PREFACE.

The fuel tests discussed in this paper were made under two Heine water-tube boilers by the boiler division of the United States Geological Survey fuel-testing plant at St. Louis, Mo. These tests began during the Louisiana Purchase Exposition, in 1903, and have been in progress since that time. The coals tested have been collected from all the prominent fields of the United States.

It has been the object to compare fuels by determining the evaporative performance of the boilers when using the various coals. Careful and complete observations have been made with each coal, so that it should be possible for engineers to determine from an examination of the records whether the conditions of the tests were favorable or unfavorable for the character of the coals tested.

The results of tests by the several divisions of the fuel-testing plant that have been published thus far appear in (a) Bulletins Nos. 261 and 290 of the United States Geological Survey, giving preliminary reports of the operations of the plant; (b) Professional Paper No. 48 of the Survey, giving complete and detailed information of the tests made during the Exposition.

The steaming tests were in charge of D. T. Randall during the Exposition period and of Walter T. Ray since that time. These men have been ably assisted by a corps of trained observers and computers, who have been alert, patient, and painstaking.

As soon as a considerable number of tests were available, it seemed desirable to begin a comparison of the various tests on the basis of such usual factors as over-all efficiency, rate of combustion, temperature of escaping gases, evaporation per square foot of heating surface, etc. It soon developed that many interesting and apparently valuable results were brought out by these comparisons, and hence they were extended. The discussions of facts led to a search for theories, and the men of the division sought out available literature bearing on the various subjects under discussion. Thus has accumulated the material here presented. It is not the work of any one, but of all.

To Messrs. Walter T. Ray, Henry Kreisinger, and Harry W. Weeks should be accorded especial credit for preparing the larger part of this material. Except for the cooperation and careful work of the chemical division, under the supervision of Dr. N. W. Lord, much of the work here reported would not have been possible.

Many of the relations submitted have been worked out at the suggestion of different engineers who were interested in particular phases of the work. Some repetition in the matter presented is perhaps unavoidable on account of the nature of the subject. Possibly much that is new has developed during the discussion, because never before have 400 boiler tests been made under two boilers (exactly alike) with coals and lignites from all sections of the United States and with all tests fired by the same expert fireman.

Acknowledgment is gratefully made to the many engineers of the country for helpful suggestions and kind criticisms that have come to me while endeavoring to direct the tests of this division so that they would prove of the greatest value to the industries of the nation, and to those engineers whose problem is still to furnish boilers of greater capacity and furnaces of higher economy. It must not be forgotten that all discussion is based on the results obtained in a single furnace with one type of boiler, but many of the results should have wide application in any furnace and under any type of boiler. Helpful suggestions or criticisms from manufacturers or engineers are again invited.

In submitting this paper I desire to thank Dr. Joseph A. Holmes, expert in charge of the fuel-testing plant, for his uniformly courteous consideration of the plans and needs of the boiler division and for the freedom which he has always allowed us in carrying on the work. We have appreciated his confidence and can only hope that it has been merited.

A STUDY OF FOUR HUNDRED STEAMING TESTS.

By LESTER P. BRECKENRIDGE.

INTRODUCTION.

By WALTER T. RAY.

Origin of the bulletin.-As the so-called "boiler tests," made by the fuel-testing plant at St. Louis, gradually accumulated, it became evident to the men making them that a great deal of incidental work was assuming the shape of a considerable compilation of plottings and calculations of various matters pertaining to the tests, and that certain details of the work were beginning to shed light on the inner reasons for the unreliability of tests of coals made under steam boilers. Much research work was being done in the boiler room simultaneously with the steaming tests proper, and of this work also certain details shed light on the same unreliabilities.

It was originally intended to insert these results in the regular bulletins of the fuel-testing plant, but the further the work went the less of a basis for general conclusions it seemed to be. For this reason publication was postponed until a large amount of matter was on hand. In order to avoid undue bulk, no account of most of the work itself is given. Many of the negative results obtained are presented, however, on the general principle that negative results are usually worth as much as positive results.

Scope of the work. An especially striking instance of the value of negative results was noted in an attempt to check up the readings of an optical pyrometer used on the combustion chamber by calculating the initial temperatures of the gases entering the boiler in each test, so as to find whether the pyrometer read too high or too low when sighted on flame. The basis used for the calculation was the efficiency noted as "72*, efficiency of boiler, per cent" in the code of the American Society of Mechanical Engineers. The calculated temperatures of combustion were ridiculous, and were very puzzling for some weeks. Meanwhile a study was made of some chapters in John Perry's book, "The Steam Engine and Gas and Oil Engines," relating to the manner in which fluids give up heat and momentum, and

1

making application of these principles to steam boilers. A development of these ideas and their application to the apparatus used in these tests constitutes what is believed to be the most valuable part of this bulletin. It is thought that a fairly trustworthy separation has been effected for most tests of the true boiler efficiency and the percentage of completeness of combustion. Apparently the former is constant, as the theory indicates, and the "practical" boiler efficiency differs from it by only a very few per cent, the main cause of the difference apparently being that as a rule all of the fuel is not burned.

