Processes in the Manufacture of Cordite

Retrieved: 12/17/2014

Key Facts:

Sulphuric and nitric acids, carbonate, acetone, vaseline and nitrocellulose, delivered by rail from MS Dumfries (Drungans works).

Nitroglycerine manufactured by the Schmid process and washed clean of acids.

Acetone added to stabilise nitroglycerine during movement, wet mix and fabrication to cordite.

Spent nitric acid recycled, spent sulphuric returned to Drungans.

Cordite blended from nitroglycerine and nitrocellulose in 'wet mix' process, to produce gelatinised paste.

Paste incorporated with vaseline and other additives, possibly mixed with nitroguanidine (Picrite) as stabiliser.

Cordite pressed into sheets and granulated then extruded in presses as 'cords'.

Cordite dried in stoving houses with recovery of nitroglycerine and acetone fumes.

Acetone recovered in stills and recycled.

Cordite dried in Drying (steeping) houses to remove remaining moisture.

Finished cordite batches mixed to ensure uniform quality and packed.

Movement to storage in magazines.

Despatch by rail from loading station to ICI Ardeer in Ayrshire or ROF Bishopton.

Baker-Perkins Incorporator used in the Dalbeattie Factory Incorporation Houses

M. Bowditch

Summary of Processes involved in Cordite Manufacture...

What is Cordite?:

Cordite was the name given in Britain to nitro-based explosives made into 'cords' or fibres that can be as fine as hair or as large as spaghetti or rods. The name Cordite is most usually applied to the mixture of two explosives, nitrocellulose and nitroglycerine, made into 'cords' and used as small arms (pistols, rifles and machine guns) and artillery propellant between the 1890s and the 1970s by the British Army and the Royal Navy. The production of cordite in both the First World War (1914-1918) and the Second World War (1939-1945) was crucial for the survival of Britain.

Please Note:

For readability, much information has been abridged. A full description, with some technical detail of key processes, is being written as a book. Some information on cordite production is irrelevant to understanding the site and readers should NOT attempt to use the process text to produce either nitroglycerine or cordite, as the process involved is highly dangerous.

The Problems with Cordite:

Cordite is far more complex to manufacture than is gunpowder, for the simple reason that the raw materials can become unstable and explode if combined at the wrong temperatures, purity and acid levels. Much of the history of cordite manufacture is marked with fires, explosions and deaths. Add to this the need to produce large amounts of cordite rapidly in wartime, and it is clear that there is a balance between speed, safety and the rate of production. The solution was to produce this explosive in comparatively small amounts in several units at the same time, all within works sufficiently large that an explosion in one unit does not destroy or shut down the entire works.

Cordite was used in artillery and small arms as propellant. That means, it exploded when correctly fused, the hot gases driving a bullet or shell down the gun-barrel and out of the gun. Propellant was and is usually cased (e.g. in brass) but naval guns of large size actually loaded the main charge in silk bags.

(1) Raw Materials:

Some prepared separately at ICI Drungans, at Cargenbridge near Dumfries and shipped in by rail. Other materials were shipped in from production sites in southern Scotland and Northern England. All were unloaded at various buildings and gantries in the two railway goods yards and stored in adjacent tanks and buildings until needed. :-

Nitrocellulose - This is cotton waste converted to a nitrate at Drungans. (MS Factory, Dumfries).

Mixed Acids (M/A) - Concentrated Sulphuric Acid (57%) and Nitric Acid (43%) - Used in nitroglycerine production to nitrate glycerine and to act as a dehydrating agent during the process.

Oleum (Fuming Sulphuric Acid) - Used in the reprocessing of refuse (spent) acids to concentrate the nitric acid and replace used sulphuric acid.

Glycerine - Reacted with acids under strictly controlled conditions. Then a product of the soapmaking industry.

Sodium Carbonate - Used to neutralise ('wash') remaining acids in the separated nitroglycerine and so make it less likely to explode when moved to other processes.

Acetone - A solvent additive that stabilised the nitroglycerine and later in the process helped to gelatinise the nitrocellulose and nitroglycerine to a paste for forming into 'cords'. Acetone is known to most women as a nail- varnish remover.

Vaseline - Otherwise 'petroleum jelly' - added to the mix as a stabiliser and gun barrel lubricant.

Graphite - 'Lampblack' to reduce the risk of static ignition in some forms of loose cordite - its use at Dalbeattie not verified, but it may have been needed for small arms ammunition.

