Jan., 1921
T H E J O U R N A L OF I N D U S T R I A L AATD E N G I N E E R I N G C H E M I S T R Y
we can meet these conditions. We work on about 35-lb. samples of coal. Just one other word in regard to the inquiry about whether the material is any good for metallurgical purposes. With Mr. Sperr in the room, I would hardly attempt to pass judgment on it in that particular. I asked a blast-furnace man not long ago to describe what a good blast-furnace coke was, He said, “After changing our minds so radically within the last few years, we are not absolutely sure as we once were.” Some will say that it is entirely unsuited for blast furnace use. It does violence, I think, to nearly everything that would ordinarily be described by a blast-furnace man as being necessary. I do not know, however, but that if we could make enough a t the rate of 35 Ibs. per run to supply a blast furnace for a week, we could find out for a certainty. I want to say in that connection that the initial incentive for all this work comes out of the great anthracite strike of 1902. We thought it would be desirable to make smokeless fuel for domestic purposes out of Illinois coal, and that has been the main idea all along. We do not know much about metallurgical coke, although I will say this, that so far as strength, and carrying the burden, and a lot of those physical conditions are concerned, it certainly looks very encouraging, but there are other conditions, like high ash, etc., which would enter into the problem. CARBONIZATION OF CANADIAN LIGNITE1 By Edgar Stansfield LIGNITE UTILIZATIONBOARD,OTTAWA,CANADA
The researches on lignite outlined i n this paper were commenced early i n 1917 b y t h e chemical staff of t h e Fuel Testing Division of t h e Mines Branch, Department of Mines, Ottawa, and t h e work is still in progress. T h e primary object of t h e investigation was t o obtain accurate d a t a essential for t h e scientific design and control of a plant for t h e carbonization of lignite on a commercial scale, rather t h a n t o design such a plant. I n t h e summer of 1918 t h e Lignite Utilization Board (of Canada was created by a n Order-in-Council of t h e Dominion of Canada, supplemented b y a n agreement as t o finances with t h e provincial governments ‘of Manitoba and Saskatchewan. T h e Board was created t o establish a n industry for t h e conversion of t h e low-grade lignites of southern Saskatchewan, and elsewhere, into a high-grade domestic fuel by means of carbonization and briquetting. The laboratory investigations of t h e Lignite Board have been carried out at t h e Fuel Testing Station of t h e Mines Branch b y members of t h e staff of t h e Board working in cooperation with t h e members of t h e Mines Branch Staff. This latter work has carried t o a logical conclusion t h e earlier work of t h e Mines Branch. T h e points essential for t h e successful carbonization of lignite, under t h e economic conditions prevailing i n southern Saskatchewan, were first decided upon, and then a carbonizer design was evolved which embodied these features. A semicommercial-scale carbonizer was erected in Ottawa, and, after many trials a n d modifications, successfully operated. It i s worthy of note t h a t t h e experience and information gained in t h e operation of t h e carbonizer a t Ottawa have been embodied by t h e engineer of t h e Puhlished by permission of Dr. Eugene Haanel, Director, Mines Branch, Department of Mines, Ottawa, Canada 1
I7
board, Mr. R. De L. French, i n t h e design of six carbonizers for a plant n o w being erected by t h 8 Board near Bienfait, Sask. This plant is expected t o t r e a t about 2 0 0 tons of raw lignite per day. This paper attempts t o trace i n outline t h e progress of t h e investigation u p t o t h e operation of t h e carbonizer in Ottawa, and t o show why this particular design of carbonizer was adopted. No full report of any stage of t h e work has yet been made, b u t t h e methods employed and results obtained i n t h e earlier stages have been published in some detai1.l T h e work falls naturally into several stages, b u t these are not chronologically distinct. T h