We have three Gas Turbine Generators (GTGs) of the GE (LM2500+G4) model in parallel, each with a site rating of approximately 27.5MW.
Inquiring about the maximum acceptable step load for this type of GTG, I reached out to GE for guidance and yet to get a response. However, I'm seeking additional insights here as well.
a) Typically during testing, we conduct load step increments using the load bank controller at 10-minute intervals, gradually increasing the load to achieve a rated Generator loading of 10%→25%→50%→75%→100%.
b) Nevertheless, the operational test procedure, particularly for load acceptance transient response tests and 4-hour load tests, dictates applying load increments from 0% to 50% and then 50% to 100%. It's important to note that speed transients should not exceed 10% and should recover to within 1% within 5 seconds, while voltage transients should remain within -15% to +20% and recover to within +/- 3% within 1.5 seconds.

Therefore, based on approach (a), the maximum step load would be 25%, whereas following approach (b) would entail a maximum step load of 50%.

Currently in PMS, the motor start inhibit function is based on spinning reserve, so when one GTG is running with a load of say 4MW, based on spinning reserve it can even start a 15MW motor. Can GTG be operable within the limits as mentioned in b) above.

 

The question I have is: When you impose a "step" load change on the engine(s) how is it done? If the turbine control system is a GE Mark* it likely has a feature called Pre-Selected Load Control, and while the Pre-Selected Load Control reference can be changed at a very fast rate, the loading rate of the Mark* does not usually the actual load on the machine so quickly. Rather, the fuel flow reference is ramped up at some rate. And, the fastest rate I've ever seen for an LM engine (and I've only worked on a few LM engines) was zero-to-full-load ("Base Load") in 30 seconds, so in the case you've outlined that would 0-27.5 MW in 30 seconds, or 0.917 MW/second, or 55 MW/minute. Which is fast, but the engine is an LM and it's capable of some pretty fast load rates--probably even faster than 30 seconds from no load to full load.

So, regardless of the turbine control system, at what rate does the actual load of the machine change during this testing? I doubt it's instantaneous, or as fast as 30 seconds from no load to full load (or in the case of a) the maximum "step change" would result in a change of (27.5MW/4) 6.875 MW in 7.5 seconds, and twice as fast in the case of b). That's just my SWAG (Scientific Wild-Arsed Guess), but you can tell us what the actual loading rate is when the load reference experiences a "step change."

Next, lets discuss starting a 15 MW motor (which is 15/27.5*100)--almost 55% of the LM's rating. Does this motor use a direct, "across the line" starter? If it's an induction motor that would be one hell of an inrush current draw--well more than 15 MW, and it would almost certainly some upset.

You didn't say if these four machines operate independent of a larger grid, such as supplying a large refinery or LNG compressor facility that is not grid-connected. If it was one single machine, while it has the ability to power a 15 MW motor if it was only operating at 4 MW if started "across the line" the inrush drawn on start-up would be above the full load rating of the machine and would almost certainly result in a decrease in LP turbine speed and generator frequency--perhaps even enough to cause the machine to be tripped due to underfrequency. AND, if the 15 MW motor is driving any kind of load at the time of starting the current drawn by that load plus the inrush current would be very, very high.

You mentioned a PMS (Power Management System) and if that controls the LM machine loading and frequency (when operating independently of a grid) then that would have to be detailed as well.

There's a lot of hypotheticals here, and not a lot of data. The packager of the engine and generator would be the first place to ask what the capabilities of the machine(s) are; followed by GE Aircraft Engine group, who can say with a very high degree of certainty what the engine can safely handle when the fuel flow-rate is increased very fast (you seem to be saying almost instantaneously). Aeroderivative engines are capable of some very fast starting and loading rates (certainly faster than many heavy duty gas turbines can be started and loaded), but I'm sure there are limits to the number of times a machine can be loaded "instantaneously" without causing serious damage. The capabilities of aeroderivative engines suit them to a variety of applications that heavy duty gas turbines just can't achieve.

There is a lot of GE-design heavy duty gas turbine controls experience here (more than 20 years' worth), but not a lot of GE LM experience. And, programmable control systems can do just about anything--it's the equipment they control as well as protect that is the limiting factor in most every case. And, while we do dabble in mechanical aspects of heavy duty gas turbines and controls--mostly as they relate to the control system--we don't usually go too much into the turbine capabilities, except as seen in control and protection schemes.

I believe GE Aircraft Engine group publishes what used to be called a IDM--Installation & Design Manual. It covers EVERY aspect of LM operation and protection and control. It is primarily intended for the packagers of LM engines to use to configure and program the control systems to operate the engines safely and to protect them. They might, for example, say the temperature in the turbine enclosure must not exceed nnn deg C--but they won't expressly say how that must be done; things like that. I'm certain maximum loading rates would be included in a document like that, but they rarely make it into the field.

