by Tony Insall
Members of the GUNNERSIDE and Swallow teams who blew up the heavy water plant. Leif Tronstad is seated in the middle, to his left is Joachim Rønneberg. © NHM
In mid-April, Boyle, by then assistant to Gubbins, asked Wilson whether he had received any reporting from SIS about the success of GUNNERSIDE and German reactions to it. Had the Germans in Stockholm connected the arrival of the GUNNERSIDE party with the operation and increased pressure on the Swedes? In a very rare indication that SOE may have been aware of the substance of intercepted signals decyphered by GC&CS, he continued, ‘I wonder whether we have covered adequately these matters by putting C fully into the picture?** There is a vast amount of enemy traffic to examine and unless section V and others know what to look for they cannot be expected to keep us informed.’ Wilson replied that he had briefed SIS in detail before the operation and afterwards, and that Section V fully understood SOE’s interest in this particular area.33
Heavy water production restarts: War Cabinet considers further measures
Initial estimates of the length of time that the heavy water plant would be out of operation did indeed prove to be overly optimistic. There were reports in early July that production would be resumed in the middle of August. It actually happened slightly earlier: Swallow reported on 3 August, less than six months after the raid, that it had restarted. Tronstad predicted that it would return to full capacity by the end of August. He suggested a bomber attack on the Skarfoss dam (which fed the power stations at both Vemork and Såheim which separately also provided a small quantity of heavy water), as a means of hampering further production. He also suggested taking steps to contaminate the heavy water with castor oil. Neither suggestion was taken up. Wilson informed Gubbins of the situation, explaining that repairs had been made more quickly than anticipated, and that scientists were unsure of the acuteness of the problem caused by renewed production, as they thought it would take a further two or three years before a weapon might be viable. He concluded that this uncertainty needed further examination by Allied scientists. If another attack was considered necessary, every possibility including air attack should be considered.34 Rønneberg heard about the resumption of production through a press report. He wrote to Wilson:
It certainly seems as if our work was not thorough enough. May I ask you one favour? If our old target should once again be in the foreground, would you please remember GUNNERSIDE? I know that all my friends would be obliged to get a chance to play the guards at Vemork a new trick.35
Wilson replied that it was unlikely that SOE would wish to stage another major operation, but if one was required, then Rønneberg would be given first refusal.
Shortly afterwards, Perrin recommended a fresh attack on the plant. He suggested that destruction could be achieved by a second sabotage operation or by a daylight bombing attack. He thought that bombing was likely to be the most effective, but noted that Tronstad had argued that such drastic measures should not be taken because of the extent of the damage that would be caused to the whole factory and thus to local livelihood, and that he had suggested it might be better to try to protect the plant if there was any chance it might be available for use by the Allies in the near future. If there was to be an attempt to stop production, he noted that SIS judged that only a daylight bombing raid would be really successful in ensuring that production could not be restarted for a long time. He suggested that the USAAF should be approached, adding that ‘it would probably be advisable that no decisions should at present be communicated by us to the Norwegian authorities’.36 However, the Assistant Chief of the Air Staff Frank Inglis did not think that bombing would be a practical proposition, and recommended that SOE should be asked to consider other means of sabotage. Ismay and the Chiefs of Staff agreed. Gorell Barnes accordingly approached SOE and asked for an appreciation of the possibility of a further sabotage operation. He stressed the view of the War Cabinet office that ‘we ought not to have regard to Norwegian post-war interests if these form an obstacle to the most effective plan. The appreciation should be prepared with this point in mind.’
On 11 October Joe Adamson, the Norwegian section operations officer, informed Wilson that although there was a dearth of precise information, they knew from Swallow that the guard forces at Vemork and Rjukan had been increased and entry to the plant would now be too difficult to force. Moreover, after the success of GUNNERSIDE, it was unthinkable that the Germans would not have increased their physical defences, blocking entrances and adding landmines and additional lighting. He judged that an attack was out of the question, but if handled with care, contamination could be effective in reducing production quite significantly without direct danger to the perpetrators, if reduction rather than stopping production completely was an acceptable outcome. He took issue with the Air Staff view, considering that a daylight precision bombing attack would be feasible and that it was the only effective answer. He added carefully that he was ‘refraining from any expression of opinion as to political repercussions in the event of such an attack’. Adamson’s advice was accepted, and Harry Sporborg so informed Gorell Barnes, explaining that SOE concluded that further sabotage would not be practical,
unless one could produce a body of at least 40 well trained and heavily armed men, who would have virtually no hope of escape afterwards. An operation of this character could not in present conditions in Norway be carried out on a clandestine basis and could really only be done on the basis of a small commando raid such as that which ended so disastrously on the previous occasion. As regards an attack from the air, we quite realise that no low level attack would be a practical proposition, but would there be no chance of a medium level precision attack in broad daylight by United States aircraft? Although this would be a very difficult operation, it does seem to be the best chance of doing really conclusive damage to the plant. As you will see from our appreciation, we think it would be possible to interfere with production, or at least to reduce it for a month or so, by insaisissable methods.†† We are pretty confident that if instructions were given, a fairly satisfactory though of course temporary effect would be created. Perhaps you would let me have instructions about this.37
Gorell Barnes rejected the idea of contamination. This was later shown to be a prudent decision. Gunnar Syverstad, an engineer at the plant and a reliable resistance contact, later tried treating two of the vats with oil, but it floated and was quickly noticed. He concluded that such contamination would have a nuisance value, but would not prevent the subsequent use of the heavy water.38 Sir John Anderson therefore requested that a bombing raid be undertaken. The Chief of the Air Staff, Lord Portal, approached the commander of the US Eighth Air Force, General Eaker, who, after a study of the target, agreed to take on the assignment.
