14. 6S: cop 25 STATE OP ILLINOIS DEPARTMENT OF REGISTRATION AND EDUCATION DIVISION OF THE STATE GEOLOGICAL SURVEY FRANK W. DeWOLF, Chief Cooperative Mining Series BULLETIN 25 GAS PURIFICATION IN THE MEDIUM-SIZE GAS PLANTS OF ILLINOIS BY W. A. DUNKLEY, State Geological Survey Division and C. E. BARNES, Engineering Experiment Station ILLINOIS MINING INVESTIGATIONS Prepared under a cooperative agreement between the Illinois State Geological Survey Division, the Engineering Experiment Station of the University of Illinois, and the U. S. Bureau of Mines [Printed by Authority of the State of Illinois] URBANA, ILLINOIS 1920 ILLINOIS MINING INVESTIGATIONS Cooperative Agreement GAS SECTION The difficulty, due to war conditions, of obtaining adequate and re- liable delivery of eastern gas-coal and of coke has suggested the wider use in gas manufacture of li$w-sulphur coal mined in the central district, com- prising Illinois, Indiana, and western Kentucky. The needs of the gas industry, and the desire of the U. S. Fuel Ad- ministration to meet those needs, has led to the appointment by Governor Frank O. Lowden, of a Technical Committee on Gas, By-products, and Public Utilities, to act in an advisory relation. The committee includes representatives of the Illinois Gas Association, the U. S. Bureau of Mines, the Engineering Experiment Station of the University of Illinois, and the State Geological Survey Division of the Department of Registration and Education, State of Illinois. Previously, some studies of the use of Illinois coal in retort-gas manu- facture and in by-product coke ovens, and of the chemical and physical properties of Illinois coal, have been conducted under the Illinois Mining Investigations, cooperative agreement — a joint agency of the U. S. Bureau of Mines, the University of Illinois, and the State Geological Survey Divi- sion. The continuation and expansion of this work has been recommended by the Technical Committee and the Fuel Administration. In response a Gas Section has been created, and experienced gas engineers, chemists, and other specialists have undertaken a program of experiment on a commercial scale to extend the use of central district coal in water-gas generators and in gas retorts. The results of the investigations will be published, and, in addition, thp nnprqtnrs nf eras plants in the region naturally tributary to central dis- used by the Technical Committee, of the progress from ill be urged to witness and participate in the tests and ir own plants new or improved practices which will fi the railroads, and assist the mines and coke ovens to ited demapds due to the war. suggestions regarding the gas experiments should be ction, Room 305, Ceramics Building, Urbana, Illinois. 3 3051 00006 3994 STATE OF ILLINOIS DEPARTMENT OF REGISTRATION AND EDUCATION DIVISION OF THE STATE GEOLOGICAL SURVEY FRANK W. DeWOLF, Chief Cooperative Mining Series BULLETIN 25 GAS PURIFICATION IN THE MEDIUM-SIZE GAS PLANTS OF ILLINOIS BY W. A. DUNKLEY, State Geological Survey Division and C. E. BARNES, Engineering Experiment Station ILLINOIS MINING INVESTIGATIONS Prepared under a cooperative agreement between the Illinois State Geological Survey Division, the Engineering Experiment Station of the University of Illinois, and the U. S. Bureau of Mines [Printed by Authority of the State of Illinois] URBANA, ILLINOIS 1920 STATE OF ILLINOIS DEPARTMENT OF REGISTRATION AND EDUCATION DIVISION OF THE STATE GEOLOGICAL SURVEY FRANK W. DeWOLF, Chief Committee of the Board of Natural Resources and Conservation Francis W. Shepardson, Chairman Director of Registration and Education Kendric C. Babcock Representing the President of the Uni- versity of Illinois Rollin D. Salisbury Geologist ILLINOIS PRINTING CO., DANVILLE, ILL. 1920 37159 — 2m CONTENTS Purpose and scope of the investigation 7 Inspection of gas plants 7 Nature of data collected 8 Summary of conditions observed '. 8 The purifying process 11 Factors affecting gas purification 11 Formulas for gas purifiers 12 Design of equipment 12 Load factor 12 Sulphur content of gas 14 Capacity of auxiliary equipment 15 Purifier operation 17 Uniformity of load 17 Temperature control 18 Revivification 19 Outdoor 19 In place 19 Small percentage of air with the gas 20 Air blown through oxide in off-box 20 Reversal of gas flow and rotation of boxes 21 Chemical control and records 22 Quality of oxide for gas purification 25 Types of hydrated iron oxides 26 Tests of oxides 26 Activity and capacity of oxides 28 Conditions found in Illinois plants 28 Equipment conditions 29 Purifying equipment in individual plants 30 Summary of capacity and load conditions 37 Effect on purifier capacities of the use of Illinois coals 41 Observed relation of oxide volume to purifier capacity 42 Rearrangement of equipment to increase caj acity 44 Results obtained in plants inspected 46 Causes of low efficiencies 50 Overload 50 Tar in the gas 51 Methods of revivification in use 55 Lack of tests and records 58 Cost of purification 5 { ) Conclusion 62 Appendices A. The Steere formula for gas purifiers 64 B. Sample record forms and computations for keeping account of purification and the performance of oxide batches 65 C. Formula of Fulweilcr and Kunbcrgcr for determining further usefulness of oxide batches 68 TABLES PAGE 1. Purifier load condition in medium-size water-gas plants of Illinois 38 2. Purifier load conditions in mixed-gas plants of Illinois 39 3. Purifying results obtained in water-gas plants inspected 48 4. Purifying results obtained in mixed-gas plants insp ectcd 49 5. Distribution of tar through beds of oxide in tv\o plants 52 6. Tar extraction apparatus in use in various plants 54 ILLUSTRATIONS FIGURE 1. E^ect of hydrogen sulphide content of gas on purifier capacity 13 2. Distribution of load in a typical water-gas plant 16 3. Observed relation of oxide volume to hourly i urifier ( a] acity 43 4. Effect of tar upon hydrogen sulphide absorj tion by iron oxide 51 5. Kunberger apparatus for testing oxides for gas purification 70 GAS PURIFICATION IN THE MEDIUM-SIZE GAS PLANTS OF ILLINOIS By W. A. Dunkley and C. E. Barnes PURPOSE OF THE INVESTIGATION For several years, the Illinois State Geological Survey Division, the Engineering Experiment Station of the University of Illinois, and the U. S. Bureau of Mines have been carrying on a cooperative investigation of coal and coal-mining methods in the central district, which includes the states of Illinois and Indiana, and the western end of Kentucky. One division of the main investigation is the use of coals of this district in gas manufac- ture. A number of bulletins have been published (see inside rear cover), dealing with various phases of this subject. Gas purification, the topic discussed in the present bulletin, is a phase of the subject which has an im- portant bearing on the use of central district coals in gas manufacture with the existing equipment and operating methods. One of the chief problems confronting the gas operator in using coals of the central district in place of gas coals from Pennsylvania, West Vir- ginia, and eastern Kentucky, is the increased amount of sulphur which must be removed from the gas before distribution. This increase may be small or large, according to the particular coal used, but even the central district coals of lowest sulphur content usually contain more sulphur and yield more to the unpurified gas, whether coal-gas or water-gas, than do the best low-sulphur gas coals from the regions mentioned, or the cokes made from those coals. Recognizing this condition, it was decided to make a study of existing purifying conditions in the gas plants in Illinois, to ascertain to what extent the use of low-sulphur central district coals would overload the purifying equipment now installed, and where changes in equipment, operating methods, or purifying materials might be necessary to enable the various plants to purify the gas from central district coals economically, should other conditions make a more extended use of these coals desirable. Inspection of Gas Plants The problem of studying Illinois gas-purifying conditions was assigned to W. A. Dunkley, gas engineer of the State Geological Survey Division, and an inspection trip was made by him to sixteen gas plants. The plants visited comprised nearly all of the medium-size plants of the State, (7) 8 Gas Purification in Medium Size Gas Plants including the suburban plants of Chicago. The urban plants of Chicago were not studied at this time, since it was felt that on account of their large size and special conditions, they might have problems which were not typical of individual plants in smaller cities. Nature of Data Collected In visiting the various plants an effort was made to secure as detailed information as possible in the time available, regarding load factor, size and arrangement of equipment for cleaning and purifying the gas, gas storage capacity, fuels used, and purifying methods employed. A few simple tests were made in each plant to determine the sulphur content of the unpurified gas and the amount of tar entrained in the gas entering the purifiers. Samples of spent purifying material and unused material were collected wherever possible, in order that information to be gained from chemical analysis of these materials might supplement the information available from inspection of the plants and conversation with the operat- ors. At all stages of the inspection, hearty cooperation was given by the gas men interviewed, and much interest was manifested by them. With the opening of the college year, 1919-1920, at the University of Illinois, C. E. Barnes, research graduate assistant in gas engineering at the University, was assigned to assist Mr. Dunkley in this study. Mr. Barnes devoted most of his time to the analytical work involved in studying the purifying materials collected during the inspection of the various plants. A summary of the results of these studies and a statement of the con- clusions that seem warranted follow: SUMMARY To summarize briefly, the following purifying conditions were found to exist in the plants inspected : 1. Low-sulphur eastern gas coals were being used in practically all of the coal-gas plants inspected. These low-sulphur coals, together with the considerable percentages of water-gas made in most of the plants, gave an average H 2 S content in the gas entering the purifiers of only 250 grains per 100 cubic feet. 2. Six of the eight straight water-gas plants inspected were using low-sulphur Illinois or Indiana coals for generator fuel. The average H 2 S content in the unpurified gas in the water-gas plants was 140 grains per 100 cubic feet. 3. In spite of the generally low sulphur-content of the gas to be purified in all the plants, 75 per cent of the water-gas plants and 50 per Summary 9 cent of the mixed-gas plants had maximum hourly gas productions in excess of purifier capacities. The computed overloads varied from 11.5 to 177 per cent. 4. Only two water-gas plants and one mixed-gas plant had average hourly productions in excess of purifying capacity. 5. Overload in most cases was due to lack of uniformity of load on the purifiers. This in turn was due to the sharp peak load and insufficient holder capacity. In some cases the load could probably have been materi- ally reduced by a little more attention to the rate of pumping gas through the purifiers. 6. Tar in appreciable amounts was found in the gas entering the purifiers in nearly all the plants inspected. The spent oxide from all the plants contained some tar. The average tar content of spent oxides from water-gas plants was 6.9 per cent and from mixed-gas plants 3.6 per cent. Tar seemed to be chiefly responsible for low sulphur absorption in some cases. 7. The spent oxides from mixed-gas plants had an average sulphur content of 37.4 per cent, and those from straight water-gas plants had an average sulphur content of 21.7 per cent. Overload and tar seemed to be mainly responsible for these conditions in some cases. In other cases, operating methods seemed to be the cause of these low absorptions. 8. Revivification in place was practiced by most water-gas plants but by few mixed-gas plants at the time of inspection. Only one plant revivified oxide in the off-box. Little trouble was reported in that plant as a result of the practice, and the operating costs were low. 9. Though several purifying installations are arranged for reversible gas flow, there seems \o be little effort to realize the fullest advantage from this arrangement. The same is true of rotation of boxes. 10. Few operators keep purifying records from which actual per- formance of a particular batch of oxide or method of operation can be definitely determined. 11. Few operators make any systematic quantitative tests on their purifiers to determine performance of the individual oxide batches. In several plants where the necessary testing apparatus is available, it is seldom used. 12. It seems that the greatest opportunity for immediate improve- ment in purifying conditions rests in the establishment of a simple but regular testing routine, together with better purification records. The analysis of fouled oxides for sulphur and tar, even if done by an outside laboratory, would, it is believed, be worth while. 10 Gas Purification in Medium Size Gas Plants 13. Total purification costs for the year 1919 varied in the plants inspected from 0.5 cents to 2.25 cents per 1,000 cubic feet of gas purified. Careful operation and good facilities for the handling of oxide were apparently responsible for low costs in several cases where the equipment was overloaded. Different conditions prevailing in different plants pre- clude the possibility of drawing comparisons as to the effects of load, etc., on costs. 