I N D U S T R I A L AiVD ENGINEERING CHEMISTRY
22
Vol. 21, No. 1
Synthetic Ammonia Plant at Ostend' F. A. F. Pallemaerts GNIONCHIMIQUEBELDE, BRUSSELS,BELGIUM
-
facture from coke-oven gas with one of the now classic s y n t h e t i c ammonia procAt the present time about a Of this type are in Operation and Others are being planned.
between water gas and cokeoven gas for h y d r o g e n A type of synthetic ammonia plant is described manufacture in the case of which is being rapidly adopted in Europe, It finds its the first coking plant in source of hydrogen in coke-oven gas and combines Belgium, the same relation three processes-the Union Chimique Belge (or Semetin favor Of coke-oven-gas Solvay & Piette) gas treatment process, the Linde hydrogen was found, and hydrogen extraction process, and the Casale ammonia the latter was t h e r e f o r e synthesis system, chosen. Nearly 100 million cubic feet of coke-oven gas are Extension of This Type of This superiority of cokeAmmonia Plant being treated per day by these processes, and more than oven-gas hydrogen is not 1000 tons of ammonia per day are being produced by necessarily universal, and in As yet the plants of this type now in operation. pacity of plants of this type localities where very cheap Thanks to the cheapness of the hydrogen recovered does not exceed 136 tons Of power is available, either from coke-oven gas and also to the general economy of hydroelectric or in the form ammonia per day, because the Casale synthesis process, ammonia is produced at gas is not available from of very cheap fuel, such as a very low cost, considerably below any published existing coke-oven plants in cheap and good lignite, the figures which have come to the writer's notice. sufficient quantity to give use of coke-oven gas may larger outputs and distillanot be the most economical; t i o n wor-ks have not yet however, according to our been systematically concentrated in Europe. However, a latest information, there is not yet a single plant where water move is being made in this direction in Belgium, where a gas for hydrogen manufacture is being produced from lignite. The following costs are found for 1 cubic meter of hydrogen series of central carbonization plants is being planned, arranged for generation of power and utilization of gaseous from different sources, under Belgian conditions, excluding by-products. The first of these central coking plants is to the cost of extraction of the hydrogen from its gaseous pribe erected in the southern Belgian coal field (Borinage) and mary material: will comprise a synthetic ammonia plant by the U. C. B. BELGIAN FRANCS DOLLARS Linde-Casale systems, the first section of which has a capacity In water gas: for 72 tons of ammonia a day. It is expected that a second From coke a t 170 Belgian francs per ton 0.22442 $0.0062 coal a t 188 Belgian francs per ton 0.214 0.0059 central plant in Hainaut, Belgium, will be completed at about In From coke-oven gas, a t 0.14Belgian franc per cubic meter, residual gas being valued a t same p r k e per calorie as the same time. original gas
Low Cost Due to Use of Coke-Oven Gas
The rapid increase in production of synthetic ammonia resulting from the new plants being built or projected cannot fail to alarm those concerned with the sale of ammonium sulfate, and experts have advanced various theories concerning the possibilities of absorption in quantities as a fertilizer. The Adriatic conference of a few months ago might be expected to have the effect of curbing further increases in production, and although such is not yet apparent, it would seem probable. I n the long run, however, the production cost of synthetic ammonia is the factor of greatest importance, and this is where, a t least under European conditions, cokeoven-gas hydrogen scores, for electrolytic hydrogen is either 1 Presented a t the Second International Conference on Bituminous Coal, Pittsburgh, Pa., November 19 to 24, 1928.
0.084
0.00234
Size of Plant
The original Ostend plant, which was really experimental as it constituted the first application of the combined CasaleLinde-U. C. B. systems, was built for an output of 16 tons of ammonia per day. It comprised three 8-ton Casale units and three corresponding Linde units, one unit of each plant being in reserve. For the first few weeks of operation, the production was low and irregular, and improvements, bearing almost entirely on the purification of the coke-oven gas preliminary to its fractionation, had to be made. Thereafter the production rose steadily until it reached an average of 18 tons, with peaks of 23 tons, giving a regular output 50 per cent above that for which the plant was intended. Four additional Casale units have since been installed, and the hydrogen plant has
January, 1929
I-VD r S T R I A L A N D ENGINEERING CHE-VIXTRY
been increased by two large units, each with a capacity of 3000 cubic meters of hydrogen-nitrogen gas; with these the plant produces 40 to 55 tons of ammonia per day. A plant of this size is small compared with the I. G. plant, the Billingham plant of Imperial Chemical Industries in England, or the Hopewell plant of the Allied Chemical and Dyes Corporation now nearing completion. However, it is proportional to the size of the country and, as already said, this type of plant especially lends itself to distribution and to economic operation under the conditions found in Europe. When coke-oven plants have been concentrated as is now being planned, additional units will be added and favorable results are expected from the enlarged plants. Thanks to the cheapness of the hydrogen, and also to the general economy of the Casale synthesis process, ammonia can be produced at the cost of about 9 American cents per kilogram of nitrogen in the form of ammonium sulfate (including the cost of the sulfuric acid and total operation costs, amortization. etc.).
