Crest of the Royal National Institution of Naval Architects - Click to return to the homepage

The Royal Institute of Naval Architects

Alfa Lavel ( Aug )


Technical Meeting -- 3 February 2016


Geoff began his presentation by saying that he became interested in diesel engines for submarines in his PhD project in the UK. The Oberon-class submarines were in their mid-life refits, and their engines had serious lubrication problems¯submarine engines are in a difficult environment.

The presentation gave an introduction to the issues experienced in using turbocharged diesel engines in submarines, especially those destined for blue-water operation, and referred to some submarine-specific historical aspects from Lyle Cummins’s 2007 book, Diesels for the First Stealth Weapon: Submarine Power 1902­45.

The Collins-class submarines are each powered by three Hedemora turbocharged V18 engines. Kockums had designed and built boats with similar engines before, but these operated largely in the Baltic, which mostly provides somewhat more benign sea states than the Indian and Pacific Oceans in which the Collins-class vessels live. They were V12 engines of smaller bore.

The Collins class vessels have experienced significant problems when snorting in rough seas. With underwater exhaust at the fin top, exhaust back pressure varies considerably and, if a wave washes over the snort mast and closes the induction valve, then the crew and engine both suffer discomfort. This led to some reliability issues, hopefully now considerably improved by careful management.


Diesel Engines Breathing in a Snorting Environment

Diesel engines, turbocharged or not, need a lot of air. The three engines in the Collins-class vessels each breathe in several cubic metres of air per second. When the snort mast dips below the water, a valve closes and the engines breathe from the engine room, reducing the pressure uncomfortably quickly. If the pressure drops to an extent hazardous to the crew, then the engines stop, disrupting the battery charge. Waves over the subsurface exhausts also cause trouble, varying the exhaust back pressure, and hence affecting turbocharger speed and boost pressure.

In Sea State 3, the significant wave height is less than 1 m; however, in Sea State 6, the significant wave height is 5 m, varying the exhaust back pressure by 50 kPa (half an atmosphere). In practice, snorting in Sea State 6 is something to avoid, short of desperately flat batteries!


NSW 3 Feb 1

How a submarine engine sees the sea state when snorting

(Drawing Phil Helmore)


Other ideas have been suggested and tried. The K-boats in WWII used steam turbines for performance, which gave the best-available power/weight ratio at the time, and there was a big snort mast and exhaust funnel to close off when submerged. However, they were a disaster, and all are believed to have sunk as a result of technical issues, not enemy action! Crews seem to have regarded a posting to one of them as a suicide mission!

Gas turbines have also been suggested, with very high power/weight ratio, relatively quiet but not efficient at sea level; back pressure makes this much worse, and they need enormous amounts of air, for both combustion and cooling. No practical arrangement has emerged.

So the diesels have it, in terms of heat engines at least.

The snorkel had been around from a Dutch innovation in the early 1930s, and had been found on a couple of Dutch submarines impounded by the Germans early in the war. Likewise, a couple of these boats had also made it to Allied havens, but the significance of the snorkel gear, which was used for ventilation, not necessarily to run main engines, was apparently not realised for several years. The snorkel (what we know as a snort mast), seems not to have come into service for running engines until later in WWII.

Common-rail Fuel Systems¯New or Not?

Diesel was clearly preferred for submarines before WWI. The fuel was less volatile than petrol or other available fuels, so the vapour was less dangerous in the event of leaks. Fuel was supplied to all cylinders from a common pump, and each injector was operated by compressed air, known as ‘air blast injection’. So early systems were effectively ‘common rail’, but not using the extremely high pressures used in modern engines.Air at around 70 bar (7 MPa) was used, and fuel was also supplied at high pressure. These systems were energy intensive and quite dangerous in the event of a leak. Rudolf Diesel is understood to have tried to produce a fuel-injection system not assisted by compressed air, without success. Brandstetter had a 1905 UK patent for an airless injector, using a spring-loaded accumulator. L’Orange designed a mechanical injection system for Deutz, with a patent dated 1908. Vickers made the real breakthrough in 1910 with a patent for a hydraulic injection system, the first to be known as ‘solid injection’.

