We are having 6 nos of Gas Turbine (GE supplied Frame 9E m/c, model no -PG9171E) controlling through Mark vi technology. All the m/c normally running on Natural Gas.

The gas fuel vent valve close feedback (L33ps1c) was lost and a sudden loss of pressure in the gas supply line occurred. This led to a change over to liquid fuel. The changeover was initiated due to a loss of FPG1 (l96fgl SRV inlet gas fuel pressure low).

The changeover signal L43FL_PRSET got picked up since the opening of SRV reached >95% for maintaining the downstream pressure FPG2. However during changeover to liquid the liquid fuel pressure also went below 10 psi which caused the signal L43FG_PRSET to pick up and the machine came back to total gas fuel despite having a low gas inlet pressure. And after 120 sec the signal L43FL_T got picked up which tripped the GT.

Want to know how gas vent valve close feedback loss will reduce gas supply pressure as there is neither any SOE or alarm observed of Gas vent valve open nor any open limit switch of gas vent valve found.

 

Hi,
Once you have solved the immediate problem, we have a customer with several Mk VI controllers and they are monitoring them and building a history (SOE) file for just this trouble shooting as well as better real-time monitoring. We have a native (read: our own code) scanner for the GSM protocol and this can run on different operating systems. I can tell you that a lot of time has been invested in the scanner! This can both help you next time and perhaps help avoid a turbine trip. Contact if interested.....

Peter

--
Peter Clout
Vista Control Systems, Inc.
2101 Trinity Drive, Suite Q
Los Alamos, NM 87544-4103

(505) 662-2484
FAX (505) 662-3956
Cell (505) 450-7810

[email protected]
http://www.vista-control.com

 

I would need to see your P&ID but it sounds like you have a block & bleed valve before the gas supply comes onboard to the SRV. Did this close on low gas pressure and open the vent? Did you have low gas supply pressure? Why did you only have 10 psi Liquid Fuel Pressure? Sounds like you have lot of problems there. Do have any SOE logs, trip history,etc. that you can post?

 

How old are these units? Is this happening to all 6?

Losing your vent valve close feedback,33PS-1 should not affect the gas pressure. Also the loss of the limit switch is not a direct logic trip. The reason it transferred was because the SRV >95% time delayed, caused by low FPG1. (Did all 6 units see a low FPG1?)

FSV352 is controlled by solenoid 20PS-1. The basic sequence for 20PS-1 is to energize during startup with L4 provided the gas shut off valve, FSV351, limit switch 33VS4-1 is closed. 20PS-1 will remain energized untill L4 drops out. Similary during startup the gas shut off valve FSV351 will open, energize 20VS4-1, once the FSV352 closed limit switch feeback is received. Once TRUE 20VS4-1 (l20vsx) latches itself and normally only drops out with L4. (There is some additional logic for gas leak test, fuel gas vent warming, fire and gas trips etc but these are in the back ground during normal operation). Once running the loss of status from any of the switches is only an alarm. Even during online operation if both the shut off valve and vent valve are simultaneously open this is not a direct logic trip. (I've had this philosophy confirmed by the OEM, and had relocate vent piping to a non-hazardous area. GE does not want their turbines to trip because of a failed switch)

In your case the vent valve is opening as can be seen by a low FPG1 the SRV >95%. I would suspect you have a problem with either solenoid 20PS, the wiring to it, or the actuating pressure. Is your site using instrument air or regulated PCV350 pilot gas off the main gas line to actuate the valves? I would check what the regulated pressure is, half a bar can make a difference. Also check the quick exhaust valves.

And remember the Alarm & Event log is your friend. Let us know what you find.

 

I've been reading, re-reading, and re-reading the initial post; watching the other responses; and then reading and re-reading and re-reading the original post again. Without the ability to see the P&IDs for these machines and the application code in the Mark VI it's very difficult to say for sure exactly what happened from the information provided. The originator didn't tell us what load was doing during this event and it's sub-events. And it doesn't appear that all turbines suffered from the low gas fuel supply pressure; just this one (I presume the "third" unit).

It seems that the logic signal driving the limit switch indicating the vent valve was closed went to a logic "0" (L33PS1C), which would indicate the vent valve was <b>NOT</b> fully closed, but was at least partially open. It would seem that along with this opening of the vent valve that the gas fuel supply pressure transmitter (96FG-1) indicated a low gas fuel supply pressure, and further that the P2 pressure went low causing the SRV to open more than 95%. (The SRV wouldn't open so much if gas fuel supply pressure was normal.) It's most likely that the cause for the fuel transfer was the low gas fuel supply pressure, and then the gas fuel preset was set to logic "1" or latched to logic "1" by the SRV opening more than 95%.

Then, when the liquid fuel supply pressure went low (for whatever reason....as liquid fuel systems are wont to do more often than desirable) the Mark VI initiated a re-transfer to gas fuel.

