January 18, 1918. THE COLLIERY GUARDIAN. 127 to the air on porous plates for two days before the first test. The following figures, giving the quantities of float and sink particles, were obtained at the con- clusion of the test occupying 20 minutes, and after immersion for 24 hours, and they clearly demonstrate that during the longer period some of the float par- ticles had been converted into sink particles. Raw Coal“ B ”— In. Per cent. Size .............. |to| ... 9’38 After 20 minutes— Per cent. Float ............ 91’88 ... 4’22 Sink.............. 8’12 ... 67’77 After 24 hours— Float ........... 84’33 ... 3’51 Sink.............. 15’17 ... 42’21 An examination of the value for the ash content of the various sized washed coals revealed an important and remarkable feature. In general it was observed that the ash content first diminished and then increased as the diameter of the particles increased. Hence there existed a particular diameter of washed coal which contained the minimum quantity of ash. Efficiency of Washing Process. Lincoln (Bulletin No. 69 Engineering Experimental Station, University of Illinois, 1913, page 57), has given the following illustration to show the method of calculating the washer efficiency. The estimation of the efficiency was based on the fact that a perfectly washed coal would contain no heavy or sink particles and that a perfect refuse would contain no float particles. Therefore it was suggested that the general efficiency of the washing process was to be regarded as the mean of the concentration of the float particles in the washed coal and the sink particles in the refuse. Washed screenings. General effici- ency of pro- - cess. Washed screenings plus refuse divided by 2. Per cent. Per ct. Effici- ency. Perct. Effici- ency. Perct. Refuse. Per ct. Float... — . ..91’29 ... — 1’2 .. Sink ... 8*71 .. — 98*79 — : 95’04 An examination of this method of calculating wash- ing efficiency, however, seemed to the author to be entirely unreliable and meaningless. In order to deduce a satisfactory method of calcu- lating the efficiency of the washing process attention must be directed to the actual practice. The object of washing the raw coal is to concentrate to the utmost the valuable ingredient (i.e., the float particles) so that the washed coal shall be a high quality fuel. Therefore it is impossible to determine for a washing plant the efficiency with which the quality of the material is improved. That is termed the 11 quali- tative efficiency ” of the process. Obviously, where the washed coal is perfectly clean, i.e., composed solely of float particles, the qualitative efficiency is 100 per cent. However, it is highly probable that the plant, in delivering a small quantity of pure coal, is rejecting as refuse quite a large proportion of the raw coal, in which case evidently a considerable loss of float particles in the refuse is taking place. Therefore a second conception has been reached, according to which, for perfect washing, the plant must have recovered quantitatively the whole of the float particles so that none escape as refuse. The effectiveness with which the float particles are recovered is termed the 11 quantitative efficiency ” of the plant. Suppose, for instance, that a raw coal composed of 95 per cent, float and 5 per cent, sink particles passes from the screens to the washer, and that during the process of washing the whole of the float particles are recovered and their concentration increased to 96 per cent., Lincoln’s method of calcu- lating the efficiency would be as follows: — Washed coal. Refuse. General effici- t------*--------(-----------------—ency of the — Efficiency. Efficiency. — process. Per cent. Per cent. Per cent. Per ct. Per cent. Float ... — ... 96 ‘ ... — ... — ... 96 Sink.... 4 ... — ... 100 ... — ... 2 It will be observed that the concentration of the float particles in the raw coal amounts to 95 per cent., and in the washed coal to 96 per cent. Therefore washing has increased the concentration by 1 per cent, of a possible 5 per cent. It is, however, a some- what abnormal estimate to assign an efficiency of 96 per cent, to that part of the process where only one- fifth of the task is accomplished. An efficiency of 20 per cent, would appear to be a more rational estimate. In regard to the quantitative working of the process, the refuse consisted of sink particles only and consequently all (100 per cent.) of the float particles were recovered by the plant. The problem was, to determine by what method the qualitative and quantatative efficiencies were to be combined so as to obtain a result which would represent the general efficiency of the process, the following being adopted: In the above case 100 per cent, of the material was recovered with the quality improved by 20 per cent. The author, therefore, considered that the general efficiency of the washing process was = 20 per cent. Two particular hypothetical cases will probably make clearer the suggested method for calculating the general efficiency of the washing process, and at the same time serve to show the error in Lincoln’s deduction. _ Case 1.—Suppose a washer receives 1001b. of raw coal containing 80 per cent, float and 20 per cent, sink particles and delivers the products given in the following table: — Raw coal. Washed coal. Refuse. Float.... 80% ... 90% (76’5lb.) ... 23’33% (3’5 lb.) Sink ....... 20% ... 10% (8’51b.) ... 76*67% (11*5lb.) Output... 100% ... 85% ... 15% Qualitative Efficiency. The concentration of the float particles in the raw coal is 80 per cent., and in the washed coal 90 per cent. Hence the concentration has been raised by 10 per cent, out of a possible 20 per cent. Therefore the qualitative efficiency x 100 = 50 per cent. Quantitative Efficiency.