Enigma enigmas during PQ17

From the Washington Naval Treaty to the end of the Second World War.
User avatar
Dave Saxton
Supporter
Posts: 3086
Joined: Sat Nov 27, 2004 9:02 pm
Location: Rocky Mountains USA

Enigma enigmas during PQ17

Post by Dave Saxton » Sat Mar 07, 2020 6:27 am

One of the puzzling aspects to Ultra Intel during PQ 17 was that PQ 17 occurred during July 1942 but the German navy as a whole had gone to the M4 Enigma no later than Feb 2 1942. The literature generally agrees that this resulted in a naval Enigma black out of M4 messages until Dec 1942. Now compounding this puzzlement are GkDos documents which indicate that Tirpitz, S&G, and PG were instructed to use M4 Neptune (referred to as Barracuda by Bletchley Park) for War Ops beginning Oct 23 1941. Denning and Hinsley have both admitted that Neptune was never broken although M4 Triton (Shark) used by the U-boats was broken episodically from Dec 1942.

Nonetheless, we have several instances that indicate Enigma intercepts played vital roles in the cases, of PQ12 avoiding destruction by Tirpitz (March 1942), Pound's decision to scatter PQ17, and of course during North Cape.

Pertaining to PQ17 there were some vital contributions given by Ultra:

*Prior to the PQ17 operation, Denning, gave an assessment in an OIC briefing that the Germans were planning a major operation involving the Tirpitz against the next PQ convoy. (Denning late in life explained that the source was Swede Intel, although at the time he did not know that the land lines through Sweden were tapped.)

*An Enigma intercept indicated that the Luftwaffe had located the convoy and were shadowing it.

*Ultra gave their opinion that Tirpitz was at sea based on a Enigma intercept from Admiral Carls to Fleet Commander Schniewind who was aboard Tirpitz.

*An Enigma intercept indicated that the cruiser covering force had been located and reported to Schniewind but that it incorrectly informed that the force included a battleship.

*It was reported that the Tirpitz sent a signal at 0740 July 4 to Kummetz to prepare to refuel destroyers as soon as the TP battle group arrived at Alta Fjord within a couple of hours.

*It was reported that a signal was sent to Tirpitz that Tovey's battle group and carrier had been located but was way too far away to intervene if Tirpitz sortied against PQ17.

There are some things which could explain these enigmas. One is that the Luftwaffe Enigma was not secure and was typically broken daily. Obviously, some of these messages in question originated with or were rebroadcast by the Luftwaffe. Another, was that in Norway, the Kriegsmarine continued to use M3 Engima, known as Hydra ( Dolphin by the British) which was broken mid 1941 but began using new bi-grams Nov 1941. It was used by minesweepers, tankers, and supply ships, even after the advent of Triton and Neptune. M4, and particularly Neptune, were typically only used by warships and submarines at sea. There was also a dock yard code (not a cypher) that the British had long since broken, that often rebroadcast Enigma messages in part or in whole, or alluded to Enigma messages. A great source for cribs.

In the case of N. Cape, the British were given a heads up that Scharnhorst was preparing to sail because Bey sent messages to minesweepers to pickup two pilots and bring them to Lang Fjord. They were later able to confirm that Scharnhorst was at sea through subsequent traffic analysis.

likewise, the text of the message from Carls to Tirpitz during PQ17 was never decyphered, but traffic analysis left no doubt that Tirpitz was at sea by the virtue that it was sent via M4 Neptune instead of by landline.
Entering a night sea battle is an awesome business.The enveloping darkness, hiding the enemy's.. seems a living thing, malignant and oppressive.Swishing water at the bow and stern mark an inexorable advance toward an unknown destiny.

