I'm having a problem with a G.E. Frame 5 not accelerating during start up.

We're reaching 19/20% and firing (on liquid), and it teeters between firing/accelerating and cranking. Only achieving flame on 2 from 4 flame scanners. It slowly climbs to 25%, and no more.

I've checked the torque converter and it's getting quite hot. The manual throttle/choke valve seems not to do anything to the speed, the pressure gauge on the line between the torque converter charge pump is only reading 1 bar, regardless of throttle/choke valve position.

I'm thinking that either the charge pump is at fault, or the torque converter is dumping oil, straight through, with possibly 20TU at fault? Although with me fully closing the manual throttle/choke valve on the drain line from the torque converter, I would have expected to see the machine accelerate?

My back to back tried to adjust the manual throttle/choke valve a few weeks ago to overcome a similar problem, so I'm not quite sure what position the valve was in beforehand, as the locking device was not re-attached.

Any ideas people?

 

what is the hydrulic pressure?
what fuel you are using? how is the firing system?
How is VCE? how much?

 

Hey,

When did this problem start? Just recently? After a maintenance outage? After a trip from load? The purpose of this question is to understand what, if anything, might have changed recently (like new fuel nozzles installed during a recent maintenance outage).

What kind of duty does this turbine see? How often is it started/stopped?

Do you have any data from successful starts to compare to these unsuccessful starts? If so, does anything jump out at you?

What speed did the unit establish flame at before when it wasn't experiencing this problem? Was it higher, or much higher, or even lower?

What kind of turbine control system is being used? Do you have any trending capability?

What Process- and Diagnostic Alarms are being annunciated prior to, during and after this event?

Does the unit uses an electric motor for starting? If so, a check of the current being drawn by the motor will tell you pretty quickly if it's putting out rated torque, or more, or less. (Amperage is directly proportional to load; minus some windage losses which are constant at motor rated speed.)

If the unit uses a diesel engine, then a check of the fuel rack should tell you if it's trying to put out a lot of torque, or just enough to maintain speed (which shouldn't be too much if the torque converter isn't transmitting any torque).

How long since the torque converter was last refurbished or replaced? They do go bad.

Do you have a spare 20TU to try?

I'm not personally familiar with this manual valve you are describing; I'd have to see a copy of the Starting Means P&ID to understand how it might work.

If the torque converter is getting really hot, I would suspect the charge pump because oil flowing through the torque converter should help to cool it. (Of course, when torque is being transmitted the oil will pick up some heat, but, again, if the oil coming from the L.O. Tank is relatively cool, then it shouldn't increase in temperature too much and it should actually be picking up some heat as it's cooling the torque converter.)

Trying to determine if the starting means is producing torque or not will tell you if the torque converter is transferring torque. The presumption is that the starting means is working properly and doesn't have any issues of its own. If it's a diesel engine, that would include making sure the air filters are clean, the fuel filter is clean, and that there's no white smoke coming from the diesel engine exhaust.

Fuel should NOT be used to accelerate the turbine--it should be torque coming from the torque converter during initial firing and warm-up. If you're not achieving flame in all the combustors like you have been in the past, then there might be some problem with the liquid fuel system (low forwarding pump pressure; dirty fuel filters; a possible issue with the high-pressure liquid fuel pump; even (unlikely) a problem with the servo-valve if the L.O. isn't maintained properly).

Have you used the manual selector valve at the liquid fuel flow divider to monitor the pressures during firing? They should all be relatively equal, and they have to be above the liquid fuel check valve cracking pressure (usually approx. 100 psig).

What kind of smoke is coming from the GT exhaust when this is occurring? Does it start having a lot of white smoke (possibly billowing white smoke) a few seconds after some flame is established? If so, then it's likely that for some reason the flame is not propagating through the cross-fire tubes and igniting all the combustion cans and hot gases are mixing with unburnt fuel in the turbine section and exhaust and causing the heavy white smoke in the GT exhaust because of incomplete combustion (just like when a lorry/truck has a lot of heavy white smoke coming from its exhaust, one or more of its cylinders isn't firing and the unburnt diesel fuel is partially combusting in the exhaust manifold to cause the white smoke).

Have you checked the ignitors (spark plugs) to see they are working correctly? Only one has to work, but if it's weak or wet then it's not going to work very well.

My money's on the torque converter/charge pump/20TU. But there's still some things you need to check before pointing the finger squarely at the torque converter/charge pump. Like changing 20TU, and checking current draw/fuel rack position (even diesel RPM, if it's a diesel engine starting means). Liquid fuel supply pressure needs to be correct and stable, and liquid fuel filters need to relatively clean, and the liquid fuel check valves need to be working correctly, as does the high-pressure liquid fuel pump and even the ignitors (spark plugs).