23
Linde process utilized the well-known Joule-Thomson effect of self-intensive cooling through expansion without external work. For certain reasons the Linde system was preferred. The Casale system was adopted after a careful study of the three then available processes. The excellence of the choice has been proved by the fact that the Casale system has been installed all over the world in the face of competition from new processes, and has now reached a total capacity of more than 1000 tons of ammonia, equivalent to 4000 tons of sulfate, a day. Let us nom go into the details of the process. Preliminary Treatment of Coke-Oven Gas
Such little experience as was available came from a small experimental plant situated a t Oberhausen in the Ruhr district, where it had been installed with the collaboration of the Linde Company, by the originator of the coke-oven-gas hydrogen idea, N r . Bronn, an engineer of the Rombacher Eisenhutten Company. A l l that was known from the Oberhausen experience was Genesis of the Type that the gas had to be dry and free from carbon dioxide; After the Semet-Solvay R: Piette Company (which has also that it probably would have to be desulfurized before since become a division of the Union Chimique Belge) de- compression. These conditions had been effected a t Obercided to manufacture synthetic ammonia, the first problem hausen through purification of the gas in iron oxide purifiers, was a cheap source of hydrogen. Coke-oven gas was chosen folloned by compression, washing under pressure nith water for the reasons already given. to remove carbon dioxide, a final decarbonation by means The next question was how to obtain hydrogen from this of caustic soda, and drying under pressure by means of source. Several processes mere investigated-fractionation, calciiiin chloride. and the iron-steam processes. This treatment was not considered very practical for The last-named processes. TT-hich were being worked at quantity production and it was proposed to replace both the time by Bamag and by the Alais, Froges et Camargue desulfurization and decarbonation by an ammonia wash Company, were carefully compared and the conclusion process, \?-orking under ordinary pessure. This was accomreached that the Alais proceps, .crhich n a s conducted in plished only after certain obstacles had been overcome, for batteries of externally heated retorts, had not been sufficiently the washing of gas with ammonia proved more difficult improved. The Bamag or Messerschmidt system had been than might have been expected. The results of this treattried out on an industrial scale, but was found to require ment were excellent, the acid constituents of the gas being a large number of converters of small capacity, considerable removed to the extent that less than 0.1 per rent carbon labor, and to yield a rather impure hydrogen gas with a low- dioxide remained, which is better than is obtained with a value residual power gas. So these processes were elimi- pressure water wash, which leaves as a rule 0.3 to 0.4 per nated. cent carbon dioxide. Of the fractionation processes, two were taken under conThe chemical drying of the gas was omitted and left to sideration, both elaborated by world-famous firms; they the refrigerating plant. were the Claude sysT h e hydrogen tem and the Linde plant was started up system. It would a l o n g t h e s e lines, be hazardous to exbut difficulties arose press an opinion as from the presence of to the superiority of heavy unsaturated either of these prochydrocarbons in the esses over the other, gas; t h e s e hydroas both are splencarbons p ol y m e r didly worked out. ized in the less cold The principles of parts of the hydroe a c h of these sysgen apparatus and tems are t o o well caused obstructions known to r e q u i r e and even local exdetailed descripplosions. One aption here. Briefly, p a r a t u s was even whereas the Claude partly d e s t r o y e d system applies exand, as a larger unit p a n s i o n w i t h ex(3000 cubic meters ternal work, by reof hydrogen-ni t r o l e a s i n g t h e comgen) was under conpressed gaseous struction, the small m i x t u r e in an exunit was not rebuilt. pansion engine A thorough investiwhich Claude calls a gation of the causes of the trouble was hydrogen motor, the Coke Ovens
INDUSTRIAL AND ENQINEERING CHEMISTRY
24
undertaken, and i t was decided to install a pressure water wash in order to remove all hydrocarbons which had a solubility coefficient higher than 1. This proved very effective, and the plant has heen running smoothly ever since. A few incidents of minor importance, however, have shown i t desirable to carry the purification process still further, for the purpose of removing oxides of nitrogen and volatile nitro
Vol. 21, No. 1