e investigation was commenced with lignite from t h e Shand Mine in t h e Souris, or Estevan area, Sask. Later other Souris lignites were studied. Now Alberta lignites, and also peat, are being tested i n a similar manner. Souris lignite when mined contains from 3 0 t o 3 5 per cent of inherent moisture, and has a calorific It loses moisture value of about 4000 cal. per gram. rapidly when exposed, and t h e lumps then disintegrate. This lignite is employed in t h e raw state, b u t it is a low-grade fuel, unsatisfactory for transportation or storage. By drying and carbonizing it, a product is obtained which may have a calorific value as much as 7 5 per cent higher t h a n t h a t of t h e original coal. SJIALL-SCALE L A B O R A T O R Y TESTS
In these experiments samples of from 3 t o I O g. were employed. This allowed very exact control of t h e conditions of t h e experiment, and also allowed a large number of experiments t o be carried o u t , under widely varying conditions, within a reasonable time. It was not possible, however, t o s t u d y t h e by-products. T h e results were used t o cut down unnecessary work i n t h e larger tests, and were also valuable as checks on t h e accuracy of control i n all subsequent experiments, and for t h e comparison of different lignites. T h e factors determined included t h e yield, analysis, and calorific value of t h e carbonized residue. T h e conditions under which t h e lignite was carbonized were varied i n order t o show t h e influence on t h e results of t h e final temperature t o which t h e charge was heated, t h e r a t e of heating, t h e pressure i n t h e retort, and t h e atmosphere in t h e retort. C O A L USED-The particular coal chosen for most of these experiments was from t h e Shand mine of t h e Saskatchewan Coal, Brick, and Power Co., Ltd. T h e sample, which consisted of a single lump of coal shipped by express from t h e mine in a wooden box, was crushed and ground t o a fine powder i n a ball mill. For convenience of manipulation, and as a preventative of t h e rapid change which a powdered coal undergoes owing t o moisture loss and oxidation, this powder was briquetted i n a small hand press. T h e briquets were cylindrical, 0 . 2 ; in. i n diameter, about 0 . 2 5 in. long, and r a n about 5 or 6 t o t h e gram. They were stored in stoppered bottles until required, and from 1 Stansfield and Gilmore, “The Carbonization of Lignite,” Trans. Roy. SOC.Can.,[3] 11 (1917), 8 5 ; [3] 12 (1918), 121. See also Mines Branch Summary Reports for 1918 and 1919.
18
T H E J O U R N A L OF I N D U S T R I A L A N D ENGIJVEERING C H E M I S T R Y
time to time moisture control determinations were made upon them. It may be noted t h a t during a period of 2 mo. t h e moisture contents fe1.l only I per cent from an original of over 30 per cent. The gross calorific value of this coal was 4260 cal. per gram. Its average analysis was as follows: Moisture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ash.................................. Volatile m a t t e r . , . . . . . . . . . . . . . . . . . . . . . . Fixed carbon.. ........................
Per cent 3 1.8
5.2
28.9
34.1
APPARATUS-The apparatus used for most of t h e experiments consisted of a cylindrical iron retort 1.5 in. high and 1.5 in. diameter, inside measurement, having a lid which was held on by a small clamp, t h e joint being rendered airtight by means of an asbestos gasket. A small inlet tube was screwed into t h e bottom of t h e crucible, and an outlet t u b e into t h e lid, t h e inlet and outlet tubes being so arranged t h a t t h e retort could be completely immersed in a n oil or lead bath. For t h e experiments under pressure a slightly larger and heavier retort was employed, with a hexagonal screw cap rendered gastight with an asbestoscopper gasket. The inlet t u b e was dispensed with, and a pressure gage and relief valve connected with t h e outlet tube. llETHOD-The coal briquets were weighed out into a quartz crucible which fitted inside t h e iron retort. T h e heating was done by immersing t h e retort i n a bath, which for tests up t o 300' C. was of oil, and for
Vol, 13, No.