User avatar
wadinga
Senior Member
Posts: 2052
Joined: Sat Mar 12, 2005 3:49 pm
Location: Tonbridge England

Re: Enigma enigmas during PQ17

Post by wadinga » Mon Mar 09, 2020 1:12 pm

Fellow Contributors,

Hi Dave. You are identifying many of the inconsistencies in the information coming from sources sometimes relying on memory and incorporating generalisations. These are my musings:

Neptun, called Barracuda by the British was introduced in May 1941, apparently specifically for the Bismarck operation, and so is not intrinsically associated with the four rotor Enigma. Whereas there was a standard set of settings to which the machine was adjusted for regular traffic, so there was an "Officer only" set of changes to these settings, to which the regular operator was not party, and handed sole access to the Enigma to the officer. He set up the machine, did the encoding/decoding himself, reset the machine to standard settings so the operator could resume standard traffic. Aside from this was a special new set of settings, solely reserved for the highest level communications. With Fleet Commander Lutjens onboard Bismarck, this could surely only be Neptun, but I have never seen this confirmed.

Generalised statements like "Neptun was never broken" are common, but are surely questionable, as equally Hinsley in the HMSO Official History says all the traffic to and from Bismarck was broken within "a few days" after her demise. There was continuous communication between Bismarck and shore stations throughout the chase including Gunter's long-winded report and multiple retransmissions with changed daily settings, giving unique opportunities for breaking. However like all the other settings Triton etc, once the settings in Neptun were changed, the Bletchley Park people had to start from scratch again. Neptun's particular protection stemmed from the fact that German warships spent most of the war sitting around in harbour, connected to more or less secure landlines, so there was no Radio traffic for the Y stations to intercept, and therefore nothing like the volume of content necessary for the decryption techniques to work.

As you observe German Air Force Enigma was far less secure and routinely broken quickly by Bletchley Park. Having visited BP just a couple of weeks ago, I saw a fancy light-up display still maintains the "old timer's story" that BP helped by indicating France as Bismarck's destination, without mentioning that the information from Jeshonnek's indiscretion came far too late for Tovey to act on it, as he was already trailing so far astern. It was Force H's intervention than sealed Bismarck's fate, and BP should not be claiming any success there.

I am ploughing through Hugh Sebag-Montefiore's description of the scraps of Enigma information available in the PQ 17 operation, and the precise timing of when scraps were available. Will revert.

All the best

wadinga
"There seems to be something wrong with our bloody ships today!"

User avatar
wadinga
Senior Member
Posts: 2052
Joined: Sat Mar 12, 2005 3:49 pm
Location: Tonbridge England

Re: Enigma enigmas during PQ17

Post by wadinga » Tue Mar 10, 2020 12:25 am

Fellow Contributors,

For Dave, supplemental.

Hinsley is contradictory himself about Neptun. In British Intelligence in WWII Vol 2 Appendix 4 in a table he says Barracuda/Neptun was in use from May 1941 onwards to the end of the war. In the same volume p 216 speaking of the run up to PQ 17 he says a message was decrypted from the Scheer TOO 1158 3rd July 1942 telling Narvik she could not decipher two signals on the Kootwyk frequency. BP couldn't decipher them either, and there was no indication it was an Offizier message, which they could handle. He says in this footnote:
They may have been indecypherable either because of an error on the part of the W/T operator or because they were encyphered using a new four wheel Enigma key that was coming into use for communication between Flag officers during fleet operations. Named Barracuda at GC&CS and Neptun by the German Navy, there is some evidence that this key, which was rarely used and never broken, was made available to the Fleet in time for the intended operations against PQ 17.

We then jump to 12th March 1943 where on p 534 Hinsley again mentions Neptun in a context where GC&CS cannot decipher signals between Tirpitz, Lutzow and various naval commands. Basically anything that you can't decipher must be in Barracuda/Neptun.

Returning to the PQ 17 operation, Hinsley says BP is decrypting Kriegsmarine Enigma but is blacked out by a settings change (probably wheel settings) for the period 1200 1st July to 1200 3rd July but read by 0500 on the 3rd. Another change at 1200 3rd July makes nothing decipherable until 1837 on the 4th July. At the end of the footnote above he says
But apart from the two received on the morning of 3rd July [ie those also unintelligible to Scheer] no indecypherable signals were received at GS & CS during that operation.
ie no mysterious "Barracuda" signals.