All of this work has been done with the hope of discovering new guiding principles for use in forecasting the results of burning any given fuel more accurately than can now be done from chemical analyses and observation of physical characteristics. It is true that at present the choice of coals in any market for any given use can be narrowed down to perhaps 20 per cent of the coals considered by a consideration of market price, freight charges, chemical analysis, heating value, and physical characteristics; but the ambition of the boiler division is to work toward that completeness of knowledege whereby about three-fourths of the 20 per cent can also be discarded after laboratory investigations.

It is also the intention of the boiler division to study the principles of fuel combustion to such advantage that much coal can be saved, and that cheap fuels can be used in certain operations where higherpriced ones are now considered necessary. Such savings will really redound to the benefit of producers of the best fuels of all grades quite as much as to the benefit of consumers. A number of new lines of work are suggested in the text; several of these it is planned to take up soon, and some of them may give data for determining the burning qualities of fuels independently of the apparatus in which the process takes place.

Perhaps the most important reason for doing the research work, and for making so many comparisons of the data obtained from all tests, was the feeling that the testing of fuels under boilers is at best only approximately comparative. In short, it was the feeling that we were on the wrong track, without knowing of a better one, which led us to do the research work.

This work was performed and the results were prepared for publication under the direction of Prof. L. P. Breckenridge, engineer in charge of the boiler division, by the following men located at St. Louis: Walter T. Ray, Henry Kreisinger, Harry W. Weeks, Charles H. Green, Robert H. Kuss, Loyd R. Stowe, William M. Park, Ralph Galt, Fred O. Pahmeyer, Raeburn H. Post, George S. Pope, Gilbert E. Ryder, Fred J. Bird, Carl J. Fletcher, and Perry Barker.

Many of the opinions given in this volume are not shared by all the authors; some of them represent a compromise. The responsibility of selecting materials and deciding what views should be taken on many points naturally fell on Mr. Ray, who was the engineer locally

in charge of steaming tests at St. Louis, and to him is to be charged the fault of any unjustified expressions of opinion or mistakes in text or figures.

It is realized that future work may show that some of the opinions advanced herein are faulty, and consequently the right is reserved to change them. Nevertheless, this work as a whole expresses the best present knowledge of the boiler division, and is given out for others to think about. It should be distinctly stated that the chemical division of this plant is in no way responsible for anything herein, although its members have been consulted on some points and have very kindly made many suggestions.

Special attention is called to the "General conclusions," page 174, and to the paragraphs entitled "Commercial considerations," page 177.

RELATIONS OF TEST DATA.

RATE OF COMBUSTION.

Fig. 1 should be considered in connection with fig. 2, the same tests being classified in both charts according to the pounds of dry coal burned per square foot of grate per hour. The figures in the small circles indicate the number of tests considered in determining the position of the points near which they are placed. In fig. 1 curve No. 1 is high for low rates of coal consumption, because it is hard to keep down the air supply. Curve No. 2 shows the per cent of stack loss. It drops at first because the air supply increases. (See curve No. 1.) It rises on the right because the gases leave the boiler at higher temperature, as shown by curve No. 3. The coals of the tests averaged on the left of the chart are in general of good quality and slower burning than those in the center; those on the right are somewhat lignitic in character. This perhaps explains why curve No. 3 is low in the center; the combustion-chamber temperatures were lower there, hence the flue temperatures were also lower. But, after all is said, the reason is in doubt, because the maximum flue-temperature range is only 50° F.

On comparing the flue-temperature curves of figs. 1 and 3 a difference will be seen in the left-hand ends. With reference to this curve in fig. 1, the coals toward the left end become better in quality and higher in "fixed carbon," so that although they burned slowly the combustion-chamber temperatures were fairly high, and consequently the stack temperatures were higher than in the center of the curve. With reference to the analogous curve of fig. 3, the coals to the left. are both poorer and slower burning, and a larger air excess was used, relatively, which made the combustion-chamber temperature lower; consequently the flue temperature was lower. Further, the lower curve of fig. 1 is perhaps drawn a little too high at the left end, and that of fig. 3 a little too low.