lt
those above t h a t temperature of lead. The lead was contained in a 4-in. length of 4-in. iron pipe with a capped end, and was heated in a gas-fired furnace which gave a very uniform temperature throughout t h e bath, and which permitted rapid heating and easy control. The temperature was followed by t w o pyrometers immersed in t h e lead. For t h e regular tests, t h e retort was plunged into t h e bath, previously heated t o t h e desired temperature. The temperature was kept constant until t h e evolution of gas ceased, and t h e retort was then removed, cooled, and opened, and t h e contents weighed and examined. I n other tests, t h e retort was slowly heated t o about 2 j 0 O C. in an oil bath, then transferred t o a just molten lead bath, and t h e temperature slowly raised t o t h e desired point. I n t h e vacuum tests, t h e pressure in t h e retort was kept below 2 j mm. of mercury by means of a good water pump. I n t h e steam tests, a slow current of steam was passed through t h e retort. I n t h e pressure tests, t h e relief valve was closed a t t h e beginning of t h e test, b u t was opened as required t o maintain t h e pressure in t h e retort, due t o t h e escaping gases, a t about 1 2 0 lbs. per sq. in. Dry coal was employed for t h e pressure series. A striking phenomenon, first observed in connection with t h e vacuum series, was later found t o t a k e place with every sample of dried or carbonized lignite. I n every case t h e residue rapidly gained in weight after removal from t h e retort, even when stored in a
Jan.,
1921
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
desiccator over sulfuric acid, its calorific value a t same time decreasing. This was later shown t o be mainly due t o an occlusion of air. All published results are, with a few stated exceptions, for weights a n d calorific values determined immediately after t h e experiment. Figs. I and a show i n graphical form t h e principal results obtained in t h e regular tests on one Saskatchewan a n d one Alberta lignite. I n every lignite tested t h e calorific value of t h e carbonized residue increases u p t o a maximum and t h e n decreases. The temperature for maximum calorific value lies between 5 joo and 6 joo C., varying with t h e lignite. But t h e yield of carbonized residue for maximum calorific value has been found t o be remarkably constant when expressed on t h e basis of t h e d r y coal taken. Five out of six samples taken from different areas in Saskatchewan and Alberta gave a maxinium value with about 67 per cent recovery, the sixth with about 71 per cent. L A R G E - S C A L E L A B O R A T O R Y TEST
I n these experiments the results determined include t h e yield and calorific value of t h e carbonized residue; t h e yield, composition, a n d calorific value of t h e gas generated; t h e yield, calorific value, and economic value of t h e t a r produced; and t h e ammonium sulfate yield available. The conditions under which t h e lignite was carbonized were, in t h e experiments here
I9
described, varied only t o show t h e influence on t h e results of t h e final temperature t o which t h e charge was heated, t h e rate of heating, and t h e moisture conditions of t h e coal treated. Further experiments have been commenced which show t h e effect of t h e pressure in t h e retort and t h e atmosphere in t h e retort. APPARATUS-The apparatus (Fig. 3) employed in most of these tests embodies three important features: Accurate temperature control. Reduction, as far as possible, of the temperature lag from the walls to the center of the charge. ( 3 ) Complete removal and easy collection of the tar vapors. (I)
(2)
The temperature control is effected by t h e use of a n electrically heated lead bath, B, with suitable thermal insulation. The bath rests on a movable platform which can be raised b y t h e screw C. The temperature is observed b y means of a pyrometer and regulated b y switches a n d rheostat. T h e reduction of lag is effected b y t h e use of a tubular retort, A. This consists of seven ra-in. lengths of 2-in. boiler tubing, mounted in a cast-iron head. No part of t h e charge is t h u s more t h a n I in. from'the walls of t h e retort, which has a capacity, t o t h e t o p of t h e tubes, of 2300 g. of pea-size lignite with about 3 5 per cent moisture content. I n later work, a castiron retort of cruciform cross-section was employed. This has a capacity of 3 joo g. COLLECTION O F TAR-A satisfactory method for collecting t h e t a r was evolved only after many weeks
20
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
of work and many failures. Not only was it hard t o remove the last traces of t a r fog, but t h e condensate was usually in t h e form of a watery emulsion, very difficult t o handle. The method employed was as follows: The hot gases leaving the retort passed down through t h e center tube of a small scrubber, D, made of iron pipe and containing three interlacing coils of wire, and passed up again through a surrounding annular space; the whole scrubber being jacketed with superheated steam. The heavy t a r oils were here condensed in a practically water-free condition, and dropped into a weighed glass beaker. The lighter oils, steam, and gases passed on and down through the simple tubular condenser E, where t h e two former condensed and collected in a receiver, the oils floating on the water and showing only a slight tendency t o emulsify. The cool gases leaving t h e condenser still contained some tar fog; they were therefore passed down through a tube scrubber, F, filled with glass beads and a thin layer of glass wool (shown shaded), through which a jet of steam from a weighed boiler was also passed. R
Vol.
