The centre

It is now necessary to deal with the physical changes which take place when coal undergoes carbonisation. In a well-filled horizontal retort, with some 3 or 4 in. space above the charge, it is probable that the periphery of the charge, which is in direct contact with the walls of the retort, very rapidly attains a high temperature. rate of heat transmission to the would, if the material were not undergoing constant physical change, be very rapid in the first few hours, but would decrease as the temperature of the charge approximated that of the retort wall. But the physical condition of coal changes during carbonisation, and particularly during the first few hours. These changes largely determine the rate at which heat is transmitted to the centre. Certain resinous constituents of coal melt readily, and it is probable that these, on fusion, tend to cement the whole charge into one mass; but, though there may be a period in which the whole charge of the retort is sticky or slightly plastic, it is impossible to imagine that there is any degree of homogeneity in the plasticity of the mass. It is conceived that the outer crust of the charge is very rapidly carbonised and cemented into coke. This may take place even while, in the centre of the charge, there

exist pieces of coal which still retain their identity. As the temperature of the mass increases there is produced a plastic layer which forms a liquid wall through which the products of distillation from itself and from the interior of the mass have to pass. This layer is probably derived from the fusible constituents, and on decomposition, froths and bubbles, producing expansion of the coal charge and giving the coke a cellular structure. Now it is the cellular nature of the carbonised mass which is the cause of the low heat conductivity of the material, and it is thus indirectly the plastic layer which impedes the rate at which heat is transmitted through the charge. It is imagined that the gases evolved during the whole of this process fight their way to the wall of the retort, either through the plastic layer or through the cellular coke already produced, and, if evolved at the side or bottom of the retort, travel between the charge and the retort wall until they reach the space above the charge. An untold hubbub must take place as the plastic layer finds its way towards the centre of the charge, and is there gradually heated to the decomposition point. Whereas the gauntlet to be run by the volatile matter during the early stages of carbonisation was drastic, now every bubble of gas produced must fight its way through tar and other semi-liquid products, until it finds an exit. Each step it takes brings it into a hotter zone, and the gas molecule which We imagine started frem the centre of the charge with a wonderful degree of complexity is so squeezed and pushed and heated and washed that by the time it finds its way through the coke it has been reduced to an appreciably simpler form. In all these reactions a certain time contact is necessary, and the final form taken by this molecule will depend largely upon whether it may be pushed or pulled out of the retort before sufficient time has been given to break it down completely.

SO

This picture of the physical changes which coal undergoes on heating indicates that it is the plastic layer which prevents heat being rapidly conducted through the coal. The problem of increasing the heat transmission of the coal substance itself will be solved by adopting suitable means for preventing the formation of the plastic layer. In the absence of the plastic layer it is possible to effect rapid penetration of

heat through the charge, and one would in consequence anticipate an increased yield. not only of gaseous therms, but also of total volatile therms. Now the plastic layer may be avoided. or at least materially reduced, by intimately admixing with the coal to be carbonised an inert absorbent material, such as coke or coke breeze. This material will avoid the formation of a plastic layer by providing an infinite number of small pores and channels through which gases may escape, and by absorbing the tarlike matter formed. Moreover, the tendency of the material to expand during carbonisation will be reduced, as the large cells which originate from the plastic layer will no longer be formed. These large cells constitute a great impediment to heat transmission, and with the particles of coal and inert material lying in close proximity during the whole carbonising period, a small celled structure will be produced, with the result that the rate of heat transmission will be considerably augmented.

By admixing coke with coal and briquetting without a binder, the rate of gas evolution is materially increased. For the purpose of demonstration equal weights of the same coal, one of which was briquetted with 25 per cent. of coke, were heated in silica bulbs in a furnace, the temperature of which was about 900° to

VOLUME & CALORIFIC VALUE CURVES OBTAINED IN THE CARBONISATION OF ORDINARY COAL & BRIQUETTED COAL COKE MIXTURE

75

25

1,000° C. The rate of gas evolution was greater in the case of the briquetted mixture, and the effect of the inert absorbent material in the coal mass is to increase the rate of heat transmission. Such an experiment demonstrates that there is a visible difference in the rate of carbonisation when a briquetted coal-coke mixture is heated on an exceedingly small scale. Some results obtained when working with quantities of the order of a retort charge were presented. In fig. I. are seen the volume and calorific value curves obtained from the carbonisation of a 75/25 briquetted coal-coke mixture compared with those obtained from the carbonisation of coal under similar conditions. It is the curve relating to the briquetted mixture corrected to 100 per cent. coal that is important, as it shows the rate of distilling the coal itself.

FIG. II.

These curves indicate that the nature of the carbonisation process is entirely changed when a briquetted coal-coke mixture is used instead of ordinary coal. The carbonising

process has been accelerated so that all the gas has been expelled from the briquettes in 6 hours, as compared with 9 hours in the case of untreated coal. As the difference between these two periods is 31 per cent., it follows that the throughput of coal as 75/25 briquettes would be greater than were ordinary coal carbonised. Striking though the diagrammatic curve may be, it is here that the thermal models employed in the first lecture show so admirably the difference in the rate of carbonisation produced by the admixture of an inert material with the coal. Fig. II. shows the thermal models obtained from the carbonisation of ordinary coal and a briquetted coal-coke mixture.

It has been shown how striking is the difference between the amount of useful work effected in the retort during the middle ten minutes of the first hour, as compared with that effected during the middle ten minutes of the last hour of the carbonising period. This difference is greatly augmented in the briquetted coal-coke mixtures. Fig. III. shows three sections taken from the thermal model obtained from the 10-hour carbonisation of ordinary coal, and two sections taken from the thermal model relating to the carbonisation of an equal weight of a 75/25 coal-coke mixture. When briquetted material has lost practically all its volatile matter, a large amount of work is still being effected in the case of the slower carbonisation process of ordinary coal.

