Members Blog

The Bay Offensive: Success or Failure?

Posted by David Craddock, on June 14th, 2023

The Battle of the Bay or, as it is more often known, The Bay Offensive, has been described as ‘…one of the decisive battles of the war … which was in the end crowned with decisive victory.’ But it never quite achieved the popular acclaim accorded to the Battle of Britain that RAF Coastal Command had asserted in their wartime Despatches.[1] To place the Bay Offensive in the wider context of the Battle of the Atlantic, it is helpful to see how the forensic analysis of the battle contained in the original Naval Staff History came about.


Towards the end of the Second World War, the task of compiling a Naval Staff History of the Battle of the Atlantic was given to Vice Admiral Kenneth Dewar who in 1922, with his elder brother Arthur, both then serving Captains, had been responsible for the unfortunate and divisive Naval Staff Appreciation of the Battle of Jutland.[2] Dewar recruited a former Fleet Air Arm pilot and historian Lieutenant Commander David Waters, who had been seconded to the Naval Historical Branch in 1946. He also brought in a retired Merchant Navy officer Commander Frederick Barley RNVR; together Waters and Barley began work on a purely chronological narrative. This approach was swiftly discarded when it became clear that the doctrine adopted and issued to the fleet following a war-gaming exercise at the Royal Naval College Greenwich in 1949 appeared to contradict all the lessons learned from the Battle of the Atlantic. This was Exercise Trident, predicated on a future (non-nuclear) war scenario in 1957 which had concluded that, although the convoy system was still seen as having value, the defensive posture of close air support was to be abandoned in favour of ‘Attack at Source’. This was to be achieved with a large task force of surface ships with carrier-born air support to interdict Soviet submarines ‘…both at source and in the waterways adjacent to their active bases.’[3]


Eric Grove assigned various reasons for this mindset, not least that the majority of officers who had direct experience of convoy operations in the war just ended had by then left the Service. ‘Many influential officers were unfamiliar with the history let alone the dynamics of the Battle of the Atlantic fought mainly by reservists.’[4] It was clear to Waters and the new head of the Historical Section Rear Admiral Roger Bellairs, who had replaced Dewar in 1948, that ‘…there was little in the genuine experience of the recent war to justify the future vision of Trident.’[5] Bellairs saw that what was needed was not a chronological record but a thorough and dispassionate ‘analytical digest of all the relevant facts of the Battle of the Atlantic’ from which lessons could be drawn – and as soon as possible. It would be another eight years, which included some considerable inter-service wrangling, before the work of Waters and Barley received the official blessing of the Board of Admiralty for publication. In his introduction Groves called it ‘the most powerful justification of the convoy system of warfare ever written.’[6] His comment suggests that the work begun under Vice Admiral Dewar was no less impartial than Dewar’s Staff Appreciation of Jutland some thirty years earlier but with none of the personal animus.


From 15th April 1941, within two months of Western Approaches Command moving from Plymouth to Derby House in Liverpool to better co-ordinate convoy escorts, Coastal Command came under the operational control of the Admiralty. As an organisation it had enjoyed none of the support given to Bomber Command and Fighter Command who were seen as taking the fight directly to the enemy. It nevertheless came to play a vital role during the Battle of the Atlantic in countering the German U-boat threat to allied merchant shipping. The focus here is on the outcomes of both the defensive and offensive stances adopted by Coastal Command and in particular the deliberate shift in emphasis announced in the Spring of 1941 from defensive to offensive operations – ‘…to seek out and destroy the U-boats wherever they are to be found, rather than wait for them to come to us.’[7]


From the outset, Coastal Command had, within the range limits of available aircraft, provided close air cover for convoys or patrols in direct support. Now, under the new directive from the Air Ministry’s Air/Sea Interception Committee, cover would only be provided when intelligence suggested that a convoy within range was being shadowed. From June 1941, the emphasis shifted to regular patrols which were judged to make the best use of aircraft flying at the limits of their operational range from airfields in Cornwall, Devon and the north of Scotland. These were intended to interdict the passage of U-boats between their bases and operating areas: the Northern Transit Area between Scotland and Iceland and the Bay of Biscay through which boats based at Lorient, Saint Nazaire and La Rochelle had to pass. But whatever the psychological effect the patrols may have had, success was limited without the means to sink U-boats who soon adopted the tactic of remaining submerged during the day, surfacing only at night. German records, to which Waters and Barley had access, revealed that during the first five months of 1942, not a single U-boat was sunk or damaged by the transit area patrols.[8] At the same it was shown that aircraft attached to or in direct support of convoys recorded three times as many sightings of U-boats per flying hour as the patrol aircraft.


