cm BOILER PLANT TESTING PLANT TESTING A CRITICISM OF THE PRESENT BOILER TESTING CODES AND SUGGESTIONS FOR AN IMPROVED INTERNATIONAL CODE BY DAVID BROWNLIE )v B.Sc. (HONOURS IN CHEMISTRY) LONDON Fellow of the Chemical Society Member of the Society of Chemical Industry Member of the Society of Dyers and Colourisfs Member of the American Chemical Society Member of the American Society of Mechanical Engineers Associate Member of the Institution of Mining Engineers Member of the South Wales Institute of Engineers Associate of the Institute of Mechanical Engineers NEW YORK: E. P. DUTTON & CO. 1922 B 7 Engineering Idbrary PRINTED IN GREAT BRITAIN AT THE ABERDEEN UNIVERSITY PRESS ABERDEEN, SCOTLAND INTRODUCTION. THERE is at present no practical and definite Code in Great Britain for boiler plant testing, and, consequently, such tests are largely carried out according to the fancy of the particular engineers engaged. I am decidedly of the opinion that the time has arrived for the adoption of a standard up-to-date Code devised on thoroughly practical lines, especially in view of the urgent necessity of fuel economy, and the fact that nearly 50 per cent, of the coal consumption of Great Britain is used for the one operation of steam generation. What is supposed to be the Standard Code in Great Britain is that of the Institution of Civil Engineers (" Report of the Committee on Tabulating the Results of Steam Engine and Boiler Trials ". Revised 191 3. Published by Messrs. William Clowes & Sons Ltd., 94 Jermyn Street, London, S.W. i. Price 3/- net). The original Committee of the Institution of Civil En- gineers for this purpose was appointed on the 29th June, 1897, and they made an Interim Report on the 25th April, 1901, followed by a final Report on the 1 4th April, 1902. The Committee was then reappointed on the I9th October, 1909, to revise the original Code of 1902, and the Report of this latter Committee is embodied in the present (1913) Code. In practice, however, this Code is ignored because it is too com- plicated and unpractical. I am of the opinion also that the Code is entirely out-of- 581122 vi INTRODUCTION date, and, with all due respect to the Institution of Civil Engineers, the devising of an Improved Code is essentially the business of two branches of engineering, chemical and mechanical. The Standard Code in America-'for Boiler Plant Testing is the " Rules for Conducting Performance Tests of Power Plant Apparatus, Code 1915," of the American Society of Mechan- ical Engineers, 29 West Thirty-ninth Street, New York, U.S.A., being the Report of the Power Test Committee, which resulted from the Council's resolution of 1 3th April, 1909. As is of course well known, this Code is at present under revision by the " Power Individual Committee, No. 4 n (Messrs. E. R. Fish (Chairman), D. D. Pratt (Secretary), A D. Bailey, W. N. Best, A. A. Carey, and E. B. Ricketts), and by the courtesy of the Secretary I have been able to study the preliminary draft of the Report of this Committee, so that any criticisms and remarks of mine in this book apply to the Final Revised Code. I think that this American Code, even as revised by the " Power Individual Committee, No. 4," is still open to criticism, although much superior to the British Civil Engineers' Code. I suggest, therefore, that the time is ripe for the devising of a Standard International Boiler Testing Code by the American, British and French Engineering and Scientific Societies work- ing in collaboration. In Great Britain the premier societies concerned are the Institution of Mechanical Engineers and the Institution of Chemical Engineers, with various other societies like the Civil Engineers, Electrical Engineers, Mining Engineers, and the Society of Chemical Industry, holding a watching brief. In America the lead would presumably be taken by the American Society of Mechanical Engineers, and in France by the Ingenieurs Civils de France. INTRODUCTION vii The present book is a contribution towards the work of devising such an improved International Code, and I have divided it into the following parts : Part I. "The Results at present being obtained on Boiler Plants in General," to show the necessity of adopting modern scientific methods in steam generation, and of devising a practical international test Code to encourage such work. Part II. "Criticisms of Existing Codes and Suggestions for Improvement." This part is divided under the following heads : 1. The necessity of a separate Code for boiler plant testing. 2. The 1 object of boiler plant testing. 3. Duration of test. 4. Sampling and analysis of the fuel. 5. Flue gas analysis. 6. The method of measuring the boiler feed-water. 7. Moisture in the steam. 8. Specific heat of superheated steam. 9. Steam or power used auxiliary to the production of steam. 10. Lbs. of water from and at 212 F. per 1,000,000 B.Th.U. 11. Various minor points. 12. The method of calculating the results. All these points are matters that could be settled imme- diately by American, British and French Committees appointed to devise the International Code, and would include the pro- vision of a list of " recommended " instruments, calorimeters, water meters, combustion recorders, pyrometers, etc. Part III. " Suggestions for New Features that may be added in the future to an International Code as the result of further discussion and investigation." This chapter includes viii INTRODUCTION the following heads, and consists of various matters which may, or may not, be added to an International Code : 1. The question of the use of a special factor depending on the quality of the fuel. 2. Labour, attendance, repairs, upkeep, interest, and de- preciation. 3. Dust and grit in the chimney gases. 4. Steam meters. Part IV. " Design of a New and Improved Code as a suggested basis for the International Code," giving, as an example, the results of an actual boiler test according to the suggested Code. For convenience and simplicity, throughout this book, the fuel under discussion is coal ; but the same reasoning and principles will apply to all fuels, solid, liquid, and gaseous. The general grounds for criticism of the Institution of Civil Engineers' Code, in particular, I consider to be as follows : 1. The Code is far too academic and not adapted to practical requirements, and it appears to be drawn up with' the idea that boiler tests are a luxury only to be carried out on rare occasions. Thus it takes up pages arguing about heat balance sheets, specific heat of flue gases, and the full chemi- cal analysis of the fuel, and almost ignores matters of vital practical importance, such as the amount of auxiliary steam or power used on the plant, the price of the fuel, and the cost of evaporation of a unit of water. 2. The Code is completely out-of-date in the methods given for carrying out the test. For example, it insists on weighing or measuring the water in tanks, even at sea, and practically omits to mention the twenty different water meters available, and also does not discuss steam meters. Although it insists INTRODUCTION ix rightly on a bomb calorimeter for fuel analysis, it recommends an instrument no one in this country, except the " Civils " Committee, has ever heard of, and as regards flue gas analysis, hand methods of the most antiquated and unpractical types are insisted upon. Automatic CO 2 Recorders, a commonplace of modern boiler plant work, are disparaged, but if used, an instrument is recommended which is many years out-of-date. 3. The Code is expressed in such a confused and compli- cated manner that it rivals the Income Tax and can only be understood with great difficulty, whilst the methods of calcu- lating the results are so intricate as to be largely unintelligible without a great effort, even to specialists on the subject. Thus it is not drawn up in any logical sequence. The first sheet deals with " General Description of the Boiler," and then the second sheet follows with data from the test. We then come to " General Description of the Economiser and Superheater," and in this way general descriptive matter, data, and calcula- tions are all mixed up together in the most extraordinary manner. The attempt also to regard the boiler, economiser and superheater as entirely separate is most confusing. I have to confess that, if only because of the continual cross references, I have been compelled to buy a number of copies of the Code and cut them up with scissors, so that all the references to each point could be stuck on one large sheet of paper, and in this way to dissect the Code into a large number of separate sheets, so as to read it easily. For example, the question of auxiliary steam or power has five different references, and the particulars relating to the calcula- tions based on the full chemical analyses of the fuel and the flue-gases are hopelessly involved. The American Mechan- icals Code is infinitely superior in this respect, and is provided with an admirable index. In studying the Civils Code at great length, one is com- 6 x INTRODUCTION pelled to come to the conclusion first, that it applies only, more or less, to the years 1897-1901, the time of its original forma- tion and little alteration seems to have been made in 1913, so that it is about twenty years out-of-date and secondly, that the Committee apparently have had in mind only academic tests on small boiler plants of one boiler or so. The attempts to apply the Code to moderate sized boiler plants, and espe- cially to very large plants, like twenty " Lancashire" boilers or equivalent, prove it to be ludicrously unpractical, as I hope to show. D. BROWNLIE. 2 AUSTIN FRIARS, LONDON, E.G. 2. March, 1922. CONTENTS. PAGE INTRODUCTION .... v PART I. THE RESULTS AT PRESENT BEING OBTAINED ON BOILER PLANTS IN GENERAL . . .... . i PART II. CRITICISMS OF EXISTING CODES AND SUGGESTIONS FOR AN IMPROVED INTERNATIONAL CODE . . .' . 49 PART III. SUGGESTIONS FOR NEW FEATURES WHICH MAY BE ADDED IN THE FUTURE TO AN INTERNATIONAL CODE AS THE RESULT OF FURTHER DISCUSSION AND INVESTIGATION . .124 PART IV. DESIGN OF REPORT SHEETS FOR THE NEW CODE . . 133 SUMMARY ..... .... 153 INDEX . ,. . . . . ^ . . . . . 157 PART I. THE RESULTS AT PRESENT BEING OBTAINED ON BOILER PLANTS IN GENERAL. THE new Code suggested in this book as a basis for an Inter- national Code is the result of fifteen years' continuous experi- ence of boiler plant testing, comprising several thousand tests, during which time the methods of testing, and calcula- tion, used have been gradually altered and improved until the Code has arrived at its present form. There has not hitherto been much reliable information available as to the actual results being obtained in practice from week to week on the boiler plants of Great Britain. Much of the data obviously only applies to special test con- ditions, where everything is particularly favourable, especially as regards attention, quality of fuel used, rate of evaporation, repair of brickwork, etc., for obtaining the best results, and such data gives an entirely false impression as to the real figures that are being obtained in practice. Thus Donkin, in his book " Heat Efficiency of Steam Boilers," gives fifty tables containing the results of 425 ex- periments on different boilers. These results are summarised on page 2 (being for the boilers only, without economisers). The figures are, in my opinion, very much too high for average working conditions. It will be noted that they apply to boilers only, without economisers and superheaters, and we have extraordinary results like 72 per cent, efficiency for ten experiments on a " Lancashire " boiler only, whereas 107 experiments on a " Lancashire" boiler showed 62-4 per , Ft A NT TESTING cent, which is still a very high figure, and such as could only be obtained with the best attention. These figures, however, show the great variation being obtained, and in 107 "Lanca- shire" boiler tests, for example, we have figures from 42-1 to 79'5 per cent, efficiency. - , ^ Type of Boiler. No. of Ex- periments. Average Efficiency. Average of Two .Best Results. Worst Result. Per Cent. Per Cent. Per Cent. Water-tube, i^-in. tube 6 77*4 8 4 T 66-6 Locomotive 37 72-5 83'3 53'7 Lancashire 10 72-0 74'4 65-6 Two-storey 9 7'3 7 6-I 57-6 ji 29 69-2 79-8 55'9 Dry back 24 69-2 757 647 Return tube ii 687 81*2 56-6 Cornish . 25 68-0 817 53'o ... 9 67-0 8ro 55'o Wet back 6 66-0 69-6 62*0 Elephant . 7 65-3 70-8 58-9 Water-tube, 4-in. tubes 49 64-9 77'5 50*0 Lancashire 40 64-2 73-0 5i-9 Cornish . 3 627 65'9 6o'o Lancashire 107 62*4 79'5 42-1 Dry back . 6 61*0 73*4 54*8 Lancashire, 3-flue . 6 59'4 66-7 52-0 Elephant . 8 58'5 65*5 54'9 Lancashire 8 57'3 74'3 45'9 Vertical . 5 56-2 76-5 44-2 W. S. Hutton, in his book, "Steam Boiler Construction" , gives data for the performance of various types of boilers " which may generally be obtained in practice with boilers having tolerably clean heating surfaces, when fired with good coal," as follows : Type of Boiler. No. of Tests Given. Evaporation. Pounds of Water from and at 212 F. Lancashire 22 8-25 to 12*02 Cornish 7 7*75 11*56 Egg-ended . 6 6-52 8-56 Vertical 15 5'57 10-21 Water-tube 73 7-02 ,, 13*40 RESULTS AT PRESENT BEING OBTAINED 3 Again, some figures supplied by Molesworth : Description of Boiler. Pounds Water from and at 212 F. Evaporated Per Lb. Average Coal. Egg-ended 8-0 Cornish 8-0 Lancashire 9-0 Water- tube 8-0 Ordinary Marin 9'5 Tubular . 9'5 Locomotive 10-5 Torpedo boat 13-0 Andre states that I Ib. of the coals mentioned will evapor- ate, under ordinary practical conditions, the following amounts of water : Quality of Coal. Evaporation. Pounds of Water from and at 212 F. Gaseous coals Bituminous-fuliginous Flaming Clear burning . . Semi-bituminous . Anthracite 6-2 5 8-00 8-75 9-10 9-20 A particularly ludicrous statement, in a recent (1920) " Chemical Pocket Book," is as follows : " The efficiency of a boiler should be as near to 80 per cent, as possible, this figure being considered excellent. A more usual figure is 70 per cent, which is quite good. The water from and at 212 F. per Ib. of combustible is a good indication of the efficiency of a boiler, and under normal con- ditions should be about 12-0 Ibs." In general, and quite apart from academic absurdities, it is not realised how bad are the figures for the average boiler plant. Thus, among practical engineers, it is usual to assume that I Ib. of average coal evaporates 7 to 8 Ibs. of water from and at 212 F., and, for example, most engine builders take a figure of 8'O Ibs. in calculating the steam 4 BOILER PLANT TESTING consumption of their engines and the size of boiler plant necessary. In my experience such figures are much too high. Some of the present authorities are also of this opinion, and, for example, the figures given in Kempe ("Engineers' Year Book, 1920 ") are as below : L Description of Boiler. Small cylindrical boiler (no economisers) . . . 45 to 55 Large cylindrical boiler with economiser . . . 60 ,, 70 Small water- tube ,, ,, . . . 50 60 Large . . 7 In very good condition ....... 70 to 75 Locomotive boiler, moderately fired . . . . 65 ,, 70 Marine boiler, well fired ...... 60 ,, 70 Most of the exaggerated opinions as to the figures for the performance of steam boiler plant seem to be due to the fact that it is not realised how important is the quality of the fuel. The performance figures of any boiler plant depend largely on this point. Thus, as I will discuss in detail later, if a coal of 12,000 B.Th.U. deteriorates, say 10 per cent., to 10,800 B.Th.U., the loss in evaporation in practice is much more than 10 per cent. The difficulties caused by the accumulation of ash, and the reduction of radiant heat causes a much greater drop in evaporation than that corresponding to the mere reduc- tion in heating value, and the net reduction is, say, 15 per cent, to I7-J per cent, instead of 10 per cent. In the same way, an increase in heating value of 10 per cent, say from 12,000 B.Th.U. to 13,200 B.Th.U., gives more than the mere 10 per cent, increase in heating value, because of the greater freedom from ash and the increased radiant heat. These facts must, therefore, be borne in mind in con- sidering the figures given on the different types of plants. For example, Bryan Donkin and A. B. W. Kennedy pub- lished, in 1897 (" Experiments on Steam Boilers," Offices of " Engineering "), the results of a series of experiments on the performance of a number of types of steam boilers comprising twenty-one separate determinations, and including " Vertical," " Tubular," " Lancashire," " Locomotive," " Water-tube," RESULTS AT PRESENT BEING OBTAINED 5 " Cornish," "Cornish Multi-tubular " and "Elephant" boilers. These experiments, however, do not give us much data of practical value, because they were carried out under ideal con- ditions with coal of extraordinarily good quality, namely, the finest Welsh steam coal. Thus the heating value of the dried coal was no less than 15,560 B.Th.U. per pound, with only 3 per cent, ash, and the theoretical evaporation of I Ib. of the dried coal from and at 212 F. was given as i6'i Ibs. of water. As typical of the figures obtained, a "Lancashire" boiler gave 70*4 per cent, efficiency without economisers, and 9-92 Ibs. of water evaporation per pound of coal. A " Cornish " boiler showed 78*3 per cent, efficiency with 1 1 '4 Ibs. of water from and at 212 F. per pound of coal, whilst a "water-tube" boiler showed 74-4 per cent, efficiency and 9-90 Ibs. of water per pound of coal. These results obviously only apply to purely abnormal conditions, and are of little use in arriving at average figures for boiler plants as generally working in industry. The firm with whom I am associated have been engaged continuously for the past dozen years or so in carrying out complete scientific investigations into the working of steam boiler plants in Great Britain, and also in reorganising existing plants or erecting new plants. From 1908 to the present time we have tested nearly 500 different boiler plants, with a total annual coal bill of about 4,000,000 tons, and made a personal examination (without testing) of about 2000 plants with a total coal bill of about 15,000,000 tons per annum. I have at the present time the tabulated results of the complete scientific investigation of 400 different boiler plants, wish a total coal bill of 3,250,000 tons per annum, the number of boilers being 1513, in forty-one different industries, as detailed on page 6. Before discussing the results obtained, I should like to give a short account of the methods used in carrying out the tests, which were in each case of the most comprehensive character, and not a mere inspection and expression of opinion 6 BOILER PLANT TESTING in the sense used by certain Fuel Economy Associations or Organisations. A detailed test of the plant was made first for one work- ing day, and a further test for one week was carried out as a Number of Industry. Plants Tested. Aniline dye manufacturers 4 Breweries 5 Calico printers 10 Carpet manufacturers 6 Cement ,, i Collieries (including several steel works) . .112 Cotton bleaching 9 mills 23 ,, piece dyeing 18 ,, yarn 9 Dyeing and cleaning 8 Electricity station i Engineering 10 Explosives 2 7 Fine organic chemicals . . . ... . 3 Flour mills 4 Food products 3 Glue manufacturers 8 Hat 3 Heavy inorganic chemicals 7 Hosiery mills 5 dyeing 4 Hospitals J India-rubber manufacturers 5 Jute mills * Lace 7 ,, bleaching 3 Laundries . . . . . * Linen mills . 2 Paint manufacturers i Paper mills .... V ... 4 1 Potteries I ... 4 Pumping station . i Residential mansions i Soap manufacturers 3 Silk dyeing and printing 2 Tanneries 7 Textile, special ! Woollen mills . ' < 36 yarn dyeing i Total . . . 400 check on the figures for the fuel used and the water evaporated, and to ascertain the weekly conditions, including all week-end losses due to stoppage of plant, etc. It will be clearly understood that the object of these tests RESULTS AT PRESENT BEING OBTAINED 7 was to find out the exact normal everyday working conditions of the plant, particularly as regards efficiency and the cost in coal for the production of a unit of steam (evaporation of 1000 gallcfns of water), so that a scheme of reorganisation could be devised for the more economical production of steam that is to say these figures represent the true performance of the plants as run from week to week, and the boiler-house staff worked the plant as usual. In carrying out the tests, the water level in the boilers and the general condition of the fires was the same at the end as at the commencement of the test. With regard to the details of the carrying out of the tests we have : Weight of Fuel Used. This was determined in the usual various ways, depending on the circumstances of the case, from weighing the fuel in barrows or bags, to weighing carts and railway trucks direct. Analysis of Fuel. As is well known, it is an extremely difficult matter to get a thoroughly average sample of coal, especially when there is variation in the quality, and we took the greatest care in this respect. Samples were taken every half-hour by the clock, and also every hour and placed in separate receptacles. At the end of the trial the accumulated hourly and half-hourly samples were broken up, thoroughly mixed, quartered, and so on, as usual, several pounds being finally taken in sealed tins for analysis. There were thus obtained two entirely independent samples of the same fuel. In the same way we took samples of coal every day during the long check test as a check on the figures for the day. The two days' samples were then analysed separately, and the average figures taken. If more than one quality of fuel was used on the test, each separate quality had two independent samples taken as above described, so that in some cases on a day's trial as many as six different samples were analysed and the results averaged. As regards the analysis of the coal, the heating value was determined by means of an oxygen bomb calorimeter 8 BOILER PLANT TESTING (" Mahler-Donkin "). The bomb type of calorimeter is, of course, acknowledged to be the standard scientific instrument for accurate heat determination, and the method consisted in burning the damp fuel (as fired) in a platinum-lined gunmetal bomb in oxygen at about 450 Ibs. pressure per sq. in. The ignition is made inside the "bomb by means of a fine platinum wire heated by outside electric contact. If platinum wire was not obtainable, fine iron wire was used, and a correction made for the heat of combustion of the iron to iron oxide. The bomb is covered with a known amount of water at a known temperature, and after ignition the rise in temperature observed with a thermometer graduated to T ^V C. As the combustion is instantaneous and totally enclosed, the heating value is obtained direct without any corrections being required. Further, the combustion is always complete, which is not the case with many inferior types of calorimeter. As I will discuss later in detail, there is much confusion with regard to the method of expressing the results. In a bomb calorimeter of this description the moisture in the coal, and the moisture formed by the combustion of the (organic) hydrogen in the coal, are driven off as steam and condensed again in the bomb, so that the whole of the heat is retained and included in the heating value. In actual practice, how- ever, when the coal is thrown into the furnace, the water is driven off as steam and tends to escape, and to pass away in the chimney base at the same temperature as the gases, say, at an average of 400 F. Thus the gross heating value as obtained by the bomb is slightly higher than the real heating value that would be obtained in practice because of this loss of heat due to steam escaping in the gases. In view of these difficulties I have, in each case, taken the higher or gross heating value, that is, the heating value as actually obtained by the combustion of the damp coal in the bomb calorimeter. With regard to the percentage of ash, this was determined by complete combustion in the muffle furnace, so that the RESULTS AT PRESENT BEING OBTAINED 9 figures given for the amount of ash represent the real non- combustible matter, not allowing for any slight loss that might take place by the volatilisation of a portion of the ash. The percentage of ash obtained from the boiler plant will, of course, always in practice be slightly higher, as it is impossible to obtain complete combustion of the ashes, and under good average conditions there is always, say, I to 2 per cent, of combustible matter in the ash as thrown away on the ash dumps. I may say, however, that in scores of boiler plants, there is a big loss in this direction, and it is quite possible to find as much as 5 per cent, or even up to 10 per cent, of combustible matter still retained in the ash. I have actually come across several instances where one works has bought ashes from another works for road-making, etc., and then used these ashes as fuel under the boilers. Water Evaporated. Various methods were used, de- pending on circumstances, to measure the water evaporated, but generally the method adopted was the use of a well-known make of pressure type hot-water meter calibrated before each test, and working between the boiler feed pump and the economisers. Moisture in Steam. Another difficulty, which I will discuss in detail later, in boiler plant testing is the fact that non-superheated steam always contains a certain amount of moisture, varying from o to 5 per cent, apart from priming, that is water from the boiler plant which has not been con- verted to steam. Theoretically, therefore, the amount of moisture in the steam should be determined and this amount deducted (with suitable temperature allowances) from the evaporation in calculating the true performance of the boiler plant The practical difficulties in the way of this are, however, very great, and on this account, most boiler tests are carried out without determining the moisture in the steam, and we have followed the general practice in this respect. Analysis of Feed- Water. Samples of the feed-water io BOILER PLANT TESTING were taken every half-hour during the day's test, and good average samples obtained in this way. The samples were analysed by the " Wanklyn " soap test method before and after boiling, giving the permanent and temporary hardness, although in exceptional cases the lull scientific analysis of the feed-water was carried out in the usual lines to determine the parts per 100,000 of calcium and magnesium bicarbonate and sulphate, including the total solids, etc. I may say that in my experience the scap test is a very convenient method, although somewhat despised by most chemists. Temperature of Feed-Water. On the day's test, the temperature of the feed-water, before and after the economisers, was taken every half-hour, or oftener if the variation was con- siderable. For this purpose we generally used calibrated mercurial thermometers, as the ordinary economiser ther- mometers supplied with economisers are not very accurate after months of work. Draught Measurement. On the day's test the draught was taken every half-hour, in various convenient positions on the boiler plant, such as the side flues of " Lancashire " boilers, the main flues, chimney base, before and after the economisers, and other similar positions. A |-in. wrought-iron pipe was inserted into the middle of the flues in each case, and the draught taken with an ordinary draught gauge of the glass " U " tube type. Temperature of the Flue Gases. For this determination a well-known make of pyrometer was used, of the thermo- electric type, with platinum-iridium and platinum junction, in a porcelain tube enclosed in a steel tube, and connected through standardised coils of wire to a voltmeter graduated direct from 212 to 1500 F. The instrument was arranged with a two-way switch, two coils of wire, and pyrometers in the flue before and after the economiser, so that the two read- ings could be taken practically simultaneously. Before each test the instruments were calibrated by determining the boil- ing-point of some substance of high boiling-point, such as RESULTS AT PRESENT BEING OBTAINED 11 aniline, using for the purpose a special form of iron condenser. As usual, on the day test the readings were taken every half- hour. Percentage of CO 2 in Flue Gases. In each case some reputable make of Combustion Recorder was fitted on the plant, and the percentage of CO 2 recorded at the rate of about 20 analyses per hour on a chart, along with the time of the analysis. One week's run was taken in this way in each case. Samples of the gas were taken from the side flues or the downtake of each boiler in rotation, allowing as a rule about twelve hours (say 240 analyses) on each boiler, so that at the end of a week a very good average was obtained. We also in many cases fitted up an apparatus for taking a large volume of the flue gases, say I 5 to 20,000 c.c., very slowly during the whole day's test, so that at the end of the trial this 1 5 to 20,000 c.c. represented fairly the average during the day. The apparatus was then taken to our laboratories and the contained gas analysed by means of the " Orsat " apparatus to determine the percentage of CO (carbon monoxide) after the CO 2 and the oxygen had been absorbed in the usual way. Steam Pressure. On the day's test the steam pressure was taken every half-hour, or oftener if there was a consider- fluctuation. Temperature of Superheated Steam. On the day's test the temperature of the superheated steam was taken every half-hour with a calibrated mercurial thermometer, as, like economiser thermometers, the thermometers in continuous use with superheaters are seldom very accurate. Auxiliary Steam, With regard to the steam or power used as auxiliary to the production of steam, in the case of an engine, the indicated horse-power of the engine was taken, and from the type of engine in use a very good idea was obtained of the amount of steam used. Thus, for the ordinary enclosed, forced-lubrication, high-speed engine for driving forced or induced draught fans, an average figure is 35 Ibs. of steam per indicated horse-power. In the case of a motor drive the 12 BOILER PLANT TESTING power used is, of course, very easily determined by measuring the amps, and the volts. The real difficulty is in connection with the amount of steam used in the form of steam jets either under or over fire-bars, and for this we used the special apparatus described later (p. 107). * Unfortunately, I cannot give the detailed figures for the performance of each of the 400 plants, since the amount of space occupied by such a mass of figures would be outside the limits of this book. In the first place, however, the true average net working efficiency of the whole of the 400 tests ?s 58 per cent., and I feel sure that these can be taken as thoroughly representative of the boiler plants of Great Britain, since, in addition to this large numbers of tests carried out in many different industries, the 2000 plants we have inspected, representing, as already stated, a coal bill of 15,000,000 tons per annum, were all found to be working on the same general lines. The figures can be divided as follows, the highest individual plant being 82-21 per cent, and the lowest 32*50 per cent: Net Working Efficiency. No. of (Boilers, Economisers and Superheaters.) Plants. i. 80 pe 2. 