Islanded Case Frame VI GT Mark V Ramp Rate


Thread Starter


There are two frame VI gas turbines installed with Mark V control system and had been running in 4 % droop mode in island condition with load of 21 & 22 MW on each machine.

There was a sudden load due to disturbance in connected load and frequency dropped to 48.75 Hz from 50 Hz within 5 Seconds. In the same 5 seconds, load increased on each GT's by 5 & 6 MW respectively. But till that time there was load throw off in the connected process load on under frequency setting of 48.75 Hz. After sudden load throw off, machine got over speed and tripped.

We are not able to find out if there is any defined load ramp rate in the Mark V, as why machine did not go to base load which if it could have, process plants would not have tripped.

Based on the information provided, there are no other generators connected in parallel with the two Frame <b>6</b> GE-design heavy duty gas turbines. And, if the two GTGs are being operated in Droop speed control while in island mode AND there is some load shedding ("throw off") scheme then it's presumed (though not included with the problem statement) there is some kind of power management system which sends signals to the GTG control systems to raise/lower load as necessary to maintain system frequency--and which also performs load shedding ("throw off") in the event of under frequency. But, the amount of load she'd (thrown off) during the under frequency condition was not stated. Further, we don't know how the power management system sends signals to the GTG control system to maintain frequency--by discrete RAISE- or LOWER speed reference signals, or by analog load setpoint signals.

Droop speed control responds immediately to the error between speed reference and actual speed. Per the problem statement, the connected load changed resulting in a frequency decrease--which would only occur if the load suddenly increased. This would, indeed, result in a load increase on the GTGs when the actual speed decreased due to the sudden increase in connected load. How does the connected load increase suddenly by 11 MW on an islanded system with two generators which continued to run for some time--until the power management system shed load (threw load off)?

So, since the connected load was changing so rapidly and mysteriously but the amount of load change did not result in the load on either GTG reaching or exceeding the Base Load value, it IS difficult to understand how the units could overspeed <I>if they were truly operating in Droop speed control without any kind of power management system--but there was load shedding (throw off) occurring by some means.

I don't have access to my workbook to calculate how much load should have increased due to a 1.25 Hz change in frequency when two machines operating in Droop speed control, but 11 MW change when the total load was approximately 43 MW doesn't seem too unlikely.

It's also very difficult to understand how two machines could have tripped on overspeed if the generator breakers were closed and there was still some load connected to the system. Are you absolutely certain the units tripped on actual overspeed--rather than a sudden excessive change in speed caused by the load shedding (throw off)? Most GE-design heavy duty gas turbines have overspeed setpoints of 110% speed--which would be 55 Hz on a 50 Hz system.

Too many things don't add up here, and there's too many unknowns. I t might be possible to see how an overspeed could occur if the generator breakers opens quite suddenly while load was high--but a properly configured Speedtronic turbine control system should be more than capable of responding to a generator breaker opening and keeping the speed from reaching anything near 110%. That is, unless Droop speed control was de-tuned to allow Droop speed control operate on an island system.

Again, there's too much we don't know and haven't been told and what we have been told doesn't add up properly--and there have been assumptions made.

However, there is typically NO load ramp rate when operating in pure Droop speed control--any change in actual speed should immediately result in a change in fuel flow-rate which should result in a load change, both nearly immediately.

Please write back with clarifications and more better information.
We had similar incident at one of our plant but with slight different configuration. We have One Distribution board catering load to down stream plants. This board is having 2 nos. Fram-VI GTs controlled by Mark-V (4 % Droop)and 1 No. double extraction type STG (30MW capacity). This board is connected to Plant grid (Not State Grid) through grid transformer. On the day of incident GTs were running at 21.4 and 21.7 MW, STG was running at 16.4 MW and Import to the board from Plant Grid was 14.8 MW. Both GTs running in Natural Gas. Suddenly this grid breaker got opened due to malfunction(Board was Islanded). Due to this breaker tripping, there was shortfall of 14.8 MW and both GTs load increased up to 26.4 and 26.9 MW and STG increased up to 18.5 MV within 4 to 5 seconds. But still there was deficit of approx. 3 MW and due to this board frequency came down to 48.75 Hz.We have external Frequency Load shedding system(Not connected to GT or STG control system), that sheds board downstream load in case of lower frequency. This system shaded 31.9 MW load. Both GT load came down to GT-12.8 MW and STG load came down to 11.2 MW immediately on load shedding. But after that Both GT load started increasing and Machine speed increased and eventually both GTs tripped on Over-speed after approximately 12 seconds of load shedding. Some of the data from Trip log is given in table below.