Nitroguanidine (Picrite) - From 1939 onwards, added to artillery cordite to reduce barrel wear and to stabilise the cordite in storage. Not proven to be added at Dalbeattie and probably added during mixing in small amounts.

In addition, large amounts of the following were needed during cordite manufacture :-

Coal - Three rail truckloads per day were needed for process steam to drive equipment and to dry the cordite. The nearest coalmines were at Kirkconnel and Sanquhar, although some coal may have come from elsewhere.

Water - 100,000 gallons per day of drinking quality water from Lochenkit Reservoir near Corsock. This was needed as wash water, for process steam and a range of other tasks. Kirkgunzeon Lade and Knock Burn provided additional water for cooling and fire-fighting, the water being pumped from intakes in the Burn and the Lade.

Goods arriving at the factory station platforms would have been unloaded for storage in warehouses or (in the case of liquids) unloaded from tankers for storage in tankage elsewhere. The Unit 1 (Southwick) layout can still be worked out.

(2) Acid Production:

Thanks to Mr. Nicholson, a former acid chemist at the works, the writer has been lent map-size plans and a sheaf of typescript Mr. Nicholson prepared in 1941-1943 detailing the structure, operational instructions and processes of the acid works at Ministry of Supply Factory, Dalbeattie, and the manufacture of nitroglycerine by the continuous (Schmid) process. The process that follows is that at Unit 1 (Southwick), but that at Unit 2 (Edingham) was exactly the same.

Mixed Acid (M/A or N/G-M/A) received in tankers from MS Factory Dumfries was stored in 3 X 80-ton mild steel tanks beside the track at the north end of the Unit 1 Railway yard. From there, it was pumped down 2" steel pipes some 1520 feet long (approx. 500 metres) to 2 x 80-tonne tanks at the charge house beside the Nitration Hill. From there, it was blown by low-pressure air along a 200 feet long pipe into the nitrator.

The nitration process as a by-product produces a much weaker acid (Nitrogylcerine Refuse Acid or N/G-R/A) mix contaminated with small amounts of nitroglycerine, water and other impurities. This 'Refuse Acid' was piped back to the Waste Acid Plant for reprocessing. In brief, steam was used to de- nitrate (free nitric oxide vapour from) the acid mix, the nitric vapours being recondensed to produce a low-concentration Nitric Acid. The low concentration acid was concentrated by being mixed with Oleum (fuming concentrated Sulphuric Acid), dehydrating the Nitric Acid and raised its concentration to the same as factory-delivered M/A, the Oleum being diluted to strong Sulphuric Acid. The reprocessed M/A acid mix was so generated was then returned to the Nitration Hills. The de-nitrated sulphuric acid or 'Dilute Oil of Vitriol' (DOV) was concentrated in gas-heated pots to drive off water, then as a more concentrated 'Concentrated Oil of Vitriol' (COV) it was cooled and shipped back to MS Dumfries by rail tanker.

Mr. Nicholson's documents form a unique record of the Waste Acid and Nitroglycerine processes and have been preserved as copies to guard against loss.

(3) Manufacture of Nitro-Glycerine:

Mr. Nicholson wrote a nine-page monograph on 'Manufacture of 'A' Nitroglycerine - Continuous Process' which forms the basis of the known process. For reasons of space and copyright, the key features are summarised here :-

Preparation at the Charge House:

Mixed acid (57% Concentrated Sulphuric Acid and 43% Concentrated Nitric Acid), Glycerine and Sodium Carbonate (washing liquid) was stored in mild steel horizontal tanks (boilers) of 160,000 lbs capacity near the Charge House.

Glycerine was delivered by rail in drums of 560 to 1680 lbs net weight, then emptied into three steam heated tanks of 29,000 lbs capacity, from which it was piped to a pair of 11,000 lbs capacity steam heated tanks in the charge house.

Wash liquids of hot, cold and soda (sodium carbonate solution) water were stored in boilers in the charge house.

High pressure air at 75 lbs/square inch (2.5 atmospheres) was delivered to a 930 cubic feet storage vessel beside the charge house, then reduced to about 28 lbs/square inch (0.9 atmospheres) low pressure air before being piped into a 100 cubic feet vessel in the ground floor of the Nitration House. This low- pressure air was used to drive the mixed acid, glycerine and wash-water solutions to where they were needed, and where necessary to aerate the mix.

A small quantity of oil was added to the process of nitration to enhance (accelerate) separation of the nitroglycerine from the mix of acids and glycerine following reaction.