14. Few different oxides are used in the plants of the State. It is believed that more experimentation on the part of gas companies, to find materials best suited to particular conditions, would be advantageous. 15. The use of low-sulphur Illinois and Indiana coals as water-gas generator fuels is general in the water-gas plants of the State. Though the sulphur content of the resulting unpurified gas is in some cases double that of gas from low-sulphur eastern cokes, the computed capacities of the purifiers is but slightly less in the former case. 16. Low-sulphur Illinois coals in coal-gas manufacture might de- crease computed purifier capacities by 25 per cent in some cases, as com- pared with the gas coals in use at the time of inspection. This decrease might be offset in a measure by more attention to selection of oxides, by making the load on the purifiers as uniform as possible, and by more attention to tar removal and purifying operation. 17. In several cases it appears that coals of higher sulphur content could be handled if existing equipment were rearranged and made more flexible in operation. In a few cases additional purifying apparatus is badly needed. The Purifying Process 11 THE PURIFYING PROCESS The purification of gas, by which in the narrower sense is meant the removal of the sulphur present in the form of hydrogen sulphide (H 2 S), is accomplished by the same process in nearly all the gas plants of the United States. It was discovered about 35 years ago that hydrated oxide of iron was a much more economical absorbent for this sulphur compound than slaked lime which had been in use since the beginning of the gas industry. Oxide of iron does not remove sulphur compounds other than H 2 S present in the gas, but since the gas from most American coals does not usually contain any excessive amount of these other sulphur compounds, it followed that purification with hydrated oxide of iron was adopted almost universally within a few years after its initial use for this purpose. Hydrated oxide of iron, when of good quality, not only absorbs a large amount of hydrogen sulphides, but also, after sulphiding, if exposed to the action of the oxygen of the air, undergoes a process of regeneration whereby iron oxide is again formed by oxidation of a considerable portion of the iron sulphides present, free sulphur being liberated. This process is usually . called revivification by gas operators. After revivification the material is again in condition to be used for purifying gas. Alternate sulphiding and revivification may be carried on until, with favorable con- ditions, the material may contain 50 to 60 per cent of sulphide. It is then usually incapable of absorbing more hydrogen sulphide on account of the clogging action of the free sulphur and the formation of more or less inert iron compounds and is replaced by new material. The chemical reactions involved in the sulphiding and regenerating, or revivifying, processes are not known with absolute certainty. Various chemical equations have been written expressing the probable final reac- tions, but it is likely that many secondary reactions really take place which are decidedly more complex than those given in the text books. Since it is the intention to confine this bulletin to the practical working phases of the purifying process, no attempt will be made here to repeat these equations or to go deeply into the theory of the reactions. The reader is referred to the literature of gas manufacture and chemistry for existing information on this subject. FACTORS AFFECTING PURIFICATION Efficiency and economy in gas purification depend upon three main factors — equipment, operating methods, and purifying material. The effect of each of these factors is more or less determined by existing plant conditions. Perhaps the most important condition affecting plant condi- tions is the load factor. Load factor as applied to purification will here 12 Gas Purification in Medium Size Gas Plants be designated as the relation between the volume of gas passing through the purifying equipment during the hour of maximum production and the rated hourly capacity of that equipment, as determined from its dimensions and from practical and theoretical considerations pertaining to operation with the usual purifying materials. Formulas for Gas Purifiers Since the advent of the gas industry, nearly 20 formulas have been propounded for the dimensioning of gas purifiers. Many of these were based upon the use of slaked lime, the predecessor of hydrated oxide of iron for purification. Many are indefinite in their terms and include an insuf- ficient number of factors to make them really applicable to present condi- tions. The formula of the Steere Engineering Company 1 of Detroit, which appeared about a year ago, is perhaps the most complete of all, and while there are still factors which will probably have to be introduced or changed in it when our knowledge of these factors becomes more complete, the formula is very useful and has been used in the computations of this bulletin. For the convenience of readers, the Steere formula and some information regarding its use are given in Appendix A. Design of Equipment In designing purifying equipment for a given plant, a number of things must be considered. One of the most important is the output of the plant, both present and prospective. In spite of electric competition, the output of most gas plants is growing rapidly; in fact, many companies are experiencing difficulty in keeping up with the demand for gas. It is important, therefore, to make the purifying equipment of ample size, but at the same time the investment is heavy and the interest on a greatly oversized equipment may offset to a considerable extent the operating advantages which might be derived from extra large capacity. Load Factor Another factor to be considered is the distribution of load during the day. The rate of output of most plants is far from uniform. Indeed, it is not unusual for some plants to put out 10 per cent of their daily produc- tion during the hour of maximum load. And since storage capacity has not usually grown apace with output, it is frequently necessary to generate and purify the gas practically as fast as it is sent out. From the figures iGas Age, Vol. 43, p. 227, 1919. Factors Affecting Purification maximum hourly purifier capacity— m cu. ft. 13 J 8 o CO o 00 rt 3 © JC 7> s Ph w Ph IX c/5 rt X C *i o Ph >% i 1 a < rt o In p (U w J o ^H 3 o pd c o Ph o CO o n '/ n bU >-, X T3 CX w ^ V ? 2 3 c Ph 9 CO |3 o c X c bO £ <*H Q T3 u c 3 CO bo 1 c o 8 hfl t3 C Hfi CO w 1 _C w * >. U u to rt Oh ed 14 Gas Purification in Medium Size Gas Plants obtained from the inspection of 16 gas plants of Illinois, of which 8 made straight water-gas and 8 mixed coal- and water-gas, it was found that the maximum volume of gas purified per hour in water-gas plants varied from 4.5 per cent to 11.2 per cent of the maximum daily output, with an average of 7 per cent; and in mixed-gas plants from 3.6 per cent to 9.3 per cent, with an average of about 6 per cent. Since the complete absorption of hydrogen sulphide from the gas by iron oxide takes a measurable time, the purifiers must be so designed that even at maximum rate of flow there will be ample time of contact between gas and oxide, even when the oxide is partially sulphided and inactive. As the laws of nearly every state require that the gas leaving the gas plant must at all times be free from any appre- ciable amount of hydrogen sulphide, the gas manufacturer must comply with this requirement by whatever means he may. Oftentimes when the purifying equipment is heavily overloaded or fuels run considerably higher in sulphur content that usual, compliance with the law is difficult and costly. As an economic matter, the purifiers must be so designed that they will hold enough oxide to completely purify the gas at the maximum rate of flow for a sufficient time so that it will not be necessary to handle the oxide too frequently. It is desirable that the oxide be allowed to remain in the purifiers long enough between revivifications so that it will absorb a reasonable amount of sulphur. The labor cost of handling oxide is one of the heaviest items of purification cost, and it is therefore desirable that an oxide take up a maximum amount of sulphur with the least cost of handling. This can be done by making the purifiers large enough, due consideration being given to investment charges. Sulphur Content of Gas The sulphur content of the gas to be purified is another factor to be considered. The gas industry has always been a particular customer in the purchase of coal, and the sulphur content has always been an important specification where there was otherwise little choice between coals. Good gas coals heretofore have contained not exceeding one per cent of total sulphur, and frequently the content of sulphur has been only .5 or .6 of one per cent. The decrease in supply of such superfine coals has caused gas operators to look about for possible new supplies, but while the industry will probably have to be reasonably particular so long as present purifying methods are in use, it will probably be necessary to use coals of higher sulphur content than would heretofore have been considered expedient. In water-gas not only the sulphur content of the generator fuel, but also the Factors Affecting Purification 15 sulphur in the enriching oil must be considered. Purifying equipment consequently will have to be designed with these matters in mind. The maximum permissible rate of gas flow through a system of purifiers does not vary inversely as the hydrogen sulphide content of the gas to be purified. Figure 1, which is plotted from the Steere formula, shows the relation existing between maximum hourly rate of gas flow through the purifiers and the hydrogen sulphide content of the gas. The curve represents the maximum hourly purifying capacity with various con- tents of hydrogen sulphide of a plant which would have a capacity of 100,000 cubic feet of gas per hour, if the gas to be purified contained 200 grains of hydrogen sulphide per 100 cubic feet. 1 It will be noted that if the sulphur content is multiplied by 5, giving 1,000 grains per 100 cubic feet, the capacity is reduced from 100,000 cubic feet to 66,500 cubic feet. Capacity in this sense pertains, of course, to the hourly rate of gas flow permissible, not to the capacity of the oxide for absorbing H 2 S. If the oxide in the purifiers could be completely fouled in either case, it is evident that five times as much gas containing 200 grains of H 2 S could be purified. It is evident, therefore, that the permissible rate of gas flow through the purifiers is not directly proportional to the absorption capacity of the oxide nor inversely proportional to the sulphur content of the gas. The rate of the chemical reaction has an important bearing, but this is not taken directly into consideration in any existing formula for the design of puri- fiers, though it is indirectly allowed for in the Steere formula. Capacity of Auxiliary Equipment While the capacity of the oxide purifiers must be designed to handle the maximum hourly load, a properly designed system may fail to accom- plish the complete purification of the gas because of conditions existing in other units of the gas-cleaning apparatus. Under favorable operating con- ditions, the purifiers are not usually called upon to handle all of the hydrogen sulphide that is originally present in the gas when generated. Water and tar vapor condensing from the gas in the condensers, and wash water in the scrubbers all remove a certain amount of ITS from the gas. In coal-gas plants the ammonia present in the gas has a very important part in sulphur removal, since it combines directly with H 2 S. There is not enough ammonia present to remove all the sulphur, but this incidental purification may remove as much as 20 per cent to 40 per cent of the ITS present in the gas. If the condensing and scrubbing apparatus is under- Ut may be obiected bv some that when the iiti9 no H 2 S at all the capacity of the purifici s would be infinite and that there should be consenuently a shrrp rise in the curve as sew n as there is any appreciable amount of H2? present in the (?p«, the time Fa tor of the reaction between H 2 S and hydrated oxide of iron comes into play, necessitating a very a preciable time of contact to purify the pas. It is often observed in practice thai it is more diffi cull to remove the lasl 10 • t of H 2 F from the gas than the first 90 per cent, emphasizi- g the fact that the permissible rate of purification is not inversely proportional to the H2? content of the g?s. 16 Gas Purification in Medium Size Gas Plants gas purified per hour— m. cu. ft. \ • \ \ > / O CO CC UJ- u. cn- D Q. Ul o 1 / < - \ x n VZL --"■ \ UJ H / V y / i / 1 zl / a <- \ \ o 1 \ L ^ "*v. V, *\ 1 ^ 1 «-' ^ (N ! \ N L I ■-" ** ■ ,/ 4 !" / f' \ "^ -* J u -— ■ --- ,,<- T \ \ \ \ i -4 V / i 1 \ N. < k ^~ **«. / 1 / J) ( < ( 3 1 »- ? -J c § i x - K — u 1 1 1 1 1 1 1 1 GAS SENT OUT TO CITY GAS MADE THRU PURIFIERS \ v / * \ / L / '\^ V 1 oc UJ X ir >3 7 / ^> U s, rj *i \ (N | A j 1 1 "*" *" A \ - ' "s c. 1 1 ,\ k 1 \ N t < ^ j - " ^ "N> i § I 5 o s c c > 1 g o Q " ffi .5 b^ p4 o £ a O o ffi -a ,0 Purifier Operation 17 sized, not only is this incidental purification diminished, but the oxide purifiers are forced to do part of the work, namely, tar extraction, which should be confined to the first-mentioned equipment. Tar and oil vapors undoubtedly have a detrimental effect on the operation of the purifiers, since they coat the purifying material and render it partly