Feb., 1918
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
TABI,E111-SOLUBILITY TESTS ON COALASH
FOR
Kz0
Sample Heated to Glow over Meker Burner for 1 Hour Total Kz0 in Ash, 4.47 Per cent How TREATED Sol. KzO Per cent H20 in Sample Washed on filter paper.. 0.10 1/2 hour boil.. 0.16 1 hour boil.. 0.23 2 hours boil.. 0.23 4 hours boil.. 0.23 7 hours boil 0.25 10 hours boil 0.28 15 hours boil.. 0.28 24 hours boil.. 0.25
..... .................... .. .............. .. .............. .. .............................. .. .. ..............................
..............................
Samples of kiln coal ash containing 4.47 per cent
KzO were heated t o a glow for approximately one hour over the Meker burner. This showed no volatilization, Two gram samples of this ash were then boiled vigorously for different periods up t o 24 hrs. t o determir.e whether this could be made water-soluble upon boiling. Table I11 shows t h a t after I hour’s boiling there is no increase in water-soluble KzO. This water-soluble is so small as t o be negligible. While this does not reproduce t h e kiln condition exactly i t is quiie evident t h a t a t t h e temperature attained b y t h e ash in passing through the kiln and for t h e period t o which it is subjected t o this temperature, no potash of t h e ash will be volatilized. Assuming t h a t go per cent of this ash in a dry process plant and 7 5 per cent in a wet plant passes up t h e flue with the gases the effect of t h e insoluble KzO in t h e ashes will affect t h e nature of t h e treater dust very appreciably. Due t o observations made while assisting in t h e research work a t t h e Security Cement & Lime Company a n d t o d a t a showing potash balances a t different plants a t hand, b u t which we are not a t liberty t o publish a t this time, it is quite evident t h a t t h e insoluble potash content of t h e dust collected from t h e kiln gases is of two sources, t h a t which passes over with t h e raw material, or partially calcined dust, and t h a t which passes over with t h e ash from the coal used in burning. CONCLUSIONS
I--K20 content of coal ash is considerable. II-K20 content of coal ash must not be disregarded in calculating t h e liberation in kilns. III--K20 content of coal ash appears in “treater dust” as insoluble K 2 0 . IV---Taking into consideration t h e KzO content of ash and t h e KzO in raw mix carried over mechanically there is apparently no “recombination,’ of the volatilized Kg0 with t h e siliceous ash particles. MICHIGANPORTLAND CEMENTCOMPANY CHELSEA,MICHIGAN
TOLUOL RECOVERY AND STANDARDS FOR GAS QUALITY B y R. S. MCBRIDE Received December 29, 1917
T h e removal of benzol and toluol from gas necessarily reduces t h e heating value a n d candle-power of t h e gas; t h e amount of reduction depends upon t h e quantity of these constituents originally in t h e gas, t h e thoroughness of washing, a n d t h e general charact e r of t h e gas with respect t o other heating a n d lighting 1
Published with permission of the Director, Bureau of Standards.