The compressed cold gas that is left is now a mixture of hydrogen and nitrogen, but still contains small quantities of methane and carbon monoxide. These impurities are almost totally eliminated by means of a liquid nitrogen wash, which removes methane entirely and leaves only 0.001 per cent of carbon monoxide. Oxygen and water vapor are also totally removed. There is no other process, t o the author’s knowledge, which produces such pure hydrogen from any non-electrolytic gas. The carbon monoxide elimination is far more complete than with any chemical process, and its chemical elimination by means of cuprous solutions was abandoned by the Linde Company as soon as the liquid nitrogen wash had been a p plied. This extremely important feature gives the Linde process a decided superiority over certain processes in which no attempt is made to eliminate the considerable reaidual carbon monoxide percentage that remains in the gas as i t leaves the expansion motor. This is, however, not so great an inconvenience with the Claude synthesis process, because part of the residual carbon monoxide is converted into methanol by a preliminary catalytic treatment, and also because the large amount of methane which is formed in the ammonia catalyst tubes is eliminated with the waste bases, as there is no recirculation. Cost of Linde Coke-Oven-Gas Hydrogen
Coke-Oven Gas Fractionation Apppsrstue
compounds. A simple physical treatment of the gas has been devised, of which excellent results are expected. The trouble due t o the formation of explosive copper acetylide has also been eliminated. The ammoniacal liquor obtained in the decarbonation treatment of the gas is distilled into the sulfate saturators. This simple process will be replaced in future plants by a system in which the ammonia is regenerated for the same purpose, in order to make all the liquid ammonia produced in the ammonia plant available for uses other than sulfate manufacture from sulfuric acid. Lat.ely, patent applications have been made for a process replacing the whole classical treatment of coke-oven gas for the recovery of ammonia and benzene by a new simplified system combined with the synthesis of ammonia. Future coke-oven plants built by U. C. B. will in all probability he without saturators or benzene scrubbers in their by-product department.
The two main factors in the cost of Linde coke-oven-gas hydrogen are energy and the cost of the coke-oven gas itself. The energy consumption per cubic meter of hydrogen at atmospheric pressure is 0.246 kilowatt-hour. The total expenses for the production of 1 cubic mctor of hydrogeu (without nitrogen and a t atmospheric pressure) amount to 0.141 Belgian franc (80.00395). The cost of the coke-oven gas which is used obviously depends or1 the available market
Plant for Manufacture of Hydrogen
This plant, built entirely by the Linde Company, is a wonder of industrial physics. The purified compressed gas is cooled in stages, first in a countercurrent cold exchanger, then in a liquid-ammonia refrigerator, and finally in the fractionation apparatus, wherc the Pictet cycle is applied together with the Joule-Thomson effect on nitrogen. This fractionation spparatus is arranged to utilize in the most effective way the cold contained in the condensed gases that are tapped off a t various points on the gas circuit as t.he temperature is lowered. The condensates which are successively removed are, in order:
-.
..
.
condensed-coke-oven gas. (2) Melha~eat temperatures of 25” to 85’ A,: it contains mainly methane, with ethane, ethylene, and unconrlensed pas. (3) Crrrhon Monoride. a lisuid mixture of urban monoxide and nitrogen, with some methane, ethylene, etc.
One of the Nitroeen Cempreasors
or OD the assumption made as to its value. In certain cases, where there is JIO sale for coke-oven gas and where its cost is taken as the differencebetween coal plus operation expenses and total income from by-products, i t may be considered ns having little or no value. This is unusual, however, and IS daily becoming more so. At the Ostend plant, where gas is sold to the city gas distribution company a t the same price per kilop.am-edorie as coal, the cost of gas is as follows: 2 3 cubic meters of coke-oven gas at 0.14 Belgian franc per cubic meter = 0 323 Belgian franc
However, as the Linde process returns 70 per cent of the calorific power of the gas in the form of a high-calorific-power
January, 1929
INDUSTRIAL AND ENGINEERING CHEbUSTRY
gas at 6500 to 8000 kilogram-calories per cubic meter; and when allowance is made for the income from this source, counting the hydrogen-free gas a t the same price per kilogram-calorie as the original coke-oven gas, the detail is as roilows: 30/100 X 0.323 = 0.0969 Belgian franc ($0.0027) The total cost of Linde coke-oven-gas hydrogen is therefore 0.00395 4- 0.0027 = $0.00605 per cubic meter ($0.1883 per 1000 cubic feet)
The author does not know of any lower cost for hydrogen. Utilization of Residual Gas
I n this calculation the residual high-power gas, vhich consists mainly of methane and ethylene and their homologs, with some nitrogen, has not kcir valued at more than t h e o r i g i n a l cokeoven gas per calorie. This is a fair assumption as long as it is not used for any other purpose than coke-oven gas itself is used. ' But this gas is m u c h more valuable than cokeoven gas. I n the first place, i t is entirely free of all sulfur compounds and other i m p u r i t i e s , a n d is absolutely anhydrous. It also is particolarly suited for certain industrial uses, such as the heating of s p e c i a l fnma.ces-the Libbey-Owens process for instance-and i t also lends itself remarkablv well to long-distance conveya.nce of gas which is being developed in condinental Europe. A simple CalculaOion shows that one million kilogmm-calories represent 121 cubic meters of rich residual gas at 8270 kilogram-calories per cubic meter or 214 cubic meters of ordinary coke-oven gas as 4680 kilogram-calories per cubic meter. The compression of these volumes to 30 atmospheres, which is generally admittcd as being