13,
No.
I
H, and into a gas holder which is not shown in t h e figure. For temperatures above 700’ C. a smaller apparatus was employed, with no lead bath. The retprt consisted of a simple piece of 3-in. boiler tube, 16 in. long. It was heated by placing it inside a tube of 3-in. bore wound around the outside with a coil of nichrome A charge of 1000 g. was taken for all experiwire. ments with this retort. The temperature of t h e lignite was observed by means of two pyrometers, one in the center and one near t h e wall of t h e retort. XETHOD-In the regular series of tests, with rapid heating, the retort was charged, usually with peasize lignite containing about 34 per cent moisture, b u t in a few experiments with dried lignite, and connected t o the purifying train which was then swept out with gas from a previous run. The lead bath, heated t o a temperature higher than t h a t desired for t h e test, in order t o allow for the cooling effect of t h e retort, was then raised t o surround the retort. The temperatures and pressures a t t h e different parts of the system and also t h e meter readings were recorded a t frequent intervals, and t h e experiments continued until t h e evolution of gas had practically ceased. The gas volumes were corrected for temperature, pressure, and moisture content, being reduced t o moist gas at 60’ F. and 30 in. of mercury. All other products were weighed, and all the products were carefully analyzed. I n a number of the experiments the gas was collected in two separate holders, and t h e two portions were analyzed separately. The gas from t h e second half of the run is much richer t h a n t h a t collected in t h e first holder. I n some tests slow heating was tried, and in others the retort was evacuated, or was kept under pressure, or a slow current of steam was passed through. The results cannot be summarized. The following are a‘ few of the most important results obtained b y the rapid carbonization of Shand lignite a t 5 5 5 ’ C. WEIGHTBALANCESHEET(Dry Coal Basis) Per cent Water of decomposition.. 11.7 G a s . , ................................ 17.0
..............
................. ................. .................
4.1
66.7 0.5
THERMAL BALANCJ~ SHEET (Heat Content of Products as Percentage of Heat in Original
FIG.3-AePARArus
FOR LIGNITE C A R B O N I Z A T I O N
The bottom half of this scrubber was water cooled. This scrubber completely removed t h e t a r fog from t h e gas. The oil first condensed on the beads acted as an oil scrubber collecting more of the tar, t h e steam prevented t h e clogging of t h e scrubber by keeping the t a r hot and fluid, and also, when condensing a t t h e bottom, carried down with i t any vapors still remailiing. The gases were thus completely cleaned, and all the liquid products, as well as t h e ammonia, from the lignite were collected in t h e vessels and could readily be weighed and examined. The tar thus collected was reasonably free from water and could be redistilled without excessive bumping or frothing. The gases leaving t h e scrubber F passed through a final cooling tube, G, through a gas meter,
Gas.
.......
Charge)
.................. .................................
Carbonized residue.. 1,oss
78.1 7.6
COMXERCIAL PRODUCTS (Yields per 2000 Lbs. of Moist Coal Charged) Gas, cu. f t Ammonium sulfate, Ibs. 10.2 Tar, imp gal . . . . . . . . . . . . . . Carbonized residue, l b s . . 910
.................. ...... .. .............
