S . IT. P A R R
582
calories for a gram of protein, the same for a grain of carboh!dratcs, and 9.3 per grain of fats. I t \\ill be observed that arrowroot giles the widest nutrient ratio. I : 20, sho\ting how largely starch enters into its composition : it is also an expensive food, costing sixteen pence per pound. Fine grain farola is prepared as a substitute costing onl) sixpence per pound. \vith a nutrient ratio. I : 43, and ought to prove a nourishing food for ordinary invalids. L \I\
COT L I
LR5ITI
(-1
UriLtol
THE AVAILABLE HYDROGEN OF COAL. Us S. 11'. P A R R . I;. 19;.
Received Jaiiuary
The volatile constituents of coal being the subject of special attention on the part of the Illinois State Geological Survey the following study has I~ecnmade in connection with that work. 'ihe available hydrogen of coal is that part of the total hydrogen which :nay ciiter into combination with oxygen for the production of heat. The carbon, sulphur and available hydrogen constitute the actual combustible mattcr. IVhile methods for the determination of carbon and sulphur are direct and accurate, there is no direct method applicable for available hy~.l~O!pl.
The usual procedure is one of calculation, aiid it is indicated in Dulong's formula for deteriiiining the heat value from analytical data thus, calories=So8o C+34460(H-
0
--)+22.5oS. T h e accuracy o f t h e factor for
8
the available hydrogen, therefore, depends upon the accuracy attending the deteri:iination of oxygen. Cut the oxygen factor is subject to error. It i c deterinined by difference and includes, therefore, the sum of all the er:ors that may have attended the work on the other constituents. This 15 of considerable moment in the case of coal high in sulphur. If the sulphur is all in the pyritic iorm, and if we assuiiie the iron otily to be an ash constituent, the ash will be too high by reason of the oxidation of the iron, the added oxygen increasing the ash factor by $5 of the sulphur COIIstituent. This error falls upon the oxygen making the percentage of that eiement too low or too high, depending on the basis of calculation. Nor can this difficulty be easily obviated by correcting the ash, since coals are very frcquently inet with high in sulphur aiid practically free from iron. In the results published by Lord 8r Haasl on jO coals of the Pennsylvania and Oliio region the results from calorific determinations compared uvitli calculations by the Dulong formula ivere in such close agreement as I
Tra:is.
h i .
Itist. Eng., 1'01. 27.
HYDROGEN OF COAL
583
to warrant the inference that the available hydrogen was correctly represented by the total hydrogen minus of the oxygen. T h e same reasoning does not apply so well, however, to other regions. Take for example, the results on the first ten coals found in the last report from the Fuel Testing Plant of the United States Geological Survey1. N’ith one exception, these are all Illinois coals, and variations in calorific values between the calorimeter and calculated results show that on the average the results by the calorimeter are 70 calories higher than those calculated by the formula. This if coniputed to available hydrogen would indicate a calculated factor for that constituent averaging 0.2 per cent. too low. A still wider variation is found in a number of instances ; for example, Yorth Dakota No. g24is 0.73 per cent. too low ; U’est Virginia Xo. 18 is 0.80 per cent. too low, and four or five others are given with an indicated error of of one per cent. In these instances the variations cannot be ascribed to sulphur . I t is not the purpose of this paper primarily to offer a more accurate method of arriving at the factor for available hydrogen, but to suggest 3 different one which may also have the advantage of indicating other properties depending upon the structure and composition of coal. The fundamental conception is based on the idea that the original substance from which coal has resulted was made up of substantially the same sort of material, and that the products of decomposition represent a progression, which at any given stage in any case bear a constant likeness of composition to the same stage in another case. Many facts suggest this hypothesis which need not be enumerated here. W e have celluloie, brown and black lignites, bituminous coals of many varieties, semi-bituminous coals, semi-anthracites, anthracites, and graphites. How are they related ? If we take a number of coals of the bituminous type, and of widely differing characteristics as to their content of volatile matter, and reassemble their products of decomposition by destructive distillation into a sort of hypothetical molecule, we can arrange the series of equations which follow. These, of course, are intended as illustrative merely, though it should be said that they have their actual counterparts in respect to percentage composition?, TABLEI. VC
Bituminous Coal.