—In dealing with the raw coal the plant rejected 15 per cent, of the weight as refuse, of which 23-33 per cent, is composed of valu- able float particles. This means that float particles amounting to , or 3’5 per cent, of the total output have been lost. But of the raw coal the valuable float particles only amounted to 80 per cent.; there- fore the plant has recovered 76-5 parts from a possible 80 parts, or 95-63 per cent. General Efficiency.—The process has recovered 95.63 per cent, of the float particles with the quality improved by 50 per cent., representing 47-82 per cent. The general efficiency of the process according to Lincoln’s method would be =83-33 per cent. Case 2.—Another washer delivered the same pro- ducts working upon a raw coal of far inferior quality. In case 1, coal fed into the washer was composed of four parts, by weight, of float, with one part of sink particles. In case 2 the coal entering the washer would be considered to be two parts by weight of float with float with one part of sink particles. The result might be tabulated thus : — Raw coal. Washed coal. Refuse. Float 66’67% ... 90% (58 51b.) ... 23’33% (8*171b.) Sink ... 33’33% ... 10% (6’5lb.) ... 76’67% (26’83lb.) Output... 100 00% ... 65% 35% Although the qualities of the washed coal and refuse are identical with those considered in case 1 the quan- tities (given in brackets for 100 lb. of material entering the washer) are different. * The general efficiency of the washing, according to Lincoln, will be the same as before that was, 83-33 per cent, despite the fact that a far inferior coal has been so washed by the plant that the yield is of the same quality as in the previous case. Calculating the efficiencies according to the author’s suggestions gives: — Qualitative efficiency = x 100 = 70 per c. . 100 — 66’67 Quantitative efficiency = 66*67 — —x ^3 33 = 87’75 per cent. General efficiency of washing process = =■ 61’43 per cent. The two examples cited above clearly indicate the fallacy in Lincoln’s method of deducing the general efficiency of the washing process; for instead of being concerned with what takes place during the actual washing, the number representing the efficiency is determined solely from the composition of the final products. It is possible for a plant, working at a relatively high efficiency to yield comparatively poor final products, and, vice versa, good final products may result from a low efficiency washer. The quality of the raw material is the determining factor. The following is a set of illustrative results for a washer : — Raw coal. Slack. p.c. p.c. Float ....... 84’5.89’5 Sink ........ 15’5.10’5 Output..... 100’00 .25’0. Washed coal. Peas. Beans. Nuts. Refuse. p.c. p.c. p.c. p.c. .92’0.....94-0.....95’0...12’5 . 8’0..... 6’0..... 5’0...87’5 23’0......22 0....20’0....10’0 One hundred pounds of raw coal entering the washer produce the following quantities of washed coal: — Weight of Weight. f A Class. Float particles. Sink particles. Lb. Lb. Lb. 25 Slack 22’38 262 23 Peas 21’16 1’84 22 Beans 20’68 1’32 20 Nuts 19’00 1’00 90 Total 83’22 6’78 Percentage composition of total washed coal 92’47% ... 7’53% The concentration of the float particles in the raw coal is 84-5 per cent., and in the washed coal 92-47 per cent.; hence, the concentration has been raised 7-97 per cent, out of a possible 15-5 per cent. There- 7*97 fore the qualitative efficiency is x 100 = 51*42 15*5 per cent. The refuse, of which 12-5 per cent, is float material amounts to 10 per cent, of the raw coal. Hence, from 100 lb. of raw coal 1-25 lb. of the float material has been lost in the refuse. Therefore the process has recovered (84-5 — 1-25) or 83-25 lb. of the 84-5 lb. entering the plant, and the quantitative efficiency is = x 100 = 98'52 per cent. 84*5 General efficiency = 50*66 per cent. Output.—It is obvious from the above example that, to calculate the general efficiency of the washing pro- cess the output must be known. As the figures for the output of the washers are of a confidential nature, it was decided to state merely the general efficiency of the washing process. By that method it was impossible for any specific information to be deduced, or even a surmise to be made as to the particular washers from which the samples were collected. The qualitative efficiencies of the different washing pro- cesses examined varied from about 25 per cent, to 75 per cent., and averaged 58-20 per cent. The author would state, however, that he considers that where the quantitative efficiency fell below 65 per cent, it might have been raised to that value by making slight alterations in the working of the plant. Perhaps 60 per cent, might be deemed good average working for trough washers. The quantitative efficiencies averaged 97-89 per cent, and varied from about 91 per cent, to 99-8. The average of 97.89 per cent, means that 2-11 per cent, of the coal is being lost. From general observation it would appear that a general efficiency of about 65 per cent, should be considered average coal washing practice, whilst 75 per cent, is certainly excellent working. The average value for the fifteen washers is 56-93 per cent., with a range of from 25-16 per cent, to 72-38 per cent. That enormous variation indicates the immediate importance of placing the practice of coal washing under scientific control. General Conclusions. The methods of treating the coal and the washing plant, which are unsuccessful at one colliery, generally cannot be adopted as a whole