Your reference
*Ultra gave their opinion that Tirpitz was at sea based on a Enigma intercept from Admiral Carls to Fleet Commander Schniewind who was aboard Tirpitz.
This surely refers only to the move up to Altenfjord and not the actual movement again PQ 17.

Only at 1918/4 was Operations Information Centre of the Admiralty able to inform units that Tirpitz had arrived Altenfjord at 0900/4 and that her escort had been ordered to refuel immediately. Presumably this is the 0740 message you mentioned above. BP broke the new Kriegsmarine settings commencing 1200/4 as early as 2000/4 but despite reading U-boat reports together with various GAF Enigma Luftwaffe messages there was no definite information re Tirpitz. It was in this information vacuum that Pound went against OIC and BP opinion who both thought the lack of warnings to U-boats and Luftwaffe to look out for friendly units meant the battleship had not sailed, and so the First Sea Lord put the scatter plan into operation. When Denning of the OIC was asked by Pound whether he would state absolutely that Tirpitz had not sailed, the former would not stake his reputation on his opinion based on a lack of indications she had sailed. Pound went for the "worst case scenario" and assumed that whilst maintaining radio silence, Tirpitz had sailed as soon as the destroyers were ready.

When Tirpitz, Hipper and Scheer did eventually sail, after the German evaluation they would not run into Tovey or carrier air attack, it was only after receiving Hitler's express permission early on the 5th July. The Flagship's report timed 1145/5 that she would leave via Rolvsoysund at 1430 on 5th July was decrypted by BP/GC&CS and relayed to the Home Fleet at 1517/5. Therefor even this critical top level signal was not sent in Barracuda/Neptun.

The reports that mistook the two funnelled cruiser "London" for a battleship in Hamilton's cruiser force came from Luftwaffe aircraft, subsequently contradicted by later reports, but creating uncertainty, and all in GAF Enigma decrypted by BP relatively quickly.

Finally in Volume 3 of his Meisterwerk Hinsley says in a footnote to the Scharnhorst sinking:

In addition to Offizier signals [which it was reading] twelve messages which were completely indecypherable were intercepted between 2130/25 Dec and 1344/26. GC&CS judged that they were encyphered in the special key reserved for fleet operations (Barracuda).

Gradually things become a little clearer but secrecy between departments at the time, intelligent people making incorrect deductions and the confusion of memory when the veil f secrecy was eventually lifted means there will always be ambiguities.

All the best

wadinga
"There seems to be something wrong with our bloody ships today!"

User avatar
wadinga
Senior Member
Posts: 2052
Joined: Sat Mar 12, 2005 3:49 pm
Location: Tonbridge England

Re: Enigma enigmas during PQ17

Post by wadinga » Wed Mar 11, 2020 8:38 pm

Hi Dave,

Further thoughts.

It would be interesting to get the verbatim details on this information
Now compounding this puzzlement are GkDos documents which indicate that Tirpitz, S&G, and PG were instructed to use M4 Neptune (referred to as Barracuda by Bletchley Park) for War Ops beginning Oct 23 1941.
They may have been instructed to use the four rotor M4 in three rotor emulation mode whilst operating in the Atlantic net (if they ever got the chance) even before Triton was adopted in early 1942. The Arctic net stuck with Heimlich/Hydra /Dolphin to the end, whereas when sufficient Atlantic U-Boats were equipped ready they used Triton/Shark, a new key for four rotor machines. Operating in the Arctic against PQ-17 in mid-1942, Heimlich was still in use, so BP could decipher with variable delays imposed by wheel and settings changes.

As this excellent page clarifies, four rotor machines were often used in more easily broken three rotor mode:-

https://uboat.net/technical/enigma_breaking.htm

BP could tell the difference between Offizier keys, versions of the regular (Allgemain) version key in use, and the completely indecipherable Bogeyman Neptun/Barracuda which was used so infrequently that there was nothing like enough volume of material to give them a chance at breaking it. Hinsley apparently suggests that the Scharnhorst sinking was the first occasion they really came up against a number of messages in Neptun/Barracuda. Hence it never was deciphered.

All the best

wadinga
"There seems to be something wrong with our bloody ships today!"

Post Reply