold nitric acid, but in the absence of air it is not affected by either sulphuric or hydrochloric acid. In the presence of air, however, copper is attacked by weak solutions of these acids. Coatings for Copper. To prevent the film of basic carbonate forming on copper-lined sinks and cisterns, and on pipes, etc., their surfaces are frequently tinned. Copper pipes are also electro-plated or lacquered. As the salts of copper are poisonous, plain copper pipes and MATERIALS AND THEIR PROPERTIES 23 cisterns are not suitable for conveying and storing water which is required for human consumption. Sheet Copper. Unless specially ordered, sheet copper is not usually made in pieces containing a greater area than 1 4 sq. feet ; it is rolled to various thicknesses to suit the many purposes for which it is required. Common sizes of sheets are, 5 ft. 3 in. by 2 ft. 8 in,, 4 ft. by 3 ft. 6 in., and 4 ft. by 2 ft. Copper Tubes are of two kinds : (a) those which are formed and joined with longitudinal seams, and (b) solid drawn or seamless tubes. In the first case the edges which are to be joined are reduced in substance so that they can be overlapped a little, and afterwards they are brought together and brazed. In seamless copper tubes the drawing process tends to make them brittle, and in order to restore ductility it is necessary for the tubes to be annealed. Copper tubes have a very large sphere of usefulness in connection with hot water supplies, and also in connection with hot water and steam heating work. Tin. There is only one tin ore, and this is known as Cassiterite or Tinstone. It occurs in the form of an oxide, the chief deposits in this country being confined to Corn- wall. Properties. Tin has a bright lustre, is very malleable, and melts at about 442 F. It is not very ductile, and the tensile strength of cast tin is only about 2 tons per sq. inch. Its specific gravity is 7 '2 9, and a cubic foot weights approximately 455 Ib. Like lead, tin is not a perfectly elastic metal. When bent, tin makes a crackling noise, and by this means it is easily distinguished from a tin and lead alloy. It is not readily affected by atmospheric air at ordinary temperatures, but it is easily converted into an oxide when heated to redness. Tin is not affected by soft acid waters. Strong hydrochloric acid and hot dilute sulphuric readily dissolve tin, and it is rapidly attacked by nitric acid. Sheet Block Tin is largely used for covering counter tops in vaults, and for covering drainers to sinks, etc. Tin is specially suited for such purposes on account of its cleanly 24 DOMESTIC SANITARY ENGINEERING AND PLUMBING appearance and its softness ; the latter property prevents it scratching and chipping glass and china ware. Block Tin Tube. This is made precisely in the same way as lead pipe, only that a much stronger machine is required on account of tin being harder than lead. Tin tubes are largely used for spirit and beer pipes, as lead and tin-lined lead pipes are not suited for this purpose. Zinc does not occur in the native state, and the principal zinc ores are found as oxides, as carbonates, and as sulphides. The last is known by the name of Blende, and is the most abundant and valuable ore. In England zinc is found in Cornwall and in Derbyshire ; it is also found at the Isle of Man. Properties of Zinc. In colour zinc is bluish-white, and when exposed to moist air it is covered with a thin film of oxide. Its specific gravity when cast is about 7, and a cubic foot weighs about 437 Ib. Zinc is a brittle metal when cold, but it is fairly malleable when heated to the temperature of boiling water. At 300 F. it can be rolled into sheets, but when its temperature exceeds about 350 F. it again becomes brittle. Its melting point is approximately 770 F., and at a red heat it boils. Zinc is readily attacked by an acid-laden atmosphere, and dissolves readily in most acids. It always contains more or less im- purity, such as iron, lead, and other metals. Sheet Zinc. The sizes of the sheets into which zinc is rolled vary in length from 7 to 10 feet, but all sheets have a uniform width of 3 feet. The life of zinc is com- paratively short in large manufacturing centres, and when laid on roofs under otherwise favourable circumstances it will only last about twenty years. In country districts it has a much longer life, provided it is properly laid and the area of one piece is not too large. In the ingot form zinc is known as spelter. Alloys. An alloy may be defined as a mixture of different metals where their union has been effected by fusion. In an alloy the metals are not in proportions which produce a definite chemical compound, for upon cooling whilst in the MATERIALS AND THEIR PROPERTIES 25 molten state it is found that the metals composing the alloy tend to separate and arrange themselves in different layers. The tendency for the constituents of an alloy to separate themselves is very noticeahle in plumbers' solder, and before a plumber takes a ladleful of solder from the pot to make a joint, he first stirs it so as to thoroughly mix the lead and tin. An alloy, however, is not a mere mechanical mixture, which tends to separate into its constituent parts under favourable conditions, and although such separation takes place to a considerable extent, it will be found that the layers are not pure metal, but that each has alloyed with it a certain amount of another nietal. Properties. The properties of alloys usually differ con- siderably from those of the constituent metals, alloys being usually harder and more brittle. Their melting points are frequently less than those of their most fusible constituents, and their tensile strengths and ductility may be either greater or less than those of any constituent metal. The chief alloys (excepting solders) with which the plumber is principally concerned, are the Copper-Tin alloys, and the Copper-Zinc alloys. From the former are made the best qualities of water and steam fittings, whilst fittings of a lower grade are produced from the latter alloys. Gun-metal belongs to the copper-tin series. A strong alloy, and one which will withstand the action of acid waters, is produced by a mixture of 90 per cent, copper and ten per cent. tin. An inferior gun-meted to the above is made by combining 85 per cent, copper, 10 per cent, tin, and 5 per cent, lead. For spindles in connection with valves, a harder gun- metal can be produced by alloying with the copper and tin a small amount of phosphorus. Brass. This alloy is obtained by a mixture of copper and zinc. Brass may be either yellow or white, the colouring depending upon the percentage of copper which enters into its composition. When the amount of copper is less than 45 per cent, the colour is white. The proportions of copper and 26 DOMESTIC SANITARY ENGINEERING AND PLUMBING zinc vary considerably, and the amount of copper may be anything from 40 to 80 per cent. Ordinary yellow brass contains 6 7 per cent, of copper and 33 per cent, of zinc. Many of the inferior brasses contain more or less lead. Common brass is unsuited for water fittings when the latter are used for sea- water and soft acid waters. The action of sea-water on copper-zinc alloys can be minimised to a great extent by including in the mixtures 1 per cent, of tin. Brass which contains iron and lead is subjected to more or less rapid corrosion when in contact with sea-water. Compositions of a few other alloys are as follows :- Red Brass. 90 per cent, copper and 10 per cent. zinc. Muntz Metal. 60 per cent, copper and 40 per cent. zinc. Aluminium Bronze. 90 to 97 J per cent, copper and 2J to 10 per cent, aluminium. German Silver. 50 per cent, copper, 25 per cent, zinc, and 25 per cent, nickel. Soft Gun-metal. 95 per cent, copper and 5 per cent. tin. The greatest tensile strength of a copper-tia alloy is approximately 17 tons per sq. inch, the relative proportions of copper and tin being roughly 88 per cent, of the former to 1 2 per cent, of the latter metal. A greater or less percentage of tin diminishes the tensile strength of the alloy. In the copper-zinc alloys the tensile strength reaches a maximum of approximately 141 tons per sq. inch, when the copper and zinc are in the proportion of 80 per cent, of the former to 10 per cent, of the latter; the tensile strength diminishes with either a greater or smaller amount of zinc. Sanitary Pottery. Although it is often thought that the materials for the production of porcelain goods are found in the neighbourhood of the potteries, such is not the case. The principal materials employed to make porcelain are calcined bone, china clay, blue or ball clay, flints, and partially decomposed granites. China clay and granites are obtained from Cornwall, ball clay from Dorset and Devonshire, and the flints are obtained from Newhaven and Dieppe. Each MATERIALS AND THEIR PROPERTIES 27 material has its own special property, such as for imparting fineness, plasticity, and strength to the clay, or for controlling the rate of contraction when firing the clay. To produce porcelain sanitary fittings from the crude materials to the finished state takes several weeks, and much skill and care is involved as they pass through the different stages of manufacture. The first stage towards the prepara- tion of the clay is the separate treatment of each ingredient, which is ground in water so as to bring it into a suitable state for mixing. Before the flints can be ground they are first calcined. In the " slip house " the ingredients are measured out and mixed in the correct proportions, and any shade or colour can be obtained by introducing into the liquid " slip " stains and metallic oxides. The " slip " now requires to be brought into the plastic state, and this is done by pump- ing it into clay presses, where excess of water is removed. After leaving the presses, the plastic clay is passed into the pug mill, from which it issues of uniform consistency and ready for the potter to mould. The clay is pressed into the moulds so as to form one homogeneous mass. A mould is made up in a number of parts, the exact number depending upon the kind and shape of fitting required, and great care is essential in joining the separate parts so as to avoid cracks appearing at the joints when the goods are fired. Before the goods are ready for the oven they must be thoroughly dried. The firing process is performed by placing the goods in " saggars," which are fireclay receptacles of different sizes and shapes and of about 1 inch in thickness ; the " saggars " afford protection from the intense heat of the fires, and subject all parts of their contents to a more nearly uniform temperature. The first firing process is carried out in the " Bisque " ovens or kilns, which are of circular or other construction, the firing being done from a number of points. The " saggars " containing the things to be fired are built up in the ovens, and when the latter are fully charged the doorways are built up and the fires lighted. By means of dampers the draught is regulated and a gentle heat maintained for about the first 3 hours ; the temperature 28 DOMESTIC SANITARY ENGINEERING AND PLUMBING being raised afterwards to about 2000 F. for another 24 hours or so. The goods when removed from the bisque ovens are in a partially baked and semi-porous state, and whilst in this condition any minor defects can be rectified, the spoiled or damaged goods being thrown aside. If the articles are to receive any form of decoration this is often done at this stage, but if the goods only require a plain finish they are taken to the dipping house to be glazed. The dipping process may consist of passing the things through the dipping-tub, in which glaze is suspended in water, the ware absorbing a sufficient quantity of the glaze to form a coating over its surfaces. Before the glaze can be con- verted by firing into a hard and practically impervious substance all moisture requires to be expelled by gradually drying the goods. To be fired, the treated goods are placed in " saggars " as before, and to protect the glazed surfaces, and to prevent them adhering to the " saggars," the goods are supported on studs. The ovens in which the glazing is completed are known as " Glost " ovens , they are similar to the " Bisque " ovens, excepting they are sometimes a little smaller and are heated to a higher temperature. The firing takes about 24 hours, and the temperature is quickly raised in order to fuse the glaze and produce a smooth and even surface. After the ovens have been cooling for about two days the contents are withdrawn. Fireclay and Brown Earthenware Goods are made from coarser materials, such as the fireclays which overlie the coal measures. Different qualities of clays are also produced by mixing the coarser with the finer qualities along with other ingredients. Earthenware Drain Pipes. There are two principal classes of drain pipes, (a) fireclay pipes and (b) stoneware pipes. The clay for the former is the more abundant and the more widely distributed, while that for the latter is chiefly confined to the counties of Dorset and Devon. Fireclays are rather coarse grained, and the pipes made from these clays depend principally for their water-tightness MATERIALS AND THEIR PROPERTIES 29 upon the quality and thoroughness of the glazing. On the other hand, the clays from which stoneware pipes are made are very closely grained and are denser than fireclays. Stoneware pipes are much superior to fireclay pipes, and the former are especially suited for resisting the action of acids. In England earthenware pipes of 12 inches diameter and less are made usually in lengths not exceeding 2 feet ; larger sizes being generally 3 feet long. In Scotland 3 feet lengths are general for all sizes of earthenware drain pipes which exceed 3 inches diameter. The thickness of stoneware pipes is usually equal to about jL their diameters, but fireclay pipes to be of equal strength would require to be a little thicker than this. The relative strength of fireclay and stoneware is stated to be in the ratio of 1 to 1 2. Earthenware pipes are generally salt glazed. Concrete Tubes for drainage work are now being made, and these are quite smooth, true in section, perfectly straight, and practically non -absorbent. They are made in steel moulds, and the concrete tubes may be reinforced with metal rods when extra strength is desired. In price concrete tubes compare favourably with earthenware pipes, and for large sizes of pipes concrete has the advantage. CHAPTEE II ROOF WORK Metal Coverings. The principal sheet metals which are used for roof work are lead, copper, and zinc. For general work sheet lead possesses important advantages, so that its displacement by other materials is confined chiefly to special cases. Lead has the advantage of adaptability in a marked degree, as it can be readily worked into shapes to fit almost any position ; it is also a very durable material, and its cost is not excessive. The principal drawback of lead is its weight. Copper possesses the advantage of lightness when compared with lead, and it can be rolled and used in much thinner sheets. Copper is a very durable metal, and is specially suited for covering domes, turrets, and similar structures. Zinc may be used in rural districts with good results, provided it is properly fixed, and removed from situations where large volumes of sulphurous acid gas are emitted. For towns and manufacturing districts zinc is unsuitable on account of the amount of sulphurous and other acids which are always present in the atmosphere of such localities. The advantages possessed by zinc are lightness and cheapness. When executing sheet leadwork the points which require consideration are as follows : 1. That the area of one piece of lead be not larger than where movement due to changes of temperature can readily take place. 2. That the lead be fixed in such a manner as will prevent its sliding or tearing from its original situation by its own weight, or being removed by the force of the wind. 3. That water be prevented from gaining access to the woodwork 30 ROOF WORK 31 supporting the lead by the former rising between the laps or passings. 4. That all woodwork supporting lead be properly laid to the required falls, and that before the leadwork is placed in position all projecting nails be punched and the woodwork swept clear of dirt. Narrow boards should be used for gutters, lead flats, etc., and these should be laid in the direction the water will flow. Turning back to the first point, lead, if laid in very large pieces, is unable to move freely on account of its weight and its softness. Thus, when a piece of lead expands or contracts, unless it can bodily move, stresses are concen- trated at one or more points ; the result is the lead begins to buckle, and is eventually cracked or torn. With regard to the maximum area of one piece of lead, this should not, as a rule, exceed 20 superficial feet. Discretion of course requires to be exercised, according to the purpose for which the lead is required and the position in which it is to be fixed. It is obvious that when lead-work is laid in exposed situations it should be in smaller pieces than when in sheltered places. The second point refers to the methods of securing lead- work in position, but these will be dealt with as the various parts of roof work receive consideration. With regard to the third point, water may gain access to, and bring about the decay of the timbers in the following ways : (a) By capillary attraction due to accumulation of dirt at gutter drips, or by drips being too small, (b) By driving wind and rain, the latter getting under the lead when the drips are shallow, or owing to the passings of joints having insufficient lap. (c) By defective workmanship. LEAD FLATS. Lead Flats usually include all lead covered surfaces which can be walked upon. On flats the leadwork is arranged in the form of bays by either introducing solid or hollow rolls. Solid roll work possesses the advantage of not being so readily disfigured when walked upon as hollow roll work, and the former can be more speedily executed than the latter. The 32 DOMESTIC SANITARY ENGINEERING AND PLUMBING chief drawback of solid roll work is the difficulty of securing the lead on inclined surfaces so as to prevent its sliding or crawling down, unless soldered dots are resorted to. As ordinarily carried out, solid roll work only permits of the lead bays being secured across their top edges and along their undercloaks. Hollow rolls permit of the lead being fastened by means of copper ties on both sides of a bay, and thus very substantial fixings are obtained. In exposed situations the leadwork is also less liable to be displaced by high winds, as no free edges are left (except under special circumstances) at the sides of the rolls. Wood cores are also dispensed with. The chief drawback of hollow rolls is their liability to disfiguration by being crushed when walked upon, and by materials falling upon them during the erection of buildings. In England solid roll work is generally adopted, whilst in Scotland hollow roll work is predominant. Fig. 3 shows a portion of a lead flat which is 17 feet in width ; the flat on three sides is supposed to be bounded with high walls, whilst a low parapet wall is represented in front. As the water must drain in the case shown towards the front wall, where a gutter is formed, a drip will be necessary to divide the flat in order to reduce the bays to a suitable length. The left side S, Fig. 3, shows how solid rolls are arranged, whilst the right side H indicates how hollow rolls are generally arranged where the drip or step dividing the flat is not more than 3 inches deep. When arranging the bays for a flat, their widths should be governed to a great extent by the widths of the sheets of lead, so as to avoid producing unnecessary scrap. As a rule the width of the bays should not exceed 2 feet when their length is about 8 feet. If the bays are short, their width, of course, can be increased, but the width decided upon should cut up the sheet of lead to advantage. Approximately 9 inches of lead are required for the under and overcloaks of solid rolls, and about 7 inches for hollow rolls. Thus if a sheet of lead has a width of 7 ft. 9 in., this will cut into three strips, each with a width of 2 ft. 7 in. Supposing hollow rolls are used, then 2' 7" 7" 2 ft., ROOF WORK 33 which would be a suitable ^vidth for the bays when of moderate length. For solid rolls we have 2' 7" 9" = 1 ft. 10 in. as the width suitable for the bays. Solid Rolls. A common method of treating solid rolls is shown in Fig. 4. The thick edge of the undercloak should be reduced with a shave-hook or other tool, or a crease will be formed in the overcloak. The overcloak is shown finished DR.R FIG. 3. Plan of lead flat showing arrangement of solid and hollow rolls. off with a width of about one inch on the flat ; the object of the lap is to stiffen the overcloak at the angle, and also to prevent it slackening to a great extent on the roll. To hold the lead in position the rolls require to be a good shape and well undercut, as indicated in Fig. 4. For general work a wood roll should not be less than 2 inches high, its widest part not smaller than If inches, and the bottom should not exceed 1 inch in width. Instead of the overcloak being treated as in Fig. 4, some- 3 34 DOMESTIC SANITARY ENGINEERING AND PLUMBING times the lap on the flat is omitted, and the edge of the overcloak is finished off on the side of the roll, about one quarter of an inch above the flat. It is supposed in the latter case that moisture is prevented from gaining access by capillary attraction to the timbers which support the lead. The possibility of leakage by this means, however, is often overrated, for where rolls have a minimum height of 2 inches it is usually impossible, under ordinary circum- stances, for capillary attraction to occur. Finishing off the overcloak on the side of a roll possesses the disadvantage of allowing the lead to slacken on it, either when the lead is walked upon or by the action of the weather. FIG. 4. Section of solid roll showing leadwork. It is desirable when laying lead on flats that the wood rolls, after being fitted by the carpenter, be finally secured in position by the plumber as the work of lead laying proceeds. Laying Lead. Before a piece of lead for a flat is laid in its place, assuming the necessary setting up and bossing in connection with it has been done, the large flat surface should be raised by either striking it with the hand or with a soft wood dresser. The lead is then laid in position, and the wood roll fixed against the upstand which represents the undercloak, and the roll made secure by nailing it down. The undercloak lead is easily pressed over the roll with the hand, and finished off with a soft wood dresser. For setting in the lead at the sides of the rolls, a proper setting-in dresser ROOF WORK 35 should be used, or an ordinary beech dresser which has been cut to suit the shape of the rolls may be used instead. To get the overcloak tight on the roll often presents a little difficulty to the young plumber, but this can be accom- plished by carefully bending the upstand well over the roll for the whole of its length ; the lead can then be made to take the shape of the roll by partially setting it in along the free edge with a large hammer and with a spare piece of wood roll, which has had any sharp edges rounded off. The overcloak can afterwards be drawn tightly over the roll, by setting in the sides with a blunt-edged dresser, and finishing off with an ordinary setting-in dresser or other suitable tool. To keep lead free from tool marks the flat dresser should be used as sparingly as possible, and if the rolls are treated in the manner described they will present a smart and clean appearance. Large flat surfaces can be left free from marks by using a planisher, which is made from a piece of scrap lead, in lieu of using a dresser. Fig. 5 shows how finished solid roll work on a flat appears where a step is necessary to reduce the length of the bays. Overcloaks should be arranged that their free edges are on the side most sheltered from driving rain. The step in a flat must be higher, of course, than the rolls, in order to prevent water following under the laps along the rolls where the latter butt against the step. If, however, it is found in a flat that the height of a step and the rolls are about the same, and that no great departure from this can be made, the ends of the rolls in the immediate neighbourhood of the step should have their height reduced so as to come half an inch or so below it. Roll Ends. The ends of solid rolls should terminate flush with the lower edges of flats, and not be cut a little short as is often done. With regard to the shape of the ends, it is unimportant whether they are cut off square or cut sloping. Before working down a roll end it is essential that all sharp edges are removed. Eoll ends A, Fig. 5, are not difficult to work down, but care requires to be taken so that the lead is not unnecessarily reduced or split at the sides by too much setting-in. The overcloak B, Fig 5, is more difficult to work 36 DOMESTIC SANITARY ENGINEERING AND PLUMBING on account of the lead requiring bossing round the lower roll, and also finishing off with about an inch lap on the flat surface. In working overlap B, Fig 5, the left side of the roll is first dealt with, and that part should be well held down to prevent it rising whilst the right side is being bossed into shape. The upper end of roll, as at C, Fig. 5, can readily be dealt with by first pressing down part of the upstand along FIG. 5. General view of solid roll work. the top of the bay, so as to enable the overcloak at C to be bent round the roll. By careful working, the lead is driven into the corner so as to take the form required. For securing the bays the undercloaks are copper nailed to the wood rolls, but this fastening is inadequate to prevent the bays creeping down unless the surfaces have little or no pitch. Soldered Dots. When lead is laid on a surface with a moderate pitch, and where solid rolls are used, soldered dots ROOF WORK 37 are frequently adopted as fixings. For a flat with a moderate pitch usually two dots are made on each bay, these being located near the lower edge. Soldered dots may either be flush wiped or raised, both forms being shown in Fig. 6. The raised form is represented by A, and the flush one by B ; FIG. 6. Raised and flush soldered dots on flats. but in each case the lead is held secure by means of strong screws, which are driven through it into the woodwork be- neath. To make the solder adhere to the screws they are first tinned. When making soldered dots, the heads of the screws should be left raised a little above the lead that the FIG. 7. Hollow roll with copper tie. solder will flow beneath, and the whole be properly sweated together. Occasionally tinned washers are used to present a large surface to which the solder will adhere, but if strong screws are used and their heads completely tinned, then washers may be discarded. The solder over the screws need not be more than about 2^ inches diameter. The principal drawback of soldered dots is that they hold 38 DOMESTIC SANITARY ENGINEERING AND PLUMBING the lead too rigid, and after a time it is found that, where they support heavy pieces of lead, the screws work their way up and through the solder, owing to the thrust and pull which are concentrated upon them. Of course where dots are well made, suitably placed, and the pieces of lead not unduly large, it may take many years before the screws work out above the surface of the solder. ) I 1 ) ) 1 1 / 7 1 FIG. 8. View of hollow roll work for lead flats. Hollow Rolls are much superior to solid rolls for sloping surfaces, as these permit of the leadwork being made secure without resorting to soldered dots. These rolls are not made so large as solid ones, their height as a rule not exceeding 1 J inches; the thickness of the lead affects the size of hollow rolls to a certain extent. In Fig. 7 is shown a section of a hollow roll with copper tie ; the latter is screwed to the woodwork, and the free end is turned between the under and overcloaks as shown. Copper ties are fixed about 2 feet apart, and are from ROOF WORK 2 to 3 inches in width. On account of the under and overcloaks being wholly in contact with each other one bay cannot slip from another. The leadwork as a whole is prevented from bodily giving way by the copper ties, which are provided at regular intervals throughout the whole length of the rolls. A plan of hollow roll work has already been shown in Fig. 3, whilst Fig. 8 gives a part view showing how the rolls may be treated where a drip or step occurs in a flat. Where the curb C, as in Fig. 8, is a plain one, the roll ends K are usually finished off by turn- ing them down on the curb as shown. If a curb is in an exposed situation it is neces- sary to add a nosing piece, over which the lead is turned, as at A, Fig. 9. Occasion- ally roll ends are treated as at B, Fig. 9, but in the latter case much more labour is involved in turning the roll on the curb or vertical surface. As a rule the method denoted by B, Fig 9, of forming the end of a roll is unnecessary, and not worth the extra labour it entails. When a step or drip occurs, as at S, Fig. 8, it is treated differently to that where solid rolls are used. In hollow roll work the under and overcloaks of steps are " clinked " or FIG. 9. Methods of finishing the ends of hollow rolls. 40 DOMESTIC SANITARY ENGINEERING AND PLUMBING ' welted " together, as at C, Fig. 10. In order to double the lead under where the roll end of the upper bay occurs, the turned end S, Fig. 8, is cut to its finished length ; the lead forming the overlap is then free at the end of the roll, and will admit of being doubled under and into position as partly shown by D, Fig. 10. The upper end of the hollow roll E, Fig. 8, or one in a similar situation, can be treated in different ways, but the method shown is the simplest and has the best appearance. Another method of treating the upper end is to turn the roll practically its full size on, and to the top of the upstand. A third method is to form a clink or welt on the upstand part. The enlarged detail A, Fig. 11, shows how the undercloak at the upper end of a hollow may be pre- pared, whilst B of the same Fig. de- notes the overcloak in a finished state. It is essential when the ends are treated as in Fig. 1 1 that the lead of the overcloak is worked well down and into the corner p ; unless this is done a leakage may occur at that point with a heavy shower of driving rain. The undercloak A, Fig. 11, is prepared by first bossing up the corner about 1 J inches high in the ordinary way, and then by forming the lead so as to enable the undercloak to be turned over inwards as shown. As regards the overcloak, that is worked into position in a similar manner to that in which solid rolls are used, excepting that from pointy, Fig. 11, the free end of the overcloak is doubled under to form part of the roll. From the particulars supplied it will be obvious why FIG. 10. ROOF WORK 41 hollow and solid rolls differ- in arrangement, see Figs. 5 and 8, when the treatment of their ends is considered. Intersecting Rolls. A part plan showing how the lead- work is arranged where solid intersecting rolls are used is given in Fig. 12. The numbers on the bays indicate the order in which they may be laid. Where the intersections occur some of the underlaps will require to be cut out so as to avoid giving the rolls a clumsy or bulged appearance at those points. The method of trimming the undercloaks is shown by bay No. 1 2 ; at the top of the centre roll part of the undercloak is represented cut away, and the free edge should be well reduced by a rasp. Bay No. 1, Fig. 12, it will be observed, has an overcloak on each side ; this is owing FIG. 11. Formation of the upper ends of hollow rolls. to the overcloaks of the side bays being turned in the same direction. The allowance of lead to cover the intersections requires to be liberal, otherwise there is difficulty in obtaining adequate laps and preventing the lead from being torn. When working an overcloak round the intersections, after any particular part is once in position it should be firmly held there, to prevent it being withdrawn when bossing at another point. After the rolls are covered the overeloaks should be made secure by copper ties which have been previously introduced. For exposed situations the ties should be about 2J inches wide, and fixed at intervals not exceeding 2 feet apart. In hollow roll work different methods are adopted for 42 DOMESTIC SANITARY ENGINEERING AND PLUMBING dealing with intersections, some involving much more labour than others to carry out. For large flats, wood cores are generally used for the centre and diagonal rolls, as at D, Fig. 13. Wood rolls in these positions simplify the work and allow the lead to be more quickly laid. No wood cores are used for small flats with intersecting rolls, as no special difficulties are presented as in larger flats which require numerous rolls. FIG. 12. Plan showing intersecting solid roll work. In Fig. 13 two methods are shown of dealing with the intersections of the rolls ; in each case wood cores are used for the centre and diagonal rolls, and the upstand of oppo- site bays is turned on these cores, the free edges of the lead meeting at the top, as on the left diagonal roll and on the centre roll Fig. 13. Separate capping pieces are required for the wood rolls, and the right diagonal roll shows a capping piece in position with the lead worked round the intersections and secured in position with copper ties. With ROOF WORK 43 regard to the hollow intermediate rolls, they may be joined with the wood cores by treating their ends in a similar way to that shown in Fig. 11 ; and the free edges of the over- cloak at the intersection may be trimmed, as in the left diagonal roll Fig. 13. The other method of treating the ends of hollow rolls, where the latter intersect with the wood cores, is illustrated on plan by cf, Fig. 13. In this case the ends FIG. 13. Plan showing intersecting hollow roll work. of the hollow rolls are turned upwards against the centre core, as shown by the enlarged detail A, Fig. 1 4. Separate capping pieces, as before, are required for the wood cores, but the re-turned ends of the hollow rolls present a rather clumsy appearance at the intersections. Fig. 14, B, gives a section showing how the ties are fixed to hold the capping pieces in position. For this class of work the ties should be fixed about 18 inches apart. In Fig. 13 the bays may be laid in the same order as in 44 DOMESTIC SANITARY ENGINEERING AND PLUMBING the previous figure, but the order may be varied, as this system of laying lead allows plenty of scope so far as the laying of the bays is concerned. It is sometimes thought that hollow roll work is more subject to leakage than where solid rolls are used ; this Hollow rolls turned against wood core. B Copper tie and capping flashing. FIG. 14. depends, however, upon the manner in which the work is done. If hollow roll work is indifferently executed, leakages at some of the corners will most likely occur ; on the other hand, if the work is properly done, and the rolls not unduly trampled upon, hollow rolls provide one of the strongest and best means for securing lead on large horizontal or moderately pitched surfaces. ROOF WORK 45 LEAD GUTTERS Lead-lined gutters, owing to their width, require little fall, and an inclination of 1 in 108, or 1 inch in 9 feet, is ample provided suitable drips are If practicable the of gutters between made. length two drips should not greatly exceed 9 feet, but for exposed situations shorter lengths are desirable. Box Gutters which have a uniform width need not be more than 1 inches wide unless they are formed be- tween two pitched roofs. A little extra width under the latter circumstances is ad- vantageous, as there is less danger of damaging the eaves when walking in the gutters. Tapering Gutters. Or- dinary parapet wall gutters which vary in width do not usually require their narrow ends more than 9 inches wide ; this width may also be reduced a little for special cases. Care should be exercised when setting out gutters so as to keep the area of the leadwork within reasonable limits. If a gutter which tapers is long, and the fall is all in one direction, its width rapidly increases ; this is es- pecially the case where the r-z 46 DOMESTIC SANITARY ENGINEERING AND PLUMBING roofs have a small pitch. Where possible, long, tapering gutters should be arranged in short sections, by locating drip-boxes at suitable points; when this is done the width of gutters can be kept within reasonable bounds. As a rule the height of drips should not be less than 2 inches, and in some districts a minimum height of 2 J inches is adopted. Of course very deep drips add to the width of tapering gutters. Fig. 15 gives a plan of a gutter which is located between two pitched slated roofs ; its total length betwen the two walls is 28 feet, and this allows it to be divided into three lengths, each 9 feet between drips, and with a drip-box U- -J FIG. 16. Drawing for ascertaining width of gutter bottom at any point. 1 foot in length. Assuming that each roof has a pitch of 45, the narrow end of gutter 8 inches wide, the drips 2 inches deep, and that each 9 feet length has a fall of 1 inch, then under these conditions the width of the gutter bottom, Fig. 15, at the wide end would work out at 1 ft. 1 inches ; and the widths of the gutter bottom at the top of the intermediate drips would be 1 ft. 2 in. and 1 ft. 8 in. respectively. Supposing, now, that each roof in Fig. 15 had had a pitch of 30 instead of 45, the remaining particulars as before the widest end of the gutter bottom would have been 2 ft. 8J in., and the widths at the top of the intermediate drips about 1 ft. 6J in. and 2 ft. 4| in. These values ROOF WORK 47 clearly indicate the influence that slow pitched roofs have on the widths of tapering gutters. A common method of ascertaining the width of any part of a gutter prior to its formation is indicated by Fig. 16, which shows the widths at different points of the gutter under consideration. In practice the sketches are made to a large scale, or, better still, full sized, so that the dimensions can be easily measured off with an ordinary rule. When a gutter gets very wide, the width of the lead to cover it can be broken by introducing one or more rolls. A section through ab, Fig. 15, is given in Fig. 17. The distance the sheet lead should turn over the fillets in order FIG. 17. Cross section of lead gutter between two pitched roofs. to avoid water following back under the slates and getting behind the leadwork, chiefly depends upon the pitch of the roof. For quick pitched roofs it is only necessary to turn the lead about 1 inch over the fillet, as on the left side of Fig. 17. Slow pitched roofs, on the other hand, should have the lead carried right over the fillet, and about 3 inches up the roof, as shown on the right side of Fig. 15. Drips. There are two principal types of drips, viz. : the square drip and the splayed drip, and these are shown in Fig. 18. The lead is a little easier to work down when the drip is splayed, and naturally some plumbers prefer it. Either form of drip is satisfactory when deep enough, and where the lead is properly worked over it. 48 DOMESTIC SANITARY ENGINEERING AND PLUMBING The chief advantage of the square drip is that the lead is more rigidly held in position than in the splayed form, but many plumbers make a poor job of square drips by getting the overcloaks raised above the wood bearing. Drips are easily worked down if the upstand lead, or that which lies on the roof surface, is first bent down and inwards ; this enables the end of the gutter to be bent down to take the form of the drip, when the lead can be readily bossed to take its correct form. Sharp " chase " or " set " wedges should not be used when forming drips, as these unnecessarily reduce the lead where strength is most required. At A and B, Fig. 18, the overcloaks of the drips are shown finished off on the sole of the lower lengths of gutter; this form of finish is desirable for drips which do not exceed 2 inches in depth, as it strengthens the lead at the angle and keeps it in position. Occasionally the overlap of drips is cut off about J inch above the lower angle, as at C, Fig. 18. The object of treating the drips in the latter way is intended bo prevent water rising by capillary attraction between their under and overlaps, and gaining access to the woodwork beneath. This source of leakage, however, is often greatly exaggerated in gutters, and it is very doubtful if the trimming of 2 inch drips, as at C, Fig. 18, serves in a small degree the purpose it is intended. The height to which water will rise by capillary attraction between two surfaces depends upon their distance apart. After gutters have been laid for a short time the surfaces at the drips, which were originally dressed close together, get a little apart, and the separation of the surfaces by natural agencies prevents capillary attraction from taking place in drips of moderate depth. Even in drips which have a depth of 1-J- inches and less, and where the overlaps are cut clear of their lower angles, the risk of leakage due to driving rain is often much greater than that which is likely to occur from capillarity. Capillary grooves are often shown and suggested for small drips, but such grooves could only be cut in fairly deep drips, and in such cases their adoption would serve no useful purpose. For drips over 2 inches deep ROOF WORK 49 they are often formed as at G, Fig. 18, where the under lap is carried to the top of the drip and not over it as in A Vertical drip with overcloak finished on sole of lower length. Splayed drip with overcloak finished on sole of lower length. Vertical drip with overcloak finished clear of lower angle. FIG. 18. and B of the same Fig. Where deep drips are wide, and there is danger of the overlaps being displaced, they can be treated in a similar manner to that shown at C, Fig. 10. Box Gutters. An ordinary form of box gutter is given 4 50 DOMESTIC SANITARY ENGINEERING AND PLUMBING ill Fig. 19, where the bottom for the whole length of the gutter is of uniform width ; the method of treating the leadwork requires no special comment, as the work is practically the same as in the gutter already described. The upstand lead y against the wall is generally about 6 inches, but in some cases it may be an inch less than that. The depth x, Fig. 19, will necessarily vary at different points, but that can be ascertained by means of a sketch which allows for the slope of the gutter and depth of drips. At the higher end of a box gutter the depth x, Fig. 19, should not be less than 2 inches. FIG. 19. Box gutter. Valley Gutters. Fig. 20 shows a valley gutter or flank, and the flat surfaces from the fillets to the angle need not exceed 4 inches in width unless the roofs have a very quick pitch. The width between the edges of the slates should be ample, to enable a man to get his feet into the gutter when climbing up it. Where the lengths over- lap each other about 4 inches should be allowed for the joint. Drip-boxes or Cesspools should not, as a rule, be less than 6 inches deep, and where practicable they should be arranged with an open end to discharge into a hopper or rain- water head. When formed in this way choked outlets are ROOF WORK 51 avoided, but many cases occur where such drip-boxes cannot be used. On all buildings where the box .type of drip-box is necessary the latter should be provided with an overflow in case the outlet pipe gets choked. There are two principal ways of arranging the outlet pipes from drip-boxes, as illus- trated by A and B, Fig. 21. In either case the outlet pipe requires soldering to the drip-box, otherwise an overflow pipe would be no safeguard to prevent leakage should a stoppage occur at any time in the outlet pipe. If a drip-box is not very deep, the overflow pipe will require to be inserted in a flattened form to enable the FIG. 20. Valley gutter, or flank. whole orifice to be below the gutter end. A simple form of overflow is given at B, Fig. 21, where the end is shown finished flush with the wall. The question is occasionally raised in connection with drip-boxes, whether those that have their angles bossed are not superior to those that are made with soldered joints ? From a practical standpoint, provided that each is well made, one is just as good as the other so far as durability is con- cerned ; generally speaking, however, unless a plumber is a very good lead-worker he makes a better job by soldering deep angles than by bossing them up. The making of drip- boxes in practice is chiefly regulated by their height and shape, the method involving the least labour being the one 52 DOMESTIC SANITARY ENGINEERING AND PLUMBING usually adopted. For example, if a drip- box like Fig. 22 is required, it is much easier to cut it out of a piece of lead and solder the corners than to boss it up. The joints could be burnt in lieu of soldering if desired. The size of drip-boxes is, of course, regulated by circum- stances, but it is desirable that they are not made unduly large or there may be difficulty in getting them into position. Flashings. To render roofs water-tight by the side of walls or other structures, lead flashings are usually employed. These flashings take various forms, and different styles of FIG. 21. Outlets from drip-boxes. work are adopted for similar purposes in different parts of the country. Plain flashings take the following forms : 1. Soakers and cap flashings. 2. Cover flashings. 3. Secret gutters. 4. Exposed gutter flashings. Soakers and cap flashings make undoubtedly the best work, so far as weathering and durability are concerned, but a large part of the work requires to be done on the roof after it is slated or tiled. Cover flashings are not suitable for exposed situations, nor for gable walls and similar structures which make greater angles than 90 with, and above, the courses of the slates. These flashings, however, are satisfactory for general work, ROOF WORK 53 View of drip-box. UNDER LAP UNDER LAP. ' Development of lead for lining drip-box shown above. FIG. 22. 54 DOMESTIC SANITARY ENGINEERING AND PLUMBING provided the situations in which they are fixed are sheltered from high winds. Cover flashings are easily fixed, but the greater part of the work requires to be done when the slates are in position. Secret gutters possess the advantage of allowing the lead- work to be done before the slating, but they are liable to give trouble by getting choked with moss, leaves, and other debris. Exposed gutter flashings are similar to the last, but they are wider, and are not liable to be choked like Secret gutters. FIG. 23. Soakers along with continuous step flashing. The form of roof construction, it will be found, regulates to a great extent the use of any particular class of flashing. In England, for example, it is the general custom (excepting the better class structures) to simply nail the slates to battens or laths, which are in turn nailed across the rafters of a roof. This kind of roof construction, unless specially prepared, does not lend itself to either form of gutter flashing, but is better suited for cover flashings and soakers, which are largely used throughout England. In Scotland roofs are generally boarded, and the slates are nailed to the boards. Such roofs are therefore well adapted for gutter flashings, and this is probably the chief ROOF WORK 55 reason why the exposed form v of gutter flashing is so largely used in North Britain. In roof work it is a decided advantage to be able to fix nearly the whole of the lead before a slate or tile is in position. A small portion of a gable wall and roof, Fig. 23, shows how soakers and step flashings are commonly arranged. A soaker is provided for each row of slates, and the lap allowed is the same as that for the slates. As a rule soakers do Fro. 24. Method of setting out continuous step flashings. not require to exceed 7 inches in width, irrespective of the size of slate ; that allows an upstand of 3 inches and 4 inches under the slates. The length of soakers can be found by adding the lap to the full length of slate, then dividing by 2, and by adding to the result -J- an inch. Thus if slates are used which are 20 inches long, and are laid with a 20 -f- 3r-\ -)+ 2 inches. Step Flashings may be made in single steps or they may 56 DOMESTIC SANITARY ENGINEERING AND PLUMBING take the continuous form, the latter being represented in Fig. 23. When setting out continuous step flashings it is essential that the lead from which they are cut be wide enough to allow the free edges of the steps to slope well backwards. If the lead is narrow, the free edges of the steps may be nearly vertical, and rain may be driven in behind the lead and leakage may result. The lead to form the steps and to overlap the soakers, as in Fig. 23, should not be less than 6 inches in width. To mark where the Fig. 25. Cover flashing. turnings for brick joints come, the strip of lead should be fixed temporarily in position, when a short straight edge can be laid along the joints and marks continued from the latter across the lead. After the positions for the turnings have been obtained, an inch or thereabout is allowed on each step for going into the joint, and the surplus lead cut out. Fig. 24 shows how a length of step flashing is prepared, and how it appears prior to turning the allowance at the top of the steps. For exposed situations, the strength of lead for step flashings should ijot be Jess than 5 Ib. per square foot. ROOF WORK 57 The thickness of the lead for soakers should be regulated by the character of the slating, so as not to unduly tilt the slates. Generally speaking the strength should not exceed 5 Ib. per square foot. Cover Flashings. In Fig. 25 a piece of cover flashing is given, and, as its name implies, the lead is simply secured on the top of the slates, and the whole width of the flashing is usually in one piece. The lead which covers the slates is FIG. 26. Secret flashing. usually about 5 inches in width, and the upstand in which the steps are formed should not be less than' 6 inches. Cover flashings should be fixed in comparatively short lengths. To prevent water following between the flashing and the slates, and leaking into the roof, the slates are tilted from the wall by means of a wood fillet, as in Fig. 25. The lead on the slates is held in position by copper or lead ties, which are obliquely fixed, and clipped round the free edge in the manner shown. The ties should not be OF THE UNIVERSITY 58 DOMESTIC SANITARY ENGINEERING AND PLUMBING less than 2 inches wide, and fixed from 24 to 30 inches apart. Lead ties are not so satisfactory as copper ones, and where the former are used they should be cut from lead weighing not less than 6 Ib. per square foot. Secret Gutters. The secret gutter flashing, Fig. 26, has a separate cap flashing, the upper edge of the latter being fixed in a groove which is cut in the stonework. Secret gutters are made of slightly varying widths, but, as the term implies, no lead on the roof is intended to be in view. Their general forma- tion is shown in Fig. 26, and the width of lead on the roof FIG. 27. Gutter flashing. surface may vary from 1J to about 2 inches. The slates cover the channel to within half an inch of the upstand lead. By means of a fillet the outer edges of the slates are raised a little, and this prevents the water following round their edges and getting beneath them. FIG. 28. Gutter flashing with roll. Another means to prevent leakage when secret gutters are used is to double the edge of the lead back on the fillet, as at A, Fig. 26. The latter mode of treatment, however, seldom serves any useful purpose, as the slater usually flattens down the free edge on the fillet in order that the slates may rest firmly on the woodwork. Secret gutters, as ROOF WORK 59 previously stated, are readily choked, and for this reason there are many situations in which they should not be used. Exposed Gutter Flashing. This is illustrated by Fig. 27, and it will be observed that it resembles to a great extent the secret gutter. In Fig. 27, however, the space between the fillet and the wall is usually about 6 inches, and in this case it is not liable to stoppage like a secret gutter as it can be flushed out with a heavy shower of rain. When a moderate volume of water is delivered from a higher to a lower roof, or FIG. 29. Arrangement of leadwork in connection with a plain dormer. where a large volume of water is likely to flow down an exposed gutter, the latter is usually modified to take the form given in Fig. 28, where the roll prevents the water rushing over the fillet and under the slates. Dormers. A method of rendering a dormer in a slated roof water-tight is illustrated in Fig. 29. The apron which lies on the slates at the front of the dormer is first fixed in position, and it should be continued over the framework, with the upper part of the structure erected upon it. Unless the apron flashing is treated in this manner there is difficulty in making a dormer weather properly along the front. Although 60 DOMESTIC SANITARY ENGINEERING AND PLUMBING the exposed gutter flashing is shown at the sides of the dormer, either soakers or cover flashings may be used instead. The cheeks or sides are each covered with two pieces of lead, which are joined by vertical " clinked " or " welted " seams. Of course the number of pieces to cover one side would be FIG. 30. Details of leadwork for dormers. regulated by the size of the dormer. Where the top is com- paratively small, as in Fig. 29, only one piece of lead] is necessary to cover it, but when large two or more pieces may be essential. The top edges at the front and sides of the dormer are shown with a simple finish, but this part may be made ornamental by introducing mouldings. In ROOF WORK 61 either case the leadwork requires to be well secured along the top edges to prevent its being blown up by high winds. Enlarged details, Fig. 30, show more clearly how the leadwork in connection with dormers is secured. The apron, or barge flashing, A, it will be noted, after passing over the top of the framework is turned upwards inside. In lantern lights the apron flashings are treated in the same manner, but instead of dressing the lead close to the framework a small cavity is sometimes left between the lead and wood- work, in order to receive and to discharge any water of condensation. B shows how the cheeks and top may be FIG. 31. Box-gutter between wall and glass roof. treated, the lead being supported along the bottom edge with suitable ties. At C a vertical clink or welt is given for joining the pieces of lead which cover the cheeks. D shows how the lead is arranged at the top of the dormer; the reason for doubling the free edge of the lead under along the front edge is for strengthening purposes. The lead should also be supported and secured by ties being inserted in the " clinked " or welted seams. The strength of lead for covering dormers or similar structures should be approximately as follows : Tops 6 or 7 Ib. per square foot. Sides and aprons, 5 or 6-lb. lead. Exposed gutter flashings, 5 or 6-lb. lead. Apron flashings should overlap the slates by about 5 inches. DOMESTIC SANITARY ENGINEERING AND PLUMBING Glass Roofs and Skylights. When a glass roof drains into a box gutter, as in Fig. 31, it will be found easier and quicker to lay the gutters, if separate flashings are fixed beneath the ends of the glazing bars. Where the gutter and flashing are in one width there may be much difficulty in getting them into position. If separate flashings are used the gutters can be laid in the ordinary way, and it is an easy matter to slide the flashings under the ends of the glazing bars, which have been slotted to receive them. The side flashings for skylights may be arranged in different ways, depending upon the circumstances of the case. If soakers are used for the sides of a fast light, which stands 4 inches or so clear of the roof, separate cap flashings are generally used for covering the top edges of the soakers, and to turn over on the upper surface of the woodwork. Where a skylight is only raised a little above the roof, separate cap flashings can- not be used, and under these condi- tions it is customary to make the soakers with a higher upstand, and to dress the latter over the upper edge of the light. Gutter or cover flashings may be used, of course, in lieu of soakers, but the principle is the same irrespective of the particular kind of flashing which may be adopted. Cornices. Both stone and wood cornices are often covered with lead in order to preserve them from disinte- gration and decay. For covering cornices similar to Fig. 32 the lead is usually cut in lengths equal to the width of the sheet, and the pieces are joined together with clinked or welted seams. FIG, 32. Lead covering for stone cornice. ROOF WORK 63 On stone cornices the lead is easily secured in position by means of lead dowels, which are introduced at intervals varying from 2 to 4 feet according to the width of the cornice. The dowels may be formed by cutting dovetailed holes in the stonework prior to the fixing of the lead ; when the latter is in position, small holes are bored into it im- mediately over those in the stonework, and the lead shaved around the holes and their edges turned upwards, as in FIG. 33. Treatment of joint of leadwork in connection with channel cornice. Fig. 32. A brass or iron mould is afterwards placed over the prepared holes, which are run full of molten lead. Another method of securing lead on stone cornices, is to make holes in the latter as above described and to run them full of lead. After the cornice is covered, tinned-headed screws are driven through the sheet lead into the dowels or plugs, and soldered dots wiped over them. When cornices, such as Fig. 32, are of wood, the best method of securing the leadwork is by means of copper ties, which are fixed in the clinked or welted joints. It is a good plan to dress the lead square over the edge 64 DOMESTIC SANITARY ENGINEERING AND PLUMBING of a cornice, and to trim it off so that it may hang free of the stone or woodwork. By doing this an even edge is preserved, and water is prevented from trickling down and disfiguring the moulding of the cornice. For covering a channel cornice, Fig. 33, the lead is laid in as long lengths as possible, and the joints either soldered or made by burning them. As these cornices are fixed level, FIG. 34. Method of covering pitched stone copings with lead. the only fall available is that obtained by cutting the channel deeper at the outlet than at any other part. The joints in the channel should be made flush on account of the small amount of fall obtainable. It is not necessary to solder the joint right across on a channel cornice, but a clinked or welted seam can be made for all except the channel. Pitched Stone Copings. Sometimes it is essential to cover pitched copings with lead, but the general method of securing the lead work with dowels or dots is not satisfactory in this ROOF WORK G5 case. Dowels or dots form rigid fixing, and although they answer for cornices more freedom is necessary for supporting heavy pieces of lead on comparatively quick pitched surfaces. A good plan of securing leadwork on pitched copings is indicated in Fig. 34. The length of each piece can be regulated to a great extent by the pitch of the coping, but in any case the pieces should not be very long. It will be observed, upon reference to Fig. 34, that the top end of each piece of lead is turned into a groove in the stonework, and in the same groove copper ties are fixed for supporting the next higher piece. Thus each piece of lead is well secured along FIG. 35. Method of securing lead on ridges with copper ties. the top and bottom edges, and yet sufficiently free to allow a little longitudinal and lateral movement. Hips and Ridges. Eolls for ridges and hips require to be raised above the slates that the lead may be able to grip them without pressing upon the edges of the slates. An old but very good method of arranging these rolls is that of fixing them on spikes, which are driven into the ridge and hip timbers. Another method is shown in Fig. 35, where the rafters are arranged to come below the top of the ridge piece, and where the edges are champered that the top of the ridge piece may be no wider than the bottom of the roll. 5 66 DOMESTIC SANITARY ENGINEERING AND PLUMBING For general work, ridge and hip rolls are similar in size to those for flats, but larger ones are used for special work. The lead wings which overlap the slates should not be less than 5 inches in width, and where ornamental wings are desired greater widths will be necessary. Fixings for Hips and Ridges. The lead on ridge and hip rolls requires to be specially well secured, or it is liable to be dislodged by high winds or to " creep " down the hips by its own weight. Copper or lead ties are frequently used, as in Fig. 35, so as to grip the lead on each wing and to keep it FIG. 36. Method of securing lead on hips or peends. tight on the roll. The ties are fixed prior to the wood roll, and spaced about 2 feet apart. On the right hip in Fig. 3 6 the lead is fastened by copper or galvanised wrought iron bands ; they take the form of the roll, and turn under the edge on each side of the wings. The bands are screwed or railed down on the rolls. Secret fixings, which consist of small pieces of lead about 4 inches long, are sometimes soldered on the under side of the lead which covers the rolls ; these fixings are dressed round the rolls and nailed down, whilst the lead for covering the rolls remains boxed up in the usual way. For good work, secret fastenings should be sunk flush with the top of the ROOF WORK G7 rolls. They are very serviceable for securing the lead on quick pitched hips where a neat form of fixing is required. Lead-headed nails, along with the use of copper ties, make simple and cheap forms of fixings, and are effective for holding the lead in its place. Bossing the lead round the ends of rolls is a practice occasionally adopted when covering large rolls on very quick pitched roofs. The rolls should be in about 5 or 6 feet lengths, and after the bottom length is fixed it is covered with lead ; the lead is left longer than the wood roll, and is then bossed over its upper end to afford a means of support. After the first length is finished another is placed in position, and covered in a similar manner to the first. This pro- cedure is followed until the hip is complete. The method of securing lead on ridges and hips is chiefly regulated by their situation and the shape of the rolls. It is clearly obvious that the same amount of fastening will not be necessary in sheltered places as in situations which are exposed to high winds. Kolls when not of a good shape may require additional fixings on that account. Kidge rolls, like Fig. 35, when covered with 5 to 7-lb. lead admit of the latter being well secured with ties, but these fixings are liable to failure unless the rolls are suitably shaped and the ties freely used. Ordinary ties, as in Fig. 35, are insufficient fixings for hip rolls, and additional fastenings should be used. Galvanised iron or copper bands when fixed on rolls do not present a good appearance, but they have the merit of being easily fixed and are effective so far as securing the lead is concerned. Secret fixings are very suitable where strong, neat work is required, but their use makes the lead more costly to lay on account of the extra labour they involve. Where two lengths of lead are joined on ridges they should overlap each other by about 4 inches ; on ridges the overlap should not be less than 3 inches. Ornamental Ridging. On many public and other buildings ridges are frequently of an ornamental character, and some particular design is either cut or formed in the wings of the 68 DOMESTIC SANITARY ENGINEERING AND PLUMBING leadwork. In these cases the rolls are much larger than those generally employed, so as to give them a prominent appearance when viewed from a distance. As the lead for ornamental ridgings requires to be wide it is often necessary to fix it in two or more widths, and afterwards to clink or welt them together. When various designs require to be bossed on the wings of the leadwork, the wings should be separately fixed to permit of the work being more readily accomplished, as suitable blocks can often be screwed down to a bench and the lead worked down over them. Torus Rolls. In Fig. 37 are shown three methods of arranging the flashings in connection with a mansard roof or similar structure. When rolls are used at the break, the lead to cover them requires to be well secured, or it will work loose upon the rolls and probably be displaced by gales. There are simpler methods than those given for fixing lead on torus rolls, but they are not so effective. A fillet is provided at the eave of the higher portion of the roof, as it is imperative that the slates be firmly laid if they are to be prevented from dislodgment in boisterous weather. At A, Fig. 37, the apron flashing is first fixed in position ; then the lead which covers the roll and is turned on the upper roof is secured along the top of the first flashing ; the wood roll is afterwards fixed, and the last piece of lead is turned upwards and over the roll, the lead being made secure by nailing it along its top edge. This method of covering a torus roll will not be suitable for every case, and unless. the lead is fixed in comparatively short lengths there will be trouble in getting it tight on the roll. A different but good method of covering torus rolls is indicated by B, Fig. 37, but it takes a little more lead than that given at A. The apron flashing, as before, is placed first in position, and on that a narrower strip is fixed for the whole length of the roll. The wood roll is then secured in position, and is covered by another piece of lead which is fixed on the higher roof. The strip of lead immediately at the back of the roll is turned forward under it, and with this strip and the lead which covers the upper part of the ROOF WORK 69 roll a clinked or welted joint is formed. To obviate a large bulge being formed along the bottom of the roll, the strip of FIG. 37. Treatment of "torus rolls" or "bottles" when in exposed situations. lead immediately behind the roll is usually thinner than the remaining part of the lead. 70 DOMESTIC SANITARY ENGINEERING AND PLUMBING When rolls are omitted, a simple but effective method of fixing the lead flashings is that illustrated at 0, Fig. 37. Turret Roofs. The leadwork in connection with turret roofs is often of an intricate nature, and especially when the turrets take a very ornamental form. It is advisable to keep the pieces of lead as small as practicable, as they can be the more readily fixed and secured than large pieces. Fig. 38 shows a ventilating turret which is located at the top of a span roof. At the base it is square on plan, whilst the louvred portion and roof take a hexagonal form. For covering the lower and plain part of the structure, the lead may be arranged in vertical bays as shown, the pieces being joined with clinked or welted joints. Copper ties would be necessary both along the bottom edge of each bay and in the seams for securing the leadwork. Before the bays are formed the apron A and flashings G- would of course be placed in position. If a bolder joint for the vertical bays is required, hollow rolls can be made in lieu of the flat seam. To fix lead on vertical surfaces, as at B, Fig. 38, hollow rolls are superior to solid ones. If, on the other hand, the rolls were diagonally arranged, solid rolls would be preferable, and the work could be executed much more quickly thair if hollow rolls were adopted. The lead flashings between the bays B and the cornice C may be covered in a number of pieces as shown. Each section of the cornice C should be covered separately, and the lead well set in to take the correct shape of the moulding, and trimmed off as in Fig. 38. For joining two pieces of lead on the cornice, clinked or welted seams may be used, the latter being formed on one side a little removed from the mitre. To cover a turret roof, such as that in Fig. 38, either solid or hollow rolls may be adopted, the latter being preferable for the shape given. In the Fig. it will be observed that the rolls are nearly vertical for a portion of their length. If solid rolls are used in such cases it is desirable that each bay be supported with secret ties in addition to the usual fixings on the vertical parts. The lead covering the cornice C is continued up the roof for a distance ROOF WORK 71 FIG. 38. Method of covering turret with sheet lead. of say 5 iDches, so that the rolls and bottom edge of the roof bays may be trimmed clear of the cornice. The upper 72 DOMESTIC SANITARY ENGINEERING AND PLUMBING end of the rolls are capped with the lead which covers the base of the finial. On large turret roofs intermediate rolls would be required, and many of the bays may require to be covered with two or more pieces of lead. Shape of Bays. In practice, the shape of the lead for covering a bay on a turret roof is not a very difficult matter to obtain. A simple method of determining the true shape is as follows : First strike a line from top to bottom down the centre of the bay. At right angles to the centre line set off a number of other lines on the whole length of the bay, at convenient distances apart. The closer the parallel lines are together the better, but 6 to 12 inches apart, depending upon the size and shape of the turret, will, as a rule, be suitable for general work. All the lines should now be measured and their lengths noted. It will be found convenient, if a rough sketch of the bay is made in a pocket- book or on a piece of wood, to show all the lines which have been struck upon the roof, together with their lengths. The centre line over the curved or irregular surface may be measured with a tape. All the lines are reproduced on a suitable piece of lead, and their exact lengths marked off. Through the measured points on the parallel lines freehand lines may be drawn, when the shape of the bay will be obtained. The allowance of lead for the rolls or other laps is afterwards added. When the shape of one bay has been produced, it is used as a template for the remaining bays ; corrections, however, may be necessary, as the dimensions of similar parts of different bays often vary slightly. Frequently the shape of half a bay is obtained in the manner above described, and cut out in either zinc or thin sheet lead ; this is simply used as a template for the other bays, and admits of any little alteration being readily made. As before, the allowance for laps and rolls is afterwards added. When the shape of a bay for a turret or similar roof requires to be produced from drawings, the exact dimensions may not admit of direct measurement, see Fig. 39. To ROOF WORK 73 determine the shape of a bay in this case, first draw to scale a part section and plan of the structure, as A and B, Fig. 39. Divide the curved line of section A into any given number of parts, making them equal as far as pos- sible for the sake of simplicity ; so far as exactness is concerned the more parts into which the line is divided the better. Next set off a line in plan B at right angles to the centre of xy, so as to divide the plan of one bay into two equal parts. From the numbers on the curved line of section A, drop per- pendiculars to pass through the plan as in B, and number as shown. Next draw a line mn as IF * g JL -f-f pg -tl n FIG. 39. Method of obtaining true shape of lead work for a bay from drawings. 