T h e over-all n e t effect of t h e errors examined in t h e preceding paragraphs i s believed to b e small a n d on t h e conservative side.
CONCLUSIONS T h e primary purpose of t h e present investigation, t h e definition of s a f e operating conditions for t h e air oxidation of curnene, w a s realized. By t h e u s e of the apparatus described, it w a s p o s s i b l e t o find t h e explosive limits of various cumene-air and cumeme-air-water mixtures a t elevated pressures. An i n c r e a s e in pressure c a u s e d a widening of t h e upper limits for both s y s t e m s studied. Water caused a d e c r e a s e in t h e upper explosive limit. T h e d a t a accumulated a r e of fundamental interest in t h e field of hydrocarbon flammability.
LITERATURE CITED (1) Armstrong, G. P . , Bewley. T . , Turck, K. H. W. (to Hercules Powder Co.). U.S. Patent 2,619,510 (Nov. 25, 1952). (2) Armstrong, G. P . , Bewley, T., Turck, K. H. W. (to Distillers Co., Ltd.), Brit. Patent 653,761 (May 23, 1951). (3) Ibid., 699,047 (Oct. 28, 1953). (4) Coward, H. F . , Jones, G. W., U.S. Bur. Mines, Bull. 503, 1-10 (1952). (5) Jones, G. W., U.S. Bur. Mines, Rept. Invest. 3787 (1944). (6) Kennedy, R. E., Spolan, I., Scott, G. S., Ibid., 4812 (1951). ( 7 ) O'Neal, Cleveland, Jr., Natl. Advisory Comm. Aeronaut. Tech. Note 3446, 28 (April 1955). (8) Zabetakis, M. G.,Scott, G. S., Jones, G. W., Ind. E n g . Chem. 43, 2120 (1951). Received for review September 4, 1956. Accepted March 15, 1957. Division 'of Petroleum Chemistry, 130th Meeting, ACS, Atlantic City, N. J . , September 1956.
Sorption of Water Vapor by Thermally Treated Lignite at Dif f e rent Re I at ive H u m id i t ies WAYNE R. KUBE University of North Dakota, Grand Forks, N.D. T h e mechanism of t h e coalification process, by which wood products a r e transformed through t h e various coal ranks, h a s received attention from numerous investigators. Generally, it i s believed that t h e woody material a d v a n c e s in rank by a slow chemical process consisting in part of dehydration and decarboxylation, occurring over geological ages. O n e avenue of approach t o b a s i c coalification mechanism and an understanding of fundamental coal structure h a s been a study of t h e forms and occurrence of moisture associated with t h e coal substance. Gauger (2) recognized that water w a s recoverable from coal from five sources, including: 1. 2. 3. 4. 5.
Decomposition of organic molecules Surface-adsorbed water Capillary-condensed water Dissolved water Water of hydration of inorganic constituents of coal
Moisture included in t y p e s 2 , 3 , and 4 h a s been studied by water vapor sorption-desorption t e s t s a s applied particularly to lower rank fuels, which a r e considered t o h a v e some gellike properties. L a v i n e (IO) summarized previous work and presented d a t a on t h e dehydration-hydration of wood and natural lignite. Larian and o t h e r s ( 9 ) extended h e initial investigation by testing p e a t and brown coal, and Zromulgated t h e p o s s i b l e classification of North American fuels in t e r n s of pore s i z e a s determined by t h e sorption studies. Gordon, Lavine, and Harrington (3) studied the effect of temperature and pressure on sorption of water vapor by lignite. Sorption s t u d i e s of t h r e e b a s i c t y p e s of natural lignite-woody, earthy, and peaty-were reported
46
INDUSTRIAL AND ENGINEERING CHEMISTRY
by T a s k e r (15) of t h e Ontario (Canada) Research Foundation for multiple desorption-sorption c y c l e s wherein slightly different sorption characteristics were noted for e a c h type. Several investigators, among them Klein (6) and more recently T e r r e s (16), showed that thermal treatment above t h e temperature required t o initiate decarboxylation in addition t o removal of t h e normally considered moisture r e sulted in artificial coalification or an accelerated metamorphism which a d v a n c e s somewhat t h e rank of t h e s o l i d fuel treated. Kube (8) exposed s a m p l e s of North Dakota lignite which had been thermally treated at various temperatures t o 950" F. in an atmosphere saturated with water vapor at room temperature, and reported that differences in resorption of water vapor existed, depending upon t h e treating temperature. Interest in t h e fundamental properties of North Dakota lignite h a s continued a t a high level b e c a u s e of t h e large reserves (some 350 billion tons) of t h i s low rank fuel (1). T h e s e d e p o s i t s represent a major untapped power and chemical source in t h e United S t a t e s . T h e present report represents a @ion of t h i s continuing interest and e x t e n d s t h e fundamental water vapor sorptiondesorption t e s t s to thermally treated lignite representing lignites originally obtained from a wide a r e a covering t h e North Dakota deposit and i t s extension into Canada. T h e major objectives of t h i s investigation were to determine: 1. The influence of thermal treatment and type of lignite on sorption of water vapor by lignite. 2. The approximate increase in coalification caused by thermal treatment at temperatures sufficiently high to initiate decarboxylation of the colloidal lignitic substance.