suitahlc for longdiskmce distribution, will require 25.4 horsepowr in the ease of the former gas, as against 45 horsepower in the case of the latter. The energy economy will tberefore be 43.5 per cent in favor of residual gas. The above advantages concern only tlie crude uses of tlie rich residual gas in the crude form. This gas, however, lends itself to even more valuable application, for its CONIponents can be collected separately and used in a whole series of chemical processes to form valuable products. Ethylene can be converted into ethyl alcohol (this application has already been put into practice in certain European plants), diethyl sulfate. or anv of the valuable haloeen addition product;. Propylene, khich accompanies its lower homolog, can be transformed into its corresponding alcohol by the same process. The Linde Company builds ethylene fractionating apparatus which will convert the 30 per cent "ethylene" obtained in the ordinary Linde coke-oven-gas fractionating apparatus into 80 to 95 per cent ethylene, the gas being thus rendered suitable for chemical processes requiring ethylene of high
25
punty. The ordinary 30 per cent ethylene can be used without rectification in many processes, especially in dcohol and ether manufacture. A Linde ethylene fractionation apparatus has been installed by the I. G., and another will soon be Norking a t the Ostend plant of the U. C . B. Ifethane has not yet found any considerable industrial use, and i t is merely used for heating purposes. A certain amount, however, is compressed into tubes and used for oxymethane steel-cutting and for the soldering of steel, lead, eta. A small quantity of hydrogen has to be mixed with the methane in order to increase the velocity of combustion, which is too slow in the case of methane alone. This use vdl probably never absorb the entire production of methane in cokeoven-gas hydroKen plants; hut without doubt additional uses
I
~
gas. Indeed,wherc-
as 1 volume of cokeoven gas will yield 0.5 volume of hydrogen, the same quantity of coke-oven gas will give 1.7 volumes of hydrogen, or more than three times as much, if methane is converted into hydrogen. It will therefore be apparent that the adn n t a g e is considerable even witliont a remunerative use for methane. It must not be forgotten, however, that in this case there is little residual gas left for heating tlie ovens, which have t o be supplied with some other gas, such as producer gas, and this gas is usually more expensive per calorie than cokeoven gas under European conditions. It will be seen from the above that, based on the residual gas having the sa,nie value as the original coke-oven gas, the price will probably be lowered considerably if the various constituents of the residual gas are exploited to the greatest possible advantage. Description of Ostend Type of Coke-Oven Hydrogen Plant
Cokeoven gas from the adjoining coking plant, which has undergone the usual tre.atment for b7-product recovery, is extracted by means of exhausters and submit.ted to a treatment in which i t is successively washed with an ammonia solution, %,it11 water in sufficient quantity to remove the ammonia carried over, then with sulfuric acid, which eliminates the last trace of ammonia, and finally with caustic soda, which fixes any sulfuric acid or carbonic acid which might be left in the gas.
26
INDUSTRIAL AND ENGINEERING CHEMISTRY
Val. 21, No. 1
The purified gas is stored in a gasholder, from which it is in part as various “residual” gases which at present are mixed exhausted by compressors which compress it in two stages and, after cold exchange, returned to the coking plant where to about 9 atmospheres. Under this pressure it is refriger- the joint residual gas is used to heat the coke-oven batteries. These coke ovens, which are of the U. C . B. (Semet-Solvay ated by means of liquid ammonia from tlie synthesis plant, aiid in passing through a cold exchanger i t is warmed again & Piette) combined type, are arranged for being heated by producer gas, rich gas, or any interinediate gas. The most to about normal temperature. At this tempera,tureit is iiow washed with water in pres- varied gaseous mixtures, depending upon the demand for sure water-wash bowers, where 1 volume of compressed gas gas from tlie town gas-distribution company, which tliey also is washed with a little more than 1 volume of water. The supply, arc used successfully. Sometimes some coke-oven older towers pass about 1750 cubic meters of gas per hour, gas is used, but the normal supply is a variable orie consistthe new towers, about 5300 cubic meters arid the correspoiid- ing oi proditcer gas (from U. C. B. automatic producers gasifvine coke breeze) and rich residual Linde gas. The ovens ine of water. Iouantit.ies . The energy of tlic water as it leaves the pressure towers &rG splendidly on this variable but ~ve1l:controlled mixis recovered in 1157draulic turbines placed on the same shaft as the pumps arid m o t o r s , each group c inp r i s i n g m o t o r , pump, arid turbine. The motor lias t o d o all tlie work a t tlic start, but its power cons u m p t i o n falls to about 40 per cent ~\-heiithe water is a d m i t t e d t.lirough the turbines. This water evolves gas as soon as it has passed the turbine, like soda water in a bottle when t h e stopper is rernoved. to be seen, proving This eas is collected Hydrogen-Nitroeen Iiypercompreueor(i tlrst, nll