111
constituents. I n this paper it is intended, first, t o present a summary of t h e more important conditions of operation which determine t h e probable effect of toluol recovery upon t h e quality of t h e gas supplied; second, t o illustrate t h e method of estimating t h e probable effect in any particular case; a n d third, t o summarize certain general recommendations as t o changes i n standards t h a t must be made i n order t h a t recovery of toluol can be carried out effectively in a large number of localities. I n this memorandum, changes of standards are considered only from t h e standpoint of toluol recovery. N o consideration is given t o any other factors which might properly in many cases make desirable a change of standards. Such matters would depend upon a number of factors, not within the scope of t h e present discussions. SUMMARY O P P R E S E N T STANDARDS O F GAS QUALITY A N D GAS COMPANY OPERATING CONDITIONS
Both the heating value and t h e candle-power of gas are used in this country as a measure of the quality of the product supplied. Usually only one of these t w o characteristics is prescribed by ordinance or administrative ruling, b u t in some cases both are fixed. I n cases where such standards have not been adopted and the quality of the gas supplied is determined by the local gas company, i t is of interest t o know what quality of gas is being supplied. This information is presented below; for convenience of consideration t h e companies are classified according t o t h e standard in force. Data are included for all American gas companies making ~ o o , o o o , o o o cu. f t . or more of coal, water, or oil gas per year and for such other companies as have been recommended t o t h e Ordnance Department for consideration b y the Sub-committee on CoalT a r By-Products. I-Gas companies in t h e following cities are expected t o supply gas in compliance with t h e candlepower requirement as follows: . ( a ) Requirements of
20
candles or higher:
New York City (including the New York Consolidated System, the Brooklyn Union Gas Company, and the King’s County Lighting Company), 2 2 c.-p. (Permission has recently been given to change to a heating value of 650 B. t. u. a t the same price or to any lower heating-value standard if a proportionate reduction in price of gas is made.) Philadelphia, Pa., 2 2 c.-p., fixed by a franchise contract with the city. Des Moines, Iowa, 2 2 c.-p. Sioux City, Iowa, 2 1 c.-p. Omaha, Neb., 2 3 c.-p. measured at the gas works, or 2 1 . 2 c.-p. measured at the city testing station, and 600 B. t. u. heating value. Charleston, S. C., 20 c.-p. and 600 B. t. u. East St. Louis, Ill., 20 c.-p. (an old city ordinance requirement) and 565 B. t. u. Northern Illinois cities supplied by Public Service Company of Northern Illinois, 2 2 c.-p. (ordinance) and 565 B. t. u. (state standard). ( b ) Requirements of 18 candles: Detroit, Mich., 18 c.-p. and 600 B. t. U . Lansing, Mich., 18 c.-p. and 600 B. t. u. net heating value. Los Angeles, Cal., 18 c.-p. and 600 B. t. u. (most of the gas
I12
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
supply of this city is natural gas which is not limited by these requirements). ( c ) Requirements of 16 candles: All cities of Massachusetts, 16 c.-p. (fixed by State statute but subject to some waiver for purposes of investigation by the State Board of Gas and Electric Light Commissioners, This Board has very recently recommended to the State legislature a standard of 5 2 8 B. t. u.). The following Massachusetts cities are of interest in this connection: Boston, Brockton, Cambridge, Fall River, Haverhill, Lawrence, Lowell, Lynn, Malden, New Bedford, Pittsfield, Springfield, Worcester. Nashville, Tenn., 16 c.-p. and 600 B. t. u. Jackson, Mich., 16 c.-p. (20 c.-p. for water gas) and 600 B. t. u. Grand Rapids, Mich., 16 c.-p. and 600 B. t. u. Peoria, Ill., 16 c.-p. and 565 B. t. u. ( d ) Requirements of candle-power less than 16: Minneapolis, Minn., 15 c.-p. and 600 B. t. u. Birmingham, Ala., 15 c.-p. and 575 B. t. u. St. Paul, Minn., 14c.-p. and 600 B. t. u. 11-Gas companies in t h e following localities supply gas in compliance with heating-value requirements as follows: ( a ) Total heating value, 600 B. t . u.: St. Louis, Mo. (This is a municipal requirement; the State requirement is 570 B. t. u.) Baltimore, Md. Tacoma, Wash. Indianapolis, Ind. Milwaukee, Wis. Hammond, Ind. Madison, Wis. Atlantic City, AT. J. Peru, Ind. South Bend, Ind. Elizabeth, N. J. Cedar Rapids, Iowa. Jersey City, N. J. Washington, D. C. Newark, N.J. Wilmington, Del. Paterson, N. J. Seattle, Wash. Trenton, N. J. ( b ) Total heating values, 585 and below:
San Diego, 550 Denver, Colo., 575 Allentown, Pa., 570 Bridgeport, Conn., 575 Chester, Pa., 570 Hartford, Conn., 575 Reading, Pa., 570 New Haven, Conn., 575 Wilkes-Barre, Pa., 5 7 0 Waterbury, Conn., 575 Manchester, N.H., 565 Portland, Ore., 570 Chicago, Ill., 565 Ardmore, Pa., 570 Oakland, Cal., 550 San Francisco, Cal., 550 Cities of New York State, 585. (Of these cities the following are of interest in this connection: Albany, Binghamton, Buffalo, Poughkeepsie, Rochester, Schenectady, Syracuse, Troy, Utica.) 111-The gas companies in the following cities have no requirements limiting t h e candle-power or heating value of the gas which t h e y supply b u t are reported t o be supplying gas of candle-power and heating value as given below: New Orleans, La., 22 c.-p. and 600 B. t. u. Jacksonville, Fla., 20 c.-p. and 580 B. t. U. Atlanta, Ga., 19 c.-p. and 600 B. t. U. Richmond, Va., 18 c.-p. and 590 B. t. u. Pawtucket, R. I., 17 c.-p. and 600 B. t. u. Providence, R. I., 17 c.-p. and 600 B. t. U. Salt Lake City, Utah, 17 c.-p. and 600 B. t. U. Houston, Tex., 17 c.-p. and 585 B. t. u. Mobile, Ala., 15 c.-p. and 600 B. t. U. Portland, Me., 15 c.-p. and 570 B. t. u. Savannah, Ga., 575 B. t. u. Battle Creek, Mich., quality not reported. San Antonio, Tex., quality not reported.
METHOD
OF
ESTIMATING
INFLUENCE
Vol.
IO,
No.
2
O F TOLUOL
R E C O V E R Y U P O N GAS Q U A L I T Y
As previously stated t h e quantity of toluol or benzol in t h e gas initially is a large factor in determining t h e quality of the gas both before and after removal of t h e toluol, since t h e conditions which make for the presence of large quantities of these aromatic hydrocarbons are t h e conditions prevailing during t h e production of high-candle-power and high-heating-value gases. I n general t h e quantity of toluol and other light oils present in water gas depends upon t h e amount of gas oil used i n t h e production of this gas. Approximately I O per cent of t h e volume of gas oil used can b e recovered as crude light oil a n d of this amount from '/& t o l/6 can be recovered as pure toluol. Coal gas made b y any of t h e usual horizontal retort processes, which are t h e only processes of coal-gas manufacture requiring particular consideration in this report, usually contains about one-fourth t o one-third of a gallon of light oil per 1000 cu. ft., depending upon t h e character and treatment of t h e coal a n d t h e quality of t h e gas. From one-eighth t o one-tenth of this light oil is recoverable as pure toluol. For each one-tenth gallon of light oil removed per 1000 cu. f t . of gas the t o t a l heating value is reduced by approximately I O t o 14 B. t. u. per cu. f t . and t h e candle-power by z1/2 t o 3 candles. However, restoring p a r t of t h e light oil removed, for example, enriching with t h e benzol fraction, may in some measure compensate for the loss in heating value and candlepower brought about b y t h e initial washing. I n fact, if a sufficient amount of additional benzol is available the candle-power and heating value can be restored substantially t o t h e original values. (The increase in quality is about t h e same per unit of volume of benzol returned as was t h e loss on removal of the light oil.) However, this practice would not generally be practicable since i t demands t h e purchase of benzol or other enriching constituents t o t a k e the place of those constituents which are permanently removed from t h e gas. I n estimating t h e loss of candle-power and heating value, t h e figures here presented are probably slightly higher t h a n would correspond t o t h e change in quality of gas a t the customers since in distributing unwashed gas there is usually considerable loss due t o condensation. From these two generalizations and a knowledge of t h e initial candle-power a n d heating value of t h e gas i t is readily possible t o estimate approximately t h e influence upon t h e quality of t h e gas of recovering different amounts of toluol or of toluol and benzol. Such estimates are, of course, not exact, but they furnish an excellent guide for readjustment of standards i n any case where this is necessary or for approximating the quantity of materials which can be obtained b y washing t h e gas. T h e following examples will make clear t h e application of t h e d a t a : ( I ) Assume water gas made from 4 gallons of oil per 1000 cu. f t . , having an open-flame candle-power of 2 0 and a heating value of 6 2 5 B. t. u. About 0.4 gallon of light oils per 1000 cu. f t . could be recovered from such gas with practically complete wash-
*
Feb., 1918
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