VOL. 2, H 0 . I
Table 1.
Ultimate Analysis, Moisture Content, and Heating Value of Six Natural Lignitesa
Lignite
Bienfait
Zap
Beulah
Dickinson
Custer
Baukol-Noonan
Ultimate a n a l y s i s , 70 Hydrogen Carbon Nitrogen Oxygen sulfur Ash Moisture, 70 Gross heating value, B.t.u./'lb.
6.9 43.0 0.8 43.8 0.4 5.1 35.8 7220
6.3 40.3 0.6 44.0
6.5 39.9 0.6 43.1 1.2 8.7 35.8 6660
7.0 37.4 0.5 47.7 1.4 6.0 41.0 6370
7.1 39.3 0.7 48.0 0.4 4.5 40.2 6670
6.8 42.0 0.8 43.0 0.4 7.0 35.6 7100
1.1 7.7 35.6 6720
"(p. 11, Table 1, 12).
3. The influence of thermal treatment on heat requirements for vaporization of residual moisture. 4. The influence of type of atmosphere on water retention in lignite.
LIGNITES TESTED T h e source of lignite and treatment after mining affect i t s physical and chemical properties. Consequently, a detailed history is p r e s e n t e d for t h e s a m p l e s u s e d in t h e s e experiments. All lignites u s e d were originally obtained by strip mining in western North Dakota and southeastern Saskatchewan, Canada. T h e s e s e l e c t e d mines cover t h e major lignite producing areas in t h i s region, and, a s shown in F i g u r e 1, extend from Bienfait, Saskatchewan, i n t h e north t o Dickins o n i n Stark County, North Dakota, i n the south. Geologically speaking, t h e lignites are from the Fort Union formation, and are of P a l e o c e n e age. Ultimate analyses of t h e various natural l i g n i t e s as well a s moisture content are given in T a b l e I. Variations in composition are c a u s e d primarily by differences in moisture and a s h content. Moisture contents as determined by t h e xylene method ranged from 35.6% for the Baukol-Noonan lignite t o 41.0% for t h e Dickinson lignite. T h e s e moistures represent as received percentage; bed moisture would b e somewhat higher.
HEAT TREATMENT T h e stoker-sized lignite, a s received, w a s stored in s e a l e d drums until thermally treated in t h e a b s e n c e of air in a rotary metal drum retort of approximately 10-pound lignite capacity at t h e Bureau of Mines Lignite Experimental Station (12). Nominal treating temperature w a s 700°F., as measured near t h e center of t h e retort. Heating w a s continued until moisture and g a s evolution were considered negligible. T h e processing w a s not t h e same as normal drying, in so far as moisture, as determined by t h e xylene method, remained in t h e solid residue until a treatment temperature of about 4 0 0 ° F . w a s reached. At 700'F. decarboxylation of t h e lignitic s u b s t a n c e w a s appreciable; from 3 9 to 45 cu. feet of g a s , chiefly carbon dioxide, w a s recovered per 100 pounds of natural lignite. Ultimate a n a l y s e s of t h e solid residue from t h e thermal treatment a r e given in T a b l e 11. Differences in carbon and oxygen contents of t h e s e r e s i d u e s may b e attributed to variations in a s h content. On an ash-free b a s i s , maximum differences in carbon concentration were less than 2 perc e n t a g e units. Oxygen-carbon ratios were from 0.16 to 0.19, further indicating uniformity of t h e ultimate composition of t h e carbonized residue.
L a v i n e ' s i n that t h e c y c l e of hydration-dehydration w a s followed, rather than t h e dehydration-hydration. T h e b a s i c c y c l e w a s repeated three times, t o determine if t h e charact e r i s t i c h y s t e r e s i s curve for the gellike natural lignite was retained in t h e thermally treated lignite.