We are operating both GTs in Droop mode only. On that day too, both machines were in droop mode and preselct/Base load selection was not present. There is no external Load set point or Discrete Raise/Lower command configured in Mark-V.<pre>
09:41:22, 25.8, 19.9, 99.52, 102.46, 49.94,
09:41:32, 11.8, 8.7, 100.93, 102.46, 36.46,
09:41:38, 11.8, 9, 101.74, 102.46, 39.49,
09:41:39, 12.5, 9.1, 102.72, 102.46, 41.11,
09:41:40, 13.1, 9.1, 103.82, 102.46, 41.56,
09:41:41, 13.5, 9.3, 105.02, 102.46, 42.04,
09:41:42, 14.1, 9.4, 105.02, 102.46, 42.86,
09:41:43, 14.6, 9.6, 106.04, 102.46, 42.58,
09:41:44, 14.8, 9.8, 107.16, 102.46, 42.22,
09:41:45, 14.9, 9.9, 108.09, 102.46, 41.92,</pre>
We are not able to understand reason for machine going unto over speed after load shedding.

Note: In past we have faced generator breaker opening at higher load in same unit but machine has always sustained and kept running on FSNL.
<i><b>In general,</i></b> the steam turbine in a combined cycle power plant is operated in a condition that just opens the control valve(s) wide open and the load is a function of steam production. And steam production is a function of exhaust gas temperature and flow, and most GE-design heavy duty gas turbines in combined cycle power plants are operated with the IGVs in a mode that maximizes exhaust gas temperature at part load in order to maximize steam production at part load. And even when load is normally lowered (automatically or manually) it takes some time for steam production to drop--both because of the latent heat in the HRSG(s) and because as load is lowered the IGVs are maximizing exhaust temperature.

So, when running isolated from a grid the steam turbine is NOT going to reduce power, but rather is going to keep producing as much power as it can because it's not really responding to frequency deviations like it should if it were in Droop speed control mode. It will contribute for some time to an overfrequency condition--made worse by the latent heat in the HRSG(s) and the IGVs maximizing exhaust temperature at part load. There's only so much the three units can do if there's not enough load and there's too much steam.

Now, this doesn't explain why the GT speed reference remained so high--that's likely because the load shedding scheme didn't reduce the speed reference at the same time the load was being shed. And, I have seen the steam turbine cause the GTs to trip on reverse power because it was producing too much power for the load and the GT control systems kept reducing load due to the high frequency relative to the speed reference when a combined cycle power plant was suddenly isolated with little or no load, and the plant ultimately tripped on overfrequency because the steam turbine just didn't reduce power fast enough because it was being operated in power mode not Droop mode.

But that doesn't seem to be the case here, based on the information provided. There's just not enough information to be able to say what happened, and a lot of times without good data there are too many perceived things that either did or didn't happen which didn't or did happen that caused problems.

I still maintain that two GTs synchronized together with both operating in Droop speed control mode independent of a grid MUST either have: (1) an external system adjusting the speed references to maintain frequency; or (2) an Isochronous load sharing scheme implemented in the two control systems with some discrete and analog signals connected between them; or (3) a seriously de-tuned Droop speed control mode in one or both control systems. It could be that the nature of the load is such that manual speed reference changes are not really necessary--but that's not very likely and that would mean the human operators were very, very diligent and knew about and could respond to most unexpected load changes (which would mean they're almost superhuman and had omniscient powers beyond mere mortal operators)--which is even more unlikely.

No; there's something which hasn't been told or which is being missed or misperceived. There's just not enough information and data to be able to say any more.
Dear CSA,

Our STG is normally running with exaction pressure control having priority and not in load or frequency control. Further, we have steam header pressure controller with Additional auxiliary boilers and HRSGs having supplementary firing facility to take care of Steam pressure.

As I mentioned earlier, Load shedding scheme is independent of GT control system and there is no connections between load shedding system and Mark-V that is why TNR is constant. But if you see from given triplog data when TNH crossed TNR and it was rising, still FSR has not reduced. Can you please explain, in which scenario, FSR will not reduce though TNH is higher than TNR?

Thank you in advance.
>Any suggestions or recommendations on subject matter?

No; there's something which hasn't been told or which is being missed or mis-perceived. There's just not enough information and data to be able to say any more.