Turboshaft engine 2PW8


Once again a relatively small turbine engine of Russian origin, the 2PW8 provided the mechanical power in a 75kW genset. I spotted this particular unit on German eBay about a week ago. The starting bid was quite high but in my opinion not completely beyond reason. Due to the size and weight of the unit, and since I’ve already got quite some similar toys, I decided to let this one go. Yet I told my friend Jürgen about it and he was immediately interested. I suggested to let the auction pass and see if anyone would place a bid - I expected noone would. In fact the auction ended without a bid and so I contacted the seller and did a little negotiating on a price that was still more reasonable to Jürgen. So this saturday we drove the 200km to the place the genset was located and picked it up. And now for some photos:

That’s the first view of the genset we took upon arrival. The unit was completely wrapped  in plastic and wax paper. The seller told us that he had to cut the plastic open to get access to the unit. Everywhere bags with desiccant were placed and no rust or corrosion, not even nicks or scratches in the paint were present. This unit was brand-new! Yet it must have been test-run because there was some discoloration in the exhaust section to be found. Due to the construction, the turbine wheel isn’t directly visible, so I decided to use my camera (that I’ve always got handy on journeys like this...) to take a shot of the turbine:

What becomes immediately clear is that the engine wasn’t run a long time for testing, the turbine disc is still shiny and the blades are in great condition. Jürgen had his trailer with him to carry the unit home in case he would actually buy it. But he didn’t have to think twice, considering the condition the unit was in. Fortunately a walk-behind fork lift was available at the place the unit was stored. The lift was barely able to carry the genset on its steel palette, but somehow we managed to place the whole thing on the trailer and secured it.


...ready to take our “prey” home...

Finally home, Jürgen is removing the bags with the desiccant.

Using Jürgen’s workshop crane, the engine is lifted off the trailer and placed on the step in his workshop. In the background, the two Ru19a-300 engines are visible. This was a somewhat risky task because the whole genset is very heavy. The label states 470kg (about 1050lbs).

All the labels were still protected with wax paper that we eventually removed. The label states “AGREGAT TYP AT-75M2-T/230-Tch/400”, the power rating and the weight of the unit. If it had been a 50 or 60Hz device, it might have been quite useful but with its 400Hz, it would be hardly of any civil use except perhaps lamps and (old) ovens...



That’s the label on the turbine engine. It’s designation is 2PW8. Unfortunately there’s very little information available on the web on this engine or the whole genset, but furtunately we’ve finally sourced a complete manual.

Since Jürgen hasn’t got an application for the genset, he decided to remove the engine and get rid of the remains somehow or the other. Yet it’s always a little sad to rip apart a completely good mechanical device. Anyway, we proceeded with disassembly.



That’s a view on the left-hand side of the unit. The large can at the right is a dynamical dust/air separator. It probably serves as an intake silencer as well (who needs a silencer...?? ;-). There are more details on this part further below.

That’s a view on the right side with the exhaust and the electrical connections.

Here’s the alternator end with its own blower, the power connector box and a sheet with the cyrillic alphabet on top ;-).

That’s the top of the engine, the gearbox section at the centre. The cylindrical part o top is a starter-generator to spool up the engine initially and then recharge the starter batteries and supply electricity (28VDC) to the accessories. The large alternator only produces 400Hz three-phase AC voltage.

And now for the details of the air/dust separator. It has got eleven of these stacks of tapered rings. Each of these ensembles of rings forms a number of circular passages that point to the front. So the air that gets drawn in, has to reverse its direction to enter the plenum in front of the compressor.

That’s a rear view of the dust/air separator. Each one of the ring stacks ends in a pipe segment that is connected to a circular cavity from which the air is scavenged throught the oval port at the left. A more detialed photo of the ring stacks is shown below.

Due to the higher mass of dust or sand particles in the air - compared to the air molecules -, these impurities won’t follow the direction reversal of the air (inertia is so kind to help us here) and hence simply flows straight through the ring stacks, being sucked together with some air out of the scavenge port. How exactly this “suction” is being produced, is explained further below. The complete dust filter is a welded construction of steel sheet and is quite heavy. This and the fact that the genset is equipped with such a device at all, leads me to the conclusion that the unit was intended for a ground-bound application.

Here’s a shot of the exhaust port. Quite unfamiliar on single shaft engines is that the gearbox is located behind the hot section of the engine. This requires the use of bleed air for cooling of the cavity between the exhaust volute and the gearbox. All the spill ports are routed into the exhaust. Left of the orifice the exhaust port of the dust filter system is visible.