SAMPLE PREPARATION Representative s a m p l e s were obtained by splitting t h e char (residue) from t h e treatment in t h e rotary-drum retort. T h e s e s a m p l e s were crushed t o 1/8 inch with minimum handling and stored in waxed, screw-cap s e a l e d b o t t l e s until prepared for t h e sorption t e s t s . T h e 1/8-inch thermally treated lignite w a s removed from t h e s e a l e d j a r s and pulverized in a Braun pulverizer, using three p a s s e s with progressive reduction in p l a t e spacing between p a s s e s . T h e 80- to 100-mesh fraction w a s s c r e e n e d out between p a s s e s and p l a c e d i n s e a l e d glass sample jars. In all s t e p s , care w a s taken t o e n s u r e minimum handling time, in order to limit exposure t o t h e atmosphere. Pulverization in the disk m i l l w a s advantageous over pulverization by hand in a mortar as done by L a v i n e (IO), in that less time w a s required (reduced exposure time) and more 80- to 100-mesh fraction could b e obtained from the same initial weight of material. Most of t h e fines produced during pulverization by hand were unsuitable, being less than 100 mesh.
-
-
HYDRATION AND DEHYDRATION PROCEDURE T h e procedure u s e d for t h e present t e s t s is as follows: Approximately 1 gram of t h e thermally treated lignite w a s placed in a stoppered weighing bottle, and weighed on an analytical b a l a n c e t o 0.1 mg. T e n samples of e a c h of t h e Zap, Beulah, Dickinson, Custer, s i x lignites-Bienfait, and Baukol-Noonan-were t h u s prepared and placed over concentrated sulfuric acid in a desiccator for 6 to 8 w e e k s at 104'F. (4OOC.). T h e weight after t h i s stabilization period w a s t a k e n a s t h e b a s e or standard weight for future
EXPERIMENTAL PROCEDURE T h e experimental procedure is modified from t h e s t a t i c method described by L a v i n e (IO), which is believed t o b e better adapted to t h i s work than t h e dynamic method u s e d by P e a r s o n (14). T h e present investigation differed from 1957
CHEMICAL AND ENGINEERING DATA SERIES
47
Table II. Ultimate Analysis and Heating Value of Solid Residue from Six Lignites Processed at 700OF.a
Lignite
Bienfait
Zap
Beulah
Dickinson
Custer
Baukol-Noonan
Ultimate analysis, % Hydrogen Carbon Nitrogen Oxygen Sulfur Ash Gross heating value, B.t.u./lb. a(Appendix 1, Table 12, 12).
4.0 71.8 1.4 13.4 0.6 8.8 12,470
3.6 67.7 1.1 11.1 1.6 14.9 11,590
3.5 67.3 1.1 11.1 1.7 15.3 11,500
3.7 68.4 1.0 11.0 1.7 14.2 11,740
4.0 72.3 1.3 13.5 0.5 8.4 12,370
3.9 70.1 1.4 12.2 0.7 11.7 12,120
tests. E a c h lignite w a s then placed in a desiccator over a sulfuric acid solution of measured concentration having a known vapor pressure. Concentration of solutions ranged from 0 (distilled water, relative humidity 100%) to about 69% by weight of sulfuric acid (relative humidity about 6.3%). P l o t s were made of t h e specific gravity of sulfuric acid vs. concentration with solution temperature as parameter, and of concentration vs. water vapor pressure (5)to facilitate determination of acid concentration. Solutions of precalculated compositioli were prepared and t h e concentration w a s checked with a calibrated hydrometer. T h e samples were kept in t h e desiccator a t 104'F. for 6 to 8 weeks, and reweighed. T h e weight gain over t h e b a s e weight (corrected) w a s taken a s t h e moisture gain. Moisture contents were then calculated on the b a s i s of the "dry" or standard weight. T h e samples were again p l a c e d over concentrated sulfuric acid (specific gravity 1.84) for .6 to 8 w e e k s for dehydration and t h e c y c l e w a s repeated for a total of three complete cycles.
TEST EQUIPMENT A constant temperature cabinet approximately 4 x 4 x 8 feet (inside dimensions) w a s constructed from insulating board. Heating w a s accomplished with a 2000-watt electric heater. A Brown Model 077 electric temperature controller with a B a l c o r e s i s t a n c e bulb having a sensitivity of O.l°F.
too
eo
+ 2 Y
0 K Y
a
60
* t
0 I a I
W
40
z
I4 -I W
a
20
0 IO
20
30
40
GRAMS OF WATER PER 100 GRAMS DRY LIGNITE Figure 2.
48
Relative humidity-water content curves for lignites at 104OF. (4OOC.)