Two developments in the Spring and Summer of 1942 did, however, begin to make an impact with the transit area patrols: an improved aerial depth charge and the introduction of the Leigh Light. Used in tandem with the existing Mark II Air-to-Surface Vessel (ASV) radar to illuminate targets at night the moment the radar return was lost in the sea clutter (at about 1000 yards or less than ten seconds flying time), the U-boat’s immunity from night attack was ended. Though losing only one submarine to a Leigh Light attack in July, The U-boat Command reversed their policy, now transiting submerged at night and on the surface by day. With Luftwaffe fighter cover now available over Biscay to counter the daylight patrols, U-boat losses to air attack by day remained at one per month for the next three months. For the Northern and Biscay transit areas combined for the period August 1942 to January 1943 a total of five enemy submarines were lost, a number soon absorbed by the rapid increase in the size of the U-boat fleet. A comparison with the attrition rate from aircraft escorting convoys or flying patrols in direct support of them over the same period shows a 250% increase in the number of U-boats sunk for a significantly lower number of flying hours. The number of aircraft lost from all causes in direct protection of convoys is also dramatically lower. The table below is abstracted from a much longer tabulation of sightings, attacks and sinkings against flying hours included among the very detailed appendices of Volume 1B.[9]


U-boat Sinkings against Flying Hours and Aircraft Lost, August 1942 – November 1943

Convoy Escort and Support Northern Transit Area Bay of Biscay
August 1942 – January 1943 1,660 1.15 4,800 6.6 12,350 34
February 1943 – May 1943 1,120 0.47 1,201 0.75 2,477 4.2
June 1943 – August 1943 2,718 1.0 12,608 3.0 1,765 3.3
September 1943 – November 1943 522 0.67 –– 8,344 8.2

*FHPS: Flying hours per sinking, ALPS: Aircraft lost (from all causes) per sinking

† 3 aircraft lost, no sinkings


The statistical evidence compiled by Waters and Barley speaks for itself and demonstrates the difficulty Coastal Command had in achieving the success in the transit areas claimed in their despatches. Another factor that changed the dynamic in favour of the U-boats over the winter months of 1942/43 was the deployment of a radar detecting device known as Metox. This converted the received radar pulse into an audible signal the pitch of which increased as the range closed and the aircrew switched the radar to a shorter range with an increased the pulse frequency. This gave sufficient warning to enable the submarine to dive before an attack could be pressed home. It had an immediate impact in reducing the number of sightings and consequently the number of attacks by day or night despite the increase in numbers of U-boats passing through the transit areas. For the three months of November 1942 to January 1943 not a single sinking was recorded, forcing Coastal Command to admit ‘…our anti-submarine patrols in the Bay had, for a time, been defeated.’[10]


It was not to be so for long. On March 5th a U-boat, U33, was attacked in the Bay of Biscay on the surface at night without any Metox warning. The same thing happened on the following day to U156 the other side of the Atlantic.[11] Both attacks had been carried out by aircraft fitted with the new Mark III centimetric ASV radar. With a wavelength of just 10cm against 1.5m for the Mark II ASV, it could detect submarines on the surface with greater range and clarity. Now a new directive was issued from U-boat Command ordering boats to surface by day only for re-charging batteries and if caught to remain on the surface and fight it out, although to be effective this tactic required the mounting of quadruple anti-aircraft guns which were then in short supply. This delayed but did not stop the passage of the U-boats – the arrival of centimetric radar, which the U-boats had no means of detecting, did make a difference in the Bay Offensive, but it was not decisive.


But concurrent events further out in the Atlantic were at last stemming the attrition rate against merchant shipping with the arrival in sufficient numbers of Very Long Range (VLR) aircraft which effectively plugged the so-called Atlantic Gap. Losses for the eight months from August 1942 to March 1943 totalled 228 ships of 1.36 million tons. With only nine Liberators of 120 Squadron Coastal Command to provide cover in the crucial area between 25° and 40°W during the winter months, Waters and Bailey rightly point out that ‘…the work of these aircraft, though deadly to the U-boats, lacked the dramatic appeal of more spectacular operations.’[12] They go on to suggest that if ever Churchill’s ‘Battle of Britain’ speech applied to others, it was to the aircrew of 120 Squadron. There is little doubt they saw the heroic work of the squadron in support of the convoys a more fitting subject for comparison with the Battle of Britain that Costal Command had attributed to the Bay Offensive.