Carbon.
Hydrocarbons.
+
Water.
-c-x
100
E
x Iao
VC
3 HzO;20% 25% ( a ) CX4HI8O3 = C,, 4- C,Hl, T ZH f 3 HZO ;27.756 20% CPIZ ( b ) C18H1803 = Ci, 3 H,O ;-- 2 9 . 2 % 19.4% CeH14 (e) C X H X O O= ~ c16 ( d ) C&z@s = CII C7H16 3 HA0;39% 16.5% Bulletin 2 9 , U. S . Geol. Survey. 2 vc = carbon of the Lolatile matter. C = total carbon. H = available hydrogen.
++
+
+
s. \v.
5 84
PARK
There is little suggestion of a progression here until ive put them upon solile common basis for comparison. This has been dotie in the last tiycj columns. The basis in the first case is that of the ratio lietween the yolatile carbon, or that which is joined with hydrogen, as a yolatile hydrocarbon, (total carbon minus fixed carhoii) and the total carbon of the coal. The ratio is further multiplied by 100 to bring it to the i n r n i o i a \\-hole number and we have for the ratio the espressioii
YC
c
Io0.
The last column expresses the ratio between the hydrogen thus joined with the carbon, and the volatiie carbon. ]Ire therefore have the expression for this column of
H --
- s 100.
\.c Now,if we locate these points, taking the first ratio for the abscissas and the second for the ordinates, and draw the curve through the points as indicated, we shall have the following :
CARBON AND NYDROGEN RATIOS Fig. I.
The final adjustment and location of this curve has not been, however, so simple as might appear, but has involved a large amount of rather
585
HYDROGEN OF COAL
tedious calculation’.
After trial of almost every possible ratio, the one vc IOO was adopted as the fundamental represented by the expression --x C element in the case. Its concordant relationship to the other constituents justifies its adoption. This characteristic indeed suggested the scheme of classification of coals recently proposed by the writer?. It may be argued in its favor that of its constituent parts the denoiiiinator is a relatively large quantity and presumably accurate as to its determined value ; the numerator has whatever error is incident to the determination of the fixed carbon or volatile matter, but its effect is divided and practically eliminated by the character of the denominator just described. It has an especial advantage over any ratios based on the values for volatile and fixe4 constituents where both terms are susceptible of marked variation and error in their determination. The curve as finally decided upon is given in Fig. 2. A brief explanavc tion of the curve is as follows : For ratios of - x 100,which are less C than 20 no location of the curve should be attempted. First, for the reason that there are no coals having ratios which fall between 12 or 13 and 20, at least, they are so rare as to be negligible, and second, because below a ratio of 20, no sufficient uniformity exists to make the ratio applicable. It should be further noted that with the ratios in the higher numbers another factor enters into the account, vuiz., that of inert volatile matter, chiefly composed of water of composition, the main curve.“A” being applicable to coals of the true bituminous type having a content of from 12 per cent. to 16 per cent. of this material. This variation is easily cared for as follows. T h e available hydrogen is first calculated by means of the main curve A. The sum of the sulphur, available hydrogen and the carbon to which it is attached (VC) is subtracted from the volatile matterS, as determined by proximate analysis. This iemainder is the inert or non-combustible volatile matter. It should be reduced to the “ash and water-free” basis to avoid the abnormal effect (ash moisture)). If of high ash and moisture. (Dividing by 100the result shows a factor for this constituent between 16 and 20 per cent., it,indicates a black lignite, and the hydrogen reading should be made from the accessory curve “c.” Similarly, the use of the sub-curves “b” and “d” is indicated by a preliminary calculation roughly made, it being only necessary to ascertain the type of coal in hand. So far as present ex-
+
1 Special recognition of assistance in this connection is due Mr. F. F. Grout, formerly of the Illinois Geological Survey, and Mr. F. K. Ovitz, of the Engineering Experiment Station, University of Illinois. * This Journal, 28, 1425. Usually called ‘ I volatile combustible matter.” I t does not include water lost at 1 0 0 ~ .
s. w.