75 , 3- 7 > r cer it. and ov er . . . . . . . 2 . ; 9 4- 65 , 5. 60 , 6. 55 , 7 50 i > i 58 69 96 80 8. Below 50 per cent. 69 Total . . . 400 I gave ("Coal Saving by the Scientific Control of Steam Boiler Plants," "Engineering," I2th and ipth July, 1918) the true average figures for 250 of these plants (1000 boilers and 2,166,000 tons of coal per annum) as follows : RESULTS AT PRESENT BEING OBTAINED 13 AVERAGE OF 250 TYPICAL BOILER PLANTS. A. Working Day's Test ; Type of boiler . . . . . "Lancashire" Number of boilers . . . . . .4 Grate area 152-6 sq. ft. Duration of test . . . . ... 9-43 hours Amount of fuel used 30,13172 Ibs. Analysis of fuel used : British thermal units ..... 11,822 Ash 11-5 per cent. Fuel burnt per boiler per hour .... 798-8 Ibs. ,, square foot of grate area per hour 20*9 Water evaporated 197.776 Ibs. ,, ,, per boiler per hour . . . 5243 ,, ,, ,, square foot grate area per hour .... 137-4 Ibs. lb. fuel .... 6-56 ,, Equivalent evaporation from and at 212 F. per lb. fuel 7-46 Ibs. Equivalent evaporation from and at 212 F. per 1,000,000 British Thermal Units . . . 631-0 Ibs. Temperature of feed-water before economiser . 116 F. after . 193 Percentage of fuel bill saved by ,, . 7*1 per cent. Draught in side flue 0-40 in. W.G. ,, ,, main flue at exit of economisers or chimney base .... o'8o ,, ,, Temperature of flue gases before economisers . 598 F. ,, after . 478 Percentage of CO 2 in flue gases from side flue of boilers by means of combustion recorder . . 7-6 per cent. Steam pressure : Lbs. per square inch (gauge) . . . 89 ,, ,, ,, (absolute) . . . 104 Temperature of saturation of steam . . . 330-5 F. ,, ,, superheated steam , . . 346*5 ,, Steam or power used as auxiliary to the produc- tion of steam . ... . . . 2 -4 per cent. Thermal efficiency : (a) Net working efficiency complete, after deducting 2-4 per cent, steam or power used as auxiliary to the production of steam 60-09 per cent. (b) Boilers only ' . 56*71 ,, ,, (c) Economisers only 4-35 ,, ,, (d) Superheaters ,, . . ... 0-51 ,, ,, B. Long Check Test (one week of 7 days) : Duration . . . . ... . 167-5 hours Amount of fuel used . . . . ' - . 128-25 tons Water evaporated ...... 184,435-0 galls. ,, ,, per lb. coal . . . . 6-42 Ibs. Taking separate industries, I gave in "Engineering" (25th July and 1st August, 1919, "Exact Data on the Run- ning of Steam Boiler Plants, No. 2. The Performance of BOILER PLANT TESTING Colliery Steam Boiler Plants") the detailed figures for the tests of 100 different colliery boiler plants, representing 570 boilers with an annual coal bill of 1,250,000 tons. These colliery plants were situated as follows : Lancashire . . ' . . * Derbyshire . Gloucestershire . Notts .... Shropshire . Scotland South Wales Staffs .... Yorkshire . Warwickshire 14 8 i 7 i 26 The true average figures for these 100 tests are given below : A, Working Day's Test : Duration of test g 68 hrs. Type of boiler . Chiefly " Lane." Number of boilers Average 5*7 Grate area 217*6 sq. ft. Number of tubes in economiser .... Average 50. Analysis of coal used : B.Th.U 10,500 Ash 15-5 per cent. Amount of coal burnt . . . ,.* ,/..*- . . 39,815 Ibs. Coal burnt per boiler per hour . . ~~~ . . 721-5 ,, ,, ,, ,, sq. ft. grate area per hour . . 18-9 ,, Draught : Chimney base '.." . . . . . 0-95 ins. W.G. Side flues . o'6o ,, ,, Temperature of flue gases : Before economiser . ... . ' . . 690 F. After f . 660 Percentage of CO 2 . . * . * 7*5 per cent. Total water evaporated 21,693 gals. Water evaporated per boiler per hour .. " . . 393 gals. Temperature of feed-water : Before economiser . 137 F. After . 154 Per cent, saving due to economisers . . . i'6 per cent. Steam pressure gauge 86 Ibs. Temperature of saturated steam 328 '4 F. ,, superheated steam . . . . 350-0 ,, Steam or power used as auxiliary to the production of steam , , i -8 per cent. Lbs. water per Ib. coal 5-44 Ibs. ,, from and at 212 F. per Ib. coal . . 6*07 ,, ,, 1,000,000 B.Th.U. 578-1 Efficiency : (a) Net working 55'52 per cent. (b) Boilers only 55*2 ,, (c) Economisers only 0-90 ,, (d) Superheaters ,, 0-62 ,, ,, RESULTS AT PRESENT BEING OBTAINED 15 B. Long Check Test (one week of 7 days) : Duration of test 168-00 hrs. Amount of coal burnt . . 235-2 tons ,, water evaporated ...... 273,990 gals. Water evaporated per Ib. of coal . . , . . . 5'ig Ibs. Approximate annual coal bill . . . . . 11,800 tons The figures can be divided as follows : Net Working Efficiency. No of (Boilers, Economises and Superheaters.) Plants. 1. 80 per cent, and over o 2. 75 ,,. . . . . o 3. 70 .. .. i 4- 65 ..... 4 5. 60 ..... 9 6. 55 .* > 2 9 7. 50 ,,,,,, 8. Below 50 per cent. ..... 35 Total .... 100 The average net working efficiency of 55*52 per cent is the lowest of any industry in the country, although not so low in comparison as is generally supposed. One reason why the efficiency of colliery boiler plants in general is lower than the average, is that there is still at work a number of the grossly inefficient "egg-ended" boilers. I will deal later with the efficiency of different boilers, but out of the 570 boilers in these 100 colliery tests, 37 boilers were of the "egg-ended" variety, 500 being "Lancashire," 31 " water-tube " and 2 special boilers. The second reason is that coal at the colliery, until a few years ago, was so cheap as hardly to be worth saving, and consequently the boiler plant was regarded as even of less importance than usual, whilst also to some extent inferior coal is used. Also in the "Chemical Trade Journal" of August and September, 1920 ("Coal Saving in the Chemical Industry"), I published the detailed figures for the tests of sixty different chemical works' boiler plants, representing 236 boilers with an annual coal bill of 620,000 tons per annum. The true average figures for these sixty tests are as follows : 16 BOILER PLANT TESTING A. Working Day's Test : Duration of test 9-2 hrs. Type of boiler Chiefly " Lane." Number of boilers Average 3*9 Grate area 156*5 sq. ft. Number of tubes in economiser Average 300 Analysis of coal used :-^- B.Th.U. . . . * . . f* . . . . 11,350 Ash . . .12 per cent. Amount of coal burnt 38,025 Ibs. Coal burnt per boiler per hour 1059-8 ,, sq. ft. grate area per hour . . . 26-6 Draught : Chimney base 070 in. W.G. Side flues 0-38 Temperature of flue gases : - Before economiser 650 F. After 450 Percentage of CO 2 8*0 per cent. Total water evaporated 2 3>358 gals. Water evaporated per boiler per hour .... 651 ,, Temperature of feed- water : Before economiser 103 F. After ....... 215 Per cent, saving due to economisers . . . 10-1 per cent. Steam pressure (gauge) 102 Ibs. Temperature of saturated steam . . . . . . . 339'2 F. superheated steam .... 384-2 Steam or power used as auxiliary to the production of steam 6-1 per cent. Lbs. water per Ib. coal 6-1 Ibs. ,, ,, from and at 212 F. per Ib. coal . . . 7-0 ,, ,, ,, ,, ,, 1,000,000 B.Th.U. 616-7 ,, Efficiency : (a) Net working 57-9 per cent. (b) Boilers only 54-2 ., (c) Economisers only . . . ... 6'i ,, ,, (d) Superheaters . . . . '. . 1-3 ,, B. Long Check Test (one week of 7 days) : Duration of test (hours) ^4-5 Amount of coal burnt (tons) . . . . . 145*62 Amount of water evaporated (gallons) .... 18918-95 Water evaporated per Ib. coal 5-89 Approximate annual coal bill 10,596 tons These figures for the chemical industry can be divided as follows : Net Working Efficiency. No. of (Boilers, Economisers, and Superheaters). Plants. 1. 80 per cent, and over o 2. 75 ,, 4 3- 70 3 4- 65 12 5- 60 9 6. 55 13 7- 50 9 8. Below 50 per cent 10 Total . 60 RESULTS AT PRESENT BEING OBTAINED 17 Taking again another group of allied industries, that of dyeing, bleaching, calico-printing, finishing, and dyeing and cleaning, in the " Textile Manufacturer" of August to March, 1921 (" Coal Saving by Modern Methods of Steam Genera- tion "), I give the detailed results of sixty-five boiler plants in these industries, representing 217 boilers with an annual coal bill of 275,637 tons per annum. The true average figures for these sixty-five tests are as follows : A. Working Day's Test : Duration of test .... ... 8*30 hrs. Type of boiler Mostly " Lane." Number of boilers ...... Average 3*3 Grate area 118-60 sq. ft. Number of tubes in economiser . . . Not stated Analysis of coal used : B.Th.U. 11,950 Ash 10-5 per cent. Amount of coal burnt 20,776 Ibs. Coal burnt per boiler per hour .... 758-5 ,, ,, sq. ft. grate area per hour . . 21*17 ,, Draught : Chimney base . . . . . . 0*83 in. W.G. Side flues 039 ,, Temperature of flue gases before economiser . 610 F. ,, after . 460 Percentage of C( 2 7-8 per cent. Total water evaporated 13,829-9 gals. Water evaporated per boiler per hour . . 504-8 ,, Temperature of feed-water : Before economiser ...... 112-5 F - After 202-5 ,, Percentage saving due to economisers . . 8-3 per cent. Steam pressure (gauge) . . ... 70 Ibs. Temperature of saturating steam . . . 316-5 F. ,, ,, superheated steam . . . 355-0 Steam or power used auxiliary to the production of steam 2*0 per cent. Lbs. water per Ib. coal ..... 6-65 Ibs. ,, ,, from and at 212 per 1,000,000 B.Th.U. . . . . . 7-60 Efficiency : (a) Networking 61-41 per cent. (b) Boilers only 56-33 ,, (c) Economisers only . . . . . 5*15 ,, (d) Superheaters . . . ." ' n8 B. Long Check Test (one week of 7 days) : Duration of test 163-93 hrs. Amount of coal burnt 83-82 tons ,, ,, water evaporated .... 167,717 gals. Water evaporated per Ib. coal .... 6-5 Ibs. Approximate annual coal bill .... 4240 tons i8 BOILER PLANT TESTING The figures for the dyeing, bleaching, calico-printing, finishing, and dyeing and cleaning industries can be divided as follows : Net Working Efficiency. No. of (Boilers, Economisers and Superheaters.) Plants. 1. 80 per cent, and over i 2. 75 "'-' . . > .... o 3- 7<> ,, ....,... 4 4- 6 5 M . . . . 12 5- G O ... . 1 . 14 6. 55 ,, ,. ,, - ~ . . . . 12 7- 5 it ii ii n . . It ,. 16 8. Below 50 per cent. . .... . . 6 Total . . . . 65^ Finally, I gave in the Annual Number (1921) of the " Papermaker " (March, 1922), the figures for the paper- making industry (" Coal Saving in the Papermaking Industry by the Scientific Control of Steam Boiler Plants"), represent- ing forty boiler plants with 1 1 2 boilers and an annual coal bill of 291,145 tons per annum. The true average figures for these forty tests are as follows : A. Working Day's Test : Duration of test . 11*12 hrs. Type of boiler . . ..... . . Mostly " Lane." Number of boilers . . *. . ' . . Average 2-8 Grate area . . . . . . 102-87 sq. ft. Number of tubes in economiser . ; . . . 216 Analysis of coal used : B.Th.U . . . n,53o Ash . . 12*75 P er cent. Amount of coal used 236,005 Ibs. Coal burnt per boiler per hour 757'9 ,, ,, ,, ,, sq. ft. grate area per hour . . . 20-6 ,, Draught : Chimney base - . . . 0*85 in. W.G. Side flues . . . ..~ . ,~ . . . 0-45 ,, Temperature of flue gases : Before econcmiser . . .... . 600 F. After ...... . . 414 F. Percentage of CO 2 . . , .. ... 8'o per cent. Total water evaporated . . ' . . . . 16,873-4 gals. Water evaporated per boiler per hour . , , . 54 I- 9 ,, Temperature of feed-water : Before economiser . ; . . - . .. . 143 F. After . . * - ". . . 270 Percentage saving due to economiser . .""j . . 11*9 per cent. Steam pressure (gauge) . v . !,_ . . 104 Ibs. Temperature of saturated steam . . . ; , _ . 340*6 F. ,, superheated steam . . . 360-0 Steam or power used auxiliary to production of steam . 3-75 per cent. Lbs. water per Ib. coal . . . ... 7-15 Ibs. ,, ,, from and at 212 F. per Ib. coal . . 7-95 ,, 1,000,000 B.Th.U. 689-3 RESULTS AT PRESENT BEING OBTAINED 19 Efficiency : (a) Net working . . . ... . 65-07 per cent. (b) Boilers only . . . . . . . 58-91 ,, (c) Economisers only . . < . . . S-oi ,, ,, (d) Superheaters . . . . . . 0-69 ,, B. Long Check Test : Duration of test . . 158-8 hrs. Amount of coal burnt 125-61 tons water evaporated 189,124 gals. Water evaporated per Ib. coal . . . . . 672 Ibs. Approximate annual coal bill ..... 7278 tons The figures for the papermaking industry can be divided as follows : Net Working Efficiency. No. of (Boilers, Economisers and Superheaters.) Plants. 1. 80 per cent, and over o 2. 75 i 3- 70 4- 65 5. 60 i 6 7 10 8 6. 55 7- 50 8. Below 50 per cent. ...... 7 Total . . . 4p_ I am also tabulating similar figures for various other in- dustries, particularly cotton and woollen manufacture, and the general results are almost identical with the four industries already given. The striking fact, as will be seen, is that, in averages, in- dividual boiler plants are working at all kinds of efficiencies, actually from 32 to 82 per cent, whilst the average for all the 400 boiler plants is 58 per cent, and the average for in- dividual industries may vary from 55 to 65 per cent The results being obtained in general can, I think, be conveniently expressed by a series of tables which I gave in " Engineering," loth to 1 7th December, 1920 ("Exact Data on the Perform- ance of Steam Boiler Plants, No. 4. Average Figures for the Performance of Some Different Types of Steam Boiler Plant "). In these tables I have given for the various types of boiler in use, first of all as much the most important the figures being obtained to-day under average conditions without any proper methods of testing and control, applying to at least 20 BOILER PLANT TESTING 85 per cent, of the boiler plants of the country. I have also given in the case of " Lancashire " and " Water-tube " boilers, corresponding figures for plants run on the most modern lines, which apply only to probably about 5 per cent, of boiler plants, and at the same time I have given figures for very bad plants, probably typical of about 10 per cent, of the plants of the country. In order to present comparative figures, the coal used throughout has been calculated as 12,000 B.Th.U. per pound gross, and 10-5 per cent, ash, which represents roughly the average quality, or slightly above the average, used through- out the country. Also, for the purpose of comparison, I have taken an average price of 403. per ton, and calculated the day's test and the week's test in pence per 1000 gallons, together with a coal bill for a standard evaporation of 20,000,000 gallons of water. Further, in taking the temperature of the feed-water (as going into the plant) an average figure of 110 E. has been taken throughout, because this is about the usual figure, being the ordinary " hot well " temperature. In studying the figures, therefore, if the temperature of the inlet water of a particular plant is different, every 11 F. can be taken as equivalent to I per cent, of the coal consumption, higher or lower. Thus, if the feed-water be about, say, 99 F. then the coal consump- tion figures will be increased about I per cent. LANCASHIRE BOILER PLANT. The adjoining table gives what I consider to be the average figures for the performance of the " Lancashire " boiler complete with all accessories, calculated in terms of standard 30 x 8 ft. boilers with average grates 6 x 3 ft. 2 in. A plant of four boilers is given because this is about the average size and, of course, for a different number of boilers the corresponding figures can easily be calculated. As regards the method of firing, hand and mechanical firing are averaged together, as there is, in my opinion, little RESULTS AT PRESENT BEING OBTAINED 21 LANCASHIRE BOILER PLANT. Bad Plant Representing, say, 10 Per Cent, of Boiler Plants at Work in Great Britain. Ordinary Aver- age Plant as being Generally Woiked To-day, Representing 85 Per Cent, of Plants at Work in Great Britain Most Efficient Plant Working under Modern Scientific Supervision or under Test Conditions, Representing only about 5 Per Cent, of Plants at Work in G eat Britain. A. WORKING-DAY TEST. i. Number of boilers working . 2. Grate area (total) 4 151-96 sq. ft. 4 151*96 sq. ft. 4 151-96 sq. ft. 4. Price of coal used (per ton delivered) . 5. Amount of coal used 6. Analysis of coal B.Th.U 40s. 36,249 Ibs. I2.0CO 403. 41,504 Ibs. 12,000 4os. 50,870 Ibs. 12,000 8. Coal burned per boiler per hour . 9. Coal burned per sq. ft. grate area per hour . 10. Water evaporated, Ibs 1 1. Water evaporated per boiler per hour . 12. Water evaporated per sq. ft. grate area per hour 13. Water evaporated per Ib. of coal . 14. Equivalent evaporation from and at 212 F. 755*2 Ibs. 19-8 Ibs. 204,000 Ibs. 4250 Ibs. in'3 Ibs. 5-62 Ibs. 6*42 Ibs 864-7 Ibs. 22*7 Ibs. 276,000 Ibs 575 Ibs. 151-3 Ibs. 6-65 Ibs. 7-62 Ibs. 1059-8 Ibs. 27 9 Ibs. 408,000 Ibs. 8500 Ibs. 2237 Ibs. 8-02 Ibs. 9*28 Ibs 15. Equivalent evaporation from and at 212 F. per 1,000,000 B.Th.U 16. Temperature of fe;d-water before econo- 535'i Ibs. 110 F. 635-0 Ibs. 110 F 856-7 Ibs 110 F 17. Temperature of feed-water after economisers 1 8. Percentage of coal bill saved by economisers 19. Draught in back flues of boilers . . . 20. Draught in chimney base . . 21. Temperature of flue gases before economisers 22. Number of economiser tubes 23. Temperature of flue gases after economisers 24. Analysis of boiler feed-water Degrees permanent No economiser Nil 0-40 in. W.G. 0*50 in. W.G. 500 F. No economiser 500 F. 12 cO 230 F. i I'D per cent. 0-35 in. W.G. 075 in. W.G. 600 F. About 320 tubes 450 F. 9 335 F. 2O'4 per cent. 065 in. W.G. 2-00 in. W.G. 650 F. About 500 tubes 310 F. 5 25. Percentage COg in flue gases (continuous record on combustion recorder) . 26. Steam pressure (average) (a) Gauge . 27. Steam pressure (average) (b) Absolute 28. Temperature of saturation of steam 29. Temperature of superheated steam 30. Steam or power used as auxiliary to produc- 5-0 per cent. 60 Ibs. 75 Ibs. 307*4 F. None 7-5 per cent. 75 Ibs. 90 Ibs. - 320-3 F. None i2'o per cent. 159 Ibs. 174 Ibs. 37 'c F. 540 F. Thermal efficiency of plant 31. (a) Net working efficiency of plant complete 32. (b) Boilers only ...... 49-2 per cent. 51*8 per cent 60 -o per cent. 54-7 per cent. 79 - o per cent. 59'5 P er cent. Nil 34. (d) Superheaters only .... 35. Cost in coal to evaporate 1000 gals, of water B. LONG CHECK TEST (ONE WEEK). (Say Two Shifts per 24 Hours.) 36. Duration ........ 37. Price of coal used (per ton delivered) . 38. Amount of coal used ... . Nil 38o-8d. i68'o hours 403. Nil 332'6d. i68'o hours 4os. 6 2 per cent. 266 -gd. 168-0 hours 403. 39. Water evaporated 40. Water evaporated per Ib. of coal . 41. Cost in coal to evaporate 1000 gals of water 42. Coal bill for 20,000,000 gals, evaporated per 149 1 80 gals. 5'55 Ibs. 386-id. ft.2 166 203 840 gals. 6*50 Ibs. 32g-6d. 404,430 gals. 7-85 Ibs. 273 -od. /22 74.8 22 BOILER PLANT TESTING difference in efficiency between the two methods. I propose to deal with this question more in detail on page 42. With regard to the following points : 1. Coal Burnt per Boiler per Hour. Calculated as 12,000 B.Th.U. coal the average-figures can be taken as about 865 Ibs. per 30 x 8 ft. boiler per hour, corresponding to about 22| Ibs. per square foot grate area per hour. A very bad plant will, as seen, burn less than this, whereas a plant on modern lines would give about 20 per cent, more duty in this respect. 2. Water Evaporated per Boiler per Hour Calculated at 110 F. inlet temperature, the average is just below 6000 Ibs. per 30 x 8 ft. boiler. This is considerably less than is generally supposed, and most steam users imagine that some- thing like 8500 Ibs. is the figure being obtained ; this, however, does not apply to more than about 5 per cent, of the plants of the country. In general, the boiler plants of Great Britain are working at nothing like their proper output, which is in- teresting in view of the fact that hundreds of works are at the same time in continual trouble due to shortage of steam. 3. Water Evaporated per Pound of Coal. This, of course, depends on the heating value of the coal and the temperature of the feed-water, but taking as usual the averages of 12,000 B.Th.U. and 110 F. the figure is 6-65 Ibs. corre- sponding to 7-62 Ibs. from and at 212 F. For all practical purposes, the figure can be tak^n as 6*5 Ibs. corresponding to 7*5 Ibs. from and at 2I2F. For bad plants and especially those without economisers, the figures are, say, 5*5 and 6-5 Ibs. respectively. Figures like 9 to 10 Ibs. of water from and at 212 F. per Ib. of coal, which are popularly imagined to apply to most boiler plants, only apply to about 5 P er cent, of the plants of the country. 4. Draught. The draught of the average boiler plant is obtained by means of a chimney (natural draught) giving a draught in the base of about 075 in. suction water gauge, which, with the average flues and economisers corresponds to RESULTS AT PRESENT BEING OBTAINED 23 about 0*35 in. water gauge in the side flues. The height of a chimney corresponding to these figures is, roughly, say, 125 to 150 ft. with average flues. In the case of bad plant, a short chimney, say, 90 to 120 ft., and cramped and de- fective flues, the draught corresponds to only about 0-5 in. water gauge in the chimney base. Without economisers this is equal to, say, 0*4 in. in the side flues. If economisers are installed under such conditions, say, an average of 320 tubes for 4 boilers, there is a very serious reduction in the draught, and in the side flues the figure would then only be about 0*20 in. water gauge. On a good plant, using mechanical- induced draught, the figure averages about 2 ins. water gauge in the flue near the fan inlet, and 0^65 in. water gauge in the side flues, much thicker fires being used. 5. Temperature of Flue Gases. The gases leaving the boilers average about 600 F. with coal of 12,000 B.Th.U. On a bad plant with a poor draught and leaky brickwork, the figure is only about 500 F., chiefly because of cold air leakages. In a most efficient plant the figure goes up to, say, 65oF., because of tight brickwork and good fires with a minimum of excess air. With too much draught or "short circuiting" of the gases in the boiler seatings, however, the temperature may go as high as 800 F. leaving the boiler. 6. Quality of Feed- Water. The average figures for feed- water are about 11 total hardness, that is 1 1 grains per gallon, and this means a considerable deposit of scale with a corresponding loss in efficiency. It is difficult to express the advantage of softening the feed water in figures of annual saving in the coal bill, but a good plant should not have have a hardness of over 5 to 6, and a softening plant is necessary in average cases to obtain such figures. A typical average bad plant has, say, 17 grains per gallon, which of course means serious scale troubles. 7. Percentage of CO 2 . The average plant is only giving, say, 7'5 per cent. CO 2 in the side flues, because of medium firing and leaky brickwork, whilst a very efficient plant is 24 BOILER PLANT TESTING about 12 per cent. It must be remembered, however, that high CO 2 does not mean efficiency unless at the same time there is no CO (carbon monoxide) present, and the figure for CO 2 is apt, therefore, to be deceptive. In averages the figures for CO are, say, cri to- 0*3 5 percent., and in good cases O'l to O'2 per cent., bad cases being over I per cent. 8. Economisers. As will be discussed later, the average saving in practice due to economisers is nothing like so great as commonly imagined, averaging about 1 1 per cent, of the coal bill instead of 1 5 to 20 per cent, as usually stated by economiser makers. For the " Lancashire " boiler plant to-day of four boilers, 30 x 8 ft., 320 tubes, 9 ft. tubes, may be regarded as representing average practice, giving 1 1 per cent, saving in the coal bill, and raising the feed-water from 110 to 230 F. As previously stated, in calculating the saving, roughly iiF. rise in the feed-water corresponds to I per cent, saving in the coal bill. The installation of economisers chokes the draught in the case of chimney draught, because of the reduction in the temperature of the gases at the chimney base. Thus, taking a typical case of a chimney 170 ft. high, with gases 600 F. in the base, giving 1*14 in. W.G. at the chimney base, if economisers are installed and the flue gases reduced in temperature to 35oF. the draught would then only be, say, 0*70 in. water gauge. In a plant run on the most up-to-date lines the saving can average 1 8 to 20 per cent, and 17^5 per cent, can be taken as a fair average figure for a good economiser installation, if modern scientific methods of control are adopted throughout. 9. Superheaters. Expressed in averages, the ordinary " Lancashire" boiler plant can be stated to be working without superheaters, and the boiler plants of Great Britain make very little use of superheating. When installed, a rough calcula- tion is, say, 0*05 per cent, saving in the coal bill for every 1 F. rise in the temperature of the steam above saturation point. A plant on modern lines will superheat the steam to, say, 170 to 200 F. above saturation point, and reduce the coal bill RESULTS AT PRESENT BEING OBTAINED 25 9 to 10 per cent, in addition, of course, to the extra efficiency in the engine or turbine. 