inactive. De- spite the recognized harmful effect of these vapors, there are few plants in which the gas entering the purifiers is entirely free from them. In some plants well equipped as to purifier capacity this condition is responsible for poor purifying results. Tar removal before purification is very important both as a matter of equipment design and of operation, and deserves more attention than it usually gets. Some kind of test by which the tar content of the gas entering the purifiers can be determined, should be made regu- larly where purifying results indicate the possibility of this trouble. Fre- quently a plant which seems to need additional purifying capacity is really in far greater need of more efficient tar-extracting apparatus. PURIFIER OPERATION Uniformity of Load Granted adequate purifying equipment, the purifying efficiencies realized will depend much upon how the equipment is operated. First, the handling of the load may well be considered. The hourly rate of gas output from a plant is, of course, out of the control of the gas manufac- turer. He must supply the demand as needed. For the sake of safety to tide over any accident to the gas-producing equipment, it is usually con- sidered necessary to keep the gas storage holders as nearly full as possible at all times. If the storage capacity is much undersize, in order to make good the output, it may be necessary to purify the gas almost at the rate of output during certain hours of the day. This may necessitate purifying the gas at a rate much in excess of the rated purifying capacity during such times. Even so, it may be possible by attention to smooth out the produc- tion curve somewhat. Figure 2 shows the output and production rates in a typical water-gas plant operating 24 hours per day. The rated capacity of the purifiers and the average hourly make are shown by horizontal lines. It will be noted that the hours of large production do not always coincide with the hours of large output. The production curve crosses and re- crosses the line of average production not only when the production is near the average for a considerable time, but also when it is averaging consider- ably above or below the general average for some time. By careful atten- tion to the rate of pumping gas these wide fluctuations could probably be prevented to a considerable extent with benefit to purifying operation. 18 Gas Purification in Medium Size Gas Plants Temperature Control Temperature control is another important consideration in purifier operation. Both excessively high and excessively low temperatures at the purifiers should be avoided. The statement is often made that at low temperatures (below about 60 °F.) the sulphiding reaction becomes slug- gish. That there is some difference of opinion relative to the effect of temperature on the purifying process, is indicated by the fact that this matter is now being studied by the Purification Committee of the American Gas Association. A temperature of about 100°F. is usually thought to give the best results, and often it is considered important that the tempera- ture be kept up in winter by artificial means if necessary. Formerly it was the practice to have the purifiers installed indoors, but the high cost of building construction as well as successful experiments with outdoor puri- fiers have resulted in a rapidly increasing number of outdoor installations. Practically all of the new installations are of the latter type. Where the temperature of the gas has a tendency to fall in winter much below the temperatures above stated, it can usually be kept up by the installation of steam coils in the purifiers or by injection of steam into the gas ahead of the purifiers. The latter practice may be questioned by some operators on account of the amount of moisture which is deposited in the purifying material, while the former may be open to objection on the ground that it dries out the oxide too much. Some outdoor installations are insulated to diminish the drop in temperature during cold weather, but this is not com- mon practice, at least in Illinois. While a fairly high temperature is usually considered advantageous in its effect on the sulphiding reaction, the temperature of the gas throughout the condensing and purifying system should not be maintained too high, lest difficulty be experienced in extracting the tar. A temperature much above 100° at the inlet to the purifiers, if maintained by the original heat in the gas as generated, would usually necessitate a temperature consider- ably in excess of this back at the tar-extracting equipment. It is usually considered that 100° to U0°F. is about the highest temperature at which tar can be extracted efficiently by most forms of tar extractors, though opinion may vary on this point. During the inspection trip to various Illinois plants by the writer, several cases were observed where the temperature at the inlet of the purifiers was around 120° to 130°, but in practically every such instance there was an excessive amount of tar being carried into the purifiers. Probably the best results will be obtained by maintaining a temperature of 90° to 100°F. at the boxes and keeping the gas saturated with water by the admission of steam to the gas, or other- Purifier Operation 19 Revivification outdoor revivification As previously mentioned, the greatest advantage of hydrated oxide of iron as an absorbent for hydrogen sulphide lies in its ability to revivify or regenerate when after sulphiding it is exposed to the oxygen of the air. Naturally the first method of revivification adopted was to remove the sulphided material from the purifying box and expose it to the air. The appearance of the material as it changes from the black iron sulphide to red or brown iron oxide, is of course a guide to the operator by which he can determine with more or less certainty when the material is reoxidized and ready for use in the purifiers again. The usual procedure in revivify- ing out of doors is to put the material in piles or windrows, perhaps two or three feet high. When the material begins to heat as a result of the oxidation process, it is raked down into a layer a foot or so in thickness, and as the surface reddens, the whole mass is turned over by shovel, this operation being repeated until the mass is of uniform color and no longer heats. This process, simple as it seems, requires considerable attention. The material frequently oxidizes very rapidly upon removal from the purifiers, and when it contains a considerable percentage of iron sulphide it is likely to ignite, especially if in a deep mass from which heat cannot escape readily. Overheating is detrimental to the further value of the material, rendering it inactive. At the same time, a moderate degree of heat promotes the oxidizing reaction without injuring the material. The operator therefore usually cools hot spots in the material by shoveling them out and exposing the hot material to the open air, which rapidly cools it, rather than by application of water, which cools the material to such an extent as to unduly retard the revivification. This method of revivification involves much handling of the material. Indeed, since it is not possible to leave material in the boxes until it is completely sulphided, it is usually necessary to handle it a dozen times or more to get a very good sulphide content. But since the sulphid- ing and revivifying reactions slow down greatly after a time, the cost of handling the material may soon offset the value of the work it is doing. Consequently, material is frequently discarded long before it contains 50 to 60 per cent sulphur, which is considered good operation. REVIVIFICATION IN PLACE Naturally gas operators looked for a method of revivifying which involved less handling of oxide, and revivification in place was the out- come. Two methods are in use at the present time, namely, ( 1 ) introduc- 20 Gas Purification in Medium Size Gas Plants tion of a small percentage of air (1 to 2 per cent) into the gas ahead of the purifiers; (2) blowing or drawing air through a box of material which has been shut off from the remainder of the purifying system. Each of these methods of revivification has its advantages and disadvantages, and there are some details of operation in both cases on which operators dis- agree. SMALL PERCENTAGE OF AIR WITH THE GAS One advantage of the first method is that it involves no danger and requires little attention. The air pump is connected to the exhauster and pumps more or less air as the exhauster runs faster or slower. One obvious disadvantage of the first method is the amount of inert nitrogen which is introduced into the gas, if an attempt is made to introduce enough air to secure complete revivification in place. Although less than 0.5 per cent of air would theoretically be required to accomplish the revivification of an oxide which was being fouled with gas containing, say, 100 grains of H 2 S per 100 cubic feet, as a matter of fact, even 2 per cent does not completely accomplish revivification. Excessive nitrogen of course dilutes the gas and requires additional enrichment, especially if the gas is made to conform to a candle-power standard. With the heating-value standard which is for- tunately rapidly replacing the candle-power standard, this effect is not so serious. The differences in operation found in different plants with this method of revivification relate chiefly to the reversal of direction of gas flow and order of the purifiers with respect to the condition of the con- tained materials. These will be discussed in the next section. AIR BLOWN THROUGH OXIDE IN AN OFF-BOX Revivification by forcing air through a box of oxide after shutting off the gas has been employed in various plants for a number of years and is heartily approved by many operators, and as heartily condemned by others. Its advantage lies in the fact that no nitrogen is admitted to the gas. On the other hand, careful attention has to be given to it while in progress, and under some circumstances it may be dangerous. O. B. Evans 1 of Philadelphia presented a paper before the American Gas Associ- ation recently, in which the experiences of several companies in the use of this method are given. From these experiences he concludes that revivifi- cation by this method is a simple operation when purifying capacity is ample and revivifications are frequent, but with overloaded purifiers extreme care must be used to prevent firing of the oxide. He believes that the best method is to recirculate air (which soon becomes chiefly nitrogen, 1 0. B. Evans, Revivification in place, presented at a meeting of the Amer. Gas Assn., October, 1919. Purifier Operation 21 since the oxygen is soon removed by the purifying material) continuously through the box and through a cooler of the contact type, in which it comes in contact with water. The water keeps the recirculated mixture saturated, and the water vapor helps to keep down the temperature in the oxide. Arrangement of valves is made whereby a small amount of fresh air can be admitted to the circulation and a similar amount of inert gas expelled from it as desired. The blower should be of sufficient capacity to reduce channeling effect and to circulate the mixture faster than the rate at which gas is passed through the box during operation. He states that shallow boxes aid considerably in successful revivification in place on account of their greater radiating capacity per bushel of oxide. With these methods Mr. Evans believes that revivification can be carried out without danger, channeling and excessive local heating being largely avoided. The suc- cessful experiences of many operators over a number of years confirm th ; s. REVERSAL OF GAS FLOW AND ROTATION OF BOXES Reversal of direction of gas flow through the purifiers and the order of the various purifiers with respect to the condition of the contained oxide, are matters of considerable importance in connection with revivifica- tion in place. In both of these matters the operator may be limited by the arrangement of his equipment. Not all plants are so arranged that the direction of gas flow in a given box can be reversed; and while a certain amount of latitude is usually allowed as to the order of purifying boxes, one finds a number of cases where the connections are so arranged that the possible number of combinations is small. Obviously, any group of puri- fiers must be so connected that any box can be shut off for refilling. In most of the older installations, when one box is off, the order of the other boxes is predetermined, only one arrangement being possible. Reversal and rotation, as applied to revivification in place, depend upon the principle pointed out several years ago by B. E. Chollar, that iron sulphide will not revivify to the oxide in the presence of hydrogen sulphide. Let us assume, as is frequently the case in new purifier installations, that the gas enters at the middle of the box and passes downward and upward through the two layers of oxide and comes out at the top and bottom of the box. The lower layer will sulphide downward and the upper layer upward. If when the sulphiding has extended say half-way through each layer, the direction of flow be reversed so that the gas enters at the top and bottom of the box and leaves at the middle, then the comparatively fresh oxide in the top of the upper layer and the bottom of the lower layer will remove the H 2 S from the gas, and any oxygen present will go on and revivify the foul upper part of the lower layer and the lower part of the upper layer. The 22 Gas Purification in Medium Size Gas Plants frequency of reversals will depend upon the degree of loading of the