The coal charged contained 3 I .8 per cent moisture. The gas had a gross calorific value of 385 B. t. u. per cu. f t . and a density of 0.94. The crude t a r had a density of 1.00. LOW-TEMPERATURE C A R B O N I Z A T I O N BY SHORT P O S U R E T O HIGH T E M P E R A T U R E S
EX-
Figs. I and z show t h a t t h e maximum calorific value of t h e residue is obtained b y carbonization a t a temperature of about 600’ C. It is clear from t h e shape
J a n . , 1921
T H E J O U R N A L OF I N D U S T R I A L AiVD E N G I N E E R I N G C H E M I S T R Y
21
T i m e ~n Minutes
FIG.4
of these curves t h a t if lignite is heated in a retort under t h e conditions usually met in commercial operations, with t h e layers near t h e wall very distinctly hotter t h a n those i n the center of t h e charge, no regulation of the average temperature of t h e mass will give a residue with t h e maximum obtainable calorific value. The amount which t h e calorific value of t h e residue falls below t h e optimum will increase with t h e thickness of t h e charge and with t h e temperat u r e gradient from t h e walls t o the center. MEmoD-Some preliminary experiments were carried out t o test t h e possibility of obtaining t h e equivalent of carbonization a t , say, 600' C., by short exposure in a thin layer t o a distinctly higher temperature. Samples of dried Shand lignite, crushed t o pass a IOmesh screen, were carbonized for a definite number of minutes in a metal box in a muffle furnace electrically heated t o temperatures of 750' t o 800" C. T h e boxes were 6 in. X 3 in. X I in., inside dimensions, of No. 18 gage sheet iron, with loosely fitting lids of t h e same metal. When making a test the muffle was brought up t o heat, and t h e lid of the box was also heated. A charge either t o half or quite fill t h e box was weighed o u t and placed in the cold box. T h e heated cover was p u t on, t h e box immediately placed on t h e floor of t h e muffle, and t h e muffle door closed. At t h e expiration of t h e desired time, t h e box with its contents was removed from the muffle, cooled as rapidly as possible, a n d t h e residue weighed a n d analyzed. No great accuracy is claimed for t h e results, which are shown graphically in Fig. 4. It is obvious t h a t the number of experiments should have been con-
siderably increased t o render the curves reliable. They do, however, show t h a t the results of such rapid carbonization follow t h e lines which theory indicates, and t h e advantage t o be gained b y further experiments was not thought t o be commensurate with t h e work involved. Comparison of the optimum results obtained with a 0.5-in. and I-in. layer with those obtained by complete carbonization of t h e same sample a t j g o ' C. and a t 600' C., show, as might be expected, t h a t t h e yield and composition of t h e residue is approximately t h e same in all cases, but t h a t t h e calorific values of 6760 and 67 50 cal. per gram obtained with temperat u r e control, fall t o 6690 a n d 6590, respectively, with t h e 0.5-in. and I-in. layers. BEARING
OF
RESULTS
ON
DESIGN
OF
COMMERCIAL
CARBONIZER
T h e primary object of t h e Lignite Utilization Board
is t o produce a domestic fuel from Souris lignite.
It
is therefore desirable, unless other reasons are found t o outweigh this, t o carbonize t h e lignite in such a way as t o give t h e residue with t h e maximum calorific value. It has been shown t h a t this is accomplished by complete carbonization a t a temperature of about 5 7 5 " C., and t h a t t h e same result can be approximated by short exposure in a very thin layer t o a distinctly higher temperature. As the object t o be attained is t o bring all parts of t h e mass t o t h e same optimum temperature, a somewhat thicker layer continually stirred should give t h e same result as a thinner layer a t rest. The economic advantage, in the way of reduction of capital cost of equipment, t o
22
THE J O C R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
be gained by t h e accereration of t h e process by the use of high temperatures is too obvious t o need amplification. N o increase i n t h e yield of by-products can be a t tained without a corresponding decrease in t h e yield and calorific value of t h e residue. The gas obtained atIthe above temperature is barely sufficient t o provide t h e heat necessary for t h e operations of drying and carbonizing t h e lignite. T h e t a r yield is also low. The plant of t h e Board is situated in southern Saskatchewan, remote from any large center of industry, Under these conditions it does not appear probable t h a t , i n t h e beginning of t h e industry, a t least, t h e possible profits t o be made from t h e full recovery of by-products will justify either t h e capital expenditure necessary for a by-product recovery plant, or t h e depreciation of t h e carbonized residue by any attempt t o increase t h e by-products. It is fully recognized, however, t h a t a t a later d a t e with a larger and wellestablished industry this policy may require revision. None of t h e results obtained give any indication t h a t t h e use of vacuum, pressure,steam, or other modified method of carbonization would have any economic advantage. Finally, it has been found t h a t Souris lignite does not soften or become sticky a t any stage of its carbonization. This is in marked distinction t o t h e behavior of bituminous coal, and permits a design of carbonizer which is simpler and cheaper t h a n can be employed for t h e latter material. DESIGN O F CAKBONIZER