74 DOMESTIC SANITARY ENGINEERING AND PLUMBING in C, Fig. 39, and at right angles to it draw a number of horizontal lines exactly the same distance apart as the divisions on the curved line in section A, and number in like order. The widths for the different parts of the bay are obtained from the plan B by taking the distances between FIG. 40. Details of lead for turret roofs. the hips and reproducing them on the lines with like numbers in C. Join the points obtained by freehand and the correct shape or development will be obtained. The allowance for rolls is then added and is represented by the dotted lines in C. The finial at the apex of Fig. 38 may be covered with three pieces of lead : the first piece, L, caps the rolls and ROOF WORK 75 covers the base of the finial ; the second piece covers the ball K, and the third the top part H. Details of Turret Roofs. Fig. 40 gives a few details in connection with these roofs. At A is illustrated how a plain cornice may be covered ; B indicates how secret ties FIG. 41. Method of securing leadwork on vertical or quick pitched surfaces with solid rolls. may be utilised for securing rather wide pieces of lead, and for keeping the leadwork close to the boarding where the lead tends to draw or fall away. The ties are soldered on the under side of the lead, and they can be either screwed to the face of the woodwork or passed through a slot and secured inside the structure, as shown in B, Fig. 40 ; the 76 DOMESTIC SANITARY ENGINEERING AND PLUMBING latter method makes substantial fixings when it is practicable to adopt it. A cross section of a hollow roll and copper tie is given at C, Fig. 40. Domes. The lead work for covering domes is arranged FIG. 42. Finial covered with three pieces of lead. and set out in a similar manner to that given for turret roofs. Solid rolls are well suited for hemispherical domes, as the bays gradually increase in width for their whole length ; if the work is properly carried out it is scarcely possible for the bays to slip, on account of the grip the under and overcloaks have upon the rolls. Wood cores should be well undercut to give the lead a firm hold, and secret ROOF WORK 77 fastenings may also be necessary in certain cases in the widest parts of the bays. Covering Vertical Surfaces. Fig. 41 shows how solid roll work may be carried out for vertical and quick pitched surfaces. Instead of the undercloaks being turned over the rolls they may be left flat, and the wood rolls planted on them and screwed down as shown by the enlarged detail A. When FIG. 43. Method of obtaining shape of lead for covering centre part of Fig. 42. the work is executed in this way the undercloak of each bay is securely held in position, whilst each roll also forms a substantial support for the bay immediately above. The overcloaks require to be made secure by copper ties, and the rolls may often be cut a little short to simplify the work, as in Fig. 41. The ordinary method of treating the undercloaks is neither necessary nor suitable for vertical or very quick pitched surfaces. The bays should be kept as narrow as possible, especially for vertical surfaces. 78 DOMESTIC SANITARY ENGINEERING AND PLUMBING Finials. These take a large variety of forms, and the methods of covering them will depend upon their size and shape. A simple form of finial which is situated at the apex of a conical shaped roof is given in Fig. 42. The flashing which covers the upper ends of the slates, and also that which covers the trunk of the finial, represent frustums of cones. A simple method of obtaining the development of a frustum of a cone sufficiently accurate for practical work is shown in Fig. 43. First draw an exact section of the frustum under consideration, say bcedb, and prolong lines FIG. 44. Shape of lead for covering base of Fig. 42. db and ec until they meet as a. With a as centre and radius ad describe the arc rs ; with the same centre and radius ab describe the upper arc tu. On rs point off the distances dff and el, making each equal to de. From H and I draw lines to a, cutting the arc tu at K and L, when KLIHK will be the development required. Allowance for laps or joints must then be added, as shown by dotted lines. In Fig. 42 the lead which covers the lower frustum is carried up a few inches on the trunk which forms the higher frustum ; as these have different pitches, an allowance must be made, as shown by the curved dotted line in Fig. 44, otherwise the lead forming the lower frustum will not girt the trunk of Fig. 42 where the overlap occurs. After the ROOF WORK 79 lead of the shape Fig. 44 is cut out, it can be bent round the structure, excepting the allowance for the overlap, which will require to be bossed into position, and any surplus lead can afterwards be cut off. After the first two pieces of lead for Fig. 42 are in position, the top may be covered by bossing a circular piece of lead in the { form of a cup, and afterwards finishing it in its place. Finials which are comparatively small in size are occasionally covered with one piece of lead, which is bossed to something like the shape required and afterwards completed in its place. In other cases, where a simple form of finial is to be covered, the approxi- mate development is cut, when the lead is partly bent and partly worked into its place. Care, however, in these cases requires to be taken when setting out the lead, so as to have it sufficiently large at all parts. Where finials are built up in sections, and are held together by iron rods pass- ing through them, the upper parts can be removed, whilst those immediately beneath receive attention. Finial Fig. 45 is shown with a circular base, and is covered with five pieces of lead. For the base A the lead may be roughly bossed to the required shape and finished in position. Should the base A be square on plan, it would be easier to cover this part with four pieces of lead with joints formed at the angles. The part B can be dealt with by bossing the piece of lead dome-shaped, and finishing it off as shown. The small column C may either be covered with a piece of lead pipe or sheet lead suitably formed, whilst the ball D may be covered as before described, and trimmed off about 3 inches down on column C. FIG. 45. Fiuial. 80 DOMESTIC SANITARY ENGINEERING AND PLUMBING For covering the conical part e a piece of pipe may be dressed to the shape required, or it may be covered with sheet lead, the edges of which are joined by burning or by other suitable means. TABLE I. STRENGTHS OF LEAD FOR ROOF WORK Gutters . . ; ; . . 5 to 7 Ib. per sq. foot. Flats . ... . ,.: ,, 6 to 8 General flashings . . . . . 4 to 7 ,, ,, ,, Under flashings or soakers . ; . 3 to 5 ,, ,, ,, Hip soakers . 3 : . . . . 2 to 3 ,, ,, ,, CHAPTEK III PIPE FIXING AND PIPE BENDING THE methods of supporting and fixing pipes are chiefly determined by the metal of which they are made, the purposes for which they are to be used, the situation in which they are to be placed, the kind of structure to which they require to be fixed, and the size of the pipes. The old method of burying small pipes in the plaster- work of walls is fast becoming obsolete on account of the inaccessibility of the pipes and the damage done when repairs require to be effected. Most of the important pipes in modern buildings are fixed either in exposed positions or in accessible situations. Of course special cases do occur in buildings where certain pipes cannot be made readily accessible, but when fixed in such positions extra precautions should be taken to prevent the pipes failing at these points. On plastered walls lead water pipes are frequently fixed to wood grounds, and the pipes can be neatly and readily secured. Where cheap fixings are necessary either malleable or tinned iron clips can be used. Lead clips are not suitable for taking the weight of a pipe, unless they are strongly made and soldered to the pipes. Cast lead lugs, when soldered to the back of pipes, make substantial fixings, and many have a neat appearance. Although lead clips and lugs can be readily procured many plumbers prefer to make them. After a suitable design has been decided upon, a pattern of the lug or clip requires to be made, when a cheap mould may be formed in either plaster or lead. When numerous fixings are required it is better and cheaper in the long run to procure gun-metal or iron moulds. 6 82 DOMESTIC SANITARY ENGINEERING AND PLUMBING SOLDER Plaster moulds are readily chipped, and their life is com- paratively short, but new ones, of course, can be readily made. These moulds require to be slowly and thoroughly dried before use if sound castings are to be produced. Lead moulds with care are fairly durable, but after a number of castings have been made they get rough, when much time is taken in trimming them to make them moderately smooth. Before running molten lead into lead moulds the latter should be well smeared with plumbers' black. Clips or lugs on small vertical lead pipes should be fixed about 2 feet apart, and a little closer on horizontal pipes. Wood Grounds are sometimes fin- ished flush with the plaster-work of walls, but they are better when fixed on the face of the plaster ; in the first case the plaster is liable to crack and to leave the edges of the boards, either by the latter shrinking or by jarring the boards when fixing the pipes. Lead pipes which are horizontally SOLDER arranged, and convey hot water, should be supported on wood fillets. When lead clips are used for securing long lead waste pipes which are alternately heated and cooled, the clips should not be soldered to the pipes, or the latter would be held too rigid. By the use of iron hooks lead pipes are often distorted or bruised, but this can be avoided to a great extent by placing a strip of sheet lead about j inch in width between the pipes and heads of the hooks. Iron hooks as a rule are not satisfactory fixings for lead pipes, but of course they possess the advantage of comparative cheapness. Large Lead Pipes. The fixings for lead soil pipes and rain-water pipes may take the form of cast lead sockets, plain or ornamental lugs or tacks, iron brackets, and lead flanges, etc. In Fig. 46 a cast lead socket is shown ; it is sufficiently large to slip over the pipe, and it is secured by soldering at both ends to the pipe as shown. Sockets make FIG. 46. Lead socket soldered to pipe. PIPE FIXING AND PIPE BENDING 83 good and substantial fixings when fixed at suitable distances apart. Sheet lead tacks, Fig. 47, only support a pipe at the back ; they are usually soldered as shown, to enable a pipe to tit close to a wall, but the soldering may be done at the front of the tacks if desired. When plain tacks are used, they are frequently made sufficiently wide to enable the nail heads to be covered by doubling part of the tack over and dressing it down, as indicated on the right side of Fig. 47. SOLDER SOLDER FIG. 47. Plain lead tacks. FIG. 48. Ornamental lugs or tacks. Ornamental lugs or tacks, Fig. 48, are cast in moulds. Cast tacks make better fixings than those made from sheet lead, as they are thicker and stronger ; they are soldered to the back of the pipes, so that no solder is exposed to view. When tacks are in pairs, as shown in Figs. 47 and 48, they are usually fixed to 5 feet centres on 10 feet lengths of pipe, and to 4 feet centres on 1 2 feet lengths, the latter being the more suitable distance apart. If single tacks are used they should be half the above distances apart, and fixed alternately on both sides of the pipes. Not less than two 84 DOMESTIC SANITARY ENGINEERING AND PLUMBING nails should be used for each tack ; for brick walls the nails would be the same distance apart as the joints, and for stone walls they should not exceed 4 inches apart. The minimum width of lead tacks for large pipes should be 6 inches. Ornamental fixings, as in Fig. 49, may be formed from one piece of sheet lead if desired. In this case, after the piece of lead has been cut to the required width, the astragals may be formed by driving the lead into grooves in an oak block by means of a blunt chase wedge or by other suitable tool. The lugs or ears may then be cut to the required design, when the centre portion may be formed by bend- ing it round a wood mandril or length of iron pipe. Should scroll ears be adopted, as in Fig. 49, these can now be shaped, and the whole afterwards soldered to the pipe down the back and round both astragals. The star or other ornamentation could either be cast, or cut from sheet lead and sweated on with solder. If desired the star may be carved out in the piece of oak on which the astragals were made, and formed by bossing the lead into it. Fixings for square lead pipes are often of an ornamental FIG. 49. Ornamental fixings. character, but as these pipes are chiefly used for discharging rain water from the roofs of buildings, the upper end of each length of pipe is prepared to serve as both joint and fixing. Socketed joints for square lead pipes are sometimes formed as in Fig. 50, where a spiggot piece is inserted and soldered to the pipe. The astragals and tacks are separately put on, and intermediate fixings are arranged in the same manner excepting that the slip joint is absent. Sockets for square lead pipes are much better when PIPE FIXING AND PIPE BENDING 85 separately cast, as they can be made larger than the pipes, and the contractions at the joints, as in Fig. 50, dispensed with. Where large lead pipes are fixed in chases or recesses inside buildings, the simplest means of supporting them* is shown in Fig. 51. The worst feature with regard to chases is that they are seldom large enough, so that the pipes are not so accessible as they should be. Flange fixings, Fig. 51, can often be adopted where pipes pass through floors. Iron and Copper pipes, owing to their rigidity, can be fixed FIG. 50. Slip joint and ornamental fixing. FIG. 51. Flange support for pipe. in a different manner from lead pipes. Small iron pipes when fixed to woodwork may be secured by clips in the ordinary manner, but they do not require to be so closely together as for lead pipes. A convenient form of hanger for fixing a pipe beneath a ceiling is illustrated in Fig. 52. The loop that directly supports the pipe is made in two parts, and bolted together as shown. Another hanger suitable for fixing to a steel girder is given in Fig. 53 ; this has a swivel joint which enables the pipe to be placed in any position relative to the girder. 86 DOMESTIC SANITARY ENGINEERING AND PLUMBING Hangers like Figs. 52 and 53 are made in malleable iron, and are comparatively cheap. Since silver-plated and lacquered copper pipes have come into more general use in good buildings for waste and service pipes in connection with sanitary fittings, these pipes are frequently fixed clear of walls so as to admit of their being readily kept clean. Two neat forms of brass fixings for light copper pipes are given in Fig. 54. That at A is suitable for brick and stone walls, and the one at B for fixing to wood- FIG. 52. Pipe hanger. FIG. 53. Pipe hanger with swivel joint. work. The part which grips the pipe is made in two parts, and is neatly joined at the front and shoulder as shown. A small dowel keeps the movable part of the clip in position at the front, whilst at the shoulder a set screw is used. Cast iron brackets make good fixings for heavy iron pipes where the latter require to be made secure along the face of a wall. Fig. 55 gives a roller bracket for supporting steam or hot- water pipes along a corridor or similar wall. Another roller fixing, Fig. 56, may be suitable for supporting large steam or hot-water pipes when fixed in a PIPE FIXING AND PIPE BENDING 87 subway or passage clear of the walls. In this case the roller is supported on a light steel girder, which is fixed across the opening. Koller fixings may also be readily designed for securing pipes beneath girders, and in other situations, when it is found advantageous to do so. Bending Pipes. The method of forming bends on lead pipes is regulated to a great extent by their size, by the radius of the bends, and by the thickness of the pipes. Bends should be made as easy as practicable, for when FIG. 54. FixiDgs for light copper pipes. sharp they unnecessarily retard the flow of liquids and gases through them ; the metal is also unduly strained during the bending process, and in the case of pipes of large diameter quick bends take longer to make. Springs are largely used for preserving a uniform bore when bending lead pipes, and they answer very well for pipes up to 1 J inches diameter, when the bends are gradually made. Bends can also be made in 2 to 2^ inch lead pipes with the aid of springs, but when these pipes are thin the back of the bends is considerably reduced in substance, and occasionally torn. 88 DOMESTIC SANITARY ENGINEERING AND PLUMBING FIG. 55. Roller bracket. Before a spring is inserted in a lead pipe the latter should first be properly straightened, and all kinks or irregularities removed by driving a bobbin or short mandril through it. When a bend is made on thin pipes the sides bulge out a little, and these should be carefully dressed to send the surplus lead towards the back of the bend. The spring can then be withdrawn by ffa^^^^rW////////Ji twisting to reduce its size and by 3 |t5Sj : ^|^^ gi vin g a P U U at tne same ti me - The twisting of the spring is not always necessary, and a steady pull will usually suffice to withdraw it provided the pipe has been properly prepared and the bend carefully made. Many plumbers are too rough with springs, and very quickly strain them by either twisting them the wrong way prior to withdrawal, or by getting them fast in the pipes. In a thin lead pipe a spring does not prevent a small buckle forming at the throat when a bend is carelessly made ; should an attempt be made to remove such buckle by dressing up the bend with the spring still inside, the inner surface of the pipe will conform with that of the spring ; under such circum- stances the spring is fastened in the pipe, and can only be withdrawn by the application of considerable force. Bends in 2 to 2-J-inch light lead pipe are better made in several stages, and at each step the bends should be brought to their normal dia- meters by drawing or driving bobbins through them. A small dummy may also be used when bending a 2J-inch pipe. FIG. 56. Roller fixing. PIPE FIXING AND PIPE BENDING 89 Slow bends can often be made on light lead pipes of 1-J inches diameter and less by first slightly flattening the sides of the pipes where the bends are to be formed. The effect of bending is to push outwards the flattened parts, when the bends can afterwards be dressed to the required size. Bends may also be easily made on lead pipe up to 2 inches diameter after loading them with fine sand. Flush and waste pipes can be neatly and quickly bent with the aid of sand, especially when the latter is warm. Prior to bending, the pipes must be well filled with sand, and their ends plugged to prevent any sand escaping whilst the bends are being made. Bends made with the aid of sand are preferable to those in which springs are used, as the inner surfaces in the latter case are corru- gated with the springs. Bending Large Pipes. When bending lead pipes whose diameters exceed 2^- inches, dummies are generally used. Large pipes flatten and buckle much more readily than small pipes when bent, and in order to strengthen the back of bends the superfluous lead which gathers at the sides should be driven towards that point. Before bending large pipes a short mandril about 1 foot in length should be driven through them to remove all irregu- larities, and also to straighten them. An ordinary bobbin does not straighten pipes when driven through them, but simply removes the creases or indentations. To make a bend of 90 on a length of 4 -inch lead pipe, not less than five stages of bending should be necessary, and each time the pipe is bent it should be worked to take its true form. A square bend may be made with less than five bending stages, but the lead is not so easily and well dis- tributed when the bend is made with less. At the point where the bend is to be made the pipe may be heated with a Swedish torch, or by other means, to a temperature of about 220 F., but the back of the bend should be kept as cold as practicable. The usual plan when pulling up the pipe is to place the knee at the point where the bend is wanted, a piece of felt or other material being used to protect the knee from the heat. If the bend is near the end of the pipe the necessary leverage can be obtained by the aid of a mandril. 90 DOMESTIC SANITARY ENGINEERING AND PLUMBING As the pipe is bent it will flatten considerably, and from this point either of the two following methods of pro- cedure can be adopted. The first method is to turn the partially made bend on one side, whilst the opposite side is smartly struck with a suitable dresser to drive the lead towards the back of the bend ; the opposite side is treated afterwards in the same manner. The flattening of the bulged- out sides makes more room in the bend for working the dummy, but it possesses the drawback of forming a hard ridge on each side of the throat. In the second method, instead of driving the lead from the sides of the bend the throat is worked up, and after this is done the superfluous lead is then driven from the sides to the back as before described. In the latter method of procedure there is less space when beginning to dummy up the throat, but there is ample space if the bend is not pulled up too much at one time. It possesses one advantage in that hard ridges are less likely to be formed. When dressing the material towards the back of a bend there is a tendency for the latter to straighten out, and this is especially pronounced when the bend is near the end of a pipe. The straightening of the bend, however, should be pre- vented as far as possible so as to avoid unnecessary labour. At each stage after the bend is dummied into shape a bobbin should be passed round it, to remove any irregularities and to make the bend of uniform bore. Bobbins are passed round bends by the application of force either at the front or behind them. To drive a bobbin round a bend a metal weight is caused to strike it ; the weight is made secure in the middle of a strong rope, and one end of the latter is passed through a central hole in the bobbin ; the weight is then drawn forwards and backwards so as to sharply strike the bobbin. Two persons are necessary to operate the weight, as the rope must be kept fairly taut or the weight is liable to knock against and disfigure the bend. When force is applied at the front of a bobbin no weight is used, but the bobbin is either simply jerked along a little at a time, by twisting the rope round a hammer shaft or PIPE FIXING AND PIPE BENDING 91 other tool, so as not to bruise the hand, or it is drawn round a bend by a steady pull, when the necessary power is obtained by the aid of a suitable winch. As regards the shape of a bobbin for drawing through pipes, it should resemble that of a pear in order to offer the minimum resistance against the sides of the pipes. In this case the hole in a bobbin is made in a lateral direction near the front, so as to enable both ends of the rope to be pulled in the same direction when drawing a bobbin through a pipe. Weights for driving bobbins through pipes are often made of plumbers' solder, and these answer fairly well and are easily obtained. Lead, of course, is too soft when used alone for weights. Brass weights may also be obtained, and some have leather washers arranged to protect the sides of the bends from being bruised when in use. Dummies. To form dummies, solder ends are often cast on steel rods and Malacca canes. Canes are suitable for straight dummies up to about 2 feet in length, but for greater lengths, and where the ends require to be bent, steel rods of f inch to J inch diameter are used. The handles of very long dummies should be fairly rigid, and may be a little thicker than the sizes given. On each steel rod two ends are usually cast, one serving as a handle when the other end is in use. Double ends also reduce the number of separate dummies, as the ends may be bent to different pitches and also differ a little in shape. A good general shape for dummies is that of an egg, but this can be modified a little as experience deems necessary. Working Drawings. When making offsets and bends much time would often be saved if full-sized working drawings were made, either on the workshop bench or on the floor. If, for example, a pipe requires to be bent to form an offset with angles as in Fig. 57, a full size and accurate drawing should first be made, and the pipe can then be bent to the drawing. With a chalk line or straight-edge set off the vertical lines AB, whose distance apart is equal to the external diameter of the pipe ; next set off the distance between lines AC equal to the offset required, and make lines CD parallel 92 DOMESTIC SANITARY ENGINEERING AND PLUMBING with those of AB. From F, or other suitable point, set off the angle of 120 with a large protractor and produce FH. Make GL parallel with FH and the lines the same distance apart as AB. Draw curves to complete the offset. The last bend may now be made. From M, or other point, draw MO FIG. 57. Working drawing. so that DM0 makes an angle of 135; draw NP parallel with MO, and complete the curves as before. The pipe can now be bent to the drawing, and when finished the plumber would be sure that it would fit its intended place. By the use of working drawings pipes can be cut to exact lengths, and the ends prepared on the bench ready for jointing in position. PIPE FIXING AND PIPE BENDING 93 Development of Elbow Pipes. Occasionally, in order to test a candidate with regard to his geometrical knowledge, Examiners of Plumbers' Work require a development of an elbow to be made. Suppose, for example, the develop- ment of an elbow similar to Fig. 58 is required. First draw to a suitable scale, or full size if required, a section of the elbow, giving it a pitch of 112J. Immediately over the FIG. 58. Method of obtaining shape of lead to make an elbow pipe. vertical section of the pipe draw a plan, and divide the circumference into any given number of equal parts exceeding six ; the more parts into which the circumference is divided the better the development will be. In the case under consideration the circumference is divided into 12 parts. To do this, first divide the circle into quadrants, and from the points say 1, 4, 7, and 10, with the same radius as that of the circle, describe curves to cut the circumference. Number as shown. From the points obtained drop lines 94 DOMESTIC SANITARY ENGINEERING AND PLUMBING parallel with the sides of the pipe to intersect with XY, Fig. 58. On the horizontal line KL transfer the numbers 1, 2, 3, 4, etc., from the plan ; make them the same distance apart, and drop perpendiculars equal in length to CYE, which represents the longest dimension of the elbow. From the intersections of the vertical lines with XY draw dotted lines parallel with KL across the verticals in MNPOM. Where the dotted lines from XY cut like numbers in MNPOM FIG. 59. Method of obtaining shape of lead for making an elbow of rectangular section. those are points in the development of the angle. It will be observed, upon reference to the plan in Fig. 58, that points 1 and 7 represent opposite sides of the pipes, and therefore the lines 1 and 7 in MNPOM are of equal length. This applies also to other points on the plan, such as 2 and 6, 3 and 5, etc. By connecting the points by freehand the upper curved line in MNPOM is formed, and represents the development of one half the angle. The lower curved line is drawn after plotting the distances from the upper to the lower side of the centre line. The space PIPE FIXING AND PIPE BENDING 95 between the curved lines is that which requires to be cut away in order to form the elbow required. The order of numbering the plan should be governed by the position of the joint, and in Fig. 5 8 the joint is shown to be in the centre of one side. The development of an elbow for a rectangular pipe is given in Fig. 59. Number the plan of pipe as shown, starting from the point where the joint is to be made. On AB set off the sides of the pipe to agree with plan, and number in like order. Make the vertical lines in dehfd, FIG. 60. Pipe bending machine by Ed. Le Bas & Co. Fig. 59, equal to cyl. From point 1 on plan drop a line parallel with the sides of pipe to cut xy in the section. From the three points in xy draw lines parallel with AB across those in dehfd. As before, where the dotted hori- zontal lines from the section cut like numbers in dehfd, these represent points in the development of half the angle. Join the points as shown with straight lines, and plot the distances on the other side of the centre line and complete the development. Bending Copper Pipes. Thin copper pipes when over \ inch diameter cannot readily be bent by hand without 96 DOMESTIC SANITARY ENGINEERING AND PLUMBING flattening a little, unless loading is resorted to. The principal materials for loading pipes are metallic lead, sand, resin, and pitch. Lead is not often adopted as it is more troublesome to use than the other materials named. Sand may be used for slow bends on the smaller sizes of pipes, but the ends of the pipe require to be well plugged to prevent the sand escaping when forming the bends. For making quick bends resin may be used, and it can be readily melted out when the bends have been made. Pitch, either alone or in conjunction with resin, is largely used for bending both large and small copper pipes. It is only necessary when using the two latter materials to load the pipes short distances past the parts where the bends are to be made. Resin and pitch are melted for loading purposes, and a paper or other suitable plug may be used for preventing them flowing past the points required. The bending of light copper pipes may be done in different ways, but where many bends require to be made, bending machines are desirable appliances, and much time is saved by their use. Fig. 60 illustrates a machine which is made in nine different sizes, and can be used for making bends on iron, brass, or copper tubes, up to 3 inches in diameter. The upper bend in the illustration has a flattened appearance, but that is due to defective shading. For making an occasional bend on a copper pipe a hole in a plank will suffice ; the sharp edges should be removed, and after loading the pipe the latter can be passed through the hole, when the bend may be gradually turned. If the pipe should flatten a little during bending, the bend may be rounded up either with a hard dresser or round- faced hammer. Copper pipes are occasionally found to be hard, and are often softened before being bent. To soften copper pipes they are frequently heated to redness and suddenly cooled. Brittleness is frequently due to the pipes not being properly annealed. CHAPTER IV PIPE JOINTS THE joints with which a plumber should be familiar are very numerous, as their construction must necessarily vary with the different materials of which the pipes are formed, the positions in which pipes are placed, and the purposes for which pipes are required. Joints for Lead Pipes. The chief joints for lead pipes are : (a) Straight and branched forms of plumbers' joints ; (b) Block joints ; (c) Flange joints ; (d) Lip joints. Plumbers should be capable of making soldered joints in any situation, for frequently pipes are much distorted in getting them into position if their relative positions are greatly altered whilst the joints are made. All young plumbers would be well advised to avoid the use of plumbing irons and spirit-lamps, etc., when learning to wipe soldered joints on the smaller sizes of pipes. To manage without their aid a plumber must be quick in the manipulation of the solder, and this is essential in order to become skilful in the art of jointing. Plumbing irons, Swedish torches, and spirit-lamps, etc., are, however, very useful appliances when used in their proper place. Preparation of Joints. When preparing the ends of pipes for jointing, this part of the work should be neatly done. The faucet end for an underhand joint should not be opened unduly wide, and the spigot end should be slightly opened after being trimmed in order that no unnecessary retardation be introduced where a joint is made. Each end should then be rasped to a thin edge, and the tarnish or smudge applied. After the shaving is done, the prepared ends are brought together and made secure, 7 98 DOMESTIC SANITARY ENGINEERING AND PLUMBING when the joint is ready for wiping. Fig. 61 gives three forms of plumbers' joints, those at A and B being shown partly in section and partly in elevation. FIG. 61. Joints for lead pipes. For soil and waste pipes their branches should curve in the direction of flow as at C, Fig. 61. The latter form of joint PIPE JOINTS 99 requires to be carefully prepared, or the solder will run through into the pipes during wiping ; or if a thick edge is left either at the front or at the side of the joint it will probably be laid bare during the jointing process. To guard against solder getting into pipes it is a good practice, when preparing large branch joints, to smear with plumbers' black the inner surfaces of the opened part, and also the end of the branch piece for about J inch on both its inner and outer surfaces. This precaution does not impair the general soundness of a branch joint, and no trouble will be caused by solder running through, provided the joint is properly prepared and is not played with too long when applying the solder. Another precaution when preparing the spigot end to prevent solder running through the joint is, first, to neatly fit the branch piece, when a mark is made on the latter around the top edge of the opened part. The branch is then removed, and a groove cut with a saw file on the lower side of the mark. Afterwards the pipes are blacked and shaved, care being taken not to shave out the small groove in the spigot end. When the prepared branch is in position, the thin edges at the top of the opening may be driven into the groove by the aid of a small hammer and blunt iron chisel. Openings for branches should stand up evenly all round. To form a large elliptical opening in a thin lead pipe, a hole may be bored with a large gimlet or other suitable tool half an inch clear of each end of the major axis. A slot may then be formed by cutting out a strip of lead between the gimlet holes, and afterwards opened up with the aid of a heavy bent pin and small hand dummy. Gauges. To prepare joints on straight pipes of uniform size is a simple matter, as the ends of the pipes only require fitting together in order that the length of the shaving can be correctly proportioned on each end. For branch joints gauges are necessary where uniformity of size is to be maintained. Gauges are often made of sheet brass, and may take the form given in Fig. 62. It is an advantage to have two of these gauges, one for small and the other for large 100 DOMESTIC SANITARY ENGINEERING AND PLUMBING branch joints. Down the centre of the gauge, Fig. 62, a number of small holes are made in which one point of a pair of compasses can be inserted for marking off the portion to be shaved. The central point at the bottom edge is useful when making right-angled branch joints, as it can readily be fixed over the centre of the opening so as to make the scribing equal on each side. For curved branch joints the central point is not of much service, as the gauge requires to be fixed a little off the centre according to the angle at which the branch joins the opened pipe. After the true FIG. 62. Gauge for marking branch joints. position of the gauge has been obtained the scribing for the joint is done from each side. Methods of Supporting Joints. The methods of staying joints for wiping purposes are very varied, and are largely governed by the sizes of the pipes, by the positions in which the joints are to be made, by the packings at disposal, and by the intelligence of the individual who is staying them. In new buildings, where bricks and timber are plentiful, these are largely used for fixings on floors and other flat surfaces : where joints are made against walls, fixing points are very useful for staying purposes. The use of strong cord and fixing points will in many cases dispense with the aid of more cumbersome fixings. PIPE JOINTS 101 When staying branch joints for soil and waste pipes, more care is necessary than for thicker lead pipes, for unless the weight of the branch be alsosupported there may be danger of the whole falling down when getting up the wiping heat. An old but excellent method of supporting branches in large lead pipes whilst wiping a joint is illustrated in Fig. 63. After the opened part and spigot end have been prepared, two pieces of wood about a foot in length and 1J inches square are placed over each other inside the pipe ; upon these two wedge pieces are fixed back to back in order to keep the spigot end from protruding too far into the opening. These packings, FIG. 63. Method of supporting pipe whilst jointing. of course, are placed in position before the branch is inserted. The weight of the branch may be principally supported as at S, Fig. 63, or in any other suitable manner. To secure the branch a piece of cord may be passed round the bend, with the ends fastened down to the bench on opposite sides of the branch. Should a branch be a long one, the wood fixings may be left in position whilst it is carried from the bench and fixed in its place. The packings inside the branch may be readily displaced by either a piece of wood or a length of iron tube. In the case of vertical pipes the packings may be displaced by a plumb-bob, but access will be necessary for regaining the packings. The object of having 102 DOMESTIC SANITARY ENGINEERING AND PLUMBING packings for the inside of branch joints in four parts, instead of in three pieces, is to allow them to pass more readily round a bend. Many methods are in use for holding brasswork to lead pipes whilst joints are made. Several forms of metal clamps and fixings may be obtained for this purpose, and some are very useful appliances. The tail-pieces of unions and taps, etc., are not difficult to stay, however, when joining them to lead pipes. When a brass tap is branched into a lead pipe, one method of staying the former whilst a joint is made is to file a groove near to the tail end and to drive the lead into it from around the edge of the opened part. A second method is to form two or three coarse threads with small stocks and dies on the tail end of a tap, and to make the hole in the pipe suitable in size for the tap to be screwed into it. This makes a simple and substantial fixing, especially if the lead round the edge of the opening is worked up and is also driven into the threads. Another method for holding a tap whilst making a joint is to fuse or " burn " the edges of the lead to the brasswork with a well-faced soldering bolt, which is heated to dull redness. The lead unites with the tinning on the brasswork, and this most readily occurs when the brasswork is well heated just, prior to " burning on," as it is termed. The worst feature in connection with this method is the frequent facing of ' the soldering bolt. It, however, securely holds the brasswork when the " burning " is properly done, and is not liable to give way if the joint is wiped in a reasonable time. To " burn on " it is important that the soldering bolt has a good face, otherwise " burning " is indifferently done. For fixing brass tail-pieces on the ends of pipes " burning on " is often very convenient, especially for the smaller sizes of fittings. If well done, " burning on " resembles a narrow copper-bolt joint, excepting that neither strip of lead nor solder is used. Resin, however, is used as a flux. Another convenient method for supporting tail-pieces is by using a number of narrow wood strips. These can be pushed through the tail-pieces into the pipe, and a reliable and simple fixing often obtained. PIPE JOINTS 103 When wiping joints some discretion requires to be exercised with regard to the solder used ; it is clearly obvious that for soil and waste pipes the joints should be lightly wiped, when compared with those for water pipes, which are required to withstand more or less considerable pressure. Many plumbers experience difficulty in making a good shaped light joint of moderate length, although they have no apparent difficulty in making a heavy joint. As a rule, the reason why many are unable to wipe light joints is owing to the use of thin, flabby cloths. Block joint. Lip joint. FIG. 64. Flange joint. To wipe light joints of moderate length a fairly stiff cloth is essential, and many good joint makers, in order to impart the necessary stiffness, insert a thin piece of zinc in the centre of their cloths. Others introduce a thin layer of cardboard for the same purpose. These stiffened cloths are used chiefly for shaping the larger sizes of joints, more pliable ones being used for getting up the heat. Block Joint. For lead soil pipes which are fixed in recesses inside buildings the block joint is very suitable, as it forms both a joint and a good fixing for the pipe. The upper edge of the hole in the wood block should be well 104 DOMESTIC SANITARY ENGINEERING AND PLUMBING rounded off as in A, Fig. 64. A lead collar is placed directly upon the wood support, and the faucet end before being opened should be cut to stand not more than f inch clear of the top of the support. When the end is opened with the tanpin, a space should be left between the lip and lead collar so as to allow the solder to now under the lip. The lead collar would, of course, be shaped, and most of the preparation for soldering done, before a length of pipe is placed in position. Flange Joint. At B, Fig. 64, a flange joint is shown; this is inferior to the block joint, although it is made very much in the same manner. In the flange joint the sharp aris only is taken off the woodwork, a lead collar being used as before, with the faucet end flanged over on to the collar. FIG. 65. Method of jointing tin-lined lead pipe. Lip Joint. The lip joint C, Fig. 64, is more suitable for gas pipes than for either water-pipe or waste-pipe work. In certain districts where inferior work is done the lip joint is often vised, on account of its cheapness, for joining brass connections to overflow pipes, to lead traps, and to hot-water service pipes, etc. There are certain times, however, when a wiped joint is no stronger than a lip joint ; in fact the latter may be much the stronger of the two. This may happen with a very short brass tail-piece when joined to a lead pipe with an underhand joint. Under such conditions the joint is principally formed on the pipe, with the result that the end of the pipe, which has been reduced in substance, is nearly laid bare. Joints for Tin-Lined Lead Pipes. When solder is used, the joints for these pipes are not prepared in the same manner as those for lead pipes, as the heat of the solder PIPE JOINTS 105 would melt the tin linings in the neighbourhood of the joints, lay bare the lead to the action of the water, and the latter in its passage through the pipes may be considerably retarded. To avoid destroying the tin linings thin brass ferrules are sometimes used, as in Fig. 65. The ends of the pipes may be slightly expanded by means of a steel pin specially made for the purpose, and the ferrule when inserted should be tight fitting ; the joint is then wiped with plumbers' solder in the usual way. During the process of wiping there is still the liability of destroying a little of the lining beyond the ends of the ferrule, although this may be avoided to a FIG. 66. Branch joint for tin-lined lead pipe. certain extent by the use of longer ferrules than that shown in Fig. 65. A method of preparing a branch joint for tin-lined lead pipes is indicated in Fig. 66. It is, however, troublesome to prepare, and it is necessary to cut in two the pipe into which the branch is to be made and to mitre the ends as shown. Each of the three ends may be slightly expanded as before described, and when blacked and shaved, a brass tee piece requires to be inserted, and a joint wiped over the ends, so as to form a combined branch and underhand joint. The dotted lines in Fig. 66 represent the brass tee piece. Another method of joining tin-lined lead pipes is illustrated in Figs. 67 and 68. In Fig. 67 a simple form of union coupling is shown, and the ends of the pipes are simply 106 DOMESTIC SANITARY ENGINEERING AND PLUMBING flanged over and brought together by means of the screwed cap and ferrule. Fig. 68 gives a special gun-metal fitting, the ends of the pipes being flanged and treated in a similar manner to that shown in Fig. 67. WASHER FIG. 67. Joint for tin-lined lead pipes. The latter is the better method of joining tin -lined lead pipes, but the special fittings make such pipes expensive. Burnt Joints for Lead Pipes. Where lead pipes are used for conveying acids, burnt joints require to be made, as ordinary soldered joints are more or less rapidly destroyed. FIG. 68. Branch joint for tin-lined lead pipes. Burnt joints are used to a limited extent for botli lead soil and waste pipes, and several forms of burnt joints for such pipes are given in Fig. 69. At A an ordinary form of lip joint is shown, and the only difference between it and a soldered one is that a strip of lead is used in lieu of solder when making the joint. Where a horizontal pipe may be PIPE JOINTS 107 turned the joint B, Fig. 69, can be adopted. C and D give two different branch joints ; the one at C may be made when the branch piece is in the position shown. D is a much FIG. 69. Forms of burnt joints. stronger type of joint, but to make it the whole branch must be free, so that it can be turned on either side whilst the lead is built up upon the joint. 108 DOMESTIC SANITARY ENGINEERING AND PLUMBING With regard to the merits and demerits of burnt joints for pipes in general work, the latter far outweigh the former. The chief disadvantages of burnt joints for lead pipes are : the time occupied in making the joints ; the limited positions in which they can be made ; and the amount of space necessary, as every part of the joint must be readily accessible when being made. Joints for Copper Pipes. The usual method of jointing copper pipes, which are used either for hot-water services or for heating apparatus, is by means of screwed socketed joints. The coarse or ordinary gas tube thread is frequently used, and the pipes require to be thick enough so as not to be unduly weakened when the threads are cut. Brass or gun-metal fittings are used, and after the joints have been tightly screwed up, the fittings and pipes are sweated together with fine solder. This mode of jointing necessitates the free use of union connections in order that all the ends of the fittings may be soldered to the pipes. Screwed joints for light copper pipes, which only permit of fine threads being cut, are not suitable for either hot-water services or for heating apparatus. Such joints are not durable as a rule, chiefly owing to galvanic action, which corrodes the soldered joints and causes the latter to leak. It may be asked that if screwed and soldered joints on .light copper pipes are liable to fail through galvanic action, why are not those in the first case, where stronger joints are made and soldered ? The reason is not far to seek. With comparatively strong pipes and fittings, the joints can be made water-tight without the use of solder, the latter serving more or less as a mere safeguard against leakage. With thin copper pipes and screwed sockets, even assuming the joints at the outset did not depend for their water-tightness upon the soldering, it is scarcely to be expected that they will remain long in that condition, owing to the varying strains to which the joints are subjected by fluctuations of temperature. As soon as a certain class of water finds its way between the threads of the pipes and those of the fittings, and comes in contact with the solder, the latter is slowly destroyed, owing to a galvanic couple being formed. PIPE JOINTS 109 Stronger joints on copper pipes, when soldered, may be liable to fail through galvanic action provided they were im- perfectly made, or depended for their soundness upon the soldering. Copper tubes of steam pipe strength do not require soldering at the joints, but such tubes are too expensive for low pressure work. Light copper tubes, however, may be safely employed for hot-water services and similar uses, provided suitable joints are adopted. The failure of thin copper pipes, as already explained, is due to the weakness of the joints and not to the thinness of the tubes themselves. Compression-Joints. The most suitable joints for thin FIG. 70. Leighs' compression joint for light copper pipes. copper tubes are those where their ends are brought together and made air and water-tight by compression. Compression- joints have been used by heating engineers on the Continent of Europe for some years, and more recently a good form of compression-joint has been patented by Leighs of Manchester. The construction of the latter joint is clearly shown in Fig. 70. The ends of the pipes to be joined are prepared by expanding one end, whilst the other has a beaded shoulder formed in it. By means of a screwed cap and sleeve piece one end is forced inside the other, and a substantial joint is formed where no jointing material is required, the soundness of the joint relying upon the closeness of the metal surfaces. A washer W is placed behind the raised bead to prevent the latter being cut or chafed when screwing up the joint. 110 DOMESTIC SANITARY ENGINEERING AND PLUMBING For preparing the pipe ends two machines are necessary, one for making the bead and the other for expanding the ends. Special fittings are also required, such as tees, elbows, etc. : these have their ends coned in order to receive the tubes upon which the beaded shoulders are formed. With regard to the soundness and strength of the compression-joint shown, it is stated by the Patentees that a 1^-inch seamless copper tube of 20 I.W.G., when connected with various fittings, proved tight with an internal pressure of 700 Ib. per sq. inch, and that a tensile force of 8J tons was necessary to pull a joint apart. Iron Pipe Joints. Wrought-iron pipes are usually joined with screwed socketed joints, whilst cast - iron pipes are LEATHER. WASHER, FIG. 71. Left and right screwed joint for " Health " water pipe. generally jointed either by flanges which are screwed together with iron bolts or by a spigot and socket form of joint. The method of making the joints, and the jointing materials employed, depend chiefly upon the purpose for which the pipes are required. For screwed joints on wrought- iron pipes a mixture of red and white lead mixed to the consistency of thick paint is frequently used. It is usually necessary, however, in the case of pipes which convey water, air, or steam under pressure, to wrap fine hemp between the threads on the ends of the pipes after painting them. Pipe- joint compounds may also be obtained for making screwed joints air and water tight, and these preparations are usually superior to mixtures of red and white lead. Fig. 7 1 shows a screwed joint for the " Health " water pipe, which consists of a wrought iron tube with a tin lining. In the centre of each socket a space is provided in which a PIPE JOINTS 111 leather washer is loosely placed. Each joint is made with right and left-hand threads, so that each socket forms a union coupling. When the pipe ends are prepared the tin lining is left a little longer than the iron tube, and is afterwards flanged over the ends. No other jointing material is required, for when a socket is screwed up the ends of the pipes are brought to press evenly against the leather washer. Special fittings are necessary in connection with the " health " water pipe, and these are supplied by the manu- facturers of the pipe. The joint used in connection with high pressure heating apparatus is illustrated in .Fig. 72. This is also a left and right screwed joint, but the pitch of the threads is less than that of the " health " water pipe. The sockets are very *. _ . v FIG. 72. Left and right screwed joint for small bore heating apparatus strong, and cavities are provided at their centres where the pipe ends meet. Special screwing tackle is necessary for high pressure joints, as one pipe end is coned to form a chisel edge, whilst the other is prepared with a true flat surface. The ends are brought together with the aid of powerful pipe wrenches, so that the sharp edge of the one cuts into the flat face of the other. No jointing material of any description is used, metallic contact being responsible for the soundness of the joint. For cast-iron pipes the spigot and socket class of joint is the most common. In Fig. 73 the general form of joint for iron waste, soil, and drain pipes is shown. Several strands of tarred-yarn are first inserted in the socket, and the upper 1 1 inches of depth run full of molten lead. To run joints properly when in horizontal situations, clay 112 DOMESTIC SANITARY ENGINEERING AND PLUMBING bands or their equivalent are essential. The bands are placed around the sockets to prevent the lead escaping when running the joints ; after the joints are run, the cooling lead contracts and leaves a small space between it and the surfaces of the pipes. After going once round a joint with a chisel and hammer, to free the edge of the lead from the surface of the pipe, the joint is made sound by well caulking it. Frequently the joints on soil pipes are indifferently made, owing to the rope-yarn being insufficiently staved. The result of this is that the metallic lead is driven into the sockets instead of being finished flush with their top edges. " Lead Wool." Instead of making joints in connection with iron soil and drain pipes with molten lead, " lead wool " may often be utilised with advantage. This material takes FIG. 73. Spigot and socket joint for cast-iron pipes. the form of thin threads of metallic lead, which are formed into strands for inserting into the sockets of pipes. The joints are partly made in the usual way by first intro- ducing tarred-yarn, the " lead wool " being used for the upper inch of the joint, and both the yarn and lead should be well staved when inserted into the sockets. Very substantial joints can be made with " lead wool," for when properly caulked it forms a compact mass of lead, completely filling every space in the depth to which it is inserted. Lead is not a suitable jointing material for iron waste pipes which discharge very hot water, owing to the unequal expansion and contraction of iron and lead. For such pipes iron borings should be used in lieu of lead, and rust joints made. Must joints are only suggested for the discharge of very PIPE JOINTS 113 hot liquids, and not for general use for waste pipes, as lead is a suitable jointing material for the majority of cases. Expansion Joints for Waste Pipes. Where rust joints' are deemed desirable for cast-iron waste pipes, provision must be made for expansion and contraction of the pipes on account of the rigidity of these joints. Bends in pipes allow a certain amount of movement to take place, and a few bends in a stack of pipes may permit of all the movement required. Where, however, a long stretch of waste pipes is rigidly fixed between two given points, and subjected to extremes of temperature, an expan- sion joint similar to Fig. 74 may be necessary to allow a little movement to take place. The construction of the FIG. 74. Expansion joint for cast-iron pipes. joint is clearly shown, the packing material in the gland being asbestos or other suitable material. The joints for iron water mains are much stronger than those for drains and soil pipes, as the former require to withstand more or less considerable internal pressure, and the sockets of water mains are also subjected to greater strain by the extra caulking they receive. Joints in water mains are made by first inserting ordinary spun-yarn, as tarred-yarn is liable to taint the water for a considerable time. The remaining space in the socket is then filled either with molten lead or with " lead wool." The depth of lead required varies with the size of the pipes, with the form in which the lead is used, and with the character of the ground through which the pipes are laid. 114 DOMESTIC SANITARY ENGINEERING AND PLUMBING For water mains of 3 inches to 6 inches diameter, if the joints are run in the ordinary manner, the minimum depth of lead for the smaller size should be 2 inches, and 2J inches deep for the larger size. If " lead wool " is used the depth of lead may be J- inch less in each case. The reason why less " lead wool " may be used than molten lead for making joints is due to the fact that the former can be firmly caulked for the whole of its depth, whilst the latter is chiefly affected by caulking for a limited depth. Turned and lored joints, Fig. 75, possess the advantages of dispensing with the use of yarn, which is subject to decay, and of permitting air-tight joints to be made by smearing the prepared ends with tallow or similar substance. In the FIG. 75. Turned and bored joint for cast-iron pipes. case of pipes which carry little pressure, and where turned and bored joints are used, the angular space may be filled with portland cement or by a bituminous composition in lieu of metallic lead. The chief drawback of turned and bored joints is their rigidity, but this, however, can be largely overcome by introducing in a line of pipes at regular intervals an ordinary spigot and socket joint. Where pipes with turned and bored joints convey liquids under pressure, metallic lead should be used as the jointing material. Spigot and socket joints for hot-water pipes may be made in different ways. A common method of making a joint is to partly fill the annular space between the spigot and socket with tarred yarn, and to fill the remaining space with rust cement. Another method is to first insert a few rings of spun PIPE JOINTS 115 yarn into which a suitable mixture of red and white lead has been worked ; several rings of tarred yarn are afterwards inserted, and the remaining inch to 1 inch of space filled with rust cement. A third method is to first insert a couple of rings of tarred yarn, and to fill the remaining space with a mixture of red and white lead, linseed oil, chopped hemp, and gold size. Packing rings for flange joints, Fig. 76, largely consist of indiarubber, asbestos, metallic lead, and corrugated metal rings. Indiarubber rings are not suitable for steam pipes, but are better suited for fixing between the flanges of cold- water pipes and fittings. Eubber is also fairly durable in contact with hot water which does not exceed 180 F., but when the temperature approaches 212 F., and over, rubber gets hard and readily cracks. Asbestos rings make good packings where flange joints are subjected to moderately high tem- peratures. Where metallic lead rings are used for flange joints, the flanges require to be truly and smoothly faced in order to prevent leakage. Lead rings are very convenient for joints which are periodically taken apart, as they can be re-used after smearing them over with grease or oil. Corrugated metal rings make good jointing material for flange joints on high pressure steam pipes and fittings. The corrugations form a series of concentric rings, and after the metal rings have been painted and covered with red and white lead they are placed between the painted flanges, and the whole securely bolted together. A little fine hemp may be added to the cementing material so as to bind it together. These metallic rings are usually of brass, and by squeezing the cementing material which is confined FIG. 76. Flange joint for iron pipes. 