ing. If none of t h e benzol was returned t o t h e gas t h e result would be a gas of about 57 j B. t. u. and IO candlepower. If t h e light oil were distilled and t h e benzol fraction were returned t o t h e gas t h e loss in heating value and candle-power would be, perhaps, one-half a s great, and t h e result a gas of about 600 B. t. u. and I j candles. I n order t o restore t h e heating value and candle-power t o substantially their original figures it would be necessary t o add benzol t o t h e extent of approximately 0 . 2 gallon per 1000 cu. f t . of gas manufactured. From t h e 0.4 gallon of light oil originally obtained, some 0 . 0 7 t o 0.08 gallon of toluol would probably be obtainable on refining, ( 2 ) Assume carburetted water gas with 3 gallons of gas oil per 1000 cu. f t . and assume a very high oilefficiency so t h a t t h e candle-power was 18 and t h e heating value 5 7 0 B. t. u. From this gas about 0.3 gallon of light oil, equivalent to, perhaps, 0.05 t o 0.06 gallon toluol, would be obtained per 1000 cu. f t . with commercially complete washing. The result of this washing would be a gas approximately 530 B. t. u. and 9 t o I O candles, which would be restored t o 5 5 0 B. t. u. and about 12 candles if re-enriched with t h e benzol portion of t h e light oil. (3) A mixture of coal gas and water gas in about equal proportion may be assumed made from water gas for which 3 l / 2 gallons of gas oil were used t o give 18 candles and 600 B. t. u. and coal gas of 580 B. t. u. and 1 5 candles. Such mixed gas would yield, perhaps, 0.3 gallon light oil per 1000 cu. ft. and t h e average candle-power would be reduced b y washing from 1 6 l / ~ t o 8 or g candles and t h e heating value from j g o B. t. u. t o about 5 5 0 B. t. u. Restoration of t h e benzol fraction would give a product of about 12 candles a n d 5 7 0 B. t. U. (4) A coal gas made from ordinary grade of gas coal
t o yield 10,000 cu. ft. of gas per t o n is assumed t o pro.duce gas of about 14 candles in t h e open flame and of 585 €3. t. u. From this approximately 0.3 gallon of light oil per 1000cu. f t . of gas could be recovered; and from i t 0 . 0 2 5 t o 0.03 gallon of toluol per 1000 cu. f t . would be available. The gas after washing would have approximately 8 candles and j j o B. t. u. which would be increased t o , perhaps, 1 2 candles and 570 B. t. 11. if t h e benzol fraction were restored. I n any of t h e above cases t h e net loss in heating value and candle-power might readily be reduced if some of t h e other constituents of t h e light oils, such as t h e solvent naphtha fraction, were also restored t o t h e gas; or t h e loss in heating value and candle-power could be made less by operating t h e washing equipment in such a way as t o accomplish only a partial removal of t h e light oils. In the latter case if removal of only 7 5 per cent of t h e quantity of light oil readily obtainable were considered satisfactory this would make t h e losses i n heating value and candle-power of only about 3/4 as great as above indicated, b u t , of course i t would also somewhat reduce t h e yield of toluol. It is probable t h a t complete washing of t h e gas with restoration of t h e benzol in most cases will be considered advisable since t h e need for toluol is considerable and very little sacrifice of toluol yield can be
113
allowed. But t h e demand for benzol is not so great and t h e restoration of t h e benzol t o t h e gas might give results a t first more satisfactory t o t h e gas users t h a n would t h e sale of this benzol with t h e slight reduction in total costs for t h e gas which might possibly be accomplished thereby. Especially might t h e restoration of benzol be necessary &'here a high candle-power standard has been in force, since otherwise t h e loss i n candle-power would be rather greater t h a n would be desirable a t one time. I n any computation, therefore, it is probably best t o assume, unless other basis is known t o be correct, t h a t commercially complete washing of t h e gas would be necessary and t h a t t h e benzol fraction of t h e light oil amounting t o approximately one-half t h e total volume of light oil removed, will be restored t o t h e gas. RECOMMENDATIONS
REGARDING
STANDARDS
FOR
GAS
QUALITY
From t h e estimates in t h e preceding section it is evident t h a t much greater difficulty is met in complying with a candle-power requirement after removal of toluol or light oil t h a n is encountered if a heating-value standard is t o be complied with. Because of this fact it seems desirable t h a t in any case where toluol is t o be removed t h e candle-power standard be altogether eliminated or be made sufficiently low so t h a t it will not interfere seriously with t h e proposed operations. Many other factors independent of toluol recovery make evident t h e desirability of eliminating candle-power requirements and substituting heatingvalue requirements as t h e primary basis of gas measurement. Therefore, t h e war is only an added influence tending t o hasten an end otherwise desirable. I n all cases where t h e