10. Steam or Power Used Auxiliary to the Production of Steam. As will be discussed later (p. 45), the amount of steam used by steam jets is much greater than is commonly supposed. The average figure we found to be 6 '6 per cent, of the steam production of the plant, being the same for both mechanical stokers and hand-fired steam-jet furnaces, the general impression being that it is only about I to 2 per cent. On individual plants the figure may be anything from 0*5 to 20 per cent, and more detailed figures are given on page 104. Mechanical forced, or induced draught generally takes about 2*5 per cent, of the production at full output. As we have no proper engineering census, it is difficult to give average figures for auxiliary steam or power consumption, because it is not known what proportion of the boiler plants of the country use such apparatus. I estimate roughly that the figure is 35 per cent, of the plants of Great Britain, and 2-5 per cent, for the steam consumption for auxiliary power for the whole of the plants of the country is probably not far wrong. I have taken the bad plants as 5 per cent., and 2*5 percent, would still be required by the most up-to-date plant. Natur- ally, in calculating the net working efficiency of a boiler plant such steam has to be deducted, as it is not useful steam, a point which will also be discussed later. 11. Efficiency of Plant. In averages, a "Lancashire" boiler plant is working at, say, 54-5 per cent, for the boilers only, and 60 per cent for the whole plant, including econo- misers and superheaters, and deducting 2 '5 per cent, for the auxiliary steam. Bad plants may be about 50 per cent, net working efficiency. As already seen, this is very much less than is commonly supposed, the usual empirical figures in general use corresponding to about 75 to 80 per cent, network- ing efficiency, which is entirely erroneous. A modern plant will give about 60 per cent, efficiency on the boiler only, corresponding to about 79 per cent net working efficiency, 26 BOILER PLANT TESTING WATER-TUBE BOILER PLANT. -' Bad Plant Representing, say. 10 Per Cent, of Boiler Plants at Work in Great Britain. Ordinary Aver- age Plant as be- ing Generally Worked To-day, Representing 85 Per Cent, of Plants at Work in Great Britain. Most Efficient Plant Working under Modern Scientific Supervision or under Test Conditions, Representing onlv about 5 Per Cent, of Plants at Work in Great Britain. A. WORKING-DAY TEST. i. Number of boilers working . . . . , 2. Grate area (total) 3. Duration of test . . . . 560-0 sq. ft. 12 hours 4 560*0 sq. ft. 12 hours 560-0 sq. ft. 12 hours 4. Price of coal used (per ton delivered) . 5. Amount of coal used . 4os. 139,765 Ibs. 403. 141,040 Ibs. 403. 137,275 Ibs. 6. Analysis of coal B.Th.U 7. Analysis of coal ash 12,000 10*5 per cent. 12,000 io*5 per cent. 12,000 io'5 per cent. 8. Coal burned per boiler per hour . 9. Coal burned per sq. ft. grate area per hour 10. Water evaporated, Ibs. .... 2911-7 Ibs. 20-8 Ibs. 898,656 Ibs. 2938-3 Ibs. 20*9 Ibs. 989,352 Ibs. 2859-9 Ibs. 20-4 Ibs. 1,081,200 Ibs. ii. Water evaporated per boiler per hour 12. Water evaporated per sq. ft. grate area per hour 13. Water evaporated per Ib. of coal 14. Equivalent evaporation from and at 212 F. per Ib of coal i8,722'o Ibs. 1337 Ibs. 6-43 Ibs. 7-46 Ibs. 20,6ii'o Ibs. 147-2 Ibs. 7'oi Ibs. 8-12 Ibs. 22,525-0 Ibs. 160-9 Ibs. 7-87 Ibs. 9' 1 1 Ibs. 15. Equivalent evaporation from and at 212 F. per 1,000,000 B.Th.U 16. Temperature of feed-water before econo- 621 '6 Ibs. 110 F. 676-6 Ibs. iioF. 759-2 Ibs. 110 F. 17. Temperature of feed-water after economisers 18. Percentage of coal bill saved by economisers 19. Draught in back flues of boilers . 20. Draught in chimney base .... 21. Temperature of flue gases before economisers 22. Number of economiser tubes 23. Temperature of flue gases after economisers 24. Analysis of boiler feed water Degrees permanent Degrees temporary 25. Percentage COz in flue gases (continuous record on combustion recorder) 26. Steam pressure (average) (a) Gauge . 27. Steam pressure (average) (b) Absolute 28. Temperature of saturation of steam . 29. Temperature of superheated steam 30. Steam or power used as auxiliary to produc- tion of steam Thermal efficiency of plant 31. (a) Net working efficiency of plant complete 32. (b) Boilers only . tf o economisers None 070 in. W.G. 0*75 in. W.G. 575 F. None 575 F. 9 2 5-0 per cent. 150 Ibs. 165 Ibs. 365-9 F. 450-0 F. 2 '5 per cent. 61*0 per cent. 59*9 per cent. 195 F. 7-4 per cent. 0*35 in. W.G. 0*50 in. W.G. 475 F. 200 325 F. 6 2 6'o per cent. 155 Ibs. 170 Ibs. 368-3 F. 530-0 F. 2*0 per cent. 69-2 per cent. 60*3 per cent. 225 F. io' 4 per cent. 0-30 in. W.G. 0-65 in. W.G. 450 F. 250 300 F. 12-5 per cent. 160 Ibs. 175 Ibs. 370-5 F. 650-0 F. 1-5 per cent. 81 -9 per cent. 65*8 per cent. 33. (c) Economisers only . . . 34. (d) Superheaters only .... 35. Cost in coal to evaporate 1000 gals, of water LONG CHECK TEST (ONE WEEK). (Say Two Shifts per 24 Hours.) 36. Duration ....... None 2'6 per cent. 333'2d. 168 hours 4-9 per cent. 5-4 per cent. 305'3d. 168 hours 7-6 per cent. 9-7 per cent. 272'od. 168 hours 37. Price of coal used (per ton delivered) . 4os. 458*2 tons 40s. 452*0 tons 403. 467-5 tons 39. Water evaporated 40. Water evaporated per Ib. of coal 41. Cost in coal to evaporate 1000 gals, of water 42. Coal bill for 20,000,000 gals, evaporated per annum (say 220 tons of coal per week) 651, 740 gals. 6-35 Ibs. 337'4d. 28,120 703,674 gals. 6-95 Ibs. 3 o8- 3 d. 25,688 815.769 gals. 7-79 Ibs. 275'od. 22,920 RESULTS AT PRESENT BEING OBTAINED 27 although individual " Lancashire " boiler plants can be worked at 80 per cent, net working efficiency, or even over. This, as seen, includes the full use of economisers and super-heaters, but allowing also for any mechanical draught with the con- sequent 2-5 per cent, auxiliary steam required to drive the fan. These figures can, in general, be said to apply to all large cylindrical boilers such as "Cornish "and "Marine" boilers, and the various adaptions of such boilers with smaller tubes used in conjunction with the ordinary standard furnace tubes. WATER-TUBE BOILER PLANT. The corresponding figures for water-tube boiler plants are given on the adjoining page, calculated for standard sized water- tube boilers with a rated evaporation of 20,000 Ibs. water per hour each, say, about 5250 sq. ft. of heating surface, with grates 14x5 ft, taking as a typical plant five or six such boilers, four working at a time. The boilers are fired by mechanical stokers, as very few water-tube boilers are hand-fired, except small boilers of, say, 10,000 Ibs. hourly evaporative capacity. When fitted with economisers, each boiler has its own separate set of economisers, the present standard practice. The water- tube boiler plant may be said to be the typical power station plant, and is also installed in many factories where high steam pressure is required. There are, of course, a considerable number of different makes of water-tube boilers on the market, some of which give more efficient results than others, but I have endeavoured to give average figures for water- tube boilers generally. 1. Coal Burnt per Boiler per Hour. Calculated as 12,000 B.Th.U. coal, the average figures can be taken as about 28,000 Ibs. (say 1-25 tons) per hour (20,000 Ibs. boiler as already stated', corresponding to 2O'5 Ibs. per sq. ft. grate area per hour. The rate of consumption of coal on the average water-tube boiler is roughly about the same, irrespective of the results being obtained. 2. Water Evaporated per Boiler per Hour. In average 28 BOILER PLANT TESTING plants the rated evaporation is being obtained, namely 20,000 Ibs. per hour, and on very efficient plants the boiler plant is often working regularly on 10 to 20 per cent, overload. It is only on very bad plants that the nominal evaporation is not being obtained, and there is not tfce same difference in evapora- tion between different water-tube boiler plants as there is with cylindrical boiler plants. 3. Water Evaporated per Pound of Coal. Calculated as usual on 12,000 B.Th.U. coal and 1 10 F. feed-water, the average figure is 7 Ibs. of water per Ib. of coal, corresponding to, say, 8 Ibs. of water from and at 212 F. On bad plants the corresponding figures are 6*5 and 7-5 Ibs. Here again, the average figures usually taken, such as 10 Ibs. of water from and at 212 F. are quite erroneous, and only apply to a few plants. 4. Draught. For the average water-tube boiler plant the standard practice is to use mechanical induced or forced draught, or a combination of both, with a short steel chimney, say, 60 to 100 ft. high. In such cases the draught in the flue near the fan inlet is only about 0*5 in. W.G., very much less than induced draught for cylindrical boilers. Roughly, the same figures apply to a most modern plant, whilst on a bad plant the draught is often greater. A number of water-tube boiler plants are worked on chimney draught only, especially small plants, but the draught obtained in practice is not much less, as a comparatively high chimney is then generally used. 5. Temperature of Flue Gases. In the average water- tube boiler plant the gases leave the boiler at about 470 F., corresponding to about 325 F. in the chimney base, which is very much less than in the case of cylindrical boilers. In the case of a bad plant the gases may go up to, say, 575 F. leav- ing the boiler, but this is exceptional, and in very good plants the temperature may be only 450 F. leaving the boiler, and 300 F. leaving the plant. 6. Quality of Feed- Water. As is well known, scale is much more serious for water-tube boilers than for cylindrical RESULTS AT PRESENT BEING OBTAINED 29 boilers, and there is considerable amount of trouble with water-tube boilers because of scale. In average cases, the total hardness can be taken as 8, and in bad plants ordinary feed-water at, say, 12 hardness is used regularly. With a modern softening plant, which reduces the make-up water to 5 to 6 hardness, and the use of engine or turbine condensate, the feed-water should not average more than 3 hardness. 7. Percentage of CO 2 . The average percentage of CO 2 on water-tube boilers with mechanical stokers is low, averag- ing about 6 per cent, less than "Lancashire" or other cylin- drical boilers. In the chain grate type of stoker the fires tend to burn thin at the back and a large excess of air passes, so that even on the 10 per cent, of bad plants the CO 2 figure is practically as good as the 85 per cent, average plants. On the 5 per cent, of good plants this error is avoided, the figure being about 1 2 -5 per cent. 8. Superheaters. The average water-tube boiler installa- tion makes much better use of superheaters than cylindrical boiler installations, and the average figures for superheat can be taken as, say, 160 to 200 F. Even a poor plant is almost invariably fitted with superheaters, as it is the custom for the boiler maker to include superheaters as part of the installation of the boiler. In a very modern plant very high figures are obtained, say 650 F. final temperature, correspond- ing to about 250 to 280 superheat. 9. Economisers. Economisers give less saving on a water-tube plant than with cylindrical boilers, because more heat is retained by a water-tube boiler, leaving less to be absorbed by the economises Thus, in average cases, the number of tubes in the economiser for a 20,000 Ibs. boiler is 200, the feed-water being heated from 1 10 F. to, say, 195 F., saving 7-4 per cent, of the coal bill. On a poor plant no economisers are installed, whilst on a most modern plant the saving reaches, say, 10-5 per cent, with a temperature of 225 F. in the feed- water leaving the economiser. 10. Steam or Power Used Auxiliary to the Production 30 BOILER PLANT TESTING of Steam. It is almost the universal custom, as already stated, to work water-tube boiler plants with mechanical forced or induced draught, and steam jets are consequently not much used. The figures for auxiliary steam or power varies from O'5 to 2-5 per cent, of the production in all classes of plants. ii. Efficiency of Plant. On the average, a water-tube boiler plant is working at, say 60 per cent, efficiency for the boilers only, and 69 per cent, for the whole plant, including economisers and superheaters, and deducting the power used auxiliary to the production of steam. Bad plants may only give a total of about 60 per cent, net working efficiency. These figures, again, are very much less than is usually supposed. It is a common belief that, generally speaking, a water-tube boiler plant is very much more efficient than a cylindrical boiler plant, and that more or less all water-tube boiler plants are efficient. This idea is entirely wrong. Figures like 80 to 82 per cent, net working efficiency, with 65 per cent, due to the boiler only, and 9 Ibs. of water from and at 212 F. per Ib. of coal, are only obtained by a few plants, although, of course, possible on most water-tube plants. As already stated, a cylindrical boiler plant will run on 77 "5 to 80 per cent, under good conditions, whilst the average is 60 per cent. It should be remembered also that the wear and tear and cost of upkeep is, on the average, considerably greater than " Lancashire " boilers. SMALL CYLINDRICAL BOILER PLANT. There are hundreds of such installations scattered about the country in small works, hotels and hydros, various public institutions, etc., with an average coal bill of, say, 10 to 20 tons a week, generally consisting of one or two small " Lancashire " boilers of some such dimensions as 1 5 to 20 ft. by 5 ft. 6 ins. or 6 ft, or one or two small "Cornish" boilers of similar dimensions, hand-fired, working without economisers, and with a very small chimney. The average figures for these plants are given on the ad- joining page. RESULTS AT PRESENT BEING OBTAINED 31 SMALL CYLINDRICAL BOILER PLANT. i. 2. 3. 4. 5. 6. 7- 8. g. 10. ii. 12. A. Working-Day Test. Number of boilers working . . . Grate area (total) ...... Duration of test ...... Price of coal used (per ton delivered) Amount of coal used . . Analysis of coal B.Th.U ..... it n i> Ash ..... Coal burned per boiler per hour ,, burnt per square foot grate area per hour . Water evaporated, Ibs ...... ,, . ,, per boiler per hour ,, ,, square foot grate area per hour .... ,, ,, Ib. of coal Equivalent evaporation from and at 212 F. per Ib. of coal . .- ...... Equivalent evaporation from and at 212 F. per 1,000,000 B.Th.U ...... Temperature of feed-water before economisers after ,, 18. Percentage of coal bill saved by economisers . 19. Draught in back flues of boilers . . 20. ,, chimney base ..... 21. Temperature of flue gases before economisers . 22. Number of economiser tubes .... 23. Temperature of flue gases after economisers 24. Analysis of boiler feed water Degrees permanent . .. . . ,, temporary ..... 25. Percentage CO 2 in flue gases (continuous record on combustion recorder) . . . . 26. Steam pressure (average) (a) Gauge 27. ,, (b) Absolute . 28. Temperature of saturation of steam 29. ,, superheated steam 30. Steam or power used as auxiliary to production of steam ....... Thermal efficiency of plant 31. (a) Net working efficiency of plant complete . 32. (b) Boilers only . . 33. (c) Economisers only . . . '. 34. (d) Superheaters ,, ' . 35. Cost in coal to evaporate 1000 gallons of water B. Lon Check Test (One Week}. (Say Two Shifts per 24 Hours.) 36. Duration ...... . 37. Price of coal used (per ton delivered) . '_'? 38. Amount of coal used . . . . . . 39. Water evaporated . . . . 40. ,, per Ib. of coal . . ' . 41. Cost in coal to evaporate 1000 gallons of water 42. Coal bill for 20,000,000 gallons evaporated per annum (say, 220 tons of coal per week) Ordinary Average Plant as Generally being WorkedTo-day. 2 " Lancashire " 40 sq. ft. 12 hours 403. 8326 Ibs. 12,000 10*5 per cent. 346-8 Ibs. i7'3 .. 48,840 2,035 1017 ,, 5 -86 Ibs. 671 , 110 F. Nil (No economisers) Nil 0-50 in. W.G. 0-25 590 F. No economisers '(i.e., 590 F.) 9 2 5 per cent. 70 Ibs. 85 316-1 F. No superheat None 54-1 per cent. II II 51 Nil 168 hours 405. 17-5 tons 21,952 gals. 5-60 Ibs. 31,886 32 BOILER PLANT TESTING Such small boiler installations are all worked more or less on the same general lines, the boiler attendant combining the work of firing the boiler with other duties, so that the attention received is not continuous. The amount of coal burnt can be taken as an average of 17 Ibs. .*per sq. ft. of grate area per hour, with grates 4x5 ft., and coal as before, averaging 12,000 B.Th.U. per pound. As regards evaporation this can best be estimated from the average figures of an evaporation of, say, 575 Ibs. of water at 110 F. per Ib. of coal, corresponding to 675 Ibs. of water from and at 212 F. Thus, on a small " Lancashire " boiler of 1 5 x 5 ft. 6 ins., the figure is about 200 gallons per boiler per hour. The draught is usually about 0-5 in. suction water gauge in the chimney base, the chimney being small, say, averaging 50 to 75 ft, and on such small mechanical draught is practically never used. Also, it is not general practice to use steam jet furnaces on these plants, but if present, the average figure for the steam consumption of the jets will be 5 to 10 per cent, of the pro- duction of the plant, with a corresponding drop in efficiency. The temperature of the flue gases in the chimney base averages about 600 F. and, as before, the average hardness of the water can betaken as 11 total hardness. Economisers and superheaters are practically never installed, but if present, the saving due to these can be calculated as already shown, namely, 11 F. rise in the feed-water, corresponding to I per cent, saving in the coal bill, and i F. rise in the superheat equals 0-05 per cent, saving. The net working efficiency is about 54 per cent, and probably does not vary more than between 50 to 60 per cent, on any individual plant. SMALL VERTICAL BOILER PLANT. This is a class of boiler largely used in many industries, particularly in engineering works, by builders and contractors, on farms and in small establishments of every description, and the average results being obtained are as given on the adjoining page. RESULTS AT PRESENT BEING OBTAINED 33 SMALL VERTICAL BOILER PLANT. A. Working-day Test. 1. Number of boilers working . . 2. Grate area (total) . . . . 3. Duration of test . . . . 4. Price of coal used (per ton delivered) 5. Amount of coal used . . . . . 6. Analysis of coal B.Th.U. .... 7- i ,. Ash . ... 8. Coal burned per boiler per hour 9. ,, burnt per square foot grate area per hour 10. Water evaporated, Ibs. . . n. ,, ,, per boiler per hour 12. ,, ,, ,, square foot grate area per hour . . . 13. ,, ,, ,, Ib. of coal . 14. Equivalent evaporation from and at 212 F. per Ib. of coal . . 15. Equivalent evaporation from and at 212 F. per 1,000,000 B.Th.U 16. Temperature of feed-water before economisers 17. ,, ,, ,, after ,, 18. Percentage of coal bill saved by economisers . 19. Draught in back flues of boilers 20. ,, ,, chimney base .... 21. Temperature of flue gases before economisers . 22. Number of economiser tubes , 23. Temperature of flue gases after economisers 24. Analysis of boiler feed water Degrees permanent ..... ,, temporary 25. Percentage CO 2 in flue gas (continuous record on combustion recorder) .... 26. Steam pressure (average) (a) Gauge 27. ,, ,, (b) Absolute . 28. Temperature of saturation of steam 29. ,, superheated steam . 30. Steam or power used as auxiliary to produc- tion of steam Thermal efficiency of plant 31. (a) Net working efficiency of plant complete . 32. (b) Boilers only 33. (c) Economisers only . . . , . 34. (d) Superheaters ,, 35. Cost in coal to evaporate 1000 gallons of water B. Long Check Test (One Week). (Say Two Shifts per 24 Hours.) 36. Duration 37. Price of coal used (per ton delivered) . .. 38. Amount of coal used . . . . V 39. Water evaporated . . . ... 40. ,, ,, per Ib. of coal 41. Cost in coal to evaporate 1000 gallons of water 42. Coal bill for 20,000,000 gallons evaporated per annum (say, 220 tons of coal per week) 3 Ordinary Average Plant as Generally being Worked To-day. 12 hours 405. 1353 Ibs. 12,000 10-5 per cent. 112-75 Ibs. 7103 Ibs. 591-9 , 5-25 Ibs. 6-01 ,, 500-9 no F. Nil (No economisers) Nil 0-25 in. W.G. 0-30 800 F. No economisers 9 2 5 per cent. 70 Ibs. 85 316-1 F. No superheat None 48-4 per cent. Nil ti 4o8*id. 168 hours. 405. 4-5 tons. 5141 gals. 5-10 Ibs. 42o-od. 34 BOILER PLANT TESTING Vertical boilers are usually worked with a short metal chimney, often aided by a steam jet in the chimney base, and hand fixed. As regards evaporation, this can be taken as being about 5*25 Ibs. of water at 116 F. per : ft>. of coal, corresponding to 6 Ibs. of water from and at 212 F. Very roughly, such boilers burn about I cwt. of coal an hour and evaporate about 60 gallons of water. The draught in the baste of the small chimney is usually, say, 0*30 in. suction water gauge, but can be higher if a steam jet is used. The temperature of the flue gases is usually very high, averaging 800 F. whilst the per- centage of CO 2 is about 5. Such plants are worked without superheaters, and of course, economisers, and further, steam jet forced draught furnaces are rarely applied. The average net working efficiency can be taken as about 48 to 50 per cent. EGG-ENDED BOILER. This boiler was invented somewhere about the year 1 780, probably by Richard Trevithick, Senior, the father of the more famous Richard Trevithick who invented the " Cornish" boiler, and, as already stated, there are actually still a number of egg-ended boiler plants at work in collieries. How many plants are still running it is not possible to say, but the num- ber seems now to be limited. The average figures for the performance can be taken as given on the adjoining page. The usual dimensions of such boilers to-day are generally 30 to 35 ft. long by 5 ft. 6 ins. diameter, with a blow-off pressure of 40 to 60 Ibs. They are built high up on the top of a large brick firing chamber, so that the bottom of the boiler is in the flames from the fire beneath, an arrangement known as a " flash " flue, and the top half of the boiler in the open air, not generally insulated in any way. There is only one large firing grate, averaging in length 6 x 6 ft. 6 ins., and in width about 4 ft. 6 ins., and the flames travel along the bottom of the boiler and straight up to the chimney, which is placed just behind the boilers. The height of the chimney RESULTS AT PRESENT BEING OBTAINED 35 EGG-ENDED BOILER PLANT. A. Working-Day Test. 1. Number of boilers working . 2. Grate area (total) 3. Duration of test 4. Price of coal used (per ton delivered) 5. Amount of coal used ..... 6. Analysis of coal B.Th.U 7- . M Ash 8. Coal burned per boiler per hour g. ,, burnt per square foot grate area per hour . 10. Water evaporated, Ibs 11. ,, ,, per boiler per hour 12. ,, ,, square foot grate area per hour 13. ,, lb. of coal 14. Equivalent evaporation from and at 212 F. per lb. of coal 15. Equivalent evaporation from and at 212 F. per 1,000,000 B.Th.U. 16. Temperature of feed-water before^ economisers 17. ,, after 18. Percentage of coal bill saved by economisers 19. Draught in back flues of boilers 20. ,, chimney base ..... 21. Temperature of flue gases before economisers 22. Number of economiser tubes . 23. Temperature of flue gases after economisers 24. Analysis of boiler-feed water Degrees permanent temporary . . . . 25. Percentage CO 2 in flue gas (continuous record on combustion recorder) .... 26. Steam pressure (average) (a) Gauge 27. ,, (b) Absolut: . 28. Temperature of saturation of steam . 2g. ,, ,, superheated steam . . . , 30. Steam or power used as auxiliary to production of steam Thermal efficiency of plant 31. (a) Net working efficiency of plant complete 32. (b) Boilers only .'.. ' 33. (c) Economisers only 34. (d) Superheaters ,, ..... 35. Cost in coal to evaporate 1000 gallons of water B. Long Check Test (One Week). (Say Two Shifts per 24 Hours.) 36. Duration , 37. Price of coal used (per ton delivered) 38. Amount of coal used 3g. Water evaporated 40. ,, per lb. of coal 41. Cost in coal to evaporate 1000 gallons water 42. Coal bill for 20,000,000 gallons evaporated per annum (say, 220 tons of coal per week) Ordinary Average Plant as Generally being Worked To-day. 4 112 sq. ft. 12 hours 408. 30,240 Ibs. 12,000 7y ; o f> io'5 per cent. 630 Ibs. 1 / 22'5 ,, 112,800 Ibs. 2,35 >, 83-9 373 4-26 355 Ibs. 110 F. Nil (No economisers) Nil o'go in. W.G. I'OO ,, ,, 850 F. No economisers "(i.e. t 850 F.) 12 5 3-75 per cent. 55 Ibs. 70 320-9 F. No superheaters Nil 34-3 per cent. Nil 168 hours 4cs. 1 82 tons 67,050 gals. 3-65 Ibs. 587'od. 36 BOILER PLANT TESTING usually averages 100 to 140 ft. The firing is carried out by hand, and the fire-bars generally are of a very heavy type, with very poor air-space. Nothing in the way of steam jet bars or other appliances seems to be used in connection with the firing of this type of boiler. Such a plant was the standard colliery practice not so many years ago. In collieries the boiler feed-water is heated by the exhaust steam of the winding and other engines, and generally goes into the boilers at about 150 to 160 F. average. As already ex- plained, however, a given temperature of 110 F. has been taken for the feed- water for comparison, and the results altered by calculation. This, of course, does not alter the essential figure of the efficiency of the boiler itself. In a typical egg-ended boiler of 30 to 35 ft. long and 5 ft. 6 ins. diameter, the amount of coal burnt is almost the same as a "Lancashire" boiler 30 x 8 ft, averaging 22*5 Ibs. of coal per square foot of grate area per hour. The amount of the evaporation calculated to 1 10 F. is only about 250 gallons, practically one-third of that of a " Lancashire" boiler 30 x 8 ft. The water, at 110 F., evaporated per Ib. of coal is only about 375 Ibs., corresponding to, say, 4*5 Ibs. from and at 212 F. The draught on such boilers is usually good, because the chimneys used are a fair height, as already stated, and the flue gas temperature is very high, say 850 F., because the flames merely go along the bottom of the boiler and straight up the chimney. The figure for CO 2 is very low, only about 4 per cent, because of the large open grates and the almost invariably bad quality of the brickwork due to the abnormal expansion and contraction. The net working efficiency is about 35 per cent, a shocking figure, and statements such as 6-5 to 8-5 Ibs. of water from and at 212 F. per Ib. of coal for egg- ended boilers are ridiculous when applied to the present average working conditions. These average figures for the various types of boilers, ex- pressed in one table for easier comparison, are as follows : RESULTS AT PRESENT BEING OBTAINED 37 -a fe So IflJ - to g - S u'o S W^03 > in 2 CO I i*pp>Ou roOO CTiMO. M -Hrt -QO o\ N o o r . M , H -3 ^ ^ 3 ^^ s M o ' s * a ^^s *8 ^ s' c 'c CO M u-> CO . 81 o n" rtCU M o'coik-M 2^-50 1QJQ M^CO vS^O -o O_0\>r>CC^(0 ^^^^^'M M^ p*;*-py\ci ^g^^^^ 00 M ts - ->oQ w JOM -g o o 5 M 10 g p o r 1 - P ' OOOMxowtx -3- uiO>ncOM covo S.d 8. S.S.8.S, S * ^^28 '6 -"S^- . 61S-: 'PI kUi.i- 1 '^. - BOILER PLANT TESTING The great importance of scientific methods in boiler plant management will be realised by the following simple coal balance sheet for Great Britain, being approximate figures allowing for abnormal circumstances due to wars, strikes, ex- change troubles and other complications. GREAT BRITAIN. (Average annual figures which will probably apply more or .less to the next few years.) so tons Coal Disposal (A) Exported, 25 per cent., as follows : Tons. Per Cent, of Total Coal Raised. i. Sold to the colonies and foreign countries . 41,875,000 1675 2. ,, ,, ocean-going steamers . . . 13,750,000 5-50 3. ,, ,, foreign countries as coke 3,125,000 1-25 4. ,, ,, ,, in the form of manufactured fuel (briquettes, etc.) . 