Return to B.I.O.S. Table of Contents
B.I.O.S. Final Report No. 1142
Item No. 30
The Wintershall-Schmaldfeldt Process For The Manufacture Of Synthesis Gas At Lützkendorf
This report is issued with the warning that, if the subject matter should be protected by British Patents or Patent applications, this publication cannot be held to give any protection against action for infringement.
British Intelligence Objectives Sub-Committee
London-H.M. Stationery Office
The Wintershall-Schmalfeldt Process For The Manufacture of Synthesis Gas At Lützkendorf.
Reported by R.J. Morley. on behalf of The Ministry of Fuel and Power.
BIOS. Item No.30. Fuels and Lubricants.
BIOS. Target No: 30/4.12.
British Intelligence Objectives Sub-Committee,
32 Bryanston Square,
The WinterShall - Schmelfeldt Process For The Manufacture Of Synthesis Gas at Lützkendorf
A description is given of the process, with sketches and flowsheets. The process consists essentially of gasification of dry brown coal dust in the entrained state; the heat of reaction is added by burning producer gas, storing heat in a regenerator and then abstracting it by recirculating saturated synthesis gas, with also the addition of oxygen at various points; sensible heat in the final gases is used to dry and disintegrate the raw brown coal. The producer gas is made by gasifying dry brown coal dust in the entrained state, using air with a little steam. The processes are now technically established, but dust losses are high and synthesis gas contains substances deleterious to Fischer-Tropsch catalyst, although not to hydrogenation of oil and tar; the plant is very large for the output and capital costs are high. In its present state of development the process appears to be less economic than the alternative Winkler process, operating under similar circumstances. The producers are particularly inefficient. Because of its dependence on the availability of a very cheap and active fuel, the process appears to have little application to British conditions.
Members of the Oil Mission attached to the C.I.O.S. Section of SHAEF G.2 Section, under the auspicious of the Ministry of Fuel and Power, visited the Wintershall A.G. plant (Target No.30/4.12) at Lützdendorf, Krumpa , near Mücheln on May 9th and 11th, 1945. A short description of this process and of other parts of the factory has already been given in Ref.1, but the present report brings together all the information on the synthesis gas process, obtained from the visit and from subsequent study of captured document. Where statements in this report are at variance with those in Ref.1, the present report should be assumed to be more correct.
The following visited the plant on May 9th, 1945: Dr H. Hollings and Mr. G.U. Hopton-(Gas Light and Coke Company) , Mr T.F. Hurley (Fuel Research Station), Dr. L.L. Newman (U.S. Bureau of Mines), Dr. R.J. Morley (I.C.I. Billingham Divisions)
The following visited the plant on May 11th, 1945: Dr. A.J.V. Underwood (U.K.), Mr. J.F. Ellis and Dr. R.J. Morley (I.C.I., Billingham Division)
The following members of the staff were interrogated: Dr.Schneeberger : Director, Dr.Schneider: Director and power plant specialist, Dipl. Ing. Dassow: Manager of gas plant (H of Dassau), Dipl. Ing. Schültz: Manager of Fischer Tropsch plant.
All staff interviewed were very co-operative, but the Mission received the impression that the technical standards of the staff were not as high as those of other plants visited. The plant was shut-down at the time of the visit, and had been out of commission since the last air-raids.
General Information On The Factory
The plant was built in 1938 and was intended originally to operate for Fischer-Tropsch synthesis only. Later a hydrogenation plant was added, hydrogenating natural oil and bituminous coal tar. The plant was unique in that synthesis gas and hydrogen were produced by the Wintershall-Schmalfeldt process, which was essentially a gasification of brown coal in the entrained state.
The factory as a whole had a bad reputation in Germany, presumably due to the extreme difficulties encountered during the first 18 months of running. The state of the factory became so bad that in October 1939 I.G. experts from Leuna (7 miles away) were brought in to help put matters right; the periodic I.G. reports (Ref.2) on this work gives a very illuminating account of the factory at this time and certain points are worth noting, in order to appreciate that the difficulties were by no means all the result of the use of novel gasification process. Apparently the whole factory was very badly designed and engineered, was poorly staffed and weakly managed. There were poor design and poor workmanship by several contractors, notably by Koppers in the regenerators, Bamag in the alkacid plants and others in the La Mont boilers. Originally the civil engineering, notably drainage and surfacing, was very poor and there was no provision at all for protecting any plant against frost. The various stages of the factory were often quite out of step, so that spares were either too numerous or else non-existent; inadequate provision was made to keep a supply of steam to the factory, when synthesis gas and producers were shut down and not producing any dry dust or any waste to heat steam. Moreover since the novel gasification processes required more than the usual amount of modifications in the first few months of running, they caused a heavy drain on what skilled labour there was, and so the factory could not be got out of the rut. There was a severe shortage of skilled engineers and foremen and the labour was mainly inexperienced, and as can be imagined discipline and morale were poor. In fact the whole reads as a classic example of how not to design, build and run a factory:
At one time I.G. had 7 staff men, 25 foremen, 140 tradesmen and 15 process operators at Lützkendorf, trying to bring order into the chaos. Gradually the position improved and the worst was over by the spring of 1940, although even in 1945 C.I.O.S. investigators still got the impression that technically the standards of the factory were not as high as those of other plants visited.
With the above as a background it will be realised that the Schmalfeldt process has not had a happy infancy, and it is quite likely that in better circumstances and with better technical attention the process would have been better developed today.
Whilst the factory as a whole was very badly damaged and seemed unlikely to run again, the gas plant itself was not so badly damaged and could readily be inspected.
Standards of Measurement
All gas quantities are measured at 0º C and 760 mms Hg, the standard figures used at Lützkendorf, unless otherwise stated. Weights expressed as t signify metric tons or tonnes. Heat units are generally expressed as T cals (=1,000 kg. cals).
General Information On The Process
The designer, Dr. H Schmalfeldt, was once a director of Wintershall and lived at Lützkendorf from the start-up in 1938 until 1940; he is now believed to be living in the Kassel area. Previous to this plant there had been an experimental unit at Ruhland-Schwarz-heide (Ref.6), which is probably the same plant referred to in Ref.1 as "Rühlen" , probably a mistaken spelling of the place mentioned during interrogation. As far as is known there is no other large-scale installation.
The principles of the gasification process were as follows.