INDUSTRIAL AND ENGINEERING CHEMISTRY
w a s used to maintain t h e s e l e c t e d temperature of 104'F. T o prevent local temperature variation, air w a s circulated with a continuous-operating revolving fan. A traverse of t h e area in which s a m p l e s were exposed with a movable thermometer showed that a uniform temperature w a s maintained. Samples in weighing bottles were placed, one for e a c h type of lignite tested, in 8-inch desiccators s e a l e d with special high temperature grease. All samples were weighed with a C hain-0-matic analytical balance.
DISCUSSION OF RESULTS Influence of T h e r m a l Treatment and T y p e of L i g n i t e on Water-Vapor Sorption. Of major interest in the present t e s t
series is t h e effect of thermal treatment on the ability of lignite t o regain moisture from atmospheres of various relative humidities. Sorption data for t h e six different thermally treated lignites would allow comparison between lignites throughout t h e lignite deposit area. Average data on moisture content for three sorption c y c l e s on lignite previously thermally treated at 700'F., a s determined in t h i s investigation, are compared in Figure 2 with moisture contents of freshly mined lignites (desorption and sorption) from L a v i n e (10) and steam-dried lignite from Gordon (3). T h e three thermally treated lignites from t h e northem part of t h e North Dakota-Southem Canada deposit (Bienfait, Baukol-Noonan, and Custer) exhibited nearly the same moisture retention, which w a s less at t h e same relative humidity than that of t h e three thermally treated lignites from the southern part (Zap, Beulah, and Dickinson). T h u s a distinction w a s made between nodhern and southern lignites, and separate average curves were plotted. However, t h e distinction is slight. T h e conclusion that t h e commercial lignites in t h e entire field are rather uniform is supported by t h e uniformity of moisture and ash-free analysis, and of tar and g a s yields upon thermal treatment (12). A typical, somewhat S-shaped sorption curve w a s obtained when the water content of t h e thermally treated lignite w a s plotted a s a function of the relative humidity Differences between t h e moisture contents (Figure 2). of t h e thermally treated lignites at the same relative humidity were less than t h o s e between t h e thermally treated and freshly mined or steam-dried lignite, as indicated by the curves in Figure 2. T h e slope of t h e sorption curves for the thermally treated lignites w a s greater than that for freshly mined or steamdried lignite, indicating that the vapor pressure increases more rapidly with water content than for the steam-dried or freshly mined lignite. Consequently, less moisture is present in thermally dried lignite. T h e effect of higher temperature level during initial dehydration is to d e c r e a s e the ability of the lignite m a s s t o assimilate water vapor. T e s t s a t near 10% humidity have pointed out that the moisture regain d e c r e a s e d with level of treatment temperature, at l e a s t until a level of 950'F. ( 8 ) . It is believed that t h i s d e c r e a s e in ability to regain VOL. 2,
NO. I
moisture i s d u e not entirely t o t h e p r e s e n c e of permanent 01 noncondensable g a s e s i n t h e c a p i l l a r i e s formerly filled with water, which prevents t h e resorption of moisture, but mainly t o c h a n g e s in t h e lignite s u b s t a n c e induced by t h e thermal treatment. T h i s i s evidenced by t h e fact t h a t more moisture is retained by t h e steam-dried lignite, which w a s treated a t approximately 390°F. (15 atm.), and s t i l l more by t h e lignite dehydrated a t 104'F. (sorption curve for freshly mined lignite). Experiments a t t h e Ontario Z e s e a r c h Foundation (15) showed that s u c c e s s i v e dehydrations a t room temperature lowered t h e ability of t h e lignite m a s s t o regain moisture. Further e v i d e n c e i n t h e present c a s e is t h a t drying a t increasing temperatures l e s s e n s the width of t h e h y s t e r e s i s loop, which is s t i l l present a t 1 0 4 ° F . but not noticeable for the steam-dried lignite or t h e lignite thermally treated a t 700" F. T h e lignites thermally treated a t 700°F. were taken through three complete sorptiondesorption c y c l e s without h y s t e r e s i s in moisture retention. At a relative humidity of 50%, which i s approximately t h e average humidity occurring in t h e s t a t e of North Dakota, a moisture content of 8 t o 94, i s indicated in F i g u r e 2 for t h e lignite dried a t 7 0 0 ° F . and about 13 t o 17% for t h e lignite dehydrated over sulfuric a c i d a t room temperature. T h i s effect of higher drying temperature is interesting b e c a u s e it h a s been assumed s i n c e t h e work by L a v i n e (IO), that drying of lignite t o less t h a n 15% moisture w a s not practical, a s t h e hygroscopic dried lignite would gain moisture to about 15% during transit or storage. Consequently, a lower moisture content may b e practical, if a relatively high temperature i s reached during initial dehydration. E f f e c t of T h e r m a l T r e a t m e n t on C o a l i f i c a t i o n . From t h e vapor-pressure maintained over the sulfuric acid s o l u t i o n