1,875 ooo 0'75 O'7S / j Total (B) Home Consumption, 75 per cent., as follows : 6. Steam generation : (a) Powei (b) Low i 7. I omestic 8. Coke from coke ovens 9. Gas works . 10. Railways 11. General purposes Total 62,500,000 25'OO 30SCS . 60,000,000 24-0 are purposes 30,000,000 I2'O . 35,000,000 I 4 -0 /ens . 20,000,000 8-0 . 18,000,000 7-20 . 15,000,000 6-0 4-80 *r v 187,500,000 75*00 That is to say, we consume 90,000,000 tons of coal per annum, 36 per cent, of the total coal raised, or 48 per cent, of the home consumption, for the one operation of steam genera- tion. In general, of this huge amount, 6,500,000 tons are being burnt at say 70 per cent, efficiency or over, 13,000,000 tons at say 65 to 70 per cent, efficiency, 15,500,000 tons at say 60 to 65 per cent, efficiency, 21,500,000 tons at 55 to 60 per cent, 18,000,000 tons at 50 to 55 per cent., and 15,500,000 tons at less than 50 per cent. RESULTS AT PRESENT BEING OBTAINED 39 Now it is possible, in averages, to work steam boiler plants a t 7 5 P er cent - net working efficiency. This is not a theoretical or fantastic figure, but a reasonable and practical basis, such as quite a number of firms have already obtained by the exercise of care and common sense, and with ordinary and well-known plant machinery and appliances. Thus, out of the 400 plants tested, 2 plants are working at 80 per cent, efficiency or over, and 9 plants at 75 per cent, or over, whilst 17 plants are working at 70 per cent, or over. There is obviously something seriously wrong with our general methods of steam generation when 69 plants are actually working at less than 50 per cent, efficiency, a disgraceful performance, whilst another 80 plants are below 55 per cent., and altogether 245 plants are below 60 per cent. We can take almost any industry in the country and if 50 boiler plants are tested, it will be found that the best plant will be 75 to 80 per cent, efficiency, and the figures can be tabulated one under the other, until the lowest plant is not more than 45 per cent, or so. It is quite a common experience to find two works in the same industry working under identical conditions, even in the same street, where the boiler plant of one is, say, 65 per cent, efficiency, and the other 55 per cent., that is, the coal bill of one works is, say, 1 5 per cent, less than the other, so that if one firm is burning 10,000 tons a year, the other man is only burning 8500 tons for the same duty. The general reason for this lamentable state of affairs is the almost complete failure to realise that steam generation is an important, interesting, and intricate branch of applied science, and that in nearly all industries there is more money to be saved in the boiler house than in any other section of the establishment. Modern scientific principles of steam generation comprise two distinct sections, namely, efficient design and equipment of the boiler plant, and scientific methods of control of the working of the plant, so that the best results are obtained. The boiler plants of Great Britain are very defective in both these sections, but the second is much the 4 o BOILER PLANT TESTING more important of the two, and the lack of interest displayed in the intelligent working of boiler plant could not be better illustrated by the fact that there is no recognised method in the country of testing the performance of a plant It should be the special object of an International Code to encourage the continual testing and scientific control of boiler plant by making the Code eminently practical. I will, of course, discuss this second section in detail in the next chapters of this book, but it will not be without interest to state here briefly the circumstances as regards the lack of proper equip- ment on the boiler plants of Great Britain. The boiler itself is of comparatively little importance, and a " Lancashire " or other cylindrical boiler plant will give practically as good results as a water-tube boiler plant. There is, however, a great lack of economisers. In the 250 plants, as already seen, the average saving due to the econo- misers was only 7' I per cent, of the coal bill. The detailed figures for the economiser performance I have given in ''Engineering," 1st November, 1918 ("Exact Data on the Running of Steam Boiler Plants, No. I, Economisers"), as follows : TABLE SHOWING RESULTS OF WORKING WITH ECONOMISERS. Average Temperature of Feed-water. Average Temperature of Flue Gases. Draught in Inches. W.G. Steam Total Division Accord- ing to Saving Obtained. No. of Plants. Pres- sure- Gauge. T he; Evapora- tion on Plant per Hour. No. of Tubes on Plant. Before. After. Before. After. Chimney Side Flue Base or Downtake of .LiDS. F. F. F. F. Blowers. per Sq. In. Lbs. No economiser at all . 95 Less than 5 / 14 134 172 5 68 371 0-85 0-41 80 I7,6i6'o 228 5 to 7 J / . 15 123 194 576 423 o'6i 0'35 79 15,340-0 203 7i to 10 / . 23 IIQ 26l 549 374 0-74 0'33 IOO 16,630 o 231 10 tO 12^ % . 33 118 243 583 403 0-88 0-39 106 29,723-0 401 iajtoi5/ . 46 no 262 58 4 379 0-99 0-41 99 19,760-0 280 Over 15 / . 24 170 284 610 375 i'35 0-60 119 23,497' 392 Total . 250 RESULTS AT PRESENT BEING OBTAINED 41 That is, 95 plants had no means of utilising the waste heat of the flue gases, and it will probably not be an exaggeration to say that 25 per cent, of the boiler plants of Great Britain have no economisers at all. Taking now the 1 5 5 plants fitted with economisers, the average saving obtained on these plants was 11*4 per cent, of the coal bill. Only 24 plants or 17 per cent, of plants fitted with economisers were saving I 5 per cent, or over of the coal bill, and only 12 plants 17 per cent, or over. It is the general rule to install economisers on rule-of- thumb lines such as, for example, 72, 96 or 120 tubes per boiler, quite irrespective of the evaporation. For this reason, and also because the draught is apt to be choked, there is in general not sufficient tubes for the best results. An average saving of say 7-J per cent, of the coal bill instead of about 17-5 per cent, which ought to be obtained, means a national loss of about 10 per cent, in the coal bill, or 9,000,000 tons of coal per annum. Also practically no use is made of feed-water heaters, so that any exhaust steam available, such as that from the boiler feed pump, economiser engine, mechanical draught engine, etc., can be usefully employed in heating the feed-water on the way to the economiser. In averages, something like 3 per cent, of the coal bill is lost in this way, say, 2,700,000 tons per annum. Also, there is a considerable annual loss due to scale in the feed-water, although it is difficult to express this loss in money. The average hardness of the boiler feed-water of the United Kingdom is about 1 1 grains per gallon, and taking the average figure of 6'5 Ibs. of water evaporated per Ib. of coal, this corresponds, at 90,000,000 tons of coal per annum, to an evaporation of 580,000,000 tons of water yearly, and a de- position in the boilers of the United Kingdom of 100,000 tons of scale and other solid material per annum, nearly 2000 tons a week. It is impossible to get the best results on any boiler plant with scale in the boilers, and we do not pay anything like enough attention to the purification of the feed-water, either by a water softening plant or otherwise. BOILER PLANT TESTING As regards mechanical stoking, probably about 25 per cent, or 22,500,000 tons, is burnt per annum by means of mechanical stokers instead of hand-firing. I dealt very fully with the question of mechanical versus hand stoking in a recent paper, " Exact Data on the Performance of Mechanical Stokers as Applied to ' Lancashire ' and other Narrow-flued Boilers," read before the Institution of Mechanical Engineers on the I Qth March, 1920. In this paper was given the detailed figures for the performance of 80 " Lancashire " boiler plants, mechanically fired. The average net working efficiency of these 80 plants was approximately 59 per cent, the boilers only being 53 per cent. In the 250 tests already mentioned 76 per cent, of the plants were hand-fired and 24 per cent, mechanical, and the figures can be divided as follows ("Pro- ceedings of the Institution of Mechanical Engineers," March, 1920, p. 275): 80 Plants Mechanically Fired. 250 Plants (76 Per Cent. Hand Firing and 24 Per Cent. Mechanical Firing). No. of Plants. Per Cent. No. of Plants. Per Cent. Over 80 per cent. . 75 to 80 , I 2 1-25 2*50 2 9 0-8 3-6 70 75 , 65 70 , 60 65 , 2 II 16 15 16 2-50 21-25 2O-OO 18-75 20-OO 13 30 47 43 I2'0 17-6 24-8 18-8 17-2 55 , 60 , 5 55 Less than 50 per cent. 80 lOO'OO 250 lOO'O The average net working efficiency of 350 hand-fired plants is about 60 to 62 per cent. So that on these figures, mechanical stoking is giving actually less efficiency than hand stoking. In my opinion, if all the plants in Great Britain are considered, there is little or no difference between me- chanical and hand stoking, and it cannot be said, therefore, RESULTS AT PRESENT BEING OBTAINED 43 that we are losing much coal, because only 25 per cent, of boiler plants are equipped with mechanical stokers. Another cause of loss is that full advantage is not taken of mechanical draught, and probably about 90 per cent, of the boiler plants of Great Britain rely on natural or chimney draught. The chimney, unless built very high, and good quality coal is used, is an out-of-date and unscientific con- trivance as a draught producer, and other things being equal, the draught simply depends on the temperature in the base, that is, the more heat is lost the better is the draught. When an economiser is installed to prevent this heat loss, the draught is choked because the temperature of the exit gases is reduced. The extent of this draught reduction is seen by the following actual example : Brick chimney 170 ft. high with a temperature in the base of 600 F. working without economisers. The temperature of the gases in a brick chimney is reduced about 2 F. for every 3 ft. in height because of the cooling action of the outside air, so that the average temperature of the hot gases in the whole of the chimney from top to bottom will be about 544 F., taking the outside air as 60 F. The draught in the chimney base under these conditions, expressed as inches suction water gauge, is very nearly given by the following formula (which includes friction losses) : where P = inches water gauge H = Ht. in feet of chimney above the firing level T = the mean absolute temperature of the chimney gases / = absolute temperature of the chimney gases P _ x ( 7 ' 6 7'9 \ \6o + 461 " 544 + 46 1/ / 7 -6 7'9 \ = 17 x \ja - TooV = 170 x (0*01458 - 0-00786) = 171 x (000672) = 1*14 in. W.G., which equals about 0*65 - 0-85 in. W.G. in the side flues of a " Lancashire " boiler. 44 BOILER PLANT TESTING If now we install economisers to reduce the coal bill, say, 17*5 P er cent., the temperature of the gases in the chimney base will be reduced from 600 F. to, say, 350 F. This will correspond to an average temperature in the whole of the chimney of about 294 F., and the draught will be as follows : 7'9\ T/ / 7- \6o + = i 7 o' 461 294 + 46 1 7-6 7-9 \ = 170 (0-01458 - 0-01046) = 070 in. W.G., which equals about 0-35 in. W.G. in the side flues of a " Lancashire " boiler. In the 250 tests the 155 plants fitted with economisers show an average drop in the flue gas temperature from 581 to 389 F. whilst the draught averages 0-97 in. W.G. in the chimney base, and is reduced to 0-45 in. in the side flue or downtake. The installation of economisers therefore causes a serious reduction in the draught and, for example, on theo- retical grounds, if more economisers were installed, and the saving increased to say 22*5 per cent, of the coal bill with the gases cooled to say 250 F. the draught would be so reduced that hardly any coal would be burnt Although, as already stated, chimney draught works very well with a high chimney and good quality fuel, so that say 0*35 in. W.G. is sufficient draught in the side flues, the great majority of boiler plants do not possess these advantages. Consequently, in spite of the fact that, in general, sufficient economisers are not installed, hundreds of boiler plants have to work with the economiser bye-pass damper partly open to allow some of the hot gases to go right up the chimney, so as to provide sufficient draught to work the plant. The main advantage of mechanical draught (forced or induced) is that the draught is quite in- dependent of the flue gas temperature, being provided by the RESULTS AT PRESENT BEING OBTAINED 45 engine or motor driving the fan, which takes the equivalent of about 2*5 per cent, of the steam production of the plant. Consequently, the full amount of heat can be extracted by the economisers, and theoretically the flue gas exit temperature could be reduced to 213 F. to absorb all the heat, and still retain all the water in the flue gases as steam. A large number of boiler plants (93 plants out of the 250 tested) probably about 35 per cent, of the plants of the country, are working with some form of steam jet furnace, either hand or mechanically fired, which can be called a variation of forced draught. Most of these steam jet furnaces are working in conjunction with natural or chimney draught but a few have mechanical induced draught. The amount of steam used by the steam jets averages about 6-5 per cent, of the production of the plant and is much too high. I published in "Engineering," i6th January, 1920 (" Exact Data on the Running of Steam Boiler Plants, No, 3. The amount of Steam Used by Steam Jets "), the results of investigation carried out into the working of 130 boiler plants fitted with steam jet furnaces, eleven different types of hand- fired furnace and eight different types of mechanically fired, comprising 437 boilers with a coal bill of about 1,000,000 tons per annum. The results are tabulated on page 46. It is generally assumed that the amount of steam used by steam jets is small, say I to 2 per cent, of the production, but it will be obvious that these figures are quite erroneous, and the amount is much more than is generally realised. One of the most striking facts is the enormous difference between the amounts of steam used by these steam jets. Thus, the lowest figure obtained was 0-50 per cent, of the production, and the highest 21-4. The differences on different plants using the same make of apparatus are almost as remarkable. Of the whole 1 30 plants, twenty-nine have coal bills of over ;iooo per annum incurred by the use of steam jets alone, whilst the number of plants with coal bills of ^500 or over is forty-seven. 4 6 BOILER PLANT TESTING The great cost incurred in average cases in the running of steam jets is not realised. Thus, taking an average sized boiler AMOUNT OF STEAM USED BY STEAM JETS. AVERAGE RESULTS FOR HAND FIRING. (Net Average 6 ; 6 Per Ceril of the Production.) Type of Apparatus. Number of Plants Fitted. Total Number of Boilers. Percentage of Production of the Plant us^d by Jets. Total Coal Bill of the Plants per Annum. Tons. Total Coal Bill used by the Jets. Tons. A. 6 17 7-6 20,200 1535-2 B. 4 7 4*5 15,400 693-0 C. 3 6 7'3 9,234 674-0 D. 18 60 6'3 137.005 8361-5 E. 7 22 8-1 55,450 4491-4 F. 2 3 3"2 5,850 187-2 G. 2 4 5'0 6,600 330-0 H. 2 6 7*7 14,400 noS'o I. 3 6 4 '4 8,525 375'i J. I 2 15-25 4,000 610-0 K. 6 16 5*9 35,346 2079-5 Total 54 149 312,010 207157 AVERAGE RESULTS FOR MECHANICAL FIRING. (Net Average 6-7 Per Cent, of the Production.) Type of Apparatus. Number of Plants Fitted. Total Number of Boilers. Percentage of Production of the Plant used by Jets. Total Coal Bill of the Plants per Annum. Tons. Total Coal Bill used by the Jets. Tons. Sprinkling Stoker. A. 25 73 5'0 140,345 7017-2 B. 16 45 5*25 95,550 5016-4 C. 7 23 5'o 30,070 I503-5 Coking Stoker. A. 4 12 2-3 21 050 4 8 4 -I B. i 3 13-8 5,750 793-5 C. 13 66 8-0 221,950 17756-0 D. i 3 7-2 4,900 352-8 E. 9 63 7*5 185,250 i393'7 Total 76 288 704,865 46817-2 plant of six "Lancashire" boilers, burning say 12,000 tons of coal per annum, equivalent to 18,000 per annum, with coal RESULTS AT PRESENT BEING OBTAINED 47 at an average price of 303. per ton. The cost of an ordinary steam jet apparatus hand-fired would be about, say, 100 per boiler, equal to 600 for the plant. Taking the average figure of 6'6 percent, of the steam pro- duction used by the jets, this corresponds to i 188 per annum as the cost cf the steam used, equal to buying an entirely new set of steam jet apparatus for the whole of the six boilers about every six months. For a similar plant a mechanical stoker equipment would cost about double, say, 2000, and in this case at 6-5 per cent, of the steam production, this would be equivalent to replacing the whole of the stokers, say, every eighteen months. In addition, also, in the latter case the cost of upkeep of the stokers has to be taken into account, whereas the hand-fired steam jet apparatus has the advantage that the fire-bars last a very long time because of the " cooling " action of the steam. Assuming that 35 per cent, of the boiler plants of the United Kingdom are fitted with steam jet furnaces, this corre- sponds to 31,500,000 tons of coal burnt per annum, and at 6'6 per cent, of the production of the plant, is equal to about 2,000,000 tons of coal used per annum for the sole purpose of steam generation to supply steam jets. Assuming that the steam jet furnace is the right method for burning 35 per cent, of the coal used for steam generation, then with proper care and attention, the amount of steam used ought to be cut down to 3*5 per cent, of the production, say within the limits of I '5 to 4 per cent. That is to say, by the careless use of steam jets, and the use of a number of furnaces of bad design, we are wasting per annum about 1,000,000 tons of coal. The question of the advisability of using steam jets at all is a matter of opinion, but in certain cases, such as for coke and coke breeze, and some varieties of refuse coal, they are very useful. Roughly, I should say that under existing conditions of burning raw coal, about 10 per cent, of the boiler plants of the country are suitable for hand-fired steam jet furnaces, whilst the question of mechanical firing is very open. 48 BOILER PLANT TESTING Finally, as regards superheaters, we do not make anything like the proper use of superheated steam. The value of superheating for fuel economy is in two directions, first, that of partial superheat to, say, 75 F. to reduce condensation losses in the steam pipe circuits,.*and, secondly, that of high superheat up to 200 F. to improve the efficiency of the steam engine or turbine. Out of the 250 plants tested, only eighty were fitted with superheaters, and of these eighty plants only twenty-five plants were completely equipped, so that the average amount of superheat on the eighty plants was 50 F, (316 F. temperature of saturation to 366 F. on the super- heater). It will not be an exaggeration to say that 5 per cent, of the coal bill, 4,500,000 tons per annum, is lost because of the failure to realise the value of superheating. Finally, I should like to point out that Great Britain is probably no worse in respect of inefficient steam generation than any other country, and this deplorable state of affairs seems to exist, for example, in America and France also. I have examined about forty boiler plants in France and judging from this short experience, and from information contained in French engineering literature, it would seem that the average net working efficiency of the boiler plants of France is not much more than 60 per cent., and certainly not over 65 per cent. I have no personal experience of boiler plants in America, but judging by the American engineering literature, especially in connection with the efforts made for fuel economy during the war, it would appear that in America also the average net working efficiency of boiler plants does not exceed 60 per cent. It is painfully interesting to reflect that at least 100,000,000 tons of coal per annum is being lost throughout the world by lack of proper methods of boiler house management. 49 PART II. CRITICISMS OF EXISTING CODES AND SUGGES- TIONS FOR AN IMPROVED INTERNATIONAL CODE. i. The Necessity of Having an Entirely Separate Code for Boiler Plant Testing. It is, in my opinion, a fundamen- tal mistake in the "Civils" Code to lump together boiler plant and steam engine tests, and this error is not committed to anything like the same extent in the American " Mechani- cals " Code, which is divided sharply into separate test codes for boilers, reciprocating steam engines, steam turbines, pump- ing machinery, compressors, blowers and fans, steam power plants, locomotives, gas producers, gas and oil engines, and water-wheels. The arrangement of the Civils Code is appar- ently a persistent relic of the days of over 100 years ago, when the steam engine was invented and developed, and when the word " engine " meant not only the actual steam engine, but the boiler and accessories as well. This point of view may have had some justification in, say, 1822, when 50 h.p. was regarded as a large size for an engine, and each engine as a rule had its own small separate boiler. In 1922, how- ever, it is obviously out-of-date, because of the size and com- plexity of the modern steam generation plant, and because of the many uses of the steam, not only for engine and turbines of greatly different sizes and efficiencies, but also for numerous other processes, such as warming buildings, drying chambers, and heating liquids, in which the condensation loss in the pipe circuits of the factory alone is an important matter. It is for this same reason also that American engineers still persist in 4 50 BOILER PLANT TESTING talking of " boiler horse-power," an unscientific term which has died out in Great Britain years ago. The only practical and scientific method is to regard the generation of steam as something entirely separate and inde- pendent from the utilisation of* steam, whether for steam engines or for any other process, and in fact one of the reasons why this country is losing 20,000,000 tons of coal per annum on steam generation is because this very point is not under- stood. In the average factory, when some endeavour is made to keep a record of the figures for the fuel consumption, such attempts seldom rise above the conception of regarding the boiler and power plant together as merely one item. Thus, for example, a paper-mill expresses the figures of its perform- ance as so many tons of paper, a brewery as so many standard forty-gallon barrels of beer, a flour-mill as so many sacks of flour, a bleach works as so many lumps of cloth, all per ton of coal. The error of this method is that it is not detailed enough, and there is not only entire ignorance as to whether the cause of inefficiency lies in the generation of steam at the boiler plant, in condensation losses in the steam pipe circuits, or in poor engine performance, but the very many different and important functions of the boiler plant are carried out com- pletely in the dark. Yet the " Civils " Code helps to perpetuate this funda- mental error by its method of regarding boiler plant and steam engine tests as something almost identical, so that they can be included in one Code. In the International Code I suggest that boiler plant test- ing be regarded as something entirely separate, and as much independent of the testing of steam engines as it is of oil or gas engines, or any other source of motive power. A separate code for boiler testing also means greater simplicity, and would be a great help in convincing every one that efficient steam generation is an operation of vital importance, and worthy of the most careful attention. CRITICISMS OF EXISTING CODES 51 2. The Object of Boiler Plant Testing. The "Civils" Code gives the unfortunate impression that boiler plant testing is a costly and troublesome luxury to be undertaken only at rare intervals, and, secondly, that there are two kinds of tests, namely, to "obtain data for scientific purposes," and "com- mercial tests " to ascertain whether the guarantee of perform- ance given by the maker has been fulfilled. It also speaks of " Comparative trials where the determination of efficiency is not the main object ". The four separate references to this point in the " Civils " Code are most confusing, and are given below (italics my own) : (1) Page 5. "When the object of a boiler or engine trial is to obtain data for scientific purposes the losses should always be measured, because they afford a valuable check on the ac- curacy of a trial. Such measurements consist in taking the temperature of the flue gases and analysing them, weighing the ash, measuring the loss of heat by radiation, etc. Though de- sirable, they are not essential in a large majority of trials, when those observations only are recorded which are necessary to ascertain whether the guarantee of performance given by the maker has been fulfilled. For such trials a shortened tabular statement has been provided under the heading ' Commercial Trials' 1 :' The note x refers the reader to page 23, which reads as follows, at the bottom of the page : (2) Page 23. "Notes: (i) The lines printed in italics re- late to data which may be omitted where a shorter form of Report for general purposes is desired (see Committee's Report, P- 4>" On turning back again to page 4 we find : (3) Page 4. " As regards the last item, the original forms intended for scientific purposes^ in which it is necessary to measure the losses, have been retained as far as boilers and reciprocating engines are concerned, but abridged forms have also been drawn up for more general use, as described in note i on page 23." Finally, on page 58 is stated : (4) Page 58". " If the object of the trial is to ascertain, for 52 BOILER PLANT TESTING scientific purposes, the rate of evaporation for a constant rate of coal consumption. . . ." A general impression also, on reading through the " Civils " Code, is that boiler plant testing is an extremely complicated and difficult operation, which involv.es a knowledge of chemistry and mathematics quite beyond the ordinary engineer, and which can only be carried out by the University graduate. These ideas are quite erroneous. Boiler plant tests must be regarded as of such vital importance that they must be carried out regularly as part of the routine of the daily running of a boiler plant, and there is nothing mysterious or difficult about them. In the International Code I would suggest one standard code for all boiler tests, and to do away with any idea of dis- tinction between" Scientific" and "Commercial" Tests, which only causes confusion and which, in any case, is wrong in principle. I would draw up the Code on such lines that the main object of boiler plant testing is to keep boiler plants at the maximum efficiency every week, all the year round, and such a Code would be flexible in the sense that it would in- clude all special tests, such as the investigation of a particular quality of fuel, of any plant, machinery, or appliance installed on the plant, and the working of the plant under different con- ditions of load. In short, boiler plant testing must be regarded as a thoroughly practical proposition which is necessary for the strictly utilitarian purpose of saving money. 