th” “..I” ...IIiPnor”” aiid r&md to the residual eas from the fra,ctionation. Tlie “dceasification eas” was dissolved by the pressure Ivater, whicli acted more effect i d y than any oil wash. It >vas therefore decided to iiist~all is not a.li carbori dioxide as iii the case of the 1Iaber-R‘;sch water wash, where a gas with 30 per cent car!nn dioxide is a sermxi refrigerator before the Tmter wash, so that bcnaeiic em of berizene recovery does pressure-washed. Indced, the gas here, besides containing was agairi recovercd. This origiiially ouly 1 to 2.5 per cent carbon dioxide, has been not eonsitme aiiy cricrgy or materials, as it consists simply very cotnplet,cly decarlronated in the ammonia wa,sh. The of static refrigerators and cold exchangers and is run 1111 object here is entirely different, mid is to remove mainly liquid animonia from the synthesis plant. acetylene present in the proportion of 4 pwts per thousand Siiice the Ostend plant has been started, the Linde Conronly. I t seeins ridiculous to have to resort t o sueh powerful paiiy Jins built a niiiriber of installations, trenting daily a total means to reniovo a disagreeable gas present in siich niiiiute of more t.lian 2.5 million cubic meters of coke-oven gas. qumtities, but the treat.iricirt is effective, besides being ab- Besides t.he extensions to tlicse various plants uiidcr prepasolutely necessary, uiitil some better means is brought into ration, or already being carried out, as at Ostend, several action. As it is, the ”degasification gas” hiis the following large new instullntions have I~een planned. Tho earliest composition: plants, cxrept for the experimental unit a t Oberhnusen, wcrc built first in I3elgiiim and tlien in France, brit of late sevcral Per em erected in Germany. It iiow looks as Ole6-S . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 oxygm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 if the Germans, viho apparently were a t first coiitent to rely Carboil monoxide.. . . . . . . . . . . . . . . . . . ti Hydrogen. . . . . . . . . . . . . . . . . . . . . . . 81 on t h r I. G. process, are seriously ctitering the field with tlie Methane . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Iinde coke-oven-gas hyrlrogeo process. Ethane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 In all, t.he Liiide Company lias provided seven units of There is no use for this gas other than to incorporate it 1600 eiibic meters of coke-oven gas per hour, and nineteen units of 5000 cubic rncters per hour-all for ammonia syninto the residua.1gas used for liesting purposes. The coke-oven gas, which is now free from berizenc and thesis. One unit for treating 7500 cubic rneters of gas per amply purified, is refrigerated for the second time by means liour is under construction, which will supply hydrogen for of cold exchangers and ammoiiia refrigerators t o a tempera- tlic Bergius process. t.ure of about -45” C. At this temperature watcr vapor is I t is worthy of notice that the Linde hydrogen-nitrogen removed sufficient.ly t.o avoid rapid obstructioii of the frac- process does away with the expensive combustion of hydrotionating apparatus due to freezing of the condensed water. gen in air, which is applied in certain plants in order to inThe gas now enters the fractionating apparatus, and leaves i t corporate the necessary amount of nitrogeii into the gaseous in part as a pure mixture of hydrogen and nitrogen with tlie mixture. Liiide nitrogen costs in power only 0.2 kilowattdesired nitrogen percentage for the ammonia synthesis and hour per cubic meter iir the larger unit, and it would be
January, 1929
INDUSTRIAL AND ENGIiVEERING CHEMISTRY
uecessafy to secure hydrogen for nothing in order to justify using 1 volume of precious hydrogen in order to obtain 2 volumes of nitrogen. I n the hydrogen-nitrogen apparatus the nitrogen, besides eoming partly from the coke-oven gas itself, is introduced by satura.tion of the hydrogen with rritrogen during the liquid nit.rogen wash. The nitrogen proportion is ingeniously regulated by acting OIL the temperature of the gas, the pressure being kept constant. The temperature itself is regulated by the pressure under which liquid nitrogen is kept in the liquid nitrogen boiler, which constitutes the last stage in tile cooling of the gas. A pressure of about 0.2 atmosphere in the uitrogen boiler insures the required nitrogen pereeuta.ge in the mixture, the gas itself being under 9 atmospheres. It has been fouud useful at Ost.end to generate hydrogen with a little less tliaii 25 per cent of nitrogeii, and to correct the mixt.ure in each synthesis unit by means of an addition of nitrogen. The Ostend nitrogen machines are of a size corresponding to the hydrogen units. They produce pure nitrogen of 99.8 per cent purity and a t the same time pure oxygen a t 99 per cent. The Linde Company builds units much larger than those at Ostend, bhe lsrgest making 3600 cubic meters oE nitrogen per hour (126,000 cubic feet per hour). I n one plant 16,000 cubic meters (860,000 cubic fcet) of nitrogen are produced per hour. Ammonia Synthesis Plant
