Dear All..
please provide your views...

G.E frame 6 gas turbine mark 5(30MW) was on full load. Load reduced to 24 MW to change over form fuel distillate to fuel gas. Gas valves were checked opened and pressure was all ok. Power factor at this time was .96 and voltage 11.14, vars 10. we have problem with one flame detector not sensing flame during normal operations.

now when the command was given to change over after 20 sec the load started reducing suddenly to 17mw then 12mw then 7mw then 4 mw then 2mw. after this point it jumped suddenly back to 17mw then 24mw.

During this time the power factor reached .76 and unity. Alarm came combustion trouble followed by loss of excitation and generator breaker tripped...

Also would like to mention we have avr problem so we regulate manually...and during the change over the voltage was being adjusted manually which was perfect......

WHAT CAN BE THE REASON OF TRIP?
COMBUSTION TROUBLE?
AND THE REASON FOR SUCH SUDDEN DROP AND INCREASE IN LOAD DURING CHANGE OVER?


thank you
detail from csa requested

 

A bit complicated, but let's take it one at a time. It very much looks like you started a fuel changeover and somehow initially you weren't getting gas flow, hence the drop in load. Then the gas came on and the load came back. The loss of excitation and breaker trip, you say the load dropped to 2mW, I think you went a bit lower and tripped the breaker on Reverse Power. The alarm "combustion trouble" for >9 sec (usually) will trip you. If you had all of those fuel problems happening, I'm not surprised that you had combustion trouble. Try a recal on your SRV & GCV to at least prove functionality, did you have enough fuel gas supply pressure prior to starting the changeover?

 

Well the pressure was checked prior starting..it was in changing conditions....well we even tried to save the turbine from tripping. by returning to distillate but that did not help to......

 

The first question to be answered (by suhail) is: Have you ever had successful fuel transfers in the past, specifically when going from liquid to gas fuels as in this case?

The fuel splitter software used in most Mark V CSPs reduces the fuel reference for the fuel being transferred from while it simultaneously increases the fuel reference of the fuel being transferred to. The fuel splitter software Speedtronic doesn't check to see if fuel is actually flowing when this reference "ramping" is occurring; it just assumes the fuel is flowing as it should be.

It's pretty clear from the data you provided that when liquid fuel was being decreased, gas fuel was not flowing. There's not likely any other reason for the loss of load.

And, then the gas fuel suddenly did start flowing, which caused the large jump in output. The low fuel flows and then sudden high fuel flows likely resulted in loss of flame in one or more combustors which caused the Combustion Trouble alarm, which , if it had persisted for more than 9 seconds (typical value) would have resulted in a turbine trip.

You talked about the generator breaker tripping on loss of excitation, but it's not clear if the turbine tripped or remained running. The generator breaker will usually trip on loss of excitation.

 

The Fuel Splitter sequencing in the Speedtronic has a ramp rate for reducing the fuel being transferred from/fuel being transferred to (usually 30 seconds from the start of the transfer after some period of "pre-fill" where the control valve of the fuel being transferred to is just slightly opened to try to fill the fuel lines to prevent a small dip in power output).

More fuel equals more torque. More torque equals more amps. More amps equals more load.

In the same fashion, less fuel equals less torque, and less torque equals less amps, and less amps equals less load.

So, if fuel isn't flowing then torque production will be reduced and amps will be lowered and load will be lowered.

Again, the Fuel Splitter doesn't check to see if fuel is actually flowing; it just ramps the references equally up and down presuming that both fuels are flowing. The Speedtronic was reducing the liquid fuel reference (and flow) and that's evident by the fact that load dropped. And it was also increasing the gas fuel reference but apparently there was not sufficient flow to keep up with the rate of the fuel transfer ramp, so that is also indicative of loss of load, and likely combustion trouble (loss of flame in one or more combustors without a flame detector) which would have caused high spreads.

It's also conceivable the sudden increase in gas fuel flow (if there was one) could have blown out the flame in one or more combustors causing the combustion trouble alarm. Or, that the decreasing liquid fuel flow with no gas fuel flow caused a loss of flame in one or more combustors and the combustion trouble alarm.

You might have ensured all manual gas fuel valves were open prior to the transfer, but that doesn't mean that gas fuel actually flowed. The pressure might have been good at the start of the transfer, but what happened to the gas fuel supply pressure during the transfer?

Where does the gas fuel supply come from? Is there a compressor to supply the gas fuel, or is there a pressure regulator to reduce the pressure from a high-pressure gas pipeline to a suitable pressure to supply the gas turbine?

I've never tried to "abort" a fuel transfer in the middle of a fuel transfer by selecting the original fuel. I don't know what would happen in that case; I'd have to look at the CSP for your unit to be sure, but I'm pretty confident that if the Speedtronic accepted the command that it would just start ramping in the other direction at the same rate, so liquid fuel flow would just start increasing at the same rate as it was decreased until the reference (and hopefully) the flow was back at 100% of reference. Simply selecting liquid fuel wouldn't have instantly returned the liquid fuel reference (flow) to what it was at the beginning of the transfer.

Again, fuel is (ultimately) load. If load drops, then fuel has dropped. It doesn't get much simpler than that.

The opposite happens when you load a unit; increasing the fuel increases the load. Or when you unload a unit; decreasing the fuel decreases the load. At any time if the fuel (flow) decreases for any reason the load will decrease; or, conversely, if the fuel (flow) increases for any reason the load will increase.

Lastly, please provide a list of all alarms (Process- and Diagnostic) that were present before, during and after the event. Because, unless there were some related Diagnostic Alarms or Process Alarms, this likely was not a Speedtronic problem.

It would also be helpful to know the SRV and GCV positions during the event, especially if the SRV went to 100% during the transfer, which would definitely indicate insufficient gas fuel supply pressure/flow during the event.

 

Dear csa

Fuel transfer in normally done according to the availability of gas for the station...and we had many successful transfer with out any problem..

on both transfers

As you say the gas was not there at the initial stage of the transfer and the suddenly started flowing...answer to that we also have pressure switch 63fg-3 ahead of SRV Which wont allow the transfer if pressure is low with an alarm..

About your say that Generator excitation loss alarm comes with generator breaker tripped is true. and agreed
Finally the breaker trip in my case and the turbine goes on cool down. but later after tripping we gave a start again on fuel gas within a duration of 20 mins..the turbine starts normally with spread high during acceleration MAx allowable 185 and spread 1 at 142. later on fsnl returning to normal values.