5 86
PARK
-
perience goes, tlie application of these subsidiary curl-es doc not eutcn:l to lower ratios than is indicated by the dotted lines at the k i t , c s c q ) t 111 the caFe of a fen- Michigan coals nitli over j per cent. snlphur. .\\ .lIL \EI,l; I1 YL)ltocrl:s cL71t\ I:-. 40
z:
O
m
cf 3J T V
rJ
=' 30
&
_1
0
>
L 0
25
+L:
g
20
[I
w
a v, T (5 ZI W
0 0 E IO
BROWrl.:rlT-
0
%
z
5 /O
/5
20
2,
I5
3
-.U
45
RflTIO OF V ' 3 L A T I L t CPRt3LN TO T O T A L C c 2 B G N ( I 1q
59
v_c_
c
2
The cnrve, therefore, is an empirical one, and is of value only as I; fits tlie case of other data not involved in its location. For \erification. therefore, a large number of results have been assembled mliicli give the ultiniate anal) sis as u ell as the calorific d u e s , and such factors tabulated as are necessary for illustrating the method. In tlie table below the volatile carbon of tlie second column is the total carbon minus the fixed carbon. vc T h e ratio of - of the third colurnii indicates t h e location 011 the curve C where may be read as ordinate the percentage ratio of the available liydrogeii to the volatile carbon. Ey multiplying, therefore, the volatile carbon by this percentage factor we have the per cent. of available hydrogen in the coal. This is the first colunin under "Hydrogen 17alues*' The second column of hydrogen values is derived from the ultimate anal! sis factors, ant1 corresponds to the expression from Dulong's formul; (H -
0
8
)
T h e last column i3 c h i \ ed from the indicated calories 11)
means of the formula
587
HYDROGEN OF COAL
+
Indicated calories- (8080 C 2250 S ) 34460 The first ten samples only have been taken as being sufficiently illustrative of the total of 54 analyses. H=
TABLE 11. Coniparison of Derived Values for Available Hydrogen.' . Total Carbon C
Description. I. 2.
6. 7. S. 9. 10. 11. 12.
13. 14. 15. 16. 17. IS. 19, 20.
21. 22.
23.
......72.16 ........69.24 ......61.13 ........62.01 .........54.06 ..........67.30 ........61.79 .........58.02 ....60.51 .......-62.20 .......62.97 ......69.85 ......71.49 ......6S. IS ........63.21 ......52.39 ..........61.80 ..........60.36 ..........60.62
Alabama, No. I.. Alabama, No. 2 Colorado, No. I . . Illinois, No. I Illinois, No. 2 Illinois, No. 3 Illinois, So. 4 . . Illinois, S o . 5 . Illitlois, No. 6 . . . . . . Indiana, No. I . Indiana, No. 2 . . Ind. Ter., No. I . . Ind. Ter., No. 2 . . Ind. Ter., No. 3.. Ind. Ter., N o . 4 Ind. Ter., No. 5.. Iowa, No. I . . Iowa, No. 2 . . Iowa, No. 3.. Iowa, S o . 4 . .
..........61.25
Average
Vol. Carbon
vc
1X.45 17.50 18.10 20.93 1S.82 15.14 17.49 17.35 13.05 19.53 20.21
19.80 21.70
20.36 19.31 15.34 15.29 18.62 21.63 20.03
Ratio vc.
Reading from
-
Curve.
25.56 25.27 29.60 33.73 32.96 22.50 28.3 29.9
21.1
21.5
25.2 17.6 17.3 19.0 18.0
21.6 16.9 16.8 17.0 24.4 19.0 1s.2
31.4 32.09 28.34 30.35 29.86 30.54 29.28 24.74 30.85 35.63 32.70
1S.a
'7.9 18.6 22.1
17.8 16.2 17.1
Hydrogen Values H H H calc. from from from Ut. Ind. Curve. Anal. Cals.