puri- fiers. Where the purifiers are being operated at normal capacity, reversal once a week is often advised. With an overload, it might be advisable to to reverse oftener. Box rotation is another means to accomplish the same end. Where there are three or more boxes in series, it would seem logical to have the clean box first to remove the H,S and the fouler boxes after, to be revivified by the oxygen which had been admitted to the gas. The following are suggested orders of rotation of a 4-box set, the changes being made when ITS appears at the outlet of the third box: 1—2 — 3 — 4 4—1—2 — 3 3 — 4—1—2 2 — 3 — 4—1 In this way the most revivified batch is placed first and the next cleanest batch is always last to take up any traces of H 2 S which may get by the previous batches at any time. Where there are only two boxes in series, especially those of the non- reversing type, it would hardly seem advisable to have the clean box first, since failure of the second box to revivify for any reason would leave no active material to intercept traces of hydrogen sulphide. As stated earlier in this section, there is no method of procedure in the matter of revivification in place, which is accepted as best practice by all operators. The only way by which any operator may arrive at a satisfac- tory conclusion is to try various methods and arrangements and satisfy himself which method is most applicable to his particular plant. Chemical Control and Records In order to secure and maintain good purifjang efficiencies, it is im- portant that the operator know at all times the status of the material in each one of. his purifiers. Knowing this, he will not only be able to judge whether his method of operation is satisfactory, but he will be able to detect differences in purifying material which might otherwise be obscured by other conditions. The extent of the system of tests and records main- tained will of course depend, among other things, upon size of the plant and the force available. Practically every plant makes the simple lead acetate paper test, but this is merely qualitative. It tells of the presence or absence of hydrogen sulphide, but gives little information relative to the amount present. The total-sulphur test carried on under the requirements of the Public Utilities Commission, at least in the State of Illinois, gives informa- Purifier Operation 23 tion only with respect to the amount of sulphur in all forms present in the finshed gas, but gives no information in regard to the performance of the individual purifiers. The introduction several years ago of the Tutwiler hydrogen sulphide burette marked a distinct advance in the matter of checking up purifier operation because it gave the gas operator a simple, easy method of study- ing his purifier performance without .the aid of a trained chemist. Un- fortunately, the apparatus has to be made of glass, and is so shaped and proportioned that it is easily broken ; and, further, unless care is taken to remove the plugs of the glass stop-cocks and insert pieces of paper around them after using, they are very liable to become hopelessly stuck in a short time. Nevertheless, with reasonable care the apparatus can be kept in good order, and the information obtained with it is very useful. One advantage of the instrument is the short time required to make a series of determina- tions. Probably after a very little practice any operator, even with no chemical training whatsoever, could make a test of the gas entering the first box and leaving each box of a four-box series, in fifteen or twenty minutes. By simple subtraction of the number of grains of hydrogen sul- phide found at the outlet of each box from the amount present at the inlet the amount of sulphur being removed by each box would be immediately known. The results of a test on a certain day might be as follows : H ,S Removed Grains FLS before boxes ioo " " after ist box 25 75 by 1st box " 2d " IO l£ " 2d " " 3d " 5 5 " 3d " " 4 th " o 5 " 4 th " Now 5 grains of H 2 S after the third box could probably be detected by lead acetate paper as ordinarily used and would perhaps ordinarily be taken as a sign to change the order or empty a box, but if this same distri- bution of the work continued for several days and the last box handled the remaining few grains of H.S were absorbed in the last box all right it might be well to retain this order for a time, since a larger absorption in the first box would be accumulating. A considerably greater sulphur absorption might be realized in the first box than would be obtained if it were emptied or reversed as soon as a trace, as shown by lead acetate paper, was visible at the outlet of the third box. Steere 1 states that 20 to 50 grains of H 2 S may be safely passed to the last box if tests are made regu- larly, and the boxes are properly proportioned. He would allow 20 additional grains to pass to the last regular box, where a catch box is pro- vided. On the other hand, let us assume that the following results were shown by the tests: •Bull 37, Steere Engineering Co., 1919. 24 Gas Purification in Medium Size Gas Plants Grains H 2 S before boxes ioo ist box 80 20 by 1st box " " " 2d " 20 60 " 2d " " " " 3d " 5 15 " 3d " " 4th " o 5 " 4 th " It would at once be evident that box No. 1 was not doing its share of the work, since if it were in good condition it could usually be expected to do at least 60 per cent of the total absorption. It would be evident that box No. 2 was bearing the chief burden and it would be high time to* empty No. 1 or reverse so that it would clear up. If such tests as the above were made faithfully day after day and carefully recorded in a book (not on loose scraps of paper) it is evident that a running record could be maintained from which, knowing the amount of gas metered per day, the actual number of grains of H 2 S absorbed by each box from change to change would be known. Knowing the number of bushels of oxide in each box, the volume of gas passing in a given time, and the number of grains of H 2 S absorbed by each box per 100 cubic feet of gas passing through, it would be very simple to calculate with a reasonable degree of accuracy the number of pounds of sulphur absorbed per bushel. A sample record and computation is given in Appendix B. Such results would be of far greater value in determining relative merits of various operating methods or of various oxides than would the usual record of gas purified per bushel between changes, because unless tests are made there is no way of knowing what proportion of the purification should be credited to each box. Usually the first box is credited with all of the purification on the theory that in the long run each box will be similarly credited, and will average up, but during this same time the sulphur content of the gas may change, or more or less tar may be carried forward, affecting sulphur absorption so that the actual performance may be entirely obscured by other conditions. As a final check on operation, analysis of the oxide for suphur and tar after each removal from the box and especially before discard would be very helpful. Often a batch of oxide is discarded when it is still capable of doing much useful work, and perhaps even more often a batch is returned to the box at considerable expense of handling when it might better have been discarded. Even the trained eye may sometimes be deceived in judging oxide, especially if it contains some tar. In a paper by Fulweiler and Kunberger 1 a method for determining mathematically 'Some of the Physical Characteristics of Ferric Oxide, Proc, Amer. Gas Institute, 1913, Vol. 8, Pt. 1, p. 476. Purifier Operation 25 whether a given batch of oxide is worth using again has been developed. The formula used is given in Appendix C. It will be observed that no knowledge of higher mathematics is required, though one does need to know what the batch has done in the past and what it is capable of doing as judged by a simple laboratory test. If the operator were in possession of the facts derived from the running record of tests above described, together with certain costs which he should know for intelligent operation of his plant, it would not be difficult to use the formula provided he could make or have made an analysis of his oxide and a laboratory hydrogen sulphide absorption test. The writers of this paper do not advise every gas company large and small, to maintain a chemical laboratory and a trained chemist. The small companies probably could not afford these refinements. The large companies already have them and know their value. We will not attempt to prescribe the minimum size of plant that can afford a laboratory. How- ever, many of the smaller plants in this and other states are links in a chain of plants. It would seem feasible for a chain of several plants to main- tain a laboratory and a chemist, to whom samples of oxide, ammonia, and other materials could be sent for examination. The results reported by him from testing oxides, together with the Tutwiler test made by the plant superintendent or other person, and adequate record of purifier changes and batch performance, all taken together, would suffice to put purifying operation on a much higher plane than it now is in the average plant. It is believed that it would also effect a real economy in dollars and cents within a reasonable time. Purification costs heretofore have usually been but a small item in the total cost of making and distributing gas, but with the greatly increased cost of labor, it is becoming especially desirable to get the greatest absorp- tion of sulphur with the least amount of handling. Revivification in place, with its obvious advantages where properly conducted, should become much more general. If difficulties are encountered, careful inquiries should be made as to why they are encountered. Only by intelligent study, aided by tests and good records, can the best results be obtained. QUALITY OF OXIDE FOR GAS PURIFICATION The quality of oxide used for purifying gas is another factor affecting efficiency and economy of purification, but this has not as yet been worked out so that it can be expressed mathematically in computing performance of a given equipment. Some research work, having as its object the deter- mination of the effects of the peculiar properties of various oxides is now 26 Gas Purification in Medium Size Gas Plants in progress, and it is hoped to throw some light on the subject in a later publication. At the present time it is recognized that oxides produced by different processes, and indeed even oxides made by the same process, show variations in performance, but just what are the causes of these variations and how they can be controlled is not now known. Types of Hybrated Iron Oxides Three main types of hydrated oxide of iron are in use for purification. These include: ( 1 ) Oxides made by rusting case iron borings in the air with water only, or with accelerating agents such as sulphate of iron, salt, etc. (2) Natural oxides, which include certain ores having the proper chemical and physical condition. (3) Precipitated oxides,* made by the chemical precipitation of hydrated oxides of iron from the salts of iron produced as by-products in certain industries or from the iron-bearing water of some mines. Each of these oxides has properties peculiar to itself, the reasons for which are not yet clear. While ferric oxide (Fe 2 ;J> ) is included in the composition of each and is the reacting material, the presence of other materials in combination with it and the physical structure of the material are of the greatest importance in determining performance. Ferric oxide by itself without water of hydration is almost or entirely non-reactive with hydrogen sulphide. Formerly the iron content of a commercial material was considered as an index to its value for purifying gas. While this may be true to a limited degree when applied to oxides of one type, it does not hold in comparing oxides of different types. Tests of Oxides Since the absorption of hydrogen sulphide is the main thing desired from the oxide, it follows that tests which will indicate the absorbing value of a particular oxide are the most logical ones to apply in valuing a material. Such tests have been devised but no test which has yet been suggested is sufficiently definite in its provisions and indicative of the results to be obtained in actual practice to thoroughly meet the require- ments of a standard test. This is evidenced by the fact that the Purifica- tion Committee of the American Gas Association is now endeavoring to devise such a standard test, which will meet the requirements of the gas industry generally. One of the simplest and best-known laboratory tests worked out thus far is that of A. F. Kunberger 1 (see Appendix C). In the Kunberger 'Some of the Physical Characteristics of Ferric Oxide: Proc, Amer. Gas Institute, 1013, Purifier Operation 27 method a small weighed sample of oxide is fouled by dry hydrogen sulphide for one hour, the water liberated by the reaction being retained in a tube containing granulated fused calcium chloride, which is weighed with the tube containing the oxide before and after fouling. The gain in weight of the two tubes (or one tube containing both oxide and calcium chloride may be used) is equal to the weight of H 2 S absorbed. Such a test is very useful in determining the relative capacities of various materials under the conditions of the test. The test is also said by some operators to check practical operations quite closely. Other operators, however, place less reliance in it, and in purchasing an oxide of unknown quality are not satisfied with anything less than a semi-com- mercial test. Such a test