The heat of reaction was originally supplied solely by burning producer gas in one of two regenerators, used alternately, the heat being stored in chequer brick until given up to a mixture of recycled synthesis gas and steam; however since the increased gas requirements, following the start-up of the hydrogenation plant, heat had also been supplied by burning oxygen in the recycled gases. Dried brown coal dust was fed into the recycled gas and steam, as it passed up and down two large generators in series. The sensible heat in the final gases was used to dry and disintegrate raw brown coal, fed into the gas stream as lumps; the sudden heating caused decrepitation and the particles became entrained. The dust was separated from the gas stream and partly used for making producer gas, the rest being fed to the generators. Part of the synthesis gas was led off, cooled and scrubbed and represented the make, whilst the remainder was recirculated, primarily to act as a carrier of heat from the regenerator to the generators.
Producer gas was also made in a separate installation by gasification of entrained brown coal dust but there was no recycling of gas of use of oxygen; air, steam and dry coal dust were fed into a tower and the sensible heat in the exit gases abstracted by waste heat boilers. The bulk of the producer gas was used for heating the regenerators of the synthesis gas units.
There were four synthesis gas units and five producers, with usually one of each out of action. The plant was very spacious, considering its capacity, although it was made up of relatively simple pieces of equipment. The four synthesis gas units, with regenerators, stocks and washers occupied an area of about 100 m x 30 m, with the vessels about 20-24 m. high. The gas boosting house was outside this area, whilst the producers occupied a separate side.
The original designed output per synthesis gas unit was 20,000 h3/hour, but the maximum obtainable without using oxygen was normally only 15,000 h3/hr. When using oxygen the maximum output was 30,000 h3/hour, although more usually 20 to 25,000 h3/hour were produced.
The maximum output of a producer was about 30,000-35,000 h3/hour.
Main Features Of The Performance
The two main features which caused operating difficulties were the relatively high content of organic sulphur ring-compounds and gum-forming compounds and the difficulty of removing dust from its suspension in synthesis gas and producer gas. Troubles in the Fischer Tropsch plant due to sulphur ring-compounds and gum-forming compounds arose presumably from the fact that about 40% of the gases evolved during drying and carbonisation was in effect not recirculated, since about 40% of the synthesis gas mixture was led off from the system before the regenerators, where presumably a good deal of cracking of such compounds occurred. Sulphur ring-compounds are difficult to remove by catalytic action, but is should be noted that for the last few weeks of operation active carbon had been used in the purification train to remove these undesirable impurities, and this it was claimed had produced a marked improvement in the Fischer Tropsch plant. Again it should be noted that these impurities were not harmful when the gas was to be used for hydrogenation.
The difficulty of removing dust from its suspension in synthesis gas and producer gas caused very poor carbon efficiencies; indeed at least one-third of the carbon in the raw brown coal left the system as a sludge, formed in the gas washers. This inefficiency however was not quite so serious as it sounds, owing to the cheapness of brown coal, which at Lützkendorf cost only 1.80 to 2.20 RM/t; thus the carbon loss in itself increased the cost of synthesis gas directly by only above 2 RM/1,000 h3. However dust also caused more of less serious difficulties due to fouling of washers, blowers, disintegrators and waste heat boilers.
The three main advantages claimed for this process over other gasification processes based on brown coal, viz. the direct use of raw wet brown coal, the suitability of dusty fuels without briquetting and the gasification of the tar formed, were closely connected with the troubles due to impurities in the synthesis gas. Gasification with entrainment did not appear to have any great advantages in itself; reaction spaces were very large e.g. in comparison with Winkler generators, and difficulties with dust would be very serious with a dearer source of fuel.
Nevertheless it cannot be denied that the Schmalfeldt process as developed at Lützkendorf was a workable process for the manufacture of synthesis gas and hydrogen from cheap brown coal and at any rate may be regarded as especially useful in extending knowledge of what can and what cannot be done in gasification processes.
Description Of The Plant
Figures 1A and 1B show a rough layout and arrangement of a synthesis gas unit, whilst Figure 2 is a diagrammatic representation of the flow, etc. Figure 3 is a photograph of one end of the synthesis gas plant, taken during the visit. Figure 5 is a rough layout of a producer gas unit and Figure 6 is a diagrammatic representation of the flow, etc.
The individual items of plant will now be considered in turn.
Brown coal was obtained from a neighbouring open mine. It was elevated and crushed in hammer mills to all below 20 mms. Everything passing the mill dropped into a raw coal bunker, from whence it was fed by Redler screw conveyors (one working, one spare) through star feeders, dropping into the side of the bottom of the gas drier. The contents of the bunker were kept under nitrogen pressure.
Typical analyses of the raw coal were:
|as received||dry basis|
|Cal. val. k.cals/kg(net)||2300-2400||Ash||12%|
The tar content was rather low for brown coal, and so the coal was not well suited to the normal treatment, i.e. carbonisation to recover tar for hydrogenation, with use of the coke in Winkler generators and boilers.
This was merely a vertical brick-lined chimney, 1.2 m. I.D. and 22 m. total height; the height from coal inlet to gas exit was 15 m. The crushed raw coal dropped straight into the drier without any conveying gas; two inlets at an angle of 30º to the vertical were available, one working and one spare: each was steam heated to prevent sticking of the coal, but were brick-lined near the drier itself. The coal met the upward rising synthesis gas and steam, then at about 1,000º C; the sudden heating caused the formation of a steam pressure inside each particle, resulting in decrepitation, and the fine coal became entrained with the gas.
No special mixing of coal and gas was attempted; coal was merely fed through the sides of the drier and the high turbulence of the gas, which had just passed round a bend, together with the explosion of each particle were sufficient to give good mixing.
The final temperature at the top of the drier was 200º to 300ºC. At full output the dry gas rate entering the drier was about 90,000 h3/hour, so that taking account of the steam present, calculated in Appendix 2 as 91,500 h3/hour, the average actual velocity of gas through the drier was about 100 m/sec. (agreeing with a statement of Dassow) and the time of drying 0.15 secs. Normally however the velocity was nearer 70 m/sec. and the time of drying 0.20 sec.
The synthesis gas and dust mixture leaving the top of the drier at 200º to 300º C passed through a rough classifier, where any large pieces of coal were separated: these were sent to a separate hammer mill, and returned to the drier; (presumably they were not returned to the raw coal hammer mills, because the mixing of hot dry coal and raw brown coal would have caused trouble).