The Nitration Hill:

The Nitration Hill was the highest of several hollow artificial hills at each Unit. The function of the earth and brick was to contain and direct upwards any accidental explosion. A two-storey structure, the Nitration House was designed with a large water tank under it and various tanks, whilst the main procesases took place on the upper floor. The following section summarises a rather complex process from Mr. Nicholson's firsthand description.

The Schmid Nitrator:

Mixed acid and glycerine were reacted in a small steel chamber about 39 inches (1 metre) high and 28 inches (70 cm) internal diameter, with a chilled brine cooling jacket and an internal mechanical stirrer. Thermometers fitted to the Nitrator allowed the operator to check that the mix was not becoming either toohot or too cold, as overheating could result in nitric oxide fume production and the risk of spontaneous detonation of the nitroglycerine. The reactor could hold 8.1 cubic feet of the mixture being reacted, amounting to about 890 lbs of acid. Reaction was continuous, with nitroglycerine rising to the top of the Schmid reactor and overflowing through pipes into the adjacent separator.

The Separator:

The acid and nitroglycerine mixture overflowing from the Nitrator flowed down a stainless steel tube to an inlet in the base of the Separator. This resembled a rectangular steel box 161.5 cm x 85 cm x 85 cm, inclined at 45° to the horizontal, with 35 corrugated plates inside it, slightly shorter than the length of the box. The acid and nitroglycerine gradually separated, the lighter nitroglycerine rising to a draw-off pipe, which was opened once sufficient nitroglycerine had accumulated. The capacity of the Separator was about 51 cubic feet, or about 5600 lbs of acids. This allowed a much slower flow than in the Nitrator, allowing the nitroglycerine to accumulate. The refuse acid (R/A) was drawn off at a lower level, initially being stored in a tank at the lower level of the Nitration Hill, then when this 'acid egg' was full, the excess spent or refuse acid (R/A) wass diluted slightly with water and sent to tanks outside the hill.

The Wash Columns and Intermediate Separators : Three columns 8 feet high were located in the Nitration House. Each consisted of glass cylinders stacked on top of one another, with stainless steel perforated diffuser-plates and rubber glands joining them. The assembly was held together by aluminium bands around the joints and vertical rubber-covered steel rods, linking the cast-iron base plate to the stainless steel head, which had a fume vent to the top of the building. The nitroglycerine from the separator was piped with injected cold water to the base of the first column, which was filled with water at at least 12° C. and was aerated with low pressure air. Fumes from the column were vented, whilst the nitroglycerin/water emulsion spilled over into a small tank of warm water - the intermediate separator. Nitroglycerine being heavier than water, the nitroglycerine sank and the wash water flowed from the top of the intermediate separator, the nitroglycerine then flowing to the bottom of the second wash column, where the process was repeated in water at 40% C. The third wash-column used 1.3% sodium carbonate solution to neutralise any remaining acid over some 12-15 minutes, before discharge down a rubber pipe to the second Wash Hill.

The Drowning Tank:

The main reason for a two-floor nitration house was safety. Underneath the house process level was a 2,000 gallon tank of cold water. If temperature in either the centre of the Schmid Nitrator or the Separator head exceeded 28° C, or a thermometer in the side of the Separator exceeded 30° C, thermostats operated to break the power to electromagnets on valves in the base of both Nitrator and Separator, so the overheated mixture was immediately discharged and doused by the water in the drowning tank, which automatically is aerated and topped up with more water. Alternatively, the discharge cocks can be opened by a lever in a blockhouse set outside the Nitration Hill, or by a switch that cuts off all power. A similar arrangement allowed operations staff in the Nitration House to use a lever or a pull-switch to drown the reacting mix within about 40 seconds.

The Final Washing House:

This held a further two columns that washed the soda and nitroglycerine in hot water at 40° C and then in cold water at at least 12° C. Nitroglycerine freezes at about 12° and can be highly unstable when thawing out. Once it had left the fifth wash column's intermediate separator, the now virtually pure nitroglycerine was collected in a lead settling tank containing about 820 lbs of the explosive. It was then discharged into three identical four-ply flannel filters into separate filter boxes and then into one of three separate 1200 lbs lead storage tanks. Up to 280 lbs of nitroglycerine at a time was removed to lead weighing tanks before being moved to the next process.

As an appreciable quantity of nitroglycerine remained in the wash waters, these were first allowed to flow over a stainless steel cataract (an inclined plate with riffles, or a series of small settling tanks) to let the nitroglycerine sink, then the wash water was discharged along a rubber-lined wooden channel into the next hill or hills, the Waste Water Settling Houses.