The US Air Force bombing attack on Vemork and its aftermath
Several weeks later on 16 November 1943, Vemork was attacked from medium altitude by several squadrons of Flying Fortresses from the Eighth Air Force. Shortly after they had finished bombing, a further attack was made by a force of Liberators which had been prevented by cloud from bombing their primary target, an airfield at Kjeller, and so diverted to their secondary objective. Unfortunately they mistook the nitrate plant at Rjukan for their target and bombed it instead. The bombing caused widespread damage, destroying the nitrate plant, most of the penstocks which supplied water to the power station, and some of the generators as well. It did not touch the high-concentration plant producing heavy water which was located in greater safety in the basement – though loss of the water supply and damage to the rest of the plant were sufficient to halt production. Of greater significance to the Norwegians was the number of civilian casualties. Twenty-one Norwegians, many of them women and children, were killed – most of them as the result of a direct hit on an air raid shelter.
The attack provoked a very strong reaction from the Norwegian government, which complained about the loss of a significant supply of fertiliser to a population which was already very short of food
(particularly following the damage to another fertiliser plant by bombing at Herøya in July), and the fact that they had neither been consulted nor informed about the attack in advance, as well as about the loss of Norwegian lives. Lie sought a meeting with Eden (who as a member of the War Cabinet would probably have been aware of the planned attack and perhaps wisely declared himself unavailable) and saw ambassador Collier on several occasions instead. He told him that ‘the destruction of the factory was something not far off a national disaster’, and he understood that it had been put on a list of exempted targets. This had been discussed in the Norwegian cabinet and he had been asked to find out exactly what undertakings had been broken and who was responsible for breaking them. However, when he questioned Hansteen, who was ultimately responsible for the Norwegian side of arrangements with Air Ministry, he was disquieted to discover that there was no written and binding agreement regarding respect for exempted targets. This was a significant admission, for it turned out that there had been different understandings of the nature of any assurances which might have been given. Lie added that the root of the trouble was that ‘both on the Norwegian and on the British side the military and air authorities concerned had been reluctant to take the civilians into their confidence, while it was notorious that on the British side the Air Ministry always tried to be a law unto themselves’.‡‡ 39 Lie followed this up with a formal written Norwegian protest on 1 December, criticising the attacks on both the fertiliser plants, which were also a potential source of revenue because some of the output was sold to Sweden. He stated that no reason had been given for the attack on Vemork, and that a bombing raid would have been highly unlikely to destroy the heavy water production there. Indeed effective destruction was almost impossible. Stressing the need for effective collaboration in future, to ensure that the nature and extent of any further bombing attacks should be proportionate to their objectives, he sought British assistance in establishing an arrangement which would henceforward permit such collaboration. When handing over the aide-memoire, Lie told Collier that while the Norwegian military authorities could not point to the existence of anything more than a verbal agreement on exempted targets, they had certainly understood and given the Norwegian government to understand, that this agreement precluded action against the Rjukan fertiliser works (as distinct from the heavy water plant close by at Vemork) without prior consultation. Though this was important, the main point for the Norwegians was to have an agreement on a list of accepted targets.
Tronstad was also shocked and bitterly disappointed by the attack on Vemork. He believed that he had been misled because he had been given an assurance by Perrin, Welsh and others that Vemork was a ‘starred target’ and would not be attacked without prior consultation.40 Tronstad reemphasised this in a letter to Hansteen, which he also handed to Perrin, stating that ‘he had constantly been told that nothing would be done without his knowledge’.41 Since Tronstad’s consistent opposition to the bombing of Vemork was quite widely known, it is surprising that any of his British colleagues in DSIR, SIS or SOE should have given him such an assurance. Some of them would have been in a position to know of the operation which was being considered, and would therefore also have known that any assurances which they might have given would have counted for nothing. It is unsurprising that there is no British record of any of these exchanges.