The mixture then passed through a simple cyclone to effect the main separation of dust. According to Appendix 3 the dust content of the wet gas leaving the drier was 150 to 200 g/h3, and that of the wet gas leaving the cyclone was about 30 to 40 g/h3, so that the separator and cyclone had a combined efficiency of about 85 to 90%. Even so this inefficiency represented a serious carbon loss from the system, since all such dust had to be removed subsequently in washers and could not be recovered from the resultant slurry. In this way some 10 to 13% of the carbon in the raw coal was lost (See Appendices 2 and 3).
In 1942 it had been planned to install multicyclones to reduce the dust content of the gas to 6 g/h3 but this project had never been undertaken. Dust recovered from the multicyclones, of course, could have been returned for gasification.
The dry coal dust contained some 18-20% ash and 60% carbon. The ash content was higher than that of raw brown coal without its water, because the dry coal dust contained nearly all the ash remaining from the gasification of dry coal dust fed to the generators; the carbon content however was the same as that of raw brown coal without its water, presumably because some oxygen and hydrogen were given up in the drier. There was of course a large ash purge from the generator system in the dry coal dust fed to the producers.
No figures were available for the grading of the dust.
Dry coal dust fell from the cyclone into a bunker, at the bottom of which were star-feeders, passing dust into pneumatic lines, working on the ejector principle. Nitrogen was used for conveying dust, at a concentration of about 3 kg/h3, to the synthesis gas units, since the conveying gas actually entered the generators along with the dust: this synthesis gas, amounting to about 10% of the make, in effect recirculated through the generator system, but not through the regenerators, the exact proportion depending on output.
These were brick-lined towers, 7.1 m. O.D, 5.5 m, I.D. and 24 m high. Two were provided for each synthesis gas unit, one being heated up by producer gas whilst the other was giving up its heat to the recycled gases; they were changed over automatically every eleven minutes with a purge to atmosphere of 1 minute after heating up.
The design of these regenerators was obviously based on that of air preheaters used for blast furnaces. They were filled with checquer, made of high quality brick, such as sillimanite or silica, to a depth of 17-18 m. Two designs of chequer had been tried; one a Brassert type and the other, which had given better results, a type made by Messrs Didier, of Berlin. The Didier type, known as "Schieffer-Strack" , consisted of hexagonal blocks, about 12 3/4" across and 7" deep, each having about twenty 11/4" diameter vertical holes, spaced at 2 1/2" centres. A plan view is shown in Figure 4A. The surface area of the channels from these dimensions amounts to about 11,500 M2/regenerator, a figure somewhat in excess of 8,540 M2, as quoted in Ref. 2 (Item 83) for 1940. These bricks were carefully stacked and lined up, so that the holes came in line. The lining bricks were well-finished, so as to permit the minimum amount of cement to be used. It was stated that there was no trouble due to dust deposition in the chequers and that the bricks stood up very well to the conditions. No figures were available for the dust content of clean producer gas or recirculated synthesis gas.
Preheated air and producer gas were fed into the top of the regenerators through ring mains. Flue gases left at the bottom of the regenerators and entered an underground line, common to the two regenerators, leading to a stack; no waste heat boiler was installed, but one was projected. Recycled synthesis gas, saturated with water vapour at 82ºC, entered the bottom of the regenerator and left it through the cupola at the top, on its way to the first generator.
The maximum brickwork temperature was 1450ºC at the top. The average exit flue gas temperature was 450ºC. The recycled gas entered at 82ºC and left at about 1,300ºC.
Double isolation valves were used on each flue gas and recycled gas line, with the portion between valves automatically vented to atmosphere, when the valves were shut. This was a safety measure, but usually it was found that the valves were a good enough seal. The isolation valves on the air and producer gas were not seen, but these may also have been double.
In line diagrams in Refs. 2&3 valves are shown on the line from each regenerator to the first generator, but by an oversight details were not obtained during interrogation. Working at 1,300º C the duty would appear to be arduous but nowhere is there any mention of difficulties encountered.
Originally there had been some trouble with erosion at the top of the cupola but this had been cured by constructional changes aimed at making the linear velocities of the gas in the cupola and off-take pipe more nearly equal.
This was a brick-lined vessel, 5.5 m I.D. and 24 m high. The special design of the cupola, with its false roof, is shown in Figure 4B. Dry coal dust, conveyed by synthesis gas at a pressure of 2.5 ats, was fed down through a passage in the centre of the cupola; in this way it was kept out of contact with the walls of the vessel, where it would slag up. The hot recycled gas and steam mixture from the regenerators was fed through ports in the false roof of the cupola.
Oxygen, saturated at 82ºC, was introduced from ring mains through ports at points near the top and middle of the generator; near the bottom, steam as well as oxygen was admitted, to avoid slagging The temperature fell from1,300º C at the top to 1,000ºC at the bottom.
This was also a brick-lined vessel, 24m high and having an I.D. of about 5.5m. On three units it was divided internally by a vertical wall, but on the fourth unit it had no such division wall. The division wall was shaped as shown in Figure 1, being curved towards the inlet and possibly hotter channel, a device presumably connected with consideration of strength and expansion; the two portions were of approximately equal area. In the unit with no division wall the gas was brought down to the bottom of the drier by an external pipe.
One-third of the total oxygen was added near the bottom inlet of the generator, whilst steam was added as required at various points in order to control temperatures and avoid slagging. The aim was to maintain temperatures throughout at about 1,000ºC.
In the opinion of Dassow the central division wall was not necessary; the unit without it worked just as well. This is understandable, since the division wall gave only higher turbulence and increased linear velocities but no greater contact time. He also said the second generator was not really required if the output was not too high, but it was necessary for the higher outputs using oxygen. This indicated that gasification was substantially complete after the second generator.
From the data given it can be deduced that the approximate gasification time (time of contact in coal in generators) was 4.5 secs at 30,000 h3/hour make and 6.0 secs at 20,000 h3/hour.