116 DOMESTIC SANITARY ENGINEERING AND PLUMBING in the corrugations against the flanges a substantial joint is made. Expansion Joints and Bends. Movement is provided for in hot-water and steam pipes by means of expansion joints and bends. The expansion joint A, Fig. 77, is made of brass or gun-metal, and is intended for use in connection with pipes with screwed joints. It consists principally of a sleeve piece which slides through a stuffing box in a long socket. Ex- pansion joints, however, are liable to leakage, and for this reason copper expansion bends are preferable for allowing FIG. 77. Expansion joint and expansion bends. movement in pipes. When expansion joints are used, suffi- cient space should be left between the socket of the sleeve piece and the stuffing-box to enable the latter to be repacked when necessary. If fixed as shown at A, there might be difficulty in removing the cap and gland to admit of new packing material being added. Expansion bends may take different forms according to the positions of pipes. B and C, Fig. 77, give two common forms of expansion bends. That at C will withstand greater strains without distortion or fracture than that at B, but its outlet and inlet ends are on different planes. Both ends of PIPE JOINTS 117 bend B are in the same plane, and this form is generally adopted for horizontal pipes which have a limited pitch. Wrought-iron expansion bends are not so suitable as copper ones for allowing movement in pipes, as the former are too rigid, and allow the strain to be chiefly concentrated on the screwed joints. When expansion bends are essential on vertical pipes, they may be bent in spiral form. Jones' expansion joint is largely used for hot-water apparatus in horticultural buildings, and it is also suitable for other places where the head of water upon the pipes does not exceed 20 feet. This joint is for pipes with plain ends, and consists of two loose collars, two rubber rings and an iron /&RASS SOCKET FIG. 78. Connection between outgo of w.c. and lead branch. band, along with two bolts and nuts. The joint is made water-tight by compressing the rubber washers, with the aid of the bolts, between the edges of the loose collar and those of the iron bands. Joints for W.C.'s, etc. There are several forms of patented joints for making connections between the outlets of w.c.'s and lead soil pipe branches, but one of the most common joints is that shown in Fig. 78. To the end of the lead branch pipe a brass socket is soldered, which permits of a good joint being made with the earthenware out-go of a w.c. The brass socket should be slipped over the end of the pipe, and the latter turned over the inside shoulder of the socket, and be finished off as at x, Fig. 78. Either portland cement or an elastic cement may be used as 118 DOMESTIC SANITARY ENGINEERING AND PLUMBING jointing material for the socket, after a ring of yarn has been inserted. The latter is the better jointing material of the two. Where w.c/s are provided with lead traps, or where short pieces of lead pipe are soldered to their outlets, they can be soldered directly to the lead branches. To enable a sound joint to be made between a lead and an iron pipe, a brass sleeve piece or ferrule is first passed over the end of the lead pipes, and soldered to it as in Fig. 79. The purpose of the ferrule is to protect the lead, and to impart sufficient rigidity to enable a caulked joint to be made. LEAD 6RAWCH t FIG. 79. Connection between lead branch and iron junction. When joining a lead soil pipe with an earthenware drain, it is better, on account of the width of the earthen- ware socket, to allow the lead pipe to protrude about half an inch or so into the bend below the edge of the ferrule, as at S, Fig. 80. This keeps the lead pipe in its place and simplifies the making of the joint. Joints for Earthenware Drain Pipes. The common form of joint for earthenware pipes is not one which admits of being readily made in a satisfactory manner. The chief difficulties in connection with it are : Maintaining a true alignment at the invert of the pipes, and in keeping the inside of the pipes smooth, and free from protruding matter PIPE JOINTS 119 at the joints. Portland cement, either neat or mixed with sand, is the usual jointing material. In Fig. 81 a couple of strands of rope-yarn, which have first been steeped in liquid cement, are pressed firmly into the socket, the remaining space being filled with cement and trowelled off as shown. The chief draw- backs associated with yarn are, it is subject to decay, and to leave cavities in the sockets where organic matter may gather, and unless yarn is carefully used it is liable to pass through the joints and cause stoppages in pipes. Where solid cement joints are made, some form of scraper or dis- placer is essential to remove the surplus cement which passes into the pipes during the making of the joints. The removal of cement from the inside of pipes often results in roughening their sur- faces, and in impairing FIG. 80. Method of connecting lead soil- pipe with a drain. the general efficiency of the drain. As any sur- plus cement must be removed immediately after a joint is made, and whilst it is still soft, the tendency is for the cement to run a little and to form a small ridge along the invert at each joint ; there is also difficulty in obtaining a true invert, and frequently the joints form a series of steps unless something is done to prevent it. 120 DOMESTIC SANITARY ENGINEERING AND PLUMBING Patent Joints. To overcome the drawbacks which accompany the use of ordinary spigot and socket joints, patented joints from time to time have been introduced. These joints possess the advantage of allowing the pipes to be easily laid, true alignments of inverts to be obtained, and many patent joints can be made in water-logged ground without in any way affecting the joint- ing material. The chief disadvantage of pipes with patent joints is FIG. 81. Common form of spigot and . . V . , socket joint for earthenware drains. their higher initial COSt. In Hassall's patent joint, Fig. 82, bituminous rings are cast both at the front and back of the socket ; other rings are also cast to coincide with these on the spigot end of the pipes, and the space between the rings is filled with cement in a semi-liquid state. The joint shown in Fig. 82 is known as a double lined one, the rings at x being omitted in the single lined form. Single lined joints admit of the use of cement in the plastic state, but when it is desirable to use liquid cement clay bands are necessary round the sockets. Other forms of patent joints are made in which composition rings are not used, and where true in- verts are obtained by means of studs in the sockets. In other cases special forms of construction are introduced at both the spigot and at the socket ends of pipes. One of the latter type is known as the "Yarrow" joint, and is illustrated in Fig. 83. To prevent the cementing material escaping from the cavity when the joint is being run, a little plastic clay is used at both the front and back of the joint, as in Fig. 83. In Ames and Crosta's joint, Fig. 84, a little clay or FIG. 82. HassaH's patent joint. PIPE JOINTS 121 cement is used at the front and back of the sockets when running in the cement, but the true alignment of the invert is preserved by studs which are formed in the bottom of the socket. Another form of patent joint, Fig. 85, differs again from CLAV FIG. 83. The Yarrow joint. those already shown. In this case the inside of the socket is made sloping, in order that the spigot end, when in position, will be raised to form a true invert. For this joint the cement is intended to be used in the plastic state, and the principal feature of the joint appears to be the provision made for centering it. Ci-Af OR OTHER PLAS-nc MATERIAL FIG. 84. Ames and Crosta's joint. When jointing earthenware pipes, special care is essential in the selection of the jointing material if they are to remain sound for any length of time. The use of portland cement as a jointing material for earthenware pipes has been responsible for many failures on account of the cement expanding and bursting the sockets. 122 DOMESTIC SANITARY ENGINEERING AND PLUMBING This is a trouble difficult to avoid in ordinary practice, for seldom is portland cement tested as regards its suitability, except in large works, where a competent clerk of works is employed. So far as earthenware drains are concerned, even assuming that reliable portland cement is used for the joints, the latter are too rigid, as any slight unevenness in the settlement of the ground results in joints or pipes being fractured. The easy manner in which earthenware pipes are damaged has been responsible for the introduction of iron pipes where reliable drains are required. Elastic Cement. The difficulty associated with earthen- ware pipes can, however, be overcome to a great extent by using a jointing material of a slightly yielding nature. Various materials, such as resin, tallow, bituminous substances, sand, chalk, etc., when mixed in certain pro- portions, may be used for producing yielding or elastic cements. Of whatever the cement be composed, to be a practical success it must not be costly, must be durable, must not creep or melt unless subjected to a high temperature, the joints must be easy to make, and the cement must not expand or contract so as to interfere with the soundness of a joint. Elastic cements may be made by erecting a suitable size of cauldron in which the necessary ingredients can be heated and mixed together. As the success of a cement depends upon the ingredients used, and the proportions in which they are mixed together, experiments should be conducted on a small scale until a satisfactory cement is produced. The following ingredients will produce an elastic cement suitable for drain pipes : 10 parts by weight of mastic asphalt. 5 parts coal-tar pitch. 5 parts fine sand. FIG. 85. Patent joint. PIPE JOINTS 123 To make the joints, a continuous stream of molten cement should be run into the sockets until they are filled, and the soundness of the joints is improved by smearing the socket and spigot ends with the heated composition before laying the pipes in position. CHAPTEE V SOLDERS, FLUXES, AND LEAD BURNING Soft Solders. The soft solders used by plumbers are alloys, composed principally of lead and tin, and these metals are mixed in varying proportions according to the class of solder required. Very fusible solders are produced by adding bismuth to the above. The composition and fusing temperatures of a few soft solders are as follows : Composition. Solder. Fusing Point. Tin. Lead. Bismuth. Cadmium. Very fine (Wood's alloy) , . 4 8 15 3 158 Fahr. Very fine 3 5 8 ... 203 55 5) 1 2 i 300 55 ) > ' ' 3 2 1 334 ,, 5 55 ' ' 2 1 i 340 ,, Fine for general work . . . | 1 1 i 370 Fine for general work . 8 7 ... Wiping solder 1 2 ... 441'" ,, The chief property of very fine solder is its low melting point. Ordinary fine solder possesses the special property which allows it to be readily " floated " so as to form a smooth and level seam. Excess of tin in fine solder causes the latter to have a rough appearance. 124 SOLDERS, FLUXES, AND LEAD BURNING 125 The property which makes plumbers' or wiping solder so useful for making joints is the plastic state in which it remains when cooling through a certain range of temperature. Much of the solder at the present time is not made by plumbers themselves, and unless soft pig lead can be readily procured for its manufacture it is usually better to purchase a reliable brand of solder than to make it. Much of the sheet lead, from which solder is made in the workshop, is not produced from pure pig lead, and frequently it contains impurity which has a detrimental effect upon the solder made from it. When making plumbers' solder, the lead and tin should be weighed out in proper proportions, and the lead should first be melted in a suitable cauldron and raised to a moderate temperature. The dross which forms on .the surface of the lead should be skimmed off, and the tin then added. As soon as the tin is melted the two metals require to be thoroughly mixed, and afterwards tested by pouring to form several pats on a clean cold stone. If the upper surface of the pats after cooling presents a white surface, with several large bright spots, the solder will be found suitable for use. A dull white surface without bright spots indicates that the solder is coarse and that more tin is necessary, whilst if the surface of the pats is covered with numerous small bright spots the solder contains too much tin. It is essential when pouring solder into moulds to keep it well stirred, or that at the lower part of the cauldron will be deficient in tin. Treatment of Poisoned Solder. As impurity in solder has a very prejudicial effect upon its working qualities, every precaution should be taken to prevent brass filings and other particles of foreign matter from getting into it. Zinc is the worst form of impurity in solder, and this metal may be introduced in the form of brass filings, or by tinning brass- work by dipping it into the solder, or by pouring solder over brasswork when joining the latter to lead pipes. There are two principal methods for purifying poisoned solder. The first consists of melting the solder, when some 126 DOMESTIC SANITARY ENGINEERING AND PLUMBING crushed rock sulphur is added; the pot should be removed from the fire, and the sulphur thoroughly mixed with the solder by stirring. The pot is afterwards replaced on the fire, and the solder slowly raised in temperature until dull redness is obtained. During the reheating process the sulphur combines with the impurity in the solder, and these rise together with particles of lead and tin to form a thick scum on the surface of the solder. The scum or crust should be entirely removed and the solder afterwards cleared by adding a little resin. A little tin may also be necessary, but this can be readily ascertained by testing the solder. In the second method, after the solder is melted it is poured on to a clean iron tray (a porcelain one would be preferable), and as it begins to set it is broken up into as many small particles as practicable. The poisoned solder is then covered with diluted hydrochloric acid and left submerged for about an hour. Any particles of zinc are readily acted upon by the acid, and the solder afterwards requires to be well washed to free it from chloride of zinc. To finally clear the solder, it is reheated and a little resin added as in the first method. Hard Solders. Hard or brazing solders are those which require heating to redness before they can be fused ; they are used for joining the harder metals and alloys, such as steel, copper, brass, gun-metal, etc. Although brazing has not been much in request in plumbers' work, it is very probable that it will become more general in the near future, owing to the displacement of lead waste pipes in the best class of buildings by light copper pipes. Brazed joints in connection with copper waste-pipe work would allow of more simple forms of fittings being adopted than many of those used at the present time. Brazing may be done by heating to a suitable tempera- ture the metals to be joined, either with a bright hot fire or with a powerful gas blow-pipe. The surfaces to be brazed are prepared by filing them clean, and the hard solder is used in a granular form with powdered borax as a flux. After the parts to be joined have been prepared, the flux SOLDERS, FLUXES, AND LEAD BURNING 127 and granulated spelter are placed on the joint, and heat applied until the spelter floats round the joint, in a similar manner to that in which fine solder is used for jointing lead pipes. When brazing brass fittings and copper pipes together some protection must be afforded the brasswork, or the latter may be fused when making the joints, as the composition of brass fittings may not differ very much from that of the spelter used. Brasswork is readily protected from fusing by plastering it over with clay, excepting the surfaces which are to be brazed. Gas blow-pipes are very useful and convenient for light brazing, and especially when oxygen is used in lieu of atmospheric air. Heat of greater intensity may be obtained by the use of acetylene and oxygen. The composition of hard solders or spelter necessarily vary according to the surfaces to be brazed. Composition. Spelter. Copper. Zinc. Tin. Hardest for iron and steel 2 1 Hard for copper, gun-metal, and hard brass . 1 1 Soft for soft brass .... 4 a , Fluxes. The principal uses of fluxes when soldering are : to prevent the oxidation of the prepared surfaces so as to enable a sound joint to be made, to assist the solder to flow, and to aid in cleansing the surfaces to be soldered. One flux is found to be more suitable than another when soldering certain metals, and the following gives the fluxes best suited for the metals and alloys used in plumbers' work : 128 DOMESTIC SANITARY ENGINEERING AND PLUMBING Metal or Alloy. Flux. Lead when coarse solder is used . Lead when fine solder is used Zinc (new) : . . Zinc (old) . . Brasswork and gun -metal : . ','' Tin and pewter . ^ ; Iron when soft solder is used . . . Iron, copper, and steel when brazed Tallow. Tallow and resin. Chloride of zinc. Hydrochloric acid. Resin. Sweet oil. Chloride of zinc or sal- ammoniac. Borax. Lead Burning. This is a term which denotes the uniting of two or more pieces of lead by fusion, and without the aid of either a flux or alloy. The cost of jointing sheet lead by " burning " is small when compared with that of soldering, and as sheet lead can be effectively jointed by " burning," the latter mode of jointing is rapidly gaining favour. Lead burning is accomplished by combining two gases which produce upon ignition a hot clean flame. The mixed gases are delivered at a nozzle or nipple, and the size of the flame is adjusted and directed on the surfaces to be joined, and fused together a little at a time. There are three different mixtures or combinations of gases used for lead burning purposes (a) Pure hydrogen gas and atmospheric air. (&) Coal-gas and atmospheric air. (c) Pure hydrogen and oxygen. When hydrogen and air are used, the former is usually generated in a machine, the latter being delivered from a container which is charged by means of a force pump. When coal-gas and oxygen are used, the former is obtained from the nearest available gas pipe, whilst oxygen may be obtained compressed in strong steel cylinders. Compressed hydrogen and coal-gas may also be obtained in cylinders if desired. A complete arrangement for lead burning is shown in Fig. 86, the hydrogen generator being on the left side, and the apparatus for supplying air on the right side of the figure. For a small generator a suitable size is 10 inches SOLDERS, FLUXES, AND LEAD BURNING 129 square, and the chambers A and B may be each about 1 ft. 3 in. high. The whole of the generator should be made of lead, but when it is intended to be a portable one it will require protecting by a suitable casing as shown. The upper chamber A is made separate from the lower one, and may be removed after taking the joint J apart. "-NIPPLE Fio. 86. Hydrogen generator and air tank. Between the large chambers A and B, a small compart- ment or safety chamber C may be provided of say 4 inches diameter. A lead grate G, which is perforated with J inch holes and 1J inches apart, is fixed about 2 inches above the bottom of chamber B. At S a 5 inch gun-metal screw cap is required, and a washout at M, which is closed when the machine is in use 9 130 DOMESTIC SANITARY ENGINEERING AND PLUMBING by means of a wood plug. A pipe P of 1 inch or 1J inches diameter communicates with the upper and lower chambers, and the bottom of the pipe should terminate not less than 1 inch clear of under surface of the grate G. From the top of chamber B a |-inch pipe is taken and turned through into the safety chamber C, and continued nearly down to the bottom as shown. A small cock D may also be fixed, so as to regulate the height of the water in the safety chamber C. To charge a generator, a few pounds of zinc, which is broken into small pieces, are distributed over the grating G, and the screw cap S replaced and made air-tight. The cock at H is closed, and into the upper compartment A water is poured. Some of the water enters the lower chamber through pipe P, but its entry is soon prevented owing to the air in B being unable to escape. To each gallon of water about three-quarters of a pint of strong vitriol (sulphuric acid) is afterwards added, so as to well mix with the water in A. The cock H is now opened, the confined air escapes, and the diluted acid flows into B and submerges the spelter, when the generation of hydrogen begins. As soon as all air is displaced from the generator, hydrogen gas is avail- able for burning. It is essential when charging a generator that the water and acid are added in the order described, for if the acid were first added an explosion would result. The reason why pipe P dips below the grate G is to enable the production of hydrogen to be controlled. When the gas begins to accumulate and to generate pressure, it displaces the dilute acid from B back into A until the acid is clear of the zinc, when the evolution of hydrogen is automatically stopped. At the same time the end of the pipe below G remains sealed, and waste is prevented, as the gas is unable to escape. It will thus be seen that when a generator is in use the diluted acid in A rises and falls according to the rate the gas is used. Care is necessary in the management of a generator if it is to work satisfactorily. Each day after use the spent or dilute acid should be discharged, and the generator well washed out with clean water. Instead of a wood outlet plug M, a stoneware cock may be used, but the former is more SOLDERS, FLUXES, AND LEAD BURNING 131 satisfactory for a portable machine, as the latter is liable to be broken. The zinc should be broken into small pieces, as a greater surface is exposed to the action of the acid than when large pieces are used. When a generator is not regularly cleansed, sulphate of zinc begins to crystallise and to choke up the pipe P. Crystals also form on the surface of the zinc and prevent the free generation of hydrogen. Should a generator, however, be allowed to get in such a state, the trouble can be overcome by washing it out with hot water, which dissolves the crystals formed. The purpose of the safety chamber C is to prevent the generator being damaged by explosion should a light be applied before all atmospheric air has been dislodged. Hydrogen alone cannot explode, but when mixed with a certain volume of air an explosive mixture is produced. If through ignorance or carelessness a light should be applied to the nipple before the air is removed, damage by explosion would be chiefly Confined to the safety chamber, as the direct passage between it and B is broken by the water seal, owing to the gas first requiring to bubble through the water before it can be delivered to the tubes. The lead for a safety chamber should only be thin, and 4-lb. lead is suitable for the purpose. The tank for supplying air, Fig. 86, is of simple con- struction, and is divided into two compartments as shown. This may be formed of galvanised sheet iron, a convenient size being 12 to 14 inches diameter, and about 3 feet in height. The height of the lower compartment should be a little less than the upper one, in order that water may not be projected from the upper chamber when the container is overcharged with air. The pipe X, Fig. 86, is made to communicate with the lower part of each compartment, so that water may be displaced from the lower to the upper compartment by means of the air-pump Y. To prevent water getting into the flexible tube and cutting off the supply of gas, the outlet pipe is carried above the lower chamber, and turned through the side as in Fig. 86. For maintaining a steady pressure of air, which is essential for good burning, the water in the upper compart- 132 DOMESTIC SANITARY ENGINEERING AND PLUMBING ment is often maintained at a constant level by continuous pumping. After a generator and air container have been charged, the gases are led through rubber tubes of any convenient length to a breeches piece N, which is provided with regulating cocks. To the outlet of the breeches piece a few feet of mixing tube are joined, and to the other end of the mixing tube a brass tube is inserted on which various sizes of nipples can be screwed. Lead burning is generally classified under four heads, according to the positions in which joints in sheet leadwork are made. These are as follows : Flat burning, horizontal burning, upright burning, and overhead burning. In flat IS FIG. 87. Examples of lead burning. burning the edges to be fused may either be placed edge to edge or overlap each other, as at A and B, Fig. 87 ; in both cases a strip of lead is used and a raised seam formed. Flat burning is the easiest kind of " burning," and is soon learned, but great care is necessary in the early stages, for at every point the lead must be properly fused or defects may arise which are difficult to locate. For burning a very hot, pointed flame is required, in order that the lead may be fused at any point without first heating the surface surrounding it. Lead strips for burning are usually of the form shown at E, Fig. 87, and are cast in clean iron moulds which may be obtained at a cost of a few shillings. A sample of horizontal burning is shown at C, and a lead strip is used to make the joint provided the edges SOLDERS, FLUXES, AND LEAD BURNING 133 of the lead are thick enough to permit of it. For thin lead the overlap is simply fused with the lead behind it. Upright burning D, Fig. 87, is much more difficult than either flat or horizontal burning, and a smaller flame is required. The edge of the overlap is fused with the lead behind it, the burning being commenced at the bottom as indicated in the figure. The joints of upright burning are much narrower than those of flat burning, and much smaller beads are formed. Overhead burning is the most difficult to accomplish, but in ordinary plumbers' work it is very rarely if ever required. Plumbers in chemical works and similar places FIG. 88. Apparatus for lead burning. are occasionally required to do overhead burning, but not nearly so often as is generally supposed. Where coal-gas is available, and compressed oxygen is used, the apparatus for lead burning may take the simple form given in Fig. 88. A cylinder when fully charged with oxygen is subjected to an internal pressure of 120 atmospheres, or say 1800 Ib. per sq. inch. As the gas leaves the cylinder it may be reduced in pressure with a special form of automatic regulator, and the arrangement of tubing and common nipple adopted as in Fig. 86. When, however, compressed gases are used an injector form of blow- pipe is more suitable, and the gas from the cylinder may be regulated by a simple pattern of adjustment valve. In Fig. 88, V represents the cylinder valve; the fine 134 DOMESTIC SANITARY ENGINEERING AND PLUMBING adjustment valve is indicated by A, and by it the necessary amount of oxygen is admitted to the tubes, and without subjecting them to undue pressure. As the valve A takes the place of an automatic regulator, care is necessary, however, in its manipulation. The rubber tube from the oxygen cylinder is joined with the injector side of the blow- pipe, and the oxygen and coal-gas mix in flowing towards the nipple. A good size of cylinder for workshop use is one holding 40 cubic feet of oxygen when under the maximum pressure of 1800 Ib. per sq. inch, but a 20 feet size is more suitable in other cases, as the weight of the larger cylinder is about 60 Ib. Oxygen for lead burning purposes can be obtained at the reduced rate of Ifd. per cubic foot in 40 -feet cylinders, and at 1-J-d. per foot in 20 -feet cylinders. In country districts the cost of carriage would require to be added. The amount of oxygen consumed varies with the kind of burning, but the average consumption may be taken at 1 cubic foot per burner per hour. Fig. 88 shows an apparatus which is very suitable for ordinary plumbers' use, and the initial outlay is only about one-third that for the generator and air cylinder given in Fig. 86. Gases when compressed in cylinders are extremely useful for burning sheet leadwork on roofs, such as that on finials, stone cornices, etc., and for burning the joints of lead-lined cisterns. A very strong rubber tube is required for the oxygen connection, as this tube is subjected to a moderate pressure. The only preparation required for " burning " is the shaving of the surfaces, and this should be neatly done. Neither flux nor plumbers' black is required. To line large cisterns with lead where the joints are burned, the lead should be arranged so that all the joints come on flat surfaces about 1-J inches or so from the angles. The lead for the sides of the cisterns should be turned inwards along their bottom edges, in order that the sides and bottoms may be joined with flat burning. UNIVHKO1 I T OF SOLDERS, FLUXES, AND LEAD BURNING 135 For small cisterns which are easy to handle the burning may be done at their angles if desired. A strip of lead should be used wherever possible, as much stronger joints are made than where two surfaces are simply fused together. CHAPTER VI SANITARY FITTINGS AND ACCESSORIES SANITARY fitting is a very comprehensive term, but in this chapter its use is limited. Owing to the rapid advances which have taken place in sanitary science, very little space will be devoted to what are now obsolete appliances. The general principles which should govern the con- struction of sanitary fittings are as follows : (a) Simplicity of form, being free from complicated and not readily accessible parts, (b) They should be made of hard, durable material, with smooth and well glazed surfaces, so as to be practically non-porous, (c) The design should not admit of unnecessary surfaces which are liable to collect dirt, and which require a lot of attention to keep them in a cleanly state, (d) Their outlets should be formed to enable reliable and simple con- nections to be made with the pipes into which they discharge. Water Closets. The best w.c.'s at the present time are the " wash-down " and " siphonic " types. There are patterns of these, however, that are not free from structural defects, but there is no difficulty in procuring a first-class article from a good manufacturer provided a reasonable price is paid for it. Other types of w.c.'s are very inferior to those already named, and a few will be described and their principal defects noted. An ideal w.c., besides satisfying the general conditions enumerated above, should be thoroughly cleansed with one flush of water, have a small amount of dry basin surface exposed to contamination, hold sufficient water to completely submerge all excrernental matter, and be trapped in such 136 SANITARY FITTINGS AND ACCESSORIES 137 a manner that an effective barrier is formed to prevent the passage of drain air into the w.c. apartment. All w.c/s which require wood or other enclosures, with the exception of the valve closet, are obsolete, the pedestal form taking the place of those in which basins and traps were made in separate parts. Wash-out Type. The wash-out w.c., Fig. 89, is a defective form of the pedestal class. The part which acts as a receiver destroys the force of the flushing water, with the result that excrement accumulates on the imperfectly flushed surfaces on the inlet side of the trap, and matter is also frequently left lodging in the trap. In certain cases where wash-out w.c/s have been fixed it has been essential to replace them with a better type within half a dozen years of instalment. Wash - down Type. For a simple form of w.c. there is nothing better at present than the wash-down type, when well designed and supplied with an adequate flush of water. In the wash-down w.c. the full force of the flushing water is utilised to cleanse the basin and trap. There are many different wash-down w.c.'s that vary in constructional "Wash-out w.c. details, such as in the size and shapes of basin, size of water surface, positions of outlets, depth of water seal, and the form of flushing rim. The size and shape of a w.c. is of considerable import- ance, for upon these its efficiency largely depends, and especially when flushed with a limited volume of water. Wash-down w.c.'s. with receding backs are, as a rule, objectionable, as these surfaces when fouled are not always thoroughly cleansed. The depth of water seal in the trap of a wash-down w.c. is limited, for when it exceeds 2 inches considerable resistance is offered by a trap, and frequently the effective removal of excrement with one flush is uncertain. On the other hand, the water seal of a trap should not be less 138 DOMESTIC SANITARY ENGINEERING AND PLUMBING than 2 inches deep in order to ensure a safe barrier against the passage of contaminated air when the seal has been reduced a little by waving out, or by evaporation, etc. A large water area in a basin is always desirable, but this in turn is limited by the form the basin takes and by the volume of water used for each flush. Thus, when a water surface is comparatively large, and when a trap has a 2-inch seal, not less than three gallons of water will be required for scouring out the w.c. A wash -down w.c. that is most easily cleansed is one that is comparatively small in size, where the water surface in the basin is small, and where the trap holds the minimum volume of water and has a small seal. Some of the w.c.'s constructed in this manner can be cleansed with two gallons of water and less, al- though such volumes are insufficient to cleanse the drains into which the w.c.'s dis- charge. Doulton's simpli- citas w.c. is illustrated in Fig. 90 ; the back is constructed fairly straight, so as to minimise the risk of soiling it. The general construction of the basin is such as to limit the water area in the trap, in order that the w.c. may be flushed with two gallons of water. The outgo is well above the floor, and admits of a reliable connection being made with the branch soil pipe. Fig. 91 gives another wash-down w.c. by the same makers. The front of the basin in this case is not curved to the same extent as in the one previously shown, and the w.c. has a larger exposed water area. This form of con- struction also has less surface that is liable to be fouled when compared with that of Fig. 90, but on account of the extra resistance offered by the increased volume of water in the trap not less than a three-gallon flush will be required to FIG. 90. Wash-down w.c. (Doulton's "Simplicitas.") SANITARY FITTINGS AND ACCESSORIES 139 give satisfactory results. To the outlet of Fig. 9 1 is attached the firm's patent Metallo-Keramic joint, so that an easy and safe connection can be made with a lead soil pipe. The connecting piece consists of a short piece of lead pipe which is Soi.orR FIG. 91. Wash-down w.c. soldered to the outgo of the w.c. To enable a lead pipe to be soldered to the pottery ware, a metallic film is deposited on the outgo, and is afterwards fired to thoroughly fix it. The lead pipe is then soldered to the metallic film. Shanks' "Modern" w.c., Fig. 92, differs slightly in FIG. 92. Wash-down w.c. (Shanks' "Modern.") construction from those already shown. The back of this closet recedes a little to prevent its getting soiled, and although receding backs are not generally a success, the one shown is an exception to the rule on account of the special design of the basin and the form of flushing rim adopted. 140 DOMESTIC SANITARY ENGINEERING AND PLUMBING The exposed water surface is of medium size, and the w.c. can be cleansed with a two-gallon flush, although for reasons previously stated a larger flush is desirable. It will generally be found that wash-down w.c.'s with S traps are more easily cleansed than those with P traps, owing to the outlets of the former admitting of a freer dis- charge. For the poorer and cheap grades of wash-down w.c.'s those with S traps should be used wherever practicable, as those with P traps often have their outlets badly formed, and with little or no pitch. A shape of w.c. that will permit of a 2J to a 3-inch seal, and be cleansed with a two-gallon flush, is given in Fig. 93. It will be observed that in order to satisfy these conditions both the basin and water surface are of restricted size. Flush Pipes should be as free from bends as possible, and the height of a flushing cistern need not be more than 6 feet above the flush- ing rim of a w.c. When flush pipes are 1^ inches FIG. 93. Wash-down w.c. diameter, and fairly straight, a flushing head of 5 ft. 6 in. gives very good results. Unless flush pipes are long, a greater head than 6 feet usually causes water to be projected on the floor of the apartment or on to the w.c. seat. In situations where a flushing head is limited to about 4 feet, and a wash-down w.c. is to be fixed, care should be taken to select a small pattern or it may not be cleansed with one flush of water. For hospital use w.c.'s are frequently constructed to enable them to be fixed clear of floors, by building into the walls corbels which form parts of the closets used. Such w.c.'s are also utilised in other public institutions where cleanliness is of paramount importance. Combination Closets. For situations where there is in- sufficient space to fix overhead cisterns the combination w.c. SANITARY FITTINGS AND ACCESSORIES 141 may be used. The flushing rims of. the basins require to be moderately large, and the outlets of the cisterns are much larger than those of overhead cisterns. On account of their silent action combination closets have been used in lieu of those with overhead cisterns, but as regards their general efficiency they are often inferior to the latter provided the flushing cisterns are fixed at suitable heights. The larger inlets of combination forms, and other features, do not com- pensate for a reasonable flushing head of the ordinary type. Valve Closets. The valve closet possesses some good points which are absent in a wash-down type, but the former has also failings from which the latter is free. The principal merits of valve closets are: (1) The large volume of water held up in the basins ; (2) The small area of basin surface liable to be fouled ; (3) The flushing power of the water in the basins when suddenly released. The drawbacks of valve closets are : (1) Their compli- cated construction ; (2) The mechanism in connection with them is liable to get out of order ; (3) They are costly ; (4) They may require to be enclosed with casings, and filth may be allowed to accumulate and remain hidden from view. When a valve closet is fixed upon a wood floor, a lead tray or safe should be placed beneath it. The valve box and trap may either be obtained in one piece of pottery ware, and with the whole of the closet placed above the floor, or the valve box and trap may be had in separate parts ; in the latter case the trap may be of lead and fixed beneath the floor, and the valve box may be either of cast iron, porcelain enamelled inside, or of lead. A source of weakness with the early forms of valve closets was in connection with their overflows, but this draw-, back is practically overcome in modern fittings, where the overflows are open and washed out each time the closets are flushed. The valve box is made as small as possible, and provision is made for ventilating it. Fig. 94 gives a modern form of valve closet, and the trapping and general arrange- ment of the overflow are clearly shown. The recharging of the basin is usually controlled by some particular form of regulated valve, the water overflowing when the water-line is 142 DOMESTIC SANITARY ENGINEERING AND PLUMBING reached. Valve closets are largely used in ships, their use being essential in many cases to prevent the backwash of water. Siphonic Closets. Owing to the limitations of both wash- down and the best forms of valve closets, the siphonic type has been introduced. This form may not at present be the acme of perfection, yet when constructed upon sound principles it possesses most of the merits of both the valve and wash- down types. Many siphonic closets are nearly silent in action, have a quick and powerful discharge, have large water surfaces and a minimum of fouling surface, and possess the merit of having their contents with- drawn by siphonage instead of depending upon their contents being dislodged by the VENTILATING PIPE (OKJWECTIONI FIG. 94. Valve w.c. where trap is fixed beneath floor. force of the flushing water, as in the case of the wash - down type. A siphonic w.c. mayalso have a deeper seal than any other form. The chief faults of siphonic w.c/s arise through defects of construction and not to the general principle embodied in them. Various means have been devised for starting siphonic action in these closets, such as by the introduction of jets or peculiarities of construction, and by either the expulsion or extraction of confined air. Many siphonic w.c.'s require special flushing cisterns to work in connection with them on account of the volume of water necessary for recharging the basins after siphonage has ceased. There are other siphonic closets with which any ordinary flushing cistern can be used, but these usually hold SANITARY FITTINGS AND ACCESSORIES 143 SOLDER' much less water than the first mentioned. The latter are recharged with water as a rule by means of an after flush compartment which forms part of the closet. For the purpose of comparison all siphonic closets may be grouped in two classes. First, those that have two traps, and second, those with only one trap. In the double trapped class, siphonage is started either by extracting or displacing air from the limb that communi- cates with both traps. The object is the same in either case, viz. to destroy the equilibrium of air pressure on the water surface of the basin, and of that in the limb between the two traps. Siphonage in the sinyle trapped type is started by momentarily re- __ tarding the first outflow of water through the trap, by introducing some particular form of construc- tion, and with or without the aid of a jet of water. Shanks' " Le- vern " siphonic closet, Fig. 95, is an example of the one-trapped class. It is simple in con- struction, has a large water area in the basin, very little surface that is liable to be soiled, and the whole of the flushing water passes through the basin. The siphonic action is due to the enlargement E in the lead outlet pipe, and is established as follows : when the closet is flushed, the water follows the curved surface at E, and owing to the abrupt change of direction at the bottom of the enlargement the water is caused to be projected towards the centre, and to produce a momentary stoppage of the flow; the brief interval of retardation is, however, sufficient . to allow the outlet leg to get fully charged, when siphonage is established, and the contents of the basin rapidly withdrawn. Provision is made FIG. 95. Siphonic w.c. (Shanks' "Levern"). 144 DOMESTIC SANITARY ENGINEERING AND PLUMBING in the flushing cistern for recharging the basin with water. This design of w.c. admits of the basin being emptied of water if a pailful of slops is emptied into it, but this may be counteracted by using an anti-slop discharge attachment in connection with the closet. Doulton's siphonic w.c., Fig. 96, represents one of the double-trapped class, and, like the one previously shown, it has a large water surface and small area that is liable to be soiled. Siphonage in this case is principally started by FIG. 96. Doulton's siphonic w.c. reducing the air pressure between the two traps by means of a small pipe which communicates with the space B and the flush pipe or cistern. It is so arranged that when the water descends in the flush pipe, the aspirating effect produced is transmitted to the limb B by the pipe A. In this manner air is withdrawn from the limb B, or rarefied to such a degree that unequal air pressures act upon the water surfaces, and cause siphonage to be established. Two traps are of course essential to confine air for this particular action, but the bottom trap also aids to a certain extent in starting siphonage by offering resistance at the outset to the escaping water, and in enabling SANITARY FITTINGS AND ACCESSORIES 145 the limb B to be charged just prior to the siphonic action being started. The inlet of the lead trap T may be obtained shorter when desired, to enable the whole of the wD PIPE RPC. upon rubber rings for their soundness are not always reliable for the outgo of a w.c. ; and where space is limited flange joints are difficult to make. Special brass union couplings with ground joints are occasionally used for soil pipe connections, to enable a lead outlet bend to be turned in any direction desired ; this form of joint is better when made beneath the water-line of the trap, in order that water may drip on the floor and indicate the presence of a defect if such occurs. Antisiphonage Pipe Connec- tions. A source of weakness in many closets is the connections between the antisiphonage pipes and the ventilating horns on closets. In the case of a ven- tilating horn being located where there may be some degree of uncertainty in being able to make a reliable joint between it and the antisiphonage pipe, the best plan to adopt is to seal up the earthenware horn, and to join the antisiphonage pipe at another point. Unless a soil pipe is periodically tested, a defect at an antisiphonage pipe connection may remain undetected for an indefinite period. If a vent horn is located at the crown of an outlet bend of a w.c. it is liable to be choked and rendered useless ; moreover, if the flush pipe also joins at the back of a closet with a horizontal con- nection, the two joints usually come too close together to admit of their being properly made. At A, Fig. 99, an antisiphonage pipe is shown which is joined with a ventilating horn at the side of the outgo of a w.c. ; this arrangement keeps the anti- siphonage pipe clear of the flush pipe, and admits of the joint being more readily made. A suitable jointing material FIG. 99. Joints for antisiphonage pipes. SANITARY FITTINGS AND ACCESSORIES 149 AwTlSYPHONAGE PlPE. is a mixture of red and white lead, linseed oil, and a little hemp. A special form of connection is given at B, Fig. 99, where the lead antisiphonage pipe is attached with a wiped joint to a brass union, which in turn is connected with the pottery ventilating horn ; these union connections usually have their faces ground together to enable the joints to be readily made and disconnected. There are many places, however, where a joint like B, Fig. 99, would be difficult to make on account of the space they occupy ; but where an antisiphonage pipe may pass directly through a wall at the back of a w.c. such a joint could easily be used. Another antisi- W.C. OUTLET phonage pipe con- nection is illustrated in Fig. 100. Here it is formed in the brass socket to which the outgo of the w.c. is joined. This latter arrange- ment, of course, can only be adopted where the pottery outgo terminates well above the floor. Where practicable one of the best methods of dealing with an antisiphonage pipe is to join it directly to a lead or iron branch instead of to the earthenware outgo of a w.c. Flush Pipe Connections. Although many methods have been devised for connecting a flush pipe to a w.c., a strong rubber cone will generally be found to be satisfactory where a horizontal joint is required. Eubber cones are comparatively cheap, and permit of reliable connections being made in con- fined situations, whilst many patented connections take up too much space. Thin rubber cones should not be used, as they are weak and are soon destroyed. Where the inlet to a flushing rim is in a vertical position, a good joint may be made by first inserting into the annular PIPE. FIG. 100. Connection between \v.e. and lead pipe beneath floor. 150 DOMESTIC SANITARY ENGINEERING AND PLUMBING space between the flush pipe and socket a little hemp, and afterwards filling the remaining space with a mixture of molten resin and tallow ; a little brickdust may also be added to increase the soundness of the joint. Molten sulphur also makes a good joint. Flushing Cisterns for W.C.'s The object of a flushing cistern is not only to regulate the volume of water used per flush, but also to reduce to a minimum the risk of polluting the water supply, by avoiding a direct connection between the water service pipe and a w.c. Flushing cisterns are also designed to prevent waste of water. There are many kinds of flushing cistern's, but they may FIG. 101. Nicholl's and Clarke's double valve flushing cistern. be divided into the following orders: (1) Single valve cisterns ; (2) Double valve cisterns ; (3) Valve and siphon cisterns ; and (4) Waste preventing siphon cisterns. The first and second classes of flushing cisterns are not often fixed at the present time, as they are rapidly being displaced by those of the third and fourth classes. It is generally recognised that only siphon types of flushing cisterns should be used, in order that the full flush of water may be utilised to cleanse a w.c. Double valve cisterns are divided into two compartments, and have a valve in each. The compartment in which the larger valve is located contains the regulation flush, and the valves are so arranged that when one is closed the other will SANITARY FITTINGS AND ACCESSORIES 151 remain open. The general arrangement of these cisterns is clearly shown in Fig. 101. A valve and siphon cistern is given in Fig. 102, but unless the lever is let go after the discharge has started the cistern will not be wholly emptied of its contents. Occasionally it will be found that when water flows freely into a flushing cistern the siphonic action is not properly broken at the end of the flush, but that continuous siphonage occurs, the water being withdrawn from the cistern as quickly as it enters it. To effectively break siphonic action it is essential that air enters the siphon at the end of each discharge. In the siphon shown, Fig. 102, a small hole is made at the top of the dome, just sufficient in size to stop siphonage at the end of the flush, but not large enough to interfere with the effective work- ing of the siphon. When the dome of a siphon is of iron there is always FIG. 102. Siphon flushing cistern. the possibility of a small aperture getting choked with rust, and causing the siphon to be somewhat erratic in its action. This may, however, be obviated to a great extent by drilling a hole in the iron dome and inserting a small brass plug which has previously been pierced. The form of cistern Fig. 102 is not what can be termed a waste preventer in the strict sense of the term, but it is nevertheless a fairly good type of cistern, and will satisfy the requirements of many Water Companies. A real waste-preventing cistern is one which is so de- signed that water cannot continuously escape through a flush pipe when resorting to irregular practices such as fastening down the cistern pull, or by holding the outlet valve clear of its seating, by putting some material beneath, or by the total removal of a valve. Waste-preventing siphon cisterns are made in many 152 DOMESTIC SANITARY ENGINEERING AND PLUMBING forms, but they depend principally for their action upon the mechanical displacement of a volume of water into the outlet limb of the siphons, or upon a pneumatic action. The pneumatic class possesses the merit of dispensing with rods and chains, and permits of their contents being discharged by the simple pressing of a button or a small piston, etc. Fig. 103 shows a displacer type of siphon waste- preventer, and its action is as follows : When the lever is pulled down the piston moves up the cylinder, displaces the water above it over the top of the siphon bend, and charges the vertical limb, when siphonage is established. It will be observed that the outlet limb of the siphon is reduced in diameter, instead of being made with a uniform bore ; some form of contraction or enlargement should always be introduced into such siphons, as it has the effect of momentarily re- tarding the outflow of water at the commence- ment of the discharge, enabling the siphon to be better charged, and of making it more reliable in action. Because the rod of the piston in Fig. 103 passes through the top of the dome, no special air-hole is essential to break siphonage at the end of a discharge, as a'ir will enter the siphon at the side of the rod, which is not perfect fitting where it passes through the dome. A waste-preventer with pneumatic action is illustrated in Fig. 104. As the cistern fills, air is confined at the top of the chamber C. At the bottom of the siphon a disc valve V is provided, which opens on a hinge during the discharge from the cistern, but falls back upon its seating after- wards. A small pipe is joined to chamber C, and the other end of the pipe is connected with the small bellows B. When the cistern is filled to the water-line, siphonage is FIG. 103. AYaste preventer (Donltons' "Speedwell "). Overflow and ball- cock omitted. SANITARY FITTINGS AND ACCESSORIES 153. established by pushing the knob K inwards in order to compress the confined air in chamber C. The effect of this further compression is to dislodge sufficient water over the outlet limb of the siphon and to bring about the discharge. Flushing cisterns are made in cast iron, wood with metal linings, and in pottery ware. Iron flushing cisterns, unless protected with a glass enamel, or other suitable coating, are rapidly corroded with many soft waters, and the closets are also discoloured with rust. Lead-lined or copper-lined cisterns are more satisfactory for soft waters, the latter being the more durable of the two. Lead, however, has the advantage of being a little cheaper than copper. Pottery cisterns present a clean appearance, and are durable ; they are more costly, however, than some of the other forms, and their use is limited in conse- quence. To prevent flushing cisterns being damaged by frost they should have slop- ing sides. Lavatories. The chief point of construction which requires consideration in connection with modern lavatories is the arrangement of their overflows. Defects which were common in the earlier types of lavatories, such as small waste outlets, and soap dishes which discharged into overflows, or into traps beneath, are mostly absent in modern lavatories. These, like other sanitary fittings, should be free from wood enclosures, and be constructed so as not to accumulate filth. Porcelain and enamelled fireclay are chiefly used in this country in the manufacture of lavatories, the latter material being used where specially strong and durable fittings are required. A good shape of lavatory is the oblong form with straight Fro. 104. Adams pneumatic action waste preventer. Ball-cock and overflow omitted. .154 DOMESTIC SANITARY ENGINEERING AND PLUMBING front and circular back, the latter having a suitable recess in which the waste outlet may be located. All lavatories, whether in cheap or in expensive forms, should be provided with skirtings, to prevent water overflowing and trickling down between walls and the backs of lavatories. The tip-up form of lavatory, once largely used, is super- seded by simpler forms which are more readily cleansed. Tip-up basins possess an advantage in having a quick dis- charge, but their drawbacks, such as the fouling of the receivers, the disagreeable odours emitted by them when not regularly cleansed, and the wear and tear upon the trunnions, overshadow any merit they possess. A weakness which is common to many present day lavatories is the form the overflow takes. Overflows should be constructed in a manner which will admit of their being easily cleansed, be open to view, and be of a good size. Hidden overflows are liable to get their surfaces covered with slimy matter, which dries and emits that stuffy smell which is peculiar to apartments in which defective sanitary fittings are placed ; they also form suitable places in which disease germs may rapidly multiply. In hospitals, schools, and other public institutions, it is of special importance that sanitary fittings have no hidden parts which may cause the dissemination of disease. In Fig. 105 two good forms of overflows are shown. That at A is formed within the basin, and made open as shown ; it is readily accessible, and is easily cleansed by pushing a brush through it. The opening above the weir of the overflow may be covered by a thin metallic hinged flap if desired. At B, Fig. 105, an exposed standing waste is shown, which serves the purposes of overflow as well. This standing waste consists of a short metallic tube which admits of easy removal for cleansing. Porcelain tubes are also used for wastes, and present a clean appearance, but they possess the drawback of being easily broken. Concealed standing wastes should not be used, even if they are made in a manner to admit of easy removal, for hidden surfaces, no matter how accessible, are not cleansed so regularly as those which are exposed to view. SANITARY FITTINGS AND ACCESSORIES 155 C and D, Fig. 106, give two defective forms of overflows ; neither admits of easy cleansing, but owing to its shortness the one at C is the better of the two. FIG. 105. Good forms of overflows. Iii public buildings, lavatories and similar fittings are often fitted with combination hot and cold water valves, but many of these appliances are more trouble than they are FIG. 106. Defective forms of overflows. worth. Combination cocks are frequently erratic in action and troublesome to repair. Separate hot and cold water cocks are, as a rule, the best fittings for lavatories and similar appliances, as they are cheaper, more durable, more easily 156 DOMESTIC SANITARY ENGINEERING AND PLUMBING repaired, and the temperature and volume of the outflowing water are easily adjusted. When lavatories are fixed with standing wastes, care should be taken to leave them in good working order. Occasionally it will be found that standing wastes when delivered from the works stick a little, and although the defect is easy to remedy, trouble has frequently resulted through neglecting it. When loose plugs are used for lavatories, they should be made of some material which is incapable of chipping the glaze. Loose plugs and chains give very little trouble under ordinary circumstances, and they are simple and cheap. Brass rods occasionally take the place of chains, to give a smarter appearance and to make a stronger fixing, but if the guides are fairly close fitting the rods are liable to stick unless some lubricant is occasionally applied. On the other hand, if the plugs drop freely upon their seatings, there may be danger of partially unsealing the traps, due to the sudden compression of the air at their inlets. This, however, applies more specially to baths than to lavatories, as the overflows of the latter would often afford relief for the air. For supporting lavatories painted cast iron or porcelain enamelled brackets and frames are very serviceable, and present surfaces which will only collect the minimum of dust. Where more ornamental fixings are required various designs of friezes and standards may be used. Ranges of Lavatories for offices, schools, hotels, works, and for other buildings, may readily be formed with separate lavatories, or with those with overlapping joints as in Fig. 107. In the range given each lavatory is provided with a different but good form of overflow. A more elaborate range of lavatories by Doultons is shown in Fig. 108. Baths. A number of the conditions which must be satisfied to produce a good type of bath are similar to those necessary in an up-to-date lavatory. Baths vary in size and shape, are made of different materials, have several grades of finish, and differ in the kind and arrangement of their fittings. SANITARY FITTINGS AND ACCESSORIES 157 The principal materials from which baths are made are cast iron and fireclay ; each material has its special merits according to the class of building for which a bath is required. The degree of perfection attained in the manufacture of cast-iron baths has done much to bring these fittings into general use, and the} 7 are produced in both cheap and ex- pensive forms, suitable for either a cottage or mansion. For most buildings cast-iron baths with roll edges are the most satisfactory kind. Iron is quickly heated when in contact with hot water, and the rolled edges dispense with the use of wood casings. To protect iron baths from rust, and to give them a FIG. 107. Range of lavatories with overlapping joints, by Twy fords Ltd. smooth and satisfactory finish, their inner surfaces are enamelled, whilst their outer surfaces are more generally painted. Metallic, vitreous, and porcelain enamels are used, and the grade of a bath is regulated by the kind of enamel used. Metallic enamel is only the application of paint, which is fixed at a moderately high temperature, the number of coats regulating the finish desired. Vitreous and porcelain enamels produce a smooth, glassy surface, which resembles that of glazed earthenware. Glazed fireclay baths are very durable, and their surfaces are easily cleansed ; these baths are specially suited 158 DOMESTIC SANITARY ENGINEERING AND PLUMBING for public and other institutions, where they are often in demand, and where substantial fittings are essential. For private houses, however, fireclay baths when only occasionally used are not suitable ; they are too heavy and cumbersome, absorb too much heat from the water, take a long time to heat through, and unless filled for a time lief ore FIG. 108. Lavatory range by Doultons Ltd. they are actually required their surfaces strike cold to the bather, especially in winter time. Other materials, such as copper, tinned steel, and enamelled sheet iron, are used to a limited extent in the manufacture of baths, and these materials have the advantage of lightness and are suitable for special purposes. For portable baths in hospitals and in other establish- ments, copper is a very satisfactory material, as it is a SANITARY FITTINGS AND ACCESSORIES 159 good conductor of heat, and readily acquires a temperature which only differs a little from that of the water inside. On account of the expense of copper the cheaper materials named are frequently used where cost is the chief con- sideration. Portable baths for hospitals are frequently fixed on rubber-tyred wheels, provision being made in the corridors for either filling them with water or for effecting their discharge. Overflows and waste outlets of baths should be of simple FIG. 109. Standing overflow and waste outlet. construction, and the traps should be fixed immediately beneath the outlets. Fig. 109 shows a good form of standing waste and overflow, and the correct position for the trap, which may be either of brass, or of cast iron and glass enamelled inside. When standing wastes are used, recesses are necessary in the ends of the baths to enable the wastes to be located out of the way of the bather's feet. A form of waste which is often used for expensive baths is given in Fig. 110. These wastes present a nice appear- 160 DOMESTIC SANITARY ENGINEERING AND PLUMBING ance when well polished or silver plated, but they are very objectionable from a sanitary standpoint, as they contain surfaces upon which filth is liable to gather and to remain unobserved. It will be noticed from the construction of the waste that water will rise in the annular space between the tubes when a bath is filled, and in consequence soapy matter will be deposited upon the surfaces when the bath is in use. Some Water Companies, however, do not allow overflows of baths to discharge into their waste pipes, but require them to be arranged as in Fig. 111. To the outlet a length of pipe is attached, and the other end of the pipe terminates in the open air, so as to serve the pur- pose of a warning pipe. This form of overflow was intro- duced with a view to minimise waste of water, but its value for that purpose is very questionable. It is, however, a very defective form of over- flow, as it does not permit of being cleansed, and it allows cold air to flow into the bathroom, along with any fine particles of matter which have been deposited on the surfaces of the pipe. For hospitals and other places where it is desirable to have special facilities for cleansing the floors and walls of bathrooms, baths may be obtained which revolve round their waste outlets. This arrangement is specially useful where space is limited, as a bath when in use may be turned from the wall of an apartment so as to allow an attendant to get on either side. FIG. 110. Defective form of standing overflow, SANITARY FITTINGS AND ACCESSORIES 161 Sinks. The uses of sinks are varied, and different materials are used in their formation. Well glazed fireclay sinks are specially suited for small houses, and they may be obtained in inexpensive forms. For mansions, hotels, restaurants, and public institutions, glazed fireclay is the most suitable material for sinks for general purposes, and for vegetable sinks. A softer material, however, is often necessary for sinks which are used for washing glass and china ware, in order to minimise the risk of chipping these goods. Where a moderate amount of care is exercised, and where sinks are fairly deep and wood drainers are used, fireclay sinks are often suitable for cleansing china ware, but for large buildings it is usually desirable to provide special sinks for this purpose. Hard wood, such as teak, and also softer woods when the latter are protected with soft metal linings, make good sinks for washing crockery ware. Metal-lined sinks are superior to those made entirely of wood, for after a time decay sets in, and the wood becomes FIG. ill. Tell-tale overflow, saturated with greasy water, and during the partial drying of the wood offensive odours are also emitted. Where sinks are used that are lined with either block tin or lead, their sides should be tapered to allow the linings to slide a little inside the wood casings when the metal expands and contracts. Sinks with vertical sides rigidly hold soft metal linings, and cause the latter to buckle and crack when alternately heated and cooled. Fig. 112 shows a wood sink, and a suitable method of lining it with either tin or lead. The metal, it will be observed, is not turned over and nailed down on the top edges of the sink, but is trimmed off a little below the top edge, and held by means of an oak capping piece which covers the free edge of the lead. When a wood sink is treated in this n 162 DOMESTIC SANITARY ENGINEERING AND PLUMBING manner either lead or tin linings will move a little and have a much longer life. Butler's sinks are frequently made of tinned copper, and are oval in shape ; or larger sinks may be of wood and be lined with tin. Drainers for sinks when made of teak are durable, but when soft woods are used they should be covered With either block tin or lead, the former metal, of course, being the better of the two. Either lead or tin is readily worked into the grooves of a drainer after it is first heated to about 200 F. either by boiling water or by other means. The metal should be rubbed into the grooves, as any attempt to drive it in will result in the metal straight- _ . , ? ening out from one, as it is driven in another groove. With regard to overflows, these should not be formed in sinks from which greasy matter is discharged, unless damage is likely to be caused should they overflow. In many sinks over- flows are ob J ecti able and are not required. Cast-iron sinks, either plain or enamelled, are also made, but the former have a dirty appearance, and the latter although cheaper than those of fireclay are less durable. Wash-Tubs. These fittings require very little comment- ing upon, as they resemble sinks excepting that they are deeper and have their fronts formed with a good slope. The most suitable material for wash-tubs is enamelled fireclay. Overflows to these fittings are unnecessary, and only a simple form of waste outlet is required. On account of the weight of fireclay tubs pedestals of the same material make the best forms of supports. Slop Sinks. In construction slop sinks are similar to water-closets, and their outlets should be readily accessible, and permit of a simple and effective joint being made with the branch waste pipes. A good form of slop sink is given SANITARY FITTINGS AND ACCESSORIES 1G3 in Fig. 113. In the basin either a brass or galvanised iron hinged grate is provided, upon which a pail can be placed. Hot and cold draw-off taps are often fixed immediately above the sink to enable water to be obtained for cleansing purposes. An overhead flushing cistern is essential for slop sinks, so as to cleanse them after receiving a pailful of slops. For hospital use special forms of slop sinks are necessary, as provision must be made for flushing out bed pans, etc. Urinals. During the last few years great strides have been made in the design and construction of urinals, and many defects which were common in the earlier types are mostly eliminated in the best modern fittings. The urinal is the most difficult of sani- tary fittings to keep in a satisfactory state of cleanliness, and more especially when cum- bersome and useless pieces of pottery are utilised in their con- struction. All urinals require a liberal amount of FIG. 113. Slop hopper. Hushing with clean water, as urea from the urine is readily decomposed, and gives off the well-known ammouiacal odours which are common in badly flushed and poorly constructed urinals. In private houses a w.c.' serves the purpose of a urinal, but for works, clubs, hotels, schools, and many other buildings, as well as in public thoroughfares, separate urinals are a necessity. Urinals commonly take the form of lipped earthenware basins, of enamelled iron and fireclay troughs, and of slate and fireclay stalls. Well glazed and constructed fireclay stalls make the best types of urinals, but these are usually 164 DOMESTIC SANITARY ENGINEERING AND PLUMBING more costly than other forms, excepting those where marble enters into their construction. From a theoretical standpoint, trough urinals when arranged to stand nearly full of water make good forms provided they are properly used and flushed, as the urine is diluted with a large volume of water. Trough urinals, how- ever, are often subjected to improper use, and they soon become offensive, owing to large surfaces which are wetted with urine and which receive no flushing water. Stall urinals which are formed of slate slabs do not admit of the whole of their surfaces being cleansed, and therefore this type is not suitable for installing inside buildings or in confined situa- tions. If, however, for economical considerations, slab urinals are adopted where they can be exposed to the external air, the division slabs should terminate about 18 inches clear of the channels in the floors, and be supported at the front by means of galvanised iron columns as shown in Fig. 114. The floor channels should be of glazed earthenware with an outlet at one end, or in the case of a long range the outlet may be better located near the centre. Slab urinals, Fig. 114, may either be of slate or glazed fireclay, the former being the cheaper of the two, but the latter can be more readily cleansed. The chief points to consider when selecting semi-circular or radial back urinals are : 1. That they are constructed in well glazed fireclay or similar ware. 2. That they are made in as few pieces as practicable, to limit the number of joints. FIG. 114. Form of slab urinal. SANITARY FITTINGS AND ACCESSORIES 165 3. That they contain the minimum length of channel. 