3.90 3.78 3.05 3.52 3.03 3.69 3.32 3.16 3.30 3.44 3.50 3.76 3.90 3.71 3.46
3.98 3.43 2.63 3.36 3.05 3.49 3.02 3.02 3.11 3.33 3.54 3.72 3.89 3.54 3.20 2.63 3.25 3.45 3.54 3.24 3.28
2.85
3.38 3.31 3.52 3.41 3.39
........
.................... ....................
3.92 3.77 3.02
3.64 2.92
3.62 3. I4 3.32 3.51 3.63 3.53 334 4.04 3.87 3.27 2.54 3.54 3.53 3.65 3.45
3.42
If we summarize the results on these 54 coals of the Fuel Testing Plant at St. Louis, we shall have a comparison of factors as follows, counting in units of l/lo of a per cent. above or below the standard which is taken as the hydrogen from indicated calories. TABLE111. Available Hydrogen : Variations in units of
1/10 of one per cent. Available Hydrogen From Ultimate from Curve. Analysis.
................. ............... ................ ............. .........
Total plus error.. Total minus error.. Highest plus error Lowest minus error.. Average of all variations..
11
7
84 3 7
12s
I .S
1.5
7 2.5
Since the above tabulations were made, analyses have been published Calculations based on data contained in Bulletin No. 261, United States Geological Survey, 1904.
588
S. W . PARR
from the Fuel Testing Plant of the Government Geological Survey', covering 70 additional coals fairly n-ell distributed throughout the TJiiiteJ States. The first 20 of these results are tabulated below sho\viiig the available hydrogen as derived by these three methods. Follo\i ing the table is a summary showing variations abovc ant1 belo\\ the staiidard which is taken as the available hydrogen calculatetl from the indicated calories as in Table S o . 3. 'r.InLI1: IV. Comparison of Derived Values for Available Hydrogen.' Description.
I.
2. 3. 4. 5. 6. 7. S. 9. IO.
11. 12.
13.
14. 15. 16. 17. IS.
'9. 20.
Yol. Carbon
'Total Carbon C
VC
. .. .. .. ... .. . . .... . . .. . . ... ... . . ... . . .. . .. .. . .. . ... .. ... . .
California, S o . I . . . . .47.j4 Illinois, S o . 6 . . . . . 56.94 Illinois, KO. 7 B... . .54.56 .5859 Illinois, S o . 7 D . . Illinois, No. S . . . . 57.25 Illiriols, S 0 . 9 A. .58.69 Illinois, 90.9 B.. . . .58.95 Illinois, S o . I O . . . .63.S3 .66.55 Illinois, S o . I I A . -64.29 Illinois, So. I I € 3 . .65.14 Illinois, So. I I C -62.52 Illinois, KO. 1 2 . . Illinois, S o . 13.. .. ..6j.S3 Illinois, No. 14... . . .5S.74 Illinois, No. I j . . . . .59.64 Illinois, S o . 16.. . .. ..67.33 .61.29 Illinois, S o . I S . . . Illinois, S o . 19 A . . ..62.76 .66.74 Illinois, KO.19 B. Indiana, KO. 3 . . . . . . .54.52
16.67 13.04 1S.31 1S.Q 17.31 1S.66 19.23 16.75 16.36 16.06 16.51 15.93 16.27 17.97 17.58 15.44 19.49 1j.26 16.6s 15.25
Ratio vc C
Reading from Curve
35,~ 22.9 33.5
14.7 23.9 16.7
Hydrogen Values. II H from from Ut. Ind. Anal. Cals.