candle-power has previously been below 18 it would seem t h a t t h e elimination of t h e candle-power requirement altogether would be reasonable; although in any event it is expected t h a t t h e company would supply a gas of a t least 8 or IO candles, which would be sufficient t o care for t h e need of those customers who must use some portion of t h e gas for open-flame lighting. I n cases where 18 or 2 0 candles or higher have been maintained regularly in t h e past, it might be undesirable t o have t h e candlepower go below 1 2 t o 14, unless open-flame lights were generally eliminated and a readjustment of t h e appliances of a 1 customers were made wherever t h e change in quality might make this necessary. For all companies which have been complying with requirements of 18 candle-power or higher, an understanding might be reached as t o t h e maintenance of a t least 1 2 t o 14 candles for such a period as might be necessary t o accomplish a general adjustment t o t h e new conditions. When coal gas is supplied either alone or mixed with very small percentages of water gas it is impracticable t o make a very rich gas since t h e character of t h e coals available i n most instances would preclude economic operation if a higher standard, either of heating value or of candle-power, must be maintained. For cities where only coal gas is supplied t h e standard could scarcely be higher t h a n about 570 B. t. u. if practically complete toluol recovery is expected. Higher heating-
1I 4
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
value standards t h a n this would probably have t o be modified for such gas supplied. If water gas is manufactured either alone or as a major constituent of t h e supply i t is entirely practicable t o make a gas of reasonably high heating value a n d candle-power initially and have after removal of the toluol a heating value of 585 t o 600 B. t. u. I n each case i t would be a question as t o which procedure was t h e more economical; t h a t is, whether i t would be better t o make the same quality of gas as h a d previously been supplied a n d supply t h e customer with a somewhat lower product t h a n formerly after t h e toluol had been removed from t h e gas, or t o make t h e . gas initially somewhat richer t h a n before b y t h e use of slightly more gas oil per 1000 cu. f t . so t h a t t h e product after washing would have substantially t h e same heating value as had previously been supplied. If t h e quality previously supplied was rather high, approaching t h e maximum of t h e range of quality permissible for efficient operation, then a n y increase i n t h e initial quality would obviously be undesirable; but otherwise a n initial increase i n quality with subsequent washing down t o t h e original might be t h e best practice. Since t h e quantity of toluol available is largely dependent upon t h e initial richness of t h e gas which has been washed there is considerable advantage from t h e standpoint of t h e Government i n having t h e richest practicable gas made initially; b u t , of course, i n a n y case t h e limits of economical operation must be clearly recognized, a n d conservation of oil might also be a n important factor. As a summary of these points t h e following suggestions are offered as desirable adjustments t o facilit a t e t h e recovery of toluol: I-Eliminate all candle-power requirements now i n force except for t h e cities where 18 candles or higher has been supplied, i n which localities reach a n understanding t h a t a t least 1 2 candles will be maintained for a period, say a year, during which time readjustments of appliances a n d substitution of mantle lamps would be accomplished t o such a n extent as t o justify complete elimination of candle-power regulations. 11-For plants making coal gas (or practically only coal gas) let t h e heating-value standard be from 550 t o 5 7 0 B. t. U . 111-For plants making water gas, either alone or as a major constituent, let t h e heating-value standard be 570 t o 600 B. t. u. monthly average total heating value, t h e adjustment being made between these limits according t o t h e economic conditions of operation. I n order t o show t h e number of companies t h a t would be affected by these several recommendations t h e following tabulation of t h e companies above listed is prepared. This tabulation does not t a k e account of a n y unusual local conditions which might affect some of t h e cases materially. The kind of gas manufactured is also indicated: W, water gas; C, coal gas; 0 , oil gas; M , mixed coal and water gas; B, by-product coke oven gas; and N , natural gas. I-Localities in which n o change of standard will probably be needed and no serious change i n t h e quality of gas supplied will probably result:
Vol.