The recycled gas had to be cooled and dedusted in order to avoid trouble with the blower and to avoid slagging troubles in the regenerator. Each unit had two washers, one working on recirculated gas (about 50-60,000 h3/hour). The washers were the same size and took the same water rate, despite the difference in gas rate. A typical arrangement is shown in Figure 4C. Each was 5.5 to 6.0 m I.D. by 22 m total height. Each was divided into two sections with its own water circulating system, each section being packed with six trays of stacked 80 mm. spiral Raschig rings. About 750 h3/hour of water were circulated round the bottom section, removing the bulk of the dust; a purge was maintained on this system to keep the solids concentration at about 80-100 g/1. Similarly 750 h3/hour of water were circulated through the top section but also through a water cooling tower.
The object of dividing into two the cooling and dedusting in each tower was presumably to remove the bulk of the dust in the lower potion, sending the purge to waste, whilst removing the bulk of the heat in the top portion, where the relatively clean conditions enabled the water to be passed through a water cooling tower and re-used.
So long as the water quantities were maintained no trouble was experienced with chokage of the packing with dust. However at each six monthly shut-down the rings were removed and washed. This job for eight washers on the synthesis gas units and five on the producers kept fourteen day men continuously employed.
Synthesis gas to be recirculated left the washer saturated at a temperature approaching 82ºC. After boosting to 0.15 ats.g. an amount of live steam roughly equal to the vapour in the saturated gas was added, and the mixture passed to the regenerators.
Blower and Control House
This building at the end of each housed the automatic control gear for the valve change over and also blowers for boosting synthesis gas, recirculation gas, oxygen, nitrogen, etc.
The gas pressure at the bottom of the regenerator was 0.150 ats.g. Since the pressure at the point of entry of coal into the drier had to be kept very close to atmospheric, to prevent gas leaking back up the coal food line, the drier and washers had to be run at a pressure below atmosphere; the pressure at the top of the washers was about -0.020 ats.g. but this varied with the cleanliness of the washers. The pressure in the drier was controlled by a damper, located between the second generator and the drier.
Control of H2/CO Ratio
This was controlled at 2.0 by adjustments of the steam to gas ration. Calculations in Appendix 2 indicate that the steam/gas ratio was about 1.0 at the exit of the drier.
Synthesis Gas Analysis
This was as follows:
|With Oxygen||Without Oxygen|
Dust Losses and Carbon Balances
As in the Winkler process the high dust content of the make gases introduced large inefficiencies and some technical and mechanical difficulties. Dust losses were a very undesirable feature of the process, and no doubt an obstacle in the way of new installations.
Carbon balances have been calculated in Appendices 1 and 3 for three periods - early in 1942, whole of 1943 and whole of 1944. Taking the synthesis gas units and producers as a whole the figures for early 1942, which were used in a Wintershall report (Ref. 4) and which Dassow thought were rather optimistic, show that 26% of the carbon did not appear in synthesis gas or producer gas, whilst the figures for the whole of April 1944 show that as much as 43.5% of the carbon did not appear in synthesis or producer gas. The figures for the whole of 1943 are probably the most reliable and taking into account other possible losses of carbon we can say that at least one-third of the carbon was lost as dust in synthesis gas and producer gas.
It was stated that the washers reduced the dust content to 30-40 mg/h3 and that Theison disintegrators reduced it further to 23-25 mg/h3. If these figures are true they represent a poor performance for Theisen disintegrator. There was a further water wash before the H2S removal plant. (Alkacid), which apparently experienced little difficulty due to dust.
Dassow stated that originally there were some difficulties, mentioned in various places above, but that now these had been largely overcome. Ash and slag gradually accumulated in the first generator and when oxygen was used the generator had to be cleaned out every six months, but if no oxygen were used the plant could be run longer. During these shut-downs other maintenance work was carried out, such as cleaning rings in the washers, repairs to brickwork. As a rule a shut-down would last 42 days-14 days to cool down, 14 days to carry out repairs and 14 days to heat up. The ash and slag mixture was white and very hard and had to be chiseled out.
Producer Gas Plant (See Figures 5 and 6)
There were five producer units, each making a maximum of 30,000 to 35,000 h3/hour producer gas, although normally making loss. Again gasification of the dust was carried out with entrainment. Each unit consisted of the producer proper, followed by waste heat boilers, multicyclones, wash tower and Theisen disintegrator.
The producer was a brick-lined tower, 5 m I.D. and 24 m high, with an internal division wall, very similar to that of the second generators. Dry coal dust was blown into the bottom of the tower, along with steam and air, the mixture passing up one side and down the other. The maximum temperature reached was 1,000ºC. The gasification time (time of contact) was about 11 secs when making 30,000 h3/hour.
The dry coal dust (18% ash, 60%C) was conveyed from the synthesis gas units to a bunker by means of nitrogen, but it was conveyed from the bunker to the products by means of air. About 15 t/hour dust was conveyed by 1,000 h3/hour air through three lines, each 125 mms I.D; assuming atmospheric pressure this corresponded with a velocity of 7.5 m/ sec. and a dust content of 15 kg/h3.
Gases from the producer through a horizontal and a vertical waste heat boiler in series, reducing the temperature to 250ºC, and raising about O.5 t steam/1,000 h3 producer gas to a pressure of 10 ats.g. Some erosion at the inlet of the first boiler was experienced and this limited the running time of a unite to six months between overhauls; some slagging up also occurred at this point and at one time consideration was given to injecting water just in front.
Multicyclones followed the boilers; the recovered dust was blown back by compressed air into the bottom of the producer, although at times some was sent to the main boilers.
The washer tower was 5 to 5.5 m. I.D. and 22 m. high, filled with ordinary 2" Raschig rings; in design it was very similar to washers on the synthesis gas units, but the water rate was only 500h3/hour. Dassow said the tower was too small and passed too much dust. A Theisen disintegrator removed most of the remaining dust, and after passing through a spray arrestor, the gas was boosted into the factory fuel gas system. Here it was mixed with tail gases from the Fischer-Tropsch and hydrogenation plants. The chief consumers of fuel gas were the regenerators of the synthesis gas units: in 1943 those consumed 73% of the T.cals in fuel gas, an amount only slightly greater than the total amount of producer gas made. So as a first approximation it can be considered that in effect the producers were run solely for the benefit of the synthesis gas units and hence all the coal fed to drivers of the synthesis gas and all the dust from the coal drying plant, etc., fed to the producers was in effect eventually consumed in making synthesis gas, even though over half the dry coal was consumed through the medium of the producers.
The pressure was 0.150 ats. g. at the bottom of the producer and 0.080 ats.g. before the Theisen disintegrator. The analysis of producer gas was-
The above analysis yields a calorific value of 1.070 T.cals/h3 (not), but the average value for 1943 was in the official records as 1.046 T.cals/h3 (net).