That’s a view of the fuel control system. It seems to be of the hydromechanical type but I’m not quite sure of the detials. This will probably require most of the effort when we’re going to get this engine operational. At the top of this photo the tacho generator is visible (black), used to signal engine RPM to the control panel.

The compressor end of the engine with the dust separator removed. There’s an oil cooler arranged around the compressor intake, probably a result of the arrangement of the gearbox at the hot section. The two cans at the compressor front contain the oil filters I guess.

After a few hours of work we finally managed to separate the power section from the alternator - without causing any damage! And that’s not a matter-of-course, considering the weight of the components.

The alternator alone weighs some nice 178kg (400lbs) but otherwise seems to be a pretty reliable device. So if anyone is in need of a syncronous generator, type “S-75-406/12”, three phase 230V, 236A, 93.75kVA, 75kW, 8000rpm, 400Hz, as-new condition, please let me know - it will be available cheap (shipping to be covered by the purchaser ;-).

Below there’s a photo of the control section of the alternator. Probably the field coil voltage is generated there and it’s controlled so as to keep the output voltage constant, regardless of the electrical load to the device. It seems the alternator is of the brushless type, so there’s no part to wear except perhaps the bearings after a long time of operation. There’s a set of twelve small fuses at the central section of the control board and a whole lot of power transistors loacted at the other side.


The engine is separated from the alternator. It’s a compact unit but still has considerable size (about 45cm in diameter, 90cm long and weighs a guessed 150kg). Yet it seems to be a completely self-sufficient unit with integrated oil system and tank. The black hose in front of the unit is the oil drain, a nice idea to have it arranged this way, simply place the hose in a suitable container and open the drain valve - no hassle, no mess of spilled oil. If the engine is placed on the gearbox flange (vertically), it reminds me of “R2D2”... ;-)

That’s the power takeoff at the gearbox end of the engine. Since there’s an open shaft seal that fits the alternator shaft coupling, a cover needs to be made that screws onto the thread at the outside of the bearing retainer to keep oil from being spilled from this orifice. I’m sure it was arranged this way to keep both splined couplings of the quill shaft (one in the drive shaft of the engine and the other inside the shaft of the alternator) lubricated with engine oil.

And here are some details that really cought me by surprise. The engine has got a tangentially mounted single-can combustor! From all its appearance, I would have bet it has got an annular or can-annular combustor. But after disassembly of the genset I followed the spark plug wire and found the small dome beneath the dust separator transfer tube with the typical cap and accessories of a single can combustor. Another detail that becomes clear in this photo is the way the dusty air is scavenged from the dust separator: Some compressor bleed air is blown through a number of small bores in a tube that’s located vertically in the transfer duct. This forms an injector pump and “sucks” a large flow of air from out of the dust/air separator with all the dust entrapped.

Unfortunately I haven’t got any specifications of the engine. What seems to be sure is that it has got a single stage radial compressor, single or dual stage axial turbine, single tangential can combustor, hydromechanical fuel system, self-contained lubrication system with the oil tank as an integral part of the gearbox and compressor-intake mounted oil cooler, high energy ignition system with a single spark plug, 28V starter/generator, integrated (epicyclic?) gearbox with an output speed of 8000rpm. The output power will be in the range of 150...200hp, main shaft speed about 30,000...40,000rpm, pressure ratio probably about 4 and air flow slightly more than 1kg/s. If anybody has got specifications or even documentation for this engine, please let me know!

Well, so far for now, I hope we’ll be able to get this beast running soon.


01/25/2004 Meanwile Jürgen was lucky to get hold of a complete documentation for his 2PW8 engine so he was able to do all the wiring to get it into operational condition. He constructed a simple control panel, yet the RPM and EGT instruments are missing. We consider to use either digital gauges modified to display the required parameters or to try to get hold of the original analogue ones. For the time being, we used two digital multimeters.

The black instrument displays EGT in degree centigrade, the larger yellow Fluke is wired up to the tacho generator to give at least an idea of RPM. The manual states that the driving shaft of the tacho generator rotates at 2800rpm at 100% engine RPM. This equals an output frequency of 41.3Hz for a two-pole generator or 82.7Hz for a four-pole machine. Unfortunately, this information isn’t provided by the manual. Since we didn’t want to detach the tacho generator from the gearbox, we simply had to find it out the hard way... The other items on the control panel are as follows: Battery voltage instrument, main power light (white), starter active light (green), engine ready for power-take-off (yellow), low oil pressure (red), main power switch (security switch at the left, automatic start button, cold motoring button, engine stop button (red).

