We have GE frame 7 gas turbines with Mark IV control systems. My question is pertaining to restarting the turbine after a stop signal has been given and the generator breaker has opened. The logic only allows a re-start if the speed is above 94%. If the speed drops below this value, the shutdown signal L94X is sealed in by speed signal L14HSX and you have to coast all the way down to firing speed before you can initiate a start again. I would think we should be able to enter the start sequence at any speed as long as there was still a flame signal present. Does anyone know why the logic is set up that way?

 

Once the speed falls much below 14HS drop out, the gas turbine is no longer self sustaining. That means in order to restart the starting means will be required. You don't say which version of the frame 7 you have, so I don't know whether you have an LCI or a separate electric starting motor, but both generally can't be energized at least until after flameout has occurred. You might have "rolling restart" capability which permits a restart prior to coasting all the way down to zero speed, but you still have to go through the purge, fire and accelerate sequences.

 

Thanks that is great information. We have a 4160v starting motor with a torque converter. After a shut down, we are able to turn the unit around and restart once speed falls below 9.5%. When starting, I believe the starting motor drops out at 60% speed so I would have thought the unit would be self sustainable above this value. I realize on shut down it is decelerating and there is a lot of fuel FSR reduction by then so I guess it would bog down. It seems the allowable restart would be something less than 94%. Do you know why the starting means cannot be re-energized below 60%? Will it damage the equipment? Thanks again for your post.

 

Bill,

I too am curious if someone like CSA or Otised could explain better the reason why most GE Frame gas turbines can't be restarted once they fall below operating speed.

We operate a Frame 7 similar to yours it sounds like. My understanding as yours was that the turbine was "self'sustaining" above approximately 60% speed. I can understand that there would be some complex logic needed to allow the unit to switch from a shutdown sequence back to a startup or accel sequence if the unit was above the 60% speed mark; or even if it was below 50% speed and would need help from the torque converter or applicable starting means. With the complexity of compressor protection etc. with variable IGV's, compressor bleed valves, inlet bleed heat and other extra's there is a lot to consider when trying to interrupt the turbine sequence.
But it seems like if a customer really wanted/needed to ability to restart on the fly it could be figured out. We recently upgraded the starting clutch system of a frame 5 unit to allow a restart of the unit above zero speed but below minimum speed. This was all done to decrease the "unavailable" time for the unit to roll down to zero speed which used to be close to 20 minutes. In this day and age time is money!

 

You may have to talk to GE to answer this. There is a lot of things that happen in a start sequence. It would not be easy to re-design the sequence to re-start from anywhere in the shutdown sequence. Once you lose flame, you are stuck with going back down to firing speed because you can't establish flame at higher speeds.

Another possible problem area is critical speeds - if you try to turn the unit around near a critical speed you might spend time at that speed rather than smoothly accelerating through it, which could damage the turbine.

I am sure there are other issues, but my memory fails me. (I used to design turbine controls for GE, but that was back in the Mark I and Mark II era, and it was on frame 3's and frame 5's, and mostly mechanical drives.)

 

otised,

you have touched on one of the larger issues, that of critical speeds. GE does not allow user control of turbine speed below 95% just for this very reason: to avoid prolonged operation at critical speeds (which they don't publish for their heavy duty gas turbines).

The originator has Mark IV Speedtronic turbine control systems, and those systems, and the Mark I and Mark II (and Fuel Regulator!) systems didn't use an acceleration reference for controlling turbine speed during starting and run-up to FSNL. They just dumped some fuel in, hoping the resulting acceleration rate wasn't excessive (neither too fast, nor too slow!). There were some vague guidelines that were published, but most of the start-up T.A.s didn't do much unless the rate was really too fast or too slow, and even that was subjective. It was very much a trial and error thing, and even ambient temperature would have an effect on the settings (until they started using temperature-corrected speed for calculating FSRs).

Beginning not with the first Mark Vs, but a year or two after they were being shipped, the start-up and acceleration scheme was changed to use an acceleration reference to control fuel to try to have a more controlled acceleration after firing and during run-up to FSNL. This greatly improved the start-ups (if it was tuned correctly) and reduced the thermal stresses on the machine.