s and t h e apparent equilibrium moisture content of t h e lignite, t h e pore radii of t h e c a p i l l a r i e s in t h e lignite mass may b e estimated by t h e Thompson equation and t h e pore sizemoisture retention relationship used t o i n d i c a t e t h e approximate rank of t h e c o a l ( 9 ) . Pore s i z e vs. moisture retention for peat, brown coal, freshly mined lignite, and t h e lignite thermally treated a t 7 0 0 ° F . i s compared in F i g u r e 3. Data for peat, brown c o a l , a n d freshly mined lignite were taken from L a v i n e (IO), and d a t a for the thermally treated lignite from the average res u l t s of t h e s i x l i g n i t e s t e s t e d in t h e present investigation. Taking 50% of t h e moisture retained for a comparison standard a s done by Lavine, i t i s found t h a t t h i s amount of bound water i s held in c a p i l l a r i e s less than 7.4 mp i n peat (Minnesota), 3.7 mp i n freshly mined North Dakota lignite, 2.5 mp in thermally treated lignite, and 1.8 mp i n German brown coa!. Study of t h e curve for t h e freshly mined and thermally created l i g n i t e i n d i c a t e s that thermal treatment results in appreciable destruction of larger c a p i l l a r i e s and to a l e s s e r extent h a s diminished t h e number of smaller capillaries-i.e., most of t h e bound water i s held in capillaries of intermediate s i z e a s compared to that of t h e freshly mined lignite. T h e over-all effect h a s been t o dec r e a s e t h e apparent average s i z e of t h e c a p i l l a r i e s available for holding moisture. B e c a u s e of t h e relationship in descending pore s i z e a t SO% moisture retained (peat, lignite, brown coal), i t had b e e n proposed that pore s i z e a s determined from sorption s t u d i e s b e used for c l a s s i f i c a t i o n of North American solid f u e l s ( 9 ) . However, s u c h a b a s i s is apparently not s u i t a b l e for all fuel ranks. Other investigators ( 1 6 ) have described t h e effects of thermal treatment a t relatively high temperat u r e s a s artificial coalification. On t h e b a s i s of smaller average pore s i z e , the thermally treated lignite may b e considered t o b e metamorphosed t o slightly higher rank than t h e fresh lignite, but s t i l l not so high a s t h e brown coal reported by Lavine. However, consideration of t h e ratios of 1957
oxygen to carbon and net h e a t i n g value t o gross heating value for t h e thermally treated lignite i n d i c a t e s that a rank between subbituminous and bituminous h a s been a c h i e v e d (11), which is a greater extent of coalification than indicated from pore size s t u d i e s . T h e rank of brown coal in c l a s s i f i c a t i o n by pore s i z e i s not logical when compared to more standard c l a s s i f i c a t i o n , a s t h e rank i s indicated higher than t h e thermally treated lignite. T h e reason for t h i s d i s crepancy is not known, but t h e brown coal may not h a v e been a typical freshly mined coal. I t s detailed history is not available. H e a t of V a p o r i z a t i o n of M o i s t u r e from T h e r m a l l y T r e a t e d L i g n i t e . Latent h e a t of vaporization of moisture from t h e
thermally treated lignite may b e estimated from sorption d a t a a t 68°F. (20OC.) and 1 0 4 ° F . (4OoC.), u s i n g t h e Clapeyron-Clausius equation a s w a s done by L a v i n e for t h e freshly mined lignite (IO). Assumptions u s e d for t h i s calculation a r e that water vapor o b e y s t h e ideal g a s l a w s ; t h a t latent h e a t of vaporization i s constant over t h e temperature range; and that t h e h e a t of vaporization i s independent of lignite type. T h e s e assumptions a r e apparently valid for freshly mined lignite. A s in t h e c a s e of t h e freshly mined lignite, the heat of vaporization w a s greater a t lower moisture c o n t e n t s for t h e thermally treated lignite, but the numerical value for h e a t s of vaporization a t t h e s a m e moisture contents were much higher for t h e thermally treated lignite. H e a t s of vaporization ranged from 1470 B.t.u. per pound a t a moisture content of 2 0 grams water per 100 grams of dry lignite t o 1900 B.t.u. per pound a t 10 grams of water per 100 grams of dry lignite, with an average of 1660 B.t.u. per pound. Lavine's d a t a ( I O ) showed t h e heat of vaporization over t h e s a m e moisture content t o b e 1080 t o 1160 B.t.u. per pound, with a n a v e r a g e of 1120 for t h e freshly mined lignite. T h e same c a l c u l a t i o n s for water vaporizing from a plane surface gave a latent heat of vaporization of 1050 B.t.u. per pound. T h e large i n c r e a s e i n h e a t required t o vaporize moisture f r o m the thermally treated lignite, amounting t o on t h e avera g e about 80% more than t h a t required for vaporization from a plane surface, indicates that t h e moisture i s tenaciously held, and that t h e mechanism of retention may b e changed. T h e i n c r e a s e i s greater than that expected from consideration of departure from a plane vaporization surface. It should b e reasonable t o a s s u m e that t h e nature of t h e water meIO0
u)
w
a
80
2
2 4
5
60
W O
zIw
a
40
w
a
3
I-
I I
20
tW
0
W I
n 0
I
1
I
I
I
I
4
8
I2
16
20
CAPILLARY
RAOIUS
IN IO-? C Y .