4. That all surfaces which are liable to be wetted with urine are effectively flushed. 5. That the minimum number of traps be used. FIG. 115. Twyford's urinal range in glazed fireclay. G. That the joints are well formed, and that the minimum length of joints come in contact with urine. 7. That no part of a channel be difficult to cleanse or be partially obscured from view. 8. That suitable flushing apparatus, together with a liberal supply of water, be provided. Single stall urinals, provided that they are of good 166 DOMESTIC SANITARY ENGINEERING AND PLUMBING design, present very little difficulty, as they can be formed in one piece of fireclay ; but when ranges of urinals are formed, joints become imperative, and the effective discharge of the urine from the stalls becomes a more difficult matter. Joints in channels of range urinals are frequently a source of weakness, and in order to overcome this defect each stall is occasionally provided with a separate outlet and trap. Traps, however, when numerous are objectionable, owing to the volume of urine they hold, and the connections between the channels and traps are frequently of defective design. In order that one trap may suffice for one or more ranges of urinals, main channels can be arranged along the front of the stalls, and covered with movable iron or brass gratings ; a small branch or subsidiary channel from each stall joins with the main channels, and the whole of the channels are made to fall to a common outlet where a trap is provided. The chief drawback of this arrangement is the length of channel required, and unless the whole of the grates and channels are frequently cleansed by an attendant they are liable to get in a very unsatisfactory state, A good form of urinal is shown by Fig. 115, where an open, continuous channel is made in the urinal itself, and where the base of the division facings terminates clear of the channel. Inlets to traps should be covered with a domical grating to avoid, as far as practicable, the accumulations of burnt matches and other matters from choking up the grating and causing the floor of a urinal apartment to be flooded. CHAPTER Vll SOIL AND WASTE PIPES Soil Pipes. Although from a legal point there sometimes appears a little difficulty in interpreting the term "soil pipe," the ordinary individual generally understands it to mean a pipe or channel above the level of the ground, communicating with one or more water-closets and a drain ; the purpose of a soil pipe being to ' convey discharges from w.c.'s to a drain. Materials. The materials of which soil pipes are made are lead, cast iron, and occasionally copper. Solid drawn lead pipes have many advantages as soil pipes; they are very durable when properly fixed, have smooth surfaces which when properly flushed are easily cleansed ; they can be securely jointed, can be fixed in long lengths ; are easily bent to suit various situations, and contain the minimum number of joints. Lead as a material for soil pipes also has drawbacks. One of the disadvantages of lead soil pipes is their higher initial cost when compared with iron pipes, although if the durability of the former, and their intrinsic value as old material, were taken into account, lead pipes would finally prove the cheaper to use. Another disadvantage is that lead soil pipes are liable to be bent and disfigured by expansion if exposed to the direct rays of the sun. This difficulty can be overcome by protecting them, when in exposed situation, with iron shields, which are made to represent square or rectangular iron pipes, but the extra cost involved would often be prohibitive. To prevent lead pipes being distorted by expansion, expansion joints are occasionally used, but the form of joint 167 168 DOMESTIC SANITARY ENGINEERING AND PLUMBING generally adopted for these pipes is not reliable, and it is also troublesome to repair. Iron Soil Pipes possess the advantages of being self- supporting, of being easily fixed, and of being cheaper than lead pipes at the outset. The surfaces of iron pipes, however, require some form of protective coating to preserve them from rusting. Drawbacks of iron soil pipes are : they are not so durable and smooth as those of lead, special bends and .other fittings are often required, the interior surfaces of the pipes get rough owing to the decay of the protective coatings, and they contain comparatively a greater number of joints. Copper Soil Pipes are too expensive for general work, but where their surfaces are tinned, and these pipes are properly jointed, they are suitable for high class work where lead pipes may prove unsatisfactory. It is no unusual occurrence for the bottom length of an unprotected lead soil pipe to be damaged when in a position where it can be kicked, and occasionally a length of iron pipe is used at the bottom of a stack of lead pipes when the latter are likely to be subjected to improper usage. Thickness of Soil Pipes. The walls of lead soil pipes should not be thinner than 7-lb. lead, and for good work a thickness equal to that of 8-lb. lead should be used. The thickness of iron soil pipes should not be less than J inch, to enable sockets of sufficient strength to be formed. A 6 feet length of 4 in. xj- in. cast-iron soil pipe should weigh not less than 60 Ib. In Great Britain it is the usual custom where practicable to fix soil pipes on the external face of outside walls. Although this practice has much to commend it under ordinary circumstances, such positions are not conducive to the best results so far as the ventilation of drains is con- cerned. Soil pipes when fixed inside buildings are protected from the inclement weather of winter, and they may be made fairly accessible. On account of the protection afforded inside pipes in cold weather ventilation is active, where with pipes in external positions ventilation would be practically stagnant. Where soil pipes, however, are fixed inside SOIL AND WASTE PIPES 169 buildings there should be no doubt about the soundness of the work or of the quality of the materials used, and every precaution should be taken to make them satisfactory in every respect. Soil pipes when fixed on the outer surfaces of walls possess the advantages of being exposed to view, and in the case of a defect less harm is likely to be done than with a defective inside pipe. Climatic conditions also regulate the positions of soil and waste pipes. In countries where it is extremely cold in winter, with the thermometer sometimes recording zero and below, there is the possibility of outside pipes getting blocked with ice. . Arrangement of Soil Pipes and their branches. The chief points to consider when arranging soil pipes are : that their branches are suitably placed, have good pitches where practic- able, and that reliable connections can be made ; that antisiphonage pipes do not cross the soil pipes unless absolutely necessary, that as few joints as possible are buried in walls, and that all branches which join a soil pipe curve in the direction of the flow. Fig. 116 shows how the crossing of pipes may be avoided where the water-closets are located immediately in front of windows, by fixing the main soil and principal antisiphonage pipes on different sides of the windows. At A, Fig. 116, all the pipes are supposed to be of lead, whilst the main pipes and junctions in B are of iron, with lead branches passing through the walls to the fittings. When iron pipes are used rust pockets should be pro- vided on the antisiphonage pipes, with means of access to enable accumulations of rust to be readily removed. It often occurs that a straight branch between a soil pipe and w.c. requires to be fixed as Fig. 117, and although this form of branch presents no special difficulty when lead soil pipes are used, the use of iron pipes and of an ordinary short junction causes a joint to come in the wall. Where the work is properly carried out, and the soil pipes are period- ically tested, joints in walls are not so objectionable, but for general work it is better to have all joints where they can be 170 DOMESTIC SANITARY ENGINEERING AND PLUMBING readily seen and easy of access. In Fig. 117 a special iron junction with long branch is shown, together with an antisiphonage pipe connection which is shown a little on one side of the branch. Soil pipes should always discharge directly into drains, i rr A FIG. 116. Lead and iron soil pipes. that they may also act as ventilating pipes for the drainage system. The upper ends of soil pipes should terminate so as to be well removed from dormers, chimneys, and other places that afford direct communication with the interiors of buildings. When a soil pipe comes too close to a dormer it should be carried up the roof, and terminate at a suitable elevation SOIL AND WASTE PIPES 171 above the dormer. If a soil pipe should terminate near the top of a chimney, drain air will frequently be found to pass down the chimney under certain atmospheric conditions, and when no lire is burning in the grate. It is therefore obvious that unless soil pipes terminate in suitable positions that the value of sound joints, good materials, and special forms of connections are very materially nullified. In large buildings, where ranges of closets are fixed immediately beneath one another, and where long branches are required, great care is necessary in the arrangement of the pipes. A number of single closets when fixed over each FIG. 117. Special iron junction. other is comparatively simple work, but when ranges of closets are required on various floors the soil-pipe work is ' of a more complex character, and both thought and skill are essential to plan and properly execute the work. Plan and section Fig. 118 show the arrangement of the soil pipe work on one of the floors of a large building, the fittings being omitted in order to make the connections clear. The main soil pipe and antisiphonage pipe are fixed on the outside face of a wall, whilst the long branch B which intercepts the connections from the closets is situated beneath the floor. The amount of pitch for branch B is limited by the depth of the floor joints, and as the distance 172 DOMESTIC SANITARY ENGINEERING AND PLUMBING between the floor and top of the branch is also very limited, the short branches from the closets should be in the form of bends, and should join at the side of the long branch B, as shown on plan. Assuming the long branch B, Fig. 118, is fixed close to the wall, and that closets with back outlets are SOIL AND WASTE PIPES 173 used, the short branches between the long branch and the closets would be nearly straight, and would practically enter branch B at right angles. Right-angled connections would of course cause waste matter at each discharge to flow in both directions in the horizontal branch, so that solid matter would often lodge on the higher side, until it was removed by a discharge from another fitting. The arrangement in Fig. 118 allows discharges to enter the main channel in a manner that all solid matter may be removed at each flush. The branch antisiphonage pipes are all shown connecting with the bends, to which the closets are also directly joined, as this method allows reliable joints to be made. To facilitate the fixing of the pipes, and to per- mit of the greater portion of the work being executed in the workshop, flange joints may be made in con- nection with lead pipes as in Fig. 119, which shows enlarged detailed connections for one w.c. Lead collars should, of course, be slipped over the ends of the pipes before the latter are flanged over on the floor. To prevent distorting any of the pipes by opening the ends of those to be flanged, the latter should first be heated with a lamp or by other means. If an ordinary brass socket S, Fig. 119, is used for a closet connection, it may require a little cutting off its plain end, so that its lower edge will stand clear of the lead collar on the floor as shown. The sockets when treated in the manner described allow simple air-tight joints to be made, besides forming suitable fixings for the bends. It will, of course, be understood that in Fig. 118 the joists are assumed to run in the right direction, thus allowing FIG. 119. Enlarged detail in connection with Fig. 118. 174 DOMESTIC SANITARY ENGINEERING AND PLUMBING the long branch to be fixed in the manner shown. If, how- ever, the pipes cannot be fixed beneath a floor as in Fig. 118, either a raised platform would be required or closets with P traps would be neces- sary. Even assuming the latter plan were adopted, a low platform may still be essential p" ^CI to give the necessary pitch where th e branches are long. Much time is saved in this class of work and the latter is much easier to carry out if full-sized working draw- ings are made. All the bends can then be made to correct pitches, branch joints prepared, and a number of joints may be made before the pipes are placed in position. Antisiphonage Pipes. Because traps are liable under certain con- ditions to have their contents siphoned out, antisiphonage pipes are provided to preserve the equilibrium of the air pressure on the inlet and outlet sides of traps, Joints on these pipes require to be carefully prepared to enable air to iiow through them without unnecessary inter- ruption. When a soil and antisiphonage pipe are fixed on a wall fairly close together, it is unimportant whether the Anti- siphonage pipe joins the soil pipe above the highest branch, GBOUWD Level-. Connections of antisiphonage pipes ill. FIG. 120. with bends on the outside face" of a wal SOIL AND WASTE PIPES 175 or whether it is continued and terminates as by dotted line in Fig. 120. Instead of joining antisiphonage pipes with the soil pipe branches in the w.c. apartments, they are frequently connected to the bends on the outside face of a wall as in Fig. 120. This method of dealing with the antisiphonage pipes possesses the advantage of simplifying the fixing of these pipes, but it should only be adopted where the w.c.'s are fixed immediately at the back of the wall, or where very short portions of the branches are left without direct ventilation. The bye-laws of the London County Council limit the connection of an antisiphonage pipe to be not less than 3 inches and not more than 12 inches from the top of a trap, so the arrangement in Fig. 120 would not often satisfy the bye-laws in ques- tion. Although the 12-inch limit may be sufficient for many cases, there are many others where this distance could with advantage be a little increased. Fur example, take a case like Fig. 121, where it would be positively absurd to arrange the anti- siphonage pipe to join the branch nearly in the middle of the wall, as indicated by dotted lines, in order to satisfy the 12-inch limit, when a few inches farther away would allow the joint to be made in a more rational place. Sizes of Soil Pipes. No formula is necessary to calculate the sizes of soil pipes, as these are entirely governed by practical considerations. The general size of a soil pipe is 4 inches diameter, not because a smaller size is inadequate to carry away the discharges from a number of w.c.'s but because this size makes an effective outlet ventilator for a drainage system, and because it may be kept in a fairly cleanly state. Soil pipes of 3^ inches diameter are better than those FIG. 121. Lead branch and anti- siphonage pipe. 176 DOMESTIC SANITARY ENGINEERING AND PLUMBING of a larger diameter so far as the cleanliness of their inner surfaces is concerned, but for a principal outlet ventilator a 3 J -inch pipe is rather small, and especially when there is a number of bends in the stack. When a stack of soil pipes is not required to act as a drain ventilator, smaller sizes of pipes may be adopted ; but special attention must be devoted to the sizes of the antisiphonage pipes, for the better a stack of pipes is scoured with flushing water, the more readily will traps lose their seals, unless adequate means be provided to counteract this. The question is sometimes asked, how many.w.c.'s may be discharged into a stack of 4-inch pipes without causing over- flow at any of the lower fittings ? The exact number would be difficult to state, as it depends upon a variety of conditions, such as the possible number of w.c.'s likely to be flushed at the same time, the type of closet used, the amount of ob- struction offered to a discharge by bends, etc., in the pipes, and also upon the arrangement of the pipes. Thus it will be clear that only a hypothetical solution can be arrived at. The time taken to flush a wash-down closet through a 1 J-inch flush pipe with 2J gallons of water is roughly estimated at 5 seconds, and if the area of the IJ-inch pipe and that of the 4-inch soil pipe are compared, the latter will be found to be rather more than 7 times larger than the former ; thus 4 2 4 4 2 2 1 ^ T 2 = T X T X ^ X - = 7~. If it is assumed for the sake of sim- J-2 1 1 O O 9 plicity that liquid matter is discharged into and through a stack of 4-inch soil pipe with a uniform velocity, then seven w.c.'s could be flushed and their discharges could meet without quite filling the soil pipe. But as the velocity of discharge through a soil pipe varies according to the height through which the matter falls, at least one more w.c. may be safely added to the number obtained. We have now 7+1 = 8 w.c.'s which may be discharged into a 4 -inch soil pipe at the same instant. It may further be safely assumed that not more than one- fourth of the w.c.'s on a stack of pipes is likely to come into use at the same time, especially when the interval of flushing is of such limited duration. Eeasoning on these lines, we now have 8x4 = 32 wash-down w.c.'s as a number which SOIL AND WASTE PIPES 177 may safely be joined to one stack of 4-inch pipes, so far as the discharging capacity of the latter is concerned. As it is not desirable to use soil pipes larger than 4 inches diameter, an additional stack should be provided in lieu of one of an increased diameter where there are too many closets for a single stack of pipes. Although in many cases antisiphonage pipes are made compulsory where more than one w.c. is discharged into a soil pipe, the value of the former pipes has often been questioned, and by some considered unnecessary. There is not the slightest doubt that, like many other things, anti- siphonage pipes have been occasionally overdone, for when the spirit of reform has taken hold of a community, whether in sanitary or in other matters, it is generally the practice to rush from one extreme to the other. Before the ventilation of soil and waste pipes received the attention it does at present, these pipes were generally unventilated ; the result of this was, that when two or more w.c.'s were joined to a stack of pipes, the discharge from one w.c. displaced a certain volume of air from the soil pipe, and diminished its internal air pressure, and as no air inlets were provided to enable the external and internal pressures to be immediately equalised, one or more water seals, which offered the least resistance, were broken, and the requisite amount of air was admitted from that source. But where two or three wash-down w.c.'s are fixed over one another on a stack of 4-inch pipes, which are carried up full bore to the roof for ventilation, the removal of the water by siphonage from any of the traps may not readily occur by a single flush, owing to the freedom with which air may enter and make good that displaced by the falling body of water. It is from such a simple case as this that the opponents of antisiphonage pipes principally draw their conclusions. ^t Numerous experiments have beeu conducted from time to time to show how traps may have their contents removed by siphouage and by other means, but many of these experi- ments possess little real value, as they are often conducted under very limited and unreal conditions. Whether the water seals of traps will be broken by siphonage or not 12 178 DOMESTIC SANITARY ENGINEERING AND PLUMBING will depend principally to what extent the outlet pipes are charged, the velocity of discharge, and upon the provision for ventilation. Taking the case of a 4-inch stack of soil pipes which receive discharges from say four w.c.'s, if the stack is con- tinued full bore for ventilation, is free from bends, and the wash-down type of w.c. is used, the flushing of the topmost w.c. may not affect the water seals of those below when no antisiphonage pipes are provided. If, however, a stack is high, and contains a number of bends, and two w.c.'s are flushed at the same time, it becomes quite possible for the combined discharges to remove the seals from the lower fittings. Where a stack of soil pipes is not provided with anti- siphonage pipes, there is often a risk of the lower traps losing their seals by discharging a pailful of slops quickly through one of the higher fittings. The top w.c., of course, on a stack of pipes does not require an antisiphonage pipe in connection with it, but where the main antisiphonage pipe is joined with a soil pipe the junction should, as a rule, be made above the highest fitting. Sizes of Antisiphonage Pipes. The size of these pipes should be governed by the general arrangement of the pipes, and by the type and number of closets that are joined to one J stack of soil pipes. Wash-down w.c.'s admit of the smallest sizes of antisiphonage pipes being used, as the discharge in leaving these closets is more prolonged than with either the J valve or siphonic types. The object to be attained in all cases is to have antisiphonage pipes of a size that will allow the necessary amount of air to pass through them, and at the same time to prevent undue air tension in any branch when a volume of water is being discharged. There is no simple formula known to the writer for determining the size of antisiphonage pipes ; judgment, along with a knowledge of falling bodies and the flow of fluids through pipes, appears to be the most reliable guide. When water is discharged from a w.c. into a soil pipe it is a common error to assume that it falls through the SOIL AND WASTE PIPES 179 latter in the form of a solid plug. It will, however, be found that a vertical soil pipe is only partially filled, and that a discharge, with the exception of excrement and paper, etc., chiefly follows the surface of the pipe. If it is assumed that 3 gallons of water are discharged into a stack of soil pipes of indefinite height in four seconds, then according to the law of falling bodies the first particle of water, if uninterrupted in its passage, would have fallen through a vertical distance of 256 feet by the time the last particle was ready ar-* to fall. In other words, the 3 gallons of water near the end of the discharge would be spread over a length of 256 feet of pipe. As resistances are encountered by falling water in a soil pipe, the height, of course, would be much less than that given. The length of 4-inch pipe which will hold 3 gallons of water is about 7 ft. 4 in. Of course the height of a soil pipe is limited, but the example simply serves to show the small area often occupied by water when the latter is falling through a pipe. Fig. 122 represents four w.c.'s which are fixed above one another and discharge into one stack of pipes ; the antisiphonage pipes are all shown to be 2 inches diameter, and this size would be ample where wash- down w.c.'s were used. If siphonic w.c.'s were fixed, the main antisiphonage pipe might be increased to 2J inch diameter for the upper half of ibs length. It is seldom desirable to use branches for antisiphonage pipes smaller than 2 inches diameter, on account of smaller sizes being more easily choked. The effect of the arrangement of soil pipes on the sizes of antisiphonage pipes is illustrated in Figs. 123 and 124. If the pipes are arranged as Fig. 123, and a volume of water is discharged from two or more of the upper fittings, the air to FIG. 122. Sizes of soil and of venti- lating pipes. 180 DOMESTIC SANITARY ENGINEERING AND PLUMBING replace that extracted from a lower horizontal soil pipe branch would require to enter through the branch antisiphonage pipes. This arrangement may necessitate the lower horizontal antisiphonage pipe branches being about 3 inches diameter. In Fig. 124 the hori- zontal soil pipe branches are shown continued and joined with the main antisiphonage pipe, and as air tension can ~$ be directly relieved in any of the principal branches by this means, all the hori- zontal antisiphonage pipes may be of smaller bore than those given in Fig. 123; neither would it be impera- tive to ventilate the branches of the w.c.'s nearest the main antisiphonage pipe, as shown in Fig. 124, provided these branches are only short. For very high stacks of 4" soil pipes which have a large number of closets con- nected with them, it may be found desirable in a few very special cases to make the principal ventilating pipe a little larger than 4 FIG. 123. Diagram illustrating the effect i nc hes diameter; but this the general arrangement has upon sizes of ventilating pipes. will greatly depend upon the arrangement of the pipes. So far as the arrangement of the soil pipe branches is concerned, that in Fig. 123 is better than that shown in Fig. 124, owing to the possibility in the latter case of matter lodging at the higher ends of the horizontal branches. In Fig. 123 the end w.c. would tend to keep the horizontal branches clear of deposit, SOIL AND WASTE PIPES 181 Unsealing of Traps. There are various ways by which a trap may lose or partially lose its water seal, such as by siphonage, by momentum, by capillary attraction, by evapora- tion, by waving out, and by the water being blown out of a trap. As already explained, siphonage is produced when the outlet water surface of a trap is subjected to a less pressure than that on its inlet side, and to prevent siphonage, all that is neces- sary is to maintain equi- librium of the air pressure on the inlet and outlet sides of a trap. Unsealing by momen- tum has reference to where a volume of water in flowing through a trap encounters insufficient resistance to prevent enough water re- maining in the trap at the end of the discharge. Traps in connection with slop sinks, or other fittings where water is discharged rapidly through them, are liable to be unsealed by momentum, unless their outlets are flattened so as , T, i ,, tO retard a little the Out- . 124. Diagram illustrating the effect the general arrangement has upon sizes of ventilating pipes. flowing water. Traps for baths, ordinary sinks, and for similar fittings, are not liable to be unsealed by mom- entum ; neither are the traps of ordinary lavatories, or those of wash-down w.c.'s when the latter are flushed in the ordinary manner. When unsealing by momentum requires to be taken into account, the best form of trap to use is the 182 DOMESTIC SANITARY ENGINEERING AND PLUMBING anti-D type. An antisiphonage pipe is not a cure for loss of seal by momentum, and the remedy lies in the provision of a suitable trap. Loss of seal by capillary attraction. When traps are not self-cleansing, fibrous matter may hang over the outlet side of a trap and remove the water from it by capillarity. The liquid rises through the interstices of the matter, and the smaller the interstices the greater the height to which the liquid will rise. As the seals of traps are comparatively small they are readily broken by capillary attraction. To remedy and to avoid this evil all that is required is a self -clean sing form of trap. Unsealing by evaporation. In this country the water seal of a trap is not readily removed by evaporation, and only where fittings are out of use for a comparatively long period, or located in heated apartments, is it likely to lose its seal by this agency. The atmosphere, except when in a saturated state, is constantly taking up moisture from any available source, so that unless traps are replenished periodically with water their seals will be eventually broken. For houses which are closed for long periods during hot weather the water seals of traps may be maintained for a much longer time by putting oil on their water surfaces. A syringe with a bent tube could be used for forcing oil to the outlet side of a trap, but very few people will take this trouble. Unsealing by waving out. Where the wind can blow directly across the end of a pipe, or down it, the air inside in the first case is under tension, and in the latter case it is subject to more or less compression according to the force of the wind. The partial unsealing of the traps of w.c.'s is very common where the upper ends of soil pipes are left plain and unprotected in any way, and where no fresh air inlet is provided for the drainage system, or where the inlet is temporarily closed. The sudden compression of the air in a soil pipe by a gust of wind depresses the water level on the outlet side of a trap, with the result that when the force is spent a portion of the water washes over the outlet in regaining its normal level by the motion imparted SOIL AND WASTE PIPES 183 to it. A similar effect may be produced when the air in a soil pipe is rarefied by a strong current passing over its open end, although this may be termed a case of siphonage. Unsealing by the blowing out of water. The traps unsealed by this means are usually those which are situated at a low point, and where the air in a system of pipes is put in compression by a falling body of water. In the case of soil pipes which are connected directly with drains, and where the latter have fresh air inlets which are either choked or automatically closed by an outward rush of air, a discharge from a w.c. at a high level is often sufficient to compress the air in the drainage system for a brief interval ; under such conditions relief is obtained by blowing out the water from one of the traps. The unsealing of traps in this manner can be prevented by making provision for a free flow of air in either direction. Waste Pipes. The arrangement of waste pipes is to a great extent the same as for soil pipes, the chief difference being that waste pipes are generally disconnected from foul- water drains by means of traps, whilst soil pipes are connected directly with a system of drains. The waste pipes from slop sinks and urinals are, however, exceptions to the general rule, and are treated in the same manner as soil pipes. In America and other countries where climatic conditions necessitate the soil and waste pipes being fixed inside build- ings, in order to protect them from frost, separate soil and waste pipes are not generally used, but all the different sanitary fittings which are located near each other discharge into a pipe common to the whole. Through British spectacles such an arrangement is often thought to be a retrograde one, but when reasoning to a logical conclusion it will be found sound in principle, provided "pipes and traps of suitable materials are used, the workmanship is all that can be desired, and special attention is paid to the ventilation of the different pipes. Where it is essential to fix soil and waste pipes inside buildings, they are as well joined together at a high elevation under ordinary conditions as to be carried down separately into a basement and there connected with a 184 DOMESTIC SANITARY ENGINEERING AND PLUMBING common drain, provided the essential conditions of good workmanship and suitable materials, etc., are fulfilled. In Great Britain, where waste and soil pipes are fixed outside buildings, their separation is quite defensible, and where soil pipes are of lead it is practically imperative that hot discharges of waste water flow through separate channels to a drain. Bath and Lavatory Waste Pipes. The lower portion of a stack of waste pipes is shown in Fig 125, where a bath and lavatory are supposed to be fixed close together on each of two floors. The main waste pipe is supposed to be of cast iron, and the branches of copper with brass fittings. All the antisiphonage pipes are of lead, and are shown inside the building ; these pipes should be carried up and terminate at a high level, or be joined with the main waste above the highest branch. For good buildings the smaller waste pipes should be of copper when discharging hot w r ater, and especially when the branches are long. The first cost of copper waste pipes is rather high, but they can be relied upon and are durable. The branch wastes should not be too rigidly fixed, but arranged that they may move in the direction of their length. Sufficient space for movement may often be ob- tained with one or more bends, but if a branch is required to be fairly straight between two fixed points, and is of moderate length, an expansion joint may be provided near the trap as at A, Fig. 125. In copper waste-pipe work a fair number of union connections is necessary, but in many cases a simple form of brazed socket joint can be adopted. Sometimes antisiphonage pipes are also of copper, and these have a smart appearance when they are lacquered and kept bright. Antisiphonage pipes, however, are not sub- jected to the same strain as waste pipes, and lead pipes are satisfactory so far as durability is concerned, and also have the merit of being readily fixed. Sizes of Waste Pipes for Baths and Lavatories. The waste pipes of both baths and lavatories should be of a SOIL AND WASTE PIPES 185 reasonable size, so that these fittings may be quickly emptied, and the discharge of water be of service for aiding in the cleansing of the drains. A waste pipe from a bath should FIG. 125. Arrangement of waste and of ventilating pipes in connection with baths and lavatories. not be less than 2 inches diameter, whilst that from a lavatory should not be smaller than H inches diameter. Outlets from fittings should not be smaller than the waste 186 DOMESTIC SANITARY ENGINEERING AND PLUMBING pipes used, otherwise the latter will not be properly cleansed, and may gradually choke up. The size of a main stack of waste pipes should be about 3 inches diameter, and in special cases a size larger may be necessary. FIG. 126. Arrangement of waste pipes for a range of lavatories. ^ Arrangement of Waste Pipes. Figs. 126 to 128 show three different arrangements of lavatory waste pipes where the whole of the branch waste pipes and traps are supposed to be of lead ; the main stack of pipes in each case is of iron. Where a trap is placed beneath each lavatory the method FIG. 127. Arrangement of waste pipes for a range of lavatories. of arranging the branches in Fig. 127 simplifies the work in connection with the antisiphonage pipes, and the long branch wastes may either be fixed above or below the floor. Pro- vision should be made by means of thumb-screws for cleansing out the long branches should a stoppage occur at any time. SOIL AND WASTE PIPES 187 In Fig. 126 the main antisiphonage pipe is shown on the inside face of the wall, but it may be placed outside if desired. Instead of having a separate trap under each fitting, occasion- ally a principal branch waste only is trapped, as in Fig. 128. The latter method is a cheap and simple one when compared with those of Figs. 126 and 127, and the short branches from each lavatory may be readily arranged to join at the side of the inclined waste pipe as in the figure given. It is always desirable, where practicable, for branches of either soil or waste pipes to join the side instead of the top of a nearly horizontal pipe ; side connections retard to a less extent the velocity of the discharging liquid, and the pipes are better flushed near the points of junction. FIG. 128. Arrangement of waste pipes. As ranges of lavatories are required in large offices, schools, and other large buildings, it is essential that every precaution be taken to prevent foul air being emitted from them, and so proving injurious to the health of any individual. Under most conditions a trap should be fixed immedi- v ately beneath each lavatory where a number of lavatories j are grouped together ; this arrangement of the trap exposes the least amount of fouled pipe surface to the atmosphere of the apartments in which the fittings are placed. When a range of lavatories is fixed in a well-lighted and detached building, which has ample permanent ventilation, and a compromise is sought between expenditure and hygenic considerations, the method of fixing the waste pipes as shown in Fig. 128 is frequently adopted. 188 DOMESTIC SANITARY ENGINEERING AND PLUMBING Another method of dealing with a group of lavatories, is to let each fitting discharge by means of a short pipe into a glazed fireclay channel immediately beneath them, the channel being laid to drain to a trap which is placed at any suitable point. This method of collecting the discharges from a range of lavatories is a very simple one, but from a health point of view it is bad in principle, for if the channels are neglected they soon get in a filthy state ; the air which circulates through the short discharge pipes is also polluted by contact with their surfaces. Sizes of Pipes for Ranges of Lavatories. Single branches, as previously stated, should have a minimum size of 1-| inch diameter; the horizontal branches should be from 2 to 2J inches diameter, according to the number of lavatories in a range. When the pipes are fixed as in Fig. 126, the short branch antisiphonage pipes which are joined with the traps may be 1 inch diameter ; a suitable size for the horizontal antisiphonage pipe would be 1-J- inch diameter, which may be increased to 2 inches diameter beyond the eight lavatory. In Fig. 127 each trap is branched into a short length of 1-J-- inch pipe, the upper end of which forms the antisiphonage pipe for each trap ; with this exception the remaining sizes would be the same as for those in Fig. 126. Slop Sink Waste Pipes. Where discharges of hot water are allowed to flow through slop sinks, the whole of the waste pipes should be of iron or other hard, suitable metal. Special branches and connecting pieces should also be used to enable reliable joints to be made with the slop sinks. Owing to the contents of a pail being quickly discharged through these fittings, it is essential that special attention be paid to the sizes of the antisiphonage pipes where two or more slop sinks discharge into the same stack of pipes. A suitable size for branch waste pipes is 3 inches diameter, and for a main stack 3 to 3-| inches diameter. For a four storey building, where a slop sink is fixed on each floor, the branch antisiphonage pipes to each fitting may be 2 inches diameter, and for the principal antisiphonage pipe 3 inches diameter. Waste Pipes for General Sinks. Waste pipes for these fittings are often subjected to unfair usage, and although lead SOIL AND WASTE PIPES 189 waste pipes and traps are serviceable for many sinks, there are other cases \vV.cre much stronger pipes and traps are essential. In places where waste pipes are likely to be subjected to rough usage they should be either of cast iron or brass or of galvanised wrought iron. In ordinary dwelling and business houses short lead waste pipes answer admirably, but they often cause trouble when of considerable length. FIG. 129. Waste pipes for sink. Of course if very hot water is not discharged through them, long lead waste pipes are also durable. Fig. 129 gives a waste pipe in connection with a scullery sink where the former discharges directly into a gully trap. The Model Bye-laws of the Local Government Board require a waste pipe to discharge on to an open channel which leads to a trapped gully grating at least 18 inches distant. The method suggested by the Bye-law is not a satisfactory one, 190 DOMESTIC SANITARY ENGINEERING AND PLUMBING as the open channel frequently gets filthy, and the force of the discharging water is destroyed. No overflow is shown in Fig. 129, for, as stated elsewhere, sinks are better without them. When a number of sinks discharge into a stack of waste pipes, and antisiphonage pipes are necessary, the latter should either terminate in the external air above the sinks, well removed from windows and ventilators, or be treated in a manner similar to that shown in Fig. 125. RUST Box. FIG. 130. Rust pockets. Rust Pockets. When iron pipes are used for purposes of ventilation, provision should be made for the interception and removal of rust. For fixing at the foot of a ventilating stack, a rust pocket similar to A, Fig. 130, is suitable, whilst the arrangement at B, Fig. 130, may be adopted where bends are necessary in iron ventilating pipes. Traps for Waste Pipes. These traps are formed in different ways, and some of the types are very defective, being deficient in seal and constructed with sharp angles which retain filth. Unless traps are of a good shape they SOIL AND WASTE PIPES 191 unduly retard the outflow of water, and are liable to be frequently choked. The water seals of traps for waste pipes should not be less than 1-J- inches deep, and as a rule they should not exceed 2 inches deep for the large size of traps. Three different forms of traps are given in Fig. 131, those at A and B being the best types at present in use. The outlet end of the siphon trap A takes different shapes to suit various situations, and this form of trap may also be obtained with long, straight outlet limbs when desired. A siphon trap is readily unsealed by momentum, but as this L..J FIG. 131. Forms of traps. loss is chiefly confined to one when fixed in connection with slop sinks, siphon traps are suitable for most of the remaining fittings. At B, Fig. 131, Hellyer's anti-D trap is given ; the throat of the trap is restricted in area, and this enables it to be readily cleansed. In cross section its outgo is nearly square, the corners being rounded a little instead of being left sharp. Owing to the shape of the anti-D trap it requires to be cast, and although it is a little dearer than a drawn lead trap it is much stronger. The type of trap at C, Fig. 131, can only be described as a poor one. It usually has too shallow a seal, the depth of 192 DOMESTIC SANITARY ENGINEERING AND PLUMBING which can neither be seen nor readily ascertained ; the water passages are of a poor form, and the principal changes of direction are too "abrupt. Traps which have movable parts in the form of flaps or balls are not suitable for fixing to waste pipes ; they are not reliable, and neither are they self-cleansing. CHAPTEE VIII DRAINAGE OF HOUSES AND OTHER BUILDINGS A DRAINAGE system should be designed upon sound prin- ciples, and constructed in a manner to prevent its being the cause of the pollution of the subsoil and a carrier of polluted air back to and into buildings. Numerous cases of typhoid fever and other illnesses many proving fatal have been due to sewage polluted water through defective drains. Sewer and drain air may also be the means of disseminating the germs of disease. It is not intended here to deal with the different forms of old brick and stone built drains, these being chiefly of historical interest, but to devote the space at disposal to modern drainage work. The value of a good drainage system is now generally appreciated, and for many buildings cast-iron drains are rapidly superseding those of earthenware for the conveyance of foul liquid matter. The substitution of iron for earthenware drains has chiefly arisen through the difficulty of maintaining the latter in a sound state for any great length of time after being laid and covered in. Very often it has been found that fine cracks have occurred at the sockets of earthenware pipes, due to the expansion of the jointing material, even when care has been exercised in carrying out the work. When joints are made with portland cement the pipes are very rigid, and if unequal settlement of the ground takes place the pipes readily fracture on account of their unyielding nature. When earthenware drains are laid, as they frequently are, with unskilled labour, and jointed with portland cement, there is little wonder that pipes and joints are readily broken, 194 DOMESTIC SANITARY ENGINEERING AND PLUMBING especially when defects occur in work which has been well supervised and executed by experienced men. As the failure of earthenware drains is frequently due to the use of portland cement as the jointing material, other substances, such as bituminous cements, which possess some degree of elasticity are now often used. In ground of moderate firmness, and which is not subject to frequent vibration, earthenware drainage work may be satisfactorily carried out, provided a suitable elastic jointing material is used, and provided, further, that the pipes are of good quality and are properly laid. The advantages of earthenware drains are smoothness, freedom from corrosion, and the durability of the material. Iron drains possess the merits of greater strength, reduced number of joints, and they can be made to remain air and water-tight for long periods after being laid. Spigot and socket joints should be used for iron pipes, and when the former are caulked with metallic lead they will yield a little should any slight settlement take place. Iron drains may occasionally be of a smaller diameter than those of earthenware, as the former may be obtained in a larger range of sizes, and may flow full and under pressure in suitable situations. The chief drawbacks of iron drains are, their inner sur- faces are not so smooth as those of earthenware, they are subject to corrosion, and therefore have a limited life. The initial cost of iron drainage work is also greater. When dilute acids are frequently discharged into a system of drains, earthenware pipes and fittings are essential, as iron would be rapidly corroded. It is sometimes contended that an iron drain in connection with a residence is liable to be attacked by a periodical discharge of dilute acid, such as may take place when spring cleaning is proceeding. An iron drain, however, is not likely to be appreciably affected at such times, as the acid used would be in a very diluted state when it reached the drain, and the greasy surface of the latter would offer sufficient protection to the metal in most cases. The minimum size of underground drain that is generally used is 4 inches diameter, but in many cases a 3-inch drain could be advantageously adopted for many branches. So far DRAINAGE OF HOUSES AND OTHER BUILDINGS 195 as the capacity of a 4-inch drain is concerned, it is frequently capable of discharging a volume several times greater than it will ever be required to discharge. When rain-water is kept separate from foul-water drains, a 4-inch main drain would be large enough to dispose of the waste discharges from a very large building. Definitions. Foul water drains are generally under- stood to be those which receive the discharges from any sanitary fitting in a building, waste water from wash-houses, and dirty waste water which is discharged into gully traps. Kain, or clean, water drains are those which receive rain- water directly from the roof of the buildings, and also subsoil water. Drainage Design. The chief points to consider when arranging a system of drains are : 1. That they are laid with self -cleansing gradients. 2. That they are arranged in straight lines between fixed points, with true alignment of inverts. 3. That principal junctions and changes of direction are made in inspection chambers, to enable any part of a system to be readily accessible. 4. That every part of a system is adequately ventilated. 5. That the main drain is disconnected from the sewer. 6. That all levels be correctly obtained. 7. That after completion the drains will be capable of remaining water and air-tight. 8. That materials are of the best of their respective kinds. 9. That all drains are laid outside buildings where practicable. 10. That all unnecessary traps are avoided. 11. That the size of drains are proportionate to their requirements. 12. That all air inlets and outlets are located in positions well removed from windows and other places that afford a direct passage for drain air into buildings. Fig. 132 represents the block plan of a detached villa residence, and a method of arranging the drains is also shown. Although the whole of the rain and waste water discharge into 196 DOMESTIC SANITARY ENGINEERING AND PLUMBING oue system, the length of foul-water drain is kept as short as practicable, and the whole system admits of good ventilation. SR INDEX. 5.R = 5oil Pi]ae. WP. Waste, Pipe. QT. GulleyTraja. R.W.P. Rain-Water Pipe. V.R Ventilating Pipe. 5.W.R 5inK Waste ?i|De. D.T. T- Disconnecting Trap. R.P/.P. FIG. 132. Drainage plan of a detached house. The disconnecting trap No. 1 cuts off a length of earthen- ware rain-water drain from the foul-water system, and the rain-water pipes in this case discharge directly into the branch DRAINAGE OF HOUSES AND OTHER BUILDINGS 197 drains. When a drain conveying clean water is disconnected from the foul-water drains as shown, a trap is not required at the foot of a rain-water pipe. In Fig. 132 there are two soil pipes shown, and it will be observed that each of these is arranged to come at the head of a section of foul-water drain, and to act as air outlet ventilators. The branch drain into which the kitchen sink discharges is ventilated by a special ventilating pipe, which is indicated on plan. It is intended that the kitchen sink waste shall discharge into an ordinary gully trap, as a grease trap is not often necessary for a villa of the size shown. Where a waste pipe and a rain-water pipe are near each other, they may both discharge into the same trap. All the chief lengths of drains in Fig. 132 are readily accessible by means of chambers, which are provided at the principal turnings and branches. A little consideration will decide the best positions for chambers, as the expense they involve prevents their general use at all junctions and turnings. Drains should not be laid close to walls if it can be avoided, and where they pass through them provision should be made to prevent the pipes taking any of the weight of the walls should any settlement take place. In large buildings, and in terrace houses, where a drain discharges into a sewer in a front street, it is often imperative to lay drains through the buildings. In such cases ample provision should be made to enable a stoppage to be removed without interfering with the floors of these places. Drains under floors should be of heavy section cast iron but no concrete foundations are necessary for these pipes provided the ground is moderately firm on which they are laid. Foundations for Drains. When laying earthenware drains, concrete is often desirable, and in some cases its use is essential. Fig. 133 shows three different ways in which concrete may be used. At A an earthenware pipe is shown laid in ordinary firm ground, where fine concrete is used for packing at the sides of the pipes after the latter have been jointed and tested. A fairly wide concrete foundation is indicated at B, Fig. 133, for poor ground, fine concrete being used as before for packing 198 DOMESTIC SANITARY ENGINEERING AND PLUMBING at the sides of the pipe. At C the drain is surrounded with con- crete. This form of construction is necessary when earthenware drains are laid in deep ground, to take the weight of the earth, and also where they are laid near the surface of the ground which is subjected to heavy traffic passing over it. When foul- water drains are of earthenware, and laid inside buildings, they should be well surrounded with concrete. The covering of drains with concrete adds greatly to their cost, so that where con- crete is freely used the differ- ence in cost between iron and stoneware drainage work may not be much. It is necessary when laying pipes that they firmly rest on the ground for the whole of their length, for if pipes simply rest upon their sockets they are liable to be fractured by the superincumbent earth. Connections with Drains, When a foul-water drain passes near the foot of a stack of rain- water pipes, the latter frequently discharge directly into a gully trap, which is joined with the drain ; if, however, a rain-water pipe is some distance from a foul-water drain, it should be treated as in Fig. 134. Here the rain-water pipe discharges into an access bend, an iron grid being provided to enable air to flow freely in or out. The disconnecting trap should be fixed as closely to the foul -water drain as possible, in order ^ - = - ; O *J :. wetted surface acb = '522 feet. As the hydraulic mean depth r is found by dividing the area of flow by the wetted surface, then r = = -0735 feet. DRAINAGE OF HOUSES AND OTHER BUILDINGS 239 To facilitate calculations being made in connection with drainage work, the following table is given : TABLE III. DATA FOR OBTAINING HYDRAULIC MEAN DEPTH, AND THE SECTIONAL AREA OF FLOW IN CIRCULAR DRAIN PIPES, WITH WATER FLOWING AT DIFFERENT DEPTHS. Depth of flow. Hydraulic mean depth r. Sectional area of flow. Full Dx -25 D 2 x 7854 i D x -296 D 2 x -632 D x -292 D 2 x -556 Dx -25 D 2 x -393 Dx-186 D 2 x -229 i Dx-147 'D 2 x-154 D= diameter of drain in feet. Bends and changes of direction retard the velocity of flow to a great extent, and the quicker a bend the greater the resistance offered. When open channel bends are used which permit of a discharge splashing over them, the velocity will be further retarded by them. Smoothness or roughness of a surface also has its effect, and it is fairly obvious the smoother a pipe surface, other conditions being equal, the smaller frictional resistance will be. Formulae for obtaining the velocity of discharge through drains are very numerous, but one of the best is that by Kutter. Where v c Jr x s (6) v = velocity in feet per second. r = hydraulic mean depth in feet. s = sine of inclination = fall length. c = a coefficient which varies with the size and condition of a pipe. 240 DOMESTIC SANITARY ENGINEERING AND PLUMBING For smooth drain pipes 181+: 00281 1 + 41-6 + 013 T \/r For general work the values of c take too long to work out by the formula given, but by the aid of Table IV., which gives values of c for varying depths of flow, the general formula is rendered convenient. TABLE IV. Diameter of drain in inches. Values of c (calculated by the writer). Depth of flow in drain. Full. | full. i full. i full. | ! 8 9 66 68 71 74 77 80 68 72 75 78 81 84 66 68 71 74 77 80 54 57 60 63 66 69 It will be observed, upon reference to Table IV, that the values of c increase with the diameter of a pipe, and are highest when a drain is flowing three-quarters full and lowest when running only one-quarter full. It is therefore of importance when deciding upon a suitable gradient that attention be paid to the probable normal depth of flow. Example 2. Find the gradients which will give a velocity of 3 feet per second in a 6 -inch drain when flowing J and full respectively. By transposing Formula 6 and substituting _ for s we have / Where 1 = length of drain in feet. h = vertical fall in feet for the given length. c, r and v as before. DRAINAGE OF HOUSES AND OTHER BUILDINGS 24 1 Upon reference to Table III. the hydraulic mean depth r = "25 x D = '25 x *5 when a 6-inch drain is running J full, and 147 x D = 147 x '5 when flowing J full. The values of c from Table IV. for the same depths of flow in a 6-inch drain are 71 and 60 respectively. When a 6 -inch drain is flowing J full the length of drain per foot of fall to give a velocity of 3 feet per .second is found by Formula 7. Where l = c X ^ x . 60 2 x -5x147x1 Substituting values, / = 02 > 7 _ 3600 x -5x147. q .-. Z = 29-4, or say 30. And gradient necessary for a 6 -inch drain to give a velocity of 3 feet per second when flowing only J full = 1 in 30. When flowing J full 1 = 71 2 x "5 x '^5 x 1 Substituting values, / = - , 7 _5041x-5x-25 ~9~ ~ ; /. ? = 70. For this case the necessary gradient will be 1 in 70. The calculations clearly show to what extent the depth of flow has upon the velocity in the same pipe, and a com- paratively quick gradient is necessary when a 6 -inch drain only flows one quarter full. For short drains where a discharge enters with a high velocity, rather flatter gradients could be used, but for long drains the formula given is a safe one to use. When the velocity of flow has been determined, the dis- charging capacity of a drain in gallons may be ascertained by multiplying the sectional area of flow by the velocity and afterwards by 6J. Expressed as a formula 16 242 DOMESTIC SANITARY ENGINEERING AND PLUMBING G = Ax^x6J (8) Where G = gallons discharged per second. A = sectional area of flow in feet. v = velocity in feet per second. J} 6J = volume in gallons equivalent to one cubic foot of water. Example 3. Find the discharge in gallons per minute of a 6-inch drain when flowing f full and when laid with a gradient of 1 in 50. By Formula 6, v = c *Jr x s. The value of s = 1 -r 50 = '02. From Table IV. the value of c will be found to be 75, and from Table III. r = Dx -296. Substituting values, we have '5 x -296 x -02. .*. = 4'05, ft. per second. The discharging capacity of the drain may now be found by Formula 8, where G = A x v X 6 J. In Table III. the sectional area of flow when a drain is | full = D 2 x '632, and for a 6-inch drain = '5 2 x '632. Substituting values G = *5 2 x '632 x 4'05 x 6 J ; /. G = 3'999, say 4 gallons per second, and the discharge per minute = 4x60 = 240 gallons. For general work the gradients given in the following table are suitable for short lengths of drains. TABLE V. . DIAMETER DRAIN. GRADIENTS. 4 inch branch from yard gully . . 1 in 24 - 1 inch in 2 ft. 4 inch main drains . . . 1 in 36 = 1 , in 3 ft. 1 in 42 = 1 1 in 48 = 1 1 in 60 = 1 1 in 78 = 1 1 in 102 = 1 in 3 ft. 6 in. in 4 ft. in 5 ft. in 6 ft. 6 in. in 8 ft. 6 in. CHAPTER IX DISPOSAL AND TREATMENT OF SEWAGE FROM MANSIONS AND HOUSES IN COUNTRY DISTRICTS IN rural districts it is often necessary, on account of the absence of a general sewerage system, to make special pro- vision for the treatment and disposal of the sewage from large isolated buildings, or from a group of small dwellings. Different methods of dealing with sewage can be adopted, but the choice of a system largely depends upon local conditions and the amount of money the owner of a property is prepared to spend. The following methods of sewage disposal and treatment are in general use for small or private works. 1. Cesspools which overflow on to land or into some available stream. 2. Bacterial systems of purification. (a) Absorption and utilisation of sewage by means of sub-irrigation, and with or without preliminary treatment. (b) Treatment in septic tanks, and subsequent treat- ment of the effluent on land. (c) Treatment in septic tanks, and the subsequent passage of effluent through contact beds or percolating filters. The first method is usually a very unsatisfactory one, on account of the nuisance caused when emptying cesspools, the frequent pollution of the soil and underground water supplies by defective construction, and the pollution of streams by overflowing sewage. Cesspools always possess objectionable features, even when soundly constructed. If they overflow on to land a rank 243 244 DOMESTIC SANITARY ENGINEERING AND PLUMBING grass quickly grows on the area dosed, and their emptying is frequently delayed on account of the objectionable nature of the operation. When cesspools are used they should be well ventilated and of water-tight construction ; their size depends upon the daily amount of sewage they are likely to receive, whether rain-water is admitted into them or not, and the length of time before emptying. Generally speaking, the greater volume of the rain-water should be excluded from foul-water drains which either dis- charge into a cesspool, or into a tank in connection with a bacterial system of sewage treatment. In a bacterial system the purification of the sewage is accomplished by minute living organisms, which, when given suitable conditions, thrive, and carry out their work. The two chief classes of bacteria which are responsible for the purification of sewage are aerobic and anaerobic ; the former require a liberal supply of oxygen for their development, whilst the latter only thrive in the absence of oxygen. For a small scheme of sewage purification to be a success, it must be of simple construction and automatic in action : for if frequent attention is necessary the chances are that a system will be neglected, and sooner or later it will result in failure. The simplest and most effective way of rendering sewage innoxious, and to purify it, is by passing it on to land which is of a good loamy nature. In the upper layers of earth bacteria are present in considerable numbers, and these attack the sewage, break it up into simple and harmless constituents, and in a form that may readily be assimilated by plant life. On account of the numerous bacteria which are present in the upper earth, the latter has been termed the " living earth," and nearly all the purification effected by a soil is done within the upper three feet ; below this depth little purification is effected by a soil, and a depth is soon reached where organisms do not appear to exist. System of Sub-Irrigation. Under favourable conditions a system of drains may be arranged so as to distribute sewage that it can be absorbed by a given area of land. The drains, DISPOSAL AND TREATMENT OF SEWAGE 245 however, should be kept as near the surface as practic- able, and the sewage re- quires to be discharged so that every part of the prepared area receives its quota of sewage. In order to effect uniform distribu- tion of sewage, a given volume must be discharged at one time ; this method prevents the overdosing of isolated parts of an area, and allows the bacteria to better perform their work. A simple sub -irrigation scheme of sewage treat- ment is given in Fig. 170. It consists principally of a septic tank S, a dosing chamber T, and a number of open-jointed field drains which are placed a certain distance apart on one or both sides of a main dis- tributing pipe. The pur- pose of the septic tank is to liquefy as far as possible the organic solids in the sewage, in order that the latter will be in a better condition for percolating into the earth. The septic tank is shown provided 246 DOMESTIC SANITARY ENGINEERING AND PLUMBING with submerged inlet and outlet, and a wall W should be constructed near the inlet of the tank to prevent its contents being unnecessarily disturbed by a sudden inrush of water. In a small septic tank a screen may be also arranged at the outgo 0, to prevent undigested solid matter passing into the dosing tank T. When sewage is first turned into a septic tank, a certain time must elapse (largely depending upon weather conditions) before the tank properly performs its work ; in other words, before septic conditions arise. A septic tank when once in working order should not be emptied, but provision should be made for removing any irreducible matter which accumu- lates in it. In the tank T a plenum automatic siphon is shown for discharging its contents into the subsoil drains; any other suitable contrivance may be used for the purpose, but anything that depends upon movable parts is usually more liable to get out of order. To the outgo of the siphon a pipe P is joined, which serves the purpose of an overflow for the tank and also as a ventilation pipe for the field drains. The subsoil drains require to be carefully arranged, and a good porous soil is essential if the system is to be a success. The main subsoil drain, which may be of ordinary 4-inch pipes, should be laid level and not deeper than 1 foot below the surface of the ground; branches should be arranged as in plan Fig. 170, that each may receive its proper volume of sewage. The open jointed branch drains should also be fixed level, and a suitable length per person is from 40 to 45 feet. The distance between the distributing branches should be regulated by the character of the soil, and will vary from 2J to 5 feet. With regard to the sizes of the septic and dosing tanks, the former should have a capacity of about one day's flow and the latter about half a day's flow of sewage. To a great extent the capacity of the dosing chamber should be governed by the capacity of the subsoil drains. From dwelling-houses where modern sanitary conveniences are in use the volume of sewage discharged will be from 15 to 25 gallons per occupant per day. By making the dosing tank fairly large, the subsoil DISPOSAL AND TREATMENT OF SEWAGE 247 drains will be better charged, and in consequence the sewage will be better distributed. Both the septic and the dosing tank should be covered, as the heat in the sewage is better maintained in cold weather, and covered tanks may be placed in positions where open ones would be objectionable. Should it be found desirable to locate the septic tank, Fig. 170, in the immediate neighbourhood of a building, it may be advisable to dispense with the disconnecting trap on the drain, and to ventilate the tank by means of a soil or other ventilating pipe. A layer of earth should also be placed upon the roof of the tank, that any escaping gases may be deodorised in passing through it. In certain cases a system of subsoil irrigation may be inapplicable, and it may be necessary to adopt artificial filters for the final treatment of the tank effluent. Sewage filters are of two types, one being termed contact beds and the other percolating filters; the filtering medium, however, is the same in each type, and commonly consists of clinker, coke, or other material which provides suitable surfaces for the growth of bacteria. These filters operate by intercepting the larger particles of suspended matter in the sewage, and by oxidising the organic solids by the action of living organisms ; if the process is carried far enough a clear and non-putrefactive effluent is obtained. Contact "beds in construction differ from percolating filters ; the walls of the former require to be water-tight, and the filtering medium varies in depth from 2 ft. 6 in. to over 4 feet, according to the fall available. The walls of percolating filters may be cheaply formed, as water-tight construction is not essential, and they may also be perforated to aid the aeration of the filtering medium. Percolating filters are not usually less than 5 feet deep, and an increased depth produces an effluent of greater purity. In either form of bacterial filter the bottom requires to be arranged that it can be effectively drained, the effluent being conducted to a suitable outlet. For the construction of a filter perforated tiles or pipes may be used for the under- drains, and upon these a layer of rough material should be 248 DOMESTIC SANITARY ENGINEERING AND PLUMBING placed ; the body of the filter may be composed of material which is of a fairly uniform size, say J inch to J inch, with all dust screened out. When contact beds are used, the sewage after passing through the septic tank is distributed over the surface of a bed, and when the latter is full the sewage is allowed to remain in contact with the material for a period of about 2 hours, whilst the bacteria effect a certain amount of purification. After the contact period the efflu- ent is discharged, and the bed remains empty for any period from say 4 to about 8 hours, in order that it may be thoroughly aerated. With percolating filters the sewage is not retained as in contact beds, but after being evenly distributed over the surface of the filter the sewage trickles down and through the mass of material, and freely escapes at the outlet. The volume of sewage with which a contact bed is capable of dealing DISPOSAL AND TREATMENT OF SEWAGE 249 varies from 15 to about 20 gallons per sq. foot per day, according to the strength of the sewage. With percolating filters their rate of working depends largely upon their depth, and the character of the sewage; the volume of sewage filtered varies from 18 to 36 gallons per sq. foot per day. When either a contact bed or a percolating filter yields an effluent of insufficient purity, further purification by secondary beds may be adopted. Where adequate fall is available, a small purification works similar to Fig. 171 may be constructed, which consist of a detritus tank T, septic tank S, automatic dosing tank A, and primary and secondary filters. The house drain D is shown discharging into the detritus tank, which intercepts any mineral matter that may pass from a yard or other surface into the drain. From the detritus tank the sewage flows into the septic tank, where anaerobic organisms attack and break up the solid organic matter. The effluent from the septic tank flows into the automatic dosing tank A, which regulates the discharge and enables every part of the primary filter to be properly dosed with sewage; the interval between the periods of discharge also allows the filter to receive the necessary supply of oxygen for the development of the aerobic organisms. From a single house the discharge of sewage only takes place at irregular intervals, and unless some method is adopted for regulating the discharge the filters would not act satisfactorily, as some parts would be overdosed whilst other parts would have little or no work to do. To distribute the tank effluent over the surface of a filter various methods may be adopted, but in Fig. 171 channels are shown which are supposed to be notched at short distances along each edge. Fixed jets or revolving sprinklers may, under favourable circumstances, be used for distributing the tank effluent, but the small orifices of jets are subject to chokage, and revolving sprinklers are usually too costly for a small system of sewage purification. After the effluent has per- colated through the primary filter, it passes to the secondary filter, and thence to the outlet drain. The area of a secondary filter may only require to be half that of the primary filter, and the capacity of a septic tank for the installation shown Of THE UNIVERSITY 250 DOMESTIC SANITARY ENGINEERING AND PLUMBING should be about equal to one and a half day's discharge. The automatic dosing tank for regulating the discharge may have a capacity of twenty gallons and over, depending upon the size of the purification works. It is not essential that a system should be so compact as that in Fig. 171, and the different units may be some distance apart if the fall of the ground should favour that arrangement. CHAPTER X WATER SUPPLY IN nature there is no water which is absolutely pure, as it readily absorbs gases, and dissolves traces of many substances with which it comes in contact. The term "pure" is only used in a relative sense, and in general pure water is under- stood to be water which contains nothing which is likely to have any prejudicial effect upon those who consume it. It is only in country districts where the average student of Sanitary Engineering is directly concerned with water at its source of supply, for in urban districts, where a scheme of water-supply has been carried out, water is delivered by a system of iron pipes, and is available in the various streets. Water Pollution. A point of importance is to deliver water to a dwelling without in any way impairing its quality and rendering it injurious to health. Some of the ways in which water may be polluted are as follows : 1. At its source, by coming in contact with decaying animal or vegetable matter. 