at another colliery. At each colliery there are unique circumstances which demand specific treatment. The nature of the coal, and of the associated and inter-bedded impurities, should control the design of the plant and the selec- tion of the washer. The more friable coals should be treated in jig washers. In the Robinson washer there is considerable trituration. Although the levigation of the material is less in the trough than in the Robinson washer, it exceeds that in jig washers. The sizing of the raw coal fed into jig washers appears to be unnecessarily close. The details of the washers need frequent readjustment to suit the vary- ing properties of the raw coal. The bin filter is the most economical and satisfactory apparatus in which to drain the washed coal. Furthermore, a compara- tively clean water is recovered, which may be used again in the washer. Settling ponds do not entirely remove the fine slime from the water. Therefore, after the water has left the settling pond, it may be filtered through a heap of rubbish. Where the specific gravity of the coal and impurity approximate to the same value, Nebel’s suggestion of drying the coal previous to washing may be examined experimentally. If the specific gravity of a flat piece of impurity exceeds 2-30, it will be released from the washer in the refuse. Should it be less than 2-30, it is doubtful whether the impurity would be discharged with the refuse or with the washed coal. Raw coal of 1J in. diameter is most amenable to a reduction of the ash content by washing. Therefore, care should be taken to prevent the production of fine material by remediable breakage. The breakage which is occur- ring at some collieries could be reduced, the investi- gation having proved that with coal of smaller diameter than f in., the concentration of the coal in the washed product is comparatively low. Therefore, the ash con- tent is high. Although the washing of material exceed- } in. diameter is efficient, the ash content was high for inter-grown coals, on account of the poor quality of the specifically light particles. On the other hand, if such coals need crushing to less than f in. diameter to free the impurity, the ash content will still be high, for the reason stated above. Hence, it is doubtful whether the process of crushing to a smaller diameter than J in. previous to washing is not unnecessary and uneconomical, as no purpose seems to be achieved. As a rule, the quality of the fine coal dust deteri- orates slightly during the washing process through the disintegration of the mineral impurities to form slimes. It would appear to be advantageous to remove the dust from the raw coal before delivering it into the washer. The dust extracted from the raw coal could be mixed with the washed coal without unduly reducing its quality. The slimes would consist then more largely of mineral matter, and could be thrown away as refuse. Float and sink tests should be made regularly at the collieries. In recording these tests, it is important to state the time the material was immersed in the solu- tion. Any deviation from the specified time caused errors in the results of the tests. In view of the fact that it is not the duty of the engineer or washer fore- man to make float and sink tests, a trained chemist should be appointed, who should be able to conduct the experiments, interpret the results of the tests, and issue reports upon which the engineer and foreman could act in order to improve the washing. Virtually, this would place the control of the washer in the hands of the chemist. Washing for 10 minutes with water is capable of removing the calcium chloride from the coal after it has been immersed in a solution of the salt for 20 minutes. Therefore, would it not be possible to pre- pare a high-grade fuel by stirring the raw coal into such a solution? The floating coal could be drawn from the surface of the solution, and washed thoroughly with water to remove the calcium chloride. As a rule, washing reduces the sulphur content of the coal. The specific gravity of the washing water serves to indi- cate the state of the settling pond. If the specific gravity tends to rise, immediate attention should be directed to the pond, which might need cleaning. The solvent action of the water does not remove an appre- ciable quantity of the soluble mineral salts from the coal. Although the calorific values are not as reliable as they might be, the results indicate that the calorific value of the air-dried coal on an ash-free basis is con- stant for the various samples from the same washer. The caking power is increased by the washing process. The fusion temperatures of the ashes yielded by the various washed samples and the raw coal do not differ . materially. Therefore, the fusion temperature is dependent on the nature of the normal ash of the coal, and not on the extraneous inorganic matter. It would appear that the concentration of the soluble salts in the washing water is never likely to rise so high as to effect an increase in the soluble salt content of the coal. The average loss of the coal with the dirt should not exceed more than 10 per cent, of the total weight of the refuse. It is .doubtful whether the value of this coal is sufficient to warrant re-washing the refuse. The. expense of washing the larger-sized coal, and the fact that the ash content was not reduced to the minimum, suggest that it might be more profitable to adopt a dry method of cleaning for raw coal exceeding 1| in. diameter. Dry cleaning presents many advantages over washing, if only it can be made a thoroughly practical and com- mercial success.