The ‘HX’ Convoy System

After my post describing the ‘slow’ SC convoys, I received many requests for information about the other series of east-bound convoys, the ‘fast’ HX series, which many Canadian warships escorted during the Second World War in the Atlantic. Here is some analysis about them.

The first convoy, HX-1, was a 15-ship group that departed Halifax on 16 Sep 39 (Canada’s Declaration of War occurred in parliament on 09 Sep and was given Royal Ascent by the Governor General a day later). HMCS St. Laurent (H89), LCdr. Arthur Mitchell ‘Boomer’ Hope, RCN, CO, and HMCS Saguenay (D79), LCdr. George ‘Gus’ Ralph Miles, RCN, CO, both River-class destroyers, formed the close escort as the convoy departed port. All cargo ships in HX-1 arrived safely in Liverpool on 30 Sep after a 14-day voyage.

In the early stages of the war, the close escort only remained with the convoy long enough for it to clear the ‘submarine danger zone’ around the approaches to the port of Halifax, about one day’s sail. The ‘ocean escort’ for early convoys was usually comprised of a heavy cruiser or an armed merchant cruiser. Later, Resolution-class battleships were attached singly to the HX convoys due to the danger of attack by German major warships. These were attached as a deterrent against attacks by German surface warships or auxiliary warships. They were not present for defence against submarines. A local escort group of minor warships would meet the convoy approximately a day’s sail from the destination and escort it into port. This distance was increased as the danger area from enemy action expanded.

All HX convoys were destined for Liverpool, a voyage of approximately 2,500 miles. The average speed of HX-1 was 7.5 knots. The last convoy in this series was the 56-ship HX-358, which sailed from New York on 23 May 45 and arrived at Liverpool on 06 Jun 45, also after a 14-day voyage. The greater distance, 3,165 versus 2,500 miles, resulted in an average speed of 9.4 knots. With the exception of the convoy system of the east coast of the U.K., the HX series was the longest continuously running convoy series of the war. Originally, the convoys were split into a slow group that sailed at the prescribed nine-knot minimum for inclusion and a faster (HXF) group of ships that were not capable of the 15 knots required for independent sailing. The HFX convoys delayed sailing by a few days and the two groups would amalgamate before entering the submarine danger zone in the Eastern Atlantic. Eventually, the dwindling number of fast merchant ships and the expanding submarine danger area brought an end to the HFX series, of which there were 17 convoys. The last one sailed from Halifax on 02 December 40 and no ships were lost from any convoy in this sub-series. Beginning with HX-208 in 17 Sep 42, the port of departure for HX convoys was shifted to New York (strangely, without a change to the series name), and the RCN concentrated on escorting the SC convoys. Once moved to New York, the average size of a convoy increased from 38 ships to 56 ships, whereas the average size of SC convoy remained at 38 ships (the opening out of the convoy schedules by 30% was identical for both ports), indicating that logistical limitations in Halifax played a part in determining convoy size. HX-300 was the largest convoy in the series with 160 ships (convoys of over 100 ships were called ‘monster’ convoys) while HX-4 was the smallest with only ten ships.

In total, 10% of the HX convoys were attacked while they were in formed groups. Of the 110 ships sunk, 96 were claimed by U-boats, five by the armoured ship Admiral Scheer, and the remaining 14 were lost to mines, aircraft, collisions, groundings, and storms. These losses totalled 0.6 percent of the 17,744 ships in the convoys. The worst convoy battle in this series occurred in March 1943 when HX-229 lost 12 of its 38 ships (31.5%), all to u-boats. A further 60 ships were sunk that had straggled behind HX convoys. Another 36 ships were listed as ‘losses out of convoy’, which were ships sunk before the ocean escort arrived, or after detachment from the convoy while en route to their various ports of destination, or after a convoy had been scattered due to attack. This brings the loss rate to 1.1%, which was considered sustainable (3% was considered unsustainable). A further 38 ships were damaged while in convoy.

HX-84 was dispersed on 05 Nov 40 when threatened by Admiral Scheer, whose five victims from that convoy included the Canadian Pacific steamer SS Beaverford (10,042 GRT), Captain Hugh Pettigrew, Master; a fast and multi-purpose ship that was one of the prides of the pre-war CP fleet. An independently routed merchantman, plus the armed merchant cruiser HMS Jervis Bay were also sunk. HX-168 was dispersed in Jan 42, likely due to the same adverse weather that caused SC-62 and SC-63 to be dispersed, but all of these 36 ships also arrived in Liverpool safely. Adverse weather made attacks by u-boats far less likely.

The HX convoy series is held up as a model of success based on the gross numbers. However, the HX convoy system relied heavily on evasive routing to prevent losses, just as did the SC convoys and all others. Of the 377 HX/HXF and BHX (a single ‘joiner’ convoy from Bermuda) convoys, for the 38 convoys that were engaged, the actual sinking rate for the entire war was 4.85%, which does not include the ships damaged (6.44%). The loss rates (sunk and sunk plus damaged) in the ‘critical period’ up to Mar 43 were 5.92% and 7.45%, respectively for intercepted convoys. Spates of enemy successes created what Stephen Roskill call “real panic” in the Admiralty: convoys HX-70 to HX-74 were all hit in Sep 40; seven of the 16 HX-107 to HX-117 convoys in Apr 41 were attacked. Clearly, the rate of loss from intercepted convoys was unsustainable. This casualty rate explains the desperation of the Allies at certain periods of the war when German naval signals could not be deciphered, particularly in 1942.

Admirals Raeder and Doenitz both strongly resented Hitler’s interference with the anti-shipping campaign when he ordered U-boats be diverted to the North Sea for the invasion of Norway, and for operations in the Mediterranean Sea and the Arctic Ocean. Without these respites and the benefits of signals intelligence, the rate of loss would have been substantially higher. The claims by a number of post-war authors that the Battle of the Atlantic was never in doubt was not an opinion held in either the British or Canadian naval headquarters. The statistics show that signals intelligence, both good and bad planning, plus strong and weak leadership all played critical roles in the eventual outcome.