2. Through badly constructed storage tanks or reservoirs, or by polluting matter being washed into them. 3. In a main distributing system. 4. By the materials of which service pipes are made. 5. Storage cisterns in houses. 6. Defective arrangement of service pipes. Well waters may be contaminated by sewage or by polluted surface water gaining access to them, or where pumps are provided they may be the cause of the well water being impaired. Lead pumps and suction pipes are frequently used 231 252 DOMESTIC SANITARY ENGINEERING AND PLUMBING in country districts, but where a water has any appreciable effect upon lead, iron pumps and pipes should be adopted. Surface water may be excluded from wells by having them properly covered, and by building the lining of the wells above the level of the surrounding ground. The lining of a well for the greater part of its depth should be of water-tight construction, in order that water may be unable to gain access without first having percolated through a large mass of earth. In the case of deep wells the subsoil water should be excluded altogether. Where a supply of water is obtained from a spring, every precaution should be taken that the tank or reservoir in which the water is stored is suitably constructed, favourably located, and that the water channels are properly protected from possible pollution. For a country house which is chiefly dependent upon the rainfall for its supply, the chief forms of pollution are due to the collecting surfaces, by the accumulation of vegetable growths, by droppings from birds (where pigeons and fowls are kept), and by metallic impurity. Tiled roofs are very susceptible to vegetable growths, and do not make such good collecting surfaces as slated roofs. Eoofs which drain into lead gutters are not suitable for collecting areas where the water is to be used for dietetic purposes, as rain- water has a very active effect upon lead and dissolves traces of this metal. With regard to the pollution of water by means of a distributing system, this may occur under favourable con- ditions where ball hydrants are in use. If, for example, polluting matter has gained access to a ball hydrant which is located at a high level, and the water is temporarily turned off, or is drawn from the higher level by an abnormal draught at a lower point, the hydrant may open and allow the polluting matter to enter and pass into the water pipes. Coal-gas from a leaky main in the immediate neighbour- hood of a ball hydrant may also enter a water main in a similar manner to the above, for when water is withdrawn from a main so as to partially empty the latter, air rushes in to take its place. Should coal-gas or other gas escape near WATER SUPPLY 253 a hydrant, it may readily x pass into a water main under the conditions stated. Lead service pipes allow certain waters to dissolve traces of the metal, and this form of impurity when taken into the human system acts as a cumulative poison. When lead is dissolved by water, lead service pipes should not be used, that is if the water is used for human consumption in any form. The following extract l clearly indicates the danger of the prolonged use of water containing traces of lead : " Acute lead poisoning, as manifested by lead colic, anaemia, paralysis, epilepsy, etc., is rarely met with as a result of the use of leaded waters, but the insidious forms of plumbism, or lead poisoning, which are much more common than the acute cases, are constantly with us. The effects produced by the small amounts of lead taken into the system are rarely so serious as to cause death, and for this reason the injurious results of the long-continued use of waters so polluted are only gradually receiving recognition. " It is believed by those who are lucky enough to escape, that the risks of this kind of poisoning are exaggerated. The contrary is quite the case. "The symptoms of chronic lead poisoning, such, for example, as are liable to ensue after the continuous use of water containing small quantities of lead, are as follows: The symptoms are usually slow in their progress ; there is general anaemia, with a consequent anaemic pallor of the skin ; there is often constipation and indigestion ; there may be loss of appetite, an unquenchable thirst, a constant unpleasant, metallic taste in the mouth, and a foul odour of the breath." Waters which dissolve lead are usually those which contain little or no carbonate of lime. Waters which contain lime carbonates do not act upon lead, owing to a protective coating being formed on the surfaces of the metal. Soft waters as a general rule dissolve lead, but of these there are exceptions. For example, Loch Katrine water, which is used in Glasgow, is very soft, and although it has an appreciable effect upon new lead pipes, after a short time their inner surfaces become coated with a film of vegetable matter, which combines with 1 New Hampshire Sanitary Bulletin. 254 DOMESTIC SANITARY ENGINEERING AND PLUMBING the oxide of lead that is first formed and prevents further action taking place. The soft water from Thirlmere, which supplies Manchester, appears to have a similar effect on lead pipes as that from Loch Katrine. In many districts storage cisterns require to be fixed in buildings where the pressure in the water mains falls too low to give a constant supply throughout the whole of a day, and unless such cisterns are formed of suitable material, are properly protected and suitably placed, contamination of the water will be the result. All storage cisterns should be provided with covers, to exclude foreign matter from gaining access to them. Pollution of a water supply by defective arrangement of service pipes is not very common at the present time, owing largely to the regulations imposed by water companies. In country districts where water is obtained from a private source irregular connections with service pipes may still be found ; such, for example, as the direct connection of a service pipe with a urinal or a w.c., instead of the direct connection being broken by means of a flushing cistern. Sources of Water Supply. The principal sources of water supply for rural districts are Rain-water, Springs, and Wells; whilst Upland surface water and Eiver water are frequently resorted to for supplying large communities. These, of course, all depend upon the rainfall for their replenishment. Rain-water is the purest form of natural water when caught in country districts which are well removed from towns and industrial centres. Rain is naturally distilled water, being slowly evaporated from the sea and from water on the earth's surface; the aqueous vapour rises into higher regions to form clouds, and upon condensation again falls in the form of rain or snow. In falling through the atmosphere rain-water absorbs any gases which may be present, and this accounts for the pollution of rain-water when caught in the neighbourhood of industrial centres, where the atmosphere is laden with impurities, such as particles of soot, sulphurous compounds, etc., and where it falls on surfaces which also intercept polluting matter. WATER SUPPLY 255 Rain is rich in oxygen, and this has the effect of increasing its solvent powers. Fresh rain-water has a flat, insipid taste, but this may be improved by filtration. For a house in a country district stored rain-water may occasionally be the principal or only convenient source of supply. When this is the case, and it is also used for drinking purposes, the following require special consideration : (a) Suitability and sufficiency of collecting area. (b) Sufficient storage. (c) Suitable filtering arrangement. Collecting Area. The surface used for collecting rain- water requires to be kept as free as practicable from all polluting influences, and no rain-water should be used for dietetic purposes which flows through lead gutters. Slated roofs make good collecting surfaces, but where a suitable roof area is inadequate to yield the necessary volume of water, specially prepared surfaces are essential. When roofs are the collecting surfaces, the most suitable channels or receivers are cast-iron gutters which have their inner surfaces well coated with a bituminous paint such as Dr. Angus Smith's solution, or other leadless coating. The interior surfaces of cast-iron rain-water pipes should be protected in a similar manner, whilst the external surfaces of either pipes or gutters may be protected with lead paint of any particular colour. Earthenware drains are .the most suitable channels for conducting rain-water to a storage tank from the rain-water pipes. Special Collecting Areas take different forms. They may be raised above the contiguous ground and arranged to fall to one end, being rendered practically impervious with a covering of either cement or asphalt. At the lower end of a prepared surface a collecting channel may be arranged which discharges into a suitable sump to intercept leaves and similar matter. Collecting areas may also be arranged as in Fig. 172. In this case the rain falls upon a grass surface, percolates through say a foot of soil, and after- wards through perforated tiles which are placed upon an impervious floor. Special tiles may readily be obtained which will support the soil, and at the same time permit of 256 DOMESTIC SANITARY ENGINEERING AND PLUMBING : ' v ?''t?'^ ' ^::-:\ ._ , : : &tf**v fte&&&AA S adequate under drainage between the tiles and water-tight floor, which should be made to fall towards a sump at any suitable point. The collecting area in Fig. 172 takes advantage of the purifying effect of the organisms in the soil, as explained in the chapter on sewage treat- ment, and also of the purifying power of grass. Where a collecting area requires to be excavated and prepared as in Fig. 172, it should be channelled all round, the bottom of the channels being lower than the 53 floor of the prepared area ; this provision | reduces materially the chances of pollu- 3 tion. So All special collecting surfaces should g be properly protected from surface pollu- ' tion by fowls, cattle, etc., by having them % surrounded with a suitable fence. Stored g rain-water is only suitable, of course, 3 where the atmosphere is comparatively ^ pure, as in most country districts. ^ Conditions affecting Yield by a Surface. For conditions represented by (c) G = '47xAx/ . . . (13) Example 4. If the rainfall of a certain district is 25 inches per annum, find the volume of water which is available for storage when the total collecting surface is 2400 sq. feet. For condition (a) G = -37 x A x/, G = -37x2400x25; /. G = 22,200 gallons. For condition (b) G = '34xAx/, G = -34x2400x25; .-. G = 20,400 gallons. For condition (c) G = 47 x A x/, G = -47x 2400x25; /. G = 28,200 gallons. Example 5. What collecting area will be necessary where the rainfall is 28 inches in order to yield 54,750 gallons -per year? O For condition (a) A = -~= ^ , '61 X/ A= 5475 37 x 28 ' /. A = 5285 sq. feet of surface. WATER SUPPLY 259 For condition (b) A = - For condition (c) A = ; A 34 x/' . 54750 34x28' /. A = 5751 sq. feet of surface. G 47 x/' 54750 47 x 28 ' /. A = 4160 sq. feet of surface. Capacity of Storage Tanks. It is obvious that the storage capacity of a rain-water tank need not be capable of accom- modating the total rainfall, as the latter is distributed over the whole of a year. Under ordinary conditions, when entirely dependent upon the rainfall for a supply, the storage capacity in this country should be equal to about 80 to 120 days' supply, according to whether a district is a wet or a dry one. A less capacity will, of course, suffice where rain-water can be supplemented by water from another source. Water Consumption. The consumption of water is usually stated in gallons per head of a population, and this varies con- siderably in different localities. In towns the consumption per head varies from 20 to about 60 gallons per day, smaller towns, as a rule, consuming less per head than the larger towns. In rural districts the consumption per head varies from less than 9 to about 20 gallons per day, the smaller value apply- ing when w.c.'s are not in use. Size of Tanks. For obtaining the size of a rectangular tank the following formulae may be used : Let P = number of persons for which storage is provided. C = gallons allowed per head per day. S = number of days' storage. I = length of tank in feet. & = breadth of tank in feet. h = depth of tank in feet. P X C X b 260 DOMESTIC SANITARY ENGINEERING AND PLUMBING .PxCxS -no*6j ' (16) x.PxCxS -^&3T6i Example 6. Determine the width of a storage tank which is 21 ft. 6 in. long, and 8 feet deep below the overflow, to hold 80 days' supply for 12 persons, where the rate of consumption is 15 gallons per head per day. _. - , i n -L By Formula 16, ^ ' , 12x15x80 Substituting values, b = , ~ , 2x12x15x80x4 43x8x25 /. 6 = 13H ft-, say 18 ft. 5 in. broad. Should a circular tank be constructed, the formula for calculating its diameter or depth may be expressed as shown below. , PxCxS / V PxCxS Where D = diameter in feet, and h, P, C and S as before. Example 7. If storage is provided for 20 persons for 60 days, what depth of tank will be essential, if its diameter is fixed at 20 feet, and the rate of consumption at 12 gallons per head per day ? Formula 18 gives h Substituting values, A^ .-. h = 7U ft., say 7 ft. 4 in. deep. Example 8. Assuming the depth of the tank had been fixed at 9 ft. 6 in. below the overflow, and the diameter of tank was required. WATER SUPPLY 261 Then by Formula 19, D = . / PxCxS V Ax6j ' and substituting values, D = / v 20 ' * 12 /. D = 17-58 ft., say 17 ft. 7 in. diameter. Purification of Rain - Water. To render rain - water sufficiently pure for drinking purposes, and also to improve its taste, it requires to be filtered. On the other hand, if rain- water is only stored for general household use, and also to PURE FOUL FIG. 173. Roberts' rain-water separator. supply sanitary fittings, a high degree of purity is not essential, and simple straining may be all that is required. Rain-Water Separators. A useful appliance for preventing the first portion of a rainfall from entering a storage tank is a rain-water separator Fig. 173 (Roberts'), which can be fixed at any suitable point in a drain leading to the storage tank. The rain-water separator shown is self-acting, and simply diverts to waste the first portion of the rain which contains the greater portion of the impurities which have accumulated on the collecting area. Its action is dependent upon a canting or tipping compartment, which is regulated to fill at a given rate. When the canter is up or in its normal position, the rain-water flows through the lower or 262 DOMESTIC SANITARY ENGINEERING AND PLUMBING foul-water outlet ; but if the canter is full of water it tilts downwards and diverts the entering water through the upper outlet; the interval required for filling the canter is the time allowed for washing the collecting surface. The canter is emptied by means of a siphon, when it tilts back to its former position some time after the rain has continued to fall. If a shower of rain is only of short duration and light, the canter will not come into action, but the whole of this water will flow to waste. Separators are made in a number of sizes, to suit either small or large collecting areas, and for either town or country use. Another form of Eoberts' separator is shown by Fig. 174, for fixing in conjunction with a stack of pipes so as to deliver the rain-water either above or below ground level. The vertical type Fig. 174, which is shown in section, contains the following principal parts. Beginning at the top, A denotes movable strainers, and B a perforated slide which regulates the How of water to the canter. The small chamber at C contains a sluice which can be adjusted to suit the area of collecting surface. E shows the course taken by the water when flowing through the separator, and J represents the canting chamber which revolves for a limited distance on a pivot ra. The small compartment F in the canter is provided with a small regulated outlet G, which discharges into the lower part of the separator. A siphon L has its PURE FIG. 174. Roberts' rain-water separator. WATER SUPPLY 263 outlet leg in chamber F, whilst its inlet leg is turned into the large chamber J. The lower portion of the separator is divided into two parts, N and 0, by a thin plate, in order that the canting chamber may divert the water on either side of the plate according to the relative position of the canter. Its action is as follows : When water enters the separator, a portion of it passes through the strainers A, and into small upper chamber C, from which it flows through the perforated slide B to the small compartment F ; when the latter is full, water overflows into the main body of the canting chamber J, and after a time the canter turns and diverts the water into chamber 0, from which it flows to storage. Prior to canting the water is delivered into chamber N, from which it is passed to waste as in Fig. 174. The small chamber F is slowly emptied by means of the small aperture G, and as the level of the water falls the siphon is brought into action, when the water from the canting chamber is also discharged. To ensure these separators working properly they require periodical attention to keep the small apertures and strainers clear, and also to lubricate the pivot on which the canter turns. When a more simple and cheaper appliance is required, a spout or channel might suffice, which can be tilted one way or the other, and so divert the water either to storage or to waste. Sand Filters. A well constructed sand filter is a very effective type for dealing with large volumes of rain-water. The arrangement of storage tank and filter will depend to a great extent upon the volume to be filtered, whether they require to be constructed upon sloping or upon level ground, and whether one or more services be required. If it is assumed that a large building is wholly supplied with rain-water, it would not be necessary to filter the whole of the water if two separate services were adopted ; viz., one to supply filtered water for dietetic* use, and the other unfiltered water for sanitary fittings, where the water is not likely to be used for human consumption, and for laundry purposes, washing vehicles, and similar uses. Bacteria are responsible for the purification effected by 264 DOMESTIC SANITARY ENGINEERING AND PLUMBING sand filters, organic matter being oxidised by the action of nitrifying organisms. Efficient nitration largely depends upon the rate of flow through the filtering medium, and upon the condition of the film of organic matter which forms on the surface of the sand. After a time the slimy matter clogs up the surface of a filter, and its filtering capacity requires to be increased ; this is done by removing a thin layer of sand, which should be replaced after washing unless new sand is substituted. When a sand filter has been cleansed, or when it is new, water should be allowed to stand upon it for about 30 hours, in order that a film of matter may be deposited upon the sand before filtration begins. Where very pure water is required the first flow through a sand filter should be rejected, or utilised for some other purpose. With regard to the depth of sand for a small rain-water filter, this should be about 1 foot, and deeper where practic- able. Each square foot of filter surface will effectively deal with 30 to 40 gallons of rain-water per day, and with these low rates the filter afea for a large house would be com- paratively small. Fig. 175 gives a plan of storage tanks and filters, together with their connections, where rain-water is supposed to be chiefly utilised as the supply for a large building. Two hundred gallons of filtered water are required per day, and the storage tank T has a water capacity of 16,000 gallons, which represents 80 days' supply. A rain-water separator E, is provided to exclude the first portion of the rainfall from the storage tank, and two filters are arranged to work independently of each other, in order that one may be in use whilst the other is cleansed. The filters may be located at a lower level than the storage tank, and the supply to each filter may be con- trolled by a ball-cock as shown. On the supply pipes to the filters three stop-cocks are indicated, which are numbered 1,2, and 3, whilst two others are provided on the outlets of the filters and numbered 4 and 5. Stop-cocks 2 and 4, 3 and 5, serve to throw the filters out of action, whilst stop-cock No. 1 is intended to be set so as to regulate the rate of flow to the filter. Adjoining the filters a small storage tank is provided 266 DOMESTIC SANITARY ENGINEERING AND PLUMBING for filtered water, which may hold rather more than one day's supply. The construction of the filters is as follows : On the floors perforated tiles are arranged, which allow the filtered water to flow into a channel, and thence to the small storage tank. About six inches of gravel are laid upon the tiles, and upon the gravel 1 foot of suitable sand is placed. To prevent the surface of the sand being disturbed by inflowing water, the ball-cocks may deliver into small earthenware channels which rest upon the sand ; in this way the water would be better distributed over the filter. A sump is provided at the outlet end of the large storage tank T, and here a sluice may be arranged in order that the tank may be emptied when desired. An overflow to the tank is provided, and is shown to discharge into the waste-water drain from the separator. The tanks and filters should be properly roofed over and ventilated, the walls and floor rendered water-tight, and ample means of access should be provided. Where concrete construction is adopted it can more easily be rendered water-tight by properly grading the aggregate. A good concrete mixture for water-tight work is 1 part cement, 2 parts sand, and 4 of broken stone, the latter being broken to various sizes. Where only a comparatively small volume of rain requires to be filtered, household filters would be the most satisfactory to use. Springs as a Source of Water Supply. As a rule springs yield a pure and wholesome water, on account of the latter having percolated long distances and filtered through con- siderable mass of earth. For houses in rural districts springs form good sources of water supply, as the water yielded by them is usually free from organic impurity, although it may contain a large amount of carbonic acid gas and dissolved mineral matter. There are two kinds of springs : (a) surface or intermittent springs ; (b) permanent or deep-seated springs. Surface Springs. A surface spring may occur either at a low point on the side of a hill, or at the side of a valley, where an impervious stratum which prevents farther downward WATER SUPPLY 267 progress of the water suddenly appears at the surface. When rain falls upon pervious strata at a high level it gradually flows downward and forward in its underground course, until it reappears at the outcrop. The volume of water yielded by a surface spring depends upon the area drained, and upon the percentage of the rainfall which percolates into the earth. As their name implies, the volume of water yielded by sur- face springs is readily influenced by wet and dry weather. Deep- Seated Springs differ from surface springs in that the water is forced to their outlets by more or less hydrostatic pressure, whilst the water of surface springs simply gravitates from a higher to a lower level, and is under no hydrostatic pressure. Deep-seated springs have their origin where the rain, after percolating through porous strata at a high elevation, eventually becomes confined between impervious strata and sinks to a lower level than its point of escape. The water of these springs usually travels long distances through porous rocks, and when it has reached its lowest point the water there accumulates until sufficient pressure is produced to force it through some fissure in the strata. Owing to these large subterranean accumulations of water, deep-seated springs yield a more nearly permanent rate of flow. Spring water is clear and sparkling, owing to its having filtered long distances through porous strata, and with having absorbed large volumes of carbonic acid gas in its passage. Where spring water is used for supplying one or more buildings in a rural district, a spring may frequently be chosen which is sufficiently high to give a gravitation supply by providing a suitable storage tank which either adjoins or is located some distance from it. With regard to the size of storage tanks, in this case it will depend upon the volume of water yielded by the spring, whether the yield is fairly uniform or not, upon the volume of water required, and whether water is stored for extinguish- ing fires. If the yield of a spring is greater than the demand, only a small storage tank may be essential ; but if the demand for water at certain times greatly exceeds the rate of supply, 268 DOMESTIC SANITARY ENGINEERING AND PLUMBING then a larger reserve will be required. Under the latter circumstances the capacity of a tank may be equal to anything from about 4 days' to 4 weeks' supply, according to the require- ments to be satisfied. The storage capacity for fire extinction cannot be definitely fixed for small supplies, as so much depends upon the special circumstances of each particular case. Owing to the freedom of spring water from organic pollution, no further filtering of this water is usually required. The outlet pipe from a storage tank should be protected by either a copper rose or a wire screen, in order to prevent any foreign matter which may enter a tank from being drawn into the pipe. A suitable overflow should be provided, and where a storage tank is large provision should be made for emptying it. If a spring be large, the supply to a storage tank may be regulated by a ball-cock, whilst in the case of a more limited supply the whole of the water may flow to the tank. When springs occur at lower levels than the buildings to be supplied, great care is necessary to guard against possible pollution of the water. The usual appliances for raising water to higher levels are pumps and hydraulic rams, the latter being specially suitable where there is sufficient water to work them. A simple method of finding the volume of water a spring will yield is that of first ascertaining by means of a stop- watch how long it takes to fill one or more buckets of known capacity. A convenient place can usually be found where a stream from a spring may be impounded, and by means of a spout or channel after the water-level has adjusted itself, the yield may be readily gauged, and expressed in gallons per minute or in any other units desired. Thus, if a bucket has a capacity of 2 J gallons and is filled in 25 seconds, the yield = 2 i x6 = 5 x ~ ZiD 2i ^!iD 6 gallons per minute, or 6 x 60 = 360 gallons per hour. Wells as a Source of Supply. In country districts under- ground water often forms the principal source of supply, and where it does not issue in the form of a spring it is often WATER SUPPLY 269 necessary to tap it by means of a well. The quality of well waters is very similar to those from springs, and depends upon the nature of the strata through which they have passed. Well water, however, is very liable to pollution unless the wells are suitably located, properly constructed, and protected by covering them. There are three classes of wells (a) Surface wells. (b) Deep wells. (c) Artesian wells. Driven tube wells may be classified as either surface or deep wells, according to the geological formation through which they are driven. Surface Wells are those which are sunk in the subsoil, A FIG. 176. Subsoil well. the water being held up by an impervious stratum as in Fig. 176. Kain after falling upon the permeable formation percolates downward until its progress is arrested, and the whole of the porous strata below the line AB is saturated unless water is withdrawn by pumping; any further water which percolates from the upper surface escapes at B in the form of a spring. The chief objection to surface wells are, their waters are often exposed to sewage pollution, owing to leaky drains and cesspools, and by surface impurities being washed by rain into the subsoil; the waters yielded by surface wells are usually very hard, and not very suitable for general household use. So far, however, as the danger to pollution is concerned, this may be reduced to a minimum by making these wells in non-polluted areas, and by constructing the upper 12 feet of 270 DOMESTIC SANITARY ENGINEERING AND PLUMBING MONKEY their depth with water-tight linings, which are continued 9 inches or so above the level of the ground. Driven Tube Wells. In soft soils a cheap supply of water may frequently be obtained by an Abyssinian tube well. These wells are formed by driving strong iron tubes to near the water-bearing stratum. The usual sizes of the tubes are from 1J to 3 inches diameter, the difficulty of driving increas- ing as the sizes of the tubes are increased. To facilitate driving the tubes should not exceed 6 feet in length, and shorter ones in some cases may be desirable. The method of driving the tubes is chiefly governed by their size and the nature of the earth to be pierced. For driving the smallest sizes a sledge hammer may suffice, or the method shown in Fig. 177 may be adopted, where a weight or monkey falls on to a driving cap which is attached to the top of the tube. Where a heavy monkey is used it is raised by means of ropes and pulleys. The first and perforated tube is 3 feet long, and is provided with a driving point, which is a little larger at A than the tube itself, in order that a hole may be made sufficiently large to just clear the sockets. In size the perforations are about J inch diameter, but when the tubes are driven into fine sand, the holes may be reduced in size by covering them with brass strainers. It is necessary when first starting to drive the tubes to see that they are exactly vertical, otherwise the well may result in failure. As each length of tube is driven in the ground another is added, and this process continues until the desired depth has been reached, when a pump is attached to the top of the FIG. 177. Method of driv- ing tubes for an Abyssin- ian tube well. WATER SUPPLY 271 tubes. When, however, a well of the driven type is used its depth will be limited to about 30 feet- Tubes can only be driven, of course, through soft loose strata, such as sand, fine gravel, etc., as in Fig. 178. Where firm strata is to be penetrated the hole for the tubes requires to be bored. After the tubes have been driven to the required depth, Fig. 178, a cavity requires to be made in which water may accumulate, or pumping would be difficult if water were drawn directly from the surrounding earth. The cavity may be formed by allowing the column of water in the tubes which form the suction pipe to fall back a number of times by destroying the vacuum, and so loosen the sand which surrounds the perforated tube. To aid in removing sandy matter which enters the tube it is better to use a common galvanised iron pump, as grit has a destruc- tive effect upon the working parts of a pump. The vacuum in the suction pipe may be broken during the formation of a cavity by providing a short length of tube with a tee,and by temporarily joining it at P, Fig. 178. A pump may then be screwed into the upper opening of the tee, whilst a stop-cock may be fixed to the remaining connection. After the pump has been primed and a few strokes made, the water rises to the outlet, when, upon quickly opening the cock, air is admitted and the water falls to loosen the earth surrounding the perforated tube. This operation can be repeated until the cavity is sufficiently large, and the temporary pump may be removed and the permanent one installed. FIG. 178. Abyssinian tube well. 272 DOMESTIC SANITARY ENGINEERING AND PLUMBING Deep Wells differ from surface wells inasmuch as the latter only tap the subsoil water, whilst deep wells tap water which is located beneath the impervious stratum which sup- 5UBSOIL WELL. \ % ^ IVH' 'i -' 'V 1 ' <<'> .,:' vVi ! i FIG. 179. Formation showing subsoil and deep wells. ports the subsoil water. The actual depth of a well does not decide the class to which it belongs, and a so-called surface or subsoil well in one locality might be much deeper than a so-called deep well in another district. Fig. 179 clearly shows the difference between surface or subsoil and deep wells. Should a storage tank take an irregular shape, as in Fig. 191, to fit some required position, the following formulae may be used : bxhx3% . . (24) 6 = x (25) (26) Where n and o = the length of the short and long sides respectively. Example 9. If Fig. 191 represents the plan of a cistern where n = 3 ft. 8 in., = 4 ft. 2 in., width 2 ft. 4 in.; find its contents in gallons if its depth is 3 ft. 6 in. By rule 24, G = (n + o) x lx hx 3, G=7|x2Jx3Jx3i 57575 : 288 ' .-. G = 199fff, or say 200 gallons. Example 10. Determine the depth of a cistern which is similar on plan to Fig. 191, to hold 250 gallons, where n = 4 ft. 6 in., o = 5 ft, and the width is 2 ft. 7| in. /~i By rule 26, h = 250 h = 9Jx2fx3i' 1280 ,-, h = 3^/9- ft-> or 3 ft., 2| in. deep. 292 DOMESTIC SANITARY ENGINEERING AND PLUMBING The actual capacity of storage tanks with an intermittent supply should not be less than twenty-four hours' requirements, and a greater reserve may be necessary in certain cases, but each case requires to be dealt with on its own merits. Domestic Filters. If drinking water is of doubtful purity the best plan is to boil it. This method, however, is not always convenient or practicable, and, moreover, water which has been boiled has a flat, insipid taste on account of its lack of aeration. Various forms of household filters are largely used for filtering water, but very few are satisfactory, and the best III PLAN FIG. 191. Plan of a cistern. types only resist the passage of micro-organisms for a limited period. Many domestic filters simply remove suspended and dissolved organic matter, but allow the free passage of germ life. As a rule where a good class of water is supplied on the constant system no filter should be used, as there is the possibility of the filtered water from a bacterial point of view being inferior in quality to the same water when unfiltered. If, on the other hand, a water supply is not beyond suspicion, the filtration of water for dietetic uses may be found desirable. In the latter case it is important that a suitable filter be selected, and that it is regularly cleansed and sterilised to keep it in a satisfactory state. WATER SUPPLY 293 Domestic filters may be divided into two classes. First, those which work under high pressure ; and second, those which operate with a low pressure. The filtering medium may be the same in each type, the latter requiring a much larger surface of filtering medium. Amongst the most reliable domestic filters are those of the VWj .METAL CASE Fio. 192. Pastenr filter. FIG. 193. Berkefeld filter. Chamberland Pasteur, Berkefeld, and similar types. Figs. 192 and 193 show the filters mentioned, and in both cases the filtering medium takes the form of a cylinder or hollow candle, fine unglazed porcelain being used for the Pasteur filter, whilst compressed silicious earth is used for the filtering medium of the Berkefeld. The construction of the filters is clearly shown, and in 294 DOMESTIC SANITARY ENGINEERING AND PLUMBING each case the water enters the metal cylinder and is forced under pressure through the porous medium, when it afterwards escapes through the glazed nozzle outlet in the case of the Pasteur, and through the bent tube at the top of the filter in the Berkefeld. The filtering medium in the Berkefeld is both thicker and more porous than that in the Pasteur, and the former consequently filters water at a quicker rate. To cleanse these filters the candles are removed, washed, sterilised by boiling them, and afterwards replaced. Owing to the slow rate of filtration by a single Pasteur filter a small reserve of filtered water is desirable. For this purpose a glass or stoneware jar, which is fixed immmediately beneath and connected with the filter, is generally adopted. Water may then be withdrawn directly from the receiver, but precautions are necessary to guard against contamination, as filtered water readily absorbs any gases to which it may be exposed. For low pressure filters of the Pasteur and Berkefeld types, several candles are arranged inside a large metal casing or cistern, the candles being joined to one common channel into which the filtered water escapes. As a number of joints are necessary with this arrangement, there is a possibility of some defect occurring through which unfiltered water may gain access to that which has been filtered. Domestic filters to be satisfactory must be simply con- structed, and admit of being readily taken to pieces for cleansing purposes. After a filter has been in use for some time, its filtering capacity is greatly reduced owing to the choking of the pores, but its normal rate may be again restored by thoroughly cleansing it. Water Fittings. Taps, cocks, cranes, or valves may be roughly divided into three classes : 1. Those which automatically regulate the outflow of water. 2. Those which are operated by hand. 3. Those which are semi-automatic in action. To the first class belong ball-cocks which are opened and closed by the falling and rising water in a cistern. To the second class the various forms of screw-down and plug taps belong. To the third class belong those which are opened by hand, but are automatically closed either by the aid of springs WATER SUPPLY 295 or by the water pressure, or by means of a weighted lever. As a large variety of fittings belong to each class, further sub- division becomes necessary in order to compare their relative merits and defects. Bail-Cocks. These taps, which belong to the first class , may be subdivided into high and low pressure forms. High pressure ball - cocks have often restricted water-ways, and where the latter are fairly large the levers are frequently compounded. Low pressure ball-cocks have large water-ways, and may either be constructed upon the equilibrium principle or with simple direct acting levers. Much annoyance and inconvenience are caused from time to time by fixing unsuitable ball-cocks in cisterns. Very frequently a form of high-pressure tap is connected with a low- pressure service, with the result that the cistern takes too long to fill. For flushing cisterns in connection with w.c.'s it is specially necessary that they fill rapidly, but this can only be accomplished by the selection of suitable taps for the pressure at disposal. Should a low-pressure ball-cock be fixed on a high-pressure service, the former will either be often out of order, or rattling sounds will be caused, due to the oscillation of the lever when the tap is nearly closed. The latter action under certain conditions is readily brought about, owing to the force which operates to close the tap not sufficiently exceeding that which tends to open it. When oscillation of a lever has commenced the water pressure is increased to a more or less extent, according to the amount of concussion produced by the quick successive opening and closing of the tap. Two common and good types of ball-cocks are given in Figs. 194 and 195, but they are only suitable on account of their restricted outlets for high-pressure services. In con- struction the principle is the same in each case, their difference being in the arrangement of the plug or piston C, which contains the washer. In Fig. 194 the piston moves horizon- tally, whilst that of Fig. 195 has a vertical motion, the water escaping from the latter by side passages which are not shown in the figure. When water is under considerable pressure the ball-cocks 296 DOMESTIC SANITARY ENGINEERING AND PLUMBING shown are somewhat noisy in action, especially when they are nearly closed owing to the water issuing with considerable velocity through the contracted orifice. This objection, FIG. 194. High-pressure ball-cock. however, can be overcome to a great extent by attaching to the outlet P of Fig. 194 a short length of tube in order that the point of escape may be submerged. Upon reference to Fig. 195 it will be observed that it does not admit of a tube being readily attached, and this is the chief drawback of this pattern of ball-cock. FIG. 195. High-pressure ball-cock. A patented ball-cock is given in Fig. 196, and in con- struction the part containing the valve is similar to that in Fig. 195, excepting that the orifice is larger. In Fig. 196 the additional power which is necessary to compensate for the larger valve orifice is obtained by compounding the levers as WATER SUPPLY 297 shown. The regulating screw S is an advantage, for by its means the ball-cock can be adjusted with precision without resorting to the practice of bending the levers. It will be observed that the mechanical advantage of lever A is con- centrated, and operates at the adjusting screw S, which forms the end of the long side of lever B. From a practical point of view the advantage derived by compounding the levers of ball-cocks is very limited, and therefore the size of the valve orifice for high pressures is also limited. The full -way equilibrium valve Fig. 197 is well adapted for low pressures. In construction it differs considerably from the foregoing, inasmuch as the water pressure is utilised in addition FIG. 196. The "Hiorlo" ball-cock with compounded lever. to the float and lever for closing the valve. At the upper pan of Fig. 197 a cup-leather C is provided to make the valve water-tight at that point. The pressure of the water is exerted both on the cup-leather C and the valve V, and as their surfaces in the case shown are equal, the total pressure acting upon each surface when the valve is closed is also equal ; under these conditions equilibrium is only destroyed by the weight of the ball and lever when the water level begins to sink. For high-pressure services, cocks like Fig. 197 are not suitable, as they are liable to produce more or less concussion when nearly closed. A drawback associated with ball-cocks which have cup-leathers is their liability to leakage should the cup-leathers get hard and dry through the turning off of water for prolonged periods. 298 DOMESTIC SANITARY ENGINEERING AND PLUMBING Of the ball-cocks shown, it will be noticed that only in Fig. 197 does the water pressure play any part in closing them, but on the contrary the pressure of the water in Figs. 194 to 196 is always acting to open them. Ball-cocks are, however, made which close with the water pressure, but these are often troublesome for high-pressure services unless a special form of construction is introduced to prevent concussion. Taps of the second class may be subdivided into screw-down and plug forms. Of the screw-down class there are many kinds, both of bib and stop-cocks, but the main principle is common to all. FIG. 197. Equilibrium ball-cock. A section through a bib tap is shown in Fig. 198. In a good class of tap the seating should be wide and a little raised in order to form a good bearing for the valve; the screw part of the tap at its lower point should be enlarged as at E, so as to distribute pressure over the greater portion of the loose valve when closing the tap. For a tap to be durable it requires to be strongly made, and especially where high pressures are concerned. A better method of testing water fittings to that frequently adopted by Water Companies is desirable, as many of the jerry made fittings which are put upon the market, aithough capable of withstanding a pressure WATER SUPPLY 299 test when they are new, fail after being in use for a com- paratively short time. For water fittings hard brass is a suitable alloy where water contains temporary hardness, but for soft acid waters gun-metal fittings should be used. A clear-way screw-down cock, Fig. 199, is very suitable for low pressures, as it possesses no loose valve which is liable to stick. It is also suitable for high pressure services. Plug taps, Fig. 200, are often used for both bib and stop- FIG. 198. High-pressure screw-down cock. taps where the water pressure is low, as well as for under- ground stop-cocks on high-pressure services. Plug taps are very suitable for joining directly with cisterns to control the smaller draw-off pipes. These taps are liable to stick when not in regular use, although this can be avoided to a great extent by properly greasing them when they are fixed ; more- over, when plug taps are in accessible positions there is very little difficulty in loosening the plugs which may have tem- porarily become fast. Bib taps of the plug type, when under a more or less considerable head of water, subject pipes and fittings to 300 DOMESTIC SANITARY ENGINEERING AND PLUMBING unnecessary strain by being too quickly closed. Bib taps when in constant use work loose, and by suddenly arresting FIG. 199. Fullway high-pressure screw-down valve. the How of water the pressure due to shock is often consider- able, and may rise to many times that which is due to the statical head of water. The quick clos- ing of plug bib taps has often been responsible for damaged pipes ; sudden shocks are far more detrimen- tal than a constant strain due to high pressure. When a quick-closing tap is fixed t0 a Fir, 200. -Plug cock. the part where the greatest pressure due to concussion occurs is near the end of the pipe. Bends in pipes influence the result, but as a rule the pressure due to any specific shock at different parts of a WATER SUPPLY 301 pipe diminishes rapidly from the end. With regard to the inte'nsity of pressure due to shock, that will depend upon the normal pressure of water at the point under consideration, the pressure during the period when the tap is opened, and the rate at which the tap is-closed. Screw-down taps are largely adopted in buildings, as these can be easily repaired, and as they are slowly closed concus- sion is reduced to a minimum. A certain amount of care is essential when fixing plug taps. Many plumbers when soldering a plug tap to a lead pipe first remove the plug from the body of the tap, in order to more quickly get up the " heat " for wiping the joint ; when the work is complete, and the water is turned on, it is often found that the tap leaks, and to remedy the defect " grinding in" is necessary. Had the plug been left in the tap during the wiping process it is very probable that the leakage referred to would not have occurred. The reason for this assumption is, that as the temperature of the whole mass of metal has been raised, the rate of expansion would be practically uniform, and upon cooling the rate of contraction would also be uniform for the whole mass. On the other hand, where the body of a tap has been raised and cooled through a big range of temperature, and the plug has not been subjected to a like action, it is quite feasible for the ground surfaces to be affected owing to the rate of contraction not being quite equal to the rate of expansion. The above description, however, more particularly applies to the cheaper class of taps, many of which are deficient in substance, and where the quality of the alloy is not all that can be desired. Spring taps, which come within the range of the third order, may be subdivided into quick and slow-closing types. Quick- action spring taps cause a great amount of concussion, and, like plug taps, are not suitable for draw-off taps on high- pressure services. The primary object of self-closing taps is to reduce waste of water, but in practice more water is frequently wasted by their use than with any other form of tap, owing to their being often out of repair. A quick-closing spring tap is given in Fig. 201. It will be noticed that the valve V opens against the water pressure, 302 DOMESTIC SANITARY ENGINEERING AND PLUMBING and therefore the latter is available, in addition to the spring S, for closing the tap. The combined forces of course require to be overcome to open the tap. This form of cock, however, will produce water-hammer in pipes when the press-knob is released, owing to the sudden closing of the tap ; where the water pressure is moderately high the valve will rebound from its seating, and a number of successive shocks may be produced before the tap is properly closed. To render a self-closing tap non-concussive its rate of FIG. 201. Defective form of self-closing tap. closing requires to be regulated. In Fig. 202 a slow-closing and non-concussive tap is shown by Glenfield and Kennedy Ltd., which operates in the following manner: Under normal conditions, when the water is on under pressure, the forces exerted by the water on both the under and upper surfaces of the piston P are equal, and the valve x is pressed upwards by the water pressure beneath it as well as by the force of the spring. At the under side of piston P a small valve y is provided, which is opened by means of a thin spindle which passes through the tubular rod R to the press-knob at the top of the tap. The use of the small valve enables the tap to be WATER SUPPLY 303 the more easily opened, as the resistance offered by the water to the opening of the valve is directly proportional to its sectional area. For example, if the diameter of a valve is f of an inch, and that of another f of an inch, the force to open the former would be four times greater than that required to open the smaller valve. When force is applied on the press-knob the small valve y is first opened; this releases the internal pressure in cylinder C, by allowing the water to escape through the FIG. 202. Glenfield and Kennedy's non-concussive self-closing tap. hollow rod K to the outlet of the tap. The differential pressure thus produced on the upper and under surfaces of piston P allows the larger valve x to be readily opened, when the water freely escapes. To enable valve x to be slowly closed water is slowly admitted into C, either through a small orifice at the top of the piston or by leakage at its sides ; upon the press-knob being released the small valve y is immediately closed, and as water gradually exerts pressure on the under side of piston P the valve begins to close. The sudden closing of x is prevented by the downward force 304 DOMESTIC SANITARY ENGINEERING AND PLUMBING of water on the top of the piston, and by equal pressures on both the under and upper surfaces being delayed through a given interval of time. Slow-action, self-closing taps which are similar in prin- ciple to Fig. 202 may be used for either low or high-pressure services, but the strength of the springs should be adjusted according to the intensity of the pressure. For low pressures a moderately strong spring is essential to overcome resistance offered by the piston P, whilst for high pressures a much weaker spring is desirable, as frictional resistance is of less importance. A strong spring has the effect of making the tap more difficult to open. Water-Hammer and Other Noises in Pipes. In describing the different forms of taps, reference has been made to those which are liable to produce shock, and by water-hammer is understood the sharp rapping sounds which are due to shock. Other noises occur in water pipes, such as buzzing sounds, but these differ from water-hammer, as the latter is accompanied by increased pressure, whilst the former are simply due to water flowing with a high velocity through an irregular or contracted orifice, and are not accompanied with any excess of the normal pressure. A buzzing sound may be produced when a screw-down tap is nearly closed, and where the valve is rather loose or does not close evenly on its seat ; if water has greater freedom to flow under one side more so than another, a rotary action is imparted to the loose valve, and thus the buzzing begins. Whistling sounds are occasionally produced by ball-cocks when nearly closed, but where these sounds are objectionable ball-cocks should be used which have submerged outlets. The rattling or clicking sounds which are produced by automatic and semi-automatic taps are a form of water- hammer, but shock of much less intensity accompanies these when compared with that of a pronounced water-hammer. To remedy a case of water-hammer its cause should first be ascertained. If a certain type of ball-cock is responsible for it, it may be necessary to change the cock for another type. Should quick-closing bib taps be the cause of water- hammer, then, if practicable, they should be replaced with the WATER SUPPLY 305 screw-down type ; but if for some reason quick-closing taps must remain, the only alternative is to provide some form of cushion on which the shock may be absorbed or relieved. For this purpose air-vessels are the most satisfactory fittings, provided they are suitably placed. Air-vessels should be fixed to the pipes where the ID) FIG. 203. Air-vessels for water pipes. Fio. 204. maximum pressure due to concussion occurs ; they must be of sufficient strength and of ample size. In Fig. 203 two different methods of fixing air-vessels are given where the flow of water is in a downward direction. The air-vessels are fixed close to the source of concussion, the plug tap in A being joined directly with the air-vessel, whilst in B the air-vessel is connected with the side of the pipe immediately above the tap. At C, Fig. 204, an air-vessel is fixed at the head of the pipe, and at D of the same figure 20 306 DOMESTIC SANITARY ENGINEERING AND PLUMBING at some intermediate point. If the air-vessel in D is some distance removed from the cause of water-hammer, it will not effectively cure the latter although it will diminish the shock. To provide an effective cushion of air, air-vessels should not be less in diameter than twice that of the pipes to which they are to be attached, and not less than 18 inches in length. Instead of using a special air-vessel, the end of a pipe is sometimes bent upwards so as to serve the same purpose, but in the latter case the capacity of the pipe is too small to be effective. It will be occasionally found that when a quick-closing tap is fixed on a branch, as in Fig. 205, a ball tap at a higher point, and which is connected with the same service pipe, is caused to vibrate and to produce sharp clicking sounds when the bib tap is rapidly closed. The concussion in the branch may be relieved by an air-vessel as shown, but as the pressure in pipe M rapidly falls when the bib tap is opened, there is also a momentary gain of pres- sure in the same pipe when the tap is quickly closed. The fixing of an air-vessel, however, in the immediate neighbourhood of a ball-cock may tend to acceler- ate water-hammer rather than to prevent it. For an air-vessel to be effective in one case and not in another may at first appear anomalous, but when the difference in construction of ball and quick-closing taps is taken into account the anomaly disappears. There is no common cure for all cases of water- hammer, and each case requires to be considered on its own merits. FIG. 205. Air-vessels for water pipes. WATER SUPPLY 307 So violent sometimes are the shocks produced by water- hammer that the sound is transmitted long distances. When several houses are supplied by one common service pipe, it is no unusual thing for water-hammer in one house to be heard in all the others. Under certain conditions a quick-action tap may produce little or no concussion, but this largely depends upon the size of the pipe and the position where the tap is joined. Thus, if a plug stop-cock is connected directly with, or close to a water main, the quick closing of the tap would not greatly affect either the pressure at the tap or that in the water main. The reason for this is rendered clear when the short distance from the plug of the tap to the main, and the relative velocities of the water in the service pipe and in the main, are taken into account. Suppose, for example, a tapping main is 6 inches diameter, and the orifice in the plug of a tap 1 inch diameter, and that the tap is joined directly to the main. Taking the velocity of the water through the tap at 9 feet per second, the velocity in the 6-inch main to yield this would be ~- = \ ft. or o 2 3 inches per second. Thus the concussion which would occur in a main by suddenly arresting such a low velocity would not be of much account. It is very different, however, in the case of long pipes, where the cocks are of the same size, for then the velocity of flow through each is equal, and the greater the velocity the greater the shock when the flow is abruptly stopped. CHAPTEK XI APPLIANCES FOR RAISING WATER HYDKAULIC rams and pumps are largely used for raising water from a low to a higher elevation, but the use of the former is limited when compared with the latter appliances. Pumps take various forms, but they may be classified as follows : 1. Single action lift pumps. 2. Double action lift pumps. 3. Lift and force pumps. 4. Plunger or force pumps. 5. Air-liffe pumps. 6. Centrifugal pumps. Numbers 3 and 4 may be either single or double acting forms, and they may also be arranged to work with one, two, or three barrels. Those in 5 and 6 scarcely come within the scope of this work, and in consequence they will only be briefly dealt with. Lift Pumps. When an ordinary lift pump is used for raising water from a well, the height to which water can be raised is limited by atmospheric pressure. The pressure of the atmosphere varies with altitude and with different weather conditions, but taking the normal atmospheric pressure at sea level to be 14| Ib. per sq. inch, this is equivalent to the pressure exerted by a column of w r ater which is 34 feet in height. As a margin of power must be on the side of the atmosphere to overcome internal resistances of a pump, the maximum height to which water can be raised by a lift pump is about 28 to 30 feet. This height should be measured from the lowest water level to the top of the bucket or plunger when the pump handle is down. APPLIANCES FOR RAISING WATER 309 Iron lift pumps may take the form shown in Fig 206. They may be fixed to a wall or wooden plank, or be made taller than the one shown, and supported by bolting them to stone flags. The bucket B in Fig. 206 is made water-tight at its sides with a cup-leather, and a valve x is arranged to open when the bucket is descending and to close when being raised. The valve V holds up the water in the pump, and prevents its returning into the suction pipe when the bucket is descending. The action of the pump is as follows : When the bucket descends, the upper valve opens, and water escapes above it ; upon the bucket being raised, the water above it is dis- placed through the outlet, and at the same time atmospheric pressure forces up the water through the suction pipe to fill the space through which the bucket has moved. The term " suction " pipe is often misleading, as it indicates that the water is raised by suction instead of being due to displacement by atmospheric pressure. When an iron pump is used the working part of the barrel should be fitted with a thin FIG. 206. Lift pump. gun-metal lining, in order to preserve the cup-leather and to keep the pump in good order. Suction Pipes. For economical considerations, and also for convenience, suction pipes are usually smaller than the 310 DOMESTIC SANITARY ENGINEERING AND PLUMBING barrels of the pumps. The effect of reducing the sizes of suction pipes is to increase frictional resistances, as the water has to flow through them at a greater velocity than through the barrels of the pumps. If, for example, a pump is of 4 inches diameter, and its suction pipe of 2 inches diameter, the velocity through the latter would be 4 times that through the barrel of the pump. Under ordinary circumstances, where a suction pipe is not very long, and where a pump is worked at a slow rate, a suction pipe whose diameter is half that of the pump will give satisfaction. On the other hand, should a suction pipe be very long, it should be increased by one size, and the area of the re- taining valve should also be as large as practicable. Suction pipes when subject to corrosion should also be of a larger size. Frequently when a suction pipe is long a pump is difficult to work. The cause of this may be due to the suction pipe being too small, or to other restricted water passages, such as the lower end of the pipe being partially choked. Should the flow of water through a suction pipe be unduly retarded, the action of the bucket will resemble that of a spring when in tension. If the flow to a pump is not as free as the rate of dis- placement from it, a partial vacuum is created, and the bucket will endeavour to fly back to restore equilibrium when the handle is quickly released. The greatest power when pumping, is required at the commencement of the stroke, as the inertia of the water requires to be overcome before the latter can be put into motion. With a single action pump like Fig. 206 the power to move the piston will vary with the upward and the down- ward strokes, as well as at the commencement of the stroke. To put water in motion in long suction pipes the power required may be much reduced by fixing air-vessels immediately beneath the retaining valves, as in Fig. 207. In form, the air- vessel for a suction pipe may be similar to that of an ordinary air-vessel, but instead of its contained air being in a state of compression it is more or less extended. If it is assumed that a pump is being worked which has an air-vessel attached, as in Fig. 207, at the beginning of each stroke the rising water will compress the confined air to a certain extent, but at the APPLIANCES FOR RAISING WATER 311 completion of the stroke the column of water in the suction pipe will tend to sink a little, and to extend the air in the vessel. This has the effect of providing a force with a spring- like action which pulls against the water, and so makes it ready to be put into motion when commencing to work the pump. Occasionally a non-return valve is fixed in a suction pipe immediately above the water-line of a well or tank from which the water is pumped. Such a valve is useful in certain cases, where the retaining valve of a pump cannot be depended upon, or where two pumps in different situations are connected with one suction pipe. A non-return valve, however, should not be used for a suction pipe which has an air-vessel attached, or the latter would be rendered useless, and in consequence the pump would be more difficult to work. Suction pipes should be arranged so that air cannot lodge in them, and this can be done by making them rise to the pumps for the whole of their length. /5ucTioM PIPE FIG. 207. Suction pipe with air-vessel attached. Pumps for Deep Wells. Where the vertical distance between the lowest water level in a well and the top of a pump bucket exceeds say 30 feet, the working part of a pump barrel will require to be fixed in the well in order that water may be raised. For a well of great depth the suction pipe should be made as short as practicable, or, in other words, the working part of a barrel should not as a rule be more than about 15 feet above the normal water-line, when the lowest water level is only about 5 feet less. In certain wells, where the water level is lowered by pumping to a considerable extent, it may be necessary for the working part of the barrel to be submerged when the highest water-line is reached. 