The design of carbonizer retort adapted t o fulfil t h e above conditions is briefly described below. T h e actual details of construction are unimportant for t h e purpose of this paper. It consists essentially of a strongly heated surface, or retort floor, inclined at’an angle slightly steeper t h a n t h e angle of repose of t h e crushed lignite. T h e material t o be treated flows down t h e heated surface from a hopper a t the top, passing under a succession of baffle plates, which control t h e thickness of t h e layer. The rate of flow of t h e material is controlled entirely by t h e rate of withdrawal from t h e bottom of t h e retort. This can be accomplished by any suitable mechanism. The retort is suitably enclosed a t t h e sides and top, and gas offtakes are provided i n t h e cover. The thickness of t h e layer is controlled by t h e difference between t h e slope of t h e retort and t h e angle of repose of t h e lignite, by t h e distance between successive baffles, and b y t h e clearance between t h e baffle and t h e retort floor. The material is repeatedly stirred by its passage under t h e baffles. T h e heated surface may be heated from below with gas. I t should be hottest a t t h e bottom of t h e retort and progressively cooler towards t h e top. T h e temperature of t h e lower p a r t of t h e heated surface may be as high as t h e materials of construction will permit. The regulation of t h e degree of carbonization of t h e lignite is entirely controlled by t h e time of its passage through t h e retort, t h a t is, by t h e r a t e of withdrawal from t h e bottom.
T’ol. 13, KO. I
S E M I C O M M E R C I A L CARBOPTIZER
Some experiments have been carried out with a very small model of t h e above design. I n this model t h e working surface varies from z in. t o 4 in. in width, is 4 f t . long, inclined a t a n angle of 4 j 0 , and is electrically heated. The bulk of t h e experiments, however, were carried out in a retort approximately 1 o . j in. wide and I O f t . long. T h e angle of inclination could be varied a t will, b u t 45‘ was found t o be satisfactory. Different materials were tried for t h e floor of t h e retort, b u t ultimately carborundum slabs were adopted. Twelve baffles were used i n t h e final arrangement; these were made of cast-iron and supported frpm t h e floor by means of end plates. The clearance under t h e baffles varied from 0.5 t o I in. The lignite was crushed t o pass 0 . 2 j-in. mesh. It was found advisable t o dry i t before treatment t o a moisture content of I j per cent or less. The capacity of t h e retort varied widely with t h e degree of carbonization produced, with t h e temperat u r e attained in t h e gas flue below t h e retort floor, and with t h e moisture in t h e lignite charge. I t may be rated roughly as equivalent t o zoo lbs. of raw lignite per hour., The results obtained, with regard t o o u t p u t , ease of control, and smoothness of operation, were regarded as sufficient t o warrant proceeding with t h e design and construction of commercial carbonizers on t h e same principle, for a plant capable of treating 2 0 0 tons of raw lignite per day, DISCUSSION
MR. R. DE I,. FRENCH: That I think is briefly what we have accomplished so far. While we do not believe that the work is at an end, yet it was successful enough in our minds to warrant us in going ahead with the construction of a plant on a commer-
cial scale. This plant is now under construction. We hope to have it in operation sometime, and when we do, we hope to
be able to say just what this process will cost in dollars and cents, and whether or not it is a commercially feasible thing t o carbonize Canadian lignite and to briquet the residue and sell it as a passing fair substitute for anthracite coal, which a week ago was selling for $ 2 2 . 6 0 a ton in the most easterly of the western cities, and at a higher price further west; I think at about $27 in Regiqa last week. Our raw coal will cost us about $41.80 at the mine. As we are in the middle of the field we should have no difficulty in getting plenty of coal at a low price. I might say that the lignite with which we are dealing is probably about as low grade a lignite as we have on this continent. It has the following analysis: RAWLIGNITE
..........................
Moisture, Ash ................................ Volatile matter.. .................... Fixed carbon
.............. ..............