IO,
No.
2
Chester, Pa. W+B Schenectady, N. Y. M Reading, Pa. W . Troy, N. Y. W Wilkes-Barre, Pa. W Utica, N. Y. W Ardmore, Pa. W Poughkeepsie, N. Y. W Allentown, Pa. W Syracuse, N. Y . M Portland, Ore. 0 Binghamton, N . Y . W Manchester, N. H. M Rochester, N. Y. M Hartford, Conn. M Chicago, Ill. W+B+N Bridgeport, Conn. W San Diego, Cal. W+O New Haven, Conn. M San Francisco, Cal. 0 Waterbury, Conn. W Oakland, Cal. 0 Denver, Colo. M Houston, Texas W San Antonio, Texas W Pawtucket, R. I. M Savannah, Ga. W Providence, R. I. M Jacksonville, Fla. M Battle Creek, Mich. M Richmond, Va. M Portland, Me. M Atlanta, Ga. M Salt Lake City, Utah M New Orleans, La. W Mobile, Ala. M Albany, N. Y. W 2-Localities in which a candle-power standard may have t o be abandoned, b u t with no serious change in t h e heating value of t h e gas supplied: Lynn, Mass. M Fall River, Mass. M Boston, Mass. M Haverhill, Mass. W Brockton, Mass. M Springfield, Mass. M Lawrence, Mass. M Malden, Mass. M Lowell, Mass. M Birmingham, Ala. M New Bedford, Mass. M Waterloo, Iowa. W Pittsfield, Mass. M Peoria, Ill. M Worcester, Mass. M Nashville, Tenn. M W+N Cambridge, Mass. M East St. Louis, Ill. 3-Localities in which slight change in heatingvalue regulations may, perhaps, be required, b u t i n no case probably more t h a n equivalent t o 5 per cent of t h e present value. (The six cities marked (*) have candle-power standards which should be eliminated also.) M B+W Buffalo, N. Y. Indianapolis, Ind. Milwaukee, Wis. M C+O Tacoma, Wash. W B+M Madison, Wis. Seattle, Wash. Cedar Rapids, Iowa. M B+M Trenton, N. J. Paterson, N. J. W St. Louis, Mo. M+B M Newark, N. J. Baltimore, Md. W+B Hammond, Ind. M W Jersey City, N. J. South Bend, Ind. M W Elizabeth, N. J. Peru, Ind. M Atlantic City, N. J. W WiImington, Del. W Washington, D. C. M *Grand Rapids, Mich. M W *St. Paul, Minn. *Jackson, Mich. M *Minneapolis,Minn. M M+B *Los Angeles, Cal. O+N *Detroit, Mich. 4-Localities in which high candle-power regulations should be changed or eliminated i n order t o permit operation on a heating-value basis; the reduction in heating value of the gas delivered would probably be a considerable percentage of t h e present value. ( I n the case of Lansing and Omaha a lower heating value t h a n now i n force would also be necessary.) New York City M Charleston, S. C. W W Des Moines, Iowa W Omaha, Neb. Lansing, Mich. M Sioux City, Iowa W Philadelphia, Pa. M Northern Illinois cities supplied by Public Service Company of Northern Illinois, M . BUREAU OF STANDARDS
WASHINOTON, D. C.