312 DOMESTIC SANITARY ENGINEERING AND PLUMBING When a single action pump is used for raising water from a well of moderate depth, and where hand power only is available, the size of a pump requires to be limited. A type of pump which has often been used in country districts for wells which vary in depth from 20 to 80 feet is given in Fig. 208. The barrel and suction pipe are of lead, the former being 4 inches and the latter 2 inches diameter. To the upper end of the pump barrel a lead head is soldered, and in turn a lead spout is soldered to the head. It will be observed that the barrel stands up inside the head at H in front of the spout, whilst towards the back of the head it is cut away. This arrangement prevents the pump being damaged by pieces of stick, or by small stones being passed through the spout and into the barrel by children ; at the same time the water can freely escape from the back. The pump rod, which is generally of wood, passes through the barrel, and to FIG. 208. Lead pump for deep wells. its upper end the guide arrangement is attached, whilst to the lower end the iron- work of the bucket is fixed. The wood rods are made in convenient lengths for handling, APPLIANCES FOR RAISING WATER 313 and are spliced together as they are passed into the pump barrel. Splices require to be well formed to prevent them failing, and to admit of the different lengths of rod being readily taken apart. A copper cylinder C should be fixed immediately above the access opening A, in order to provide a suitable place in which the bucket can work. A non-return valve is provided at V, and this can be renewed or repaired by means of the opening at A. There is no difficulty in balancing a handworked pump like Fig. 208, as the wood rods are buoyed up with the water and the pump is double acting in principle; that is, it will raise water with both its upward and its downward strokes. A casual glance at the pump may not make this clear, but when it is considered that at each downward stroke a volume of water must be displaced by the rod, then obviously the upward stroke must displace that much less. By making the diameter of the pump rod equal in area to half that of the barrel, the volume of water due to the full length of the stroke may be equally divided and displaced by the upward and downward motion of the rod. A pump similar to Fig. 208 can be worked with a given power which would be totally indequate to operate one with iron rods, provided the conditions with regard to size, and the height through which water requires to be raised, are equal. For supporting lead pumps oak bearers are often em- ployed ; these are fixed on each side of the pump barrel, and pieces of oak board which have been cut to fit around the barrel are nailed across the bearers. At a support two lengths of barrel are joined, and a lead flange is fixed upon the woodwork and soldered to the barrel. A convenient length for lead pump barrel is 8 feet, and such a length can be properly supported in the manner described. In this class of work it is usual to tin the prepared pipe ends, either with a Swedish torch or with a soldering bolt, and to fix the opened ends downward. The latter precaution is necessary to guard against solder entering the barrel when making the joints. To protect the upper part of the pump it should be cased in, and a suitable insulating material should also be used when a pump is fixed in an exposed situation. 314 DOMESTIC SANITARY ENGINEERING AND PLUMBING When a 4-inch pump is used to raise water through a big height, long handles are essential to give the necessary leverage, but at the same time long levers limit the length of the stroke and are cumbersome to use. Double action pumps take different forms, but when they are arranged to deliver double the volume of water of a single action type through a given height, then the power to work them requires to be increased. These pumps as a rule are operated by some form of motive power. The form shown in Fig. 208 will only raise the same volume as a single action pump, but it possesses the advantage of dividing the power to raise the water between the upward and the downward strokes. Lead pumps, of course, are not suitable for raising water for diet- etic purposes if the water has any corrosive action on this metal, but as a rule, well waters contain tem- porary hardness and have no action on lead. Lift and Force Pumps. Fig. 209 gives a lift and force pump, and this type is commonly employed for raising water to the higher parts of buildings when the water supply is derived from a well. The pump is fixed to an oak plank, which in turn is secured to a wall or other structure. As already stated, when a pump is fixed above ground level the vertical distance between it and the water in a well is limited by atmospheric pressure ; so far, however, as the height to which water above the pump can be raised, this is controlled by the mechanical advantage of the lever, and the force a person can bring to FIG. 209. Lift and force pump by Nicholls and Clarke. APPLIANCES FOR RAISING WATER 315 bear upon a handle when working it continuously for a given time. In a lift and force pump the water is put in motion both in the suction pipe P and in the delivery pipe D with the upward motion of the bucket ; the chief resistances to the downward stroke are those due to the stuffing box S, and the cup-leather of the bucket. The guide G keeps the bucket rod E in a vertical position, which is essential when sliding through a water-tight stuffing box. The delivery D, like the suction pipe, should not be less than half the diameter of the pump, or the latter will be difficult to work. Where a delivery pipe is long, and where water is raised through moderate heights, an air-vessel should be attached immediately above the non- return valve n. An air-vessel has the effect of diminishing shock at the commencement of the stroke, and of reducing the power to put the water in the delivery pipe in motion. An air-vessel of a large size should be used, in order that a large portion of the water at each stroke may enter it. The confined air in a vessel on a delivery main acts like a spring in compression, its force per sq. inch being equal to that of the column of water which presses upon the air-vessel. A non-return valve /* is essential for a lift and force pump, to prevent the water in the delivery main from exerting pressure on the bucket when the latter is descending. At T, Fig. 209, water can be obtained directly from a well, and the tap also admits of the delivery main being emptied. In Fig. 210 a plunger or force pump is given. With this form of pump the water is displaced from the barrel with the downward stroke, and the delivery pipe joins at the bottom of the pump as in the figure. So far as the amount of power for working a pump is concerned, the plunger type possesses an advantage over the lift and force form, for with the former the energy required is divided between the upward and the downward strokes, whilst with the latter practically the whole of the burden comes upon the upward stroke. The bucket of a plunger pump is often formed with double cup-leathers, and the non-return valve and air-vessel may be arranged as shown. A draw-off cock should be 316 DOMESTIC SANITARY ENGINEERING AND PLUMBING provided just above the non-return valve, in order that the air-vessel may have its air supply renewed when necessary. When lift and force pumps are used for raising water from deep wells, double barrelled types, Fig. 211, are frequently adopted. When pumps have double barrels the rods balance each other, as one rod is arranged to ascend whilst the other descends ; moreover, double barrelled pumps can be of smaller DELIVERY diameter when compared PIPE. with single forms, and the power necessary to work them can be better utilised. Pumps for deep wells are generally provided PLUNGER, with wheel and cranks in lieu of levers, as the as former admit of the better utilisation of energy when working them, and they can be operated from two sides at the same time by adding the extra handle H, as in Fig. 211. The wheel W should be about 4 feet diameter, and be of moderate weight so as to serve the purpose of a fly- wheel. 5uCTIONJ PIPE. FIG. 210. Plunger type offeree pump. Double barrelled pumps do not require sucli large air- vessels as single pumps, as the delivery of water is more nearly regular. It will be observed in Fig. 211 that twice the length of the crank determines the length of the stroke, and that with an ungeared pump a stroke is completed at each revolution of the wheel. In some cases, where water requires to be raised through more or less considerable height, and where only hand power APPLIANCES FOR RAISING WATER 317 FIG. 211. Double barrelled lift and force pump. 318 DOMESTIC SANITARY ENGINEERING AND PLUMBING is available, gearing is resorted to. The effect of gearing is to reduce the working power by spreading the effort over a longer period ; in other words, for each complete stroke of a pump the wheel will make two or more revolutions, according to the ratio of the gearing adopted. For farms and other places, where horse power is available, that form of energy may be utilised for raising water by providing suitable gearing. When it is necessary to raise large volumes of water through a great height in a limited time, some form of motive power becomes necessary. In country districts windmills or wind engines can often be utilised for raising large volumes of water at a compara- tively small cost, but for these appliances to be effective they require to be fixed in exposed situations. Wind engines are arranged to automatically adjust themselves to suit either moderate or high velocities of wind, and also when blowing from any direction. The size of a mill is governed by the amount of work to be done, and by its relative position to surrounding objects which may obstruct the wind. For example, a 10-foot mill in an exposed situation may be equal in power to a 16 -foot mill which is located in a somewhat sheltered position. Koughly speaking the power of a wind engine varies according to the square of its diameter, when other conditions are equal. It is occasionally found that after a pump has been installed it gets more difficult to work, but the cause as a Kiile is not difficult to discover. For example, if a lift and force pump requires more energy to work it than it should do, the first thing that should be done is to ascertain on which side the cause has been introduced. This can be readily done by opening the draw-off tap so as to empty the delivery main, when a smart stroke or two will indicate whether the bucket works too tightly in the barrel, or whether the suction pipe is partially choked. Should the suction side of the pipe be found satisfactory, then it is obvious that the fault is on the delivery side. Under the latter circum- stances it is possible that the air-vessel has been water-logged, or if this is not the cause, then the delivery main should be APPLIANCES FOR RAISING WATER 319 examined to see if it has been flattened at any point; the non-return valve at the foot of the delivery pipe should also receive attention. Centrifugal Pumps. The ordinary type of centrifugal pump is of simple construction, the water being raised by an impeller which is keyed to a shaft, and which revolves at a high velocity inside a metal casing. The impeller usually contains six vanes, and the water enters at the centre and leaves at the tips of the vanes. To utilise as far as practicable the energy due to the revolving mass, the space into which the water is delivered upon leaving the impeller is gradually enlarged towards the outlet of the pump. This form of construction has the effect of reducing the velocity of the water as it approaches the outlet, and of converting its kinetic energy into pressure. This form of centrifugal pump delivers a constant stream, and is very suitable for raising large volumes of water through a comparatively small height. The efficiency of centrifugal pumps diminishes as the height of the lift is increased. High Lift Centrifugal Pumps. During recent years great improvements have been effected in the construction of centrifugal pumps, and the limited heights to which water could be economically raised by earlier types has been largely overcome by constructing these pumps in compound form. Instead of one chamber, two or more are provided, each having its own impeller, which is keyed to one common shaft. After water is delivered from the first impeller, it passes into the second chamber, and thence through the rising main or subsequent chambers. Single impeller pumps of special design will raise water through a height of over 150 feet. Air-Lift Pumps. - - These can only be used for wells which have a considerable depth of water, and under the best conditions air-lift pumps only have a small mechanical efficiency. For this type of pump, air is delivered under a certain pressure through a nozzle at the botton of the rising main, the lower end of which must be submerged to a greater depth than the height through which water is to be raised. An air compressing plant is essential to operate these pumps, 320 DOMESTIC SANITARY ENGINEERING AND PLUMBING and when sufficient air at the necessary pressure is delivered through the submerged nozzle, the downward pressure of the water in the well is sufficient to overcome that of the air and water in the rising main, when water is caused to rise and escape at the outlet. The air and water do not mix, but form themselves into alternate bands or short columns in the rising main. Efficiency of Pumps. If a pump were perfect it would raise in a given time a volume of water equal to the space traversed by the plunger, multiplied by the number of strokes in the time under consideration. In practice less water than the above would be raised on account of a certain volume slipping back during the closing of the valves and by leakage at the sides of a bucket. If a pump is in good condition it may have an efficiency as high as 99 per cent., and on the other hand its efficiency may fall to, or below, 50 per cent. when in a poor or indifferent state of repair. Lever pumps as a rule are not so efficient as wheel pumps, as the latter have a more nearly uniform stroke, whilst the former are more jerky in action and are not always given a complete stroke. For the purpose of calculation lever pumps will be assumed to have an efficiency of 85 per cent. This is a fair basis to work upon, and a standard which any good form of pump should satisfy when not worked exactly under ideal conditions. Lever Pump Formula. The lifting capacity of an ordinary lever pump upon the basis stated can be obtained by the following formula : 416 Where G = volume of water raised in gallons. I = length of stroke in inches. d = diameter of pump in inches. n = number of strokes per minute. t = time of pumping in minutes. Example 11. An ordinary lift pump is 3J inches diameter, how many gallons of water would it raise in half an hour with a 9 -inch stroke, and when worked at 20 strokes per minute ? APPLIANCES FOR RAISING WATER 321 By Formula 27, G (3 J) 2 x 9x20x30 33075 Substituting values, G = -^ g - = , .'. G = 159 gallons, which are raised in half an hour. Assuming the diameter of a lever pump is required for raising a given volume of water in a certain time. By transposing Formula 27 we have d= /GX416 (28) v Ixnxt Example 12. Determine the diameter of a lever pump which will raise 65 gallons in 20 minutes, if it has a 7-inch stroke and when worked at the rate of 22 strokes per minute. Formula 28 gives d = \/ Gx416 . Ixnxt Substituting the values given, d = 7 X *2t2i X .*. d = 2'96, or say 3 inches diameter. The power which is necessary for operating a lever handled pump can be obtained, when the length of each part of the lever from the fulcrum, and the resistances to be overcome, are known. In Fig. 212 an ordinary pump handle is shown which represeats a lever of the first order. By the following general formulae any one of the four values which are represented by the symbols can be found when the other three are given. In pump calculations the actual weight of a handle is not taken into account, as the bucket nullifies to a great extent the advantage derived by the extra weight of the longer arm. (29) . , (30) 21 322 DOMESTIC SANITARY ENGINEERING AND PLUMBING (31) (32) Where P = the resistance to be overcome when pumping. y = length of short side of lever in inches. x = length of long side of lever in inches (length to centre of grip, as in Fig. 212). F = force or effort applied near end of lever. , Example 13. Suppose the resist- 4 ance to be overcome in pumping is equal to 80 lb., and the lengths of the short and long side of lever are 6 inches and 30 inches respectively, what force should be exerted on the end of the lever ? In Formula 3 1, F = and upon given, substituting the values 80_x_6. 30 ' /. F = 161b. With a force of 16 lb. on the long end of lever the latter would be in a state of equilibrium, assuming that its weight need not be taken into account. In order to make the latter form- ulae applicable to pump work, it is necessary to ascertain the value of P. The actual resistances to be overcome when pumping vary with the different types of pumps. In lift-and-force pumps frictional resistances are greater than those of ordinary lift pumps, and lift-and-force pumps when geared offer still greater resistance by friction. Formulas for Lift Pumps. The total resistance to be FIG. 212. Sketch illustrat- ing mechanical advantage of lever. APPLIANCES FOR RAISING WATER 323 overcome when raising water by an ordinary lift pump Fig. 206, may be obtained by the following formula : ^ /Q o\ -IB" (33) Where P = total resistance in Ibs. to be overcome. d = diameter of pump in inches. h = height in feet through which water is raised. When the right side of the equation in 33 takes the place of P in Formulae 31 and 32, we obtain the following : ' ' (34) and by transposition, d = ^ . . . . (36) When an ordinary lift pump is used for raising water for supplying a house, its diameter as a rule does not exceed 4 inches. Where the short and long parts of a lever are in the ratio of 1 to 6 a 4-inch pump should not be difficult to work. Example 14. If a 4-inch lift pump raises water through a height of 24 feet, and the short and long sides of lever are 6 inches and 36 inches respectively, determine the effort which must be applied at the end of the lever. -D T? i OA ^ By Formula 34, F = 4 2 x9x24x6 576 Substituting values given, 1 = ^5x3Q = "25 ' /. F = 23 gV, or say 23 lb., which must be applied at the end of the lever. The effort that can be applied by a person is limited, and as a rule this will vary from 18 to 25 lb. when pumping continuously for fairly long periods. The length of a pump handle is also limited, for when a handle is very long it cannot be raised sufficiently high to utilise the full length of its stroke. It is also cumbersome to work. The diagram Fig. 213 shows the distances through which 324 DOMESTIC SANITARY ENGINEERING AND PLUMBING the hand travels with handles of varying lengths when the levers are moved through 90. The vertical distances through which the hand is raised are also given. With a handle 2 feet long it will be seen that the distance through which the hand travels is 3 ft. If in., whilst the vertical height is 2 ft. 3 in. For a 4-foot handle the length of the J ! / / ' /' ;i / /i /i , i / /i /i / / ' i / 1 / Vv x/ -^ /"*! / , / / / / V' 7 y ^v 7 >= /; i/ -' /I / CO /I / ' / / / / / ^'" jff ' /' J > x / ' & i 1 / 1 / - "^" "*" X" 1 x , / f x ^' ^J 1 x ^"^ $ V x ^ x ' / i ___-'*"" i ^~-~" 1 FIG. 213. Diagram illustrating limiting lengths for pump handles. arc and the vertical distance are 6 ft. 3f in. and 4 ft. 5J in. respectively. In Fig. 213 the short lever arm is 6 inches, and when this is turned through 90 it gives a stroke of 8J inches. Should the length y be less than 6 inches the stroke would be shorter than shown, unless the handle were moved through a greater number of degrees. Formulae for Lift-and-Forqe Pumps. To find the diameter APPLIANCES FOR RAISING WATER 325 of a lift-and-force pump for raising water through more or less considerable height, and where the power and leverage are limited, the following formula can be used : 50x#xF Where d diameter of pump in inches. x = long side of lever in inches. y = short side of lever in inches. F = force applied on pump handle. h = height through which water is raised. Example 15. Determine the diameter of a lift-and-force pump for raising water through a total height of 75 feet, when the short and long parts of lever are 5 inches and 33 inches in length, and when a force of 24 Ib. is to be applied at the end of the lever. / By Formula 37, ^ = V 50x#xF 9xTx? _ Substituting values given, d = /5Qx 33x24 ^ 19x75x5 /. ^ = 2*35 inches diameter. As this is an odd size the nearest stock size would be selected. Transposing Formula 37, we have (38) 50 xx Example 16. Supposing a lift-and-force pump, Fig. 209, is of 3 inches diameter, and is used for raising water through a height of 60 feet; what force must be exerted on the end of the lever when the short and long sides are 5 inches and 32 inches respectively ? By Formula 38, F = ^^A2iy- OU X X a , ... .. ,, 3 2 x 19x60x5 513 Substituting values given, 1 = -- 50^32 -- = l6~ ' /. F = 32 T y, say 32 Ib. Example 17. Give the volume of water which would be raised by the pump in the last example in fifteen minutes, if it 326 DOMESTIC SANITARY ENGINEERING AND PLUMBING has a 7 -inch stroke and is worked at the rate of 22 strokes per minute. By Formula 27, G = Substituting values given, G = d^xlxnxt 416 3 2 x7x22xl5 10395 416 208 ' = 49*9, or say 50 gallons. Wheel-pumps have an advantage over those with lever handles, as the former give a full stroke with each revolu- tion of the crank they better utilise the energy imparted to them, and are less jerky in action. By the aid of Fig. 214 the mechanical ad- vantage of a wheel will be made clear, besides helping to explain the meaning of the symbols used. Formula for Single Action Lift - and - Force Wheel Pumps. For lift- and-force pumps like Fig. 211, and where the rods are balanced, the following formula may be used : (39) (40) FIG. 214. Illustrating mechanical advantage of wheel handle. 50xE /Fx50xE d=\J ig xh/xr Where F = force in Ibs. applied to handle of wheel. r = radius of crank in inches. E = radius of wheel as in Fig. 214. d = diameter of pumps in inches. h = height through which water is raised. APPLIANCES FOR RAISING WATER 327 It will be observed, upon reference to Fig. 214, that the length of the crank r determines the length of the stroke, which is equal to twice that of the crank. For convenience the radius E of pump wheel is often limited to 14 or 15 inches, but where the resistances to be overcome are large a bigger radius may be adopted, or gearing may be resorted to. The usual form of gearing is shown in Fig. 215, where the FIG. 215. Gearing for double barrelled deep well pump. pump wheel is connected to a shaft on which are small cog wheels, which in turn react on larger cog wheels to which the crank rod is joined. The ratio of the gearing is obtained by counting the number of cogs in each wheel. Thus, if the upper and smaller wheel contains 6 cogs, and the larger or lower wheel 15 cogs, then the gearing is in the ratio of 15 to 6, or 2 J to 1 ; in other words, 2J revolutions of the fly wheel are necessary to complete one stroke of the pump. For purposes of calculation the higher number will be expressed 328 DOMESTIC SANITARY ENGINEERING AND PLUMBING as a ratio of the smaller, and for the case referred to this ls ^7. Z 2 Formulae for Geared Lift-and-Force Pumps. equals 7. , . . (42) Where = the ratio of the revolutions made by the wheel to one made by the crank. The remainder of the notation as before. Example 18. Find the force which must be applied to the handle of a 3-inch diameter lift-and-force wheel-pump, in order to raise water through a height of 120 feet, when the radii of the crank and wheel handle are 4J and 18 inches respectively. As the pump in the example is not geared, the force to work it will be found by Formula 39. Where F = ^ -- . 50 xK Q ,,.,,. , . 3 2 xl9xl20x4J 513 Substituting values given, F = - ^ ^ -- = -=- ; OU X -Lo .-. F=102f Ib. The working shows that the pump is unsuited for manual labour where only one or two persons could operate it. If it is assumed that two men could be employed to work the pump, each would require to exert about 51 Ib. on the wheel handle. Example 19. Suppose, now, we desire to find the diameter of a pump which can be worked by one man when exerting a force of 28 Ib. on the wheel handle, in order to raise water through a height of 120 feet. Assume the wheel is geared in the ratio of 3 to 1, and that the radii of the crank and the wheel are 4J inches and 18 inches as before. By Formula 42, d= Xhxqxr APPLIANCES FOR RAISING WATER 329 c ,,.,,.., , . , / 28 x 5 x 18 ,- Substituting the values given, ^ = V 2 x 120xx4 = ' ' /. d=2'64, or say 2| inches diameter. Should the lifting capacity of a geared wheel-pump be required, it may be directly obtained by the following : _ 190 Where G = gallons raised. d diameter of pump in inches. r = radius of crank in inches. n = number of revolutions made by fly wheel per minute. t = time in minutes. q = the ratio of gearing adopted. B = number of pump barrels. Example 20. Find the volume of water which should be raised in 35 minutes by a double-barrelled lift-and-force pump of 2 inches diameter, where the fly wheel makes 30 revolutions per minute and where the crank has a radius of 4J inches. The wheel is also geared in the ratio of 2J to 1. By Formula 43, G = * xrx * x * xgxB . j.y \j Substituting the values given, 190 19 : , or say 80 gallons. Hydraulic Rams. Where sufficient water is available for working hydraulic rams, they are very suitable for auto- matically raising water for supplying mansions, farms, hotels, hamlets, etc. These appliances may be divided into two classes, viz., those which raise a portion of the water which operates them, and those which utilise the energy from one source in order to raise water from a separate source. The latter are occasionally termed ram-pumps. Each class of ram varies widely in constructional details, but when well 330 DOMESTIC SANITARY ENGINEERING AND PLUMBING constructed rams will work with a very small head of water. A head of 18 inches and a little less is sufficient to operate certain rams of the first class, provided they are properly fixed. For every foot of working head, water may be raised by a ram through a height of over 50 feet, but the efficiency of this appliance rapidly falls when only small heads are used to force water to high elevations. FIG. 216. Walter Simpson's hydraulic ram. A section of a ram by Walter Simpson, Aberdeen, is given in Fig. 216, the inlet being at I and the outlet at 0. The ram contains two valves, the larger one D being known as the dash valve, and the smaller one V acts as a non-return valve. It will be observed that the dash valve is closed with an upward motion, whilst the non-return valve opens in the direction of the flow. For its action, the ram depends upon the velocity of the inflowing water being suddenly arrested by the closing APPLIANCES FOR RAISING WATER 331 of the dash valve D, when the pressure in- side the ram is suddenly raised. At the period of maximum pressure a small volume of water is forced through the valve V into the air-vessel, from which it passes through the outlet into the rising main. Although water escapes into the air-vessel, the resist- ance to its entrance is sufficient to cause the water in the supply pipe to recoil when the dash valve opens by its own weight. As soon as the energy which produces the recoil or backward motion has been ex- pended, the water in the supply pipe again regains its forward flow, and after a brief interval, during which a certain volume escapes through the open dash valve, the latter is again suddenly closed and the operations repeated. In Fig. 217 a ram is shown in position, together with the drive and delivery pipes. This illustration will also aid in making clear some of the points which require consideration in ram work. For the suc- cessful working of a ram, the drive pipe plays a very important part; if this pipe is too short, the dash valve beats too rapidly, and the ram will not do effective duty, owing to insufficient resistance being offered to the recoil of water when the dash valve is closed. To make the point clear we will assume that the length of drive pipe to a ram is, say, 15 feet; and that the working head is *10 feet, which is repre- sented by H in Fig. 217; let the height through which water requires to be raised be, say, 100 feet. Under these conditions 332 DOMESTIC SANITARY ENGINEERING AND PLUMBING the working head is 10 feet and the resistance to be over- come is that due to a head of 100 feet. Now, in order for the ram to raise water, the pressure inside it when the dash valve is closed will require to exceed that due to the head on the delivery side. When a ram is working, two forces operate to relieve the increased pressure which is due to the sudden closing of the dash valve; one is the water which escapes through the retaining valve and into the air-vessel, and the other is the recoiling water in the drive pipe. If, now, the resistance offered to the recoil of water is less than that due to the opening of the retaining valve, the dash valve will beat, and the ram will appear to be working, when in reality no water is being raised. In the case under consideration, where the length of the drive pipe is only 15 feet, the resistance offered to the recoil would be insufficient, and in consequence the increased pressure would be principally relieved by the drive pipe, and the recoil would occur at a quicker rate. For reasons stated it becomes obvious that a long drive pipe is essential, and as a general rule it should not be less in length than the height through which water requires to be raised. Long drive pipes require to be of adequate size, otherwise the dash valve will not close quickly enough. When the recoil in a drive pipe takes place, the water is affected through the whole of its length, so the longer the pipe the greater the internal resistances,' and the longer the interval between the beats of the ram. When laying a supply or drive pipe, it should have a gradual rise from the ram to the source of supply as in Fig. 217, and where bends are necessary they should be made as easy as practicable. A rose, or strainer, should be provided at the inlet end of the drive pipe, to prevent any matter passing into it and so interfering with the working of the ram. Either iron or lead drive pipes may be used according to the size required, but when cast-iron spigot and socket pipes are adopted the joints should be made with rust cement. Lead is not a suitable jointing material in this case, as the joints require to be made with a material which imparts a greater degree of rigidity. Yielding joints on a drive pipe impair the efficiency of a ram. APPLIANCES FOR RAISING WATER 333 When rams have a high working head they are subjected to considerable strain, and as a rule it should not exceed 30 feet. Where possible the supply of water to a ram should be sufficient to maintain a constant head upon it, or, in other words, the supply tank T, Fig. 217, should be always full. When, however, a ram is supplied by a spring or stream which has a varying yield, a much larger supply tank will be neces- sary than where a large and fairly constant volume of water is available. Should a supply tank become emptied by the outflow exceeding the rate of inflow, the ram will cease to work, and the water available will flow through the open dash valve and thence to waste. To obviate this waste of water a float valve may be attached to the end of the drive pipe in the supply tank, in order to automatically shut off the supply to the ram when the water level has been lowered to a certain point. During the refilling of the tank the float valve opens, and the ram can then be restarted by holding down the dash valve for a few seconds, or by means of a pumping valve at the supply tank. The reason why the ram does not restart itself is due to the gradual closing of the dash valve when the supply is being cut off. Air-vessels for rams should be of a large size, in order that water at each beat may be first directly discharged into them, without much compression of the contained air. Rams occasionally fail to raise water owing to air-vessels getting water-logged; the beating of the dash valves may continue, but the resistance on the delivery side is too great when it is necessary to first put the water in the rising main in motion. When a ram is in constant use the air-vessel should be re- charged with air at regular intervals of, say, once a week. In a ram the water is under more or less considerable pressure when it enters the air-vessel, and therefore the capacity of water for the absorption of air is increased, and, owing to the water in the air-vessel being continually changed, the latter, in consequence, is gradually deprived of air. To renew the air, one or two cocks are provided at the base of an air-vessel, and when these are opened and the 334 DOMESTIC SANITARY ENGINEERING AND PLUMBING valves on the drive and delivery pipes are closed, the vessel is readily charged with air. In many rams a small air or sniffle valve is provided for making good a portion of the air, but it cannot be entirely relied upon for automatically supply- ing the requisite volume, although it will lengthen the interval between the periods of recharging as above described. An air or sniffle valve opens and admits air during the recoil of the drive water, and it is forced along with water into the air-vessel at each beat of the ram. The principal causes for rams getting out of order are due to faulty dash or retaining valves, to air-vessels getting water-logged, to defects in the pipes, to air in the drive pipe, and to an insufficient supply of water. Within certain limits a ram can be made to use less water by shortening the stroke of the dash valve by adding one or more washers at K, Fig. 216, but the volume of water raised is also diminished, and the pulsations of the ram are made at a quicker rate. If a ram has never given satisfaction, this may be due to structural defects, or to the drive pipe being irregularly laid so as to permit of the lodgment of* air, or to the drive pipe being too small, or to insufficient length. The leakage of a non-return valve, or the lodgment of air in a drive pipe, causes the dash valve to remain closed. A defect in the lower part of a delivery pipe would also have a similar effect. Water-logged air-vessels, and drive pipes which are too short, allow pulsations to continue without raising water. When a ram is newly started, the water will, of course, rise in the delivery pipe to the same level as that in the supply tank. At this period, provided there is an ample supply of water, the dash valve will be closed, and in order to start the ram it will be necessary to open the dash valve several times, by hand, until water is forced through the delivery pipe to a height which offers sufficient resist- ance, to bring about the recoil of the water in the drive pipe. With regard to the volume of water a ram will raise, this can be calculated when its efficiency for the given conditions is known. When a certain volume of water flows through a APPLIANCES FOR RAISING WATER 335 drive pipe, its energy is usually calculated in foot-pounds. This is obtained by multiplying the weight of water by the height through which it falls. For example, if the working head on a rani is 10 feet, and 100 Ib. of water are delivered to it per minute, the energy in the drive water for the time given is equal to 100x10 = 1000 ft.-lb. Assuming that 10 Ib. of water are raised in the same interval of time through a height of 90 feet, then the energy to raise this volume through the height given, when frictional resistances are neglected, is equal to 10x90 = 900 ft.-lb. But as the drive water contains 1000 ft.-lb. of energy, then (1000-900) = 100 ft.-lb. which are absorbed by friction and by leakage, etc. Under these circumstances the efficiency of the ram ., , 900x100 ftA would be JQQQ = 90 per cent. Of the 100 Ib. of water which are delivered to the ram only 10 Ib. are raised, the remaining 90 Ib. escape through the dash valve and flow away to waste. For making calculations in connection with rams the following formulae may be used : GxHxe r~ -JT-; '. ' (44) G-* :.- - . - - - (45) Where G = gallons supplied to ram in any given time. q = gallons raised during the same interval of time. H= head of water upon ram. h = height through which water is raised. e = efficiency of ram. In Fig. 217 H and h are indicated. Only approximate values of e can be given, as these vary with different ratios of H and h, and with different makes of rams. The efficiency of a ram is also influenced by the manner in which it is fixed. 336 DOMESTIC SANITARY ENGINEERING AND PLUMBING TABLE VI VALUES OF e Where g = 2 3 4 5 6 9 12 15 20 e = 88 85 81 75 72 63 5 -4 3 ! Example 21. A ram when working with a full beat requires a supply of 12 gallons per minute. Assuming the ram has an efficiency of 85 per cent, when raising water through a height of 120 feet, and when the working head is 19 feet, how many gallons should it raise per hour ? Supply per hour = 12 x 60 = 720 gallons. , GxHxe By Formula 44, q Substituting values, q = h 720 x 19 x '85 120 .-. 2 = 96'9, or say 97 gallons per hour. Example 22. In 24 hours a ram delivers 140 gallons through a height of 160 feet. If the working head is 15 feet, efficiency of ram 64 per cent., determine the rate of supply per minute the ram will require. The volume of water raised per minute Using Formula 45, G = n 160 x -097 Substituting values given, G = -y= ^ ; /. G = 1'6 gallons per minute as the volume required. Example 23. Find the efficiency of a ram which raises 950 gallons per day through a height of 75 feet, when the rate of supply is 18,000 gallons per day and when the working head is 7 feet. By Formula 46, *= Substituting values given, e = GxIT 75x950 18000x7 /. e = '565, or per cent. APPLIANCES FOR RAISING WATER 337 The following table gives approximate sizes of drive and delivery pipes for rams : TABLE VII Water delivered to ram per minute. Diameter of drive pipe. Diameter of delivery pipe. 1 to 3 gallons. 1 inch. inch. 3 8 1^ 1 8 16 2 16 24 2 i 11 24 32 3 li 32 50 3i ]i 50 60 4 2 A hydraulic ram of the second type, by Keiths, Blackman & Co. Ltd., is shown in Fig. 218. With this appliance, a pure and limited supply of water can be raised by means of impure water, when the latter is obtainable in sufficient volume. The two waters are not able to mix, as each is supplied to separate parts of the appliance, and any leakage at the pistons is free to escape to the exterior of the ram. From the figure it will be observed that the lower part resembles an ordinary ram, whilst the upper part resembles a single action pump. Where practicable, the pure water should flow by gravity to the appliance, but "if this is not possible, then it may be raised a few feet through a suction pipe. The action of the ram pump shown is as follows, where the pure water will flow by gravitation to the ram. When the piston K is in the position shown, water flows through the non-return valve V 1 , and fills the cylinder C. The impure water which supplies the motive power enters at A, and when the dash valve D is rapidly closed, the energy due to concussion is exerted on the piston P p which is directly joined by means of rod K to the upper piston P 2 . With each beat of the dash valve, the pistons are caused to rise, and to displace the pure water from cylinder C, through the retaining valve V 2 and into the air- vessel ; from the latter, the water flows through the delivery pipe to the point desired. During the recoil of the water in 22 338 DOMESTIC SANITARY ENGINEERING AND PLUMBING PURE WATER^ i SUPPLY PIPE. * OUTLET FIG. 213. Keith and Blackman's hydraulic ram pump. APPLIANCES FOR RAISING WATER 339 the drive pipe the dash valve opens, the pistons descend, and the cylinder C receives a fresh charge of pure water. These operations are repeated so long as the ram continues to work satisfactorily. To aid the downward motion of the pistons, a weighted lever operates on the rod at W. It is now clear why pure water should flow by gravity to the ram, for as it requires to be raised by the downward motion of the pistons, the power available is only that obtained by means of the weighted lever. CHAPTEK XII HYDROSTATICS AND HYDRAULICS Hydrostatics is that branch of science which treats upon the equilibrium of fluids. In this case it is confined to the pressure of fresh water when the latter is at rest. If a cistern forms a cube, and each side is 1 foot long, the weight of water the cistern would hold when full is 624 Ib. This value also represents the pressure which would be exerted on the bottom of the cistern, but as its sides are acted upon as well, the total pressure exerted by the water is not synonymous with its weight. Under the force of gravity, pressure is due to head of water, and upon any unit area in a horizontal plane the pressure is the same. For example, if a cistern is 4 feet deep, and is filled with water, the internal pressure on the bottom per square foot of surface equals 624x4 = 249*6 Ib. On a vertical surface the pressure varies with the depth of water, being zero at the water level, and a maximum along the bottom edge. In Fig. 219, the lengths of the horizontal broken lines which are enclosed by the triangle ABC represent pressures at different depths, BC being equal to AB, or to the depth of water in the tank. To calculate the pressure on a vertical surface, it is necessary to know the average head of water which acts upon that surface. For example, the average head on the side AB, Fig. 219, is -y- , or half the depth of the water. Should the Lt average head be required for a portion of a vertical surface such as ef, Fig. 219, then this would be l ~I 2 where \ and h 2 represent the heads upon e and / respectively. If \ = 2 feet 340 HYDROSTATICS AND HYDRAULICS 341 and 7fc 2 = 3 feet, the pressure acting upon each square foot of 24-3 surface between the points given equals - - x 62*4 = 156 Ib. a As the pressure per sq. foot per foot of head is equal to 62*4 Ib., the pressure per sq. inch for each foot of head equals 62-4 .00 IK _ = -4331b. If a closed receptacle, such as a cylindrical tank or boiler, be supplied with water from an overhead cistern, the pressure of the water is transmitted over the whole internal surfaces of the vessel, and the intensity of the pressure at any point is A m c FIG. 219. Diagram illustrating water pressure. proportional to the head of water above that point. On any given horizontal surface, pressure is transmitted over the whole area with undiminished force, and it acts at right angles to the surfaces. The length and diameter of a supply pipe in no way affects the pressure transmitted upon a surface, when water is at rest, although both size and length materially affects the discharging capacity of a pipe. From this, it becomes clear that, if pressure is transmitted upon a horizontal surface by means of a supply pipe, the total pressure on that surface is equal to the weight of water contained in a cistern whose plan is the same as the surface, and whose depth is equal to the head of water under consideration. 342 DOMESTIC SANITARY ENGINEERING AND PLUMBING For determining pressures the following formulae may be used : P = A x 62-4 x h ' .' : ""'"' . " . ; ; ; ' ,' . (47) . .; .--; >;-,..; (48) . . . . ...-'- .' (49) Where P = total pressure in Ibs. p = pressure in Ibs. per sq. inch. A = sectional area of surface in sq. feet. a = sectional area of surface in sq. inches. 7i = head of water in feet. For the formulas given it is necessary to either find the area of a surface or to substitute values which produce it, but modified formulae may be used for special cases. Thus, to determine the pressure on the inner surfaces of cylindrical vessels, the formulae may take the form given below. Formulae for determining the total pressure on the Sides of Cylindrical Vessels P = Dxl96xAxL . . . '. ' . -. . (50) . . , . . (51) Formulae for obtaining total pressures on the ends of Cylindrical Vessels and on Pistons, etc. :. ... , .. :| . (52) . ; ? . :\ . (53) p from which, h = ^2^34 C 54 ) Where P = total pressure in Ibs. D = diameter in feet. d = diameter in inches. h = head of water in feet. L = length of pipes or cylinder in feet. I = length of pipes or cylinder in inches. A few worked examples will aid to show the range of problems to which the above rules may be applied. Example 24. Find the total distributed pressure, and the average pressure per square inch, on the vertical surface of a copper hot-water tank which is 1 ft. 9 in. diameter and HYDROSTATICS AND HYDRAULICS 343 3 ft. 6 in. high, when the head of water above the centre of the cylinder is 36 feet. Using Kule 50, P = Dx 196 x^xL. Substituting values given, P = If x 196 x 36 x 3| ; /. P = 43,218 Ib. total distributed pressure on vertical surface. For the second part of the problem use Formula 49. Where p = hx '433, ^ = 36x433; .\p = 15-58, or about 16 Ib. per sq. inch. Example 25. If a 6-inch diameter drain is tested for soundness by the hydraulic test, determine the total force which tends to displace the stopper, if the latter is subjected to a mean head of 12 feet of water. By Formula 53, P = & x '34 x h. Substituting values given, P = 6 2 x '34 x 12 ; .*. P = 146-88, or nearly 147 Ib. Example 26. What is the total distributed pressure which acts upon one side of a cistern which is 6 ft. 6 in. long, 4 ft. 6 in. wide, and 3 ft. 9 in. deep ? The highest water level is 3 inches below the top of cistern. 3' 9" 3" Average head on side = - ^ - = lf feet, and area of surface pressed upon = 6Jx3J. Using Formula 47, P = A x 624 x h. Substituting values given, P = 6 J x 3 x 624 x If ; /. P = 2484-3 Ib. Example 27. A dead-weight safety valve is loaded to the extent of 6 Ibs. Assuming the valve orifice is -f- inch diameter, find the head of water which will exert pressure equal to the load given. Formula 54 gives A = -^ " Substituting values given, h = X * / x ** = 4517, or say 45 feet 344 DOMESTIC SANITARY ENGINEERING AND PLUMBING In the last problem the load on the valve represents the pressure which tends to close it, and therefore takes the value of P as shown. Hydraulics is that branch of science which treats on liquids when in motion. To put water into motion, a certain amount of pressure is required, which depends upon the nature and magnitude of the resistances to be overcome. If a house is supplied by water from a tank in an elevated situation, the pressure of the water at any point in the supply pipe, provided the water is at rest, is equal to the vertical distance between that point and the surface of the water in the supply tank. Should a draw-off tap, however, be partly opened, the water in the pipe is put into motion, and the pressure is reduced. If the tap is opened wide, the pressure in the pipe is still further reduced. The principal factors which require taking into account when ascertaining the flow of water through pipes and orifices are as follows : (a) Pressure absorbed in overcoming the inertia of the water and putting it into motion. (6) The form the outlet orifice takes. (c) Pressure absorbed by pipe friction. (d) Pressure absorbed by sudden contractions in pipes, and by abrupt changes of direction. When pipes are of considerable length, the pressure absorbed by (a) and (b) is a negligible quantity, as it is so small when compared with that absorbed by pipe friction. On the other hand, the pressure absorbed by (a) and (b) when pipes are short is of more importance, as it may represent a large percentage of the total pressure. The form of an orifice affects the rate of discharge, because the stream lines converge to a more or less extent to form a contracted neck vena contrata just beyond the opening through which the stream lines issue, and instead of the sectional area of flow being equal to that of the orifice, it is only equal to the area of the contracted part. It is not, of course, possible to measure the point of greatest contraction for different forms of orifices under ordinary conditions, but, as there is a Definite relation between the area, of contraction HYDROSTATICS AND HYDRAULICS 345 and that of an orifice, the latter is expressed as a ratio of the former, and termed a coefficient. Thus, for a short tube of 1 inch diameter, the stream at the point of greatest contraction measures *9 or ^ inch diameter, and the ratio of the area of .92 the tube to that of the contracted neck is -^- = '81. Suppose now we require to find the discharge through a short tube, when the water at the point of greatest contraction has a velocity of v feet per second, and where A equals the area of the tube in feet. The volume of water discharged by the tube would equal '81 xvxa = cubic feet per second. Mul- FIG. 220. Illustrating head of water above orifice. tiplying by '81 makes the necessary correction for the stream lines in this case. Coefficient for orifice in thin plate . . = *62 good shaped nozzle . . ='94 short tube where length equals 2 to 3 diameters = '81 short tube where length equals 4 to 12 diameters = '77 short tube where length equals 13 to 24 diameters =*73 Plow of water through Orifices and Short Tubes. The maximum velocity with which water issues through an orifice 346 DOMESTIC SANITARY ENGINEERING AND PLUMBING or short tube in the side of a cistern, Fig. 220, is nearly the same as that acquired by a body which has fallen from rest through a height h, which is measured from the centre of the orifice to the water surface. The velocity of a falling body is found by the general formula v = V2fii -- . - - * (55) Where v = velocity in feet per second. g = force of gravity = 32*2. li height fallen through in feet. For finding the discharge in gallons, the velocity formula may be modified to take the form beneath. G = d 2 xl6-3xcxV . . (56) Where G = gallons discharged per minute. d = diameter of orifice in inches. c = coefficient which varies with the form of orifice. (See page 345.) h = head of water above centre of outlet. Further simplification is possible for formulae in connection with any special form of aperture ; thus, for a short tube, where c = '81 the two constants may be multiplied together, when we have 16*3 x 81 = 13*2. In practice the decimal may be omitted and the value taken as 13. Formula for Short Tube. G = d 2 xl3xVA . . . . (57) / - G - ' By transposition d = /u ^ . . . (58) A ' .-.. (59) Example 28. If a short tube of 1J inch diameter is under a constant head of 2 ft. 6 in., find its rate of discharge. (See Fig. 220.) By Formula 57, G = d 2 x 13 x Vh. Substituting values given, G = (1 J) 2 x 13 x V2-K G -5 X 5 13 158. ~4 X 4 T' IS" .'. G = 32*09, or say 32 gallons per minute. HYDROSTATICS AND HYDRAULICS 347 Example 29. What head of water would be necessary to discharge 15 gallons per minute through a short tube of 1 inch diameter ? By Formula 59, A = 15 /. h = 1-33 feet, or say 1 ft. 4 in. Flow of Water through Long Pipes. When water is flowing through pipes of more or less considerable length, the chief resistance is that offered by the surfaces of the pipes. The velocity of a particle of water varies according to its distance from the surface of a pipe, its velocity being greatest at the centre, and the least against the surfaces of the pipe. Because the velocity throughout the cross section is not uniform, the size and condition of a pipe have a marked effect upon its average, or mean velocity of flow. Speaking generally, when water is flowing through a pipe its mean velocity is proportional to the square root of its hydraulic mean depth, to the square root of the pressure head, and inversely proportional to the square root of its length. By the aid of the following formulae many problems in connection with long pipes may be solved. These make allowance for bends in pipes, when the latter are laid or fixed in the usual manner. Formulae for Long Pipes. (60) < 63 > 348 DOMESTIC SANITARY ENGINEERING AND PLUMBING Where G = gallons discharged per minute. d diameter of pipe in inches. li head of water in feet. I = length of pipe in feet. /=a coefficient which varies with the size of pipes, and is given in the following table : TABLE VIII VALUES OF / Lead pipes from ^ inch to 1 inch diameter /=210 , li , U , . . /=260 Iron > u 14 3 , . /=210 M 2 , 3 j i /=330 } II 34 , 54 , 5 7 ' /=460 /=570 To use Formulae 60 to 62 easily, a knowledge of logarithms is necessary, but for those who are not familiar with this branch of mathematics, problems may be solved by the aid of the table below. TABLE IX Diameter pipe. 5th power diameter. Diameter pipe. 5th power diameter. 4 inch. 03125 3| inch. 525-218 2 2373 4 1,024-000 1-0000 4i 1,845-281 li 3-0517 5 3,125-000 14 7'594 54 5,032-843 2 32-000 6 7,776-000 24 97-656 64 11,602-906 3 243-000 7 16,807-000 When a pipe is used for conveying water from a storage tank T to a point P, Fig. 221, and when discharging full bore, the head absorbed by friction at different points of the pipe is represented by the vertical distances between the hydraulic grade line No. 1, and the horizontal line L; the latter repre- HYDROSTATICS AND HYDRAULICS 349 sents the level of the water in the supply tank. The vertical distances between No. 1 hydraulic grade line and the water pipe indicate the pressure at any point when the pipe is discharging full bore. The hydraulic grade line simply indicates the level to which water would fall in vertical tubes, were it prac- ticable to obtain them suffi- ciently long and to join them with the pipe in question. In Fig. 221 the pipe is supposed to be of uniform bore from end to end, and the hydraulic grade line is shown to form a straight line from T to P. A true hydraulic grade line, however, like the one shown, can only be ob- tained when the whole of the pipe line is below it. Example 30. Find the gallons discharged per minute by a 3-inch cast-iron pipe at the point P, Fig. 221, when the head of water and length of pipe are as shown. Total head of water avail- able above P = 360 -100 = 260 feet. Length of pipe given is 6500 feet. By Formula 60, G= /rf*xj V 350 DOMESTIC SANITARY ENGINEERING AND PLUMBING In Table VIII. it will be found that the value of / for a 3-inch pipe is 330, and in Table IX. the 5th power of a 3-inch pipe = 243. 26Q ' Substituting values, G = ^/ 243 .-. G = 56'6, or say 57 gallons per minute. In this calculation the whole of the head has been utilised in discharging 57 gallons per minute at point P. Suppose, however, a discharge under a given pressure is required at P; under the latter conditions the whole of the 260 feet of head would not be available for forcing the water through the pipe, and the hydraulic grade line would require to be raised. Example 31. What diameter of pipe would be required to discharge 100 gallons per minute under a pressure of 50 Ib. per sq. inch, at P, Fig. 221, when the length of pipe and head are as shown ? The equivalent of 50 Ib. per sq. inch is 50x2-31 = 115-5 feet head. The value 2'31 represents the head in feet which exerts a pressure of 1 Ib. per sq. inch. For delivering the volume required, the head available will be 360-(100 + 115-5) = 144-5 feet. Now by Formula 62, d= The value of /, however, is a variable quantity, and we do not know the precise value to assign to it when beginning the problem. It is therefore necessary to select a trial value of/, and if it does not agree with the diameter obtained, as shown in Table VIIL, the problem must be reworked with either a lower or a higher value, as may be found necessary. We will assume the diameter required is somewhere between 3J and 5 inches, and for this range the value of / in Table VIII. is given as 460. / Substituting values, d = I 5/ 100 2 x6500 460" x 144-5 ' HYDROSTATICS AND HYDRAULICS 351 Working by logarithms Log. 100 2 =4 Log. 460 =2-6628 Log. 6500 = 3-8129 Log. 144'5 = 2-1599 7-8129 4-8227 4-8227 5th root 5)2-9902 5980 Antilog. -598 = 3-963; .-. d = 3'9*63, or say 4 inches diameter. This problem may be worked by ordinary arithmetic with the aid of Formula 63 and Table IX. By Formula 63, c? 5 Substituting values as before, d 5 = fxh' 100 2 x6500 460 x 144-5 ' .-. d 5 = 977-8. Upon reference to Table IX. it will be found that the 5th power of 3J = 525; this value, however, is too low, and the diameter which agrees with the next higher value is the size required. As before, the pipe required is of 4 inches diameter. When water is discharged at the point of escape under a pressure of 50 Ib. per sq. inch for the conditions shown in Fig. 221, the hydraulic grade line is raised to the position occupied by the straight line No. 2. If a pipe line follows the general configuration of the earth's surface, as in Fig. 222, it should form two sections of different diameters. The first section would be from A to B, and the other from B to C, each having its own hydraulic grade line as shown. If a certain volume is required per minute at C, it is obvious that a similar volume must first be delivered in the same time at B. When comparing the two sections, that from A to B is 8200 feet long, and the maximum head above the latter point is 35 feet; for the section BC the total length is 650 feet, and the vertical distance between B and C is 185 feet. Thus the head for the shorter section is more than 5 times that for the longer section, and it becomes obvious 352 DOMESTIC SANITARY ENGINEERING AND PLUMBING that, in order for the longer section to deliver a volume equal to the discharging capacity of the shorter section, the former must be of a larger size. Should a pipe from A to C, Fig. 222, be of uniform bore, the length BC would be never fully gorged. A straight line from A to C, as shown by the dotted line, would not be a true hydraulic grade line, because it falls below the pipe which rises to B, and under such circumstances would have a negative value. Assuming that a pipe of uniform bore were used for the conditions given in Fig. 222, the maximum head avail- able for forcing water through the pipe would be 35 feet, which is the vertical distance between point B and the level of the water in the tank. Example 32. Assuming a line of pipes follows the general contour of the ground surface as in Fig. 222, determine the sizes of the pipes to deliver 120 gallons of water at C, when the head and lengths of pipes are as shown. Total head for section AB = 250 -215 = 35 feet. Length of pipe given = 8200 feet. Pipe for Section AB. By Formula 63, d* = 91*1 It will be necessary to use. a trial value for /, and if we assume the pipe required will be between 5 and 7 inches diameter, the value of / from Table VIII. is 570. HYDROSTATICS AND HYDRAULICS 353 Substituting values given, d 5 = * X " ; From Table IX. we find that a 5J-inch diameter pipe when raised to the 5th power = 5032*8, but, as this value is not large enough, the diameter which agrees with the next high value is the one required ; .*. Pipe for section AB will be 6 inches diameter. Pipe for Section BC. Head upon C from point B = 215 -30 = 185 feet. Length of Section BC = 650 feet. By Formula 63, ^ = 5l^- Z . fxh As this pipe will be smaller than the one for section AB, we will assume that its diameter will lie somewhere between 2 and 3J inches. For this range of sizes the value of /, Table VIII., is 330. Substituting values given, d 5 = In Table IX. we find that the nearest diameter when raised to the 5th power to agree with the above value is 3 inches ; /. Pipe for section BC will be 3 inches diameter. Short Pipes. When pipes are short, the head to generate velocity at entry requires to be taken into account, and also that absorbed by branches and special fittings. For these conditions the head absorbed by friction may be obtained by the following formula : ' * * * (64) Where h = loss of head in feet. , d = diameter of pipe in inches. G = gallons discharged per minute. e a coefficient from Table X. 2 3 354 DOMESTIC SANITARY ENGINEERING AND PLUMBING TABLE X. For a plain pipe end in a tank . . . ,, trumpet shaped end of pipe in tank right- angled branch in pipe .. ,, plug tap when branched into pipe . ,, screw down tap when branched into pipe 00 OO VO VO rH C<1 -* 00 i-H i-H rH rH rH II II II II II Tdia. SN^^5w*sC^SR?SSS^S?^^ FIG. 223. Flow of water through pipes. If we assume that water is withdrawn from an overhead cistern, as in Fig. 223, where the draw-off pipe is comparatively short, the resistances offered by the tap and by the pipe end HYDROSTATICS AND HYDRAULICS 355 in the cistern may influence the discharge to a considerable extent. It is not possible to directly calculate the discharge by a short pipe, as the different resistances are not known at the outset. An indirect method is therefore adopted for calculating the discharging capacity of short pipes, which consists of first ascertaining the total head to yield an assumed discharge. When a certain head is known to give a required discharge, the discharge for any other head can be obtained by pro- portion, so long as the conditions with regard to size and length of pipes remain unaltered. The discharging capacity of pipes and fittings varies directly as the square root of the pressure head. If, therefore, it is found that a water pipe gives a discharge of 6 gallons per minute when under a head of 9 feet, the same pipe would discharge under a head of 1 foot x = 2 gallons per minute, and for a V9 head of 16 feet - ~ = = 8 gallons per minute. Example 33. Determine the volume of water that would be discharged in 10 minutes by a plug tap, as in Fig. 223, when the tap is subjected to a constant head of 12 feet. The dia- meter of the pipe is 1 inch, and its length will be assumed to be 28 feet. The principal resistances to be considered are those due to (a) Water entering the pipe. (b) Length of pipe. (c) Plug tap. As it is necessary to ascertain the head absorbed for an assumed discharge before the actual discharge can be cal- culated, we will take the assumed discharge at 10 gallons per minute. Head absorbed by friction. (a) Head absorbed by water when entering pipe. By Formula 64, h = Value of e from Table X.^1'58. 356 DOMESTIC SANITARY ENGINEERING AND PLUMBING , , ... .. 7 10 2 xl-5S Substituting values given, h = ^ - ; -L X \) I /. h = -59 feet. (b) Head absorbed by length of pipe. By Formula 61, ^=S Value of /for a 1-inch pipe in Table VIII. = 210. 10 2 x28 Substituting values, h = .-. h = 13-33 feet. (c) Head absorbed by plug tap. G 2 xe Formula 64 gives h^j-^. Value of e from Table X. = 14. Substituting values given, h = ^ -^= ; -L X ^0 / .\h = '52 feet. Total head to discharge 10 gallons per minute = '59 -fl3'33 + 52 = 14-44 feet. In the example only 12 feet are available as the pressure head, so the discharge for this head 10 x \/F2 = -r= = 91 gallons per minute ; /. the volume discharged in 10 minutes = 9-1x10 = 91 gallons. Suppose, now, we work Example 33, by omitting the resistance due to the water entering the pipe, and also that offered by the plug-cock. By Formula 60, G = Substituting values, G = /v /l 5 x21Qxl2 ; /. G = 9'48 gallons per minute. In ten minutes the discharge = 9'48 x 10 = 94'8 gallons. HYDROSTATICS AND HYDRAULICS 357 The simple and the more abstruse method of solving Example 33 only gives a difference of 94'8 91 = 3'8 gallons, which is not very much. The difference, however, is more marked when the ratio of the diameter to the length of the pipe is less, as the following example will show : Example 34. Find the discharge per hour by a 1^-inch lead pipe when arranged in a similar manner to that given in Fig. 223, when the head of water upon the plug cock is 7 feet, and when the length of the pipe is 10 feet. Assume a discharge of 20 gallons per minute. (a) Head absorbed when entering pipe. By Formula 64, h = Substituting values, ^ :.h = 467 feet. (b) Head absorbed by length of pipe. G 2 x/ Formula 61 gives " = ^ ^5. 202 x 10 Substituting values, h = /. h = 2-025 feet. (c) Head absorbed by IJ-inch plug cock. 20 2 xl-4 Substituting values, h = .'. h = 414 feet. Total head absorbed in discharging 20 gallons per minute = (-467 + 2-025 + -414) = 2-906 feet. As the actual head available is 7 feet, the discharge for this would be 20 x /7~ .--_^ =31*04 gallons per minute ; /. Gallons discharged per hour = 31'04x 60 = 18624 gallons. 358 DOMESTIC SANITARY ENGINEERING AND PLUMBING If now we try to solve the problem by omitting the resist- ances (a] and (c), and ascertain the discharge directly by Formula 60, /d b xfxh we have G = y' - 4 . xJk*' i n /(li) 5 X 260x7 bubstitutmg values, G = ^/ - y~ - .