This is a Casale plant. The Citsale system lrss become very widely distributed in Enrope, Russia, and .Japan. More than 1000 tons of anrnioiiia a day are produced by it. This shows very remarkable development, as the first Casale unit, wliioh produced 2 tons of ainnioiiia per day, was started a t Terni in Italy as late as the autuinn of 1021. From 2 t.oos tlic capacit,y of the Casalc unit was iucreased sueccssively t,o 6, 20, and to 30 tans. Altogether t.went>y-fourplants are a t present in operation, distributed all over the world: ten in France, three in Japan, one in Russia, one in Spain, two in Switzerland, two in Relgium (where anot.lier large plaut, is being planned), three in It.aly, oiie in England, one io the United States (origii~allyerected a t Niagara Falls but now being transferred to Delle, CV. Va.). The Casale synthesis plant runs on hydrogen from any source; nine plants run 011 coke-oven-gas hydrogeu, teu on electrolytic hydrogeii, three on mater-gas hydrogen. The equipment is standardized so that a 5, 8, 16, 24, or 30ton a day Casalo unit CRII be purchased complete. At. Ostend there were zit first. three &ton units; a t present there are seven. In tlie iicw plarits being planned by tlie U. C. 13. Company, 30-ton units are eouteniplated, altliougl~ no unit of this size has yet been built. The experience with the 20-ton unit, which is in action in several plants, and which has an output of 26 to 32 tons (actually renched in one of the Japanese plants), shows that a produetion of 35 to 40 t.ons may be expected of tlie 30-ton unit. Wlieii the iierr large unit goes into operation? it will, to the best, of the writer’s knowledge, be bigger than any ammonia uuit, iiichiding the Haber-Bosch I. G. unit. As is xvell known, the Cnsale process works under a pressure of 750 atmospheres. Of late it lias been thought in various quarters that i t was advantageous to carry out the synthesis at l o r w pressures, 200 ntmosphores or even less. In the miter’s opinion, there is much to be said against this view as the use of high pressure has very decided advantages. ,One of the most important is that the entire aininonia output is available in the form of liquid ammonia, the latent cold content of which is extremely useful and permits very con-
27
siderable energy savings in other parts of the plant, especially where coke-oven gas has to be treated. The elimination of the huge refrigerat,ing machines, which vould be required in the absence of available liquid ammonia, alone constitutes a considerable cconomy of energy, as has been mentioned in the description of the coke-oven gas treatment. The availability of the total output in the form of liquid ammonia also renders complete removal of benzene possible without any appreciahle energy consumption, and besides, there are still othcr uses of this liquid ammonia in the modern srnnionia derivatives plant, siicli as refrigeration of the i~itrousfumes in the ammonia oxidation nitric acid plant, etc., which made the liquid state of the Casale ammonia an extremely valuable asset. to the process. As regards power consumption in compressing the gas, it has been repcatedly sbown that there is very little difference in the energy consumed once the pressure exceeds 200 or 300 atmosplieres. For instance, tlic ratio of the energies required for compressing an ideal gas t o 729 and to 243 atmospheres is 6 to 5, so that,, whereas tlrc Casale process c o n s ~ ~ ~ n1.2 e s kilowiitt-hours per kilogram of ammonia for compression, a process working under 300 atmospheres will still consuinc 1.05 kilowat,t-bours. This difference, small as it is, is compensated by the absence of a recirculating pump in t,lie Casale cycle, recirculation being induced in tlie latter system by the injection of the fresh gas, under a drop of about 50 at,niosphercs pressure, Wheu the further energy savings resulting from obtaining all the ammonia in tlie liquid form is added, the advantage of the Casale system as regards power consumption is obvious. The above comparison is e ~ d made y when it is remembered that the isotliermal compressiou enerby is proportional to the logarithm of the pressure. The high-pressure equipment of a Casale unit is so compact that it meiglis far less tiinn any unit of the same capacity working under a so-oalled mrdirrm or low pressure,
Synthesis Tubes, Ammonia Condensers and Receivers
At Ostend the hydrogen-nitrogen mixture produced in the Iinde hydrogen plant is sent to a gasholder, whence it is exhausted by the compressors. These compressors (there are six of them) each compresses 1200cubic metersof hydrogennitrogen mixture per hour; it must he remembered that they are the srnaller type OS unit. The compressed mixture is admitted as such, without aiiy purification or desiccation, into the synthesis cycle, which coniprises an oil separator, a synthesis tube, a water-cooled condenser, and a c0ndensitt.e receiver. At Ostend recirculation is still being effected in most of the units by means of recirculat.ion pumps, but these
28
INDUSTRIAL A N D ENGINEEEUNG CHEMISTRY
Vol. 21, No. 1
are being replaced hy a new injector system which simplifies the circuit considerably and does away with the upkeep and attendance, as well as power consumption, necessitated by these pumps. Further simplifications have been brought in a t the pioneer plant a t Terni, which is run as an experimental p h n t hy the Societa Italians Ricerche Indwtriali Sii, this company being the research organization of the Casale Company and founded by Luigi Casale. These simplifications consist in the elimination of the o i l separator (as a result of tlie suppression of