If there was an architect/engineer responsible for power plant design, they should be able to answer your questions--and they might even be answered already in a Plant Operations Manual they might (should) have provided.

 

Thank you for your reply,
Some quick answers to your questions,
1) There is no time interval mentioned in the operational test procedure. I believe it is also 10 minutes
2) Actually it is a 12MW motor started across DOL
3) The facility receives no power from grid. (only GTG's generated power)

Based on functional description from PMS manual,

 

Received the reply from GE,
"Higher the step load, higher will be the turbine speed oscillation (and consequently generator frequency) and longer will be the recovery time.
50% step loads will exceed +-10% frequency variation and the recovery time to reach +-2.5% frequency will exceed 5 seconds time.
Higher the MW of the turbine at the beginning of the step load, better will be the reaction (less oscillation and quicker recovery time), this means that if generator is at 0MW with breaker closed and you give 8MW step load, the frequency variation will be higher than if you take the same step load when generator is already at 15MW.
If customer require those step loads, it’s possible that 81U and 81O will need to be adjusted, otherwise I’m not confident that the GTGs will be able to do procedure b)."

 

I agree that doing an 8MW load accept beginning at 15MW will result in less speed variation than doing an 8MW load accept starting from 0MW. However, the top end can also be a problem if not connected to a grid and running in isoch mode, where100% power is defined as hitting a limiting regulator. You typically cannot do a large load accept to 100% power where 100% power is defined as running on a limiting regulator. In fact, while running in isoch mode you need to avoid running on a limiting regulator even in normal operation. Otherwise, the power turbine will roll off in speed. The power that can be achieved before hitting a limiting regulator will vary depending on among other things ambient temperature and pressure. However, you mention a site rating of 27.5 MW. Is 27.5MW being defined as 100% power regardless of ambient temperature? If so, the 50% load accept would be going from 13.75 MW to 27.5MW. That certainly will be a challenge for a +G4. (An LM6000 would be capable of that scenario). The max power for a +G4 on a cold day is in the neighborhood of 35 MW so the chances of performing that load accept is greater on a cold day than on a hot day. On a very hot day max power will be a lot lower than the peak of the tent curve and the chances of performing a 13.75MW to 27.5MW load accept would be extremely difficult if not impossible.

Also, the IDM does provide guidance on load accept capabilities but typically the IDM is not given to customers. However, given load accept scenarios and details of the installation including generator inertia, GE should be able to provide estimates of the response of the engine for those scenarios.

 

I don't understand the situation here--I understand the question, but I don't understand why there is a question in the first place. One would think this was all discussed and agreed to by all parties during the design phase of the power plant.

Also, the @Selk says there is a PMS and we don't know exactly how the PMS is sending signals to the three LM units--except there is some mention of 4-20 mA signals. We don't know what governor mode the LM units are running, if one is running in Isoch mode and the other(s) in Droop mode, or some kind of Isoch Load Sharing Mode or ???

Lastly, is this some kind of test that is being done for the first time?

Are these units being commissioned at this time--meaning this is some kind of acceptance test?

These are two-shaft machines and I would think they were chosen for this application because of their fast start and load abilities--and that all of this has been discussed and agreed to long ago. If so, push the button and go--and see what happens.

Originally I thought the original poster was talking about instantaneous load acceptance (also called "load throw-on" and several other local colloquialisms depending on where in the world the power plant might be located), and that's kind of difficult to do for most locations because to instantly "dump" a large amount of load on a turbine-generator requires some creative high-voltage switching, and most likely tripping another machine off the grid.

Again, if this was a power plant design criteria that the A/E (Architect/Engineer) agreed to provide, and wrote a manual to describe the power plant's capabilities, it's pretty unlikely the supplier of the engines would have agreed to the conditions and requirements if the equipment wasn't capable of meeting the requirements. I've been in similar situations, where a test I was asked to perform was highly likely to fail and didn't want to execute the test. Two of the tests (at different sites) failed--miserably. But, the supplier and designers and plant owners got together and worked out a solution that was acceptable to everyone (meaning acceptable to the bankers who had lent the money to build the plant) and the tests were ultimately successful. One site, well, ..., that one was an unmitigated disaster and led to death threats and caused a six-month delay in hand-over of the equipment to the utility. No further tests were ever performed at that site. And, there was a shake-up of plant management. A couple of other sites actually passed the tests on the first attempt--a surprise to me (and others!).

There's too many things we don't know about this situation and requirements.