E-class Submarines for the RAN

HMA Ships AE1 and AE2 were E-class submarines for the Royal Australian Navy (the “A” indicating Australia) and were delivered just before World War I. The E-class boats were submersibles, designed to spend most of their time on the surface. To charge batteries, the boat was surfaced and the hatches opened. This was the usual mode of operation until well into WWII.

The engines were large, taking up most of the hull cross-section in the machinery space, and each drove a generator to charge batteries, but also a shaft to the propellers. This arrangement, as opposed to the diesel-electric mode we expect now, also persisted well into WWII.

NSW 3 Feb 2

Vickers engines in submarine E23 around the beginning of WWI

Note lots of exposed moving parts and pipework joints

(Photo courtesy DST Group)


The Bosch Injection Pump

In 1922, Robert Bosch entered the fuel injection business. He designed a ‘solid injection’ pump, with camshaft-driven pressure and timing, and designed to be supplied with fuel at low pressure. The pump and injector were separate items, one high-pressure pump per cylinder, and volume production began in 1927. It was so successful that it became a standard for more than 50 years. Almost everyone used Bosch’s ‘solid injection’ system under licence, or copied it with variations to evade patents. Vickers, however, persisted with their own designs until they built their last engines in 1943. From this point on, there were few high pressure ‘common rail’ fuel systems in volume production until the evolution of electronically-controlled systems in the 1990s.

Performance of Turbocharged Diesel Engines in a Snorting Environment

Turbocharged engines have been accepted best practice for all kinds of shipping, for many decades. They are the most economical power source in both power/weight ratio (at least among piston engines) and in fuel consumption. However, they present problems for submarines because they don’t respond well to dramatic changes in pressure, especially exhaust back pressure.

Turbocharged engines for submarines appear to have been introduced by Gustav Pielstick, chief designer at MAN, in the Type IX U-boat for the Kriegsmarine (the Navy of Nazi Germany). Those boats did not snorkel, but even on the surface they experienced a phenomenon known as “the following sea problem”. When waves rolled over the casing, the exhaust was submerged, the back pressure went up, the turbines slowed down, the pressure in the hull fluctuated wildly, and sooty exhaust leaked into the boat, causing discomfort at best for the crew. This was a problem for all the boats, but much worse when they tried turbochargers. The exercise was repeated in the Type XXI U-boats with limited success, once again beaten by varying pressures.

The MAN M9V 40/46 of 1939

This diesel engine was a 9-cylinder version of the MAN 40/46, having a bore of 40 cm and stroke of 46 cm, giving 58 L per cylinder (522 L total ¯ i.e. a big engine) and an output of 1640 kW at 520 rpm. At the time, Pielstick was also working on a higher-powered version, with a gear-driven supercharger prior to the turbocharger, which was supposed to be good for over 2.2 MW. This may not sound all that impressive today for an engine of over 500 L, but it was at fairly low speed. Interestingly, the OKM (Oberkomando der Marine­German Navy High Command) couldn’t see any need for that much power and the “supercharged and turbocharged” project was cancelled. For the turbocharged engine, the valve timing was changed to reduce the valve overlap, so that any water in the exhaust could be blown out during starting and to help cope with the fluctuating back pressure. Boost pressure was not very high because of the back pressure ¯ the turbocharger would have needed plenty of margin to cope with the variable pressures. However, the reduced valve overlap caused the scavenging to be poor and exhaust temperatures unusually high for a moderately-powered engine. The turbocharger was fairly quickly replaced by a gear-driven mechanical supercharger.

The MAN M9V 40/46 of 1941

The MAN M9V 40/46 of 1941 had the gear-driven mechanical supercharger fitted to resolve the “following sea” problem. This produced a successful submarine engine that coped with the conditions, but a price must have been paid in fuel consumption and therefore range. However, at this stage of the war the submarine could safely sit on the surface to charge batteries, and could transit on the surface, only having to submerge if something appeared over the horizon.