usually consists in fouling two oxides, one of known practical performance, the other the unknown material, with unpurified gas, in a pair of small purifiers. These purifiers may contain from a few quarts to a few bushels of the materials in one or more layers. Sometimes two sets of two or more purifiers each are used. The purifiers are followed by gas meters to measure the gas passing through each oxide. The test usually consists in passing gas through both materials at the same rate and noting the volume of gas purified by each until the time when some hydrogen sulphide passes one material as shown by a test with lead acetate paper at the outlet. The rate of gas flow at the beginning of the test is generally at least twice that usual in practice. Sometimes when one material begins to pass hydrogen sulphide the rate of flow is reduced until absorption is complete and the test continued at the new rate until H 2 S again passes the material. The test may be continued until both materials are entirely fouled and in some cases the materials after revivification are tested again. Such a test would seem to possess advantages over a strictly laboratory test, since the same kind of gas is used that has to be purified in the works purifiers and the rate of fouling is nearer to that obtaining in practice. On the other hand, in small test installations where the surface of contact between the oxide and the box is usually relatively much greater than in practice, there is danger of the gas passing up the sides of the box to a considerable extent and the excessive rates of purification are likely to cause channelling. Another disadvantage of tests of this kind is the time required to complete them. With rates of fouling approaching those in practice, weeks or even months may be required to foul the materials. In all tests of oxides, the great difficult)- is to interpret the results fairly and with certainty. At present those gas companies who make extensive tests on oxides interpret the tests as best they can in accordance 28 Gas Purification in Medium Size Gas Plants with their own particular conditions and the opinions of their own engineers, and until a standard test with carefully specified equipment and testing procedure is worked out, it will be very difficult properly to evaluate commercial purifying materials. Activity and Capacity of Oxides The existing overloaded condition of the purifying apparatus in many gas plants necessitates more attention to the activity or speed of oxides than has hitherto been given to the subject. A satisfactory test must indicate the relative activity of the materials under test. At the same time capacity of oxides will remain an important consideration. The Steere formula allows about 6 minutes time of contact of gas with oxide, assuming all the space occupied by the oxide to be free space. The actual free space in a layer of oxide will of course be considerably less than the total volume, depending upon the coarseness of the material, a factor which is continu- ally changing as the material is used and becomes more and more clogged with sulphur. A new oxide sponge might have 60 to 70 per cent free space, and therefore the actual item of contact would be considerably less than 6 minutes as mentioned above. If, as is sometimes the case, the time of contact in a heavily loaded purifying system is reduced to 2 minutes or less, it is evident that the rapidity of the material may have a very im- portant bearing on its value for such a plant. It is quite likely that various factors could be worked out applying to various types of purifying oxides, which introduced into formulas for purifying capacities would materially alter the hourly capacities permissible with an equipment of a given size and arrangement. This subject needs further study. For the present the differences in oxides will not be considered in studying the purifying con- ditions in Illinois plants. Computations will be made, assuming that all oxides are the same. In drawing conclusions and suggesting remedies for certain cases, the possibility of using more rapid oxides must be borne in mind. CONDITIONS FOUND IN ILLINOIS PLANTS Having considered the conditions affecting purifying efficiencies, let us see what the actual conditions are in Illinois plants, so far as can be determined from the information collected. As was mentioned earlier in this bulletin, sixteen of the medium-size plants of the State were visited. Information was gathered relative to load conditions, size and kind of equipment, operating methods, and to a Purifier Operation 29 certain extent, the results obtained. This latter information was supple- mented by the results of analyses of spent oxide samples collected in the various plants. Upon returning to headquarters the information obtained was tabu- lated, purifier capacities were computed, oxide samples were analyzed, and the results tabulated. Half of the plants visited were mixed-gas plants. Their problems were somewhat different from those of straight water-gas plants, so the data and results have been tabulated separately. In accord- ance with an oral agreement under which results were obtained, the tabu- lations contain no plant names. The writers will be glad to inform any operator as to the designation of his own plant, in order that explanations may be made in case of misunderstanding or difference of opinion as to results obtained. Equipment Conditions The purifying equipment conditions in Illinois plants are probably the same as those existing in similar plants elsewhere. There is the usual combination of old and new equipment which results from piecemeal con- struction. The old equipment has often been outgrown, but in many cases is still serviceable, and the usual policy has been to increase capacity by adding another purifier, usually of the outdoor type, to the older indoor equipment, retaining the latter in service. Such a condition exists in about half of the plants inspected. While it is of course advantageous to get as much economical service as possible from a given unit of equipment, it appears in some cases that the old equipment is a drag on the new. In making additions in some cases, the minimum cost of the addition has been looked after, rather than operating economy. The connections are so arranged as to permit both the old and new equipment to be used at the same time, but with little regard to the flexibility of the system. The arrangement is often so fixed as to permit only one order of the boxes in series. In some cases there is a certain latitude for arrangement among the old units, but usually the new equipment either precedes or follows the old in a fixed position relative to it. Most of the new purifiers are equipped for deep layers of oxide and valved for reversible flow, whereas most of the old boxes have shallow oxide layers and straight flow. In no case which we recall is there a com- bination of old and new equipment in use, provided with arrangements for perfect flexibility as to rotation of boxes and reversibility of flow. In about half the plants are the boxes uniform as to size and type and com- pletely flexible as to arrangement. 30 Gas Purification in Medium Size Gas Plants The computation of purifier capacity for many of the plants is a rather complicated matter. The Steere formula is applicable primarily to boxes arranged for reversible flow, though the formula can be applied to straight-flow boxes by the use of an appropriate factor. In some plants where there is a combination of reversible and straight-flow boxes, it is necessary to use one's judgment in the selection of the proper factor, and opinions of two engineers might differ as to the proper value of the factor to be taken. Purifying Equipment of Individual Plants The following is a brief description of the purifying equipment of the various plants inspected, together with remarks relative to special condi- tions which seem noteworthy and pertinent. Plant No. 1. The purifying equipment of this straight water-gas plant consists of three cylindrical steel outdoor boxes, each 35 feet in diameter and 13 feet high. Each box contains two layers of oxide 4 feet deep. The arrangement is entirely flexible, permitting the boxes to be used in any desired order. The direction of gas flow in each box is reversible, the gas entering between the layers and leaving at the top and bottom or vice versa. Southern Illinois coal is used as generator fuel in this plant and the gas at the purifier inlet contained 180 grains of H 2 S per 100 cubic feet at time of inspection. The computed hourly capacity of the purifiers is 240,000 cubic feet, or nearly 2.2 times the maximum hourly make reported. The installation is therefore oversize and with good oxide would permit the use of coal of considerably higher sulphide content, if other conditions made it desirable. The cost of purification per thousand cubic feet in this plant should be very low, unless the capital charges on an equipment so considerably oversize are excessive. No figures relative to capital charges are available. Analyses of spent oxide from thib plant indicate that best results are not being realized from this equipment, the sulphur absorption per bushel not being nearly so high as in several other water-gas plants that are much more heavily loaded. Plant No. 2. The purifying equipment in this plant consists of two cylindrical dry-seal straight-flow indoor boxes, 18 feet in diameter and 9 feet deep. Each box contains two 4-foot 3-inch layers of oxide. The generator fuel used at time of inspection was eastern coke and the H.,3 content of the unpurifled gas was 100 grains per 100 cubic feet. The computed hourly capacity of the boxes was 39,800 cubic feet of gas, which could theoretically be increased to about 52,000 cubic feet per hour by arranging for reversible flow, while the maximum hourly make was re- ported as 110,000 cubic feet. The sulphur absorption per bushel in this Purifier Operation 31 plant was low, as would be expected in a plant so overloaded. The situa- tion was complicated by insufficient condensing apparatus and by a relief holder having but one connection, giving but slight opportunity for any cooling of gas in the holder. At the time of inspection the gas was enter- ng the boxes at a temperature of about 120°F. and was carrying much tar tog. Analyses of spent oxide, however, indicate that this is not a year- around condition. The purifying condition in this plant were the most unfavorable observed in the State, and it is greatly to the credit of the operators that they succeeded so w T ell in supplying clean gas to the public. The shortcomings of the present cleaning and purifying s) r stems are realized 'by the management, and extensive improvements are projected. Plant No. 3. The equipment of this water-gas plant consists of three dry-seal, oblong, indoor boxes 16 feet by 12 feet horizontal section, and 12 feet deep. Each box contains two 5-foot layers of oxide. These boxes were originally designed for reversible flow. The superintendent conceived the idea that the capacity would be increased by making straight- flow boxes of them. The computed capacity was thereby reduced from 55,200 to 40,800 cubic feet per hour with gas containing 190 grains per 100 cubic feet. Whether decrease in efficiency has resulted from the change would be extremely difficult to determine, since recent improve- ments in the tar-extracting apparatus have probably increased efficiencies to a far greater degree than the change referred to decreased them. Since the purifiers are now overloaded nearly 100 per cent, a return to the former arrangement which would involve very little expense, is suggested. Plant No-. 