By March the Northern Ireland based 120 Squadron had been reinforced with a further eight Liberators and with British and US VLR squadrons flying from Iceland, Newfoundland and the Azores in conjunction with both British and American carrier-born escorts. In the first three weeks of May, a further twenty-three U-boats were sunk in the mid-Atlantic by air and surface escort and on 23rd of that month Admiral Dönitz called a halt to the U-boat campaign in the North Atlantic.[13]


Coastal Command’s Bay and Northern Transit Area Offensives did not win the Battle of the Atlantic, but by taking the fight to the enemy, within the range limitations of of available aircraft, the development and successful deployment of the Leigh Light and the skilful use of the Mark II and Mark III ASV radar they did much to demonstrate that given aircraft with sufficient range, the battle could be won.

Extract from Eric J. Grove (ed.) The Defeat of the Enemy Attack on Shipping 1939-1945 , NRS Volume 137, 1997 p. 115.

In the first three weeks of May 1943 the Bay Offensive made five U-boat kills, including a supply U-boat. In the same period convoy escorts destroyed 23 U-boats with the result that the enemy withdrew from attacking the North Atlantic trade convoys on the 23rd of the month. The U-boat Command Records make it clear that this move was dictated by the convoy escorts. The May 1943 casualties in the Bay were the result of the adoption of faulty tactics by U-boats crossing the Bay. These tactics were persisted in throughout June and July and there were many casualties. In June two U-boats were sunk and seven damaged; in July eleven were sunk and six damaged; and in the first two days of August four were sunk. On August 2nd the enemy abandoned these faulty tactics. Though no technical antidote to the cause of surprise attacks by aircraft had been evolved, the U-boats by correct passage tactics were thereafter almost immune to attack. Except for this enemy error the Bay Offensive would not have inflicted these casualties. Like the Northern Transit Area patrol it was begun originally as a deliberate policy, ‘ a new conception.’ Both were in fact analogous to the patrol scheme of ‘ the ingenious Mr. Richard Hall Gower of 1811 ‘ noted in Chapter 1. One of the arguments put forward in 1942 to justify the intensified Bay Offensive was a comparison between the surface area of the Bay of Biscay with that of the Atlantic. Operational research was at this time only nascent. The collection of accurate data was difficult, nor could it be certain that it was complete. In addition lack of precision in the definition of the data used resulted in statistical analyses that tended to obscure the results of operations. For instance, statistics of shipping losses grouped together ships sunk under escort and ships sunk ex-convoy and so gave a false measure of the defensive value of sailing ships under escort; by failing to indicate whether ships sunk in convoy had air as well as surface escort, the significance of air escort was further obscured.

[1] Despatches on Operations of Coastal Command, Royal Air Force in Grove (ed.), The Defeat of the Enemy Attack on Shipping 1939-1945 (Aldershot, Ashgate for Navy Records Society, 1997), p.100

[2] A less contentious version of the Dewar brothers’ Staff Appreciation was published without attribution as Narrative of the Battle of Jutland (London HMSO, 1924).

[3] The proceedings of Exercise Trident were issued to the fleet as CB 04520/1 which are held at the National Archives under ADM 239/489-90. This quote is from Folio 55 in Eric J. Grove (ed.), Introduction p. xiii.

[4] Ibid. p. xvi

[5] Ibid.

[6] Ibid. p.9

[7] Coastal Command Review, 1941, ibid. p.102

[8] Ibid. p.104

[9] See Table 7 ‘Hours Flown, Aircraft Lost and U-boats Sighted, Attacked and Sunk by Shore-based Aircraft Operating in Atlantic (North of Latitude 40°N, East of Longitude 45°W, August 1942-November 1943’ in Grove (ed.), p.478

[10] Coastal Command Despatches, ibid. p.106

[11] U-boat Command Diaries for March 1943, ibid.

[12] Ibid. p.105

[13] Ibid. p.115

One comment on “The Bay Offensive: Success or Failure?

  1. Paul H on

    Very interesting, David. The defensive nature of the convoy system and of Coastal Command did not resonate well with either naval tradition or the personality of our Prime Minister. Persistent aerial attempts to defeat the U-boat at source were never very successful. The Valentin post-war submarine pen attacks using Disney bombs shows their futility. I believe that the important role of escort carriers also deserves much more credit. There will never be any data to show how many U-boat attacks were foiled by Fleet Air Arm biplanes loitering ahead of each convoy. Do you have any information on the ‘Pumpkin’ Leigh Light as tested on the Fairey Swordfish?


Leave a Reply

Your email address will not be published. Required fields are marked *

About David Craddock

David Craddock began his working life in the Merchant Navy as a cadet with P&O. He came ashore to study graphic design and later a BA in History with the Open University. Over five decades he has combined both interests with a career in exhibition and graphic design specialising in historic interpretation. He is a Trustee of the Britannia Museum at the Royal Naval College Dartmouth and now devotes most of his time to writing. His first book What Ship, Where Bound? A History of Visual Communication at Sea was published by Seaforth in February 2021 and was well received on both sides of the Atlantic.