A U-boat was small enough that it could escape detection unless an aircraft happened upon it and, even then, given the speed of the aircraft around, if they were keeping a sharp lookout, the boat would have had time to dive with minimal risk. The German Navy kept the submarines at sea for longer periods using the ‘milch-kuh’ (milk cow) submarine tankers, much as we extend the range of fighter aircraft today.

All these vessels seem to have lived on the surface, only submerging to hide from reconnaissance aircraft or to prepare for a stealth attack.

Royal Navy Supercharger Philosophy

The Royal Navy also adopted the supercharger solution for all of its submarines, both for diesel electric submarine power and later for emergency gen-sets in nuclear boats, right up to and including the Upholder class, built in the 1980s.

As a result, for the Upholder class, they took the turbocharged Paxman Valenta minehunter engine, which also powers other vessels and was one of the commoner railway locomotive engines (the HST/XPT had these, though they have now been re-engined with MTU engines), but specified supercharging instead of turbocharging. They now have a unique engine in their submarines (there are about ten in the world!), something that seems to happen to many of us in the submarine community…

Closed-cycle Diesel Engines

The first closed-cycle diesel (CCD) seems to have been proposed at Stuttgart University in 1940, and the idea was sufficiently developed for a complete system to be installed in U798 in 1941. This was an exquisite Daimler-Benz V-20 engine of 1.1 MW. It used high-pressure oxygen tanks for oxygen storage, and oxygen was added to the exhaust gas and recirculated. The excess exhaust was cooled and pumped overboard. However, the project was cancelled by the OKM because of “competing priorities”!

Several attempts to develop CCD have been made over the years, including one involving BAE Systems (descendant of Vickers), and at least one of these systems used argon as a ‘padding gas’ to improve thermal efficiency. However, overall there are systems with more advantages. Kockums of Sweden used Stirling engines successfully, and Japan bought the Kockums system, but fuel cells seem to be the front-runners for air--independent propulsion (AIP) now.

MAN M6V 40/46 with KBB Turbocharger

The turbocharged submarine diesel engine was also tried in the Type XXI U-boat. By 1942, Dönitz was becoming concerned about the loss of tactical advantage because of the low submerged speed and, of course, as radar emerged as a threat, their losses mounted alarmingly, and an engine that could snorkel to charge batteries whilst submerged became important.

Pielstick’s combined supercharger/turbocharger solution was back on the agenda and, by 1943, , the M6V engine fitted with both gear-driven supercharger and KBB turbocharger (built by Kompressorenbau Bannewitz GmbH) and an aftercooler, appeared for the Type XXI U-boat. This engine claimed a power output of 1.5 MW at 520 rpm (cf 1640kW for the M9V turbocharged version), and the impressive fuel consumption of 214 g/kWh.

The MAN engines in the U-boats were originally directly coupled to the shafts. Reversing was accomplished by stopping engines, shifting the camshaft to switch the valve timing, then restarting the engine backwards.

The Type XXI was an electro-boat, with a big motor on the same shaft, which doubled as submerged propulsion and as generator for battery charging when driving with the diesels. This could also reverse the boat, so the second camshaft position wasn’t needed for reversing. Pielstick kept the camshaft shift gear, and arranged for two different timings, with 150° overlap for surface running and 44° for snorkelling, when the speed was limited by the masts and less power was needed. I suspect the impressive fuel consumption was only achieved with the 150° timing.


Looking back at the MAN M9V 40/46 in comparison, it was a big engine (522 L) and at 520 rpm and low boost pressure, air consumption would have been about 3­4 m3/s for each engine. The boat was of 1819 t submerged displacement, so about half the size of the Collins class, and the engines were consuming about the same amount of air, so the internal air pressure would fall about twice as fast if the snort mast closed. The snorkel needed a valve like a modern snort mast to avoid flooding the boat if a large wave came over it, so the pressure would drop very rapidly if this happened.