4. This is an up-to-date small water-gas plant. The purifying equipment consists of two cylindrical reversible How outdoor boxes 14 feet in diameter and 10 feet high, each containing two layers of oxide 4j/2 feet deep. The computed maximum hourly capacity is 25,000 cubic feet of gas, containing 175 grains H..S per 100 cubic feet, and the reported maximum hourly make is 22,000. Since no air is admitted for revivification the reversibility of these boxes is a matter of less consequence than would otherwise be the case and the actual capacity is probably less than the computed. The plant is seldom operated more than 12 hours per day. In addition to these working boxes there are four square indoor water-seal, straight-flow boxes, 10 feet in diameter and 4 feet deep, designed to hold one 3-foot 6-inch layer of oxide each. These boxes are relics from a former coal-gas plant. They are connected by a center seal, so that only three of them can be in service at one time. The feasibility of putting these boxes into service so as to permit the use of generator fuel (bituminous) of higher sulphur content has been considered. Inasmuch as the coal suggested for use gives about 400 grains of I I 2 S in the gas, and 32 Gas Purification in Medium Size Gas Plants the two boxes now in use would have a computed maximum hourly capacity of 22,600 cubic feet of gas of that sulphur content, it hardly seems that the slightly increased hourly capacity to be obtained from the small old-type boxes would warrant the expense of putting them into service and the inconveniences attending their use. Plant No. 5. The purifying equipment of this plant was originally designed for coal gas. Extensive necessary repairs to the coal-gas equip- ment, difficulty of securing efficient retort-house labor, and the economies realized from making water gas from bituminous coal have resulted in at least temporary abandonment of the coal-gas equipment except in so far as it could be utilized in handling the water gas. The purifiers include two oblong dry-seal indoor boxes 15 by 10 by 8 feet deep, each containing two layers of oxide Zy 2 feet deep, and one square dry-seal box 20 by 20 by 5 feet deep, containing one layer of oxide 4^ feet deep. The boxes are all of the straight-flow type. The large shallow box can be used only as a catch box, being necessarily last in the series. The computed capacity of the purifiers is 32,700 cubic feet of gas per hour, and the overload during hour of maximum make is about 22 per cent. The present arrangement of the boxes fulfills present needs quite well. By arranging the two deep boxes for divided and reversible flow, the capacity could be increased to about 36,500 cubic feet per hour. The purifying costs reported are quite low. Plant No. 6. The purifying equipment of this water-gas plant con- sists of two rectangular indoor boxes 16 by 12 by IIV2 ^ eet > eacn contain- ing two 5-foot layers of oxide, in parallel with two boxes 20 by 20 by 5 feet containing one 4-foot layer of oxide each. The deeper boxes are arranged for reversible flow, but the shallow boxes have straight flow only. The computed capacity of the system as now arranged is about 60,000 cubic feet per hour. The overload at the time of maximum make is about 67 per cent. It is understood that a rearrangement of the boxes whereby all the boxes would be in series, has been considered. If such a change were practicable from other considerations, the capacity would be increased slightly but probably not sufficiently to pay. On the other hand, if plenty of overhead room is available, and it were practicable to double the depth of the two shallow boxes, making two layers in each box, and to arrange the whole system for reversible flow throughout, the capacity would be increased to over 100,000 per hour or to about the present maximum hourly production of the plant. Plant No. 7. This large suburban water-gas plant is equipped with four rectangular water-sealed indoor boxes each containing two 5-foot layers of oxide. The gas flow in each box is reversible, gas entering at the Purifier Operation 33 middle and leaving at the top and the bottom of the box or vice versa. Rotation of the boxes can be accomplished in only one direction, as: 1 — 2 — 3 — 4, 2 — 3 — 4—1, etc., but not 4 — 3 — 2— 1, etc. The computed hourly capacity of the system is about 206,000 cubic feet. It is about 40 per cent overloaded at time of maximum production. The purifying costs in this plant are the lowest reported by any plant inspected. The spent oxide from this plant showed a sulphur content rather above the average. The tar content was higher than the average found in all the plants and indicated that an excessive amount of tar had been allowed to enter the boxes at some time during the life of the oxide. At the time of inspection, the gas entering the boxes was reasonably clean. The indi- cations are that purifying results could be improved somewhat if the load curve, shown in Figure 2, could be smoothed out, making the production through the boxes more uniform. Plant No. 8. This suburban water-gas plant has two sets of purify- ing boxes in parallel. One set consists of four rectangular indoor boxes connected by a center seal which permits the use of only three boxes at a time. Each box is 24 by 24 by 7 feet deep and contains two layers of oxide 3 feet deep. The other set of boxes consists of four boxes, each 16 by 16 by 4!/2 feet deep. Each box contains a single layer of oxide 2 1 /) feet deep. Three of the boxes are interchangeable, viz., any one of them can be made first box, but the position of one box fixes the order of the set. The fourth box of the set is a catch box and is always at the end of the series. All the boxes in both sets are arranged for straight upward flow. The computed maximum capacity of the two sets of boxes, as now arranged, is about 100,000 cubic feet of gas per hour. The capacities of the two sets are so different (about 4 to 1 ) that any rearrangement by placing in series would probably not be feasible. The set of larger boxes has a present capacity of about 80,000 cubic feet of gas per hour. If a new valving system could be installed, putting the fourth box into use and permitting reversal of flow, the capacity of the set would be approximately doubled, giving the entire system a total capacity of 180,000 cubic feet, more or less, per hour, which is about equal to the present maximum flow through the purifiers. Plant No. 9. This plant produces about 60 per cent water-gas and 40 per cent coal-gas. Each kind of gas has an independent condensing system and the gases are mixed at the inlet of the purifiers. The purify- ing equipment consists of six boxes. Of these, four indoor boxes are arranged in two pairs, the members of each pair being connected in parallel and each pair acting as one divided-flow but non-reversible box. The paired boxes are 16 feet square in horizontal section and contain one layer 34 Gas Purification in Medium Size Gas Plants of oxide each, three feet deep. The remaining boxes include an outdoor cylindrical steel box 25 feet in diameter and 10 feet high, containing two 4-foot layers of oxide, and an indoor catch box 16 by 16 by 4 feet deep, con- taining one 4-foot layer of oxide. The paired units and the cylindrical box can be arranged in various orders asl — 2 — 3, 3 — 1 — 2, and 2 — 3 — 1. The direction of flow except in the cylindrical box is non- reversible. The history of the development of this installation is not known. It seems likely that the original installation consisted of a four- box set arranged with a center seal as is common in old installations. In adding to the original equipment, an arrangement was made which is more flexible than is found in many plants. The computed hourly capacity is about 70,000 cubic feet per hour, which is ample for present needs. Plant No. 10. This mixed-gas plant puts out about 90 per cent coal gas on the average. The peak load on the purifiers is rather sharp, the maximum hourly production being about 9.5 per cent of the maximum daily production. The purifiers include two cylindrical outdoor boxes 15 feet in diameter and 12 feet high arranged for reversible flow. Each box contains two 5-foot layers of oxide. The computed maximum capacity is about 30,000 cubic feet per hour and since the maximum hourly produc- tion through the boxes is said to be 75,000 cubic feet, there is a large load factor. The storage capacity in this plant is about 65 per cent of the max- imum day. Water-gas and coal-gas are cleaned separately and mixed at the inlet of the boxes. The coal-gas production will probably average around 540,000 cubic feet per day, with an hourly production of about 20,000 cubic feet. The maximum daily gas output from the plant is reported to be about 800,000 cubic feet, which indicates that the maximum water-gas production may be about 32.5 per cent. The overload on the purifiers then is evidently due to the pumping of water-gas through the purifiers at an exceptionally high rate during peak load. It would seem, however, that if the maximum hourly output of the plant for a peak lasting say four hours at a time averaged no greater than the reported maximum make through the boxes, namely, 75,000 cubic feet per hour, and if the holders were full at the beginning of the peak, there would be no necessity of pumping gas through the purifiers so fast, even were it ncessary to retain the city holder two-thirds full at all times. Even were the load curve smoothed out as much as possible, with the existing storage capacity it is likely that there would be a considerable overload during maximum hours. The logical extension of the purifying system would seem to be the instal- lation of another box similar to those now in place. Another such box would bring the hourly capacity up to about 50,000 cubic feet per hour. Purifier Operation 35 Plant No. 11. This small mixed-gas plant has an average output of about 200,000 cubic feet of gas per day, of which about one-fourth is water-gas. The maximum hourly purification is about 12,000 cubic feet. The purifying equipment consists of two cylindrical outdoor purifiers, 15 feet in diameter and 12 feet high, containing two 5-foot layers of oxide each. The flow is divided and reversible. The computed capacity of these purifiers is about 30,000 cubic feet per hour ; therefore the system is much underloaded. It would be interesting to know the capital charges on an oversize system of this kind, but no figures are available. The size of the purifiers is ample to care for the growth in output for some years to come. Plant No. 12. The purifying equipment of this plant which produces only 10 to 15 per cent water-gas, consists of four water-sealed indoor boxes 16 by 16 by 7.5 feet deep, each containing two layers of oxide 2.75 feet deep. Three of the boxes are arranged for rotation, the possible arrange- ments being 1 — 2 — 3, 2 — 3 — 1, 3 — 1 — 2, but not the reverse. The fourth box acts as a catch box and is always in last position. The computed capacity of this installation is about 44,800 cubic feet. If the boxes were valved for divided reversible flow, and 2 per cent of air used for revivification in place, the computed capacity would become about 61,000 cubic feet per hour. At present the installation is about 25 per cent overloaded at time of maximum hourly production, but since the load is being handled well and the cost of purification is low, there is little reason for making a change. The use of higher sulphur coals, if desirable or necessary for other reasons, might make the suggested change advisable. Plant No. 13. The original purifying equipment of this plant, which makes about 40 per cent water-gas, consisted of two rectangular, water- sealed, indoor boxes, each 16 by 24 by 5 feet deep and containing one layer of oxide, 4 feet 4 inches thick. A cylindrical outdoor box, 15 feet in diameter and 12 feet high, containing two 5-foot layers of oxide was subse- quently installed. The new box is of the divided-flow reversible type. On account of tar trouble, the first rectangular purifier was emptied and refilled with shavings to act as a shavings scrubber. The remaining rect- angular box and the new outdoor box have a combined capacity of about 26,000 cubic feet of gas per hour. As is common in piece-meal installa- tions of this kind, there is little opportunity for rotation of boxes; indeed in this plant there is only one arrangement possible. The rectangular box is always first in series, and the only change possible is reversal of flow in the second box. The overload during maximum hour is only about 15 per cent and there seems little reason for making a change on that account. Greater flexibility with respect to rotation of boxes would be desirable and would probably permit a more nearly complete fouling of the oxide and a 36 Gas Purification in Medium Size Gas Plants reduction in purification labor. If a regular shavings scrubber were installed and the box now used for tar extraction were put into service again as a purifier, the capacity of the installation would be increased to about 35,000 cubic feet of gas per hour. Plant No. 14. This plant makes about 95 per cent coal-gas and the purifying results obtained are considerably above the average. Seven boxes are in use, of which three boxes handle coal-gas exclusively, while the remaining four handle a smaller part of the coal-gas and all the water-gas. The purifiers are arranged in two parallel groups. One group consists of two 24 by 24 by 5 foot boxes, each box containing one layer of oxide 4 feet 6 inches in depth, and one 24 by 24 by 12 feet box containing two oxide layers, each 4 feet 6 inches in depth. These boxes are straight flow, but the sequence can be changed. The other group of boxes consists of four rectangular boxes, each 20 feet in diameter and containing one 2 foot 6 inch layer of oxide. These boxes are connected by a center seal and one box is always off. The computed hourly capacity of the first group is about 72,000 cubic feet- of gas per hour, and that of the second group about 23,000 cubic feet per hour. Since the maximum hourly production is reported to be 70,000 cubic feet, the system is not overloaded, at least with gas of the present sulphur content. In case an increased capacity were necessary, it might be desirable to build up the two shallow 24-foot boxes to conform with the existing deep box of the same area. Such an enlarge- ment with valves so arranged as to give complete reversibility would increase the hourly capacity of this group from 72,000 cubic feet to 157,000 cubic feet per hour, over 150 per cent of the present total capacity. Sim- ilarly, the four shallow boxes of the other group, if built up to double depth and arranged for complete reversibility, would have a capacity of about 100,000 cubic feet per hour. Plant No. 15. This small mixed-gas plant produces about 40 per cent water-gas. The purifying equipment consists of four old-type, indoor, water-sealed boxes, connected by a center seal which permits the use of only three boxes at one time. The boxes are 10 by 8 by 4 feet and each con- tains one 3-foot 6-inch layer of oxide. The calculated capacity of the installation as now arranged is only about 6,000 cubic feet per hour, while the maximum load is reported to be 15,000 cubic feet, and the average load 7,000 cubic feet per hour. It is evident, therefore, that the installation is heavily overloaded. If the fourth box could be put into service except for the short interval required to change a box, the capacity would be increased to about 7,500 cubic feet per hour, which would still leave 100 per cent overload. A new purifying system, or the addition of one up-to-date box with proper