Per cent 3I .8 5.2
28.9 34.1 4260
You can see it is a very wet lignite and hasn’t a particularly high calorific value. Practically all our work has been carried out on this lignite because we started with it and because %e wished t o compare our results we have endeavored to stick to it all the way through. PROF.E. P. SCHOCH(of the University of Texas, Austin, Texas, who presented the following resum6 of “A Process for the Economic Manufacture of Fuel from Texas Lignite”) : Lignites are characterized by a high water content, the property of “slacking” on exposure to air, and a high content of
Jan., 1 9 2 1
T H E J O C R N A L OF INDL’STRIAL A N D ENGINEERING C H E M I S T R Y
carbon dioxide (7 t o 8 per cent in Texas lignites). It is this 3 2 t o 40 per cent incombustible volatile matter which causes briquets made from raw lignite t o explode in the fire. Hence lignite must be retorted to render it fit for briquetting. The question arises: What is the most economic extent of retorting? For our experimental study of this question, the lignite used was obtained in the open market in Austin, but all of it was from the same mine. The lignite thus obtained was of rather mediocre quality. To our knowledge better lignite can be obtained even a t this mine and certainly in other localities, but what we used is representative of much of the lignite now sold in Texas; hence, the figures presented below may be considered t o be safe for all commercial lignites in Texas, b u t low for specially good lignites. I n our first set of experiments we retorted lots of I O lbs. each in powdered form with constant stirring and fractionated the gas evolved as the temperature was raised. These experiments revealed : ( I ) The fact t h a t the evolution of carbon dioxide ceases abruptly a t about 525 C. ( 2 ) T h a t t h e per cent by volume of carbon dioxide in t h e gas collected u p t o this temperature is from 23 t o 33 per cent. , (3) T h a t t h e other constituents of t h e gas evolved up t o 525 C. have high calorific powers, so t h a t t h e mixture has a calorific power of 410 B. t. u. (4) T h a t all t h e t a r is evolved with this gas. These results were obtained also with a different kind of a lignite from a totally different field. The gas fractions obtained a t temperatures higher t h a n 5 2 j o C. have heating powers of 410 B. t. u. per cu. f t . or less, and the total amount of gas obtainable by retorting a ton of this lignite is not more than 6500 cu. it. (the lignite from another region gave 6900 cu. f t ) , with a n average heating power of t h e whole gas of 410 B. t. u. This result is in marked contrast with the 10,000cu. f t . of 400 B. t u. reported heretofore. The coke left after complete retorting has a n ash content of 25 t o 28 or even 30 per cent and a heating power of 10,000 B. t u. or below. The relatively poor quality of this coke and the fact t h a t the gas obtained with i t would have t o be enriched t o make i t fit for “city use” led us t o consider the feasibility of retorting the lignite with a maximum temperature of 525’ C. It was evident t h a t by removing as much as possible of t h e large per cent (about 30 per cent) of carbon dioxide from t h e gas obtained up t o 5 2 5 ’ C., its heating power could be raised substantially, and a simple trial showed t h a t this could be done readily t o such a n extent a s t o make t h e gas directly fit for “city use.” To t r y out this whole procedure on a sufficiently large scale, we constructed a n apparatus which retorted I IOO lbs. of lignite per 24 hrs. and purified all the gas. Theretort was a 6-in. castiron pipe placed vertically and surrounded by a brick furnace 7 ft. high, with gas burners a t the bottom. The low temperature required made i t easy t o operate in such a manner as not t o injure the iron retort; its life is likely t o be great. The amount of gas obtained was 2 2 j o t o 2500 cu. f t . per ton of raw lignite with a heating power of 525 t o 540 B. t. u . ; the yield of coke was goo lbs. of r1,ooo B. t. u. (or more!), and the yield of d r y t a r was z per cent. T h e carbon dioxide was removed down t o 2 per cent by means of potassium and sodium carbonate solution. Calculation shows t h a t the amount of lignite needed as fuel for retorting is about 7 . 5 per cent of the lignite retorted. The coke comes out of the retort a t a temperature just high enough for briquetting, and not so high as t o take fire on exposure to air. The advantages of this procedure are: ( I ) A coke of the highest heating power obtainable. ( 2 ) A gas immediately usable in city mains. (3) The maximum amount of tar obtainable. (4) A cheap retort with large capacity, operating under mild conditions, and yielding t h e coke a t a temperature a t which i t can be easily and immediately handled for briquetting.