the hot gases as they leave the catalyst. The catalyst is contained in an annular space in tlie core of the tube, and is thermally insulated from the inside wall of the pressure tube hy the concentric heat exchanger and the cold gas which enters the tnhe. This arrangement, which is protected hy several Casale patents, has proved exceedingly effective. The reaction is regulated in an ingenious way. As below 750 atmosplicres pressure the violence of the catalytic reaction would he such as to overheat the catalyst, there is left in the reaction mixture sufficient uncondensed ammonia to keep the system just slightly endothcrmal, so that it may be regulated by means of an electrically heated resistance placed in the axis of thc catalyst tube, and over which the gases pass just before entering the catalyst, after they have been heated by temperature exchange with the gases which leave the catalyst. In practice the voltage aod amperage are kept constant, the speed of recirculation being regulated so as to maintain temperatures and regular production. The energy consumed hy the resistance is from 0.27 to 0.18 kilowatt-hour per kilogram of ammonia. The presence of the necessary quantity of uncondensed ammonia in the gaseous mixture before catalysis is insured by simply water-condensing the reaction gases. KO simpler means could be devised. I t is apparent that the system as a whole is marked by the utmost simplicity, as all special condensation arrangements are controlled with artificial refrigeralion and no special purification treatment of the gas, either catalytic or chemical, is necessary.
Nitrogen Apparatus
The Catalyst
the recirculation pumps) and the incorporation of watercooled ammonia condenser and condensate receiver into one singlc apparatus. The two S t o n onits working under this system a t Terni are very simple and occupy the small space of 12 X 30 feet, for the two compressors. Compressors a t Ostend have been modified in order to he able to receive their hydrogen as it comes directly under the pressure of 9 atmospheres from the Linde fractionating apparatus. This arrangement permits of a further energy saving amounting to one-third of the compression energy, so that the power requiremcnt is reduced from 1.22 to 0.82 kilowatt-hour per kilogram of ammonia in the 6-ton units, and to 0.76 kilowatt-hour in the 20-ton units, the process becoming thus from day to day more economical. It has been suggested that some further purification he applied to the compressed hydrogen before admitting i t into the synthesis tube; but apart from the fact that any serious attempt in this direction would complicatc the process, there would he little advantage in prolonging the activity of the catalyst, as its replacement is only a matter of a few hours. Besides, there is no object in prolonging the life of the catalyst beyond a year, as it is not advisable to let a catalysis tube work for more than a year without testing its mechanical condition. The most striking part of the Casale system is the catalyst tube, which is so arranged that the pressure tube, which is very similar to a big naval gun, is kept a t a temperature a t least 100’ C. below the temperature at which the alteration of the structure of the metal under the effect of hydrogen begins. The efficiency of the Casale design is shown by the fact that a catalyfiis tube, which has been in continuous action for six vears. was tested a t 1500 atmosDheres and withstood the test. ’ This remarkable result is achieved hy so directing the
The effortsof Doctor Casale have tended toward producing
a cheap and robust catalyst rather than a delicate, lowtemperature catalyst. The process of manufacture, which lras been patented, consists in burning steel turnings in oxygen, in the presence of certain activators which are introduced in thc form of certain very cheap materials. The temperature a t which the combustion in oxygen takes place is so high as to drive off any phosphorus or sulfur. When all the metal is burned and the oxide solidifies, the mass is allowed
One of the Air Compressom
to cool and is then broken up and crushed into pieces of a proper size. The catalyst is introduced in the form of an oxide, and is reduced in the catalyst tube hy means of hydrogen-nitrogen mixture. Catalyst which has lost some of its activity is regenerated in the same way. This catalyst, which is made of the cheap-
January, 1929
INDUSTRIAL AND ENGINEERINC CHEMISTRY
est mterials, enters into the cost price of the ammonia to the extent of a few Belgian centimes only per kilogram of ammonia. It would be wrong to think, howevcr, that this catalyst is not very active. It is a t least remarkably free from sulfur, analyses showing i t to contain only one-fourth as much as a reputed specially active catalyst used in a recent American process. The writer wishes here to pay a tribute to the originator of this simple and effective synthetic process, Luigi Casale, who, while a student a t the University of Turin, undertook the study of what was then, except for Doctor Haber's work, an entirely new subject, and carried thraugh, with the most meager financial support, to a successful conclusion its iodustrial application. The death of Doctor Casale, two years ago, was a great loss to the scientific and industrial world, for he no doubt would have made further important contributions to the advancement of knowledge in synthetic processes. Fortunately, however, his examplo inspired his disciples and callahorators, who now carry on his work so