Please write back to let us know what you find. And, even if it seems irrelevant, please provide all of the requested information if you need more assistance.

 

> what is the hydrulic pressure?
> what fuel you are using? how is the firing system?
> How is VCE? how much?

Hydraulic pressure is 89-90 Bar.
Liquid fuel/diesel.
Firing system is good, as far as I know. We started last week absolutely fine on gas, and changed over to liquid/diesel. We haven't started this machine on liquid recently.
VCE, unsure?

When we see the 2 flames (A + B), it's only for a few seconds. Would I be right in saying that the MKIV has to see the flames for 10 seconds in order to move further through the start up sequence?

 

The problem has started recently. We have changed fuel nozzles, dual fuel, mechanically atomizing. New liquid fuel and purge check valves.

The machine normally runs at between 4 and 8 Megawatts. Usually started and ran for a couple of months at a time.

No data jumps out at me from previous starts.

I'm not too sure what speed the unit achieved flame at during previous successful attempts.

G.E. MKIV Control system, no trending unfortunately.

There are no alarms showing that would be in line with what I am seeing. Low lube oil, and low lube oil temp, but these are known faults with the switches.

We use a Detroit diesel engine for starting. The engine seems to be functioning correctly, and does not seem to be labouring.

To the best of my knowledge, the torque converter has never been changed, unit build date, 1991.

We don't have a spare 20TU unfortunately. On the same drain line as 20TU, just after (downstream) there is a manual valve. On the starting P&ID it states this valve should be set/closed or opened to achieve 800RPM, then locked in that position. Even playing with this valve and closing it fully, the turbine will not speed up.

Previously there has been a pressure gauge fitted on the supply line to the torque converter from the charge pump. This gauge is reading around 10PSI on ratchet, which I assume is the oil making it's way through the system from the A.C. lube oil pump. When we try to start or crank the machine, this gauge doesn't rise. I wasn't sure if this was due to the charge pump being inoperative, or if the torque converter was just dumping oil straight through to drain. Either of these problems would explain the heat, but it would be too coincidental for 20TU and the manual valve to be at fault, surely!? Leading me back to the charge pump?

All pressures at the selector block on the flow divider are equal during the start sequence, and expected pressures.

Unfortunately we do not have fuel gas available right now to try a start on that.

 

Okay, Jimbo76,

Mark IV. Dual fuel. Runs mostly on gas fuel. No problems starting on gas fuel, except for a similar problem a few weeks back.

Dual fuel machines that don't run very frequently on liquid fuel are prone to lots of check valve problems--liquid fuel check valves and purge air check valves. Consult the Liquid Fuel and/or Fuel Purge Piping Schematic drawings (P&IDs) to find the Tell-tale Leak-off on the drawing. It's almost always physically located on the right side of the Turbine Compartment, under the walkway. It should be draining into the gas turbine Drains Tank, usually through a "scupper" or "cup" so that flow out of the pipe into the scupper can be seen--when someone is watching, that is.

When you try to start the unit on liquid fuel, you need three people--two at the manual selector valve of the liquid fuel flow divider (one with a paper and pencil), and one near the Tell-Tale Leak-off. When the unit is firing and with the manual selector valve at position 1 (combustion can #1) when the fuel pressure builds up to something more than 100 psig and stabilizes (hopefully it will build up to something more than 100 psig!) have the person without the pencil and paper call out the pressure to the person with paper and pencil to record the value, then the person without the paper and pencil will shift the valve to position 2 and call out the value to the person with the paper and pencil to record the value, and so on, through the first ten positions. (I think the 11th position is the liquid fuel pump discharge pressure, which should be just slightly higher than the individual fuel nozzle pressures, and the 12th position is the liquid fuel forwarding pump pressure (the pressure upstream of the high-pressure liquid fuel pump); nice to know these pressures, especially the liquid fuel forwarding pump pressure ahead of the high-pressure liquid fuel pump.)

The person at the Tell-tale Leak-off could be using a smartphone with video recording capability to record if any liquid fuel comes out of the Tell-tale Leak-off. If there is any flow, that indicates that at least one or more of the liquid fuel purge check valves is leaking (one of the most common causes of liquid start-up and operational problems--<b>especially</b> for dual fuel machines that don't run very often on liquid fuel. The leaking check valve(s) will be those fuel pressures which are significantly lower than the others--when there is flow coming out of the Tell-tale Leakoff during firing.