-. G = 37*17 gallons per minute. And discharge per hour = 3717 X 60 = 2230*2 gallons. For this problem the two methods give a difference of 2230-2 -1862-4 = 367*8 gallons per hour. The latter example clearly shows that when the discharging capacity of short pipes is required the total resistances should be taken into account, and this is especially necessary when the pipes are of large diameter. For pipes, however, of 1J inches diameter, and for smaller sizes, and where their lengths exceed 400 diameters, the smaller resistances may be omitted. Formula 60 under such conditions will give their discharge with sufficient accuracy. Suppose the diameter of a service pipe is required for filling a cistern in a given time, when the water in the main is under a known pressure as in Fig. 224. The type of ball- cock will affect the rate of discharge to a certain extent, but if a full- way cock is used, and the service pipe is of moderate length, the retardation offered by the tap when fully open may often be neglected. A ball-cock, of course, begins to close before the normal water-line is reached, unless special provision is made to prevent it. The pressure head above the point of delivery may be found by first converting the Ibs. pressure per sq. inch into equivalent head, and afterwards deducting the vertical distance between the main (where the pressure is taken) and the point of discharge. Example 35. If a house is supplied on the intermittent system, find the size of pipe required to deliver 15 gallons per minute, when a pressure of 40 Ib. per sq. inch is recorded where the service pipe joins the main. The length of the pipe is 150 feet, and the vertical distance between the main and point of discharge 45 feet. (See Fig. 224.) HYDROSTATICS AND HYDRAULICS 359 To solve this problem it will be necessary to assume that the pressure is constant in the main. For the conditions given the pressure head above the ball tap will equal (40 x 2'31) -45 = 47-4 feet. G 2 x/ ^/jr*-' "r~ ifti d S^ li "H**t 1 S j X II Assume the size required ies between 1 inch and 1 J inch iameter in order to obtain value for/, which according \ 1 o Table VIII. will be 260. i Substituting values given, | /5 15 2 xl50. ! i 260x47-4' i i /. f? 5 = 2'73. i On reference to Table IX. i b will be found that the I learest size when raised to "^ ts 5th power, which satisfies $ 73, is H in. diameter ; ? /. 1J inch is the diameter equired. The theoretical size lies etween 1 inch and 1 inch - iameter, but of course the i / ____!_ ^T y^^l^^^^^^^S^^3^^^^^^^^^/ &&^$*^" ^\ / 5 it- ^ ^-S FIG. 224. Flow of water through service pipes. higher commercial size would be adopted, and this would com- pensate to a great extent for the retarding influence of the tap. 360 DOMESTIC SANITARY ENGINEERING AND PLUMBING Another form of problem may now be attempted, where the pipes are arranged as in Fig. 225. Example 36. Determine the sizes of the pipes to deliver simultaneously 6 gallons per minute at C and 5 gallons per minute at D. The vertical distance between point C and the A. *5!555^vvmmvmw^ Length of Pi|ae to / Brunch J, 16 feeiv Ungfh of Branch, 4 feet \ I \ I \ --{-- \ \ \ \ \ ' r M ' 1 Length of branch t\ from J, 65 feet". No 2. FIG. 225. Flow of water from an overhead cistern. average level of the water in the supply tank is 11 feet, and that between point 13 and the source of supply 43 feet. Branch No. 1 is 24 feet long, No. 2 branch 65 feet long, and the length of the draw-off pipe to J is 18 feet. The first thing to consider is, that during the period of draw-off adequate pressure is maintained to properly supply both taps. To attain this end it is essential that the hydraulic HYDROSTATICS AND HYDRAULICS 361 grade line for the main draw-off pipe shall not fall below the bend at B. Suppose the maximum fall is fixed at 3 feet, then the line AB will represent the hydraulic grade line for the main draw-off pipe. The hydraulic grade line for branch No. 1 is shown by the line from B to C, and that for branch No. 2 by the line from B to D. Having decided upon the hydraulic gradients, the sizes of the pipes can now be determined. Size of Main Draw-off Pipe. The head available for the main draw-off is the vertical distance between A and B, Fig. 225, and this pipe will require to discharge 6 + 5 = 11 gallons per minute. The length of pipe to branch J is given as 18 feet. By Formula 63, d 5 = II 2 x 18 Substituting values given, d 5 x o /. d = 2-792. From Table IX. we find that 2792 lies between the 5th power of a 1-inch pipe and that of a IJ-inch pipe. .\ size of main required = 1^ diameter. The value of /was assumed, and it is found to satisfy the answer obtained. Had it been too large or small this part of the problem would have required reworking. Size of Branch No. 1. The head available for this, above the point of discharge, is the vertical distance between B and C. Then #- For trial value of / assume the pipe required does not exceed 1 incli diameter, when by Table VIII./=210. Substituting values, d 5 = -- ^ ; ZLO X o ..#=514. Upon reference to Table IX. '514 lies between the 5th power of a f -inch pipe and that of a 1-inch pipe ; .-. size of No. 1 branch = 1 inch diameter. 362 DOMESTIC SANITARY ENGINEERING AND PLUMBING Size of Branch No. 2. The head above the draw-off tap on this branch is the vertical distance between B and D, Fig. 225. Assuming the diameter is less than 1 inch, then by Table VIII. /= 210. a u 1 4.- i M 5 2 x65 Substituting values, d 5 = In Table IX. it will be found that "193 lies between the 5th power of a |-inch pipe and that of a f inch pipe ; /. size of branch No. 2 = j inch diameter. Collecting the sizes, we have Main supply pipe to J . . 1 J inch diameter No. 1 branch . . . 1 No. 2 . . . . f In each case the pipe is a trifle larger than necessary, so that each pipe will deliver rather more water than asked for in the question. THICKNESS AND STRENGTH OF PIPES The thickness of a cast-iron pipe for withstanding a given pressure is usually determined by some empirical formula, which provides a margin of strength for slight variations of thickness and other inequalities. So far as a lead pipe is con- cerned, its strength may be considered without much error to be directly proportional to its thickness, and inversely pro- portional to its internal diameter. The thickness of wrought iron and copper pipes is governed more by the form the joints take than by the internal pressure they are required to with- stand. In the case of lead pipes a minimum thickness is necessary to resist crushing by external forces and to allow for the making of bends. HYDROSTATICS AND HYDRAULICS 363 Formulae for Lead Pipes 2x*xS d Pxrf fe ~2^7 ' t = dxY pxdxF 2xS (65) . (66) (67) (68) Where P = bursting pressure in Ibs. per sq. inch. p = safe working pressure in Ibs. per sq. inch. S = ultimate strength of metal per sq. inch. F = factor of safety. d = diameter of pipe in inches. The factor of safety varies from 5 to 10, and these values indicate that the maximum safe working pressures are from 5 to 10 times less than those which produce fracture. TABLE XI. AVERAGE TENSILE STRENGTH OF METALS PER SQUARE INCH Copper 31,000 Ib. Wrought iron 50,000 Ib. Cast iron . 18,000 ,, \ Lead . 2,600 Cast steel . 62,000 Tin . 4,500 The formulae for lead pipes may be checked by the follow- ing tests, which were carried out for the writer : TESTS ON LEAD PIPES Test No. Bore of pipe. External diameter. Weight per lineal yard. Thickness : Bursting walls of pressure per pipe. sq. inch. 1 ^ inch. 92 inch. 7 Ib. 21 inch. 1960 Ib. 2 i 1-23 11 ,, 24 1500 3 i 1-54 16 27 1340 364 DOMESTIC SANITARY ENGINEERING AND PLUMBING Using the results of these tests for obtaining the ultimate strength of lead, we have by Formula 66, Substituting values from Test No. 1, 2x-21 : = 2333 Ib. per sq. inch. Substituting values from Test No. 2, q _ ISQOxf = 2x-24 ' /. S-2343 Ib. per sq. inch. Substituting values from Test No. 3, s _ 1340x1. = 2x-27 ' /. S = 2481 Ib. per sq. inch. For each test the ultimate strength of lead has a different value, but this is chiefly due to the fact that lead pipes are seldom perfectly true in section, especially when the pipes are of small bore. The strength of a pipe of course is only equal to that of its weakest side. In Table XI. the strength of lead is given at 2600 Ib. per sq. inch. Test No. 1 gives within 11 per cent, of that value, and Test No. 3 within 5 per cent. From these results we may infer that a higher factor of safety is necessary for small bore pipes than for those of larger diameter. Example 37. Find the maximum safe working pressure for a ^-inch lead pipe which weighs 7 Ib. per yard, when a factor of safety 8 is adopted. By Formula 67, p = The value of t in the table of tests = '21, whilst s in Table XI. = 2600. HYDROSTATICS AND HYDRAULICS 365 Including these values and the others given, 2 x -21x2600 p= T^r /. p = 27'3 Ib. per sq. inch. Example 38. Determine the thickness for a Ij-inch diameter lead pipe which is to he subjected to a maximum pressure of 45 Ib. per sq. inch, when a factor of safety 6 is adopted. Using Formula 68, t= pxdxY . 2xs o , . ., , . , . 45 x H x 6 Substituting values, t= - ; .-. = -078 inch thick. Although a pipe of the latter thickness would withstand the internal pressure, it would be far too thin for a water pipe, its weight per lineal yard being a little less than 5| Ib. From Example 38 we see that Formula 68 is not suitable for ascertaining the thickness of pipes for withstanding either medium or low pressures. For example, a IJ-inch lead waste pipe which weighs 12 Ib. per yard has a thickness of '156 inch, and with a factor of safety 6 would be capable of withstanding a pressure of 90 Ib. per sq. inch. Formula for Cast-iron Pipes The thickness of cast-iron pipes up to 9 inches diameter can be determined by the formula below. * = [{(-00014xd)xGt?+50)} + -27] . . . (69) Where t = thickness of metal in inches. p = water pressure in Ibs. per sq. inch. d = diameter of pipe in inches. Evximple 39. Find the thickness of a 6-inch diameter cast- iron pipe which is subjected to a pressure of 150 Ib. per sq. inch. By Formula 69, * = [{(-00014xd)x(p + 50)} + -27]. Substituting values, * = [{(-00014 x 6)x(150 + 50)} +-27], = [(00084x200) + '27]; ,-. t -438, or say ^g- inch thick. CHAPTER XIII DOMESTIC HOT WATER SUPPLY Movement of Heat. Heat moves in three ways : (a) by conduction ; (b) by convection ; (c) by radiation. To heat water in an apparatus the first and second forms of heat motion come into operation. When the surfaces of a boiler are exposed to the action of heat, a certain amount of the latter is transmitted through the plates. In other words, heat is conducted through the metal walls from the fire to the water side of a boiler, and, in turn, heat is absorbed by the water in contact with the heated surfaces. The transference of heat from receptacle to receptacle which contains water is accomplished by convection, when suitable passages are provided through which the water can circulate. Circulation of Water. Movement, or circulation, of the innumerable particles of which water consists takes place as soon as they differ in weight, the more heated and lighter particles being displaced by those of greater density. In order for water to freely circulate between two receptacles, such as a boiler and a tank, two separate paths are essential. One path is provided through which the heated particles escape after being heated, and the other conveys the cooled particles to the source of heat. The Tank System. Fig. 226, although not often installed, possesses one or two favourable points. In the first place the position of the hot-water tank admits of a cheap form of tank being used, and this, accordingly, reduces the cost of a com- pleted system. Instead of a cylindrical tank, either a square or rectangular form may be adopted. Another advantage possessed by the tank system, is that a free outflow of water can be obtained at the highest draw-off taps, on account of the 366 DOMESTIC HOT WATER SUPPLY 367 water being withdrawn directly from the overhead tank, and owing to the comparatively short lengths of pipe through which it has to flow. The tank system, however, has drawbacks. The tank can be emptied by any of the draw-off taps should the supply to FIG. 226. the feed cistern fail, and when water is withdrawn at a tap a mixture of hot and colder water is frequently obtained. The latter is a bad feature in connection with the tank system, for the resultant temperature at the point of escape may be considerably less than that of the water in the tank. Of course, when the draw-off first begins the water which oc- 368 DOMESTIC SANITARY ENGINEERING AND PLUMBING cupies the upper part of the boiler will be at a higher tern* perature than that in the upper part of the tank, and if only a small volume is withdrawn the hottest water may issue at the point of escape. On the other hand, when larger volumes are required the heated water in the boiler is soon replaced with colder water, and this flows to the point of draw-off" and mixes with that from the overhead tank. With regard to the volume of the water which will travel by each path, this will depend entirely upon the resistance offered, the greater volume naturally flowing by the route which offers the least obstruction. As a rule the taps at the higher levels will discharge a greater percentage of hot water directly from the tank than taps at lower points. In Fig. 226 the flow and return circulating pipes are shown to take a very favourable course between the boiler and the tank, but in practice the course is usually more circuitous than shown. The length of the circulation pipes retards the move- ment of the water, for as a rule these pipes are only of small bore. Details of Tank Systems. It will be observed in Fig. 226 that the draw-off branches are only taken from the flow-pipe, and if the best results which this system is capable of yielding are to be obtained, the flow-pipe will require to terminate at a fairly high point in the tank. In some cases a separate draw- off pipe is adopted which is joined half-way up the tank. With regard to the cold supply to tank T, it will improve matters, where the relative positions of the pipes are as shown, if the water is deflected sideways when it enters the tank by means of an elbow or tee. This arrangement prevents the cold water upon entering the tank from taking a direct course to the return pipe when hot water is being withdrawn. The connections to the boiler are shown at the top, the return being continued by means of a tube to near the bottom of the boiler. It is not practicable in every case to make the boiler connections in the position shown, and it is often necessary to either join the pipes at the side or at the back. A high and low connection to a boiler is not essential to make the water circulate, but to define the course the water should take, DOMESTIC HOT WATER SUPPLY 369 Although high and low connections with boilers are generally made, it is not uncommon to find that reversed circulations occur. As a general rule, it will be found when an intended flow-pipe acts as a return, and vice versa, that some form of retardation has been introduced in connection with the flow-pipe. The flow-pipe connection should provide as free a passage as practicable for the escape of the heated water, and when the flow-pipe is either joined at the side or at the back of a boiler, the length of the horizontal pipe in the immediate neighbourhood of the boiler should be reduced to a minimum. If a portion of a flow-pipe must be horizontally arranged, this should be introduced if possible above, and not on a level with, a boiler. Where a tank system is adopted, and it is deemed necessary to shorten the lengths of the dead branches in order that hot water may be readily obtained after the opening of a tap, the piping may be arranged as in Fig. 227. The branch circuit should be kept as short as possible, and the pipes should have a gradual rise towards the tank. To enable water to circulate through the branch in the direction indicated by the darts, a stop tap s should be introduced immediately above the branch at B. It is only essential for a small portion of the water to circulate through the secondary circuit, and this can be adjusted by means of the stop tap S. Cylinder System. For heating water for domestic purposes the cylinder system is the more generally adopted. It chiefly differs from the tank system in the following respects : (a) A cylindrical tank takes the place of the square or rectangular one, as a circular shape is better suited for resisting internal pressure. (b) Shorter circulation pipes are employed on account of the cylinder being located nearer to the boiler. (c) The cylinder is arranged to remain full or partially charged when the cold supply fails. In Fig. 228 a cylinder system is shown which is suitable for a small dwelling, and all the connections are arranged to give good results. The air or expansion pipe is shown to terminate above the roof, but when in the immediate neigh- bourhood of a supply tank it may be turned over so as to 24 370 DOMESTIC SANITARY ENGINEERING AND PLUMBING discharge into it as shown by dotted lines. There is not often any objection to the latter method, excepting where the temperature of the cold water is liable to be appreciably raised, and when cisterns are fixed in positions where the free escape of steam would be objectionable. All draw-off branches in a cylinder system are joined to the air-pipe, (excepting special cases) and it is desirable that TE&HE. &KHBBML. ( WITH SECONDARY CIRCUIT ^ FIG. 227. their connections be made well below the bottom of the cold supply cistern, to prevent air entering when draw- off taps are opened. To enable a system to be emptied of water, a pipe is frequently connected with the boiler as in Fig. 228, the free end terminating at any suitable point. A stop tap C should be provided on the cold supply pipe so that the water can be turned on and off as required. On account of the diversity of views which prevail with DOMESTIC HOT WATER SUPPLY 371 regard to the positions the connections with a cylinder system should be made, it may be well at this point to discuss some of the methods of arranging the pipes, and to point out any merit or demerit they possess. In Fig. 229 the cold supply is shown joined directly with the boiler, this method being largely adopted in certain districts. In this case^the principal thing to consider is, the final temperature of the water when discharged at any point when compared with that in the cylindrical tank. When a tap is opened, water will flow to the point of escape from any available source, and when the pipes are arranged as in Fig. 229, the cold water upon entering the boiler can flow to the cylinder through either the flow or return pipe. If only 372 DOMESTIC SANITARY ENGINEERING AND PLUMBING a comparatively small volume of water is withdrawn at a time, the cold supply connection as arranged will answer fairly well; when, however, larger volumes are withdrawn, the hot water in the boiler is soon displaced, and cold water passes through the flow pipe and mixes with the heated water at the top of the cylinder. Should the flow connection be made immediately above the return, the water in the upper part of the cylinder may not be unduly cooled during the period of withdrawal. It is, however, generally desirable for a Wash-out Cock FIG. 229. Cylinder system Avhere cold water supply joins directly with boiler. flow-pipe to be connected at a high point of the cylinder, as this allows a limited volume of very hot water to be quickly obtained after a fire is first lighted. If, on the other hand, the flow connection is made at a low point, then the heat trans- mitted from the boiler in a given time is diffused throughout a greater mass of water, and consequently it is raised through a smaller range of temperature. No fault can be found with the connections in Fig. 229 so far as the heating of the water is concerned, but the fault occurs when a moderate volume of water is withdrawn. DOMESTIC HOT WATER SUPPLY 373 It is frequently contended that cold water should not directly enter a boiler as in Fig. 229, on account of the latter being subjected to great variations of temperature, and, in consequence, a greater amount of strain. This point, however, so far as range boilers are concerned, is not of much importance, for in practice the life of a boiler does not appear to be affected by any particular arrangement of the cold supply. The principal advantage offered by joining the cold supply to the cylinder, is that unnecessary mixing of hot and cold water can be prevented when water is withdrawn. When the cold supply is joined, as in Fig. 228, the entering water is well diffused over the lower portion of the cylinder, and the upper and hotter water is more nearly uniformly displaced. In Fig. 230 the flow connection to the cylinder is shown joined at a low point, and a by-pass is pro- vided between the flow and the air-pipe. For a simple system, a by-pass offers no special advantage, its use being only intended to con- vey a portion of the hottest water directly to the top of the cylinder. As a rule this can be better accomplished by joining the flow-pipe at a higher point. Secondary Circuits. Although for the smallest installa- tions secondary circuits are not required, they are desirable in large buildings to avoid long dead branches, and to permit of hot water being withdrawn immediately after opening a tap. A secondary circuit is shown in Fig. 231, and the lower end is joined about 6 inches below the top of the cylinder. In order for water to circulate through the secondary circuit, the latter must be arranged that air cannot accumulate. It is not essential that a circuit should fall from the main air or StobCbck. * Simplifying 40# = (150 - a?) x 76. 40a;= 11400 -763. 116a=11400. From which, x ...... =98 -/$ gallons. Therefore the volume of hot water at 160 F., which is necessary to produce the volume required, may be taken as 98 gallons. Assuming a few hours are required to heat this volume of water, the capacity of the cylinder may be increased by 40 per cent, when the final capacity = 98 + 39 = 137 gallons. 400 DOMESTIC SANITARY ENGINEERING AND PLUMBING Sizes of Cylindrical Tanks. From the following simple rules either the capacity in gallons or one dimension can be obtained when the remaining particulars are given : Where G = capacity in gallons. k = height or length of tank in inches. d = diameter in inches. Example 47. Find the volume of water a cylindrical tank will hold when its diameter is 2 ft. 9 in. and its height 4 ft. 6 in. By Formula 78, G= , ... ,. n 33x33x54 Substituting values, G- = .-. G = 166, say 166^ gallons. Example 48. A cylindrical tank of 80 gallons capacity has a diameter of 20 inches, determine its height. TT 17 1 >7Q 7 353 XG Using Formula 78, h = -^ . 353 x 80 Substituting values, h= 9Q 2Q ; .-. h = 7 Of inches, say 5 ft. 10^ in. Example 49. What diameter of cylinder would be required to hold 65 gallons when its height is 4 feet ? By Formula 80, ^ = v h Substituting values, ^= /v / 353x65 ; 48 = 21-8 in., say 1 ft. 9} in. DOMESTIC HOT WATER SUPPLY 401 Square and Rectangular Tanks. The capacity of these tanks or any dimension can be obtained when the remaining values are given. . ...... (81) t _277xG Ixh ' '- <> Where G = con tents in gallons. 1 = length of tank in inches. b = breadth in inches. h = height or depth in inches. Sizes of Pipes for Systems of Hot- Water Supply. The sizes of circulating pipes between a boiler and a cylinder are usually determined by arbitrary rules; the chief thing is to provide a passage which will not unduly retard circulation. Where draw-off branches are taken from secondary circuits, the latter should be sized to properly supply the number of taps which are likely to be in use at the same time. For a small system which is supplied with soft water f-inch circulating pipes are usually satisfactory. If these pipes are long they should be increased in size. Larger circulation pipes are also desirable when they require to be trapped, in order to compensate for a slower rate of movement of the water. In hard water districts where deposits occur in pipes and boilers, larger circulating pipes should be used than in soft water areas. An increase of one size is usually sufficient where the conditions are similar. The return pipe, as a rule, is not materially affected by deposit, the salts which escape from the boiler being usually precipitated in the flow pipe. For large systems where independent boilers are used, the minimum size of circulating pipes should be 1^ inches diameter. Generally speaking, feed pipes to cylinders should not be 26 402 DOMESTIC SANITARY ENGINEERING AND PLUMBING less than one size larger than the principal draw-off pipes ; a large feed pipe is specially necessary where there is only a short vertical distance between the top of the hot water and cold supply tanks. For comparatively small systems the principal draw-oft pipe from a cylinder does not require to exceed 1 inch diameter, but in the case of large installations where pipes are long, their sizes are better ascertained by the aid of hydraulic formula, when the special requirements of each system can be taken into account. Steam Apparatus for Heating Water. Steam, when avail- able, as in many large buildings, is a very suitable and convenient agent for heating water, and it possesses the special advantage of being able to raise large volumes to a relatively high temperature in a limited time. Properties of Steam. To convert water at 2 1 2 F. into steam at the same temperature, requires 966 B.T.U. per pound of water. The British thermal unit is generally expressed by the abbreviation B.T.U. , and represents the amount of heat necessary to raise one pound of water from 39 to 40 F. or, say, through one degree. The heat which is necessary to convert water at any temperature into steam at the same temperature, is known as the latent heat of steam, and this value varies according to the pressure of the steam. For steam at atmospheric pressure (equivalent temperature 212 F. ), its latent heat is 966, and for pressures less than that of the atmosphere, the value of the latent heat increases. On the other hand, when the pressure of steam exceeds that of the atmosphere, its latent heat is less than 966 ; in other words, less heat is necessary to change water from the liquid into the gaseous state. As the heat which is stored in steam is given up upon its condensation, its value as a heating agent is apparent. For example, 1 Ib. of steam at 15 Ib. per sq. inch, (gauge pressure) when condensed and when the water of condensation is cooled to 150 F., gives up 939 + (250 -150) = 1039 B.T.U. The value 939 is obtained from Table XIII. and is the latent heat of steam for the pressure given. From the same Table, the value 250 will be found to DOMESTIC HOT WATER SUPPLY 403 represent the temperature of the steam when subjected to a pressure of 15 Ib. per sq. inch. Boiling Point. The boiling point of water is a variable quantity. At sea level water boils in an open vessel at 212 F., but if it is subjected to greater pressure by confining it in a closed vessel, the boiling point is raised. At high altitudes, water in an open vessel boils at less than 212, and the same result is obtained when water is confined in a vessel, and when the air pressure in the vessel is reduced by an air pump or by other means. TABLE XIII. PROPERTIES OF STEAM Pressure in Ib. per sq. inch. Tempera- ture F. Latent heat. ITS:;'*'* Latent heat. Atmospheric 212 966 16 252 938 1 216 963 17 254 936 2 219 961 18 256 935 3 222 959 19 257 934 4 225 957 20 259 933 5 228 955 21 261 931 6 231 953 22 262 930 7 233 951 23 264 929 8 235 949 24 266 928 9 238 948 25 267 927 10 240 946 26 269 926 11 242 945 27 270 925 12 244 943 28 272 924 13 246 942 29 273 923 14 248 940 30 274 922 15 250 939 31 276 921 Heating Water by Steam. There are two general methods of heating water by means of steam (a) the direct method, where live steam is introduced into the water ; and (b) the indirect method, where the steam and water are kept apart by arranging steam-heated surfaces which are surrounded with water. The direct method is limited in its application, on account of the noise usually caused when steam and water come together, and the increase in volume of the water due to 404 DOMESTIC SANITARY ENGINEERING AND PLUMBING condensation of steam. This mode of heating water, however, is often suitable for industrial purposes, and for heating water in swimming ponds as well as for other special uses. For general work indirect steam heating is necessary, but less heat is abstracted from a given weight of steam than with the direct method, One of the simplest forms of indirect steam heaters is shown in Fig. 247. A copper coil in the cylinder takes the place of a boiler, steam being admitted at the upper part, whilst the water of condensation drains to a steam trap T, which is located in any convenient place. The feed and draw- FIG. 247. Steam heater. (Union connections omitted. ) off pipes in connection with the cylindrical tank are arranged as in systems where boilers are used. If the water of condensation could be arranged to gravitate to the boiler, a steam trap may not be necessary, but where steam is taken from a high-pressure boiler, and its pressure is reduced before being admitted to a heater, a steam trap is imperative unless some other contrivance is introduced. The tubes of steam heaters or calorifiers, as they are frequently termed, take different forms and they are arranged in a variety of ways. So far as the amount of heat which is abstracted from a given weight of steam is concerned, one form DOMESTIC HOT WATER SUPPLY 405 of heater is as effective as another. It is, however, customary when comparing steam heaters to speak of one form as being more efficient than another form, and the term should not be taken to indicate that one form of heater uses less steam than another to do a specific amount of work, but only, that a par- ticular form of tube arrangement will condense more steam in unit time per unit area than another form. In other words, one form of heater will do work quicker than another, FIG. 248. Steam heater or calorifier. (Union connections omitted.) but the consumption of steam is practically the same with different types for the same work done. Generally speaking, straight plain vertical tubes transmit heat more slowly than other forms per unit area, and the reason for this is, that instead of the water of condensation spreading evenly over the whole surface of the tube, it trickles down them in streams, and leaves a large area not wetted. Coils are therefore largely adopted for heaters, as the retardation introduced by changes of direction permits of a greater tube 406 DOMESTIC SANITARY ENGINEERING AND PLUMBING surface being wetted, and the rate of heat transmission is consequently increased. Instead of placing a coil inside a tank, as in Fig. 247, the heater is often fixed as an independent unit, as in Fig. 248. This is preferable, for when a heater requires to be repaired it can be readily detached from the tank. In Fig. 247 the tank is shown with a removable top, in order to render the coil accessible. The heater in Fig. 248 consists of a number of Washout FIG. 249. James Pyle & Co.'s steam heater with automatic device for controlling steam supply. tubes which communicate with steam spaces at the ends. Flow and return pipes are arranged in the ordinary manner. A safety valve may be necessary at the top of the heater, but this principally depends upon the strength of the casing and the pressure of the steam. When coils are used for heaters, their length, as a rule, should not greatly exceed 150 diameters, otherwise their lower parts will be useless for transmitting heat. Thus for a coil of 1-inch copper tube its maximum length should be about DOMESTIC HOT WATER SUPPLY 407 12 feet. Where a greater length of tube is necessary to transmit the requisite quantity of heat, two or more coils may be formed. To obviate overheating, and to prevent waste of steam, it is essential that a heater be provided with some device for automatically controlling the steam supply. There are various means of accomplishing this end, one method being shown in Fig. 249, where the steam valve is operated by a steam trap T. To the inlet A of the trap a spindle is joined, the other end of the spindle being joined with the steam valve B. The trap is of the expanding and contracting type, and is more clearly shown in Fig. 251. 5teom Outlet Inlet for water of Condensation , 5team Inlet FIG. 250. Section of James Pyle & Co.'s automatic steam valve. The automatic valve in Fig. 249 is brought into action as follows : So long as there remains a certain difference in temperature between the steam and the water to be heated, the steam is condensed, and its discharge effected by the opening of the valve of the trap. When the temperature of the water in the heater is raised, the valve of the steam trap is gradually closed, and in turn operates to close the steam valve B. In this manner the supply of steam is controlled to suit the rate of condensation, and when the water has reached the temperature for which . the appliance has been adjusted, the supply of steam is cut off. Figs. 250 and 251 give sections of the automatic steam valve and trap, which are shown in position in Fig. 249. 408 DOMESTIC SANITARY ENGINEERING AND PLUMBING In the trap Fig. 251 a copper tube is arranged in the form of a bow, and when it is subjected to increased temperature the rate of expansion tends to straighten the tube and to close the inlet valve. Upon the tube cooling, contraction takes place, and the bow regains its normal position when the valve is again opened. In order to make this form of trap sensitive to changes of temperature, the copper tube is some- times charged with a volatile liquid. Steam Traps take many forms, and their use becomes necessary in steam heating work, where the water of con- densation cannot be returned directly to the boiler. The primary purpose of a steam trap is to deal with the water Inlch Adjusting 5crtfur. J5 Ootleh FIG. 251. Section of steam trap which is used in conjunction with valve shown in Fig. 250. of condensation, and to avoid unnecessary waste of steam. As a rule the most effective form of steam trap is the box type, and when circumstances will permit of its use it should be adopted. The initial cost of the box type may be higher than other forms, but it is by far the cheapest in the end. Fig. 252 gives a very good form of box trap, by Lancaster and Tonge, where the opening and closing of valve S is accom- plished by a quick screw motion when the float E falls and rises. To the top of the float an adjustable air-valve N is attached, to which is joined a tube, which terminates near the bottom of the float. The small orifice in the upper part of the tube is to admit of the escape of air, which would otherwise be confined in the float and interfere with the working of the DOMESTIC HOT WATER SUPPLY 409 trap. To start the trap, water is poured into it after the cover is removed, until the overflow or outlet is reached. As water enters the float through the aperture F, it begins to sink and to open the valve, when the water of condensation can be discharged. So long as water only flows into the trap, the ball Fie. 252. Lancaster and Tonge's steam trap. will remain submerged, but when steani reaches and enters the float, the water is displaced at N and through the small aperture F in the float, until the latter is rendered buoyant and the valve S is closed. After a short time the steam in the float is condensed, and water again enters until its buoyancy is destroyed. If, in the interval, water has accumulated at 410 DOMESTIC SANITARY ENGINEERING AND PLUMBING the valve, it is immediately discharged, but when steam appears the float is again raised and the valve S closed. Heat transmitted by Steam Heated Coils when surrounded with Water. The amount of heat transmitted through tubes is governed by the temperature or pressure of the steam, by the initial temperature of the water to be heated, by the form the tube takes, and whether the whole of the heating surface is effective or not. As regards the pressure of steam, the higher it is the greater is the difference between the temperature of the heating medium and that of the water to be heated, and consequently the quicker the rate of heat transmission. When water is being heated in a calorifier with steam, the rise of temperature will not be at a uniform rate, but quickest at the commencement when the water is cold. If, for example, we assume that a volume of water requires to be raised from 44 to 180 with steam at 15 Ib. pressure per sq. inch (equivalent temperature 250 F.), the difference in temperature between the steam and the cold water is 250 44 = 206, and the difference between the steam and the hottest water 250-180 = 70. As the maximum difference of temperature coincides with the maximum rate of heat transmission when other conditions are equal, it is obvious that as the temperature of the water is raised the rate of heat transmission will accordingly be reduced, and for the case given will be at a minimum when the temperature of the water is 180. For purposes of calculation it is usual to take the average difference of temperature in order to simplify matters, and although this may not give results which are strictly correct they are usually sufficiently accurate for ordinary work. Thus where the temperature of the steam is 250, and that of the cold and hottest water 44 and 180 F. respectively, the average- difference of temperature' for the steam and water will be 250-1^=138. 2 The following Table gives the approximate number of heat units which are transmitted per minute per square foot of sur- face per degree difference of average temperature for short coils. DOMESTIC HOT WATER SUPPLY 411 TABLE XIV. Heat transmitted in B.T.U. per Steam pressure per square inch. minute per sq. foot of surface per degree difference of temperature. 5 Ib. (temp. 228) 15 (temp. 250) 30 (temp. 274) 4-5 The above values are for an initial temperature of 44 and a maximum temperature of about 180. By the aid of the following formulae calculations in connection with steam heaters may be made : /-, . . . (85) G = vP - Vsk^r ~- - < 86 > 38xGx(T-0 m Where G = gallons of water heated. T = temperature of hottest water. t = temperature of cold water. P = temperature of steam. U = B.T.U. transmitted per minute from Table XIV. d = external diameter of tube in inches. / = length of tube in feet. m = time in minutes heating water. Example 50. Determine the length of coil when formed of 1-inch copper tube which will raise 130 gallons of water in 20 minutes from 42 to 180* F. Assume the steam is supplied at 15 Ib. per sq. inch. 38xGx(T-0 By Formula 80, 1= , T\TT\" 412 DOMESTIC SANITARY ENGINEERING AND PLUMBING In Table XIV. the values of P and U are given as 250 and 6 respectively. Q , ... f . , j 38 x 130 x (180-42) Substituting values given, I = 7 _38x 130x138 : 139x6x20 ' /. Z = 40HJ- ft., say 40 ft. 10 in. Example 51. If a steam heater contains a coil of 1-inch copper tube which is 10 feet long, how many gallons of water would it raise per hour from 44 to 180 F. when supplied with steam at 30 Ib. gauge pressure ? Using Formula 86, G ' Values of P and U from Table XIV. are 274 and 8 respectively. Substituting values given, x 8x60x1x10 \ / a = \ 1 38 x (180 -44) 162x8x60x10 38x136 .-. G=150jf$, or say 150 gallons. Example 52. Ascertain how many minutes it will take a calorifier to raise 350 gallons from 42 to 178 F. when the total length of the 1-inch copper coils is 50 feet and when the steam pressure is 30 Ib. per sq. inch. 38xGx(T-/) P--~YxUxdx/ 2 / Substituting values given, 38 x 350 x (178 -42) DOMESTIC UOT WATER SUPPLY 413 38x350x136 fftl - .._ * 164x8x50 ' .-. m = 27|4, or say 27 1- minutes. As the arrangement of heating surface in calorifiers is a very important factor as regards the rate of heat transmission, it is necessary that experiments be carried out, in order to ascertain the actual value of U, for any special form of heater. Indirect Systems of Hot Water Heating. The systems to be dealt with in this case are confined to those where hot water is the indirect heating medium, and where range or independent boilers form the direct or primary heaters. Indirect heaters of this type are only suitable for waters which cause incrustation difficulties, as they are much more expensive to instal and much slower in action than direct heaters. It has been previously stated that the deposition of lime salts from temporary hard water chiefly takes place when water has its temperature raised to over 180 F. An indirect system is therefore designed that the water which is withdrawn from it will not readily have its temperature raised to 180 F. The amount of solid matter which is precipitated from water will depend upon the nature and amount of hardness the water contains, upon the temperature to which the water is raised, and upon the volume used. For example, supposing that 500 gallons of hot water are required in an establishment per day, that the water contains 15 degrees of hardness, and that when raised to 212 F. 10 degrees of hardness are eliminated. Of course it is not likely that the whole of this volume under usual conditions would be raised to anything like 212, and for our case we will assume that only 20 per cent, or 100 gallons reaches that temperature. Upon this basis the amount of solid matter precipitated in a period of three months (91 days) would be 100x10x91 = 91,000 grains, or 13 Ib. It is therefore obvious that, unless some measures are adopted to prevent the deposition of lime salts, a boiler may rapidly be destroyed. Fig. 253 gives an indirect system for heating water, where 414 DOMESTIC SANITARY ENGINEERING AND PLUMBING B represents an independent boiler, H the indirect heater, C the outer cylindrical tank from which water is withdrawn at the various taps. A small tank T supplies the boiler and Secondary Return. FIG. [253. Indirect heating[system. indirect heater with water. It will be observed that no water is withdrawn from the indirect heater, the same water being heated over and over again, and any loss is made good by DOMESTIC HOT WATER SUPPLY 415 means of the small supply tank T. From the top of H an air- pipe is taken and terminates as shown, the small tank serving the purpose of an expansion as well as a supply tank. It is only necessary for the bottom of tank T to be just above, or on the same level as, the top of the indirect heater, for the less the head the smaller will be the maximum temperature to which water in the indirect heater can be raised. In an open vessel water at sea level boils at 212, but when a boiler or indirect heater is subjected to pressure the boiling point of water is increased. For example, suppose the vertical distance between the top of the indirect heater and the level of the water in the supply tank T, Fig. 253, is 8 inches, the boiling point would be approximately 213. Should, however, the vertical distance between the supply tank and the boiler in an ordinary system be 35 feet, the boiling point at the lowest level would be raised to 250. The further application of heat when the boiling point is reached does not increase the temperature of the water, but the latter is converted into steam at the same temperature. It will thus be clear that, in a system like Fig. 253, the water in the indirect heater H can never get much hotter than 212 F. ; moreover, as a difference of temperature must exist between the water in the indirect heater and that in the cylinder C before heat can be transmitted from the former to the latter, the maximum temperature in C will be less than 212. As a rule the water in the outer cylinder will not greatly exceed 170 F. By arranging the indirect heater as in Fig. 253, the flow and return connections can be conveniently made at the bottom of the cylinder. Where an independent boiler is used, one with a water passage beneath the fire bars is preferable, although little solid matter can be deposited, owing to the water being seldom renewed. The cold supply pipe from an overhead cistern, and the secondary returns, are arranged as in other systems. The following table gives the temperature at which water boils when subjected to varying heads ; 416 DOMESTIC SANITARY ENGINEERING AND PLUMBING TABLE XV. Head of water. Boiling tempera- ture F. Head of water. Boiling tempera- ture F. Head of water. Boiling tempera- ture F. ft. 8 in. 213 23 ft. 9 in. 240 46 ft 10 in 259 3 216 26 1 242 49 2 261 5 3 219 28 5 244 51 6 262 7 7 222 30 9 246 53 10 264 9 11 225 33 248 56 1 266 12 3 228 35 4 250 58 5 267 14 6 230 37 8 252 60 9 269 16 10 233 39 11 254 80 10 281 19 2 235 42 3 256 103 11 292 21 6 238 44 7 257 128 2 303 It is important that an indirect heater contains sufficient surface to dissipate the heat as quickly as received. If we assume when heating the water in a system that the area of the indirect heater is equal to the boiler surface, the former would readily impart to the cold water surrounding it all the heat the boiler could transmit. When, however, the temperature of the water in the indirect heater and that surrounding it do not differ much, the area of the indirect heater may be too small, and unless the fire were checked the water may soon boil in the heater. As a rule the surface of an indirect heater should be from 3 to 4J times the area of the direct heating surface of the boiler, the smaller value being adequate for boilers with controlled draught, and the larger value for boilers without automatic control. To increase the surface of an indirect heater without FIG. 254. Indirect heater. DOMESTIC HOT WATER SUPPLY 417 increasing its capacity, cross tubes may be inserted as in Fig. 254. This would increase the initial cost of the indirect heater, but a saving would be effected on the size of the external cylindrical tank. Fig. 255 gives a section of the " Sylphon Automatic Temperature Regulator " for hot-water boilers, by the National Leuer FIG. 255. " Sylphon " automatic temperature regulator by National Radiator Co. Ltd. Radiator Company Ltd., and Fig. 256 shows the appliance in position. The regulator consists principally of a metal bellows B, and a lower sealed vessel R which is joined to the bellows by means of a central tube T. The bellows is fully charged, whilst the vessel beneath is partially charged with a fluid which is volatilised at a comparatively low temperature, and this supplies the energy for operating the appliance. When a certain tem- 27 418 DOMESTIC SANITARY ENGINEERING AND PLUMBING perature is reached the confined liquid is converted into the gaseous state, when the internal pressure expands the bellows B, and imparts motion to the weighted lever on the top of the appliance. To one end of the lever arm a chain may be joined as in Fig. 256, when upon the expansion of the bellows the draught damper is closed whilst the check damper in the flue is opened. Under these conditions the draught is checked, for the air supply to the fire is diminished, and, simultaneously with this, air is admitted into the flue by the opening of the check Check Darner. ^..J- Draught Damper FIG. 256. Automatic regulator attached to boiler. damper. When the vapour reassumes its liquid state, owing to a cooling action having taken place, the bellows contracts and the position of the dampers is reversed. In this manner the rate of combustion is automatically adjusted to suit the demand made upon the system. By shifting the weight on the lever the temperature at which the regulator can be brought into action is said to vary from 90 to 190. Collapse of Copper Cylinders. Hot-water cylinders collapse when the metal of which they are made is not sufficiently rigid to withstand a reduced internal pressure, and although cylinders DOMESTIC HOT WATER SUPPLY 419 collapse under different conditions, the real cause is the same in each particular case. A very common cause of cylinder collapse is due to the air or expansion pipe being trapped, so that the free escape of air from the system is prevented. Fig. 257 will aid in making c A FIG. 257. Illustration showing how a cylinder may be collapsed. this clear. Supposing an expansion or air-pipe is fixed immediately beneath a ceiling, as in the figure shown, it is quite possible for this pipe not to rise for the whole of its length. Should the pipe be of lead and be supported with hooks or clips, after a time it will sag between the points of support, or, when the pipe passes through a ceiling to a floor 420 DOMESTIC SANITARY ENGINEERING AND PLUMBING above, the vertical part may slip down a little, and produce a sag as at A, Fig. 257. In either case the air given out when heating the water would lodge in the air-pipe. Should water be withdrawn, the accumulated air would be relieved, and pass along with the water to the point of escape. If, how- ever, the water in the cylinder is heated, and none is with- drawn, the air accumulates in the horizontal pipe. Should overheating take place, steam also gathers at the highest point, when water is gradually forced from the cylinder back through the feed-pipe to make room for the steam. The reason why the steam and air cannot escape through the air -pipe is rendered clear when one considers that before air can be dislodged the water in front of it must be first displaced through the dip at A, Fig. 257. As the displacement of the water must raise the column in C, the pressure due to the latter would exceed that due to the head of water in the supply cistern. The least line of resistance under these conditions is offered by the supply pipe, and through that pipe, either part or nearly the whole contents of the cylinder may be dislodged, provided the generation of steam continues. So long as the steam pressure is maintained inside the tank nothing serious happens, but when the system begins to cool condensation of the steam takes place, when a partial vacuum is produced. If the cylinder at this period is unable to resist the atmospheric pressure, the sides give way and a collapse is the result. Another way in which a cylinder may collapse is when the feed and expansion pipes are blocked with ice, and an attempt is made to withdraw the water by opening the sludge or emptying cock. In a similar manner the upper cylinders in connection with tenement buildings, when treated as in Fig. 237, occasionally collapse. For example, if the water in the supply pipe to the cylinders is frozen near the cistern, and the upper part of the air-pipes are also blocked with ice, the open- ing of a draw-off tap at a low level would cause water to be removed from the upper cylinders, when the latter would collapse. The most common cause of cylinder collapse is probably due to expansion pipes getting locked with air. DOMESTIC HOT WATER SUPPLY 421 Prevention of Collapse. When the cause is known it is a simple matter to guard against a cylinder being collapsed. In the first place, a collapse can only be brought about when the external pressure overcomes that inside, and to obviate this the air or expansion pipes should be arranged and sized, that air can freely enter or escape from them. If pipes are in exposed positions, they should be protected from frost by well covering them with hair felt or other suitable material. A Float Valve may be fixed at the top of a cylindrical tank to prevent collapse, but as a rule it is not necessary. These valves are arranged to open and to admit air as soon as the water level inside the tanks begins to fall, but they close as the water level is raised. Vacuum valves are also used for the same purpose ; these are intended to open and to admit air should there be a tendency for a partial vacuum to be formed. The value of the latter class is doubtful, for should they stick a little at a critical time a sufficiently reduced pressure may occur to bring about a collapse. Noises in Boilers. These are generally produced by (a) Boilers which confine air at their upper parts. (&) Overheating of the water. (c) Partially choked circulating pipes. In the first case, when air is trapped inside a boiler, during the heating of water, a volume of air escapes, and produces a rumbling sound as it rises through the water to the outlet. The air of course is replenished through the renewal of the water. In the second case, the noises produced are similar to the above, only that they may be more pronounced. When water is overheated due to local circuits occurring within a boiler, steam is generated as this water rises and reaches a higher level owiug to its being subjected to a lower pressure. Over- heating, as already stated, is chiefly the result of retarded circulations, or it may be caused by a too powerful boiler being used. Partially choked circulating pipes often produce thumping sounds. It is seldom that a flow-pipe is totally blocked with lime deposits, although its bore may be reduced in some cases 422 DOMESTIC SANITARY ENGINEERING AND PLUMBING to one-third its original size. The return pipe, as a rule, is not affected to any great extent with deposited matter. When a flow-pipe is totally or partially choked, very violent thuds may be produced, although an explosion would not occur so long as the boiler return remained clear. The boiler, of course, is subject to damage by being burned, and by being subjected to considerable strain. Boiler Explosions. The most common cause of boiler explosion in connec- tion with hot -water ap- paratus is due to pipes being choked with ice. Stop-cocks in circulating pipes are oc- casionally responsible for it where safety valves have not been provided. It is also possible, under favour- able conditions, for an ex- plosion to be caused by cold water gaining admission to a boiler when the latter has been deprived of water and when heated to redness. From experiments it has been found that so long as the pressure due to the sudden generation of steam can be relieved through the supply pipe (assuming all other passages to be choked), an explosion will not take place. On the other hand, when the feed-pipe will not afford the necessary relief, owing to a screw-down stop tap acting as a non-return valve, the pressure due to the sudden generation of steam may cause the boiler to explode. The latter cause may be rare, but the possibility exists under the conditions named. Safety Valves. The question is often asked, should boilers in connection with hot water supplies be provided with safety FIG. 258. Dead- weight safety valve. DOMESTIC HOT WATER SUPPLY 423 valves ? No doubt they should be used where they are some- times absent, but in many cases safety valves can be safely dispensed with. For example, when either wrought iron or copper boilers are used, and where no trouble is caused by the deposition of lime salts, and provided boilers and pipes are arranged on inside walls, well away from the influence of frost, then under these conditions safety valves are not of much importance. Where, however, pipes are not properly protected, or where there is danger of them getting blocked with ice, or choked in any other manner, safety valves should be used. When cast-iron boilers are adopted, safety valves should be used, and in this case their use should be made compulsory, for the explosion of a cast-iron boiler may be attended with disastrous consequences. With regard to the position in which safety valves should be placed, no hard-and-fast line need be laid down, and each case should be treated upon its own merits. Where stoppages are not likely to occur in the primary circulating pipes, a safety valve may be fixed in any convenient place, either at the boiler or on one of the circulating pipes. From a general standpoint a safety valve should be fixed directly on a boiler. This, however, is not always practicable, especially with range boilers, and in many cases it is either necessary to fix a separate pipe to the boiler for receiving the safety valve, or to fix the latter to one of k the circulating pipes. Where trouble is caused by saline matter being deposited from water, a safety valve may, with advantage, be joined with a return circulating pipe, and as near to the boiler as possible. The latter position, under the conditions given, is the least likely to be affected by deposit, and if fixed directly to the top of the boiler there is the possibility of its being rendered useless. In the case of boot boilers, where soft water is used, safety valves may be fixed directly to the boiler, but when they are fixed in flues they require to be protected with a suitable form of loose cover. A safety valve, as a rule, should not be fixed on the top of a range, for in such a place it is liable to be knocked, and caused to leak. Safety valves for boilers take various forms. The dead- 424 DOMESTIC SANITARY ENGINEERING AND PLUMBING weight type, Fig. 258, is commonly employed, and it is simply constructed, the valve orifice being covered with a metal plug which is screwed to the top of the outer casing; over the latter loose weights are placed, the load being governed by the pressure at which the valve is to come into action. Another type of dead-weight valve (Jeffrey's patent) is shown in Fig. 259. This form of construction is intended to make the valve less liable to leakage if it should be given an accidental knock, or be jarred in any way. Instead of VULCANITE RUBBER WASHER. FIG. 259. Jeffrey's dead-weight safety valve. 5oFT CAP. \ .WASHER FIG. 260. Croydon relief valve. using a hard metal to metal facing, as in Fig. 258, the valve seating is covered by a vulcanite rubber washer in a cast-iron cap, the latter in turn being pressed upon by a rubber pad which is inserted in the upper casing of the valve. It will be observed that the upper part of the valve is hollow, and by introducing lead of shot the safety valve may be loaded to any reasonable extent. "When fixed, this valve should not be subjected to the heated products of combustion, otherwise the rubber pad may be soon destroyed and the valve caused to leak. DOMESTIC HOT WATER SUPPLY 425 Fig. 260 gives a different type of valve to either of the above. In this case, relief is brought about when the soft metal cap gives way. Another valve which is similar in principle has a thin mica disc in lieu of the metal cap. Such FIG. 261. Macintosh's mercury gauge and relief. valves, however, cannot be adjusted with the same precision as the dead-weight type. Spring safety valves are also largely used, the spring being either in a state of tension or compression. Generally speaking they are not so suitable for kitchen boilers as the dead-weight type. 426 DOMESTIC SANITARY ENGINEERING AND PLUMBING An arrangement which can be used in many cases, and which is less likely to get out of order than any form of valve, is the mercury gauge and relief (Macintosh's patent), Fig. 261. The glass tube T contains mercury, the pressure of which must exceed the normal pressure of water in the apparatus. This arrangement is intended to be joined with a flow-pipe F, and should the mercury be subjected to increased pressure, the necessary relief is afforded by the mercury being dislodged from the tube into the receiver E. To charge the tube with mercury the plug P is removed at the top of the gauge. The chief drawback to this type of gauge is its cost. Neither would it be suitable for fixing on circulating pipes, which are likely to get choked between the boiler and the point where it is introduced. CHAPTEE XIV LOW PRESSURE HOT-WATER HEATING APPARATUS IN the British Isles, where provision is made for warming buildings other than that by open fires, low pressure hot-water apparatus is frequently installed. For general residence and horticultural work this mode of heating has much in its favour, as the temperature of the heating surfaces can be regulated to suit different requirements, and a mild and humid atmosphere can be maintained. For warming large buildings, low pressure steam may be a more suitable heating medium, and in the case of factories and workshops where steam is often available it becomes unnecessary to provide a different heating medium. Either live or exhaust steam is suitable for warming purposes. In large rooms, where a big number of people congregate, a heating system should be installed which can be readily cooled down should the place get overheated. When the amount of heat that is stored in a volume of hot water is compared with the amount of heat contained in an equal volume of steam, it will be found that the latter contains less than one-hundredth the amount of the former, when the temperatures of the water and steam are 180 and 228 F. respectively. Thus it is evident that water is com- paratively slow in giving up its heat, and therefore not so suited as steam where quick fluctuations of temperature are desired. Systems of Piping. There are three general ways for arranging the piping in connection with low pressure hot- water apparatus (a) One pipe system. (&) Two pipe system, (c) Overhead, or drop system. 427 428 DOMESTIC SANITARY ENGINEERING AND PLUMBING In the one pipe system the main circuit is everywhere of the same diameter, whilst in a two pipe system the mains are sized according to the amount of heating surface to be served. The overhead or drop system consists mainly of vertical returns, and represents a modification and combination of the one and two pipe sytems. Fig. 262 shows a one pipe system when arranged for a single storey building, and it will be observed that the flow and return connections to and from each radiator join the same FIG. 262. One pipe system. main. In this system the cooled water from the heating surfaces mixes to a certain extent with the heated water from the boiler, and where a circuit is very long the water near the end may have a very low temperature. Where, however, the pipes are of a suitable size, the connections properly arranged, and the circuits not unduly long, the drawback mentioned is not so apparent, as the hottest water accommodates itself in the upper parts of the horizontal pipes, whilst the colder water occupies the lower parts. To obviate any unnecessary cooling of the water in a one LOW PRESSURE HOT-WATER HEATING APPARATUS 429 pipe system, the returns from the heating surfaces should be joined at the side of the horizontal mains, in order that the colder water may be delivered directly to the lower parts. It is important, when arranging mains, that no air locking shall take place, and to avoid this air-relief pipes require to be fixed at the highest points. Kadiators when connected to horizontal pipes also require to be provided with some form of air relief. In Fig. 262 the main from the boiler to point A constitutes the flow, whilst that part of the circuit from A to the boiler forms the return. B may also be made the highest part of the circuit, the return starting from that point. The feed cistern of an apparatus is usually fixed in any convenient place above the highest heating surface, and the supply pipe may either join the boiler or the return pipe, as found most convenient. It is not necessary to fix a feed cistern more than 2 or 3 feet above the top of the highest radiator, and to exceed this serves no useful purpose, but it subjects an apparatus to unnecessary strain. With regard to the point where a flow-pipe should join a radiator, it matters little in many cases whether it be at the top or at the bottom of the radiator. Generally speaking, the best results are obtained by joining the flow-pipe to the top, although the bottom connection is the neater of the two, especially when the branch is rather large. The inlet of each radiator should be controlled by a stop valve, in order that the temperature of the water can be modulated as required. When a building is two or more storeys high, a good method of piping for a one pipe system is that shown in Fig. 263. Where possible the heating surfaces on the upper and on the lower floors should be over one another, so as to reduce the number of risers which serve them. The return risers should join at the side of a main, whilst the flow riser in the majority of cases should be taken from the top. Each pair of risers forms a secondary circuit from the main circuit, and under ordinary conditions a good circulation through the risers is assured. In Fig. 263 the boiler is assumed to be centrally placed, and to the left and right main circuits are provided. When a 430 DOMESTIC SANITARY ENGINEERING AND PLUMBING building is large, the number of principal circuits can be increased, and this practice is commendable, as the lengths of the circuits may be diminished, and better results obtained. One section can also be put out of use without interfering with the working of the remainder of the installation. The cold water supply in Fig. 263 is shown joined at the end of one of the circuits, where an air-pipe is also pro- vided. A one pipe system has the special advantage of being FIG. 263. One pipe system with flow and return risers. immune from short circuiting, and in consequence can be adopted where a two pipe system would result in failure. A two pipe system is shown in Fig. 264, and the manner in which it differs from a one pipe system is in its circuit being graded and in the riser returns joining the main return. The maximum size of the flow-pipe starts at the boiler, and its size is decreased as the heating surface is supplied, until the head of the circuit is reached. From the latter point the circuit forms a return, and it is increased in size as the branch returns are joined with it. The chief drawback of a two pipe system is its liability to short circuit, and for a portion of a system to be either rendered useless or be very much impaired. Short circuiting takes place when too much resistance is encountered by the circulating water, but this may be due to LOW PRESSURE HOT-WATER HEATING APPARATUS 431 the improper grading of the pipes, to the latter having insufficient pitch, or to pipes being dipped or trapped. The chief merit of a two pipe system, when well designed and properly installed, is that the cooled water from the heat- ing surfaces is delivered directly to the return, instead of mixing with, and cooling, the heated water in the flow-pipe. This point is very important in large buildings where long circuits are imperative. For a two pipe system to be a success, the pipes require to be properly sized, and be given a moderate pitch. Dips must FIG. 264. Two pipe system with flow and return risers. be avoided, and suitable fittings used, and unless these pre- cautions are observed short circuiting will take place to a more or less extent. Pitch of Pipes. Where practicable, heating mains should have a pitch of 1 inch in 10 feet, and for small sized branches the pitch should not be less than 1 inch per foot. When a main is of a comparatively small bore, and fixed quite level, the movement of the water through it is considerably retarded. Overhead or Drop System. A very suitable method of piping for high buildings, and where the heating surfaces can be arranged to come above one another, is the overhead or drop system, Fig. 265. In this system the flow -pipe passes directly from the boiler to a high elevation, where an air-pipe is provided. From the highest point one or more overhead 432 DOMESTIC SANITARY ENGINEERING AND PLUMBING horizontal pipes are taken (depending upon the size of the building), and from the latter, vertical returns are dropped to pass close by the heating surfaces. No work, as a rule, is put upon the flow-pipe, and no other air-escapes in the form of valves or pipes may be necessary, other than that at the head of the flow-pipe. Air is chiefly given out from water when the ^m^^ FIG. 265. Overhead or drop system. latter is heated, and the liberated air rises and escapes at the highest point. After the water leaves the boiler a cooling action sets in, with the result that little air is given out from the water when in the return pipes. To the vertical return No. 1, horizontal connections are shown between it and the radiators, but these should only be short when this form of connection is adopted. A better mode of arranging the con- LOW PRESSURE HOT-WATER HEATING APPARATUS 433 nections is that on the vertical return No. 2, where tees are used which aid the circulation through the surfaces. Owing to the large amount of vertical piping in an overhead system, a quick circulation is ensured, and in consequence the pipes can be smaller than in other systems, where a greater proportion of horizontal pipes are used and where the circula- tion is slower. The horizontal overhead and low level returns require to be graded as regards their size, according to the amount of work put upon them. The vertical returns which serve the radiators are of uniform bore from end to end. It frequently occurs when arranging the piping for an FIG. 266. One pipe system where circuit dips beneath doorways. installation that obstructions require to be passed. To attain this end the formation of one or more dips may be unavoidable, but as they impede circulation they should be avoided as far as possible. The best system of piping where dips are im- perative is the one pipe system, but as regards the general arrangement of pipes for all cases no hard-and-fast rules can be laid down. In Fig. 266 a case is shown where a couple of doorways come in the line of piping, and in order to pass them the flow- pipe is carried up and over them. The remaining part of the circuit from the air-pipe forms a return, and instead of taking the latter back and over the doorways it is dipped beneath them as shown. By rising the flow-pipe and keeping the return-pipe low the circulating head is increased, and this overcomes to a 28 434 DOMESTIC SANITARY ENGINEERING AND PLUMBING great extent the retardation introduced by the trapped return. At A, Fig. 266, an air-pipe is shown in order to liberate air, which tends to gather at that point either when charging the apparatus or when the latter is in use. The Circulating Head is the vertical distance between the highest part of a flow-pipe and the fire-bars of a boiler, and the power to produce circulation is the difference in density between the ascending and descending columns which consti- tute the flow and the returns. Another one pipe system is shown in Fig. 267, where the FIG. '267. One pipe system where circuit dips beneath doors. pipe requires to be dipped beneath a number of doorways. In this case radiators are fixed on each of two floors. The flow- pipe passes directly from the top of the boiler to the upper floor, and it is supposed to be carried up to near the ceiling in order to give extra power for circulating the water through the dipped return. Only two air-pipes are necessary, as at x and y, for any air which finds its way into any other part of the circuit will rise and accumulate in the radiators. From the latter air can be periodically released by opening the air-cocks. Sizes of Pipes. The following table gives the aproximate LOW PRESSURE HOT-WATER HEATING APPARATUS 435 amount of heating surface supplied by different sizes of pipes : TABLE XVI. Internal diameter of pipe. Square feet of surface, chiefly horizontal pipes. Square feet of surface, horizontal and vertical pipes. Square feet of surface, vertical pipes. 