Turbocharged diesel engines seem to have disappeared from diesel-electric submarines in the western world and not reappeared until the 1980s. What do we learn from this? There’s nothing new under the sun (or sea)…

Technology that didn’t work before, might work later because we know a bit more about how things work. This applies to both common-rail fuel injection and, at least for the submarine community, turbocharging.

In comparison to the MAN M9V 40/46, modern submarine engines tend to be relatively high-speed engines (1300­1800 rpm):

  • Hedemora VB is 210×210 mm (7.27 L per cylinder)
  • Pielstick PA4 is 200×210 mm (6.6 L per cylinder)
  • MTU 396 is 165×185 mm (4.0 L percylinder)

The Collins-class Submarine Engines

Why do we have the engines we do in the Collins-class submarines? Basically, they were specified in the early 1980s and the contract was signed in 1987, with off-the-shelf or low-risk items specifically required.

Pielstick had, by 1983, decided on a combined turbocharged and mechanically supercharged system, but it had not yet been tested. It promised significantly poorer fuel consumption than a turbocharged engine but better than a pure supercharged engine, and claimed more stable running when snorting in a seaway.

MTU had demonstrated the 12V396SB configured for snorting, but was said to be good for <1 MWe for a then-untested 16V396SB version.

Hedemora had demonstrated the V12B configured for snorting, and promised that the V18B would work as well giving 1.4 MWe. However, they had never actually built a turbocharged V18B14SUB, but had built lots of V18B engines for industrial power and marine generator sets, and had tested a V12B against submarine-type conditions. They had built a number of V12A (smaller bore) submarine engines for the Royal Swedish Navy, the latest examples being turbocharged. Therefore the V18B14SUB was accepted as an off-the-shelf design, which it really wasn’t. But then, nothing else seems to have met the specification either.

Now we have all 19 of the Hedemora V18B14SUB engines in the world! The V indicates the vee configuration, the 18 is the number of cylinders, the B is the larger bore (210×210 mm), the 14 is the speed (1400 rpm), and the SUB category is the monolithic engine (not bedplate mounted) for submarines.

NSW 3 Feb 3

Hedemora V18B engine at the Submarine Training and Systems Centre, WA

(Photo courtesy DST Group)


A supercharged engine of the time might achieve 275­295 g/kWh, combined super- and turbo-charging claims 265 g/kWh, and turbocharged engines achieve about 230­240 g/kWh when snorting. It was felt that a turbocharged engine was needed because it promised significantly better fuel economy than the alternatives. In 1987 a long range was specified, and Kockums was naturally more concerned about achieving the range than about fuel costs and pollution, but we can see that on paper at least, the turbocharged engine is the best option on all counts if it can be made to work reliably. The technology was really 1970s, with individual jerk pumps for fuel injection and a hydraulic governor.

Defence has a research engine, the HAD V6B, built from various parts by Hedemora Australia. The camshafts and the turbocharger are almost the only new parts to special order. This engine was chosen by DST Group, as they expect to be living with it in Collins class boats for another 20 years, and hoped to learn some things that can be directly applied to the Collins class, as well as improving their understanding to place them better to judge the offerings for the next class.

Diesel Engine Technologies for Undersea Platforms

Marine platform (including undersea) lifecycles are of the order of 20­30 years, while those in the automotive and transport industries are of the order of 5­10 years. Given the evolution of technology in recent years, especially in the car industry, why does it take so long to get the new methods into marine platforms? Mainly because there are vastly different market volumes and drivers.

Drivers for technology in the automotive and transport sectors include competition, production cost reduction pressures, development time, product refinement, fuel economy, and the requirement for emissions control. Some of these are beginning to matter in Defence, but are not key drivers.

Additional undersea environmental factors include salt water being present in both fuel and air, dynamic exhaust pressures due to wave action, and dynamic inlet pressures. These things are the keys, specifically for submarines, but are not important or even recognised as issues in other markets.

Modern Common-rail Fuel Systems

So let’s look at what we could get with a modern electronic fuel-injection system.