From figures given in Appendix 2 it appears that 35 to 45% of the carbon entering a producer was lost as dust to the washers. Apart from showing the poor efficiency of the multicyclones this figure shows that gasification within the producer was by no means complete. The dust content of gas entering the multicyclones was as high as 200 to 250 g/h3 wet gas, and 150 to 200 g/h3 leaving them. Presumably the multicyclones did not remove more dust, because the dust at this stage was very fine, much finer than the dry coal dust leaving the drier of the synthesis gas units. (On the Winkler generators at Zeitz the multicyclones reduced the dust concentration from 300 to 60 g/h3).
In Ref. 5 Koppers refer to the Schmalfeldt producers as being little more than distillation and cracking units, with practically no gasification of solid carbon; calculations in Appendix 2 confirm that this is no exaggeration. The cause of this is uncertain, but may lie in the fact that all free oxygen has disappeared by the time the coal particles have been stripped of their volatile matter, and the subsequent reactions between H2O, CO2 and the coke particles are comparatively slow.
Coal Drying Plant
Three of the boilers installed in the power plant could use only dry coal, and a coal drying plant, fired by coal and producer gas, was installed to provide this dry coal. The plant was also used to a certain extent to supply dry coal to the producers, when for any reason the driers of the synthesis gas units were unable to dry enough coal.
Purchased Coal Dust
A small quantity of dried brown coal dust, containing 12% H2O and 53% C, was also purchased and used partly on the boilers and partly on the producers.
Dassow gave the following figures for labour.
|Coal Transport and preparation||70|
|Fitters and labourers||80|
The shift men worked 56 hours/week and the day men 60 hours/week. Most of the men were German workers.
Sulphur and Fischer-Tropsch Difficulties
Schmalfeldt gas caused considerable difficulties in the Fischer Tropsch process, which used a cobalt-kieselguhr catalyst, due to the relatively high content of thiophene and other S-ring compounds, as well as gum forming compounds. This, in the opinion of Dassow, was due to drying and gasification being done in the same apparatus: he reckoned there would be no difficulty if the drying were done first, say in a Büttner drier, separately from gasification. S-ring and gum-forming compounds would tend to be destroyed in passing through the regenerator at 1,350ºC and to a lesser extent in passing through the generators at 1,000ºC. Since only 60% of the synthesis gas was recirculated and 40% was bled off as make immediately after the drier, therefore 40% of the gaseous products of drying did not pass through either regenerator or generators, and 40% of the carbonisation products did not pass through the regenerator, although all carbonisation products passed through a zone at 1,000ºC. It is not certain whether the deleterious impurities were evolved in the drying or in the carbonisation, but in either case we should expect an appreciable fraction of them to escape with the make gas and that such gas would contain more of them than would gas derived from a gasification system, e.g. Winkler generators, where drying and carbonisation were carried out in a separate process.
Schültz said that whereas the organic sulphur in coke water gas was about 90% CS2 and COS and 5-10% thiopene, organic sulphur in Schmalfeldt gas contained about 20% thiopene.
The raw synthesis gas contained about 1 to 2% H2S. The only purification steps originally installed were to remove the bulk of the H2S in an Alkacid plant, followed by Luxmasse (30% soda, running at 160º to 280º C). The latter however passed too much organic sulphur, especially ring-compounds, and gum-forming compounds, so in 1940 an oil-scrubber was installed after the Alkacid plant. The oil-scrubber also enabled the 6 to 8 g/h3 benzol in the gas to be recovered. Later when gas outputs were increased the Alkacid plant was backed up by oxide boxes, running cold and using bog ore. At this stage therefore the purification train was: water wash, Alkacid, oxide boxes, oil wash and Luxmasse reduced the organic sulphur content from 40-100 g/100h3 to O.5 g/100 h3 at best, but usually only to 1.5 to 2.5 g/100 h3. This however was still not good enough and in 1944 an active carbon plant was installed before the Luxmasse; this greatly improved matters and reduced the organic sulphur content after the Luxmasse to 0.3 g/100 h3 or less. The plant was run with this improvement for only two months before bombing stopped operations, but during this period the Fischer-Tropsch plant showed greatly improved performance.
It should be noted however that there was no difficulty in using Schmalfeldt gas for hydrogenation, where sulphur was not a poison. Gas for this purpose was taken from after the oil wash.
Raw Materials and Services
In Appendix 2 flowsheets are deduced for synthesis gas and producer gas. From this and other data derived from the reports for April 1944 the following have been drawn up, for a generator output of 20-25,000 h3/hour synthesis gas.
(a) Synthesis gas, including producer gas. It is assumed that all the fuel gas to the regenerators was producer gas; if this were actually the case then with the achieved dust losses the synthesis gas units could not dry enough coal, so that separate coal drying plants would have to be used. However in any process using synthesis gas there will be a certain amount of by-product gases, which could be used as fuel gas; at least 300 T.cals/1,000 h3 synthesis gas would be available because of the 3.5% CH4 in synthesis gas. If such fuel gas is valued at producer gas value then it is permissible to calculate as though all the fuel gas to the regenerators was producer gas.
(b) Synthesis gas only, i.e. plant between regenerators and second generator (inclusive)
(c) Producer gas only
(d) Without use of oxygen When not using oxygen it was said that the producer gas requirements were much the same in total quantity, but amounted to about 2,300 T. cals/1000 synthesis gas made, because they were spread over a smaller quantity of gas; the complete combustion of 100 h3 O2 to CO2 would be the equivalent of about 500 T. cals producer gas.
In addition about 44 kwh/1,000 synthesis gas was used for boosting synthesis gas to the purification plants and for boosting producer gas from the producers to the various points of usage.
Synthesis gas, produced in the manner described above, was used directly for the Fischer-Tropsch plant, after suitable purification. Hydrogen for the hydrogenation plant was produced from the same gas but taken off as a separate stream after the oil scrubbing. It was compressed to 8 ats. and the Co converted with steam to H2 and CO2. CO2 was removed by water washing at 8 ats. and CO removed by copper liquor scrubbing at 200 ats. The final pressure was 700 ats and the final gas composition 91-92% H2, 4% N2, 3.5% CH4.