The Wash Water Settling House and Mud Washing:

From the hills present on site at Unit 2 (Edingham), it appears that the wash water from the Nitration Hill and the wash water from the Final Washing House, went to separate Wash Water Settling Houses. Nicholson in his account fails to mention a key point - the nature of the Mud Washing that he mentions in the title to his section. Mud containing nitroglycerine is shown as being a secondary product of settlement. There are three possible sources for this mud :-

Decomposition products of the very small quantity of mineral oil used as a separation accelerant.

Impurities from the wash water - unlikely, as the water would have been of drinking water purity.

Addition of some clay-like substance to act as a flocculant, gathering the oil and nitroglycerine traces and forcing them to precipitate.

For these reasons, it is inferred that a small quantity of some kind of pure mineral clay (possibly Kaolin or Fuller's Earth) was added to the wash water as it entered the Wash Water Settling House and before settlement took place.

Chute Separation and Puddled Pond:

The water leaving the Wash Water Settling Houses goes through an underground drain pipe into a wooden chute or trough and then into a final pond lined with puddled clay.

Nitroglycerine Desensitising:

The nitroglycerine has to be desensitised (stabilised) by mixing with alcohol, acetone or dinitrotoluene, or blended with ethylene glycol nitrate to keep it liquid. Dalbeattie used acetone, from the evidence of the layout and the words of former staff. It is unclear exactly where or how this was carried out, but the Final Washing House has a trolleyway at a high level and a narrow-gauge railway at low level. This is important because it indicates that a comparatively unstable product (i.e. nitroglycerine) was being transferred at high level, whilst a more stable product (acetone/nitroglycerine emulsion) was leaving at low level, probably to go to the Burette Houses.

It is inferred that acetone was added to the nitroglycerine in the Final Washing House either in the three storage tanks or after weighing. The argument for the storage tanks is that they could have had a graduated quantity of acetone run into the tank when empty, into which just enough nitroglycerine was added to reach a given volume or mark and so a set level of concentration. The weighing tank would have then been used to ensure that a given weight of nitroglycerine/acetone mixture was placed in an individual bogie tank for despatch to the burette houses.

From Nitroglycerine to Cordite:

Following manufacture, the nitroglycerine would have had to be moved to the 'Wet Mix' or Burette Houses for mixing with the nitrocellulose. The Sanford account of this in his 'Nitro Explosives' is of interest for several reasons :-

The nitrocellulose was placed in double-walled lead-lined troughs or mixers, these being kept at a steady temperasture of 40° to 45° C by water circulating in the tank walls as the nitroglycerine was slowly added and carefully mixed with the nitrocellulose. Acetone previously added to stabilise the nitroglycerine enhanced the gelatinisation process and would make the cordite easier to work.

Sanford describes the use of high-pressure steam injected into tanks of cold water as the way to ensure a reliable 60° to 80° Centigrade hot water supply. This would appear to have been the system used at Dalbeattie, judging from the arrangement of pipe bearers and tank bases at the entrance of the Burette Houses; the complexity of the tank bases may indicate recirculation of the cooling water.

Just how the mixing took place is not clear, but there is evidence that cordite at Powfoot and Ardeer was blended by powered stirrers a bit like commercial bread dough mixers, in a very wet environment to guard against acid fumes and fires. The alternative was to use a wooden paddle to stir this 'Devil's Porridge', another 1914-1918 war expedient. The process took about half an hour to complete, until the mix became semi-transparent and the nitrocellulose had been completely dissolved. Some of the bases have heavy studs that could have been used to secure some kind of motor.

The mixture would have been 65% guncotton, 30% nitroglycerine and 5% vaseline for Cordite MD, the type using acetone in its production. Later in 1943, solventless cordite processes became available but do not appear to have been carried out at Dalbeattie.

Cordite Additive Mixing:

This is described by Sanford as being carried out after gelatinisation is complete, but it is unclear whether it was used to try and provide a common quality of product, or whether dyes, vaseline (and possibly picrite) were added at this point. The writer's best guess is that the additives were blended in at the Incorporation Houses before the mixture was taken in bags to the Paste Rolling Houses or the Press Houses.

Cordite Milling Section:

A different bogie would have been used to move the bags from the Burette Houses to the Cordite Milling section. Each Unit had seven main Milling houses - three for Rolling, three for Die Pressing to produce the strands or cords of cordite, and one house split into two (the Double Sided House) for Paste Rolling and Flaking. The Incorporation and Die Pressing buildings were paired together. This indicates that the processes were amongst the most dangerous other than for the Nitration and Burette sections.