Modern common-rail systems use extreme pressures, in excess of 2000 bar (200 MPa) and carry significant volumes of compressed fuel in the rails. Double sheathing of all high pressure fuel lines is therefore required. It is essential that high-pressure fuel cannot leak into the atmosphere.

With electronic fuel-injection control, there is an engine-control unit (ECU) like on your modern car engine. It is well suited to deal with transients, controls both the start and end of injection, allows multiple injections, controls fuel consumption and emissions, gives quicker and more accurate response to changes in demand, gives better response for variable ambient conditions and can be optimised for lower noise.

Safety Issues

With fuel supply at around 2000 bar (200 Pa), containment is critical. All components, including pumps, need to be safely contained. MTU have developed (or are developing) a submarine variant of their 4000 series engine (it is mentioned on their website). However, many in the submarine community remain nervous of such high pressure systems.

Performance Issues

Such an engine would require significant work on the control system and turbocharging arrangements. The control system would need load control, not just the conventional speed control. MTU are well capable of this, but one would need to be reassured that they were applying it to cope with open-ocean sea states. Likewise for the Pielstick engines, which may be favoured by a French supplier, although Pielstick is now back in the MAN stable after many years in French hands.

Nuclear Power?

A lobby is in favour of changing to this option. However, there seems to be no immediate prospect of doing so. There is a lack of Australian expertise for support, and a lack of political will to go nuclear ¯ on land or sea!

It is clearly not an option for SEA1000, as tenderers have already been selected. However, the performance comparison is not as black-and-white as the claimants make out. The crew go stir-crazy if they are confined underwater for too long. A nuclear-powered vessel can steam around the world without refuelling, but having to feed and water the crew is then the dominant problem. Nuclear boats are steam-turbine driven and, because all the machinery cannot be shut down, may not be as quiet as a conventional electric boat which can shut down nearly all machinery when necessary. The Collins-class vessels have, at times, sneaked undetected through US defences and been able to pick off US vessels in RimPac exercises.

Battery Storage

Lead-acid still seems quite likely to be a viable contender in the immediate future. Nickel-metal hydride (NiMH) has probably been overtaken. Lithium-ion or lithium-polymer versions also look promising.

However, different battery types cannot be a direct replacement in existing vessels due to differing mass and dimensions, and testing in a Collins class vessel would present real problems.

Other technologies are emerging, and we need to be open to change this time.

Air-independent Propulsion

All tenderers for SEA1000 are known to have some technology in this area. We will likely want something, and a fuel-cell fit shows much promise. All non-nuclear air-independent propulsion possibilities require both fuel and oxygen supplies, which provide another sort of hazard. The timing of the new class, and the options available suggest some flexibility for future development.


We looked at the emergence of the diesel engine as the universally-preferred means of propulsion and power generation in submarines from the beginning of the 20th century until the emergence of nuclear power. The use of pressure charging, the pros and cons of turbochargers and the issues that arise were considered and the emergence of electronic control and the use of common-rail fuel systems. These also have issues for submarine applications, and we have looked at where the technology might go from here, including weighing the pros and cons of nuclear power and non-nuclear air-independent propulsion and power-generation systems.


Question time elicited some further interesting points.

Ceramic liners have never been really looked at for the Hedemora diesels for the Collins class. Hedemora were, generally, open to discussion, and did change a number of aspects of the design. However, this is not typical of most diesel engine manufacturers.

There have been issues with fuel and water mixing in the Collins-class vessels. However, this is so in many navy vessels, both submarines and surface ships. Fuel tanks need to breathe and, in rough seas, they breathe in salt-laden air and sometimes salt water. Also, submarines displace fuel with salt water, and rough seas lead to mixing. All ships require water separators for their fuel systems.

The vote of thanks was proposed, and the certificate and “thank you” bottle of wine presented, by Greg Hellessey.

Survitec July

Bentley Systems


Furuno June 2018

IQPC Smart Ships


UBM April 4th 2018