Figure 3. Moisture-pore size relationship a t 104'F.
(40OC.)
CHEMICAL AND ENGINEERING DATA SERIES
49
n i s c u s for both thermally treated and freshly mined lignite is similar for t h e s a m e moisture contents. However, t h e increase i n average latent heat requirements for vaporization i n t h e freshly mined lignite is only 7% greater than that from a plane surface. T h u s , t h e moisture in t h e thermally treated lignite may be considered more tightly retained than i n t h e natural lignite. P e r h a p s t h e more weakly held water in t h e natural lignite is a s s o c i a t e d with t h e gel structure of the lignite, which is evidently destroyed during thermal treatment a t relatively high temperatures, a s indicated by loss of h y s t e r e s i s in t h e sorption curves. Water retained b y t h e gellike structure should follow closely t h e u s u a l sorption l a w s for gels, and t h e h e a t of vaporization should not differ considerably from that calculated from a plane surface. T h i s c h e c k s well for t h e freshly mined lignite, a s t h e latent h e a t of vaporization from t h e freshly mined lignite is within 7% of that from a plane surface, but d o e s not hold for t h e thermally treated lignite. Kreulen (7) in recent work distinguishes between a strongly assimilated moisture a s s o c i a t e d with t h e inner surface of lignite and weakly assimilated water a s s o c i a t e d with t h e gellike structure of t h e lignite. T h e strongly a s similated moisture d i d not follow t h e usual sorption l a w s nor did t h e calculated h e a t of vaporization correspond t o t h a t usually obtained for water, similar to t h e behavior noted in t h i s investigation. Influence
of T y p e of Atmosphere on Water Retention.
Dehydration over pure sulfuric a c i d for 6 t o 8 w e e k s d o e s not remove a l l moisture. T h e quantity of moisture remaining after t h e establishment of a n “apparent” or “practical” equilibrium weight is problematical, and only a relative value for t h e moisture content of t h e lignite may b e obtained. As there is disagreement over k h a t is actually measured a s “moisture” ( 4 ) , e s p e c i a l l y with low rank fuels, t h e present method is believed adequate for t h e purposes of t h i s in5.6
s3 46
4.4
4.0
WEIGHT LOSS WEIGHT GAIN
+z
HELIUM
3.6
NITROGEN
I
+
I
OXYQEN
I-
=
0
:
1.6
I3
A
ACKNOWLEDGMENT
0 4
e
12
16
zo
24
2e
32
EXPOSURE, DAYS Figure 4. Weight change of thermally treated lignite dependent upan atmosphere at 104OF. (4OOC.) TINE
50
vestigation, which is e s s e n t i a l l y comparison of water retention a t definite humidities. In addition, similarity of methods allowed comparison with the work of previous investigators a t t h e University of North Dakota. In preliminary t e s t s , t h e thermally treated lignite w a s found t o increase i n weight over pure sulfuric acid after a n initial loss. T h e initial loss w a s due t o removal of moisture unavoidably gained during handling, storage, and pulverization of the samples. T h e weight increase, about 0.4% of the b a s e weight, w a s constant after two 6- to 8-week periods i n desiccators, allowing prediction of weight gain for subsequent c y c l e s over pure acid. Consequently, a correction factor could be calculated for e a c h lignite to b e subtracted from t h e g r o s s weight after e a c h 6- t o 8-week t e s t period t o give t h e “corrected” b a s e weight. Repeated t e s t s indicated that t h e correction was valid, in t h a t weight gains a t relatively low humidities could be predicted and reproduced. Without t h e correction, a standard weight could not be maintained. At high humidities, t h e correction factor w a s not a s important, a s it w a s a small fraction of t h e total weight gain. B e c a u s e of t h e thermal treatment in t h e rotary drum retort, t h e lignite had become activated sufficiently t o s o r b oxygen under t h e t e s t conditions. Previous work with natural