3. Duration of Test. The duration of the test is a matter of the greatest importance in determining the true performance of a boiler plant, and the " Civils " Code is very vague on this point. All the definite instructions it gives are as follows (P. 9):- "The approximate duration of the trial should be fixed before commencing it, and should be a multiple of the period elapsing between the times of cleaning the fires ; it should never be less than three hours, and should be as long as possible in order to eliminate error in the measurement of the thickness of the fuel." CRITICISMS OF EXISTING CODES 53 It must be obvious that the test has got to be of sufficient duration to allow of the elimination of errors. Thus, I pre- sume even the Civil Engineers' Committee would agree that, for example, the figures of a test of one hour's duration would be worthless as a true indication of the average working of a boiler plant. To allow an official test of only three hours is absolutely ridiculous, as every one must know who has had much practical experience of boiler plant testing, and no re- liance whatever could be placed upon such a test. What is meant by the statement that " the duration of the trial should be fixed before commencing it " and " should be a multiple of the period elapsing between the time of cleaning out," I am at a loss to understand. The American " Mechanicals " Code is infinitely more sensible, definite and practical on this point, as follows : Page 43. " 44. The duration of tests to determine the effici- ency of a hand-fired boiler should be at least ten consecutive hours. In case the rate of combustion is less than 25 Ibs. per sq. ft. of grate per hour, the tests should be continued for such a time as may be required to burn a total of at least 250 Ibs. of coal per sq. ft. of grate. Tests of longer duration than ten hours are advisable in order to obtain greater accuracy. "45. In the case of a boiler using a mechanical stoker, the duration, where practicable, should be at least twenty-four hours. If the stoker is of a type that permits the quantity and condition of the fuel bed at beginning and end of the test to be accurately estimated, the duration may be reduced to ten hours, or such time as may be required to burn the above noted total of 250 Ibs. per sq. ft. " In commercial tests where the service requires continu- ous operation night and day, with frequent shifts of firemen, the duration of the test, whether the boilers are hand-fired or stoker-fired, should be at least twenty-four hours. Likewise in commercial tests, either of a single boiler or of a plant of several boilers, which operate regularly a certain number of hours and during the balance of the day the fires are banked, the duration should not be less than twenty-four hours. " The duration of tests to determine the maximum evaporative capacity of a boiler, without determining the efficiency, should not be less than three hours." 54 BOILER PLANT TESTING It will be noted that the American Code insists on at least ten consecutive hours, and that the three hours allowed by the " Civils " Code for the complete test is in the American Code only allowed for the comparatively unimportant opera- tion of determining the maximum'evaporative capacity of the plant, quite irrespective of the efficiency. Surely the common-sense guide to the duration of the test is the actual practical working conditions of the given boiler plant. If, for example, the boiler plant starts up at 8 a.m., runs full output until 12 midday, partially shuts down for the dinner hour until I o'clock, and then runs full output again to 5 o'clock, the only proper course is to run the test right through for nine hours, that is from 8 a.m. to 5 p.m., including the partial stop in the dinner hour. In a colliery, for example, or under colliery conditions on an experimental plant, the maxi- mum "winding period" may be from 6. am. to 2 p.m., in which case the test would be carried out for this period. Certain industries, such as flour-mills and paper-mills, as a rule, run right through twenty-four hours a day on steady load for six days, in which case, of course, the test can be carried out at any time. I would propose that in the International Code no test should be regarded as official if of less duration than eight hours, and in every case longer than this, or the full working day or shift, is much preferable. In the few cases where the complete working day or shift is less than eight hours, I would allow the lesser time, but would attach little importance as a rule to any test of less than six hours' duration. It would not be possible to include the American figures of ten hours and twenty-four hours in an International Code, because, for example, in this country the average working day is now only eight hours. On this point of the duration of the test a second very serious matter for criticism, in both the " Civils " Code and the American " Mechanicals " Code, is the total omission of all reference to the figures for the performance of the boiler plant CRITICISMS OF EXISTING CODES 55 when starting and stopping, when banked up, and on light load at night* and during the week-end. Most boiler plants do not suddenly start up at full load, run for a test period, and then suddenly shut down again. The usual practice is to run on intermittent loads during the night, for keeping buildings warm and perhaps working the factory at much reduced load ; and also for the boiler plant to remain banked up under pressure during the week-end, if only to be able to work the pumps in case of fire. Speaking in averages, anything from 10 to 30 per cent, of the annual coal bill is usually absorbed in this way, and a test carried out in the spirit of both the Codes only applies there- fore to the 70 to 79 per cent, of the coal burnt during ordinary working hours. I found out very soon, by practical experience, that it is necessary to carry out a long check test to include the essential elements of the day's test, namely, the amount of water evaporated and coal burnt, together with the heating value of the coal, and consequently out of the 400 tests, 365 tests have a long check test of one complete week in addition. In the International Code, therefore, I would suggest a long check te^t of a complete week of 1 68 hours, that is, including the full week-end, and the combination of the two tests, namely, a day test of not less than eight hours, and a full week's test, will give, in my opinion, a much more satisfactory test of a boiler plant both from a practical as well as from an academic and scientific point of view. In the 400 tests, thirty-five plants were tested during the day only, either because it was impossible to carry out a week's check test without a great deal of trouble, or because of the express wish of the client. In the remaining 365 tests, 185 plants gave a slightly inferior result as compared with the day's test, and 1 80 tests showed a somewhat better result. There is always some loss by cooling during the week-end, but this is often counter-balanced by the fact that a boiler 56 BOILER PLANT TESTING plant may have been forced during the day, so that the efficiency is less than at night when on easy load. 4. Sampling and Analysis of the Fuel. Both Codes give, on the whole, very clear instructions as to the sampling of the fuel, and the necessity 5. " Duplex Mono," front view, with door open. FIG. 6. " Duplex Mono" with clock and other mechanism exposed whilst the machine is actually running. FIG, 7. Original " Mono'' automatic gas analysing machine for CO and unburnt gases only. CRITICISMS OF EXISTING CODES 69 and there are now the following additional instruments put on the British market since 1913 : " Cambridge Electrical " (British) " Hays" (American) "W.R. " Combustion Indicator (British). There must be almost an equal number of different makes of CO 2 Recorders on the American market, and there are also a number of continental machines, chiefly French, German and Dutch. To insinuate that all these instruments, most of which have been on the market for years, and in many cases tested and certified correct by the National Physical Laboratory, cannot be used for determining accurately the percentage of CO 2 on a boiler test is, in my opinion, not only ridiculous, but grossly unfair to most makers of CO 2 Recorders. It might be thought therefore that, as far as the "Civils " Code is concerned, the section relating to Flue Gas Analyses was originally drawn up in the years 1897-1902, when there may have been a good excuse for ignoring the CO 2 Recorder, and that the Revision Committee of 1913 had practically left the original instructions alone, although they were by this time hopelessly out-of-date. In spite of this, however, it is explicitly stated in the introductory letter (p. 4) that one of the sections revised was the sampling and analysis of the flue gases. The " Civils " Code, as seems to be usual when any modern appliance is considered, throws doubt on the accuracy of all CO 2 Recorders, and states they must be checked, not only before the trial, but during it as well (!). It may be re- marked that to test the accuracy of most CO 2 Recorders, all that is necessary is to let the instrument run on air for a few minutes to ensure that the chart record is exactly 0*0 percent. CO 2 . To talk of testing a CO 2 Recorder "during " the trial, that is every few hours, is childish, and one is compelled to come to the conclusion that most of the members of the 70 BOILER PLANT TESTING " Civils " Committee have had little or no experience with CO 2 Recorders. I would like to guarantee that the results given by the average CO 2 Recorder are far less liable to error than the clumsy methods of hand anafysis recommended by the " Civils " Code, apart from the fact that the average CO 2 Recorder will give twenty analyses in the same time that it takes to carry out three or four analyses by hand. In the International Code I would make it compulsory to use a CO 2 Recorder working at a proper speed of say fifteen analyses per hour, but preferably twenty analyses, so that on the day's trial at least 150 CO 2 determinations will be carried out. Further, I would suggest that the CO 2 Re corder be worked day and night on the plant for the whole week's check test, say six or seven hours in turn, taking flue gas from essential portions, such as the downtake or side flue of " Lancashire " boilers, or entrance to the main flue of each water-tube boiler, the main flue, chimney base, etc., so that during the whole test about 2 500 analyses of CO 2 would have been carried out. There is very little trouble in obtaining re- sults like this with a CO 2 Recorder, and in fact it is much less trouble to get 500 analyses with a recorder than to carry out ten analyses by hand. I have used CO 2 Recorders on every one of the tests of 400 boiler plants during the last twelve years or so, and it is possible to take a CO 2 Recorder equipment on to a boiler plant and have it recording the percentage of CO 2 on a chart at the rate of twenty analyses per hour in fifteen minutes from the time of arrival on the plant. For convenience, it is best to have the instruments taken out of their original iron cases and installed in special wooden cases, so that they can easily be carried about and hung up on the wall close to the boilers. Most CO 2 Recorders use a trickle of water (3 to 5 gallons per hour) to drive them, and the easiest method is to place on the top of the recorder a small galvanised iron cistern, specially made CRITICISMS OF EXISTING CODES 71 to fit, and holding about 2 gallons of water. All that is needed to work the instrument is a bucket of water, the tank being filled, the water allowed to work the CO 2 Recorder through a tap, and run down into the bucket underneath, when it can be returned to the tank about every half-hour. At the points of analysis ^-in. W.I. pipes are inserted in the flues, and provided with rubber corks through which a piece of glass tube is inserted. The CO 2 Recorder is then hung up near the chief points, and connected to the glass tube by means of a thick india-rubber tube. The other end of the CO 2 Recorder is connected in the same way to another piece of i-in. W.F. pipe inserted in the chimney base or adjacent main flue. A convenient way to do this is to take several lengths of |-in. W.I. pipe and lay them on the floor tempor- arily for the test. By this arrangement a continuous circula- tion of flue gas is ensured, that is to say, if, for example, the point of analysis is the side flues of a " Lancashire " boiler, the draught in the chimney base pulls a continual current of flue gas from the side flues through the CO 2 Recorder, and there is no inaccuracy due to " lag " in the pipes. As already stated, this question of CO 2 is also of the greatest importance in the regular working of a boiler plant, and it is most unfortunate, to say the least of it, that the present standard Boiler Testing Codes should not only be out-of-date in this respect, but should also disparage the CO 2 Recorder in the most unjustifiable manner. In the 400 tests since 1908, with which I have been as- sociated, there is included approximately 400,000 analyses of CO 2 by means of CO 2 Recorders, and the average figures of CO 2 for all these tests is only 7-5 per cent, and I should estimate that at least 90 per cent, of the boiler plants of Great Britain are unprovided with CO 2 Recorders in running order. The figures for the 400 plants are divided as follows : BOILER PLANT TESTING Percentage of CO 2 . No. of Plants. Corresponding Figure Expressed as a Per- centage. 12 per cent, and over . . . 10 2'5 ii ,, , ,, . 10 2'5 10 , . ..^39 9-9 9 . . . 48 I2'I 8 .- . 60 re*2 .^ 3 7 M 91 2^*4. 6 , : . . 58 14-7 5 M > 12*9 Less than 5 per cent. . 27 6-8 Total 394 The amount of calculated loss on a boiler plant due to excess air is shown by the curve, Fig. 8, and the average OF fVCt- L.OS S FIG. 8. Curve showing the fuel loss for corresponding percentage of CO 2 . figure of 7*5 per cent CO 2 corresponds to an approximate calculated loss of 1 1 -o per cent, in the coal bill, taking 14 per 1 Six plants not determined. CRITICISMS OF EXISTING CODES 73 cent, as the maximum CO 2 obtainable. That is to say, on the annual coal bill in Gre it Britain for steam generation of 90,000,000 tons, the loss due to low CO 2 is no less than 9,900,000 tons per annum. With regard to the analysis for CO, there has been no option but to use hand methods, since the new " Duplex Mono," already described, is only just coming on the market. The recommendations of the " Civils " Code with regard to the hand methods to be used for the analysis of flue gases are as follows (pp. 68-69) : "It is essential that the temperature of the flue gases should be taken at the same point as that from which the sample for analysis is drawn. Great care must be taken to avoid their dilution by air leaking between the boiler and the surrounding brickwork, through cracks in the brickwork or through ill-fitting damper frames or other openings in the ex- ternal flues. " Since the gases cannot be assumed to be homogeneous, an attempt must be made to obtain an average sample through- out the width of the flue ; it should be remembered, moreover, that (i) the gas must be drawn into the sample tube at a uni- form rate per hour, (2) the gas in the dead space between the flue and the sample tube must be eliminated, (3) the gas in the sample tube must not be allowed to diffuse back again into the main tube or be drawn back into the main current by sudden changes of pressure." l The note l is as follows, together with the illustration : " Condition (3) is not satisfied by any of the ordinary forms on the market. That shown in Fig. 6 (see next page) has been designed by Mr. G. Nevill Huntly to fulfil (i), (2), and (3). The gas is drawn in at A, the dead space is cleared by sucking at B ; the rate at which the gas is drawn is fixed by the distance C-D, and this can be increased by joining a piece of glass tube with rubber to D. The gas in E cannot be sucked back and cannot diffuse back. The two 3-way taps greatly simplify the transference of gas for analysis." 74 BOILER PLANT TESTING Mercury The Code then goes on to say : "It is often necessary and advisable to carry out the analysis of the gases at once. One of the most convenient arrangements for this purpose is the ' Orsat ' apparatus, as it requires no supply of pure water* and no bottles of chemicals. : " A calibrated 'Orsat' appar- atus with mercury as a confining liquid gives satisfactory results for carbon dioxide, but is not suitable for determining carbonic oxide. The commercial pattern is not recommended for the de- termination of oxygen. Instruc- tions for use are issued by the makers of the different instru- ments. " If the gases are analysed on the spot the determination of carbon dioxide alone is advisable ; this will permit of tests every fifteen minutes. If a fuller analysis is deemed necessary a series of continuous samples should be drawn off into small tubes over mercury and analysed either during or as soon as pos- sible after the trial." The reason of all these com- plications is not very clear, but seems to be due to the fallacy in the " Civils " Code of trying to calculate the efficiency of a boiler plant from the analysis of the flue gases. It is most extraordinary that anyone should suggest that it is necessary to collect samples of flue gas over mercury. The samples taken according to the methods suggested would be so small in size that they could not possibly be an average of the vast volume of flue FIG. 9. (Fig. 6 in the "Civils" Code.) CRITICISMS OF EXISTING CODES 75 gases in an ordinary boiler plant, and if we are going to use mercury to take samples even approximating to a true average, we should require hundredweights of mercury, not ounces. All this worry about "dead space," " diffusing back," "changes of pressure," "temperature of the gas," etc., is simply a waste of time on a practical boiler test. It is quite simple to obviate it by drawing two or three samples of flue gas into an " Orsat " or other hand apparatus one after the other, and only analysing the last sample. In any case there is no need to use mercury, and a solution of glycerine and water will give results just as accurate after it has been in contact with flue gas for a short time and become saturated. The makers of the " Orsat " apparatus will be considerably astonished to learn that " the use of mercury as a confining liquid" is recommended, and also that the "commercial pattern " whatever this may mean is not suitable for the determination of oxygen. In any case, to carry out only four analyses per hour is nothing like sufficient to get a proper average, and, as already stated, a CO 2 Recorder will give about twenty analyses per hour. It is very little use to take " snap" tests at quarter of an hour intervals, because the composition of the flue gas varies almost continuously, as a CO 2 Recorder soon shows. In order to collect average samples of flue gas over a period, the "Civils " Code states, on page 69 : " For a permanent apparatus the arrangement of collect- ing-tubes illustrated by Mr. Breckenridge ' 2 is probably the best. It averages gas both for temperature measurements and gas samples. It must, however, be built into the flue, and hence is only suitable for a permanent installation. Mr. Breckenridge has shown, however, that a single steel tube closed at the inner end and perforated with a series of holes 3 throughout its length takes a good average sample : it is readily withdrawn for cleaning, and is the most practical form for temporary trials. A very much larger quantity of gas must be drawn from the flue than is taken for analysis. The sample is drawn from a small tube joined to the main aspirating 76 BOILER PLANT TESTING tube ; the latter may be of half-inch bore, and the current may be conveniently produced by a steam- or water-ejector. A test should be made for gas-tightness before and after trial." The notes at the bottom of page 69 are as follows : " 2 A study of four hundred 1 " steam tests made at the Fuel Testing Station, St. Louis, Mo., 1904-1906. 'United States Geological Survey, Bulletin 325, Washington, 1907, 'page 157." " 3 The area of these holes must be small as compared with that of the bore of the pipe, otherwise less gas will enter the tube near its closed end than elsewhere." According therefore to the "Civils" Code we are recom- mended, for the collection of large samples of flue gas, to use some method devised years ago by Professor Breckenridge of U.S.A., no further information being given. Presumably when a boiler test is contemplated every one has to write out to the United States Geological Survey. Accordingly I wrote to Washington; U.S.A., and I am informed (May, 1921) very courteously that the publication in question has been out of print for a very long time and is no longer available from any official source in America. It will give some idea as to the practical value of the " Civils " Code when it recommends an American method described in some publication twenty years old, and which has apparently been out of print for years in the land of its origin. I fail to understand what is the need of all this trouble about collecting large samples of flue gas, which is a perfectly simple operation. For a few pounds one can buy a very efficient gas collecting apparatus, which will collect say 15,000 to 20,000 c.c. of flue gas at any desired speed. For example, we have the " Hays" Gas Collector, as illustrated, Figs. 10 and II. In this apparatus a water supply pipe is connected to the valve WV and the large tank of the collector about half filled with water, with a pint of engine oil poured on the top, so that the water in the collector does not absorb any CO 2 . The inlet for the flue gas is through the valve GV on a J-in. FIG. 10. " Hays " Automatic Gas Collector. [To face p. 77. CRITICISMS OF EXISTING CODES 77 pipe. To work the collector the flue gas valve GV is closed t the valve GC opened and the supply of water turned on by opening the water valve WV so as to slowly fill the whole of the FIG. n. Working principle of the " Hays" Automatic Gas Collector. collector with water until it begins to flow out of the overflow OP, when the valve WV is shut off. The valve GC is then closed and the flue gas valve GV opened. 78 BOILER PLANT TESTING Water will then flow from the tank T into the flow regu- lator R and be discharged through the drip DC. The rate of the water, that is, of the collection of the flue gas through GV, will be absolutely steady, depending on how much DC is open and is quite regardless of ; 4he level of the water in T. If this flow regulator is not provided, a simple bottle or cylinder filled with water and allowed to empty itself will not give a true average sample over a number of hours because, as the vessel empties, the rate of flow of the water diminishes as the "head " of the water is less in the apparatus. These gas collectors can be installed permanently at different points of the plant, and one instrument retained for carrying about for temporary installation at any other points of the plant. The sample of gas for analysis is withdrawn by the "Orsat" apparatus through the valve GC and after analysis the gas content is expelled and the apparatus is ready for use again. The content is very large, about 17,500 c.c., and the rate of collection can be fairly rapid, say, 2000 c.c. per hour. This large sample of gas can then be analysed at leisure for CO 2 , CO and oxygen, and the figure for CO 2 will be a useful check on the CO 2 Recorder. I would suggest, therefore, that the International Code insists upon a large sample of gas being taken continuously, say 2000 c.c. per hour, and this large sample' be then analysed for CO 2 , CO and oxygen. The next question is the points at which to draw the samples of flue gas, and this concerns also the methods of calculation. In the " Civils " Code flue gas analysis is intended primarily, as already stated, as a basis for the " heat balance " of the plant and therefore it is necessary to draw samples of gas from the chimney base, that is the final exit of the plant, so as to calculate from the analysis of the gas the amount of heat lost by the plant. I propose to discuss in detail the method of calculation in later pages, but will say here that in my opinion this method of calculation is not so accurate, and CRITICISMS OF EXISTING CODES 79 is infinitely more complicated and troublesome, than the simple method based on the actual heat in the coal. The difficulties of the " Civils " Code method are shown by the following paragraph from the code (p. 70) : " Since the amount of heat carried away by the flue gases is proportional to their volume, and this volume is at any instant inversely proportional to the amount of carbon dioxide by volume, the mean percentage found from a recording in- strument or from the analysis of an average gas sample is not exactly that required. The error from this cause, with a boiler fired from a mechanical stoker, will probably be under cri per cent, of CO 2 ; with a hand-fired boiler it may be as high as 0-5 per cent.; with, say, 10 per cent, of carbon dioxide in the flue gas this would mean errors of I per cent, and 5 P er cent, respectively in the heat balance, and renders unnecessary a higher accuracy than cri per cent, on the carbon dioxide ; for this accuracy to be reached the gas sample must be collected over mercury. Considerable errors may occur if water is used, and the next best fluid to mercury is a mixture of equal volumes of glycerol and water, when the error would probably not exceed O'4 per cent. " The apparatus used for gas analysis, therefore, must be correct to cri per cent, for carbon dioxide and oxygen, 0*05 per cent, for carbon monoxide, and cro2 for methane. These correspond roughly on an average heat balance with about 0*25 per cent, of the total heat available. "When boilers are fired by mechanical stokers, the gas samples may be drawn directly into an analysing apparatus, but when the firing is by hand continuous collection is neces- sary to ensure correct results. "When all the gas aspirated passes into the collecting- vessel the volume of the aspirating-tube must be very small compared with the volume of the vessel into which the gas is drawn, otherwise the sample collected will contain little besides the gas lying in the tube when the collection was begun." I suggest in the International Code to abandon entirely this " heat balance sheet " method of calculation, and to take the samples of flue gas as near the furnace as possible, so as to get proper information as to the state of the firing. That 8o BOILER PLANT TESTING is to say, the sample pipes, say ^-inch W.I., should be placed in the downtake or side flues of " Lancashire " or "Cornish" boilers, in the furnace exit of water-tube boilers, the front uptake on marine boilers and so on. Samples can of course be taken at various othfer points as already described, to detect air leakages, and an efficient permanent installation for a CO., Recorder, with connection to take samples for hand analyses, has pipes with valves connected to all the points, so that any point on the plant can be switched on to the CO 2 Recorder at will, whilst at the same time the gas is filtered from dust and dirt through a filter, and a continuous current of gas maintained through the circuit to do away with " lag" errors. 