If there isn't any flow coming out of the Tell-tale Leakoff during firing, then if VA-19 is working correctly the liquid fuel purge check valves are working correctly. And, if there are imbalances between the 10 fuel pressures recorded during firing then the liquid fuel check valves of those cans with higher or lower pressures than the others are suspect.

Atomizing Air pressure is also very important, and if the unit has a Booster AA Compressor you should examine all of the check valves in the piping and the solenoid-operated valves to make sure they are all working correctly. (This will likely require a few tubing fittings, a source of dry, compressed instrument air (or a nitrogen bottle with a proper regulator) to "stroke" the solenoid-operated valves.) Have a look at the Atomizing Air Piping Schematic drawing for details.

Lastly, if fuel nozzles don't get properly purged after operation on liquid fuel, the distillate will carbonize in the nozzles and cause poor atomization and flow restrictions which will also affect the ability to establish and maintain flame.

The logic in most digital GE turbine control systems says that flame must be present in one or more combustors with flame detectors during firing for approximately two (2) seconds before the fuel is reduced to the warm-up value. (Usually, a little extra fuel is put into the machine during firing to help establish flame, and then a couple seconds after flame is detected (to help get flame ignited all around the machine through the cross-fire tubes) the fuel then gets cut back to warm-up to try to prevent thermal stress on the hot gas path parts. Sometimes, when the fuel is cut back to warm-up, flame is lost in one or more combustors--and while this happens frequently, it's not normal, but it does happen.

The machine then goes through a 60 second warm-up period at the reduced fuel flow-rate then starts increasing the fuel flow to help accelerate the unit.

My guess is that the check valves and/or the atomizing air flow-rate is insufficient at firing to establish flame and to maintain flame during warm-up and acceleration. There is insufficient energy from the hot gases at low speeds to accelerate the turbine <b>without</b> an assist from the starting means through the torque converter. You may also be experiencing a problem during starting on gas fuel that is being overcome be excessive gas fuel (which increases the internal temperatures and therefore the stresses on the hot gas path parts).

On these older dual fuel machines there is usually a lack of maintenance that contributes to these kinds of problems. Machines which don't run often on liquid fuel also have problems with bacteria growing in fuel lines and filter canisters and in the suction strainers of the forwarding pumps. Sometimes flow through the system after periods of inactivity can cause the growths to move through the system, and I've seen it cause the liquid fuel check valves to stick open causing low pressure at the liquid fuel flow divider and poor atomization (pressure atomization--not air atomization) at the fuel nozzle.

One more thing, most of the air-operated valves on the machines use compressor discharge air to actuate the valves. To do this, the compressor discharge air is passed through pressure regulators upstream of the solenoid valves. The air is usually moist and the regulators can easily fail--and quite often do--causing the actuators to be damaged or little or no air to get to the actuators through the solenoid valves. These air pressure regulators rarely, if ever, get any attention during maintenance activities. (If the plant has a reliable source of dry instrument air it's always a good idea to use that instead of compressor discharge air for these air-operated valves--including the compressor bleed valves.)

Finally, if the unit isn't experiencing problems starting on gas fuel but is when starting on liquid fuel--again, it could be because over the years someone has been mucking with the gas fuel start-up constants because the torque converter/starting means has been slowly failing over time. So, while it may seem it's working correctly on gas fuel, it might not actually be. It may even be that the gas fuel you're using now might have a higher heating value than the fuel which was originally being used and is also contributing to the seeming lack of problem when starting on gas fuel. Many times these problems are rarely the result of one thing--especially on older machines, and even more especially on older dual fuel machines which don't run very often on liquid fuel, or which start even more rarely on liquid fuel. A lot of minor problems can be masked if the machine is started on gas fuel and then transferred over to liquid fuel, then back to gas fuel. In the case of failing liquid fuel check valves, that would result in high exhaust temperature spreads which might not be too excessive--but if no one saw any fuel coming out of the Tell-tale Leakoff no one would be the wiser. Also, failed or failing liquid fuel check valves can usually appear to be working okay when switching to liquid fuel during full speed operation; again, it would likely appear as an elevated exhaust temperature spread when running on liquid fuel but it might not excessive. Seemingly minor issues can be huge problems when trying to start on liquid fuel.

Please write back to let us know what you find. Also, this topic has been discussed MANY times before on control.com (that's like how your preferred World Wide Search engine found this site), and there's a 'Search' feature that will return lots of relevant results (though it's recommended you use the Search 'Help' feature before searching; it's not intuitive).

VCE is "old" Speedtronic lingo for FSR.

 

There have been a few problems with the foot valve for the charge pump, unfortunately that would mean a tank drain.