Although the order for the operation was not given through the usual channels, it is worth taking account of how the Air Ministry’s German target committee (the ultimate body normally responsible for target selection) weighed up conflicting considerations when assessing the priority which should be given to Norwegian targets throughout the war. It shows the extent to which the Air Ministry responded to requests from other departments in the selection of targets, and did not necessarily select them itself. There were no discussions about the need for any prior consultation with the exiled governments of countries where targets were being considered for attack. There is only one reference to Vemork, in December 1941, when the committee was told that the ‘heavy water’ factory in Norway had been suggested as a target. The officer responsible reported that he had been in contact with the experts, who had informed him that such an attack was not yet merited. Shortly afterwards, the MEW representative told the committee that the Norwegian government had complained about RAF attacks on fish oil factories, since this was the sole source from which the population could obtain their fats. He sought the opinion of the Admiralty and Coastal Command. They both stated that these factories were useful alternative targets to aircraft of Coastal Command or the Fleet Air Arm, but were not regarded as of prime importance. After further review it was determined that these targets would not be attacked. There was a discussion about the NS anniversary meeting to be addressed by Quisling in Oslo on 25 September 1942, when it was agreed that it should be attacked. (This was successful.) There were also discussions about an attack on the fertiliser plant at Herøya, but the MEW reported that it would not be completed and therefore constitute a worthwhile target before 1943, so it was taken off the target list. In October 1942 there was consideration of a proposal for bombing cement factories which were wholly engaged in supplying material for German fortifications. It was agreed that the importance of these targets, combined with the tactical difficulty of locating them, did not justify attacking them.
In late 1942 there was discussion of targeting the molybdenum mines at Knaben. The chairman of the target committee concluded that the plant, being small and difficult to locate, constituted a most unsatisfactory bombing target. The committee agreed that the Knaben mine was more suited for action by SOE. However, the subject was raised again in January 1943 because of the increased importance of molybdenum from Knaben to the German economy as a result of the denial of supplies from North Africa. CCO and SOE held out little prospect of a successful attack, and the same applied to bombing by normal methods. It was suggested that Bomber Command might consider the possibility of a low-level attack by Mosquitoes. The subsequent attack did have some success, but insufficient to put it out of production. It was later attacked by a larger USAAF raid on 16 November 1943, the same day as the attack on Vemork, though again damage was limited. There was consideration of a possible attack on the Holmenkollen hotel in Oslo as it was being used by the German Air Force, which was dismissed unless a case could be made that such an attack would have significant political value. There were also reviews of the U-boat pens in Bergen and Trondheim. Although the pens in Trondheim were considered to be less solidly built and therefore more vulnerable to bombing, the RAF preferred Bergen. When the pens there were attacked in October 1944, there was poor visibility. Most of the force did not drop their bombs as they could not identify target markers. However some bombs fell on the central part of the town, causing civilian casualties and damaging buildings of cultural and historical value. In April 1943, there was an assessment of Dutch protests about casualties to civilians, following attacks near Haarlem. It was agreed to re-evaluate targets with a view to eliminating those where attacks might lead to similar results in future.
These records give a useful insight into the way in which important targets were selected (or sometimes discounted), and the attention which was sometimes paid to the concerns of Allied governments. This does not mean that Vemork would have been treated any differently, given the importance which the War Cabinet attached to the destruction of the heavy water plant there, but it shows that while the Air Ministry may have been a law unto itself – after all, it had a job to do – it was quite capable of taking reasonable requests into consideration.42 But there is nothing in Air Ministry or Cabinet Office files to show that there had been any prior discussion about placing Vemork or any other Norwegian target on a list which would exempt them from aerial attack. There is, though, evidence to show that both were keen to ensure that the Norwegians were not made aware of what was being considered. Moreover, in his unpublished history of the Norwegian section which was written after the war, Wilson
stated that the USAAF bombing of Vemork had been carried out by them ‘without consultation with their allies’.43 Unless Wilson was referring to a lack of consultation with Norwegian Allies rather than British Allies, then in view of his involvement in preparing the appreciation which concluded that bombing was the only viable means of destroying the plant and its stocks, which led to Sporborg’s recommendation that a USAAF attack should be reconsidered, this assertion is hard to understand.
As a result of the protests over Vemork, there were protracted negotiations about an agreed list of targets in a process which involved the Air Ministry, MEW and later the Admiralty – because of the possibility of the navy shelling coastal targets and the involvement of carrier-launched aircraft. Indeed in June 1944, Lie complained about an attack by naval aircraft on shipping and oil tanks off the Norwegian coast, in the course of which a fish oil factory north-east of Stadlandet was hit. This was not on the draft list of agreed targets which was being discussed. Lie pointed out that it was producing entirely for Norwegian consumption, that it was not a special or urgent case, and that it could not be considered a German military or naval objective. An agreed list, which included Knaben and the mine at Sulitjelma, was eventually finalised on 11 October 1944. It made a number of concessions to take account of Norwegian concerns. The attack on the Bergen U-boat pens (which, as a military target, was outside the scope of the agreement) took place shortly afterwards. However, when making his protest to Collier, Lie acknowledged pragmatically that there was no evidence that there had been any breach of the undertaking and he was complaining only because the Crown Prince had instructed him to do so. He made clear that he wanted to avoid any further recriminations.44