6. The Method of Measuring the Boiler Feed- Water. In order to measure the amount of water evaporated there are two general methods that can be adopted, namely, (i) weigh- ing the water or measuring its volume in tanks, and (2) using a water meter. A third possibility, that of measuring the output of the plant as actual steam by means of steam meters, will be discussed later (p. 131). The " Civils " Code gives a most elaborate account, occupy- ing no less than eleven pages (pp. 38-49), of the methods to be used, insisting on the tank method only, even at sea. The reference to water meters is as follows (p. 46, No. 6, "Feed Meters"): " Feed meters are not recommended for scientific trials, but as some makes appear to be capable of giving results within I per cent, of accuracy, they may be usefully employed in many cases when it is desired to obtain an approximate idea of the normal performance of the boilers ; they should, however, be calibrated before and after the trial with water at the temperature of the feed. A " Venturi " meter, for instance, or a notch gauge, may be used for the continuous measure- ment of water when the quantity is large, and the flow is fairly steady, as giving fairly accurate results." This is altogether a most remarkable paragraph. It is stated that some meters "appear to be capable" of giving CRITICISMS OF EXISTING CODES 81 results accurate to I per cent, in which case they can be used when " approximate " results are obtained. If any water meter is accurate to I per cent., then in my opinion it is probably more accurate than the " Civils " method of weighed tanks, and much more suitable for all boiler trials, no matter how "scientific". I have had a fairly long experience of boiler trials carried out by means of tanks, and even when the greatest care is taken, very few trials are accurate to I per cent, by this method. The operation is in practice extremely clumsy and tedious, and the observers quickly get tired of the monotonous operations, so that errors soon tend to creep in every time a tank is filled and emptied. Further, whilst a meter can be calibrated "before and after the trial " and any errors detected, it is impossible to do this with the tank method, although the error is likely to be as much, if not more, than a water meter. As with CO 2 Recorders, the above paragraph may have applied to 1897-1901, when the "Civils" Code originated, but to say that it applies at the present time, or even in 1913 when the Code was revised, is quite wrong. There are at the present time nearly twenty different makes of boiler feed meter, British and American, on sale in this country ; apart from many continental meters, and for the " Civils " Code to maintain that all these meters are not accurate enough for the very purpose for which they are specifically designed, namely, boiler testing, is a very strong statement, to say the least of it, and one which is made without any apparent justification. The American " Mechanicals " Code is, as usual, much more up-to-date and states the following (p. 12): " 9$. Water Weighing and Measuring Apparatus. (i) Feed-water. Wherever practicable the feed-water should be weighed, especially for guarantee tests. The most satis- factory and reliable apparatus for this purpose consists of one or more tanks each placed on platform scales, these being elevated a sufficient distance above the floor to empty into a receiving tank placed below, the latter being 6 82 BOILER PLANT TESTING connected to the feed pump. Where only one weighing tank is used the receiving tank should be of larger size than the weighing tank, to afford sufficient reserve supply to the pump while the upper tank is filling. If a single weighing tank is used it should preferably be of such capacity as to require emptying not oftener than every five minutes. If two or more tanks are used, the intervals between successive empty- ings should not be less than three minutes. Measuring tanks calibrated by weighing may also be used. " In tests of complete steam power plants, where it is required to measure the feed-water without unnecessary change in the working conditions, a water meter may be employed. Meter measurement may also be required in many other cases, such as locomotive and marine service. The accuracy of meters should be determined by calibration in place under the conditions of use. " If a large quantity of water is to be measured, an automatic water-weigher, a rotary, disk, or Venturi meter, a weir, or some form of orifice measurement may be employed. In any case the measuring apparatus should be calibrated under the conditions of use, unless its design is such that standard formulae and constants may be applied for determining the discharge. If recording mechanism is employed in connection with orifice or weir measuring apparatus, make sure that its record is reliable." This statement is, however, not quite fair to water meters in general, and most meters can be used quite well for the smallest boiler plants. It will be noted that the wording as regards "Venturi" and " Notch " meters is very much the same in the two Codes, and it is very curious that the "Civils" Code has quite a number of references to specific American conditions, such as the "Barrus" calorimeter (p. 57), the " Breckenridge " method of gas collection (p. 75), and the question of three hours' duration of the test (p. 53). It is interesting to give the names and types of the various boiler feed meters on the British market. Such meters are divided into two general classes : (a) open, non-pressure types, and (<) closed, pressure types. CRITICISMS OF EXISTING CODES 83 The first class consist of (i) Automatic Water Weighing Machines in which a small tank is continually filled with water to a certain weight, when it is released, and the water falls into the boiler feed tank, the number of such small weighed tanks being recorded on a train of wheels mechanism. Of this type there is the "Avery Automatic Water Weigher," the " Leinert Meter " and the " Sarco Tippling Meter". (2) " V Notch Meters" in which the water flows through a " V " notch of given dimensions, and, by means of a float mechanism, a continuous record is kept of the height of the water in the notch, the amount of water passing being proportional to the height in the notch. Of meters on this principle there is the ''Bailey Meter," the "Kent V Notch Recorder," the " Lea V Notch Recorder," the " Paterson Fluxograph" and the "Rheograph Water Flow Recorder". (3) "Weir Meters!' on the same principle as the last, but using a " weir " instead of a ''V" notch, represented by the " Simmance-Abady Precision Meter". In the second class, the meter is placed between the boiler feed pump and the economiser, or, in fact, at any point of the circuit between the feed pump discharge and the boiler feed valve. In the case of an injector the meter must be on the suction side, and can also, if necessary, be on the pump suction. Pressure meters are divided into the following classes : (1) Piston Meters in which the pressure of the water actuates in its travel a double acting piston, so that each stroke of this piston represents a definite amount of water, the number of strokes being recorded by a train of wheels mechanism. Meters of this type are the " Kennedy," " Sarco " and " Worthington Duplex". (2) Rotary Meters, in which the pressure of the water actuates a rotary displacer, wheel, disc, or other appliance, one revolution corresponding to a definite amount of water, and the number of revolutions being recorded by means of the usual train of wheels mechanism. This type is represented 84 BOILER PLANT TESTING by the "Kent Uniform" Meter, the " Leeds " Meter, the "Sarco Disc" Meter, the " Siemans Disc" Meter, the " Siemans' and Adamson " Meter, and the " Worthington Turbine " Meter. (3) " Venturi Meters" on the principal of the " Venturi " tube, such as the "Bailey Fluid Meter 4," and the ''British Thomson-Houston " Meter. A water meter has two very great advantages over any tank method. In the first place, it is much more convenient and reduces the trouble and worry of boiler testing to an astonishing degree. Secondly, it has the very great advantage that it enables a boiler plant to be run on the only possible lines for the highest of efficiency, namely, that of continuous testing all the year round. It is absolutely essential that a weekly record be kept of the water evaporated on a boiler plant, together with the amount and analysis of coal burnt, and other vital figures, and the water meter is practically the only possible method of doing this. The " Civils " Code in this respect is not devised' on practical lines, and is obviously only intended for an occasional test, as it would be almost impossible to carry out even a week's trial on the methods recommended. Eor trials at sea, and for locomotives, the advantages of the closed type of meter will be obvious. The trouble is, of course, that all water meters are not equally accurate, and it would be rather a delicate matter for me to attempt to give an opinion as to the accuracy or other- wise of any individual meter. I would suggest that this be one of the investigations to be carried out by the Committees engaged in devising the International Code, and that a list of " approved " meters be issued after such investigations are complete ; any one of these approved meters, under suitable conditions, to be allowed for use on an official test, in addition, of course, to the tank method if desired. Whatever meter is used, it must be installed with a bye- pass arrangement, and a testing tank. The method of instal- lation I recommend, from long practical experience, for per- CRITICISMS OF EXISTING CODES manent installation for a pressure type of meter is illustrated in Fig. 12. A is the water meter and F a small calibrated test tank, (say 6000 Ibs. capacity). Normally, the feed-water flows srppy/nuE .&R*LUL Siipg Type, fit Sy\HPLlH^ C/^STIN6. TESTING TANK.. FiG. 12. Typical installation Recommended for a boiler feed meter (pressure type). through the stop valve B b the sampling casting E, the meter A, the combined stop and check valve C and on to the boiler plant. Bj is an ordinary parallel slide stop valve, E is a simple "sampling casting" of my own design, con- taining a thermometer socket to read the temperature of the feed-water actually passing through the meter, whilst at the 86 BOILER PLANT TESTING same time a tap is provided to take a sample of the water. The valve C is an ordinary stop valve, but loose on the spindle so that it acts also as a non-return (check) valve. The object of this is to obviate any back pressure action on the meter and make the travel,in the meter absolutely " for- ward " only. On the bye-pass is a safety valve D as a safe- guard to the meter, and in this position it acts for both bye- pass and normal feeding. In case of necessity the meter can be shut out of the circuit by closing valves B x and C and opening valve B 2 . In order to test the meter when running normally, all that is necessary is to shut valve C and open valve B 3 to the test tank. I devised and used the above method long before I ever read the American " Mechanicals" Code, but in this Code the following almost identical method is recommended (p. 155) : " Calibrating Water Meters. 227. Referring to Fig. 2, two tees A and B are placed in the feed pipe and between them two valves C and D. The meter is connected between the outlets of the tees A and B, and the valves E and F are placed one on each side of the meter. When the meter is running, the valves E and F are opened, and the valves C and D closed. A small bleeder G is kept open to make sure that there is no leakage. A gage is attached at H. When the meter is tested, the valves C, D and F are closed, and the valves E and I are opened. The water flows from the valve I to a tank on platform scales. In testing the meter, the water is throttled at the valve I to obtain the desired rate of discharge, the gauge meanwhile showing the working pressure. The piping leading from the valve I to the tank is arranged with a swinging joint, consisting merely of a loosely fitting elbow, so that it can be readily turned into the tank or away from it. When the desired speed has been secured, the end of the pipe is swung into the tank at the instant the pointer of the meter is opposite some graduation mark on the dial. When the required number of cubic feet are discharged, the pipe is swung away. The tests should start and stop at the same graduation mark on the first dial, and continued until at least 10 or 20 cub. ft. are discharged for one test. The tank is weighed before and after filling. CRITICISMS OF EXISTING CODES 87 ' 228. The water passing the meter should always be under pressure so that any air in the meter may be discharged through the vents provided for this purpose. Care should be taken that there is no unnecessary air drawn into the feed- water. The meter should be tested before and after the trial, and repeated calibrations should be made to obtain confirmative results. " 229. Fig. 2 1 and the description apply to a piston meter, but any other type of meter carrying water under pressure may be calibrated in the same manner." F IG . !3. Meter calibration. (American Mechanical Engineer's Code, Fig. 2.) The note x says : <(1 Reproduced from 'Trans. Am. Soc. M. E., J vol. 24, p. 724, fig. 1 1 8." By such methods the accuracy of the meter can be checked at any time in less than half an hour. The " Civils" Code might very well have stated that one of the advantages of the tank method is that it is always certain in the sense that any error is small, whereas if a meter does go wrong, the error may be very great and uncertain, as it may be high or low, or commence suddenly. By using a testing tank, therefore, as 88 BOILER PLANT TESTING described, in conjunction with a water meter, the possibilities of error are reduced to a minimum, and comparable, for example, only to the possibility of making a mistake in the actual number of tanks used on a test according to the " Civils " Code. For' a very large boiler plant and a per- manent installation, I recommend, as an absolute certainty, two different makes of meter in series with testing tank, as the expense of an extra meter is trifling in comparison with the advantages obtained. 7. Moisture in the Steam. Another of the great practical difficulties in the way of a scientifically accurate boiler plant test is that steam, unless superheated, always contains some water. The amount is generally I to 2 per cent, but may be anything from zero to even 5 per cent. Theoretically, of course, such water must be deducted, as it is included as steam (with the full absorption of latent heat) in the amount of water evaporated, and if not deducted, makes the efficiency of the boiler plant too high. Both the " Civils" and the American ''Mechanicals" Codes state that this moisture in the steam must be determined, and deducted in calculating the efficiency. The " Civils " Code, however, goes on to point out the difficulties as follows (p. 52): " No. 1 5. Measuring the Moisture in Steam. Up to a certain limit, depending on the steam velocity, the moisture can be measured by one of the forms of calorimeter in the market ; these instruments are not generally reliable when the moisture exceeds about 2 per cent, as it then appears to creep along the walls of the pipes and does not all enter the calori- meter. It is therefore desirable to provide the steam-pipe with a steam-drier, and to measure the quantity of water which is discharged by an automatic trap ; and also to measure the moisture of the steam after it has passed the steam-drier. Opinions differ as to whether it is best to collect the steam by a perforated tube fixed across the diameter of the steam-pipe or by a tube arranged to collect from the centre only of the steam-pipe, but the former method is in more general favour." CRITICISMS OF EXISTING CODES 89 Further references to the question of moisture in steam are (p. 78):- " Full information should be given as to the method of determining the weight of the moisture present in the steam, and how it was trapped, or collected, and weighed." l The note l says : "See paper by Dr. W. C. Unwin, < Proc. Inst. Mech. E.,' 1895, P. 31." The chief difficulties are that the ordinary steam calori- meters, whether of the throttling or separating type, are not very accurate at the best of times, and particularly, as stated, when the moisture is over 2 per cent. ; there is no known satis- factory method of obtaining a true average sample of steam; and further, the steam also has generally a violent swirling motion as it passes along the pipes, especially in the neighbourhood of bends, valves, etc., which does not make the sampling any easier. The " Civils " Code is compelled, therefore, to recommend the laborious proceeding of inserting a special steam drier in the steam pipe, measuring the amount of water discharged from the drier by means of an automatic trap, and then deter- mining the moisture left in the steam, by means of a steam calorimeter, after it has passed the steam drier. In this con- nection it may be pointed out that the modern type of steam- drier such as the " Stefco " and the " Tracy " are not placed in the steam pipe circuits, but in the boiler itself. There is, consequently, no trouble in fitting them, and they will separate the moisture so that the steam contains certainly less than I per cent, and may be absolutely dry. It is also very difficult to know where to take a sample of steam issuing from a boiler. If it is drawn from the vertical branch pipe a few inches above the boiler shell there is trouble due to " showers " of condensed steam, and the percentage of moisture shown is apt to be too high. On the other hand, 90 BOILER PLANT TESTING the sample must be taken close to the boiler to avoid inac- curacies due to cooling and condensation. The American " Mechanicals " Code states (p. 18) : "9/#. Steam Calorimeters. The most satisfactory in- struments for determining the ^amount of moisture in steam are calorimeters that operate upon the throttling principle, or that combine the throttling and separating principles ; the orifice used being of such size as to throttle to atmospheric pressure, and the instrument being provided with two ther- mometers, one showing the temperature above the orifice and the other that below it. If no commercial make of calorimeter is available on a test, an instrument of the throttling type can be made of pipe fittings as shown in Appendix No. 11. In- struments working on the separating principle alone may also be employed ; also certain forms of electric calorimeters. See 'Trans. Am. Soc. M. E./ vol. 28, p. 616." The Appendix 2 is a detailed description of a throttling calorimeter. As regards the method of sampling the steam the American " Mechanicals " Code has the following (p. 35):- " C. Sampling Steam. 28. Construct a sampling pipe or nozzle made of ^-in. iron pipe and insert it in the steam main at a point where the entrained moisture is likely to be most thoroughly mixed. The inner end of the pipe, which should extend nearly across to the opposite side of the main, should be closed and the interior portion perforated with not less than twenty i-in. holes equally distributed from end to end and preferably drilled in irregular or spiral rows, with the first hole not less than half an inch from the wall of the pipe. "The sampling pipe should not be placed near a point where water may pocket or where such water may affect the amount of moisture contained in the sample. Where non- return valves are used, or where there are horizontal connec- tions leading from the boiler to a vertical outlet, water may collect at the lower end of the uptake pipe and be blown up- ward in a spray which will not be carried away by the steam owing to a lack of velocity. A sample taken from the lower part of this pipe will show a greater amount of moisture than a true sample. With goose-neck connections a small amount CRITICISMS OF EXISTING CODES 91 of water may collect on the bottom of the pipe near the upper end where the inclination is such that the tendency to flow backward is ordinarily counterbalanced by the flow of steam forward over its surface ; but when the velocity momentarily decreases the water flows back to the lower end of the goose- neck and increases the moisture at that point, making it an un- desirable location for sampling. In any case it should be borne in mind that with low velocities the tendency is for drops of entrained water to settle to the bottom of the pipe, and to be temporarily broken up into spray whenever an abrupt bend or other disturbance is met. " 29. If it is necessary to attach the sampling nozzle at a point near the end of a long horizontal run, a drip pipe should be provided a short distance in front of the nozzle, preferably at a pocket formed by some fitting, and the water running along the bottom of the main drawn off, weighed, and added to the moisture shown by the calorimeter ; or better, a steam separator should be installed at the point noted. " 30. In testing a stationary boiler the sampling pipe should be located as near as practicable to the boiler, and the same is true as regards the thermometer-well when the steam is super- heated." The use of a steam calorimeter at all is a laborious, hot and generally most unpleasant job, and I am afraid that if the " Civils " Code is to be following in this respect, there will be in practice more trouble in determining the moisture in the steam than in testing all the rest of the plant put together. As a consequence, in very few boiler trials is the moisture in the steam determined, and practically all the 400 boiler tests we have carried out have been average in this respect. I would suggest that in the International Code, the deter- mination of the moisture in the steam be abandoned entirely, as the results are dubious and not worth the trouble, and that this question be investigated by a future Committee with a view to settling the point definitely. It certainly seems a feasible proposition to insert steam driers in the boilers them- selves, as already mentioned, since these modern driers in- crease the economy of the boiler plant, and the steam could then be taken as free from moisture. 92 BOILER PLANT TESTING In plants fitted with superheaters there will of course be no inaccuracy, but, unfortunately, superheaters are compara- tively little used in this country. Thus out of the 400 plants only 114(285 per cent.) were fitted with superheaters, and most of these only partially equipped. For plants without superheaters the efficiency results would therefore be a little too high. In practice, however, by a little care this error can be re- duced to a minimum by keeping the water at a reasonable height in the gauge glasses. Most boilers in Great Britain are working much below their proper rated output, so that moisture in the steam is not excessive, and for testing any plant for guarantee the amount of evaporation can be specified so that the conditions with and without the appliance will be the same in this respect. As usual, of course, the moisture determinations, by means of a steam calorimeter, can always be added to the International Code by arrangement for any particular test. 8. Specific Heat of Superheated Steam. With regard to this point, the " Civils " Code gives the following (p. 50) : "No. 13. Specific Heat of Superheated Steam. Experi- ments have for some years past been carried out with the object of ascertaining the amount of heat that is represented by various degrees of superheat, but definite results have not yet been obtained. " As far as our present knowledge extends the value 0-48 may be adopted for the purposes of this report. Every 100 F. superheat then represents about 4 per cent, of the total heat of the steam, and an error of croi in the value of this specific heat wonld not affect the results of the calculations by more than o-i per cent. 1 " The note *, at the bottom of the page, is as follows : " l The values given in Marks and Davis's Tables are generally accepted. A chart giving ' mean specific heat ' of superheated steam over a wide range of temperatures and pressures will be found in ' Principles of Thermodynamics,' by G. A. Goodenough. 2nd ed. London, 1912." CRITICISMS OF EXISTING CODES 93 This hardly seems to be the right point of view in 1922. 0*48 is the specific heat figure as originally determined by Regnault and Him. Mini's formulae was : Specific heat at constant pressure is = 0-4304 + 0*0003779 T (T = temp. F.) 0-500 0450 40 50C TEMPERATURE 700 750 Of SUPEK HEATED FIG. 14. Curve showing the specific heat of superheated steam (Knoblauch and Jakob's figures). This, however, is only correct for atmospheric pressure. The figures for the specific heat of superheated steam at different pressures have been determined with great accuracy by Knob- lauch and Jakob, of the Royal Technical University, Munich, and are given in any book on steam tables, in the form of the curve Fig. 14, and I would suggest, therefore, that in the International Code these figures be taken as official for 94 BOILER PLANT TESTING the calculations. The figures vary from about 0*45 to 0*685 > thus, in the examples I have given in the specimen Complete Report according to the suggested International Code (p. 149), the temperature of the superheated steam is 559 F. and the boiler' pressure 162 Ibs. absolute, with a corresponding saturation temperature of 364 '2 F. From Knoblauch and Jakob's table, the specific heat of super- heated steam for these conditions is 0-54, wh'ereas of course the "Civils" Code would take it as 0-48. The difference may not be great, but we might as well use the accurate figures, especially when it is no more trouble. 9. Steam or Power Used Auxiliary to the Production of Steam. One of the most serious defects in both the American " Mechanicals " and the " Civils " Codes is the scanty attention given to the question of steam or power used auxiliary to the production of steam, arid this fact alone largely destroys the real value of any boiler plant test carried out according to either of the codes. To illustrate this point, it is best to start with the simplest possible boiler plant, namely, one boiler, hand-fired, with injector feed and natural chimney draught. Apart from the infinitesimal amount of heat taken as energy in working the injector (the latent heat of the steam used being returned to the boiler) all the steam produced from this boiler is useful steam ready for the factory. That is, if 5000 Ibs. of water is evaporated per hour, 5000 Ibs. of steam is ready at the boiler stop valve for useful work, and the real net working efficiency of the plant is calculated on this 5000 Ibs. evaporation. But if we now add to this simple boiler plant any appliance to help in generating the steam, and this appliance takes some steam directly or indirectly to work it, then this latter steam must be deducted from the evaporation in calculating the real net working efficiency of the plant. For example, if a steam jet furnace, hand or mechanically fired is added to the plant, and the amount of steam taken by the steam nozzles is 10 per cent, of the production of the CRITICISMS OF EXISTING CODES 95 boiler, that is, 500 Ibs. per hour, then the real net production of the boiler plant is only 4500 Ibs. useful steam per hour (5000 - 500) in spite of the fact that the boiler is still evaporat- ing 5000 Ibs. The net working efficiency must be calculated on the 4500 Ibs., that is, the real amount of steam available for useful work. This may sound elementary and obvious, but it is at any rate not clear to the Committees who devised the two Codes, and it is conveniently ignored in practically every boiler plant test that has ever been published. Thus at the present time we have firms making appliances for steam genera- tion, advertising broadcast figures of tests carried out with their particular appliances, in which results of 75 to 80 per cent, boiler plant efficiency are shown, and the fact that the appli- ance itself may be wasting say 2-J to 20 per cent, of the steam production of the whole plant is coolly ignored. It surely must be obvious that all steam used in connection with the production of the steam must be deducted, and from the point of view of real net working efficiency the boiler plant must be regarded as being in a closed box, into which a certain amount of heat as coal or other fuel is thrown at one end, and a certain lesser amount of heat as useful steam comes out at the other end, as shown in Figs. 15 and 16. The references given to this vital point in the " Civils " Code are so confusing that one can only say that no definite instructions are given at all. These references dissected from the Code are as follows (p. 7) : " It is particularly desirable that arrangements should be made for supplying the steam used by auxiliary apparatus, such as steam blast, fans, pumps, etc., from a separate boiler entirely disconnected from that under trial, 1 separate feed-measuring apparatus being provided if it is desired to ascertain the quantity of steam thus used. If such auxiliary boiler cannot be blanked off during the trial, the pressure in it should, if possible, be maintained the same as that in the boilers under trial, in order to minimise leakage through the stop- valve ". 