lignite dehydrated over sulfuric acid h a s shown that no sorption of other than water vapor occurred even after standing for periods up t o 3 y e a r s (15). To c h e c k on oxygen sorption, a s e r i e s of t e s t s w a s performed with e s s e n t i a l l y a dry oxygen atmosphere over t h e lignite samples. Average results of a duplicate s e r i e s on “dry” Custer lignite, run at 104’F., a r e presented in Figure 4 . In t h e oxygen atmosphere over pure sulfuric acid, the weight of s a m p l e increased uniformly after a stabilization period of 15 days, with no apparent d e c l i n e i n rate of increase after 17 additional d a y s ’ exposure. T h u s , it i s believed that oxygen retention by t h e thermally treated lignite i s responsible for t h e weight i n c r e a s e noted over concentrated sulfuric acid i n air-filled desiccators. T e s t s in different inert g a s atmospheres showed t h a t varying percentages of moisture may be removed, depending upon t h e atmosphere. R e s u l t s of a duplicate s e r i e s of t e s t s with dry C u s t e r lignite over concentrated sulfuric a c i d in atmospheres of helium, nitrogen, and argon a r e compared in Figure 4. A s expected, a net weight l o s s occurred in t h e inert g a s atmospheres. Weight l o s s in t h e inert g a s atmospheres w a s inversely dependent upon t h e molecular weight of t h e gas, t h e greatest l o s s occurring in t h e first day. Moisture l o s s in t h e helium atmosphere w a s much greater than in either the nitrogen or argon atmosphere, being about three times that in nitrogen and f i v e t i m e s t h a t in argon. T h e increased l o s s in moisture in helium i s greater than expected and may b e related t o modification in sorption centers or retaining properties of t h e water in t h e capillaries. Similar results were observed i n thermal drying of natural lignite in inert g a s streams (4). In nitrogen or argon atmospheres, little difference in weight w a s noted after t h e initial exposure period. Actually a slight gain occurred after about 20-day exposure, d u e perhaps t o exposure t o air during handling. T h e 6- t o 8week exposure in t h e a c t u a l t e s t s e r i e s would b e sufficiently protracted t o e n s u r e a c l o s e approach t o equilibrium conditions.
INDUSTRIAL AND ENGINEERING CHEMISTRY
Lignite s a m p l e s were furnished by t h e Bureau of Mines, Department of Interior, L i g n i t e Experiment Station, Grand Forks, N.D., courtesy of W . H. Oppelt, supervisor, Utilization Section.
VOL. 2, NO. I
LITERATURE CITED
(1)
Brant, R., U. S. Geol. Survey, Circ. 226 (1953). (2) Gauger, A. W., Am. Inst. Mining Met. Engrs. Trans. 101, 148 (1932). (3) Gordon, M., Lavine, I., Harrington, L., Znd. Eng. Chem. 24, 928 (1932). (4) Hoeppner, J., Fowkes, W., McMurtrie, R., Bur. Mines, Rept. Invest. 5215 (1956). (5) “International Critical T a b l e s , ” Vols. 1 and 2, McGraw-Hill, New York, 1933. (6) Klein, H., Braunkohle 29, 1 (1930). (7) Kreulen, D., Brennstoff-Chem. 36 (17/18), 266 (1955). (8) Kube, W., Proc. N. Dakota Acad. Sci. 6, 8 3 (1952). (9) Larian, M., Lavine, I., Mann, C., Gauger, A., Ind. Eng. Chem. 22, 1231 (1930).
Thermal Properties of Soybean
(10) Lavine, I., Gauger, W., Zbid., 22, 1226 (1930). (11) Oppelt, W., Proc. N. Dakota Acad. Sci., 6, 36 (1952). (12) Oppelt, W., Golob, E., Cooney, J., Kube, W., Bur. Mines, Mineral Research Invest., Prelim. Rept. 99 (1954). Kamps, ., T. W., (13) Oppelt, W. H., Kube, W. R., Chetvick, hi. €I Golob, E. F., Bur. Mines, Rept. I n v e s t 5164 (1955). (14) Pearson, R., master’s thesis (unpublished), University of North Dakota, 1950. (15) Tasker, C., Ontario Research Foundation, Div. of C h e m Research Lignite, No. 23 (1933). (16) Terres, E., Brennsloff-Chem 33, l(1952). Received for review August 29, 1956. Accept ?d February 7, 1957. Investigation supported by a grant from the National Science Foundation, Washington, D.C. (G. N. Hickox, program director for engineering sciences).