23
PROF. PARR:I would like t o ask Mr. French if he expects sufficient binder for his briquet t o come from the tars One of his numerical factors especially interests us H e says 7 per cent of heat is lost in the final accounting for the heat. If he finds it possible t o locate, with sufficient accuracy, those percentages of heat in the various constituents, and then say pretty accurately here is 7 per cent of heat unaccounted for, we would like t o know about it. It is one method of getting a t the exothermic quantity of heat. Seven per cent of 4000 cal would be somewhere within the range where we think the measurement of quantity of exothermic heat resides. T h a t factor, 7.6, is exceedingly interesting t o our work. MR. FRENCH:A remark of Prof. Schoch’s reminds me I should mention some things myself. We found exactly the same things in the beginning of our work t h a t he did. We never got 10,000 cu f t of gas or anything like it. I suggest t h a t some of those high figures may be due t o the method of carbonization, because I know of one case where a man was actually operating The a carbonizer so designed t h a t they fed moist coal t o i t moisture t h a t was driven off passed through the hot charge and what you got was a gas producer on a small scale. This person may have got IZ,OOO or 20,ooo cu. f t . of gas, but he was getting it a t the expense of his residue. I judge from Prof Schoch’s remarks t h a t he was primarily after gas. We were after residue, and i t appears t h a t with our own carbonizers we had just about enough t o operate the carbonizers, and not much more. M r . Stansfield ran a series of experiments in t h e small retorts under pressure, vacuum, and with a steam atmosphere, b u t none of these seemed t o show any advantage, and he went back t o practically atmospheric pressure. In answer t o Prof. Parr’s question on tars, we took the t a r and distilled i t a t 325’ C. On t h a t basis, we got what we called “available binder,” a quantity of pitch representing 2 . 5 t o 3 per cent of the carbonized residue, and t h a t is not sufficient. It is probably not a quarter of what is required. It takes a large quantity of binder t o make residue briquets, because physically the residue more nearly resembles charcoal than i t does coke. I imagine i t will be similar t o some coke which Prof. Parr has here. Answering Dr. Porter, the water is the water of constitution. It is dry coal. It is dried a t 105’ C., and t h a t is the water left after drying. Returning t o Prof. Parr, so far as loss of heat is concerned, I would prefer t h a t *Vr. Stansfield should answer t h a t question himself, because I do not know very much about his calculations, except t h a t I have a number of them, and I know t h e loss of heat always runs around t h e figures given. T H E COMMERCIAL REALIZATION OF T H E LOW-TEMPERATURE CARBONIZATION O F COAL By Harry A. Curtis INTERNATIONAL COALPRODUCTS CORPORATION, IRVISGTON, NEWJBRSBY
The’ p r o c e s s h e r e i n d e s c r i b e d w a s d e v e l o p e d f o r c o n v e r t i n g b i t u m i n o u s c o a l into a u n i f o r m , s m o k e l e s s f u e l r e s e m b l i n g a n t h r a c i t e in p r o p e r t i e s . It was r e c o g n i z e d at the outset t h a t the problem was one i n w h i c h small-scale t e s t s a l o n e would n o t yield the necessary data f o r p l a n t d e s i g n , and while m u c h v a l u a b l e i n f o r m a t i o n has b e e n s e c u r e d i n s m a l l a p p a r a t u s , t h e d e v e l o p m e n t of the p r o c e s s has b e e n very l a r g e l y t h r o u g h u s e of commercial-size u n i t s . F o r the past f o u r and a half years large-scale e x p e r i m e n t a l w o r k has b e e n c a r r i e d on in p a r a l l e l w i t h l a b o r a t o r y tests. The e x p e r i m e n t a l plant, as finally d e v e l o p e d , has a c a p a c i t y of a b o u t roo t o n s of r a w c o a l p e r d a y , b u t