valving, would perhaps be the best solution. The present Purifier Operation 37 single boxes, if built up to double height and arranged for complete revers- ibility of flow, would have a computed hourly capacity of about 22,600 cubic feet. Such conditions as head room available, strength of supports of the present system, condition of present boxes, and cost of the improve- ments would of course determine the best way to increase capacity. The fact that the purifying labor cost per 1,000 cubic feet in this plant is about double that of any other plant inspected (other small plants included) indicates that there is sufficient economy to be realized in labor charges to offset a considerable capital charge. Plant No. 16. This plant makes about 93 per cent coal-gas. The purifying equipment is rather unique among the installations inspected. The system includes two outdoor concrete boxes, each 40 by 26 by 14 feet 4% inches deep, each containing one 53-inch layer and one 62-inch layer, two indoor boxes each 30 by 20 by 5 feet deep, containing one 4^-foot layer, and two indoor boxes 20 by 20 by 4% feet deep containing one 3-foot 9-inch layer each. The greater part of the purification is done in the two concrete boxes. These are placed first in the series and the order cannot be changed, nor is there any rotation of the two boxes though they are valved for divided reversible flow. Coal-gas only is purified in the con- crete boxes. The old type indoor boxes follow the concrete boxes and the water-gas enters the inlet of the former. The old purifiers are all single- layer straight-flow boxes and are all arranged in series. The computed capacity of the whole system is about 190,000 cubic feet of gas per hour while the maximum hourly purification is reported to be only 40,000 cubic feet. It is evident, therefore, that the capacity is ample for some time to come. This plant was the only one inspected in which the oxide was revivified in the box by an air blast. Only the oxide in the large concrete boxes could be revivified in this manner. The way in which revivification is conducted in this plant will be discussed later. Summary of Capacities and Load Conditions Tables 1 and 2 give the summarized data relative to load conditions, storage capacity, purifier capacity, etc., of the plants visited. 38 Gas Purification in Medium Size Gas Plants Table i. — Purifier load conditions in medium-size water-gas plants of Illii 3 u 3 ^ £i J, 3 X IS 43 1 a I a 3 o 03 fa 1 1 0) Ig ■?3 ft "3 o J, 03 cd 8 >> a '5 rt f a 13 iw x'C 03 =3 £ » >, K 03 c 03 ■« >>£ o5 _>> 6 5- • ^ I ■ 3 1 a a . C/3 03 .,->' X ° 3 >. t« o5 'Tj'o fa,0 S? "«*- 9*' 3 O rt r°~ X >- O ta* 3 ^ O a! fee £ 3 S 03 -i_> **-< O -M ° v- - «3 3 fc 3 <3 aa l 1650 1800 69 110 C - 1800 R- 60 103 180 18463 10250 241 45.6 28.6 2 1400 2000 58 110 C - 500 R- 112 30 100 4420 2210 40 277.0 145.5 3 617 712 26 80 C - 300 R- 75 53 190 4625 6500 41 196.0 63.7 4 170 227 7 22 C - 100 R- 50 66 175 2225 9810 25 88.0 28.0 5 360 430 15 40 C - 290 R- 30 72 130 3127 72SO 34 119.2 44.8 6 1450 1600 60 100 C - 973 R- 80 66 125 6285 3920 54 186.0 111.6 7 3880 6000 161 280 C - 5000 R- 100 85 100 17125 2865 206 136.0 78.1 8 3200 4100 133 175 C - 500 R- 100 15 125 10900 2660 104 168.8 79.0 Aa 'erages 61 141 5686 145.8 72.4 1 In holder capacity column, the abbreviations Holder, respectively. 'C" and "R" stand for City Holder and Relief Purifier Operation Table i. — Purifier load conditions in medium-si^e mixed-gas plants of Illinois. 39 5J 3 >■ >> | S-. I ft o J, ft o cS ft en r CD o I 6 . 0£h b en El- I- o vfci 1 £ (O-j-i £ ,£ >, ^ 1 rf 4^ ft«w cij O CtJ +j" o S > 2+- >, c ti E C CD U CD ft, cti "^ cd ctj"£ Sft W ft .S g £2 PQ « II en ^3 3gs deft *g CD 03 . CD'rj ^ft H140 / /I, / / / O K'I20 // / / S 1 E 100 £ /c \ ^ An 7r Q H 60 /! D /: / 3 4 Z / o f/ >" 20 < ^ ^ / 20 00 GO 00 '.V. 100 H XX) 180 00 22< 300 26( MO BUSHELS OF OXIDE IN USE. Figure 3 — Observed relation of oxide volume to hourly purification capacity. through a series of boxes containing new oxide at such a rate that hydrogen sulphide could be detected almost immediately at the outlet of the system. In testing oxides on a laboratory scale, this condition is found and is made use of to determine the relative rates of reaction of various oxides. The time of contact called for by the Steere formula is approximately 6 minutes. Many comparatively slow oxides will not give a test for 44 Gas Purification in Medium Size Gas Plants hydrogen sulphide immediately on a gas containing 100 grains per 100 cubic feet unless the time of contact is reduced to approximately 0.5 minutes. Other more rapid oxides will absorb all of the hydrogen sulphide from gas containing 100 grains of H 2 S for several hours with this time of contact. As the oxide becomes fouler, however, the rate of absorption slows down and eventually some hydrogen sulphide will pass by the box. This will happen more quickly, other things being equal, in a purifying system that is overloaded. The result is that in an overloaded plant, other conditions being the same, more frequent changes will be necessary to keep the gas clean. When the frequency of changes becomes excessive the operator usually has one of the following choices; namely, to enlarge the purifying equipment, to rearrange existing equipment, to improve opera- tion by increasing the absorption per bushel through more complete revivi- fication in place, in some of the ways already discussed, or to find a more active oxide. Sometimes in the more extreme cases, only the first alterna- tive will prove a feasible, permanent remedy. REARRANGEMENT OF EQUIPMENT TO INCREASE CAPACITY As just mentioned, the capacity of a purifying installation may some- times be materially increased by minor changes whereby the existing equip- ment can be used in a more advantageous way, or the existing equipment may be enlarged, without changing its position or increasing the ground space occupied. Again, internal changes are possible whereby the puri- fiers may be made to accommodate a greater volume of oxide. Frequently, where the existing equipment is in good condition, such changes may be made at a fraction of the cost of an entirely new installation. Where additional capacity is needed it w T ould often pay to consider (1) whether the best possible performance is being obtained from the present equipment, and (2) whether some minor changes would not secure sufficient additional capacity to defer the installation of new equipment to another time. Of course, in this as in any other construction, there is always some uncertainty as to the relative cost of installation now or a few years hence. Some of the rearrangements which may be made have been suggested in the detailed descriptions of the purifying installations inspected. In the first place, it may be well to consider whether the amount of oxide is the maximum that the boxes will accommodate or whether the layers are of the greatest thickness permissible. In several instances it was found that the available space for oxide in the boxes, even allowing for the trays and a reasonable amount of free space, was considerably greater than Purifier Operation 45 would correspond to the number of bushels said to have been purchased for the boxes. Again the amount of free space allowed may be excessive. In some cases the trays could be relocated without unduly diminishing the free space and with a distinct increase in capacity. Take, for example, a 4-box set of purifiers arranged for reversible flow, each box being 15 feet in diameter and 12 feet high, containing two layers of oxide each \y 2 feet in depth. The hourly capacity, according to the Steere formula, would be 61,845 cubic feet, if the gas contained 200 grains of H 2 S per 100 cubic feet. Now, assuming that the depth of each layer could be increased by 6 inches, the capacity would become 66,262, an increase of capacity of 14 per cent. If the space available for oxide is being utilized to the fullest extent, then rearrangement may be in order. As has been suggested in several cases, reversible flow may help considerably. Let us assume an installation of three rectangular boxes, each 25 by 25 by 12 feet, each containing two 5-foot layers of oxide equipped for straight flow only, and with a given sequence as ABC, BCA, CAB. Such an installation purifying gas containing 200 grains of H 2 S per 100 cubic feet w T ould have a maximum hourly capacity, according to the Steere formula, of 133,000 cubic feet of gas. Let us assume that this installation is overloaded and that it is desired to increase its capacity about 35 per cent. This increase may be accomplished in either of two ways, namely, by making the three existing purifiers perfectly flexible as to arrangement and reversibility or by install- ing a fourth box of the same size as the existing boxes, arranged for straight flow only. Leaving out of consideration for the present the difference in operating cost which would be in favor of the former arrangement, let us consider -the probable relative costs of the two arrangements. At present prices, the cost of a new box of the size given, arranged for straight flow only, would probably be somewhere between $12,000 and $16,000. The installation of three 6-inch reversing valves, together with alteration of the manifold whereby the three existing boxes could be made entirely flexible as to arrangement and reversibility, would cost somewhere between $6,500 and $11,000, depending of course upon the amount of work and material that would be required and the amount of material from the old manifold that could be applied to the change. It is evident then that if the existing purifiers were in good condition and conveniently arranged for operation, rearrangement would be decidedly cheaper than the addition of another box of the same type. The economy of operation would also be distinctly in favor of rearrangement. A number of installations have been observed 46 Gas Purification in Medium Size Gas Plants in which such a change with corresponding increase in capacity was apparently feasible. In a few cases where old 4-box sets with center-seal connections are in use, a change in the connections whereby all the boxes could be used at one time would be advantageous. A few cases have also been observed where a small increase in capacity could be realized by putting into series boxes now arranged in two parallel groups, but in most of the cases of this kind observed, such a change would hardly give enough increase in capacity to pay unless all the boxes were made reversible and arranged so as to be used in any desired sequence. In most cases where two parallel groups are in use, one of the groups consists of shallow single-layer boxes. Frequently these shallow boxes are of con- siderable cross-sectional area and where substantial foundations exist and plenty of head room is available, it might be feasible to build up the boxes to double height, installing an additional layer of oxide. Such a rebuilt group of boxes, if arranged for reversible flow and for rotation with the boxes of the deeper group, also equipped for reversible flow, would usually increase the capacity of the system very materially. For example, in Plant No. 6 already described, such a reconstruction, if feasible, would increase the capacity by about 66 per cent. Of course, in making any alterations of the kinds described several things have to be considered. It would obviously be unwise to go to con- siderable expense to alter the connections of or reconstruct boxes which through long service had become unsound. And it might be unwise to pro- long the use of boxes so arranged that the cost of operation was excessively high on account of inconvenient location or poor facilities for handling oxide. In any case it would be advisable to carefully compare the cost of rearrangement or reconstruction with the cost of new equipment necessary to give an increased capacity equivalent to that expected from the proposed change. The possibility that the space now occupied by purifiers especially when in substantial buildings, might be used to advantage eventually for some other purpose should also be considered. Since present practice is almost unanimously in favor of outdoor purifiers, it might be obviously unwise to perpetuate the use of valuable buildings for this purpose. The economic as well as the physical features of such a change need considera- tion. In all cases where possible changes have been suggested in this paper, it is to be understood that only the results to be expected from such changes have been considered. The considerations just named and physical conditions existing in the various plants might make the suggested altera- tion entirely impracticable. Each case would have to be considered care- fully by itself. Purifier Operation 47 RESULTS OBTAINED IN PLANTS INSPECTED As will be seen clearly in Tables 3 and 4, purifier capacity alone will not necessarily insure good purifying results. Some of the best results found in Illinois plants at the present time are in overloaded plants. A study of Tables 3 and 4 shows how difficult or impossible it is to harmonize the results actually obtained with the conditions under which they were obtained. In general, one cannot but be impressed by the differ- ence between the results shown and those generally considered as typical of the best practice. Text books and treatises on gas manufacture usually state that spent oxide will contain from 50 to 60 per cent of sulphur, thereby implying that this degree of sulphiding is usually attained before the oxide is discarded. Yet the results shown in the tables indicate that in the average Illinois plant, at least, not nearly the usually accepted standard of performance is actually realized. In the water-gas plants, excluding Plant No. 5, which was a mixed-gas plant when the materal was fouled, the average percentage of sulphur in the spent oxide was only 21.7 per cent. In the mixed-gas plants, on the other hand, including Plant No. 5, the average was 37.4 per cent. These results are somewhat lower than those reported by Mr. Evans in the paper already referred to. The average absorption in the water-gas plants studied by him was 35 per cent sulphur and in the coal-gas plants 44 per cent. The latter figure is, we understand, for straight coal-gas, whereas the results reported by us in Table 4 are for mixed-gas plants. It is probable, too, that the plants whose results are reported by Mr. Evans are considerably larger in size than the plants in Illinois inspected by us, and therefore the conditions were probably more favorable for good results. Even so, it is apparent that the results gener- ally obtained fall considerably below those considered as good standard practice. In comparing Tables 3 and 4 one is impressed with the very consid- erable difference in sulphur absorption in the two gas-making processes. A number of reasons for the difference may be suggested. Coal-gas usually contains considerably more sulphur per unit volume than does water-gas, and according to the law of mass action, the greater concentration of the H 2 S in coal-gas increases the rate of absorption. The coal-gas production is also more uniform and the peak loads on the purifiers usually represent a greater rate of water-gas purification rather than of coal-gas, so the average load conditions with coal-gas are more favorable. The ammonia in coal- gas, of which traces pass through the purifiers, may also assist materially in the purifying process by keeping the oxide alkaline. A section of the Purification Committee of the American Gas Association is now studying the effect of this factor. 