The raw blasting gelatine dough arrived in one of three bunded (banked) enclosures linked by covered passages to the adjacent Incorporation House, to be stored in flimsy asbestos cement houses until required for use. These Expense Magazines have not survived as all were demolished, but their sites are marked by squares of acid resistant gritless asphalt.

Incorporation Houses:

The milling process started with the incorporation of additives, done in twelve individual rooms that faced out onto a service veranda. The incorporation process produced a more solid product. This sheet or slab product was then removed and taken round to the Block Pressing houses.

The machines were powered by shaft-drives from motors in a passage that ran the length of the building. The fumes from the cordite processing - a mixture of acetone and a little nitro-glycerine - were collected by ventilators and may have been condensed to liquid for return to 'wet mix', but this is not clear from the evidence. The writer was surprised to discover that the same passage was used to house the ducts (and, possibly, condensing equipment) used to extract the fumes and possibly to recover the acetone.

Block Pressing Houses:

The purpose of these presses, sited in eight rooms with double doors and a split-level floor, was to turn the slabs or sheets of cordite into blocks the right size and consistency for extrusion in the Die Presses. The same rear- passage system for shaft drives survives, the studs in the passage floor where the motors were fixed. It appears from remarks of Malcolm Bowditch and Leslie Hayward that the machines in the Block Pressing House were designed with a dump tank of water above the press, so that the presses and the operators could be drenched in seconds in case of fire or acid fume generation. The presses had collection tanks for this water, which remain as stark evidence of the hazards.

Apparently this area was the most fume-laden part of the Cordite Milling Section and only mature women were normally tasked with this responsible and dangerous task. There are references to 'The auld maids in P.6' (Pressing 6), presumably mature unmarried women, but the Ardeer works is known to have borrowed women from the Blending and Packing section for this very unpopular and dangerous duty, if staff fell sick. The risks of acetone addiction from the fumes were serious enough for staff to be rotated to other duties to reduce their exposure; accounts of Ardeer tell of staff becoming intoxicated, at least one so badly that he died on the way to hospital.

Die Pressing (Extrusion) Houses:

The blocks of cordite were run from the Block Pressing Houses by bogie to the Die Pressing Houses, where the presses extruded the consolidated paste into strands or 'cords' like oversized spaghetti that were cut to length (if fairly large) or rolled onto reels (if of small cross-section) before removal for drying. The purpose of this extrusion or pressing was to give a large surface area for ignition by the detonation charge when fired in a gun.

The rooms in which this task was done are the largest in any of the process houses. They had massive extractor fans and ducting near the ceiling to remove the fumes. Motors set in corner rooms of these Houses drove individual drive shafts to each room to power the process machinery. The finished 'cords' were removed in boxes to the Stoving Houses by bogie, using the narrow-gauge railway.

Stoving Houses and Acetone Recovery:

Although the acetone was a good stabiliser of the nitroglycerine, and a material help in gelatinisation and in pressing, it had to be removed from the finished cords or they would rapidly deteriorate. The solution was to lay the cords out on perforated zinc trays in rooms which were heated by radiators, with a fume extraction system to remove the acetone and any inadequately- blended nitroglycerine. The fumes were condensed and then distilled to recover and recycle the acetone, any nitroglycerine being returned to the Nitration Hills for reprocessing.

Drying Houses:

A similar process to Stoving but at a lower temperatures was used to remove any remaining moisture so that the cordite was left completely dry. S

Blending and Packing Sheds:

The sorting or blending by the women packers was a matter of taking mixtures of pressed cords from various batches so that each case held a standard quantity and quality - presumably, so the good, the bad and the indifferent were so mixed as to provide a product with consistent behaviour. The job was very dull and repetitive, so many of the women sang as they worked to pass the time. Sorted, the cordite was checked before being boxed and taken by cart down to the magazines until time for despatch.

The Magazines:

These are heavily built, well-ventilated structures of concrete with strong steel doors, located mainly to the south of the works, towards Edingham Moss. Manifestly designed for holding explosives, they have external weatherproofed switch frames for lighting and are sunk into deep trenches. During the years from 1947 to 1965, it is suspected that they would have held Naval mines and other ammunition. Matthew Taylor of Edingham Farm has confirmed that they were built at the same time as the rest of the works, Gordon Nicholson, a chemist from the acid works, explained that the explosives were gradually accumulated until a trainload was available, then were loaded at the railway siding to the south of the railway for despatch to Ardeer works and possibly eleswhere.

2006 Richard Edkins, Dalbeattie Internet.