Flow balancing of fuel system, gaseous fuel.

Mk-1533 through 1535 discussions.

Flow balancing of fuel system, gaseous fuel.

Postby j79guy on Sun Jan 11, 2009 8:07 pm

The Industrial Avons are vulnerable to any flow imbalance in the fuel delivery system, on natural gas, or any gaseous fuels. Flow imbalances cause high EGT thermocouple spread, and may excite a third order flexural frequency of the HP. Turbine disk. When bad enough, long enough, the HP disk may fail, and even with low thermocouple spread, the uneven flame tube loading causes accelerated turbine blade shroud tip abutment face wear. We here flow balance the fuel delivery system as a whole, including the manifold, pig tails and injection nozzles, to get the flows within 1% of each other. However of late, I've been thinking of incorporating a "pinch valve" on each pig tail, to allow tweaking of each fuel nozzle while on the fly, to reduce the EGT thermocouple spread to a minimum.
Any thoughts on this?

Robin
j79guy
Site Admin
 
Posts: 256
Joined: Sat Jan 10, 2009 5:06 pm

Re: Flow balancing of fuel system, gaseous fuel.

Postby AgentJayZ on Mon Jan 26, 2009 9:17 pm

A good idea, it seems. But to see if it is really necessary, a test is needed.
A pressure sensor is needed at the inlet of the fuel manifold, and another is needed at each end.

If the fuel pressure does have a significant drop from inlet to end (and that's the assumption we have) then your idea is a good one. If it does not, and we can only be sure by testing... then pressure drop with distance from the inlet is not the cause of you flame tube temperature spread.
AgentJayZ
 
Posts: 9
Joined: Sun Jan 11, 2009 8:37 pm

Re: Flow balancing of fuel system, gaseous fuel.

Postby j79guy on Mon Jan 26, 2009 9:38 pm

I agree. However, I recall the situation of the Mk-1900 Industrial Spey, we had high thermocouple spread, and the fuel nozzles flow tested ok, within 1% of each other. The spread was still there. So investigating furthur, we flow tested through the whole nozzle, including the manifold and flex line. Bingo, we found a huge variation in flows from one whole "nozzle" to the next. Turns out the 90 degree bends in the fuel flex lines were the single biggest restriction in the system, especially those with a bit of gobby weld inside these sharp bends. By simply grinding the welds smooth inside, a significant reduction in T/C spread resulted.
My idea of the pinch valves on the Avon were to adjust for individual nuances in each combustor "feed system", for lack of a better term. While on the fly, the pinch valves could be tweaked to give the lowest T/C spread, regardless of absolute fuel flows. A spectral analysis of the disk flexural frequencies would be in order, to verify we weren't hurting the engine, but I suspect the system would work great.

Robin.
j79guy
Site Admin
 
Posts: 256
Joined: Sat Jan 10, 2009 5:06 pm

Re: Flow balancing of fuel system, gaseous fuel.

Postby TillamookTurbine on Mon Jan 26, 2009 10:39 pm

Since this is gaseous fuels and it sounds like you don't have to comply with a type design, like in aircraft, can you modify the fuel system to incorporate a sonic choke point in each of the feed legs? Then the mass flow is set by the upstream pressure to the choke and not by any restrictions downstream of the choke. We use these all the time in rocket testing to decouple the propellant feed system from the combustion chamber. Otherwise combustion perturbations will get amplified by the feed system and drive the combustion in to instability. Saturn V 1st stage engines had to incorporate attenuators in the propellant lines that look like hydraulic gas over oil accumulators to dampen the "pogo" instability.

To get the choke two work, though, you need a pressure ratio of about 2:1 across the venturi to sonically choke it (at least for air) I'm not sure for methane; depends on the speed of sound.

Terry Spath
TillamookTurbine
 
Posts: 63
Joined: Mon Jan 12, 2009 5:56 pm

Re: Flow balancing of fuel system, gaseous fuel.

Postby j79guy on Tue Jan 27, 2009 8:40 pm

Terry,

With industrial turbines, especially lease engines, that you own, you can do any darn thing you want. That's why I love working on aeroderivitive industrial turbines, 'cause they're all ex-fliers, and now we can modify the heck out of them. The aircraft guys can't believe the stuff we do, the liberties we take and yes, I've wrecked a few engines. But, overall, we've made a pile of great mods and upgrades, and in some cases have tripled the overhaul life of some models, without effecting the output negatively. Conversely, we have incorporated upgrade mods, that result in up to a 23% increase in output, without a reduction in overhaul life. I enjoy flying in the face of OEM engineering and modifying where we see an obvious deficiency. And yes, we see a consideral amount of design deficiency in all makes and models, none are exempt. As our reputation grows with a particular customer, they slowly come around to the notion that hey, maybe these guys really do know what they're doing, and these modifications they're suggesting might actually work! It's tough, as they have been brow-beat for years into thinking that the OEM knows best, what God and (Insert manufacturer name here.) hath wrought, let no man put asunder. (Ok, rant over. Whew! you must have touched a nerve!)