29
is no concentrated carbon dioxide available in the coke-ovengas hydrogen process as there i s in the Haber-Basch process, so that a different source had to be used. Although this mlfate process produces sulfate a t a lower cost than from sulfuric acid, i t is only operating on a small scale a t Ostend notwithstanding that the U. C. R. Company is the largest sulfuric acid manufacturer in Belgium, because the primary material used is a valueless by-product of another industry of the U. C. B. Another new feature, also patented, is that the evaporation of the concentrated sulfate solution, which is the most expensive item in the gypsum process, has been replaced by a simpler precipitation process. A serics of other developments are under experimental investigation a t the Ostend plant, which is used as a pilot plant by the U. C. B. Company in its researches upon the new synthetic industries. Conclusion
To summarize, a new approach t o the synthetic ammonia manufacture problem has been opened up that in practice has proved remarkably satisfactory and cheap. Its develop ment ha5 in turn made possible a series of derived processes, mainly concerned with the transformation of the various residual gaseous by-products of cokeoven gas. Among the several new products which have been sought are those of the fatty alcohols and some of their derivatives.
Twelfth Chemical Industries Exposition The Twelfth Exposition of Chemical Industries, which will he held a t the Grand Central Palace, New York City, the week of May 6 , 1929, draws together chemists, engineers, manufacturers, and others interested from forty industries which arc dependent in their operations upon a chemical change in the nature of the material or are under chemical control. A rcceiit special study of the following of these industries
Part Of Sulfate Plant
that progress Ius not stopped at his death, as is shown by the numerous improvenionts which have since been made and the synthetic. methanol process now working a t Terni with a production of 5 tons per day.
Cement arid lime Pine chenlicais Ceramics, day, glass Food products Coal-tar products Glue, gelatine Coal by-products X e a v y chemicals Electrochemical pmdiicts Leathor B ~ p l o s i v e s ~ e l l ~ l ~ s eOils (vegetable and aiiimal) Fe*tili'ers Paint snd varnish
I'etioieurn refining Pulp and paper Rubber Scrap SUE-
Wood chemicals
indicates that they alone produce 17 per cent of all the products manufactured in the United States and employ 22.6 per cent of the total capital invested in manufacturing. Today these plants are producing 8 billion dollars' worth of finished prodThe Sulfate Plant ucts-an increase in production during the past thirty years of It is an easy matter to transform liquid ammonia into 6 billion dollars. The invested capital of this group is above sulfate by means of sulfuric acid, but it is difficult to do so 7 billion dollars and the average plant investment represents while the average investment for all other industries in a simple way without using complicated devices such as $843,000, is $l83,WO per plant. The chemical process industries are mechanical stirrers in the saturators, or blowing air or gas among the highest capitalized; a rayon plant-in a new indusinto them for different purposes such as has been successfully try-requires a minimum of $3,500,000; an average petroleum developed by the Is. C. B. Company. Some difficulties refinery has over 3 million dollars invested, and cement plants over 2 million. with this method were a t first encountered, owing partly average These industries employed a million persons, ranking third in to the high temperature maintained in the saturators and any group of American industries. They have over 2 million to the large production per unit capacity of the apparatus. electric motors installed, consuming over 6 million kilowattBut these troubles were overcome in a short time, and now hours of electrical energy. In this use of electricity they rank second in America, which is the greatest producer and user of 100 tons of sulfate are made per day in one small saturator electricity in the world. These industries consume over 9X of about 9 feet in diameter. This sulfate is centifuged and million tons OF coal per year, occupying first place in this respect. neutralized with ammonia in a special way, and is then dried In a study of the chemical industry made during June last. a in a rotary drier by means of combustion gases from a coke New York banking wmpany found that the American chemical industry. favored by its large domestic market, has grown to be furnace. the largest in the worid, with an annual production of 82,278,This method of manufacturing sulfate will soon be replaced 000,000. This growth is attributed to the more numerous a t Ostend by a new process similar to the gypsum process. and prosperous population and the rapid expansion of industiics The system, which is covered by a series of U. C. B. patents. in which chemicals are used. The United States is not indei s special in that i t uses an ammonium carbonate solution pendent of the products of other countries and while it is the second largest exporter in the world, it is also the largest importer. made in a new way from combustion gases of a central pul- In 1927 chemicals imported were valued at $198,903,000 while verized coal boiler plant. It will be remembered that there exports amounted to $184,133,000.