1 in. 40 50 75 u 75 90 130 H 130 150 250 2 230 360 600 24 360 500 800 3 520 650 1100 4 900 1280 2500 5 1600 2000 6 2300 3460 8 4600 6150 10 8000 10,400 12 11,520 14,000 With regard to the size of branches and radiator con- nections, these are given below. Less than 50 sq. feet heating surface 1 in. diameter. More than 50 and less than 80, 1 Over 80 1J The sizes of branches should also be regulated by their general arrangement, for where they are rather long, and portions lie flat, they may advantageously be increased in size. Heating Surfaces take the form of pipes, coils, and radiators. Either of the two latter is used when a large area of heating surface requires to be concentrated in a com- paratively small space. Pipes are well suited for warming works and horticultural buildings, but for residences, offices, and similar buildings, radiators are preferable. Eadiators usually have their surfaces vertically arranged, and they possess advantages over other forms of heating surfaces in that they are neater in appearance and collect less dust. Kadiators are made in many plain and ornamental forms, so as to suit any particular position in a building. The best type of radiator, when the heating surfaces are exposed, is the 436 DOMESTIC SANITARY ENGINEERING AND PLUMBING column or loop class, where the sections are spaced to enable the whole of their surfaces to be readily cleansed. Fig. 268 gives a single column radiator by the Beeston Foundry Co., each section when 36 inches high contains 3J square feet of surface. Two and three column radiators of the same height contain per section 4 and 5f square feet of surface respectively. FIG. 268. Single column radiator by the Beeston Foundry Co. Ltd. There are two methods of joining the sections of radiators. One is by the use of tapered nipples, which are inserted in the upper and lower openings of the sections, and where wrought-iron tie rods are used to bind the sections together. The other, and better method, is by left and right-hand screwed nipples, the sections being tapped to suit. For simple systems of ventilation, the cold inflowing air is often warmed by means of ventilating radiators. When a LOW PRESSURE HOT-WATER HEATING APPARATUS 437 good type of radiator is selected, this method of inlet ventila- tion has much in its favour, as the air supply can be readily controlled and all parts may be arranged so as to admit of their being readily cleansed. Fig. 269 shows a simple form of ventilating radiator. In FIG. 269. Ventilating radiator by the Beeston Foundry Co. Ltd. this case baffle plates are fitted on both sides of an ordinary single column radiator, and a base is provided to admit fresh air at the back. The baffle plates are held in position by lugs which are cast on them, and they can be readily removed for cleansing purposes. A ventilating radiator of the flue type is illustrated in Fig. 270, and the fresh air may be arranged to enter at the back as shown, or through the floor beneath. This radiator 438 DOMESTIC SANITARY ENGINEERING AND PLUMBING is also designed that every part may be seen and readily cleansed. A drawback of many ventilating radiators, however, is their hidden parts, where dust and other matter can accumulate. FIG. 270. The " Marshall" ventilating radiator by the Beestou Foundry Co. Ltd. For this reason ordinary loop radiators are often used, either with air grates immediately beneath them or in a wall behind. Of course when the ordinary type of radiator is adopted and where baffle plates are not used, the air is not warmed so well, and there is greater likelihood of draughts being felt. Fig. 271 gives a hinged radiator. This is a very useful form for LOW PRESSURE HOT-NVATER HEATING APPARATUS 439 many situations, as it can be swung from the wall and so enable the space behind to be readily accessible. Comparative Value of Heating Surfaces. The relative amount of heat emitted by a surface largely depends upon the form it takes, upon its roughness or smoothness, and to some extent upon the kind of metal which forms the surface. FNJ. 271. The " Hospital Hinged" radiator by the Beeston Foundry Co. Ltd. When heating surfaces are exposed to view, the air of an apartment is warmed in two ways, viz. by radiant and by convected heat. Radiant heat passes from its source in straight lines, and is absorbed by the cooler surfaces of walls, furniture, and other objects. The intervening air space is not very appreciably warmed by radiant heat, as air alone is not warmed by direct 440 DOMESTIC SANITARY ENGINEERING AND PLUMBING rays of heat. Air, however, in buildings always contains floating particles of matter which absorb a certain amount of heat, and these in turn give up some of the heat they have received to the air which envelopes them. Converted heat is that which is absorbed by air coming in contact with heating surfaces, or by contact with objects which have been warmed by radiant heat. Eadiant heat can be intercepted by means of a screen, whilst convected heat, which is conveyed by the circulating air, cannot be cut off by this means. Small pipes emit more heat per unit area of surface than larger ones, when the difference in temperature between the pipes and the air surrounding them is the same, and when the rate of circulation is also equal. Heating surfaces when grouped close together, such as in coils and radiators, emit less heat than where they are farther apart ; in the former case a large percentage of the radiant heat is not utilised for warming purposes, as it cannot get away, but is simply radiated and re-radiated from surface to surface. The efficiency of heating surfaces is also affected by their height, for air currents, upon being warmed by contact at a low level, absorb less heat as they ascend. From the above it will be clear that the most efficient radiator is the single column class, when the sections are a reasonable distance apart. Polished copper, brass, or nickel plated pipes emit less heat per unit area than other surfaces. The positions in which the greater percentage of the heating surfaces should be placed are in the coldest parts of a building. These, obviously, are the windows, external walls, and near doorways. In buildings of considerable width it is necessary to fix heating surfaces in other positions than those stated, in order that warmth can be evenly distributed through- out the whole space. Kadiators should be fixed about 6 inches from walls where practicable, otherwise there is greater likelihood of the latter being soiled. Air, upon being warmed, leaves the heating surfaces with more or less considerable velocity, with the result that particles of dust which the air contains stride LOW PRESSURE HOT-WATER HEATING APPARATUS 441 against any adjoining wall and discolour it. Discoloration of walls may also be avoided by the use of light shields, which are occasionally placed on the tops of radiators to divert the air currents towards the centre of the apartments. Radiator Valves. A suitable pattern of valve is that FIG. 272. Angle valve by the National Radiator Co. Ltd. shown in Fig. 272, and it allows a neat and simple connection to be made. Whatever form of valve is used, one should be selected which does not unduly impede the movement of the water. Air-Valves. To admit of the escape of air from pipes and heating surfaces, air-valves are frequently used, These 442 DOMESTIC SANITARY ENGINEERING AND PLUMBING may be divided into two classes: (a) Those which are periodically opened by hand; (b) those which are automatic in action. The latter are suitable for a circuit where an air-pipe cannot be fixed. A simple form of automatic air- valve is given in Fig. 273. Its action is as follows : If we assume that the valve is closed by the ball being partially submerged, the FIG. 273. The "Ideal" automatic air valve by the National Radiator Co. Ltd. air passes through the water to the upper part of the valve. Should the accumulation of air be continuous, the water is displaced from the small pocket, when the ball falls by its own weight, opens the valve, and permits the air to escape. The discharge of air is followed by water and the ball is again buoyed up and the orifice closed. Feed Cisterns. The size of feed cisterns should be sufficient to accommodate the increased volume of water, when the latter is raised in the apparatus to its maximum LOW PRESSURE HOT-WATER HEATING APPARATUS 443 temperature. Approximately, water expands Y V f i ts bulk when raised from 40 to 212 F. The highest water-line of a cistern should be a few inches beneath the overflow, and the ball-cock should be arranged to close when a cistern contains only a few inches of water. As regards the size of a supply pipe which communicates between the feed cistern and the apparatus, this should not, as a rule, be less than 1 inch diameter. A dip or trap is necessary in a feed-pipe to prevent hot water circulating back to the supply cistern, and when the circulation in a system is liable to be sluggish a deeper trap than usual should be formed. Calculation of Heating Surface. Windows, roof lights, and external walls are the principal cooling surfaces in build- ings, and the renewal of air for ventilation is also responsible for more or less considerable absorption of heat. Doorways, floors, internal walls and crevices, also account for loss of heat, and in countries where the cold is very intense these are also taken into account. In the British Isles, however, it is usually sufficient to consider the principal heat losses, such as those due to glass surface, external walls, ventilation, and the ex- posure of buildings. To cover minor heat losses an allowance of 10 per cent, of the above is usually ample. Discharge of Air through Flues. To calculate the actual volume of air which will constantly pass through a flue or opening when the movement of air is dependent upon natural agencies is impossible. The volume of air which will flow through an ordinary chimney varies considerably, the actual discharge being governed by the height, size, and form the flue takes, by the difference between internal and external temperatures, by the freedom with which air can enter a room, and by the kind of walls in which the flues are formed. It is thus clear that discretion requires to be exercised when ascertaining the probable average discharge of air through any opening or extract shaft. On an average, when a fire is burning the discharge of air per sq. foot of flue area is about 11,000 cubic feet per hour. If a building is already erected, the velocity of air through an opening can be determined by an air-meter, but the records will vary considerably. It is usual, however, to take a number 444 DOMESTIC SANITARY ENGINEERING AND PLUMBING of readings, the mean velocity being used for the basis of calculation. In this case the volume of air discharged by a flue will equal the area of its cross section, multiplied by the mean velocity in feet. When an air-meter cannot be used the following formula will aid in ascertaining the discharge of air through an upcast shaft : . (88) Where Q = discharge in cubic feet per hour. a = area of cross section of flue in inches. li = height of flue in feet. T = temperature of air in flue. t external temperature of air. Example, 53. If a 15 in. by 9 in. duct is 24 feet high, and the external and internal air temperatures are 45 and 60 F. respectively, determine the approximate volume of air this duct should discharge per hour. By Formula 88, Q = 80 x a X Substituting values given, , K n /24x (60-45) V J Q = 80xl5x9x-844; /. Q = 9115 cubic feet. Assuming that a heating coil had been placed at the base of the duct in order to raise the escaping air to 80 F., the velocity of air through the shaft would have been increased. Under these conditions the discharge should be 460 + 45 ' Q = 80x 15x9x1-289; .-. Q = 139,21 cubic feet. Heat to Warm Air. To raise 1 cubic foot of air from 30 to 31 F. requires '01928 B.T.U., so that the volume of LOW PRESSURE HOT-WATER HEATING APPARATUS 445 air which can be raised through 1 degree by one heat unit will be , niQOQ = 51'86 cubic feet. In order to simplify matters 'OLuZo we will take the latter value at 50 cubic feet, and this will be sufficiently accurate for practical work. Thus the number of heat units necessary to make good the loss of heat due to ventilation can be obtained by multiplying the total volume of air in feet by the temperature through which the air is raised and by afterwards dividing by 50. Heat absorbed by Walls. The heat lost by walls varies according to their thickness, to the class of material used, to the treatment of their surfaces, whether cavities are formed in them or not, and according to their relative exposure. The values given by different authorities vary, but the most reliable information on this matter known to the writer is that given in the German work by Recknagel and Eietschel. Tables XVII. and XVIII. are from that work, but the values are converted into English Units by Professor Kinealy and given in book Formulas and Tables for Heating. TABLE XVII. Loss OF HEAT THROUGH BRICK WALLS IN BRITISH THERMAL UNITS PER SQUARE FOOT OF SURFACE PER HOUR, PER DEGREE DIFFERENCE OF TEMPERATURE, THE BRICKS BEING 8J X 4 X 2 IN., WITH f IN. MORTAR JOINTS Outside walls. Inside With additional stone face. With air wall, space Thickness both of 2-4 of wall. No plaster. One side plas- tered. sides plas- tered. 4 in. thick. Sin. thick. 12 in. thick. inches plas- tered. brick 52 49 43 1 , 37 36 33 31 29 26 25 H , 29 '28 26 25 23 21 21 2 bricks 25 '24 22 20 19 19 21 22 21 19 18 17 16 3 2 19 18 17 16 15 14 31 16 16 15 14 13 13 4 14 14 ... ... ... 12 4 , 12 12 ... ... i 446 DOMESTIC SANITARY ENGINEERING AND PLUMBING TABLE XVIII. Loss OF HEAT THROUGH STONE WALLS IN BRITISH THERMAL UNITS PER SQUARE FOOT OF SURFACE PER HOUR, PER DEGREE DIFFERENCE OF TEMPERATURE Total thick- ness of wall. Sandstone. Limestone. Total thick- ness of wall. Sandstone. Limestone. 12 inch 45 49 32 inch 26 28 16 39 43 36 24 26 20 35 38 40 ,, 22 24 24 ,, 31 35 44 21 23 28 28 31 48 19 21 HEAT LOST BY GLASS SURFACE This is given in the following table along with other partic- ulars. The values given by the two authorities differ a little, but not to any considerable extent : TABLE XIX. Authorities. Kind of surface. Recknagel and Rietschel. German Government standard. Single windows 1-03 T09 Double ,, : . 472 518 Single skylight 1-092 1-118 Double 492 621 Doors . 410 414 Fireproof floor 124 ,, ceiling ... 145 Plaster 1'6 to 2 '6 in. thick 615 ,, 2-6 to 3'2 in. ,, 492 ... (1) (2) (3) The total heat lost in B.T.U. by walls and glass can be obtained by multiplying the exposed area, by the difference of the air temperature on the two sides, and afterwards by a suitable value from Tables XVII. to XIX. According to Professor Kietschel the values in Tables XVII. and XVIII., and also those in column 2 of Table XIX., should be increased as stated below. When the exposure is a northerly one, and the winds are important factors, increase by 10 per cent. LOW PRESSURE HOT- WATER HEATING APPARATUS 447 Where a building is heated during the daytime only, and is not an exposed one, increase by 10 per cent. Where a building is exposed, and only heated during the daytime, increase by 30 per cent. Where a building is heated intermittently during the winter months, and with long intervals of non-heating, increase by 50 per cent. In order to arrive at the amount of heating surface to warm a building, we must next know how many heat units are emitted by such surface. The precise amount of heat emitted chiefly depends upon the form the surfaces take, and upon the difference in temperature between the surfaces and the atmosphere which envelopes them. Professor Carpenter of Cornell University, in his work Heating and Ventilating Build- ings, gives the following values, which are reproduced in the table below : TABLE XX. HEAT UNITS EMITTED PER SQUARE FOOT OF HORIZONTAL PIPE SURFACE PER HOUR, FOR DIFFERENT RANGES OF TEMPERA- TURE BETWEEN THE SURROUNDING THEM HEATING SURFACE AND THE AIR Total B.T.U. per square foot per hour. Ti^flfV^-^. _- ,, Jjinerence 01 temperature Diameter of pipe. degrees F. 6 in. 4 in. 2 in. lin. 40 49-6 56-2 59 77 50 64-5 73 77 100 60 79-8 90 95 124 70 95-2 108 113 148 80 112-0 127 133 173 90 128 147 153 199 100 147 167 175 228 110 166 188 198 257 120 184 208 219 287 130 203 230 242 318 140 223 252 266 346 150 244 276 291 378 160 265 300 316 410 170 286 324 341 443 180 307 348 367 475 190 330 375 393 512 200 356 403 415 552 448 DOMESTIC SANITARY ENGINEERING AND PLUMBING The amount of pipe heating surface can now be obtained by first ascertaining the total heat losses per hour, and after- wards dividing by a value from Table XX. which agrees with the conditions given. Example 54. If a room contains 200 sq. feet of external sandstone wall which is 20 inches thick, 60 sq. feet of glass, and 10,000 cubic feet of air at a temperature of 30 F. are passed into it per hour, determine the area of 4-inch pipe surface which will maintain an internal temperature of 60 F. when the average temperature of the water in the pipes is 160 F. Heat absorbed by Air. The air required for ventilation is 10,000 cubic feet per hour, and to raise this from 30 to 60 requires oU Loss by Wall Surface. In Table XVIII. the loss through a sandstone wall when twenty inches thick is '35 B.T.U. per sq. foot per hour per degree difference of temperature. The loss therefore by 200 square feet, when the difference in tempera- ture between the internal and external surfaces is (60 30) = 30, will be 200 x 30 x -35 = 2100 B.T.U. Loss of Heat by Glass. The heat lost by glass for an ordinary window according to column 2, Table XIX., is 1*03 B.T.U. per sq. foot per degree difference of temperature per hour. Therefore the loss due to 60 feet, when the difference in temperature between the two sides is (60 30) = 30, will be 60x30x1-03 = 1854 B.T.U. The heat absorbed by air, exposed wall, and glass equals 6000 + 2100 + 1854= 9954 B.T.U. Adding, for minor losses due to doorways, ceilings,floors, etc., 10 per cent, of the above = 995 B.T.U. The total heat losses will equal = 10,949 B.T.U. LOW PRESSURE HOT-WATER HEATING APPARATUS 449 Heat emitted by Pipe Surface. In Table XX. the heat emitted by a sq. foot of 4-inch pipe surface per hour, when the difference between the pipe and air of apartment is (160 60) = 100, is given at 167 B.T.U ; . ' . , 10949 ,, 94 /. heating surface required = ^ =60 sq. feet. By following the calculation it will be evident that the provision for ventilation has a big influence on the amount of heating surface required. The area of exposed wall and the amount of glass are also very important factors. In the British Isles a heating installation should be capable of maintaining an inside temperature of 60 F. when the out- side air is 30 F. In America and other countries, where the cold is intense in winter, an inside temperature of 70* is usually required when the outside air is at zero. To calculate the heating surface required for an internal temperature of 60, and an external temperature of 30, when the average temperature of the water is 160, the following simple formula may be used : Where K = total pipe surface in sq. feet. Q = cubic feet of air passing through apartment per hour. W = area of exposed wall surface in feet. G = area of glass in feet. Badiator Surface. Eadiators emit less heat per unit area than horizontal pipes, and when the former are used their heating surface should be increased over that of pipes as follows : For plain 1 column radiators add 5 per cent. 9 10 " AV/ Example 55. A large room contains 10,500 cubic feet of space, 870 square feet of exposed wall surface, and 250 square feet of glass. If the air of the room is changed 3 times per hour, 29 450 DOMESTIC SANITARY ENGINEERING AND PLUMBING determine the amount of heating surface, when single column radiators are used to maintain an inside temperature of 60 when the external air is 30 F. Average temperature of water in pipes 160 F. By Formula 89, E = + + ' 10500x3 , 870 , 250 Substituting values given, E = T -- '"TT IT' Total pipe surface =2287 sq. feet. For single column radiators, plus 5 per cent. . . . 114 sq. feet. /. total radiator surface = 2401 sq. feet. Heating Surface for Drying Rooms. When low pressure hot-water apparatus is utilised for drying rooms, the following formula may be used for obtaining the heating surface required, to maintain an internal temperature of 80 when the outside air is 30 F., and when the average temperature of the water in the pipes is 170 F. : Ra3 140 + T + 20 + 3~ ' ' (90) Where E = total area of pipe surface in square feet. Q = volume of air passed through room per hour. W = area in feet of external walls. I = area in feet of internal walls. G = area of glass in feet. To obtain satisfactory results in drying rooms, free ventila- tion is necessary, and the entering air should be diffused as evenly as possible throughout the whole of the space. Example 56. A drying room measures 20 feet in length, 12 feet wide and 12 feet high. The area of the external wall surface is 144 sq. feet, and that of the internal walls 624 sq. feet. There is no glass. If the ventilating arrangement provides for 5 air changes per hour, find the heating surface required to give an inside temperature of 80 when the outside air is 30 F. LOW PRESSURE HOT-WATER HEATING APPARATUS 451 Using Formula 90, B = +++ . ,,.,.. , . -p 20x12x12x5,144,624 Substituting values given, K =- .. JA + -^-+- 140 20 .-. E = 154-5, say 155 sq. feet. The area of a pipe surface may be found by multiplying its circumference by its length, or the following rule may be used : rfxllx* ( , 42 Where A == area of surface in feet. d = external diameter of pipe in inches. 1 = length of pipe in feet. For approximations, it may occasionally be desirable to calculate the heating surface for warming a building from its cubic capacity. This method is not accurate for isolated apartments, as the chief factors, such as exposed walls, glass surface, and ventilation, differ much in different parts of a building. The following Table gives the approximate number of cubic feet of space warmed to different temperatures by one square foot of pipe surface, when the outside air is about 30 F., and water in the pipes not less than 160 F. : TABLE XXI. Kind of building heated. Inside temperature Space warmed by 1 square foot of Fahr. surface. Workshops and factories 50 130 cubic feet. Warehouses . 55 100 Churches and large rooms 60 86 Living rooms 60 58 65 50 Entrance halls 70 42 f 75 34 Drying rooms . . . . 80 85 27 22 I 90 18 452 DOMESTIC SANITARY ENGINEERING AND PLUMBING Boilers for Low Pressure Heating Apparatus. For medium and large-sized installations, independent cast-iron sectional boilers are superseding those of wrought iron, as the former are more durable, and their heating surfaces can be more advantageously shaped and arranged. Brickwork settings are dispensed with, and sectional boilers also possess the advantage of portability, as they can be taken through narrow openings. Small heating systems require less powerful boilers than the cast-iron sectional type, and for these many other forms of independent boilers can be readily obtained. The selection of a boiler should be governed to a great extent by the class of fuel to be consumed. Where a fuel such as anthracite is used, smaller and more tortuous passages between the heating surfaces are permissible, in order to extract as much heat as possible from the fuel consumed. Soft bituminous coals require a simple form of boiler, owing to the amount of soot deposited in the flues. All boilers require ample provision in the form of soot doors, to enable the heating surfaces to be periodically freed from soot. Heating surfaces of boilers are either direct or indirect ; the former are exposed to the fire, and receive the flame impact, the radiant heat from the burning fuel, and the heated products of combustion. Indirect surfaces are those which do not face the fire, but receive their heat only from more or less flame impact, and the heated products of com- bustion. Indirect surfaces absorb considerably less heat than direct surfaces, but they are useful as they deprive the products of combustion of much heat, and prevent them escaping from the boiler at an unnecessarily high temperature. In connection with indirect boiler surfaces there is a tendency to overvalue them. It is often stated that the value of indirect surfaces is about one-third that of direct surfaces, but this valuation in most cases is considerably overrating them. A general comparison of the value of direct and indirect boiler surfaces is unsatisfactory, for very much depends upon the ratio of the grate area to the heating surface and the design of the boiler. From a test made with a boiler which had a relatively large amount of indirect heating surface, the amount LOW PRESSURE HOT- WATER HEATING APPARATUS 453 of heat transmitted through each sq. foot of surface per hour worked out at 2500 B.T.U. The difference in value of the heating surface in contact with the fire and that most distant was considerable, and whilst a sq. foot of surface most favourably located would transmit 10,000 B.T.U and over per hour, a sq. foot farthest removed would probably not transmit 250 B.T.U in the same time. Thus, for this case, some parts of the indirect surface would have not more than one-fortieth the value of the best direct surface, area for area. FIG. 274." White Rose " boiler by Hartley & Sogden. Intermediate parts of the indirect surface would have a higher value, but the example will indicate the difficulty of assigning a true value to the indirect heating surface of a boiler. Fig. 274 gives one of Hartley & Sugden's cast-iron sectional boilers, where the sections are vertically arranged, and joined together by nipples and bolts. The upper part of each section is constructed to form a centre and two side flues. At the fire-box the flames and heated products of combustion envelope, to a more or less extent, the surfaces immediately 454 DOMESTIC SANITARY ENGINEERING AND PLUMBING overhead, and afterwards they pass into and along the upper side flues towards the front of the boiler, and thence through the centre flue back to the chimney. There are numerous waterways which increase the heating surface, and the grate area is proportional to the number of sections a boiler contains. This type of boiler is made in different sizes. Their lengths vary from 3 ft. 4 in. to 6 ft. 10 in., and the sections have a uniform width of 2 ft. 7 in., and they are catalogued to heat from 575 to 2500 sq. feet of radiator surface. Another type of cast-iron sectional boiler by Lumby Sons, Wood & Co. Ltd., is given in Fig. 275. This is a larger type than that given in Fig. 274, and, as will be seen, contains proportionally more horizontal heating surface. The course of the flames, and products of combustion, after passing between the surfaces immediately over the fire-box, pass along towards the front of boiler, and thence to the chimney through the top horizontal flues. This pattern of boiler is made in various sizes, the greatest length being 6 ft. 2 in. and the shortest 2 ft. 8 in. The catalogue ratings of these boilers vary from 1600 to 4000 sq. feet of radiator surface. There are many makers of cast-iron sectional boilers, and the boilers illustrated are only intended to show the general form they take, not to infer that they are superior to those produced by other firms. The "Trentham" Cornish boiler, Fig. 276, is suitable for large high buildings. It is circular in form, being made of f -inch wrought-iron plates. For this boiler, brickwork settings are required, the flues being arranged that the heated products of combustion, after leaving the combustion chamber, can pass under the lower portion of the boiler towards the front, and afterwards back over the upper part to the chimney. A water- way bridge is formed, and the heating surface can be further increased by providing cross tubes in the combustion chamber as shown. The " Trentham " boiler, owing to its shape, is suitable for withstanding high pressure, and it is made in sizes from 2 ft. 8 in. to 5 ft. diameter, and from 4 to 18 feet in length. These are catalogued as being capable of heating from 900 to 10,500 feet of 4-inch pipe. A good draught is required for boilers with long and LOW PRESSURE HOT- WATER HEATING APPARATUS 455 tortuous flues, and the height of a chimney should exceed the length of the horizontal flues. \ i Fig. 277 gives a simple and effective form of wrought-iron boiler when the cross tubes are arranged in the manner shown. 456 DOMESTIC SANITARY ENGINEERING AND PLUMBING The chief objection to. this type of boiler is its height, and this limits its use on that account. Section. Elevation. Fro. 276. Trenthani boiler by Lumby Sons, Wood & Co. Ltd. LOW PRESSURE HOT-WATER HEATING APPARATUS 457 For a small heating installation the dome-top boiler, Fig. 246, or one of a similar grade, may be adopted. The drawback of this type is its low efficiency as a heater, but it has the advantage of a low initial cost. When the heating FIG. 277. " Goliath " boiler by Lumby Sons, Wood & Co. Ltd. surfaces of a boiler are mainly of a vertical nature, and where there is a large clear passage to the outlet flue, the greater portion of the heat given out from the fuel escapes into the chimney. Boiler Draught Regulator. An automatic device, which is 458 DOMESTIC SANITARY ENGINEERING AND PLUMBING shown in Figs. 255 and 256, is very useful for controlling the draught of a boiler, and for economising fuel. To prevent unnecessary loss of heat from boilers and main pipes, these should be covered with a good form of insulating material. Various substances are used for this purpose, but all allow a certain amount of heat to pass through them. Sizes of Boilers. A common method of estimating the power of a boiler is to assume that each sq. foot of its surface will transmit sufficient heat to supply about 35 sq. feet of radiator surface. Taking the average amount of heat emitted by a single column radiator as 160 B.T.U. per sq. foot per hour, then each sq. foot of boiler surface should transmit 160x35 = 5600 B.T.U. per hour. For a normal rate of firing, the latter value is very much higher than can be obtained in practice, and such basis for rating boilers which differ so widely in construction is a very inaccurate and misleading one. The amount of heat which is transmitted by a sq. foot of boiler surface is not a fixed quantity, but depends upon (a) the design of the boiler ; (b) the ratio of the grate area to the heating surface ; (c) the rate of firing ; and (d) the class of fuel used. In operation, the condition of a boiler flue is also an important factor. More failures have resulted in connection with heating systems by installing boilers which are too small than by any other means. Most boiler catalogues are considerably over- rated, but many makers introduce a kind of saving clause in which they recommend that a larger size be selected than the one listed to do the work required. In the following Table is given the heat value in British thermal units per pound of different fuels (Moles worth's) : TABLE XXII. Coal average 14,000 B.T.U. Coal Lancashire (steam) 13,900 B.T.U. Welsh (steam) . 16,000 ,, Newcastle . 14,000 ,, ,, Welsh (medium) 13,900 ,, Coke .... 12,500 ,, Scotch 13,500 Coke (gas) . 10,000 LOW PRESSURE HOT-WATER HEATING APPARATUS 459 It is not often that 50 per cent, and over of the heat in the fuel is transmitted through the surfaces of a low-pressure boiler, owing to the latter not being fired to advantage. The fuel is often added at long intervals, the fire-box being sufficiently commodious to hold in many cases from 5 to 10 hours' supply. These slow rates of combustion give a low efficiency, and for best results a moderately high rate of firing is imperative. For obtaining the size of boilers the formulae beneath are given a = ^^ (92) wxfxc R= axwxfxc u Where a = area of boiler grate in feet. R = total heating surface of radiators in sq. feet. w = heat units emitted per hour per sq. foot of radiator surface. w = weight of fuel in Ib. consumed per sq. foot of grate. c = a coefficient, the value of which depends upon the type of boiler and the rate of firing (see Table XXIIL). /=heat value of fuel (see Table XXIL). For radiators, the average value of u may be taken as 160, and for values of pipes see Table XX. TABLE XXIIL VALUES OF c FOR DIFFERENT FORMS OF HEATERS AND DIFFERENT RATES OF FIRING For good sectional self-contained boilers .... Cornish " Trentham " boilers ,, vertical self-contained boilers with cross tubes . ,, self-contained dome-top boileis c= -5 to -65 c='4 to -55 c='5 to -6 c='35 to -45 The fuel consumption w to agree with the values of c is also given. For very low rates of firing the values of c will 460 DOMESTIC SANITARY ENGINEERING AND PLUMBING decrease, the rate of firing for working conditions varying from 3 to 12 Ib. of fuel per sq. foot of grate per hour. Boilers set in brickwork w = Q to 9 Ib. Sectional boilers with winding flues w = *J to 10 Ib. Self-contained vertical boilers w = 8 to 12 Ib. Example 57. If a system contains say 1500 sq. feet of heating surface, and a sectional cast-iron boiler is adopted, find size of boiler required to satisfy a rate of firing of 8 Ib. of fuel per square foot of grate per hour. Assume the coal will yield 14,000 B.T.U. per Ib. By Formula 92, a= R *.^ . wxfxc For the rate of firing given, the value of c will be say -5. 1500x160 Substituting values, a = /. q = 4'29 sq. feet of grate. Should a higher rate of firing be adopted a smaller size of boiler would suffice, but the stoking would require to be done at shorter intervals. For the boiler and conditions given in Example 57, the fuel consumed would amount to 4'29 x 8 = 34'32, say 34 Ib. per hour. Example 58. The fire-grate of a boiler like Fig. 277 has an area of 1*8 sq. feet. Find the amount of heating surface this boiler will serve, if coke, which has a heat value of 12,500 B.T.U. per Ib., is the fuel burned, and when the rate of firing is 9 Ib. per sq. foot of grate per hour. Using Formula 93, E = The value of c for this boiler from Table XXIII. for the rate of firing given will be about *5, _ l-8x9x!2,500x-5 and substituting values, E= - ~TfiO " ' .% E = 632-8, say 633 sq. feet. Chimneys. Failure of boilers is occasionally due to defective draught owing to the chimneys being too small or LOW PRESSURE HOT- WATER HEATING APPARATUS 461 too short, containing too many bends or being formed of long lengths of exposed metal pipes. Circular chimneys offer the least resistance, and for small boilers where a flue is formed of metal pipes, the minimum diameter should be 5 inches. "Where possible, a boiler should be joined with a brick chimney, as the latter is less influenced by the weather. The effective area of a chimney is less than its actual area owing to the soot which accumulates on the surface. To determine the size of a chimney for a heating installa- tion, the following formulas may be used : For installations containing less than 700 sq. feet of heating surface / (94) When an installation contains 700 sq. feet of heating surface and over /.rr> . . i > (95) Where d diameter of chimney in inches. h = height of chimney in feet. K = total heating surface of radiators and pipes in feet. In the case of a square chimney, the length of one side may be considered equivalent to the diameter of a circular one. Example 59. Determine the size of a chimney when its height is 36 feet for a system containing 850 sq. feet of heating surface. By Formula 95, ^=, JK , ^8x850 , 9 Substituting values given, a = */ -- =- T A /. d = 8'29, say 8 inches diameter. APPENDIX HYDEAULIC MEMOEANDA 1 Imperial gallon of water 1 cubic foot of water 1 inch A column of water 1 inch square and 1 foot high A column of water 1 inch diameter and 1 foot high The capacity of a 1 foot cube The capacity of a tube 1 inch square and 1 foot long The capacity of a tube 1 inch diameter and 1 foot long The capacity of a tube 1 foot diameter and 1 foot long The capacity of a sphere 1 foot diameter 1 cubic foot sea water . 1 inch 1 Imperial gallon . 1 American J5 )> - 1 cubic foot of water 1 Imperial gallon . 1 American , . 1 cubic foot . , 1 Litre of water )t * 1 cubic meter of water = 277*274 cubic inches. = 62-37 Ib. = -036 = '434 = "34 = 6*232 Imperial gallons. 0434 034 = 4-9 = 3-263 = 64-001 Ib. = '037 = 1*2 American gallon. = '83 Imperial = 231 cubic inches. = 7'48 American gallons. = 4-543 Litres. = 3-8 = 28-375 = '22 Imperial gallon. = '264 American = 61 cubic inches. = -0353 cubic foot. = 220 Imperial gallons. = 264 American APPENDIX 463 WEIGHT OF A CUBIT FOOT OF WATER AT DIFFERENT TEMPERATURES Temp. deg. F. Weight Ib. per cub. ft. Temp, deg. F. Weight Ib. per cub. ft. Temp, deg. F. Weight Ib. per cub. ft. 32 62-42 110 61-87 190 60-31 35 62-42 115 61-81 195 60-2 40 62-42 120 61-71 200 60-08 45 62-42 125 61-65 205 59-93 50 62-41 130 61-56 210 59-82 55 62-39 135 61-47 212 59-64 60 62-37 140 61-38 220 59-58 65 62-34 145 61-29 230 59-31 70 62-31 150 61-2 240 59-03 75 62-27 155 61-1 250 58-75 80 62-23 160 60-99 260 58-46 85 62-18 165 60-84 270 58-17 90 62-13 170 60-78 280 57-88 95 62-07 175 60-66 290 57-58 100 62-02 180 60-55 300 57-26 105 61-96 185 60-43 400 53-63 WEIGHT OF A SQUARE FOOT OF DIFFERENT METALS, FROM T V INCH TO 1 INCH THICK, IN POUNDS Thickness, inch. Wrought iron. Cast iron. Steel. Copper. Zinc. Tin. Lead-. tt 2-5 2-3 2-6 2-9 2-3 2-4 3-7 $ 5-0 4-7 5-1 5-8 4-7 4-8 7-4 A 7'5 7-0 7-6 8-7 7-0 7'2 11-2 i 10-0 9-4 10-2 11-6 9-4 9*6 14-9 T 5 * 12-5 117 12-8 14-5 11-7 12-0 18-6 1 15-0 14-1 15-3 17-2 14-0 14-4 22-3 T 7 6 17-5 16-4 17-9 20-0 16-4 16-8 26-0 i 20-0 187 20-4 22-9 18-6 19-3 29-7 A 22-5 21-1 23 25-7 21-0 21-7 33-4 1 25-0 23-5 25-5 28-6 23-4 24-1 37-1 II 27-5 25-8 28-1 31-4 25-7 26-5 40-9 1 30-0 28-1 30-6 34-3 28-0 28-9 44-6 H 32-5 30-5 33-2 37-2 30-4 31-3 48-3 1 35-0 32-8 35-7 40-0 32-7 33-7 52-0 M 37-5 35-2 38-3 42-9 35-1 36-1 55-7 i 40-0 37-5 40-8 45-8 37-4 38-5 59-4 464 DOMESTIC SANITARY ENGINEERING AND PLUMBING WEIGHT OF ONE SQUAKE FOOT OF METALS New stan- dard wire gauges. No. Wrought iron. Ib. Steel. Ib. Copper. Ib. Tin. Ib. Zinc. Ib. Lead. Ib. 1 1T92 12-24 13-7 11-32 11-23 17-75 2 10-97 11-26 12-63 10-42 10-35 16-45 3 10-02 10-29 11-53 9-52 9-45 15-03 4 9-22 9-47 10-61 8-76 8-70 13-83 5 8-43 8-66 9-70 8-01 7-95 12-64 6 7 "63 7-84 9-78 7-25 7-20 11-44 7 6-86 7-04 7-90 6'52 6-48 10-29 8 6-36 6-53 7-32 6-04 6-00 9-54 9 5-72 6-13 6-58 5-43 5-40 8-58 10 5-08 5-22 5-85 4-83 4-80 7'62 11 4-61 4-73 5-31 4'38 4-35 6-91 12 4-13 4-24 475 3'92 3-89 6-20 13 3-66 3-76 4-21 3-48 3-45 5-49 14 3-18 3-26 3-66 3-02 3-00 4-77 15 2-86 2-94 3-30 272 270 4-30 16 2-54 2-60 2-92 2-41 2-40 3-81 17 2-14 2-19 2-46 2-03 2-02 3-21 18 1-91 1-96 2-20 1-81 1-80 2-86 19 1-59 1-63 1-83 1-51 1-49 2-38 20 1-43 1-47 1-64 1-36 1-35 2-14 21 1-28 1-31 1-47 1-22 1-24 1-92 22 I'll 1-14 1-28 1-05 1-04 1-66 23 95 97 1-09 90 89 1-43 24 87 89 1-00 83 82 1-30 25 79 81 91 75 74 1-18 26 71 73 82 67 67 1-06 27 65 67 75 62 62 97 28 58 60 66 55 54 87 29 54 55 62 51 50 81 30 50 51 58 47 47 75 APPENDIX 465 WEIGHT OF CAST-IRON PIPES IN LBS. PER LINEAL FOOT Bore, inches. Thickness of metal. {in. fi". iin. tin. fin. iin. Iin. Hi". 1| 4-3 6-9 9-8 13-0 ... 2 5-5 8-7 12-3 16-1 ... 3 8-0 12-4 17-1 22-2 ... 4 10-4 16-1 22-1 28-3 34-9 . 5 12-9 19-8 26-9 34-4 42-3 ... ... 6 15-3 23-4 31-9 40-6 49-7 ... - 7 27-1 36-8 46-7 56-8 ... ... ... 8 ... 30-8 41-6 52-8 64-3 ... ... ... 9 ... 34-4 46-0 58-9 71-7 ... 10 ... 51-4 65-1 79-0 93-3 ... ... 11 ... ... 56-4 71-0 86-4 101-8 ... ... 12 ... 77-3 93-7 110-4 127-4 ... 14 89-6 108-4 127-5 147-0 15 115-7 136-1 156-8 177-7 16 123-1 1447 166'6 188*7 18 137-9 161-8 186-2 210*8 The above weights are for plain pipe ends. For either a socket or a flange joint allow 1 foot of pipe. 466 DOMESTIC SANITARY ENGINEERING AND PLUMBING WIRE AND PLATE GAUGES Equivalent diameter or thickness in the fraction Equivalent diameter or thickness in the fraction of an inch. of an inch. No. No. New standard Birming- ham Ameri- can New standard Birming- ham Ameri- can wire wire wire wire wire wire gauge. gauge. gauge. gauge. gauge. gauge. 7/0 500 21 032 032 0284 6/0 464 ... 22 028 03 0253 5/0 432 ... ... 23 024 025 022 0000 400 454 46 24 022 022 02 000 372 425 409 25 02 02 018 00 348 38 365 26 018 C18 016 324 34 325 27 016 016 014 1 3 3 289 28 014 014 0122 2 276 284 257 29 013 013 on 3 252 259 229 30 012 012 01 4 232 238 204 31 on 01 009 5 212 22 182 32 0108 009 008 6 192 203 162 33 01 008 007 7 176 18 144 34 009 007 006 8 16 165 128 35 008 005 0056 9 144 148 114 36 007 004 005 10 128 134 102 37 0068 ... 0044 11 116 12 09 38 006 ... 004 12 104 109 08 39 005 0036 13 092 095 072 40 0048 0032 14 08 083 064 41 0044 15 072 072 057 42 004 16 064 065 05 43 0036 17 056 058 045 44 0032 18 048 049 04 45 0028 ... 19 04 042 036 46 0024 ... 20 036 035 032 INDEX Abyssinian tube wells, 270. Access openings for drains, 200. Action of acids on copper, 22. on lead, 5. on metals, 2. on tin, 23. on zinc, 24. Action of air on lead, 4. Action of water on boilers, 387. Adjustable boning rods, 226. Air, flow through vertical shafts, 443. heat to warm, 444. Air inlet valve for drains, 218. lift pump, 319. test for drains and other pipes, 230. Air valves, 441. automatic, 442. Air-vessels for hydraulic rams, 333. for pump delivery pipes, 315. for pump suction pipes, 310. for water pipes, 305. Alloys, 24. composition of, 25. properties of, 25. strength of, 26. Aluminium bronze, 26. Anti-D trap, 191. Anti-flooding traps, 208. Anti-siphonage pipes, 174. connections of, 148. effect of arrangement on sizes of, 179. sizes of, 178. Appliances for raising sewage, 210. for raising water, 308. Arched flue boiler for kitchen ranges, 388. Area of pipe surface, calculation of, 451. Arrangement of anti-siphonage pipes, 175. of soil pipes, 169. of waste pipes, 186. of water service pipes, 281. Artesian wells, 273. Atmospheric pressure, 308. Automatic air-valve, 442. damper regulator for boilers, 417. flushing tanks, 219. steam valve, 407. Bacterial systems of sewage purifica- tion, 248. Ball taps, 295. Earning, 21. Basement drainage, 209. Baths, 156. fireclay, 157. iron, 157. overflows for, 159. waste outlets for, 159. waste pipe for, 184. Bending pipes, 87, 95. dummies for, 91. springs for, 87. weights for, 91. Bends for drains, 199. Boilers, 387, 452. action of water on, 387. arched flue, 388. boot, 390. chimneys for, 460. connections of range, 368. Cornish, 454. deposition of lime salts in, 390, 413. dome top independent, 392. draught regulators for, 417, 457. explosion of, 422. formulae for large, 459. formulae for range, 394. heating surfaces of, 452. local currents in, 390. noises in, 421. rauge, 388. rate of firing, 460. removal of scale from, 391. sectional, 453. sizes of, 458. vertical, 457. Boiling point of water, 403, 416. Boning rods and sight rails, 224. Boreholes, 273. Box-gutters, 49. Brass, 25. Brazing, 126. British thermal unit, 402. Burnt joints, 102, 106, 132. Calorifier, 405. 467 468 INDEX Capacity of cylindrical tanks, 398, 400. of square tanks, 401. of rain-water storage tanks, 259. Cast-iron pipe formula, 365. water-tanks, 286. Cement, elastic, 122. Centrifugal pumps, 319. Cesspool or drip-box, 50. Cesspools, 243. Chambers for drains, 200. sizes of, 201. Chimneys for boilers, 460. Circuits, dipped or trapped, 376, 386, 433. Circulating head, 434. Circulation of water, 366. how reversed, 369. Cisterns, 285, 442. overflows for, 288. safes for, 287. size and capacity of, 290, 401. wash-outs for, 289. water storage, 285. w.c. flushing, 150. Closed water storage tanks, 286. Coatings for iron pipes, 20. Co-efficient, of contraction for orifices and short tubes, 345. Collapse of copper tanks, 418. Combination w.c.'s, 140. Compound water main, calculation of, 351. Compression joint for copper tubes, 109. Concrete tubes, 29. Conductivity, 2. Connections for anti-siphonage pipes, 148. for drains, 198. for flush pipes, 149. for w.c.'s, 117, 147. Connections of drains with sewers, 214. of pipes with hot-water tanks, 371. of pipes with range boilers, 368. of pipes with towel rails, 374. Connections of water service pipes with mains, 283. Constant water supplies, 280. Consumption of water, 259. Contact beds for sewage treatment, 247. Convected heat, 440. Copper-lined storage tanks, 286. ore, 22. properties of, 22. soil pipes, 168. Copper tubes, 23. heat transmitted by, 410. joints for, 108. Cornish boiler for heating apparatus, 454. Cover flashings, 57. Cylinder system of hot-water supply, 369. Cylinder - tank system of hot-water supply, 383. position of overhead tank, 385. Damper regulator for hot-water boiler, 417. Dead-weight safety valves, 423, Deep-well pumps, 311. Defects of leadwork, 31. Definition of drain, 195. of soil pipe, 167. Density, 1. Deposition of lime-salts in boilers, 390. prevention of, 413. Development of drip-box, 53. of elbow pipes, 93. of frustums of cones, 77-78. Diameter of compound main for given discharge, 352. of water pipes for given discharge, 350. Dipped or trapped circuits, 376, 386, 433. Direct heating surfaces of boilers, 452. Discharging capacity of drains, 235- 242. of short tubes, 346. of vertical air shafts, 443. of water pipes, 347, 349, 355, 357. ' Disconnecting chambers, 203. traps, 205. Domes, lead covered, 76. Domestic filters, 292. Dormers, lead covered, 59. Double acting pumps, 314. barrelled pumps, 316. Dr. Angus Smith's composition, 21. Drain flushing, 218. laying, 223. stoppers, 233. testing, 229. testing-machine, 233. track, timbering of, 228. ventilation, 216. Drainage design, 195. Drainage of basements, 209. of buildings, 193. of stables and byres, 214. Drainage plans, 196, 213. Drainers for sinks, 162. Drains, 194. bends and junctions for, 199. chambers for, 200. INDEX 469 Drains, connections for, 198. discharging capacity of, 235, 242. foundations for, 197. gradients for, 240, 242. joints for, 111, 118. manhole covers for, 204. sizes of, 194. traps for, 204. velocity of flow through, 239. Drawings, working, 91. Drip-box, 50. development of lead for, 53. overflows for, 51. outlet pipes for, 52. view of, 53. Drips for lead gutters, 47. Drop or overhead system of piping, 431. Drying rooms, heating surface required, 450. Ductility, 1. Dummies for bending pipes, 91. Earth filter for rain-water, 256. Earthenware cisterns, 285. drains, 28. Elastic cements, 122. Elasticity, 2. Equilibrium ball tap, 297. Expansion bends, 116. joints for hot-water and steam pipes, 116. joints for waste pipes, 113. Expansion of water, 443. Explorer for drain testing, 235. Explosion of boilers, 422. Fault in strata, 273. Feed cistern for heating apparatus, 442. Filters for rain-water, 256, 263. domestic, 292. Fixings for copper and iron pipes, 85. for lead pipes, 81. Flanks, 50. Float valves for hot-water tanks, 421. Flow of air through vertical ducts, 443. of metals, 2. Flow of water through drains, 239. long pipes, 347. orifices, 345. Flue type of radiator, 438. Flush pipes, 140. connections of, 149. Flushing cisterns for w.c.'s, 150. Flushing drains, 218. Flushing tanks (automatic), 219. Flushing tanks, capacity of. 223. mechanical type of, 222. plenum type of, 220. sizes of siphons for, 223. vacuum type of, 219. Fluxes, 127. Formulae for boilers, 394, 396, 459. cast-iron pipes, 365. chimneys, 461. cylindrical tanks, 400. drainage work, '237, 239, 240, 242. Formulae for flow of air through shafts, 444. for flow of water through pipes, 346, 347, 353. for heating surfaces, 449. for hydraulic rams, 335. for lead pipes, 363. for pumps, 320, 322, 326, 328. for rectangular tanks, 401. for steam heaters, 411. for water collecting surface, 258. for water pressure, 342. for water storage tanks, 259, 291. Foundations for drains, 197. Frustum of cone, development, 77, 78. Fuel, calorific value of, 458. Fusibility, 1. Galvanic action, 285. Galvanising, 21. | Gauges for preparing joints, 99. ; Gearing for pumps, 327. j German silver, 26. i Glass roofs, lead flashings for, 62. Glazes for pipes, 21. Gradient, hydraulic, 349, 352. Gradients for drains, 240, 242. | Grease traps, 206. Grenades for testing purposes, 235. : Gully trap, 189. i Gun-metal, 25-26. Gutter flashings, 58. Hard solders, 126. water, 275. Head absorbed by friction in pipes, 347, 353. Head and equivalent water pressure, 350. Head producing circulation of water, 434. Heat absorbed by walls, 445. convected, 440. emitted by pipe surfaces, 447. latent, 402. lost by glass surfaces, 446. losses, 443. movement of, 366. 470 INDEX Heat, necessary to warm air, 444. radiant, 439. transmitted by copper coils, 410. value of fuel, 458. Heating capacity of range boilers, 393. surface for drying rooms, 450-451. Heating surfaces, 435. calculation of, 443. of boilers, 452. relative value of, 439. Heating water by steam, 402. High pressure ball-cocks, 295. Hips, lead covered, 65. Hollow rolls for leadwork, 37. Hot - water apparatus (low pressure), 427. Hot- water systems (domestic supplies), 366. cylinder, 369. cylinder-tank, 383. indirect, 413. sizes of pipes, 401. steam heated, 402. tank, 366. their drawbacks when cylinders are located some distance from boilers, 376. Hot-water supplies for large buildings, 381. for small buildings, 366. for tenement buildings, 378. Hot-water tanks, sizes and capacities, 400-401. Hydraulic grade line or gradient, 349, 352. mean depth, 236. memoranda, 462. Hydraulic ram pump, 337. Hydraulic rams, 329. air-vessels for, 333 drive pipes for, 331. formulae for, 335. sizes of pipes for, 337. Hydraulic test for drains, 229. Hydraulics, 344. Hydrogen generator, 129. method of charging, 130. Hydrostatics, 340. Impurities of lead, 3. Independent boilers, 392, 452. Indirect heating systems, 413. surfaces of boilers, 452. Intermittent water supplies, 280. Intersecting roll work, 41. details of, 44. hollow rolls for, 43. solid rolls for, 42. Iron. 14. Iron, malleable cast, 19. properties of, 18. Iron drains, 194. Iron pipes, 19. joints for, 110. weight of, 465. Iron soil pipes, 168. Joints for copper pipes, 108. compression, 109. Joints for drains, 118. earthenware, 118. iron, 111. patent forms of, 120. Joints for iron pipes, 110. expansion, 113, 116. flange, 115. health water pipe, 110. high pressure, 111. rust, 112. spigot-and-socket, 111, 114. turned and bored, 114. Joints for lead pipes, 97. block, 103. burnt, 102, 106. flange, 104. lip, 104. Joints for soil pipes, 111. and branches, 98, 118. Joints for tin-lined lead pipes, 104. screwed forms, 106. soldered forms, 105. Joints for w.c.'s, 117, 139, 147. Joints, gauges for, 99. packing rings for, 115. supports and fixings for, 100. Junctions for drains, 199. Latent heat of steam, 402. Lavatories, 153. overflows for, 154. ranges of, 156. waste outlets for, 154. waste pipes for, 184. Lead burning, 128. apparatus for, 129, 133. cost of oxygen for, 134. Lead compounds, 5. properties of, 4. Lead flashings, 52. channels forms, 58. cover, 56. merits and demerits of different forms of, 52. soakers, 55. step, 55. Lead gutters, 45. box form of, 45, 49. drips for, 47. INDEX 471 Lead gutters, fall of, 45. plan of, 45. section through, 47. tapering forms of, 45. valley or flank, 51 width of, how ascertained, 46. Lead laying, 34. Lead-lined cisterns, 286. Lead ores, 3. Lead pipes, 6. formulae for, 363. joints for, 97. machine for making, 8. tests on, 363. Lead poisoning, 253. Lead soil pipes, 167. Lead, strengths for roofwork, 80. Lead traps, 11, 13. Lead-wool, 112. Leadwork on cornices, 62. on dormers, 59. on linials, 78. Leadwork on flats or platforms, 31. arrangement of rolls, 33. soldered dots for, 37. view of hollow roll- work for, 38. view of solid roll-work for, 36. Leadwork on glass roofs, 62. hips or peends and ridges, 66. stone copings, 64. torus rolls or bottles, 68. Leadwork on turret roofs, 70. details of fixings for, 74-75. shape of bay, how obtained for. 72-73. Leadwork on vertical surfaces, 75-77. Lever pumps, formulae for, 320. Levers for pumps, 324. Lift-and-force pumps, 314. formulas for, 324-328. Lift pumps, 308 formulae for, 322. Litharge, 6. Lustre, 2. Local currents in range boilers, 390. Loss of heat through glass, 446. walls, 445, 446. Loss of water from traps, 181. Low pressure ball-cocks, 295. Machine, lead burning, 129. lead pipe making, 8. lead rolling, 15-17. pipe bending, 95. tapping and drilling, 284. Malleability, 1. Malleable cast iron, 19. Manhole covers for drains, 204. Manufacture of lead pipes, 7. Manufacture of sanitary pottery, 26. of sheet lead, 13. of traps, 13. Mechanical advantage of levers, 322. of wheels, 326. Metal coverings for roofs, 30. Metals and their properties, 1-2. weight per sq. foot, 463. Methods of fixing sight rails, 225. of laying drains to given gradients, 223. Movement of heat, 366. Muntz metal, 26. Noises in boilers, 421. Non-return valve, 378. One pipe system, hot water heating, 428. Overhead or drop system, hot water heating, 431. Overflows for baths, 159. for cisterns, 288. for lavatories, 154. for sinks, 162. Packing rings for joints, 115. Patent joints for drains, 120. Percolating sewage filters, 248. Permanent hardness in water, 275. Pipes, bending of, 87. heat emitted by, 447. iron, 19. joints for, 97. lead, 7. machine for bending, 95. pitch of, for heating apparatus, 431. sizes of, for heating apparatus, 435. soil, 167. suction, 309. surface area of, 451. waste, 183. weight of cast iron, 465. Piping systems for heating apparatus, 427. Plans for drainage work, 196, 213. Plug taps, 299. Plunger pumps, 315. Poling boards, 228. Pollution of water, 251. Preparation of joints, 97. Pressure exerted by water, 340. Pressure of water and equivalent head, 350. Prevention of cylinder collapse, 421. Properties of alloys, 25. copper, 22. iron, 18. lead, 4. 472 INDEX Properties of metals, 1. steam, 402. tin, 23. zinc, 24. Protective coatings for copper, 22. iron, 20. Pumps, 308. air lift, 319. centrifugal, 319. deep well, 311. double acting, 314. double barrelled, 316. efficiency of, 320. formulae for, 320, 322, 326, 328. gearing for, 327. lift, 308. lift-and-force, 314. limiting lengths of levers for, 324. plunger, 315. power to work, 321. Purification of rain water, 256, 261. Purifying solder, 125. Radiant heat, 439. Radiator surfaces, comparative value of, 449. Radiator and towel rail connections with domestic supplies, 374. Radiators, 435. discoloration of walls by, 441. flue, 438. swinging, 439. valves for, 441. ventilating, 437. Rainfall, 257. Rain water, 254. filters for, 256, 263. separators for, 261. storage tank capacity for, 259. Rain-water drains, 213. Range boilers, 387. Ranges of lavatories, 156. w.c.'s, 147. Red lead, 5. Relief valves for hot-water tanks, 425. Resistance to the flow of water through pipes, 344. Reversed circulation, 369. Ridges, lead covered, 65. River- water, 274. Rust-pockets, 190. Safes for cisterns, etc., 287. Safety valves, 422. Sand filters for rain water, 263. Sanitary fittings, 136. pottery, manufacture of, 26. Scale in boilers, 391. Secondary circuits cylinder system, 373. Secondary circuits cylinder - tank system, 385. tank system, 369. Sectional boilers, 453. Septic tank, capacity of, 246, 249. Service pipes, arrangement of, 281. calculated diameters of, 358, 360. Sewage lifts, 210. Sewage treatment, 243. bacterial system of, 248. contact-beds for, 247. percolating filters for, 248. sub-irrigation system of, 244. Sewer connections, 214. Sheet lead, 13. Short circuiting in hot- water pipes, 430. Short pipes, head to generate velocity through, 353. Sight rails and boning rods, 224. method of fixing, 225. Sinks, 161. waste pipes for, 188. Siphon traps, 191. Siphonic latrines, 146. w.c.'s, 142. Sizes and capacitv of cylindrical tanks, 400. of square tanks, 401. Sizes of anti-siphonage pipes, 178. of boilers, 458. of chambers for drains, 201. of cisterns, 290. of pipes for domestic hot water supply, 401, of pipes for heating apparatus, 435. of pipes for hydraulic rams, 337. of rain-water storage tanks, 259. of soil pipes, 175. of waste pipes, 184, 188. Skylights, leadwork on, 62. Slate cisterns for water storage, 285. Slop sinks, 162. Smell or chemical test, 231. Smoke rocket, 234. test, 231. Soakers and their arrangement, 55. Soft water, 274. Softening water, 276. Soil pipes, 167. arrangement of, 169. thickness of, 168. Soldered dots, 36. Solders, 124. composition of, 124, 127. hard, 126. soft, 124. treatment of poisoned, 125. Solid rolls for roof work, 33. ends for, 35. INDEX 473 Sources of water supply, 254. Specific gravity of aluminium, 1. of copper, 22. of iron, 18. of lead, 4. of tin, 23. of zinc, 24. Spelter, 127. Spring taps, 301. Springs as water supplies, 266. deep seated, 267. surface or subsoil, 266. Springs for bending pipes, 87. Standing wastes for cisterns, 290. Steam apparatus for heating water, 402. properties of, 403. traps, 408. valve, automatic control, 407. Steel storage tanks, 285. Step flashings, 55. Steps in flats, 39. Stone copings, lead covered, 64. Stop-cocks, 282. Storage cisterns, 285. Strength of alloys, 26. of copper, 22. of iron, 18. of metals, 363. of pipes, 362. of steel, 19. of tin, 23. Struts for timbering trenches, 228. Suction pipes for pumps, 309. Supports for joints, 100. Systems of piping for heating ap- paratus, 427. Tables (see Contents), xiii. Tank system of hot water supply, 366. Tanks, cause of their collapsing, 418. their size and capacity, 259, 290, 400. Temporary hardness in water, 275. Tenacity, 2. Tensile strength of metals, 363. Testing appliances for drains and other pipes, 232. air gauges, 234. explorer, 235. grenades, 235. smoke machine, 233. smoke rockets, 234. stoppers, 232. Testing drains, 229. Tests on lead pipes, 363. Thickness of pipes, calculation of, 365. of soil pipes, 168. Tidal traps, 208. Timbering for trenches, 228. Tin-lined lead pipes, 9. joints for, 104. Tin ore, 23. tubes, 24. Tinned lead pipes, 8. Torus rolls, 68. Towel rail and radiator connections with secondary circuits, 374. Trapped circuits, 376, 386, 433. Traps, 190, 204. anti-D, 191. disconnecting, 205. grease, 206. gully, 189. loss of seal from, 181. siphon, 191. steam, 408. tidal or anti-flooding, 208. Treatment of sewage, 243. Trough closets, 146. Tube wells, 269. Tubes, heat transmitted by, 410. Turret roofs, 70. Two-pipe system of hot water heating, 430. Underground stop-cocks, 282. Unit, British thermal, 402. Unsealing of traps by capillary attrac- tion, 182. by evaporation, 182. by momentum, 181. by siphonage, 174, 181. by water being blown out, 183. by waving out, 182. Upland surface water, 274. Urinals, 163. Valley gutters, 50. Value of heating surfaces, 439. Valve w.c.'s, 141. Valves, air, 441. automatic steam, 407. float, 421. non-return of reflux, 378. radiator, 441. relief, 424. safety, 422. vacuum, 421. Velocity of falling bodies, 346. of flow through drains, 239. Vena contracta, 344. Ventilating radiators, 437. Ventilation of drains, 216. inlet valve for, 218. of soil and waste pipes, 177. Walings for timbering trenches, 228. Wash-outs for cisterns, 289. 474 INDEX Wash tubs, 162. Waste outlets for baths, 159. for lavatories, 154. Waste pipes, 183. bath and lavatory, 184. sink, 188. sizes of, 184, 188. Waste preventing flushing cisterns, 152. Water available from surfaces, 257. collecting area, 255. consumption, 259. expansion of, 443. Water-closets, 136. combination, 140. connections of 117, 139, 147. flushing cisterns for, 150. latrine, 146. ranges of, 147. siphonic, 142. trough, 146. wash-down, 137. wash-out, 137. valve, 141. Water fittings, 294. ball-cocks, 295. full way taps, 300. plug taps, 299. screwdown taps, 298. spring taps, 301. Water hammer in pipes, 304. Water, hardness of, 274. pollution of, 251. pressure of, 340. service pipes, 281. softening of, 276. storage cisterns, 285. Water supplies, 254. constant and intermittent, 280. rain water, 254. river, 274. spring, 266. Water, upland surface, 274. Weight of cast-iron pipes, 465. of lead for roofwork, 80. of metals, 463, 464. of water, 463. Weights for bending pipes, 91. Wells, 268. artesian, 273. borehole, 273. deep, 272. surface or subsoil, 269. tube, 270. Wheel-pump formulae, 326. White lead, 6. Wind engines for pumps, 318. Wire and plate gauges, 466. Working drawings, 91. Wrought-irou cisterns, 285 Zinc, 24. Printed by MORRISON & GIBB LIMITED, Edinburgh LIST OF BOOKS ON ARCHITECTURE, BUILDING CONSTRUCTION, ETC. THE ARCHITECTS' LIBRARY. Edited by F. M. SIMPSON, F.R.I.B.A., Professor of Architecture in the University of London. A HISTORY OF ARCHITECTURAL DEVELOPMENT. By F. M. SIMPSON, F.R.I.B.A. Three Vols. Medium 8vo. Vol. I. ANCIENT, EARLY CHRISTIAN, AND BYZANTINE. With 1 80 Illustrations. I2s. 6d. net Vol. II. MEDIAEVAL. With 257 Illustrations. 2is. net. Vol. III. RENAISSANCE. With Illustrations. [/ the Press. BUILDING CONSTRUCTION. BY BERESFORD PITE, F.R.I.B.A., F. T. BAGGALLAY, etc. Two Vols. [/ the Press. AN ENCYCLOPAEDIA OF ARCHITECTURE. By JOSEPH GWILT, F.S.A. Revised (1888), with Alterations and Considerable Additions by WYATT PAPWORTH. With 1700 Engravings. 8vo, 2 is. net. A TEXT-BOOK OF THE HISTORY OF ARCHITECTURE. By A. D. F. HAMLIN. A.M. With 229 Illustrations. Crown 8vo, 7s. 6d. THE ART OF BUILDING A HOME. A Collection of Lectures and Illustrations. By BARRY PARKER and RAYMOND UN WIN. With 68 Full-page Plates. 8vo, IDS. 6d. net. RIVINGTON'S NOTES ON BUILDING CONSTRUCTION. Medium 8vo. Part I. 695 Illustrations. IDs. 6d. net.. Part II. 496 Illustrations. IDS. 6d. net. Part III. MATERIALS. 188 Illustrations. i8s. net. Part IV. CALCULATIONS FOR BUILDING STRUCTURES. With 551 Illustrations. 135. net. ADVANCED BUILDING CONSTRUCTION. By the Author of " Rivington's Notes on Building Construction." With 385 Illustra- tions. Crown 8vo, 45. 6d. ELEMENTARY PRACTICAL BUILDING CONSTRUC- TION. Stage I. By FRANK WILLIAM BOOKER. With 727 Illus- trations. Crown 8vo, 2s. 6d. BUILDING CONSTRUCTION. By EDWARD J. BURRELL. With 303 Working Drawings. Crown 8vo, 2s. 6d. LONGMANS, GREEN, & CO., 39 PATERNOSTER Row, LONDON, E.G. BRICKLAYING AND BRICK - CUTTING. By H. W. RICHARDS. With over 200 Illustrations. Medium Svo, 35. 6d. SANITARY ENGINEERING WITH RESPECT TO WATER- SUPPLY AND SEWAGE DISPOSAL. By LEVESON FRANCIS VERNON-HARCOURT, M.A., M.Inst.C.E., Emeritus- Professor of Civil Engineering and Surveying, University College, London. With 287 Illustrations. Medium 8vo, 145. net. THE VENTILATION, HEATING, AND MANAGEMENT OF CHURCHES AND PUBLIC BUILDINGS. By J. W. THOMAS, F.I.C., F.C.S. With Diagrams. Crown 8vo, 2s. 6d. THE VENTILATION, HEATING, AND LIGHTING OF DWELLINGS. By J. W. THOMAS, F.I.C., F.C.S. With 72 Illus- trations. Crown 8vo, 6s. HEALTH AND HYGIENE. HEALTH IN THE NURSERY. By HENRY ASHBY, M.D., F.R.C.P. With 25 Illustrations. Crown 8vo, 35. net. HEALTH IN THE HOUSE. Twenty-five Lectures on Ele- mentary Physiology. By Mrs. C. M. BUCKTON. With 41 Illustrations. Crown 8vo, 2s. THE LAWS OF HEALTH. By W. H. CORFIELD, M.A., M.D. Fcap. Svo, is. 6d. ELEMENTARY PRACTICAL HYGIENE. By WILLIAM FURNEAUX. With 146 Illustrations. Crown Svo, 2s. 6d. HYGIENE. By J. LANE NOTTER and R. H. FIRTH. With 99 Illustrations. Crown Svo, 45. 6d. PRACTICAL DOMESTIC HYGIENE. By J. LANE NOTTER and R. H. FIRTH. With 84 Illustrations. Crown Svo, 2s. 6d. SCHOOL HYGIENE AND THE LAWS OF HEALTH. A Textbook for Teachers and Students in Training. By CHARLES PORTER, M.D., B.Sc., M.R.C.P. (Edin.), of the Middle Temple. With 121 Illustrations. Crown Svo, 35. 6d. WORKS BY GEORGE VIVIAN POORE, M.D. COLONIAL AND CAMP SANITATION. With u Illustra- tions. Crown Svo, 2s. net. ESSAYS ON RURAL HYGIENE. With 12 Illustrations. Crown Svo, 6s. 6d. THE DWELLING-HOUSE. With 36 Illustrations. Crown Svo, 35. 6d. THE EARTH IN RELATION TO THE PRESERVATION AND DESTRUCTION OF CONTAGIA ; together with other Papers on Sanitation. With 13 Illustrations. Crown Svo, 55. LONGMANS, GREEN, & CO., 39 PATERNOSTER Row, LONDON, E.G. UNIVERSITY 195099