Below the control panel, the high-energy ignition exciter (left, light grey device) and the electro-mechaical startup sequencer and starter contactor (right) are located.

This cross-section shows the internal arrangement of the engine (click the image for a larger TIF version). The compressor (1) is a single-stage radial unit with an intake guide vane system. The combustion section (2) equipped with a single, tangential can-type liner and a simplex (!) burner nozzle. The hot gas flows through a spiral volute to the single stage axial turbine section (4) and the exhaust duct. The casing (3) is cooled by compressor bleed air. The gearbox (5) is a simple spur gear with a reduction ratio of 4.656:1 and it provides mounting pads for the fuel and oil pumps, the tacho generator and the starter-generator. RPM at the power takoff shaft is 8,000. Here are some technical specs:

Rated power output: 120hp (a little under-statement I guess)

100% rotor speed: 37,250 rpm

Maximum continous EGT: 650°C

Maximum intermittent EGT: 880°C

Maximum time for startup: 120s

Weight including accessories: 232kg

Air mass flow: 1.8kg/s

Compressor pressure ratio: 3.0

Specific fuel consumption: 750g/hphr (=approx. 1kg/kwh...efficiency?!?)

Jürgen is excited just before the first real start-attempt of his engine. He cold-motored it several times before, checked the oil supply and bled all the air from the fuel system. I was pretty much excited as well, especially since we didn’t know the exact freqency that equals 100% rpm. Pretty scary...

And then...cold motoring for a short time...automatic start activated...ignition. The Russian bear broke loose! It’s hard to describe the intense sound this engine produces. During accelerating from about 30...70%, the compressor produces a ground-shaking hum to burst into a terrific roar afterwards. The hum somewhat reminds me of an Allison 250, but this beast is much louder. It’s simply evil. I’m pretty sure, if I ran my GTP30 side-by-side with this monster, the Garrett wouldn’t be heard at all. I guess the 2PW8 emits at least half of its power as noise... ;-).

Now here are four video clips of the runs. I adjusted the camera a little bad so our heads are clipped in the first two videos (which isn’t a real loss...). The first start (1.6MB) was terminated at approx. 70% rpm as the exhaust flux started to stir up the small stones the place in front of Jürgen’s hall is paved with. I was afraid this debris would be drawn into the compressor so I signalled Jürgen to stop the engine

For the consequent starts, we placed a sheet metal plate on the ground in front of the exhaust port. The second start (1.9MB) was continued up to governed RPM. Due to the extreme noise and the excitement, we simply forgot to memorize the instrument readings. So we conducted a third run (1.7MB), this time the camera pointed at the control panel. It becomes clear that the tacho generator is a four-pole machine since governed RPM equals approx. 82Hz. EGT maxed during startup just short of 700°C and dropped slightly below 300°C while the engine was idling - not too bad. Yet it is to consider that it was pretty cold this day, about 0°C, so EGT during “standard conditions” will be somewhat higher.

A little later we started the engine for a fourth time (2.1MB) to let it run a little longer until the “engine ready” light came on. This time the camera was placed at the other side and it becomes clear how intense the jet efflux is - the metal sheet is wobbling and slowly creeps from the engine even though a stone is placed at the front to keep it from moving. Jürgen is going to construct an exhaust duct to do away with the problem of the exhaust stirring up the ground. It might be a good idea to point the duct slightly upward.

What’s quite interesting is that the engine consumed four litres (one gallon) of diesel fuel for just the four starts to be seen in the videos. Considering this and that almost no smoke is visible, the engine seems to be burning the fuel quite cleanly.

To view the videos, you will need the DIVX codec. You can get it here: http://www.divx.com/divx/

If the audio remains quiet (what’s a turbine video worth when the sound is missing??...), please search the web for “tsunami codec pack” and start the installation program. Select “Custom” installation and be sure to check only the “Windows Media Audio Decoder” box. Of course you may select other decoders as well in case you need them. After this procedure, the videos should play properly.

It’s really sad that the camcorder sound wouldn’t even get near the actual sensation of noise this engine produces... I guess if I hadn’t already had some experience with other gas turbines, I would have s**t my pants while starting this one ;-).