Oil M e a l
JOEL 0. HOUGEN’ Rensselaer Polytechnic Institute, Troy, N.Y.
A
nnual production of s o y b e a n s in the United S t a t e s exc e e d s 350,000,000 b u s h e l s a year (7), probably about o n e half (6) the world total. Within the United S t a t e s s o y b e a n s are grown widely in the Middle West. Cultivation of soyb e a n s is a major effort in t h e corn belt s t a t e s ; over 20,000,000 acres (5) are devoted to t h i s crop in that region, almost all harvested for the beans. Two major raw materials a r i s e from the soybean: oil and meal. Food products, principally margarine and shortening, utilize about 80% of t h e oil; the remainder is used in soap, paint and varnish, and m i s c e l l a n e o u s nonfood products. Soybean meal i s consumed for the most part a s feed for livestock (about go%), smaller amounts being used for soybean flour, plywood glue, and paper coatings, and in various industrial protein-containing materials. Soybean o i l meal i s defined a s the ground residue which remains after the oil is removed from the soybean, regardless of the p r o c e s s of extraction. L i v e s t o c k feed i s the primary market, e x c e p t for a small outlet for industrial uses. Accepted specifications for soybean oil meal obtained by the two standard methods of processing, expelling and solvent extraction, are (1): for meal from either source, carbohydrates and fiber 7 % (maximum), nitrogen-free extract 27% (minimum), and moisture 12.5 % (maximum). Minimum p e r c e n t a g e s of protein and fat differ, depending on the source of the meal. For hydraulic and expeller meals t h e s e are 41.0 and 3.5 and for extracted m e a l s 44.0 and 0.5, respectively. Following oil extraction the soybean meal is p r o c e s s e d for solvent removal and to improve i t s nutritional properties. T h i s usually involves raising the temperature of the meal by adding thermal energy through the walls of a container, by direct steam injection, or both. Data useful in the d e s i g n of apparatus used to p r o c e s s particulate materials are not overabundant. T h i s is particularly true for materials which are biological in origin and h e n c e susceptible to alteration upon exposure to air, moderate temperatures, or both. T y p i c a l of such material is soybean oil meal. ‘Present address, Research and Engineering Division, Monsanto Chemical Co., St Louis, Mo.
1957
T h e objective of t h i s work w a s to determine experimentally the thermal properties of a typical soybean oil meal, including thermal conductivity, thermal diffusivity, and h e a t capacity.
PROPERTIES OF MEAL, EXPERIMENTAL APPARATUS, AND PROCEDURE T a b l e I l i s t s some of the physical and chemical properties of the soybean oil meal u s e d in t h e s e studies. T h e meal w a s extracted in conventional H a n s a M e u h l e type extractors using Skellysolve B as the solvent. Following t h e oil extraction t h e solvent w a s removed by heat and by direct steam injection a t about 200 O F . Solvent removal w a s considered complete. R e s i d u a l o i l content w a s found to b e 0.7 to 0 . 8 % , a s shown in T a b l e I. Toble I. Properties o f Soybean Oil Meal after Solvent Removal Size
Oil content, 70 Moisture content, % Total protein, 70 Water-soluble protein, 7’’ Variety of bean
Will p a s s 48 mesh 0.7-0.8 13.2 51.55 (dry stock b a s i s ) 39.57 of total protein* (dry stock b a s i s ) Lincoln (probably)
eDetermined by tentative procedure recommended by Soy Flour Association, revision of Dec. 10, 1946.
T h i s meal may be further processed before being soid. It i s frequently packaged in metal c a n s prior to final treatment. Meal which w a s canned but had not received i t s final (heat) treatment was supplied by the A . E. Staley Manufacturing Co., Decatur, Ill. Thermal conductivity w a s obtained by using the concentric cylinder technique with a material of known conductivity as the reference. (Figure 1). Concentric copper tubes were arranged a s indicated, with provisions for producing an atmosphere of saturated steam within the innermost tube. T w o small iron-constantan thermocouples were located a t points about diametrically opposite e a c h other on e a c h tube a t the mid-point of the vertical sections. T h e reference material used w a s Grade 6 Spheron carbon black furnished by Godfrey L. Cabot, Inc., Boston, Mass. At a mean temperature of 118’F. and a bulk density of 20.2 CHEMICAL AND ENGINEERING D A T A SERIES
51