Unfortunately, the acceleration rates that GE chose were usually too slow for most people, especially if they had peakers, so the rates were jacked up usually.

If you have a Frame 7E/EA with an electric starting motor, you can watch (some can even trend, if they have a starting motor current transducer and trending capability) the torque being provided by the starting motor up to the point the torque converter is de-pressurized (and current is directly proportional to torque). That's usually 150-170% of rated right up to the point 20TU-1 is de-energized. So, the starting means is providing a pretty good assist right up to the point that it's "de-coupled."

I would think it would be possible to write some application code or sequencing <b>for those units that use an acceleration rate reference during starting</b> to "catch" the unit during a fired shutdown. (But, once the flame goes out, the unit needs to drop below minimum firing speed and go through a purge cycle and fire at the proper firing speed.)

The problem comes if you catch it at a critical speed, it will likely hover there for a while whilst the Speedtronic figures out if the starting motor should be energized and how to start adding fuel to increase acceleration, etc.

And, the whole issue of undocumented critical speeds also comes into play.

Again, it's likely possible, but it would also likely take some creative and possibly complicated sequencing along with a good bit of trial-and-error, with the potential for some damage to the turbine either because of high vibrations and/or possibly dumping too much fuel in to try to increase the acceleration rate. There is a good bit of inertia, especially on units with air-cooled generators, and that has to be overcome before acceleration can begin again.

But, not on a Mark IV-equipped machine. The compute horsepower isn't there, and they don't use an acceleration rate reference during starting.

It's been said many times before on control.com: Once you own the machine, and it's not covered under any kind of warranty or long term service agreement, it's your machine. You just have to have the cojones to own the responsibility for the consequences of any changes you might make.

It's rumored that GE is being more receptive to the ideas proposed by Customers/owners and is considering more requests that would have been rejected out-of-hand not that long ago. So, write a proposal, submit it to your local sales rep, and see what happens. Remember, GE is going to charge for the "development" of something they may end up offering to others as an option someday. But, that's capitalism!

Turn the tables on them by throwing some numbers together to say that they can market this to others and make X dollars on your idea, and offer to let them work out the bugs on your machine. Capitalism shouldn't always be a one-way street, should it?

 

Yes, the way the power market is today every second counts. Most of our starts are "Fast Starts" because they are trying to get into the 15 minute load blocks. Sometimes they want us to turn the machine around right after a shutdown but they are not aware of the time it takes to roll down back to firing speed. I would think there would be a market for GE to look into this.
Thanks for your post.

 

Thank you gentlemen for the information. This is my first time on this forum and I wish I had used it for many questions and discussions I have had in the past. Thanks again for sharing your vast knowledge.

 

CSA,

You are mostly correct in your analysis. One slight error is that I know Mark I and Mark II did have an acceleration limiter during startup that tried to hold acceleration to 1% per second. There was an "accel" indicating light that would come on any time acceleration was limiting VCE (the analog version of FSR).

There hasn't been much call for this from the customer side on this issue that I know of. It would take a lot of work and simulation to design this feature. I don't know how often there is a need to abort a shutdown sequence or how much it is worth. Most of GE's development money goes to improving efficiency and output and reducing startup time since there is a clear competitive payoff there.

One way to get GE's attention on this would be to put in requirements in bid specifications for turning around a shutdown sequence. I don't know if Siemens, Alstom or MHI can do this - if not, then there may be little incentive to do anything other than take exception to the requirement.

Another way might be to request a quote from GE's Conversions, Modifications and Upgrading group.

 

A thank you to CSA and Otised for your insight. I definitely understand the limitations of the computing horsepower of the "legacy" speedtronic systems, although I missed being able to experience them based on my time in the industry. I appreciate your answers that I think confirm the ideas I had in my head anyway about sequencing. Again my experience is with what I work with which is MKVI, and did not think about the evolution that the controls "code" has gone through.
Like you said, if a customer or the industry wanted it, I am sure GE would provide, as long as the money was there!