H
calc. from Curre
2.45 3.12 3.06
32.2
17.1
j.22
30.2 31,s 32.6 26.20 24.60 24.90 24.7 25.5 24.7 30.60 29.50 22.90 3r.So
1S.00 17.3 17.0 20.50
3.12 3.23 5.27 3.43 3.62 3.49 j.70 3.38 3.58 3.22 3.23 3.69 3.37 3.39 3.62 2.94
21.1
25.0
28.0
22.10
21.70 22.0
21.3 22.00
17.90 1S.40 23.90 17.30 25.6 21.7 '9.3
_
Ayerage.
. . ... . . .... . . .. . . . .
.
Available Hydrogen from Curve.
Total plus errors . . . . . . . . . . . . Total minus errors . . . . . . . . . . . . . Highest plus error.. . . . . . . . . . . . . Lowest minus error. . . . . . . Average of all errors per sample I
.
.
. . .. . . .
26 69 5.5 5.5 1.4
2.42 2.9s j.07 3.37
2.71
3.21
3.22 j.2j
3,39
3.44
3.23
3.49
3.34 3.36 3.40 3.23 3.24 j.15 3.13 3.37 3.47 2.79 3.26 3.03
3.54
_
.
3.173
3.306
TABLEV. Summary of Comparative Factors Sho-sing Variations in Units of
..
.
2.49 2.91 2.Sj 3.1s
1/10
3.63 3.60 3.43 3.36 3.29 3.43 3.43 3.75 2.82 3.14 3.07 -
3.293
per cent.
From Hydrogen Analysis.
4 142
1.3
9.5 2. I
U. S. Geol. Survey, Bulletin ?io. 290. * Calculations based on data contained in Bulletin No. 290, Cnited States Geological Survey, 1905.
589
ELUTRIATION O F FINE POWDERS
A further comparison may be made by tabulating the averages from the various data made use of in the above computations. TABLEVI.
H
H
3.39
3.28
3.42
3.5'
3.46
3.69
3.96 4.09 3.30
3.95 4.06 3. I 7
3.90 4.16 3.29
3.54
3.41
3.61
H F r o m Curve
......
Average of first 20 (Table No. 2) Average of 50 from Bulletin No. 261, TJ. S. Geol. Survey Sverage of 50 from report of Lord and Haas.. Average of 12 from Mich. Geol. Survey Average of first 20, as in Table No. 4 Average of 7 0 from Bulletin No. 290. U. S. Geol. Survey.
................ ........................... .. ...............
From Ult. Anal. From Ind. Calories.
It will be seen from the above that out of 180coals the extreme of error by use of the curve is much within that resulting from the ultimate analysis; and that on the average the error represented by the former is less than that resulting from use of Dulong's formula. There is here opened up the possibility of making the essential factors as derived by ultimate analysis equally accessible with those of the proximate method, the real value of chemical work on coal being greatly enhanced thereby. Further interest in this connection would center in the derivation of the factor for tota1,and fixed carbon, a discussion of which must be reserved for another time. UXIVERSITY O F ILLINOIS, C R B A N A , ILL.
[CONTRIBUTION F R O M THE O F F I C E O F PUBLIC ROADS, U. S. DEPARTMENT O F AGRICULTURE'.]
AIR ELUTRIATION OF FINE POWDERS. BY ALLERTON S. CUSHMANAND PREVOST HUBBARD. Received February 4, 1907.
In recent years the production of extremely fine ground material for use in the various arts has made it highly desirable to obtain a standard method for separating and measuring particles which will readily pass through the finest meshed sieves. All the methods heretofore used depend upon the elutriation of the material with water flowing at different velmities, each velocity allowing the sedimentation of particles within the range of certain definite dimensions. Various forms of apparatus have been constructed on these principles, among which the best known are those of Schoene2 and Hilgard3. Recently an improved form has been described by Binns4. I n all the methods of water elutriation, however, it Published by permission of the Secretary of Agriculture. Bulletin d e la societe iniperiale des naturalistes d e Moscou (1867), Part Bulletin 38, Bureau of Chemistry, Department of Agriculture, 60. 'Trans. Am. Ceram. SOC.,8, 244 (1906).
l
2
I,
324.