Thus the admixture of an inert material with coal increases the rate of gas evolution. The question arises whether this enhanced rate of gas evolution is accompanied by the superior thermal yield of gas which is to be expected from the results of the laboratory investigation previously described (Lecture II.).

Carbonising trials with briquetted coalcoke mixtures were carried out in an experimental, but up to the time of going to press it had been impossible to operate the setting in such a way as to obtain through the retort retort walls the required quantity of heat to effect the degree of tar cracking necessary to give the enhanced thermal yield of gas.

Very beneficial volatile therm yields have been obtained, however, in these experiments, the results of which are sufficient to demonstrate at least that an entirely new field is open for obtaining a more economic distribution of the therms of the original coal. As many as 8 to 9 additional volatile therms, which are several times more valuable than non-volatile therms, have been obtained from the particular coal under test. These additional volatile therms are obtained at the expense of coke, and it will subsequently be necessary to account theoretically for them. To redistribute the thermal energy of coal in this newer way will constitute an important contribution to the coal conservation problem. The tar obtained in these experi

The question of carbonising briquetted mixtures has been studied from the point of view of obtaining the maximum yield of volatile therms, but other investigators have recommended the addition of inert matter to the coal in order to obtain a resulting solid fuel of high density and with a high

degree of combustibility.

That this solid fuel is indeed superior to coke has been al mirably shown by E. C. Evans. It possesses a greater heat conductivity than ordinary coke, owing to the fact that the absence of the plastic layer has prevented expansion of the coal, with the result that a small-celled structure is produced. In consequence, combustion proceeds with greater velocity, while the powdered ash, instead of containing as much as 10 to 20 per cent. of combustible matter (as is often the case with coke) is practically free from carbon. Many

COKE

COKE FROM

BRIQUETTED COAL COKE MIXTURE

FIG. V.

investigators are engaged in a study of this fuel, and for several years now the author has worked upon this question, and recently in collaboration with E. R. Sutcliffe and E. C. Evans.

mass.

It was originally decided to admix inert material with coal in order to break up the plastic layer which prevented heat being readily transmitted through the Firstly, it is essential to admix the correct proportion of coke, as is seen from fig. IV., which shows sections of briquetted fuel from (a) 100 per cent. coal, (b) 85,15 coalcoke mixture, (e) 75/25 coal-coke mixture. It will be observed that while the formation of the plastic layer is reduced in the case of 85/15 coal-coke mixture, it will only be avoided when as much as 25 per cent. of coke is admixed with the coal. That there is practically no plastic layer formed when the right proportion of coke is used in the briquette may be proved by the fact that if the distillation process is arrested at any time after the third hour the volatile matter remaining is evenly distributed throughout the briquette. The importance of this observation in supporting the contention. that the plastic layer is avoided in briquette carbonisation is apparent.

SO

Fig. V. shows sections of ordinary coke, and that obtained from the carbonisation of a 75/25 coal-coke mixture. These sections show the particular cellular structure typical of ordinary coke, and the very dense small-celled coke produced when the formation of the plastic layer is avoided. The larger cells of the normal coke require a considerable amount of binding material, and a coking coal is one in which sufficient binding material remains after the period of expansion to allow the coke cells to be adequately cemented together during the shrinking process. With the production of a small-celled coke, such a large quantity of binding material is not required, and it is conceivable that a large proportion of the parent substance from which the binding material originates is distilled away and converted into volatile therms. This binding material is believed to be the source of the additional therms produced in the rapid carbonisation of coal and in briquette carbonisation. In order to develop this hypothesis an experiment was carried out, in which equal weights of coal (one being briquetted with 25 per cent. of coke) were heated in silica bulbs. The only dif ference between this experiment and the one previously described is that on this occasion the silica bulbs were heated directly by means of Meker burners, in order to de

velop a very large heat potential in relation to the weight of coal taken. In the previous experiment special means were adopted to ensure that the total available heat was limited, and inore in accord with large-scale practice. The rate of gas evolution is practically the same in both cases, owing to the fact that the quantity of heat available is so great that the coal charge is heated just as rapidly as the briquetted mixture. Under such conditions of drastic heating practically the whole of the binding material is converted into volatile matter.

It is often said that laboratory carbonisation of coal is of little value in forecasting the results eventually to be obtained on the large scale. That the laboratory distillation fails is due partly to the fact that the furnace is out of all proportion to the weight of coal charge used, and the amount of coke binding material evolved as volatile matter is higher than would be realised on the large scale. Further, the size of the retort in relation to the coal charge is also usually out of keeping with large-scale conditions, and a degree of tar cracking is practised which causes the results to be materially different from large-scale experience. The difficulty of designing laboratory apparatus to yield results comparable with large-scale practice is generally realised, but the author does not know whether the underlying cause of these differences is appreciated.

The question arises as to what is the maximum yield of volatile therms that may be obtained in the carbonisation of coal. This the author cannot answer, but one is

faced with the fact that a small-scale laboratory experiment has yielded nearly 100 therms of volatile matter from a coal which under normal conditions of working has given only 90 and this yield should be aimed at on the large scale. It may be pointed out in passing that if this hypothesis be true, then the usual test for volatile matter in coal is of very limited utility for computing the value of a coal for gas-making purposes. The test is carried out by heating very rapidly an exceedingly small quantity of coal; and under these conditions. of drastic heating the whole of the volatile matter is evolved, the coke is useless, and the result is one appreciably different from that eventually realised on the large scale. What is required to be known in computing the value of a coal is the proportion of volatile matter, which is recoverable as volatile therms under the conditions of the practical process to be adopted.

(To be continued.)