Comparison of Schmalfeldt Process With Winkler Generators
It is of interest to compare the Schmalfeldt process with Winkler generators, since they are to a large extent alternative processes for manufacturing synthesis gas from brown coal.
In a Winkler generator all the heat of reaction for gasification is developed in situ by the combustion of oxygen, apart from a little introduced as superheat in steam. In the Schmalfeldt process, as originally designed and run, no oxygen was used but all the heat of reaction for gasification was introduced by combustion of air, via the formation and subsequent combustion of producer gas; however in the Schmalfeldt process as more recently run some of the heat of reaction was also added by the direct combustion of oxygen.
To simplify the comparison it is assumed that Winkler generators are gasifying dry brown coal, instead of the more usual grude or brown coal coke. Then in both processes no tar is recovered but is cracked during the gasification process.
In the following table performance figures for Winklers are derived from Ref.7, based on Leuna experience. Unit costs are typical of those met with in the neighbourhood of brown coal mines, and items marked thus (x) include their own capital charges. Both types of plants are assumed to be on reasonably high output.
Whilst the figures in the above table must be treated with a certain amount of reserve, owing to uncertainties inherent in estimates of this nature, certain features become clear. Even with fuel as cheap as 2 RM/t the Schmalfeldt process is somewhat more expensive in running cost and appreciably more in capital cost. Despite its poor carbon efficiency direct fuel costs are no more in the 'Schmalfeldt process, even after allowing for costs on that part of the plant used for drying; this is because inefficiency costs so little with coal at this price, whilst drying costs in special driers are not low. Oxygen is a formidable item in the cost of Winkler gas. The costs of power and cooling water in the Schmalfeldt process are high, mainly because about four times as much gas has to be boosted, cooled and dedusted. Labour and maintenance costs are high in the Schmalfeldt process, because of the large and scattered plant, varied operations and many points of control. This is also reflected in the capital costs, which are probably three times that of the Winkler plant proper, i.e. excluding coal drying and oxygen manufacture. This can readily be appreciated from the comparison given in the following table.
Ref.1 gives the cost of raw water gas in 1943 as 20.4 RM/1000 h3, to which has to be added 6.6 RM/1000 h3 for boosting and purification: the former figure is equivalent to 27.4 RM/1000 h3 H2+CO, although it is not clear whether capital charges are included.
Ref.1 gives the capital cost as 22,600,000 RM, specifically excluding railways, water supply & power plant; it probably excludes roads, drains, etc. but may include the oxygen plant.
Additional costs, which should be debited to the Schmalfeldt process, but are not included above, have to cover (a) the cost of disposal of the large quantities of muddy effluent, (b) the cost of the higher inert content of water gas (CH4+N2 = 6%, as against 3% in Winkler gas) and (c) the expensive purification treatment if the gas is to be used for Fischer Tropsch synthesis.
As coal becomes more expensive so does the difference between the costs of the two processes widen. Thus with raw brown coal at 4 RM/t the difference would be increased by about 4 RM/1000 h3 H2+CO, allowing for the coal used in making steam, power, oxygen, etc.
Nevertheless remembering what was said on p.3, that the process has been developed at Lützkendorf in very disadvantageous circumstances, it is yet possible that the Schmalfeldt process could equal the Winkler process in cost whatever the cost of brown coal. Obvious improvements could be brought about (a) by making fuel gas by Winkler generators, running on air and the coarser fraction of the dried dust, replacing the Schmalfeldt producers, which are very inefficient and (b) by installing multicyclones on the synthesis gas units to reduce dust losses. The Schmalfeldt synthesis gas process could then be imagined as saving the cost of coal drying and much of the cost of oxygen as saving the cost of coal drying and much of the cost of oxygen manufacture in the Winkler process, by a process involving a high capital cost of the gasification plant proper and higher costs for labour, maintenance and cooling water.
It cannot safely be argued that because the Schmalfeldt plant at Lützkendorf was never repeated that therefore, in German opinion, there are more economic processes. Reasonable operating results were not obtained at Lützkendorf until 1942, thus doing something to eradicate earlier bad impressions, due to the initial disadvantageous circumstances. After that time now factories, requiring gasification of brown coal, ceased to be built. Moreover German economy was such as to put a premium on tar, so this would give a preference to a Winkler process, using brown coal coke after tar recovery.
It is clear however that Schalfeldt process can find no application under British conditions. We have no suitable fuels which are cheap enough to permit large dust losses or which cannot be gasified more economically in conventional ways.
Appendix 1 Carbon Balances
From the official Betriebsbericht, together with certain assumptions, the following balances have been drawn up.
The relatively low output and air raids no doubt account for the high production of fuel gas compared with April 1944 (see below) and hence to the considerable use of the coal drying plant.
Appendix 2 Material Balances
Sufficient information was obtained from the visit to enable a good deal of information on material balances, etc., to be deduced, but the various assumptions made in the following calculations should be borne in mind.
(a) Synthesis Gas
Consider a unit on normal output, and assume a recirculated synthesis gas rate of 35,000 h3/hour for a make of 25,000 h3/hour; about 2,500 h3/hour oxygen would be used and about 60.8 t/hour raw brown coal (based on 1943 data) or 29.1 t/hour of water-free brown coal.
Now raw coal, on a dry basis, contained 60% C, 4% H, 20% O, 3% S, 1% N and 12% ash. The analysis of dry coal dust leaving the drier was given as 60% C and 18% ash, but no figures were given for the other constituents.
The basic data are as follows:-
(1) O2 found as CO (35%) and CO2 (18%) in synthesis gas = 3,125 + 4,500 = 7,625 h3/hour
(2) O2 added as say 96% oxygen = 0.96 x 2500 = 2400 h3/hour
(3) H2 found as H2 (49.5%), CH4 (3.0%) and H2S (1.5%) in synthesis gas = 12,375 + 1,500 + 375= 14,250
(4) Carbon found as CO (25%), CO2 (18%) and CH4 (3.0%) in synthesis gas = 11,500 h3/hour = 6.15 t/hour
(5) H2 present in raw brown coal = 4% x 29.1 = 1.16 t/hr = 13,000 h3/hr
(6) O2 = 20% x 29.1=5.82 t/hr = 4,070 h3/hr
Since about one-third of the coal was completely gasified in the synthesis gas generators at least one-third of the H2 and O2 in the raw brown coal would appear in synthesis gas, even if no evolution of H2 and O2 occurred in the drier. In practice we should expect from the temperature conditions some evolution of CO2 & H2S in the drier. There is insufficient information available to estimate exactly this quantity, but it has been assumed that 40% of the oxygen in raw brown coal, excluding water, appeared as CO2 in synthesis gas. With this assumption a complete balance can be struck and it will be shown that a reasonable analysis of dry coal dust results.