48 Gas Purification in Medium Size Gas Plants c^ ■lj -no jaj J9d -dj o © 00 o © C CN ;soo uonnoyuncl p3}oj, -' ~ - •%} -no p\[ jsd — }soo - o O oo 00 Ov © I/) © © lO spu9^eui uox^Boyuntj ■}j -no p\[ J9d — s;soo c r^ o co O o © © VO t^. joqB[ uouBoyunj T3 T3 • ;>, UOI}B9yiAIA9.I C cd * cd^ 6£o to o Ov CO jgd pgyund sb3 [bioj, vO cn oo vO vO vO o 00 lO Ov sbS pgyundun •Ij-noooi -lad S e H "S-K) o 00 o 3 t^ o CN © © Mi © apixo iuads 00 " ^ vO CO vO >* Ov lO A\ip "nq J9d -sqj — jtjj. CN CN lO © -* c CN gpixo }U9ds Aip CO CO CN rn CN © r- Ov o- •nq J9d -sq[ — .mqdpg vO "■> t^ »o VO CN ~ © Ov sis-eq Aip }U9D J9d CM ■* © vO o OO t^ CO. Ov 9pIXO }U9dS ui jb^l 00 CO CN M sisTjq cn O t^ ^, £ © CN 00 1^ A\ip ;u90 J9d — apixo }U9ds tii -lnqcqng 00 ^ vO lO o CO CO M CN CN pqsnq J9d "sqj vO - o> « CO CO lO rf o — gpixo ^uads iqSis/^ CO •>* VO CN CO vO CO CO © CO <* wj ui ,'r w M hn ex bfl C c c c C V-. O o o asri ui apixo jo pui^j X ,Q ^_- X X £> J_, ^_: 13 CO 0) cd 3 tJ 3 fl 3 3 cd ^) s J3 cd Pi tf £ ^ p^ ^ ^; ^ uoi^Dtipoid O C Ai_moq -xbui ")} -no vO lO 5 vO VO vO ]Aj J9d gpixo spqsng A^lOBdBO vO o o O C-) © © 00 00 jgyund -x-eui jo ^U9D t^ vO vO VO lO jad- — uq g>puu 'x-biaj ^ t^ Ov 00 00 co vO ^ •lj -no H o o ^ © o © C © ■ — uononpojd m - ^ b£ g < Ap-ep 93BJ9AY CO CO l^td ~ cn c^ Tf lO vO t^ 9C tn .2 rt rt^ >>+3 en ^"2 a 'J'xo *S O O +j -PC/3 w £ ^ O OJhrl O -.-. ni I-- ■p ° a •°cS§ CD ^-" O O S rt « ;hh Purifier Operation 49 $ ■*»' •}j *no "j^ jgd s^uao — isoo SuiXjund pajoj. o o 00 —i O CN o 1/3 O SO O <* to o o o IO CN o CN Os O 00 w * H ~ CN *-' •q.j -no •p\[ J3d S^U30 }SOO o o 00 o SO On IO o o o o o> s[bij3^bui uop-eoyunj 80 ■y -no ■y$ J3d <}U30 — ^soo O 00 o o so X o o o CN CN aoq"B[ uoi^Bogianj CO lO 00 Th CO LO >> >. >. ^, >» >. c a a C a C o o o c o o CO x x x X X x: X rZ o o o ( o .a ' {) £ UOiq.B3yiA.IA.3J JO poqj3J^ a> o c CJ a. o o O T3 "1 — T3 •: T3 — w w -►-> +i 3 3 3 3 r! O o O o c o O a T -;i -no -jai — -nq 00 CN oo 00 X <* CO CN jgd payund sb3 ibjoj. CN sb3 psyundun c CO O OS o o o CN c IO SO o o 00 Os CO ■%j -no 001 -xsd S Z H 'SJQ CN t*j CN CO "tf apixo qusds c- 1*5 •<* cn CO c CN CN c CO Ajp -nq aad -sqj — jbj, WJ CN ^< CO CN CN CN apixo qusds Aip CN r*. tn <- - oc 00 r- CO o •nq J9d - sqt 'anqd[ny c CN N -t 00 CN <0 CN r^. CN d CN <* SISBq Ajp }U3D J3d 1 = co rt CN o r^ 00 Tf SO SO 3plXO ^U3dS UI JBJ, CN lO cn CO >o * CO sisBq Ajp ^ "0 CN t -t "* -f ^U30 J3d — spixo quods ui mqdfrig 3 o C^l CO - CN SO CO CO •nq J3d -sen C" 00 ■* CO o a CN c o ^ — apixo qusds ^q3j3A\. If cn io O o 1^ oc o 1^ IO so ' j : bo - M • - DO . M t c '{ T3 V- 5) be .5 •.3 15 - 1 js C a *& c E- : 3S11 ui apixo JO pUl\J ,t] ~ - & T ' s ^ ?,% H^ ■d 9 Bin >/ eg O 3 c ffi2 5 C 3 Pt c^ pj pe M uoiqonpoid Of r^ A[.inoq - xbui -qi no o CO -t l^ e ir. ^< >n CO JAI aad spixo s[aqsng AqiOBdBO 00 O rj< lO c © SO aoyund -xbui jo quso 9 lO ^ -t PO -* .iod — uq 'XBUI 3>p3p\I a* cn CO 2 *"" IO CN CO CN — uoiqonpojd C o o 1 o lO § § S o o rt , * £ 4, p ii'l VO 00 00 ™ t» ^ ro 00 o 00 tN cs cs c ^ fO ■* VO ? > a i r i I B sJ ro ^h ^ «4H O. o O Jpq Purifier Operation 53 cerning the effect of tar in coating oxide and decreasing its absorption capacity. Of course, the ability of the material to revivify is also affected, so that tar is a disadvantageous thing to have present at any and all times. It was thought at first that the reason for the comparatively high sulphur percentages in some rather tarry oxides might be due to the tar- extracting action of a portion of the oxide batch, permitting the remainder of the batch to foul more completely on tar-free gas. It was expected that analyses of oxide from different levels in an upward-flow box would show a concentration of tar in the lower part of the box. Samples of oxide were therefore taken in two plants, one water-gas and the other coal-gas, at various levels. Each sample was collected from several points at the same level, to insure a representative sampling. Analyses of the samples and the levels where taken are shown in Table 5. It is surprising to note that the t.ar is not concentrated at the bottom. In the water-gas plant the greatest percentage of tar was found in the bottom of the upper layers, while in the coal-gas box, the top of the batch contained the most tar. Likewise, there is no apparent relation between the percentages of tar and sulphur in these batches. The highest percentage of sulphur in each case is coincident neither with the lowest, nor with the highest percentage of tar. One naturally concludes that each batch must have absorbed the larger part of its sulphur in each case before the tar was present to any great extent. The concentration of tar in the upper part of the batch may be caused by tar condensing out of the gas, due to the cooling action of the purifier box cover, and dropping down into the oxide. Of course, in prac- tice the accumulation of tar is relatively slow, and since the rate of sulphur absorption in a given part of an oxide batch is slowing down as the cen- centration of iron sulphide increases, it is difficult to determine in a partic- ular case to just what extent the accumulation of tar is affecting the per- formance. The means of removing tar differs considerably in different plants. Table 6 shows the gas-condensing and scrubbing equipment in use in the different plants. For convenience, the final tar-extracting apparatus in each case is printed in italics. It will be noted that the shavings scrubber appears to be the favorite tar-extracting equipment in water-gas plants, though the P. & A. tar extractor and the bubble washer are preferred in a few plants. The writers of this bulletin have had no opportunity to study the relative merits of these various types of apparatus under conditions which were comparable. As will be noted in the table, the amount of tar in the gas at the inlet of the purifiers varies considerably even with the same type of tar extractor, and it seems to be possible, under favorable con- ditions, to obtain practically complete extraction with any one of these 54 Gas Purification in Medium Size Gas Plants fl c o " H . d "3 o i-. . C al 03 E?-S a _• o be w Oh 2 '+3 "o 9, Ph &$ co "S CN ■*-^ — ~-- rtjju o *s < O CO 00 v6 O fO Ov r-t- O OOv o vo CO ro CN Ov oo Ov O © O00 < Od o CO Ov 00 o a "3 o a p4 t is era to 2* 1 o< d* <-> cocx, cu : Ih 03 P tN 1 ^p CO t3 c 0, 2 ft u 1-' CO CN^ "1 o 03 a o O a Ih CO CO 03 W CO CD O O T) B O pq ■d Ih O 0 00 ^< M5 vO <0 «o 1* 00 CO CO CN CN CN CN O S «3'x 11 (D 4) 0) cu £ oj c 03 a. - es •d *'O r d o moo pq aj ^ fO cj CN c> CN ro — < ~h ,-1 -h ,-< i-H^>O^N"^rtrtrtfO io,i-< CNr^vOCvt^oO^OOOOi^covOi-^CNOO »-"0 '-< ~* ^ CN ro CN CN CN CN H i-l -h h i-h CN 0000\OC^TJ iO "1 o o o o © oo £E (XI > N o OtOiOiOOOiOOiOiOiOOiOiOiOOOOOOOOOOOinOiO 0«; v_< *-< CN i-< CN CO •>* lO \0 I )0\0-hcncO'1 , »OvOi Appendix B 67 Where a box is emptied during the month, the total absorption by that batch during the month up to the time of emptying could well be entered in the column, preferably in red ink. Likewise the number of the batch substituted for the batch removed should be entered in the same column ahead of the daily entries. The total absorption by each batch while in a box is entered in the batch record, a suggested form of which follows : Batch Record Batch No / Kind of oxide Natural Bought from Date purchased 1-3-20 Wt. oxide per bu 25 Price per bu. (on ton) . . No. bushels in bat h 2500 Wt. new sponge per bu. (air dried) 35 $20 Ion Date put into purifier Box No. Date taken out of purifier Pounds H2S absorbed Equivalent lbs. sulphur Lbs. sulphur per bu. oxide Remarks 1/12/20 A C /, B C 7/6/20 2/15/21 8/17/21 2/25/22 6/20/22 23500 23000 20000 18000 8000 22100 21600 18820 16900 7540 8.85 8.65 7.54 6.76 3.01 8/1/20 3/16/21 41.0% sulphur 9/30/21 3/15/22 c; rded Totals 92500 86960 34.81 % sulphur in spent oxide .5 % tar in spent oxide. .2.0 % cyanides. Note: --If laboratory facilities are available, the percentage of sulphur in the oxide could be checked each time before the batch was returned to use. This would give mere accurate total results than the sum of the Tutwiler tests, though the latter are very useful in enabling the operator to know at all times the condition of his purifying material. The sulphur equivalent to a given weight of H^S is found by multiplying the latter by 16 and dividing the product by 17. The batch throughout its usefulness should be accompanied by a batch number, prhferably of metal, which is laid on the box cover or otherwise significantly placed while the batch is in the box and is stuck into or laid upon the batch while the latter is revivifying out of doors. The use of such a number will avoid mistakes as to the identity of a given batch. The above forms could of course be printed in book form or made up as desired. The absorption record would require a page per month, while the batch record might well occupy a page per*batch. The number of entries per batch would depend upon the practice in a given plant. Ordinarily one would not expect to handle a batch more than four or five times, perhaps less, but there are conditions which make many more handlings necessary. Any space remaining on the page after the entries might well be devoted to an extension of the remarks giving some information relative to the operating methods employed, rapidity of the material as to fouling and revivification, presence or absence of tar, fineness of the material before and after use, caking, etc. Also, if any laboratory tests of the material had been made prior to or during use, the agreement of these tests with practical results observed might eventually be valuable. 68 Gas Purification in Medium Size Gas Plants APPENDIX C Determination of the further usefulness of a given batch of oxide, according to the formula 1 of Fulweiler and Kunberger: 2 Average cost of purification ==A^ per M cu. ft. of gas Cost of new oxide =B^ per bushel Residual value of old oxide =C$ per bushel Cost of removing, revivifying, and replacing =D. Bulletin 2. Bulletin 4. Bulletin r>. Bulletin 6. Bulletin 7. Bulletin 8. Bulletin 9. Bulletin 12. Bulletin 13. Bulletin 91. Bulletin 100. ENGINEERING EXPERIMENT STATION URRAXA. ILLINOIS Coal mining practice in District VIII (Danville), by S. O. Andros, 1913. Coal mining practice in District VII, by S. O. Andros, 1914. Coal mining practice in District 1 iLongwall), by S. O. Andros, 1914. Coal mining practice in District V, by S. O. Andros, 1914. Coal mining practice in District II. by S. O. Andros, 1914, Coal mining practice in District VI. by s. o. Andros, 1914. Coal mining practice in District III, by S. O. Andros, 101."). Coal mining practice in District IV. by S. (). Andros. 1P1.">. Coal mining in Illinois, by S. O. Andros, 1915. (Complete resume of all the district reports.) Subsidence resulting from mining, by L. B. Young and II. II. Stock. 1916, Percentage of extraction of bituminous coal witb special reference to Illinois conditions, by (\ M. Young. 1017. U. S. BUREAU (IF MINES WASHINGTON. I). C. Bulletin 72. Occurrence of explosive gases in coal mini's, by X. II. Darton, 1!>1.~>. Bulletin 83. The humidity of mine air, by R. Y. Williams, 1014. Bulletin 99. Mine ventilation stoppings, by R. Y. Williams, 1915. Bulletin 102. The inflammabilitv of Illinois coal dusts, by J. K. Clement and L. A. Scholl, Jr., 1916. Bulletin 137. Use of permissible explosives in the coal mines of Illinois, by J. R. Fleming and J. W. Koster. 1917. Bulletin 138. Coking of Illinois coals, bit F. K. Oritz, 1917. Technical Paper 190. Mfithane accumulations from interrupted ventilation, with special reference to coal mines in Illinois and Indiana, by II. f. Smith and Robert J. Hamon, 1918. 1 Bulletins listed in italics apply directly to the problem of use of central district bituminous coals in place of eastern coal and coke.