9 6 BOILER PLANT TESTING CRITICISMS OF EXISTING CODES 97 98 BOILER PLANT TESTING The above note l refers to page 70, and is as follows : " No. 30. Steam Used by Steam Jets and Fans. Many mechanical stokers are provided with steam jets under the bars, others require steam for actuating the mechanism of the fans. The ^necessary steam may, under special conditions, exceed 10 per cent, of the quantity pro- duced, and should be subtracted from the weight of feed-water (unless the efficiency of the heating surface, is in question) because it is not available for useful work ". There is also the following : Page 74, paragraph " Line 3 ". " If mechanical stokers are used, the name of the stoker should be stated, and also whether it was of the sprinkling or coking type, unless the information has already been given in Line i. It is desired, if possible, to give particulars of the power needed to work the stoker and how supplied." Page 74, paragraph " Line 4 ". " If any system of forced or induced draught is used it should be carefully explained, and the positions and any peculiarities of the draught gauges be described : data as to the power needed or steam used should be supplied, if possible." Page 78 "Line 30". " Wherever possible the amount of steam employed in producing the draught (whether an actual steam jet is used or whether the steam is used in working an engine for driving fans) should be given, and it should be determined by an independent test. The steam which is blown through nozzles, can, however, with a reasonable degree of accuracy, be calculated by the formula on page 71 (Sect. 30)." It is obvious, therefore, that the " Civils " Code in general regards the determination of the steam or power used as of little or no value, although " it is desirable if possible " to undertake it. In fact, so little importance do they attach to this question that they do not even trouble in the calculations to give an example, although a number of pages are devoted to explaining how to calculate a "heat balance sheet" from the flue gas analysis, a matter of relatively little moment. Thus also on page 5 it is stated that : CRITICISMS OF EXISTING CODES 99 " The principal measurements are those pertaining to the weighing of the fuel and the determination of its quality ; to the weighing of the water evaporated, and to the measure- ment of the power produced. The data thus obtained suffice to determine the thermal efficiency of a boiler." In the American " Mechanicals " Code the following appears (p. 47, " Correction for Steam or Power Used for Aiding Combustion. The quantity of steam or power, if any used for producing draught, injecting fuel, or aiding com- bustion, should be determined and recorded in the Table of Data and Results. This should also be recorded by foot- note below the table, a statement showing whether or not a deduction has been mide from the total evaporation for steam or power so used, and if such deduction has been made, the method of computing it." The American Code, therefore, only considers auxiliary steam or power as a matter for a foot-note, and leaves it apparently to the fancy of the engineers in charge as to whether they bother to deduct it or not. The various points of the plant, where such auxiliary steam or power is used (already given in Figs. 15-16), con- sidered in detail, are as follows ; (1) Mechanical Coal and Ash Handling. The amount of steam or power taken here is not excessive. An average sized plant of three " Lancashire " boilers handling, say, 90*0 tons of coal per week will take approximately, say, 3 h.p. reckoned as continuous working. The determination of the h.p. and its equivalent in steam production on the plant, can be determined without trouble. Thus, if a small non-con- densing steam engine is used, the figure of 30 Ibs. of steam per i.h.p. can be taken, and the h.p. of the engine calculated in the ordinary way. Such a result will be accurate to 10 per cent, which is near enough as this particular item is only a small one. (2) Mechanical Stoker or Moving Hand-fired Bar Drive. Here again, the power is not particularly excessive, averaging ioo BOILER PLANT TESTING I to 2 h.p. per boiler in most cases, say 30 to 60 Ibs. of steam per boiler per hour, or 0-4 to O'8 per cent of the pro- duction, and can be calculated as in item (i). (3) Steam Jets. The question of steam jets is undoubtedly the worst feature of this question of auxiliary steam, and the " Civils " Code does not realise the importance of it. It ad- mits in one paragraph that the amount may be 10 per cent of the production, and then in another, contents itself with a pious expression that the amount should be determined if possible. There is not the slightest indication given that it is just as essential to determine this figure as it is that of the amount of water evaporated and the coal burnt, and that without the figure any boiler test is practically useless. As further showing the minor importance attached to this point by the " Civils" Code we have the following (p. 71) : " The steam which is blown through nozzles can, with a reasonable amount of accuracy, be calculated with the help of the formula : Q - Ibs. of steam per minute = P x a where P = the steam pressure above that of the atmosphere (i.e. it is the gauge pressure) plus 7-5 Ibs. and a is the sectional area in square inches of all the nozzles. This formula gives too high results for pressure below 50 Ibs. per square inch. P should be measured near the nozzles. If the steam is superheated, the weight of steam, as found above, must be multiplied by the square root of the ratio of the absolute temperatures (t + 459) of the saturated and superheated steam." I maintain that such an empirical formula is absolutely worthless for determining the real amount of steam used by nozzles. In the first place, it is impossible to determine with any reasonable amount of accuracy the area of a number of nozzles. For example, a given steam jet furnace may have anything from 6 to 64 nozzles per boiler, and these nozzles soon wear larger by the friction of the steam, the holes being then, as a CRITICISMS OF rule, irregular in shape. It is a hopeless job, under these con- ditions, on an average sized boiler plant of, say, four boilers, with anything from 100 to 200 nozzles, to get the real area of all these nozzles. I once tested a boiler plant of thirty-two "Lancashire" boilers with 14 nozzles per boiler, that is, a total of 448 nozzles on the plant, and the " Civils " Code seriously suggests that the only way in testing this plant would be to try and measure the irregular area of 448 different nozzles worn by the steam, each nozzle having to be taken off by means of pliers, and then replaced. Also, such nozzles are supplied as a rule from each boiler independently by a small steam pipe, which may be -J to I in. in diameter, and this pipe is provided with a stop-valve which is often worked partially open, so that the actual area of the nozzles is in any case not a criterion of the amount of steam passing. The American " Mechanicals " Code recommends the use of steam meters, but allows also various empirical methods, as given below (p. 1 3) : 9'l V PLANT TESTING Lbs. of Dry Steam Per Hour. ;'' 430 615 " 930 I2OO 1400 1560 2180 . ' ., 'ffc . 2640 3050 TABLE I. DISCHARGE THROUGH ORIFICE i IN. DIA. AT 100 LBS. PRESSURE. Pressure Drop, Lbs. Per Sq. In. 4 I 2 3 4 5 10 15 20 "The water-glass method affords an approximate means for determining the steam consumption of auxiliaries, and for measuring the leakages of steam and water from the boiler and its connections. (See Appendix No. 3 for description of water-glass method.)" The empirical "water-glass" (i.e., gauge glass) method is described as follows (p. 154) : *'() Water-Glass Tests of Leakage. 224. To deter- mine the leakage of steam and water from a boiler and steam pipes, etc., the water-glass method may be satisfac- torily employed. This consists of shutting off all the feed valves (which must be known to be tight) and the main feed valve, thereby stopping absolutely the entrance or exit of water at the feed pipes to the boiler ; then maintaining the steam pressure (by means of a very slow fire) at a fixed point, which is approximately that of the working pressure, and ob- serving the rate at which the water falls in the gage glasses. It is well, in this test, as in other work of this character, to make observations every ten minutes, and to continue them for such length of time that the differences between successive readings attain a constant rate. In many cases the conditions will have become constant at the expiration of fifteen minutes from the time of shutting the valves, and thereafter the fall of water due to leakage of steam and water become approxi- mately constant. It is usually sufficient, after this time, to continue the test for two hours, thereby obtaining a number of half-hourly periods. When this test is finished, the quantity of leakage is ascertained by calculating the volume of water which has disappeared, using the area of the water level and the depth shown on the glass, making due allowance for the weight of one cubic foot of water at the observed pressure." CRITICISMS OF EXISTING CODES 103 The vital importance of a proper method of determining the steam used by the nozzles is best shown by giving some indication of the amount of steam that is being used by them in practice. I have already given (p. 45) some data on this point, and in my experience, the average consumption is about 6 -5 per cent, of the evaporation of the plant. Of the 400 plants tested, 153, that is 38 per cent, were fitted with steam nozzles, and the figures for each of these 153 tests are given in the columns of figures on pages 104-7. With regard to the method to be used for determining the amount of steam used by steam jets, in the International Code I suggest the alternatives of a steam meter, or a surface condenser. The only matter for criticism on this point in the American J< Mechanicals" Code is that it does not make the use of steam meters compulsory, and allows alternative em- pirical methods, but the " Civils " Code does noj: allow steam meters at all. Steam meters will be dealt with more par- ticularly on page 131, but they are especially convenient for determining with great accuracy (say to I per cent since the demand is steady) the amount of steam used by nozzles, when these nozzles are all fed from one main supply pipe. I would recommend strongly, as the ideal arrangement for a modern boiler plant, that the steam pipes be so designed that the whole of the auxiliary steam of the plant be passed through one auxiliary steam main pipe, as a branch from the main steam pipe over the boilers. On this auxiliary pipe should be installed a steam meter, ^which would thus give a continuous record of all the auxiliary steam. Also, because the amount used by steam nozzles is much the greatest and requires special watching, I would recommend further that all these nozzles be supplied by one pipe branching from this auxiliary steam main pipe, and on this pipe a second steam meter be installed. IO4 BOILER PLANT TESTING I go>5 8 1 hn 1 ju 3 Ijfjjg p 1 E '5 PQ SJ W " <;i O iii> i * 3 3 "o ' T3 'o MS]j c J3 '*J J eat; 5 ^5 c >^ o i i a 3 1 ilii u * Hi i S c^ 1 ^ 1 w z s 345 Explosives .... Mechanical 6 6861 48-78 21-40 267 >j .... )( 2 5915 53-81 17-60 393 Jute mill .... ' 2 2848 40-81 17-20 356 Colliery .... Hand 2 3173 47'79 15-25 394 > .... Mechanical 7 4343 40-77 15-10 355 Paper mill .... ,, 3 5144 47-91 13-80 245 ... ? j 3 4311 55 >:t 4 13-50 365 Hosiery manufacturers Jt 2 2838 46-88 12-25 Explosives . . * . ?> 15 6533 57-39 IX "35 325 Engineering Hand 2 4761 50-33 10-90 334 Paper mill . . . M 5 4995 49-77 10-80 167 >> Mechanical 2 6583 59HO 10-40 Colliery .... Hand 7 9728 46-47 10-30 295 ' 6 6108 52*00 10-30 375 Pottery .... H 2 1698 45-97 IO'2O 231 Woollen mill Mechanical 2 7002 55*97 10*00 376 Paper mill . . . 3 5408 45*94 9'6O 3*3 Cotton yarn dyeing H i 5393 50-94 9'60 389 Paper mill .... Hand i 2970 9*58 64 Cotton mill . Mechanical i 3929 66-39 9-40 76 Explosives . M 5 5933 65-58 9-25 259 Woollen mill t) 4 6316 54-50 9'tio 346 Engineering Hand i 5404 48-64 9-00 360 Paper mill .... Mechanical 2 5601 47-46 8-90 286 Colliery .... Hand 6 3561 52-60 8-88 338 ,, .... Mechanical 8 6229 49-35 8-64 66 j .... Hand 3 4294 66-30 8-60 88 Woollen piece dyeing . M 2 6126 64-87 8-50 130 Calico printing . .-.. 4 429 1 62-14 8-50 33 Explosives . . . - C Mechanical 8 6363 69-13 8-36 382 .... Hand 5 4123 44-80 8-07 367 Colliery . tt 3 4249 46-73 8-06 156 Explosives .... Mechanical 3 7382 8-00 268 Colliery . . . Hand 3 3813 53"8o 8-00 366 Chemical manufacturers Mechanical 4 2990 46-85 8-00 359 Colliery . . v . Hand 9 4297 47'5Q 7-90 230 Glue manufacturers . -. J} 2 3071 56*01 7-88 275 Colliery . . . '. tt 3 4483 53' 2 5 7-76 308 Paper mill . - . . M 2 4699 51-36 773 344 > ... Mechanical 2 6431 48-81 7-60 385 Glue manufacturers . ~'.. . Hand 2 4879 44-12 7-60 244 10 Colliery Explosives . . . . Mechanical 16 445 4347 55-I5 76-11 7-50 7-47 328 " ' * ' ' " 9 4955 50-11 7-30 CRITICISMS OF EXISTING CODES 105 ill. "O bi) >2 J3 c W .2 | i 1 s E '5 M '3 . |f|| s f 1 "o *o a. ^ bio's "S c 1 II |P| st 1 a 1 -ill cj- w o 153 Calico printing . Mechanical 4 6769 60-14 7-30 69 Woollen piece dyeing . M 3 7737 65-96 7-20 51 Engineering 3 5946 67-07 7-02 312 ,, ... Hand 2 4530 50-95 7-00 340 Flour mill .... ii I 4585 48-98 6-90 283 Residential mansions . 2 3474 52-69 6-84 98 Colliery .... M 8 5000 63-98 6-65 87 Explosives .... Mechanical 13 7401 64-98 6-co 225 Glue manufacturers Hand 3 2754 56-19 6*50 129 Woollen mill Mechanical 4 6915 62-21 6-50 48 ... M 2 6714 67-23 6*50 390 Colliery .... Hand 5 5852 42-33 6-40 347 ' > }j 5 4331 48-44 6-40 335 Flour mill .... Mechanical 2 3554 49*46 6-31 318 Cotton piece dyeing . Hand 2 50-66 6-30 62 Chemical manufacturers Mechanical 3 5943 66-40 6-25 307 Engineering H 4 4211 5i-39 6 -02 235 Woollen mill |f 4 7357 55-74 6'oo 1 60 Hand 2 3159 6-00 269 Hosiery manufacturers Mechanical 3 6046 5374 5*95 357 Colliery .... Hand 3 4869 47*67 5 '90 96 Tannery .... i 8805 5-80 109 Colliery .... j } 6 634 1 63-17 5*75 190 Explosives .... Mechanical 4 7702 57*7 5*60 336 Glue works .... M 4 2940 49*42 5-58 330 Food products Hand 2 3731 50-04 5-50 243 Engineering M I 4836 5*50 38 Cotton piece dyeing . Mechanical 3 6421 68-47 5*40 271 Paper mill .... Hand 3 4274 53*47 539 144 Dyeing and cleaning . Mechanical 7 6216 60-86 5-30 181 Paper mill .... ii 2 3368 58*33 5*30 63 Colliery .... Hand 3 8035 66-39 158 Explosives .... M i 6136 59-82 5-00 223 Dyeing and cleaning . Mechanical 4 6601 56-30 4*97 254 Cotton piece dyeing . . ,, 3 6255 54*96 4'93 12 yarn 5> 4 735 74*05 4*9 379 Colliery .... Hand 2 3895 45-60 4-80 256 ,, tj 3 4963 54*73 4-80 215 Paper mill .... M 2 4798 56-65 4-80 91 Explosives .... Mechanical 10 6927 6474 4*52 273 Paper mill .... Hand 2 53'29 4*47 395 Hosiery dyeing . Mechanical 2 2212 40-06 4*47 291 Cotton yarn dyeing M 2 5754 52-47 4*40 118 Tannery .... Hand 2 4167 62-86 4*35 io6 BOILER PLANT TESTING I oSo s >>4J t^ . bi c 1 1 .SJbU I CO D o 1 b CQ l-J W l<*o ii 1 , 3 1 *0 "o QH ^ wg-g c 2-S ja " " ' ''o OJ ^P p-2 1-1 6 J2 S 2 >S J o^ a 3 o ^ 11 VH ex, llt^ 0i W cu 55 Paint manufacturers Hand I 4982 66-78 4-30 107 Woollen mill Mechanical 3 6073 63-25 4-30 186 > ... Hand 2 4486 50-96 4*20 67 Special textile . . t . Mechanical 6 6144 66-23 4*20 240 Aniline dye manufacturers . Hand 4 4757 55-34 4*20 211 Paper mill . )? 6 7326 56-76 4-20 303 Woollen mill .._. . - . Mechanical 2 5789 51-68 4-12 79 Cotton mill . . . * . Hand 2 8372 65H4 4-10 23 Colliery .... ?J 5 938i 71-84 3-97 226 Explosives . .'..:-. M 4 7716 56-18 3-97 294 Mechanical 2 4463 52-09 3-92 249 Woollen mill J) 2 3598 55-05 3-80 104 Cotton mill .... Hand 4 3618 63-49 3-80 125 Paper mill ". . . . Mechanical 6 5366 62-47 3 'So 368 Colliery . . M 7 4907 46-66 3'8o 229 Dyeing and cleaning . 5 5071 56-03 3-8o 343 Heavy inorganic chemicals . M 3 2555 48-88 3-5 306 Dyeing and cleaning . 3 3182 Sl'5 1 3-50 200 Woollen mill n i 7463 57-37 3-50 170 Hosiery ,, . . . II 3 4376 58-94 3-50 198 Cotton piece dyeing Hand 57-43 3-50 77 Paper mill .... n 5 6536 65-57 3-50 80 Dyers and cleaners Mechan'cal 4 65H 1 3-5 57 Explosives . M 16 9528 66-74 3'47 317 Colliery . . Hand 6 5590 50-78 3HO 360 Linen mill .... .. 4 3626 3'4 213 Woollen mill . Mechanical 2 3167 56-70 3*40 70 Cotton mill . . . . tt 2 4305 65-95 3 '4 2 Cotton yarn dyeing . >? 2 11456 80-09 3'25 16 Dyeing and cleaning . "~ ;, Hand 2 5093 73-26 3-20 238 Hat manufacturers Mechanical I 7457 55-45 3*20 133 Woollen mill . . - 5j 3 4659 61-76 3-20 159 Colliery . . . . Hand 10 5586 59*76 3-18 124 Wcollen mill . . | .- Mechanical 3 6382 62-53 3-10 14 Food produces Hand 5 7652 73*96 3-10 183 Colliery .... Mechanical 2 8487 56*93 2-96 35 1 Cotton piece dyeing M 3 3257 48-11 2-/5 82 Cotton mill .... n 3 8i93 65-25 2-75 358 Colliery .... Hand 4 3472 47-50 2-70 279 Woollen mill Mechanical 2 3073 52-83 2-70 1 08 Dyeing and cleaning . t 4 4081 63-20 2-70 18 Engineering Hand 4 7609 73-02 2-60 363 Woollen mill Mechanical 2 3016 47-02 2-60 27 M Hand 6 4161 70-10 2-50 CRITICISMS OF EXISTING CODES 107 4> ^Eo ti uj _u g'c/2 rt S 1 1 1 c '5 1 'o (Q iff! ii p D Q "8 OH uJ fer 'S C Tfl ^ * 1 'O o 1 IS c n w 2 3 ^ D C >. > "S 6 3 rt > |-a ^ o V %& 304 Woollen mill Mechanical 5 6037 51-66 2'5O 214 Cotton mill . M 5 7557 56-66 2*50 169 Lace mill Hand 2 59-01 2-50 173 Woollen mill Mechanical 2 7062 58-62 2-50 164 Hat manufacturers H I 8734 5Q-54 2'5O 22O Explosives . Hand I 4165 2*49 252 Paper mill . M I 2976 54-98 2*48 > Mechanical I 7906 51-03 2-25 197 Engineering ,, 2 4735 57H7 2'IO 4 Cotton mill . j? 2 9190 79-66 2'IO 352 Colliery Hand 5 4716 48-09 2-00 369 Explosives . jj 673 46-65 2-00 222 Food products M 2 1786 56-32 i -go 4 1 Colliery Mechanical I 12900 68-24 1-70 56 Woollen mill {J I 4659 66-75 1-50 20 Paper mill . f 5 2 8059 72-11 1-30 370 Colliery Hand IO 5679 46-52 I'OO 266 Soap manufacturers )? 2 3480 53-86 I'OO 189 M |( 2 2213 57-82 I'OO 288 Woollen mill Mechanical 2 7054 52-55 o-55 86 Cotton piece dyeing it I 5057 65-06 0-50 For ordinary testing purposes the objection to the steam meter is that the steam nozzles are almost always supplied by a separate small pipe from each boiler, a most wasteful and unscientific method, which makes it difficult to determine the amount of steam used. Consequently, even if only two or three boilers are tested, the steam meter has to be coupled up to each boiler in turn, a very troublesome proceeding, although when once connected, this method is decidedly the best. In testing the 400 boiler plants, I have used the surface condenser method, and have devised for the purpose the apparatus illustrated in Fig. 17, and in the photograph of Fig. 18. This apparatus has been in use for many years, and found to be most satisfactory. In Fig. 17, A is a sheet-iron cylinder, water-jacketted, io8 BOILER PLANT TESTING with bolted lid B. The steam jet appliance, nozzles, pipes, etc., is detached from the furnaces of say one boiler, and placed bodily inside the cylinder A, being attached by a short pipe to coupling a, b, c y d, as most convenient, in the lid B. The steam jets are then connected to the boiler, the whole CRITICISMS OF EXISTING CODES 109 apparatus being stood in the firehole close up against the boiler, so that the steam is blowing through the nozzles in the cylinder A exactly as it does when attached to the boiler furnace. The steam then blows down through the pipe C and through the coil D, open to the air at E. A con- tinuous flow of cold water enters the jacket 'by the pipe F and leaves by the pipe G, and then flows into the large cooling vessel H. As a consequence the steam from the nozzles is condensed, and runs out at E, where it is weighed. In working the apparatus, the valve on the steam supply to the nozzles is opened the same as usual, and as soon as a steady stream of condensed water is issuing from E, the test is carried out at half-hourly intervals by putting under the flow at E a weighed bucket at the half-hour by the clock, and collecting all the condensed water for weighing until the next half-hour. In this way numerous tests can be carried out, under any different conditions as regards the amount of turn of the steam supply valve, boiler pressure, etc. One method (which is not to be recommended) that has been used to determine the amount of steam used by the nozzles is to place the nozzles in a weighed amount of cold water, and to pass in the steam for a given time, so that the water does not get hot enough to lose weight by giving off steam. The increase in weight then represents the amount of steam passed by the nozzles, but there is some difference of opinion as to whether the same amount of steam issues from the nozzles in the air under the furnace as it does when immersed in cold water. The " Civils " Code also mentions the method of having one separate boiler devoted to supplying only the steam nozzles, and measuring the amount of water evaporated in this boiler to determine the steam produced. This, in my opinion, is not only quite unnecessary, but means also in most cases altering all the pipe circuits for the steam nozzles, as already described. If we are going to go to all this trouble, the steam meter is obviously the best method. Further, it i io BOILER PLANT TESTING must be remembered that the efficiency of the auxiliary boiler plant may be different. (4) Mechanical Draught. This is also an important item, as mechanical draught may take anything from say I to 3 per cent, of the production. /The typical installation of induced draught driven by direct coupled high-speed non- condensing engine absorbs about 6 h.p. per " Lancashire " boiler, 30 x 8 ft, and the consumption of the engine is about 35 Ibs. of steam per i.h.p., corresponding to an evaporation of 215 Ibs. per hour, or say, in averages, 2*75 P er cent, of the production. In most cases it is sufficiently accurate to calculate the i.h.p. of the engine under average conditions in the ordinary way, but in special cases the actual steam consumption can be determined by condensing the exhaust steam or by a steam meter. If the exhaust steam is used to heat the feed-water, this heat added to the plant must of course be deducted in calculating the steam taken by the engine. When the fan is electric driven, the power can of course be obtained easily, but it must be calculated back to steam production of the plant, as if the current was generated on the spot. If the current is purchased from outside sources the cost must be calculated in the terms of coal. It is very difficult to get the figure for the h.p. when the fan is driven from a line shaft, a bad practice in any case, and often the only way is by taking all the details of the fan, revolutions per minute, size, etc., and obtaining the h.p. from a similar fan driven by steam engine or motor. (5) Various Other Uses of Auxiliary Steam. There are various other points at which auxiliary steam or power may be used, such as the boiler feed pumps, economiser scraper engine, water-softening plant, electrolytic processes for the prevention of corrosion, and so on, and in each case the actual steam taken can be calculated without trouble on the same general lines, and added to the total amount of auxiliary steam. As will be seen, all these items added up together make a very formidable figure, and place an entirely different CRITICISMS OF EXISTING CODES in complexion on the figures for the real net working efficiency of a boiler plant. 10. Lbs. of Water from and at 212 F. per 1,000,000 B.Th.U. Both Codes give the figure for the Ibs. of water per Ib. of coal, and also the Ibs. of water " from and at 212 F." per Ib. of coal. This is, however, not sufficient, and I suggest to be included in the International Code a new and additional figure, namely, Ibs. of water from and at 212 F. per 1,000,000 B. Th. U. It is of course little use to give merely the figure of Ibs. of water evaporated per Ib. of coal, since the temperature of the feed- water going into the plant may be at a temperature of anything from 32 to 212 F., and each 11 F., means a difference of approximately I per cent in the coal bill. Both Codes get over this difficulty by calculating the evaporation "from and at 212 F.," that is, taking the evaporation of water per Ib. of coal if the feed-water was always 212 F. This, however, still leaves untouched the other and equally important error, namely, the heating value of the fuel, which may vary from 7500 to 14,500 B.Th.U. per Ib. The mere figure, as given in the codes, of the evaporation " from and at 212 F." means, therefore, very little. For many years I have calculated accordingly the figure "Ibs. of water from and at 212 F. per 1,000,000 B.Th.U.," which obviates both these difficulties, and I suggest that this new figure be included in the International Code. 1 1 . Various Minor Points. (a) Both Codes insist upon reading the barometer during a boiler trial, so as to get the absolute steam pressure. Thus in the " Civils " Code (p. 78, line 34) is stated : " This (the absolute pressure) is obtained by adding the atmospheric pressure to the gauge pressure." It seems to me to be a waste of time to read the baro- meter, and for all the difference it makes in the calculations, the absolute pressure may just as well be taken as 15 Ibs. plus the gauge pressure. It certainly seems curious for the " Civils " Code to insist on including barometer readings, and ii2 BOILER PLANT TESTING yet at the same time, as already seen, taking an arbitrary figure of 0*48 for the specific of superheated steam, irrespective of its temperature. (b) The new figure for the latent heat of steam is 970-8, which is included in the American " Mechanicals " Code. In the " Civils " Code, however, the old figure of 966 is used as follows (p. 86) : " Line 48 is line 47 multiplied by the evaporation factor, and the evaporation factor is equal to the total heat required to evaporate a pound of steam under boiler conditions, divided by 966, />., it is equal to 1 ^ ." In the International Code 970 '8 should be adopted. (c) As regards temperature measurements (" Civils " Code, pp. 64-67), it is not necessary to fill the sockets with mercury, as thick oil will do just as well up to 500 to 600 F., and it is not vry convenient to take flue gas temperatures up to 600 F. with mercurial thermometers, if only because these are so fragile, even when in armoured cases. In the Inter- national Code I would suggest a list being given of approved resistance and thermo-electric pyrometers, which are much more suitable for boiler testing work. (d) The " Civils " Code (p. 37) recommends a most re- markable implement as an aid to boiler plant testing, namely, " a tool (Fig. 19) for gauging the thickness of the fire in the grate," illustrated as follows : FIG. 19. (Fig. 2 of the " Civils " Code.) CRITICISMS OF EXISTING CODES 113 The Committee who drew up the " Civils " Code seem to have been much worried as to possible inaccuracies due to different amounts of coal being in the fires, at the commence- ment and at the end of the test. It is quite possible, of course, if a boiler fire is filled up with coal at the beginning of the trial, and allowed to burn empty at the end, that a considerable error several cvvts. of coal per " Lancashire " boiler will creep in, and the plant will appear to be doing better than is actually the case. The solution of this difficulty is, however, merely the elementary one of looking at the fires when commencing, and having them the same at the end, and anyone with any common sense at all will not be in error more than say I cwt, equal to about 0*5 per cent, on the weight of the coal. It is quite easy to have an error as large as this in merely weighing the coal. Such simple methods, however, will not do for the " Civils " Code, even if the fact that 10 to 20 per cent, of the steam produced by the plant being blown away by steam nozzles is regarded as of no importance. On page 9 of the Code we have : " In most cases the principal observer will be able to measure the thickness of fuel to within I in. As the error may be in opposite senses at the beginning and ending of a trial, it may amount to the weight of a layer of fuel 2 ins. thick. If C is the weight of green coal, which will form I cub. ft. of incandescent fuel, the total error should not exceed C x A x - - = - - Ibs. Therefore if W is the number of 12 6 Ibs. of fuel per hour, A the square feet of grate covered by fuel, when its thickness is measured, and n the percentage of looCA error admissible, the trial must last hours. C may be taken as 20 Ibs. for large coal and 30 Ibs. for small slack. In making use of this formula, however, it is necessary to have some regard to the quality as well as the size of the fuel. When it contains much dirt or makes a pasty clinker, the bars, if not self-cleaning, have to be cleaned at short intervals by the firemen, and at each cleaning there is loss of heat and combustible matter. The duration of the trial and the times ii4 BOILER PLANT TESTING of cleaning should therefore be so arranged as to give this loss the same average value that it would have if the trial were in- definitely prolonged. For instance, if the fuel were such as to make cleaning necessary every 4 hours, it would be unfair to make a 5-hours' trial.; 8 hours would be the proper time ; or, if it were not possible to have the trial longer than 5 hours, a more accurate result would be obtained by working for 4 hours only and cleaning the fire-grates only once." Accordingly, we are recommended to use a " tool," a huge steel poker, i to i-J- in. diameter, and presumably 8 or 9 ft. long, as illustrated in Fig. 19. When the trial starts we have apparently (p. 13) to open the fire-doors, and in the blinding heat, rummage about in each fire taking the thickness at various places, and recording the same. This procedure is to be repeated at the end of the trial, and presumably a calculation made for different thicknesses of the fires before and after. It is not stated whether another calculation will also have to be made to compensate for the loss of efficiency caused by having the fire-doors open and allowing cold air to enter whilst the measurement is in progress. Taking, for example, the case already mentioned, of a plant of thirty-two " Lancashire " boilers, do we understand that before com- mencing a test, which can be of 3 hours' duration (" Civils " Code, p. 9), it is necessary to insert this " tool " in sixty-four different furnaces ? I should estimate that each determination would take at least 4 minutes, and that after about four fur- naces one man would be exhausted, so that with relays of men, the measurement of the thickness of the sixty-four fires would take, say, 4 to 5 hours, or longer than the duration of the test allowed. Or perhaps we have in such cases to have a whole battery of " tools ". It would be very interesting to know if any member of the " Civils" Committee has ever tried to measure the thickness of a fire with the " tool," which he joins in recommending for this purpose. (i) On page 7 of the " Civils " Code there is stated : " If it is desired to ascertain whether cold water is lodging in the bottom of Lancashire or other boilers of similar type, a CRITICISMS OF EXISTING CODES 115 horizontal tube should be screwed into the front part of the boiler, or preferably the front manhole door. The stopped end of the tube should be carried sufficiently far back to avoid the possible lodgment of cool water at the front end which may arise from the bottom flue not extending right up to the front of the boiler." What this means is not clear. 