Thus, by assumption, O2 obtained from coal = 0.40 x 5.82 t/hr = 2.35 t/hr = 1,630 h3/hr.
Hence, by O2 balance, O2 out = 7,625 = O2 in = 2400 + 1630 + 1/2 (steam decomposed) steam decomposed = 7,190 h3/hour = 5.78 T/hour.
Hence, by H2 balance, H2 out = 14, 250 = H2 in = 7190 + H2 obtained from coal H2 obtained from coal = 7,060 h3/hour = 0,63 t/hr.
Lastly it is assumed that two thirds of the sulphur in raw coal is volatile.
It is now possible, by algebraic methods and without further assumptions, to deduce the following flowsheet, calculated to show the required quantities of C, H, O and S gasified and also to show the same analyses for dry coal dust fed to the generator and leaving the drier respectively.
The truth of the assumption, that 40% of the O2 in raw brown coal appeared as CO2 in synthesis gas, can be tested only by seeing how the above flowsheet fits other known facts. It will be seen that the analysis of dry coal dust is very close to that given quite independently by Dassow: thus for C, 58.6% compares with 60%, and for ash, 18.0% compares with 18%. It should be noted that the flowsheet assumes complete gasification of C, H, and O in the dry coal dust fed to the generator. From the above balance the gases evolved in the drier, apart from water vapour, would consist of CO2 334 h3/hr. H2S 343 h3/hr. H2 3,700 h3/hr.
Since the calculation of these quantities involve the accumulated errors in the whole balance they are subject to considerable error. The amount of H2 evolved is impossibly high and the amount of CO2 evolved is low. The real cause of these errors lies either in the over-estimation of H and under estimation of O in the gas or the reverse in the coal. However these errors do not seriously affect any conclusions in this report.
The above flowsheet also confirms that about one-third of the dry coal dust was fed to the generators.
According to the official monthly report for April 1944 27,375 t steam were charged to the synthesis gas units, (definitely excluding the producers), for a make of 45,900,000 h3 synthesis gas. This is equivalent to 0,6 t. steam/1000 h3 synthesis gas. On the other hand it was stated quite frequently that recirculated synthesis gas was saturated at 82ºC when entering the regenerators at 0.15 ats g. i.e. contained 0.62 t water vapour/1000 h3 gas. Since 1.4 volumes of synthesis gas were recirculated for 1 volume made, the quantity of water vapour entering the regenerators was therefore 0.87 t steam/ 1000 h3 if synthesis gas made. The fact that only 0.6 t water vapour/1000 h3 gas. Since 1.4 volumes of synthesis gas were recirculated for 1 volume made, the quantity of water vapour entering the regenerators was therefore 0.87 t steam/1000 h3 gas. Since 1.4 volumes of synthesis gas were recirculated for 1 volume made, the quantity of water vapour entering the regenerators was therefore 0.87 t steam/ 1000 h3 synthesis gas made. The fact that only 0.6 t steam/1000 h3 synthesis gas made was charged against the whole synthesis gas plant means that recirculated synthesis gas leaving the washers must have been saturated at a temperature approaching 75ºC and containing about 0.59 t water vapour/1000 h3 synthesis gas made; for undoubtedly some of the 0.6 t steam was used in the form of oxygen saturated at 82ºC (about 0.06 t) and an unknown amount of steam added directly to the second gasifier; there were also no doubt miscellaneous usages of steam on auxiliary equipment. It is concluded that the total quantity of water vapour entering the generators was about 1.0 t steam/1000 h3 synthesis gas made, both in the form of water of saturation and in the form of direct addition of steam. Steam actually reacting with carbon was quite a small proportion of this, viz Q.23 t/1000 h3 synthesis gas made, according to the amount of steam decomposition deduced above.
The steam balance for 25,000 h3/hour synthesis gas made then becomes water vapour and direct steam added steam decomposed.
This mixture contains more than enough heat and water vapour to permit saturation of recirculated synthesis gas leaving the washer to a temperature of as high as 82ºC.
(b) Producer Gas Insufficient information is available to deduce an accurate flowsheet for the producers, but a rough flowsheet has been deduced as follows.
Consider a producer making 30,000 h3/hour producer gas, of composition 12% CO2, 0.3% O2, 16% CO, 16% H2, 2% CH4, 53.7% N2. The equivalent carbon content was 0.161 kg/h3 or 4.83 T/hour. The dry coal dust rate to the producer is not known accurately, but a figure of 0.5 t/1000 h3 producer gas was quoted by Dassow, which agrees with the figure of 15 t/hour also quoted by Dassow in connection with the rates through the pneumatic feed lines. With a carbon content of 58.6%, to be consistent with the analysis deduced above, this corresponds with a carbon feed of 8.79 - 4.83 = 3.96t or 45% of the feed; this is high, but later it will be that this was probably the case.
It will be remembered that it was suspected in the calculations on synthesis gas that less H2 was evolved in the drier than appeared from the calculations; if the H and O contents of dry coal dust were not 2.7% and 18% as deduced but 3.5% and 17.0% respectively, the hydrogen balance for the producers would be correct, and there would be a small positive steam decomposition. This could well be so, and tends to confirm the air rate and negligible steam decomposition deduced above. It is noteworthy that the bulk of the hydrogen appearing in producer gas may be regarded as derived from coal rather than from steam; moreover the C/H ration found in the gas is about 9, which is close the C/H ration found in brown coal tar. This bears out the contention of Koppers (see P.14) that the Schmalfeldt producers are little more than distillation and cracking units, with practically no gasification of solid carbon.
The only figure available for steam usage is 4,622 t for 47,950,000 h3 producer gas, charged in the monthly report for April 1944; this corresponds with 96.5 kg/1000 h3 producer gas or 2.9 t/hour for 30,000 h3/hour. Although small this is ample to permit the calculated very small steam decompositions.
Thus the steam content of wet producer gas was about 2.8 t/hour = 3,500 h3/hour, so that the wet producer gas rate was 33.500 h3/hour.