Back to the OP; the fuel flow speeds within the distribution manifold and flex lines of natural gas fueled industrial turbine units is quite low. For example, the engine of the OP, a Rolls Royce Mk-1533 Avon, consumes approximately 130,000,000 Btu/Hr. at full power. There are eight fuel injection nozzle feed flex lines, so each is flowing 16,250,000 Btu/Hr or 270,833 Btu/Min. If natural gas has 1000 Btu/Cu-Ft, then 270.8 Cu-Ft per minute is moving through each flex line. Each line is 3/4" on the ID, so you can see the velocity is quite low. Any turbine engine with shrouded turbine blade tips, is susceptible to uneven combustor loading, and the fewer combustors, the more critical this becomes. If combustor loading is not even, the turbine blade tips move relative to each other, and eventually may interlock, via a sawtooth wear pattern at the shroud tip abutment faces. When the blades interlock, their collective flexural frequency drops down into the operating range of the engine, and within short order, will tear out of the disk. I've seen it, and it's not pretty. In extreme cases, the Hp Turbine disk may excite on a 3rd order frequency and fail as well.
The only indication we have of combustor loading is the downstream exhaust gas thermocouples. A high temperature spread between probes is an indication that the combustors may not be loaded evenly. To date, users monitor this thermocouple spread, and when it gets too high, will pull the fuel nozzles, clean them, and cross their fingers. If fibreoptic borescope inspection reveals nothin out of the ordinary, and the observed vibration levels check out, I am keen to start adjusting the individual flows in each fuel delivery flex line, to "power balance" the combustors. The fly in the ointment is the thermocouple placement. If all eight thermocouples were placed in exactly the same orientation to the upstream combustors, in terms of insertion depth and angle, all would be well. But they are not. A couple probes are at an unusual angle, and thus indicate a different temperature profile than their neighbouring probes. Therefore it will prove to be impossible to completely eliminate thermocouple temperature spread, but we should be able to minimise it.
One of my customers is keen on the idea, so I wish to try it out on their next engine that comes into the shop for overhaul. Should be interesting.

Robin
j79guy
Site Admin
 
Posts: 256
Joined: Sat Jan 10, 2009 5:06 pm

Re: Flow balancing of fuel system, gaseous fuel.

Postby TillamookTurbine on Thu Jan 29, 2009 5:29 pm

I'm very familiar with the "OEM knows all" syndrome. Take the engine that we experimented with, P&W F100-200, and you can really debunk that myth. Number one, the talented guys that designed it in 1970 are retired, dead, or have forgotten what they knew. Number two, the OEM had very few engineers that really knew the "big picture" and it's certainly no better 40 years hence. For example, one of our consultants on the RASCAL propulsion project was a retired
compressor expert from Pratt. He was very helpful on compressor questions but honestly had no idea how the CEN (variable area nozzle) worked.

We found the real hotbed of knowledge on this engine was at the USAF overhaul facilities. We had the pleasure of working with Edwards AF base engine shop. It's mostly technicians and mechanics with few if any professional engineers. However, many of the technicians have 20 years experience with that engine. Many of them have taken it upon themselves to really get educated to the point that they have a very thorough engineering understanding. It's my experience that they know the engine far better than Pratt does.

Terry
TillamookTurbine
 
Posts: 63
Joined: Mon Jan 12, 2009 5:56 pm

Re: Flow balancing of fuel system, gaseous fuel.

Postby j79guy on Fri Jan 30, 2009 6:21 pm

Great, get 'em on this board!

Robin.
j79guy
Site Admin
 
Posts: 256
Joined: Sat Jan 10, 2009 5:06 pm

Re: Flow balancing of fuel system, gaseous fuel.

Postby masih on Mon Jun 01, 2009 11:25 pm

Is it possible that it causes this problem for shape of fuel manifold in natural gas fuel engine?
As you know the fuel manifold is dead end .
masih
 
Posts: 2
Joined: Wed Jan 14, 2009 11:13 pm

Re: Flow balancing of fuel system, gaseous fuel.

Postby j79guy on Thu Jun 04, 2009 9:10 pm

Masih,

Possibly, on certain engine models. The Avon has a fairly large manifold, with the main tube inside diameter quite large. The Spey does not, and has a too small main tube, causing the end fuel burners to be "starved" for fuel. The best solution, is to build the main tube like the LM1500, which is large in diameter, a true circle with two joins and a central fuel inlet flange.
The GE LM2500 Natural gas manifold has a generous main feed tube diameter, but is an "end wall" type arrangement like the Avon.
You really are only limited by common sense. Some time ago when we converted two Avon units from Liquid to Natural Gas fuel, we scratch built the fuel delivery system, and made the main feed tubes quite large in diameter. We also used flex feed lines to the injection nozzles instead of the hard "pig tail" lines, which are engineering overkill.

Robin.
j79guy
Site Admin
 
Posts: 256
Joined: Sat Jan 10, 2009 5:06 pm


Return to Avon

Who is online

Users browsing this forum: No registered users and 1 guest

cron