12. The Method of Calculating the Results. The methods given in both the Codes for calculating the results of a boiler test are, in general, so completely involved and com- plicated, that it is impossible to criticise them in detail in any reasonable length. It seems to me that the fundamental error in these methods is the attempt at all costs to evolve a " heat balance sheet," which, in my opinion, is not necessary, and in any case is inaccurate. In dealing with this difficult subject, I propose, for the sake of simplicity, to describe first the method of calculation which I suggest be embodied in the International Code. I hope to show that this method is quite simple and practical, that there is no particular mathematical knowledge required at all, and that the complete figures of an official trial can be worked out in twenty minutes on first principles without any empirical formulae. In this connection the " Civils " Code especially falls into the error of giving ready-made formulae and symbols, which means that most engineers use these in a rule-of-thumb way without understanding the principles in- volved, and errors are bound to result. Taking again the specimen test results on page 146, the es- sential figures, apart from simple averaging, division, etc., which require no consideration, are that 105,328 Ibs. of water at I2IF. are evaporated on the plant, the average temperature after the economiser being 296 F. and the temperature of the super- heated steam averaging 475 F. with 147 Ibs. gauge pressure (162 Ibs. absolute). At the same time 15,960 Ibs. of coal with a net calculated heating value of 11,715 B.Th.U. per Ib. are burnt, and the total amount of auxiliary steam (or power expressed as steam) is 13,639-9 Ibs. (12-95 per cent.). n6 BOILER PLANT TESTING The essential figure required, to which all other figures are merely subsidiary, is the net working efficiency of the com- plete steam generation plant, that is to say, for every 100 Its of fuel delivered to the plant, how many Ibs. of fuel are actually used for the production of useful steam, and how many Ibs. are wasted. Thus, when we say that the average net working efficiency of the boiler plants of Great Britain is 60 per cent, we mean that out of every 100 Ibs. of coal burnt, 60 Ibs. are used to produce useful steam and 40 Ibs. are wasted in radiation, imperfect combustion, leaky brickwork, loss of heat in the ashes, auxiliary steam used on the boiler plant itself and so on. We have, therefore, in the calculations, to determine the actual number or British Thermal Units of heat given to the plant in the fuel, and the actual number of British Thermal Units of heat coming out of the plant as useful steam. By the steam tables we find that the total heat required to convert I Ib. of water at 32 F. into steam at 162 Ibs. absolute is 1 194 5 B.Th.U. The feed- water entering the plant is, however, 121 F., that is to say, less heat is required than 1194*5 to convert I Ib. of water at 121 F. to steam at 162 Ibs. absolute. The " Civils " Code assumes that the specific heat of water is the same at all temperatures, that is to say, the British Thermal Unit of Heat is the amount of heat required to heat I Ib. of water i F. irrespective of the initial temperature of the water. Or, in other words, that it takes the same amount of heat to raise the temperature of I Ib. of water from, say, 32 to 33 F. as it does to raise it from, say, 200 to 201 F., and therefore the amount of heat required to convert I Ib. of water at any initial temperature (/) above 32 F. into steam at a given pressure can be got by subtracting 32 from (/), calling the figure B.Th.U., and then subtracting it from the total heat figure from 32 F. In the example, therefore, according to the " Civils" Code, the total heat from 121 F. is 121 - 32 = 89, subtracted from 1194-5 = 1105-5. This, of course, is erroneous, and the specific heat of water is different for every CRITICISMS OF EXISTING CODES 117 F. The heat required to raise I Ib. of water from 32 to 33 F. is a maximum, and the specific heat very slightly falls with the temperature to about 130 F. and then rises again as shown on the curve, Fig. 20. Thus, for the given example of 121 F. the amount of heat to be subtracted is not 121 - 32 = 89 B.Th.U., but actually 88 B.Th.U. This is not of course a very serious matter, but as it is no more trouble to be accurate, I suggest that in the International r* I'lSo rioo SPECIFIC HEAT- I'ooo FIG. 20. Curve showing the specific heat of water at different temperatures. Xr Code the formula /& = foi/f- 35 is used in calculating the heat (//) at a given temperature (f) to be subtracted from the total heat at 32 F. This formula is quite simple, and gives a very near approximation indeed to the mathematical curve. Thus at 121 F. h = (1-017 x 121 F) - 35 = 123-057 - 35 = 88 B.Th.U which subtracted from 1194-5 gives the figure of 1106-5. ii8 BOILER PLANT TESTING In the same way for the economiser calculation, the total heat from 296 F. is (296 x 1-017 - 35) = 266-0 B.Th.U., so that the total heat from 296 F. is 1194-5 - 266 = 928-5 B.Th.U. Also, the steam is superheated to 475 F., that is 475 ~ 364-4 (saturation temperature from the steam tables) = 1 1 06 F., and from Knoblauch and Jakob's Superheat Tables the specific heat of steam at 475 F. and 162 Ibs. absolute is 0-54. 1 10-6 x o 54 = 597, so that the amount of heat required to raise I Ib. of water at 121 F. to steam at 162 Ibs. absolute, and superheat it to 475 F. is 11065 + 597 = 1166-2 B.Th.U. This gives us all the figures necessary for the amount of heat absorbed by I Ib. of water under the different conditions of the boiler, economiser and superheater. In the test 105,328 Ibs. of water required 15,960 Ibs. of coal, so that I Ib. of water = 5 = 0*15152 Ibs. of coal. 105 320 Also, since the net heating value of I Ib. of coal is 11,715 B.Th.U., 0-15152 Ibs. of coal will contain 11715 x 0-15152 1 775'O B.Th.U., which represents, therefore, the actual heat put into the plant. The net percentage of heat taken by the different essential portions of the plant is now as follows : (1) Boiler Only. Since the amount put in the firehole was 1775-0 B.Th.U., and that absorbed by the boiler is 928-5 B.Th.U. (i.e., heating the water from 296 F. to saturation at 162 Ibs.), the percentage of heat in the original coal absorbed , 928-5 x 100 by the boiler is - = 52-31 per cent. (2) Economizers Only. Again, since the heat put in the firehole is 1775-0 B.Th.U., and that abstracted by the boiler and economisers is 1106-5 B.Th.U. (i.e., heating the water from I2IF. to saturation at 162 Ibs.), the percentage of heat absorbed by the boiler and economiser is 1 106-5 x IPO _ 62 . 33 per cent Since the boiler was 52-31 per cent, the economiser only will be 62 -3 3 - 52-31 = 10-02 percent. CRITICISMS OF EXISTING CODES 119 (3) Superheater Only. Again, as before, taking the original figure of I775'O B.Th.U., the amount of heat ab- sorbed by the boiler, economiser and superheater is 1 166*2 B.Tn.U. (i.e., heating the water from 121 F. to saturation at 162 Ibs. and superheating to 475 F.), the percentage of heat absorbed by the boiler, economiser and superheater is 1166-2 x 100 - - - = 65 70 per cent., and as the figure for the boiler and economiser together is 62-33 P er cent., the figure for the superheater only is 6570- 62*33 = 3-37 per cent. (4) Net Working Efficiency. We have seen, therefore, that for 100 Ibs. of coal put in the firehole 52-31 are absorbed by the boiler, 10-02 by the economiser, and 3-37 by the super- heater, a total of 6570, the other 34*30 parts being wasted by radiation, inefficient combustion, leaky brickwork, etc. This figure of 6570 per cent, is not, however, the net working efficiency of the plant, because although 6570 per cent, of the heat of the coal is absorbed by the boiler, economiser and superheater, this heat is not given out of the plant to the factory as steam. 12-95 P er cent - f it: * s taken up by the plant itself, as already seen, for steam jets, induced draught, boiler feed pump, etc. The real net working of the plant is 6570 x 12-95 therefore - - - = 8-50 - 65 70 = 57'2O per cent. (5) From and at 212 F. Calculation. In order to cal- culate the Ibs. of water evaporated per Ib. of coal, assuming the water was 212 F. (instead of actually 121 F.), all that is necessary to do is to multiply the Ibs. of water at 121 F. per Ib. of coal, that is, ^- = 6 -60 x 1106-5 (the total heat from 121 F. to saturated steam at 162 Ibs.), and divide by the latent heat of steam (970-8). That is - - =7'5i Ibs. of water from and at 212 F. This figure can be checked by the " factors of evaporation " figure from any engineering pocket book. 120 BOILER PLANT TESTING To get the " evaporation from and at 212 F. per 1,000,000 B.Th.U.," if 7-52 Ibs. from and at 212 F. are evaporated per Ib. of coal, the heating value of which is 11,715 B.Th.U., then 11,715 B.Th.U. is equivalent to 7-52 Ibs. and 1,000,000 is equivalent to 7'52 x 1,000,000^ 11715 (6) Saving in Coal Bill Due to Economizers. In the gross efficiency figures, the boiler and ecortomiser is 62-33 per cent, and the economiser only is IO'O2 per cent, so that expressed as a percentage the saving in the coal bill is IO'O2 x TOO = 1 6- 1 per cent, which can also be checked by empirical tables found in most pocket books, or econo- miser makers' catalogues. (7) Saving in Coal Bill Due to Superheaters. In the same way the gross efficiency of boilers, economisers and superheaters is 6570 per cent, and the superheater only 3-37 per cent, so that the saving in the coal bill due to the super- heater is 3 ' 37 X IOQ = 5-1 per cent 6570 These figures are the essential ones for the true performance of any boiler plant That is to say, for every ico Ibs. of coal put into the plant, 57*20 Ibs. are being used to produce useful steam, the other 42-8 Ibs. being wasted, 8-5 Ibs. due to auxiliary steam and 34-3 parts to various losses not yet analysed. One of the losses is radiation from the whole body of the plant. In my opinion, it is not essential to determine this experimentally, and I would not propose to include it as part of the International Code. It can, however, be determined if required by running a test on the plant for a number of hours, the longer the better, with the main steam pipes to the factory shut down and supplying just enough coal to the plant to main- tain the full boiler pressure without blowing off. The amount of coal burnt in comparison with that of the full working trial will give the radiation loss, which may correspond to anything CRITICISMS OF EXISTING CODES 121 from 5*0 to lO'O Ibs. for every 100 Ibs. of coal taken by the plant The " Civils " Code (p. 53, also p. 79) seems to regard this as an essential part of the test, but the method proposed is not at all clear, and as far as can be made out, seems to apply to the boiler only, and not to the whole plant. Of course, the radiation loss should include the boiler, econo- miser, superheater and all accessories. The other sources of loss are faulty firing, that is, excess or deficiency of air passing through the fires, cooling of the plant due to leakage of cold air, heat losses in the ashes, and insufficient heat absorption by the boiler, superheater and economiser, that is too high an exit temperature in the chimney base. The "Civils" Code, and to a lesser extent the American " Mechanicals " Code, get in to a terrible tangle in calculating these losses, and, in fact, approach the whole of the calculations on the assumption that these losses, and not the amount of heat put into the plant, are the most important part of the calculations. It is stated in the " Civils" Code (p. 5) that the measurement of the losses are " a valuable check on the accuracy of a trial," but my point is that in any case the loss figures are not accurate, even if they were necessary. For example, we have in the " Civils " Code (p. 78, 1. 23, pp. 80-85, 11. 39-40, p. 86, 11. 51, 52, 54, 55, 56, also pp. 88-90) the particulars dealing with the amount of heat passing away from the plant, and the heat carried away by the excess air, in half a dozen pages of complicated calculations. The basis is the full analysis of the flue gas (CO 2 , CO, O and difference) to represent all the gases leaving the plant during the test, and then the calculation of this volumetric gas analysis to parts by weight. It is then assumed that all the carbon in the coal appears in the flue gas as CO 2 and CO, and following on a complete analysis of the coal for the per- centage of carbon, hydrogen, etc., it is calculated how many Ibs. of dry flue gases leave the boiler flues per Ib. of fuel burnt. This calculation is based on dry flue gases, b.ut 122 BOILER PLANT TESTING they are not dry as they contain, as already seen, an enormous amount of water, equal to 8*95 times the weight of the hydrogen originally present in the fuel, together with the natural moisture, but making allowance for this we arrive at the figure for the weight of air* drawn into the flues per Ib. of fuel. By another calculation on the percentage of carbon, hydrogen, etc., in the coal burnt, the theoretical weight of flue gases is calculated, and in this way the excess air is determined. In order to try and reduce the labour of these calculations an empirical formula is recommended (p. 85), namely, Heat carried away by dry gases _ Constant x a x (T - t) per Ib. of fuel burnt C Where a = percentage of carbon by weight in the fuel after correction for unburnt material in the ash. C = percentage of CO 2 . With regard to the constant, this is explained as follows : "The constant for flue gas at about 500 F. may be taken as O'6a5 and the error will not exceed I per cent, on the heat balance, which is well within the limits of experimental error." With regard to the question of the loss in heat in the hot ashes (p. 86, 11. 42, 42^), this is, in my opinion, a matter of little importance. The " Civils '' Code has to assume that the temperature of the hot ashes as withdrawn from the fires is 2000 F. and the specific heat of the ashes is 0-3, and all these assumptions reduce the " heat balance sheet " to a farce. In order, therefore, to make this calculation for the heat pass- ing up the chimney, we have to make a complete and laborious chemical analysis of the fuel, in addition to the heating value by the bomb calorimeter, an analysis of the ash, an analysis of the flue gases, and a whole series of complicated calculations, and the results when obtained are based largely on assump- tions. I would suggest that in the International Code all this be cut out as not essential. The information obtained by the CO 2 Recorder will give us without any trouble, by the curve, Fig. 8, a figure for the CRITICISMS OF EXISTING CODES 123 loss in heat due to faulty firing, which is practically that obtained at so much expense and trouble by the method of the "Civils" Code. Thus, for example, in the specimen trial the figure for CO 2 is 6'O per cent., and by the curve this corresponds to 1 6 per cent, loss in coal, taking 14 per cent, as the maximum CO 2 obtainable in practice. Another extraordinary feature of the "Civils" Code is that the whole of these laborious calculations have to be gone through all over again for the economiser, and yet a third time for the superheater. I am at a loss to understand what is the reason for suggesting that the economiser and superheater should be regarded, from the point of view of gas analysis, as separate from the boiler, and the essential figures for the economiser and superheater can be worked out in a few seconds by the method suggested on page 115. And to fur- ther add to the confusion, the " Civils" Code then proceeds to give (p. 80) an entirely fresh and empirical formula for calculating out efficiencies without the aid of steam tables, in spite of the fact that anyone can buy steam tables for a few pence, and the method suggested is admittedly not accurate. 124 ' PART III. SUGGESTIONS FOR NEW FEATURES\VHICH MAY BE ADDED IN THE FUTURE TO AN INTER- NATIONAL CODE AS THE RESULT OF FURTHER DISCUSSION AND INVESTIGA- TION. i. The Question of the Use of a Special Factor, Depend- ing on the Quality of the Fuel, to be Used in Calculating the Net Working Efficiency of the Plant. Up to the present time, in most boiler plant tests the efficiency has been calculated simply from the actual amount of heat present in the fuel, and no allowance has been made for different qualities of fuel, and for the different theoretical efficiencies possible. For example, if one plant is using the finest washed nuts with, say, 3 per cent, ash and 14,000 B.Th.U. per lb., and another plant is using merely refuse coal with, say, 35 per cent, ash and 7000 B.Th.U. per lb., and no auxiliary steam is used in each case, the efficiency is calculated in the same way in both instances, that is, on the amount of heat actually present in the coal. Thus, the first plant may be working at 72 per cent, net working efficiency and the other at 59 per cent, and according to the methods of calculation generally adopted, the first plant is regarded as doing very much better than the second. Stated in this way the results are completely misleading, because the fact is ignored that with the good coal the price may be, say, 2 los. per ton, and the possible efficiency 80 per cent, and with the inferior coal and the same amount of skill and attention, the efficiency can only be 65 per cent, but the price is 1 55. per ton. SUGGESTIONS FOR NEW FEATURES 125 The reason is, of course, that with inferior coal the per- centage of ash is so great that, even with mechanical stokers, it is not easy to prevent excess air passing through the fires, it is much more difficult to burn the coal with the evolution of radiant heat, and an excess of heat is lost in the ash. This is, however, very unfair to the plant using inferior coal, both scientifically, as well as from a practical and business point of view. For example, to look at the matter in another way, take a case of a plant burning 200 tons a week of the inferior coal at 255. per ton, with a net working efficiency of 59 per cent., and with an annual coal bill therefore of i 2,500. If expensive coal of 14,000 B.Th.U. at 2 los. per ton is substituted, it is quite true that the efficiency is increased to 72 per cent, and the amount of coal burned is only 163 tons per week, yet the net result of this is to increase the fuel bill of the factory from 12,500 to 20,375 P er annum. Accord- ing to the usual methods adopted, however, as in both the Codes, the plant is now doing better, in spite of the fact that the practical result would be to lose 7875 per annum ! This, of course, is an absurdity, and I have in the past tried to get over this difficulty to some extent by always giving the cost of evaporation of 1000 gallons of water, although this has the defect that it is dependent on prices of fuel, which vary con- siderably in different neighbourhoods. I would suggest, how- ever, that this item be included in the International Code. Some makers of appliances for steam generation understand the point very well, and many of the remarkable results that are published as having been obtained with various appliances will be found, on investigation, to be really due to the fact that especially good quality coal has been used. If such coal had been used on the original plant, the same results might have been obtained without the appliance at all. Thus, it is a favourite method to take, for example, a hand- fired plant with natural draught, and simple fire-bars burning average coal of moderate price and quality, arid to fit on some appliance to burn cheaper fuel with the object of showing a 126 BOILER PLANT TESTING saving. The only fair and reasonable method of proceeding is to burn the same quality of coal as before, and if further trials are carried out with cheap refuse coal or expensive good quality coal, then to have at the same time analogous trials with these coals on the plant as originally" working. It is amazing the number of tests published that trust to the steam user to ignore these elementary facts. A given furnace stated, for example, to save, say, 30 per cent, of the coal bill, will be found on inves- tigation to have been tested with cheap refuse coal, of which only a limited supply is available, as against average priced coal on the original plant. If, in the later case, the same re- fuse coal had been used, whilst the results might not perhaps have been as good as with the special furnace, the real saving would have been about 10 per cent, instead of 30 per cent. When this is pointed out the reply is usually, as I know from experience, that the test has been carried out according to the usual practice, and sufficient allowance has already been made in calculating the efficiency from the actual heat value. On pressing the point, however, refuge is then as a rule taken in the "Civils" Code, in which no allowance has been made for different qualities of fuel, although in the test in dispute no attempt has been made to carry it out according to this Code. What is required is that there should be added to the ordinary calculated efficiency a further quantity X, which would vary according to the heating value and quality of the fuel. The value of X would be expressed as a curve, which I would suggest calling the Standard Curve cf Efficiency Cor- rection for the Diminishing Heat Value of the Fuel, so that the value of X would increase as the value of the fuel decreased. This would put all boiler plants on a real comparative basis, and give proper credit to the man who is burning a cheaper and inferior fuel, and doing a national service as well. Such a curve could only be obtained experimentally, and this I suggest would be one of the points of investigation for the International Committees. SUGGESTIONS FOR NEW FEATURES 127 What would be necessary would be for, say, twenty or so thoroughly representative coals to be taken, from the highest to the lowest quality, and a complete series of experiments carried out under various conditions of steam generation on, say, " Lancashire," " Marine," and " Water-tube " boilers with, for example, (i) rated evaporation, (2) 20 per cent, overload, (3) 20 per cent, and (4) 50 per cent, below rated load, both with hand firing, and different types of mechanical stoker. If a whole series of trials were carried out in this way, on a well equipped experimental plant, sufficient data would be obtained to elaborate such a curve with a considerable amount of accuracy. Based on the experience of over a thousand tests, I would suggest the curve as given in Fig. 21 (next page). I do not pretend that this curve is accurate, because, as already stated, it would need to be based on much experimental work, but it is of interest as illustrating the principle. Thus, in the Specimen Test, the value of X with coal of 1 1 ,7 1 5 B.Th.U. would be 5 -o, and this would then be added to the figure of 59-4 per cent, for the net working efficiency. The result, 64-4 per cent, which would be called the C.D.H.V. net working efficiency (corrected for diminishing heat value). In the example just mentioned the plant with 14,000 B.Th.U. coal, and 72 per cent efficiency, would have added the figure of 0*5, making 72-5 per cent, efficiency; and the plant with* 7000 B.Th.U. and 59 per cent, efficiency would have the figure of 23-5, making 82-5 per cent. The latter plant would, therefore, be much the best by calculation, as it is in practice. 2. Labour, Attendance, Repairs, Upkeep, Interest and Depreciation. Another difficulty is that of the cost of work- ing the plant, quite apart from the fuel bill. If, for example, a plant of ten " Lancashire " boilers has only four men per shift looking after it, together with reduced labour at night, and a wage bill of, say, 20 per week, working on 60 per cent, efficiency, and the plant is then reorganised to work on an efficiency of 75 per cent, but the wage bill is increased to 128 BOILER PLANT TESTING ,30 per week, this extra 500 per annum ought to be de- ducted in calculating the net result. In the same way, if /9ooo /eooo (fcooo <5ooo \ 13000 I2ooo II 000 \ \ \ \ \ s s \ *\ \ Sooo 8000 X, s s \ s x ^ x X, X ' X X, 6