No reliable analyses of final dust were given us, but there is reason to believe that the figures given in Ref.4, of 50% C and 50% ash, were too optimistic and too low in carbon.
(c) Combined Material Balance
As a check on the above calculations it is interesting to use them to calculate the combined material balance for synthesis and producer gas and hence derive a figure for the dust losses on the synthesis gas units.
It will be remembered that the 1943 figures were used to deduce the quantity of raw brown coal fed to the synthesis gas driers. Thus for 330,774,000 h3 synthesis gas and 522,633,000 h3 producer gas.
The above of course assumes no carbon losses in the coal drying plant, or any form of gaseous losses, of which there are several potential sources.
(d) Fuel Gas Requirements
In the whole of 1943 the synthesis gas units required 1,780 T.cals fuel gas/1,000 h3 synthesis gas, and in April 1944 1,715 T.cals/1,000 h3; in each case the fuel gas consisted of a mixture of producer gas and tail gases from the Fischer-Tropsch and hydrogenation plants.
Appendix 3 Dust Losses
Ref. 4 is a note written at Lützkendorf in early 1942 when a project requiring larger gas quantities was being considered; the note was intended to show what alterations would be necessary to the washer water system and for this purpose gave figures for the dust emission from the synthesis and producer gas plants. Dassow said these figures were rather optimistic and calculations based on other data confirm this.
An output of 90,000 h3/hour synthesis gas on 3 units and of 120,000 h3/hour producer gas was considered.
(a) Synthesis Gas Each unit was taken as losing 3.5 t/hour dust containing 60%C, when making 30,000 h3/hour synthesis gas, i.e. 70 kg. carbon lost 1,000 dry synthesis gas.
Synthesis gas composition was taken as 25% CO, 20% Co2 and 3.5% CH4, corresponding to 260 kg carbon/1,000 h3 synthesis gas.
(b) Producer Gas
The composition of dust in the final gas was stated to vary considerably, the ash from 45 to 55% and carbon from 50 to 70%; these figures are difficult to believe, since 50 + 55 = 105% and 45 + 70 = 115%. However the average composition was taken as 50% carbon and 50% ash.
The amount of dust was then calculated from the analysis of the dry coal fed to the producer (18% ash, 60% C) and of final dust.
Producer gas composition was taken as CO + CO2 + CH 4 = 32% (figures given on p.8 indicate only 30%), corresponding to 172 kg. carbon/1000 h3 producer gas. Hence the carbon lost as dust.
(c) Overall Carbon Balance
From the above figures and assuming no other carbon losses the following balance can be deduced, showing carbon in t/hour.
It should be noted that the total carbon loss of 25.7% calculated above is considerably less than the figures of 39 and 43.5% calculated in Appendix 1 from official reports of achieved data. Taking into account Dassow's statement that the figures in Ref.1 were too optimistic and the fact that some carbon losses could have occurred other than as dust, e.g. as leakage of synthesis gas into a regenerator, coal and gas used during starting operations, it is considered that at least one-third of the carbon was lost as dust.
Allocation of achieved dust losses between synthesis gas and producer gas is very difficult. From the material balances in Appendix 2 however we can deduce the following:
(a) Synthesis Gas
Carbon loss = 37,350 t for 330,774,000 h3 synthesis gas of 113 kg/1,000 h3 dry synthesis made. This agrees with the figure of 70kg/1,000 h3, deduced from Ref.4, allowing for the much greater overall losses achieved in 1943.
(b) Producer Gas
Carbon loss = 3.96 t/30,000 h3 producer gas = 132 kg/1000 h3 dry producer gas
This is greater than the figure of 74 kg/1000 h3, deduced from Ref.4. It is however inline with dust concentrations given by Dassow (see below) and with the amount of dust fed to the producers. The achieved 1943 efficiencies cannot be accounted for by lower losses on the producers, without allowing higher losses on synthesis gas.
In Appendix 2 it is deduced that with a synthesis gas unit on normal output the wet gas rate at the exit of the drier and cyclone was 123,000 h3/hour and the dry dust leaving the drier was 29.5 t/hour.
Using the above figure of carbon loss of 113 kg/1000 h3 dry synthesis gas made, the carbon loss for 25,000 h3/hour would be 2.82 t/hour and so the dust loss, with 60%C, would be 4.70 t/hr.
The figure of 160 g/h3 quoted by Dassow may have been referred to the dust content before the cyclone and after the separator, the difference being due to large material removed in the separator.
Also in appendix 2 it is deduced that with a producer unit on full output the wet gas rate was 33,500 h3/hour and dust leaving the Multiklon was 6.66 t. Thus the dust content after the Multikon was 199 g/h3 wet gas. This is to be compared with Dassow's very rough figure of 200 g/h3 for gas before the boilers and Multiklon and it should be noted that even the optimistic figures of Ref.1 lead to a dust concentration of about 1354 g/h3 wet gas after the Multiklon.
We conclude from a consideration of all the evidence that the following figures are probably near the truth:
Ref.1 C.I.O.S. Report Item No.30 File No. XXXII-90 by H.Hollings. Wintershall A.G., Lützkendorf.
Ref.2. C.I.O.S Document. Bag 2168. Target 30/4.03. Items 74 to 206. I.G. documents mostly dealing with Lützkendorf in 1939-40.
Ref.3. Oel and Kohle, 22.6.42, 38, 685. A.Thau.
Ref. 4. Wintershall document, abstracted during the visit but not taken away.
Ref.5. C.I.O.S. Document. Bag 2719. Target 30/6.12. Item 1/211. Notes on meeting between Koppers and Brabag.
Ref.6. C.I.O.S. Document. Bag 2719. Target 30/6.12. Item 1/232. Notes on conversations between Koppers and Brabag.
Ref.7 B.I.O.S. Final Report No.333 by R.J. Morley Winkler Generators for Manufacture of Water Gas.
Ref.8. C.I.O.S. Document. Bag 4182. Target 30/4.12. Item 3. Lützkendorf. Operating report for 1943.
Ref.9. C.I.O.S. Document. Bag 4182. Target 30/4.12. Item 4. Lützkendorf. Operating report for April 1944.
CIOS. Documents mentioned above will be made available at:- Board of Trade, German Division, (Documents Unit), Lansdowne House, Berkeley Square, W.1.
All applications for permission to inspect these documents should quote the appropriate Bag.Number.