We don't know if there is a second limit switch on the vent valve to indicate fully open. The originator only told us about a limit switch that indicates fully closed, and when it's not actuated then the valve could be either partially open or fully open--but without a limit switch to indicate fully open we wouldn't be able to know anything other than it was not fully closed.

We don't know where this vent valve is located with respect to the gas fuel supply pressure transmitter, so we don't know how it's opening could (or could not) affect the pressure transmitter and possibly erroneously create a low gas fuel supply pressure condition even though one did not exist. <b>HOWEVER</b>, the fact that the SRV did go more than 95% open does indicate that there was, in fact, a low gas fuel supply pressure condition on this machine. (The SRV should be able to maintain P2 pressure reference without going open more than 95% even at full load, or Peak Load, if so equipped.)

Some sites use different types of filters (coalescing, etc.) and/or strainers upstream of the SRV to protect against the ingestion of liquids or solids through the gas fuel system. If these filters and/or strainers get plugged, then the flow-rate through them will decrease causing the pressure downstream to decrease causing the SRV to open more than normal, possibly leading to some of the issues the originator noted. But, without being able to see the P&IDs it's very difficult to know what the configuration is or what else might have happened.

But, for whatever reason it does certainly appear there was a loss of pressure (flow) upstream of the SRV on this particular unit, which led to a transfer to liquid fuel, and then some kind of loss of liquid fuel pressure, which led to a transfer back to gas, and then an inability to maintain fuel flow and load and a trip on loss of flame. All very understandable if the gas fuel flow/pressure to this machine dropped, regardless of vent valve position (I find it difficult to believe the vent valve could be so large that so much fuel was flowing out of it that the SRV would have to go open to maintain P2 pressure). And, again, if the liquid fuel pressure/flow wasn't sufficient, then the re-transfer back to gas would occur. And, if there was insufficient flow to maintain load and flame, then the unit would trip.

We would need to be able to see the application code in the Mark VI to see what was driving the solenoid valve.

It would also have been nice to know what the gas fuel supply pressure was doing one minute prior to the initial low gas fuel supply pressure condition, or five minutes prior to the initial low gas fuel supply pressure condition, or one hour prior to the initial low gas fuel supply condition. In other words, was the gas fuel supply pressure trending down, causing the SRV to trend open for some period of time prior to the initial low gas fuel supply pressure condition? Should a vigilant control room operator have been able to alert an outside operator to a potential problem prior to the event?

 

No first time in this GT only we accoutered such type of FPG1 low problem . And we are not sure whether that vent valve got opened or not (as its open limit switch l33ps2o not picked up) but FPG1 goes down below 295 psi (& due to that change over to liquid fuel took place). & as liquid fuel also having low pressure (found problem in FV 155) it again came back to gas despite of low gas pressure & at that time SRV opening reached above 95 % hence again change over occurred and after 120 sec the signal L43FL_T got picked up which tripped the GT.

We already checked 20PS solenoid v/v, wiring of it is ok. We are using instrument air for valve operation (Normal operating pressure of instrument air is 8.0 kg/cm2, which is supplied by utility dept.) & actuating pressure is also found ok. After tripping we checked 20PS operation by forcing it’s command we found every thing normal without doing anything after trip only close feedback limit switch (L33ps1c) wiring found loose & due to that only it’s feedback hunting observed.

 

Most of the time, limit switches on GE-design heavy duty gas turbine applications indicate one condition. In the example of the closed-position limit switch, L33PS1C, it indicates FULLY CLOSED. if the valve is anything other than FULLY CLOSED, then it's NOT FULLY CLOSED. It doesn't have to be fully open to be NOT FULLY CLOSED.

And, in my experience, the signal names for limit switches on a single valve are similar. For example, a valve with both a closed position limit switch and an open position limit switch would have signal names like L33PS1C and L33PS1O, the 'PS1' portion indicating the limit switches (33) are on a single valve, the 'C' for FULLY CLOSED and the 'O' for FULLY OPEN.

If the valve were neither fully closed nor fully open both limit switches could conceivably be a logic "0". Since most actuators and valve plugs have to move through a range, or arc, of travel, there would be a "deadband" between fully open and fully closed.

So, I question whether you are comparing the limit switches from the same valve, or from different valves, and whether you are aware that the logic states will generally go to logic "1" only at the ends of travel (FULLY OPEN or FULLY CLOSED) and will be logic "0" when NOT FULLY OPEN or when NOT FULLY CLOSED.

The Device Summary should list the conditions when the limit switches should change state.

But, I don't believe that the position of a vent valve (unless it's also venting P2 pressure--and can vent a <b>LOT</b> of P2 pressure!) can affect the position of the SRV. Or that vent valve has to be able to vent a lot of gas fuel supply pressure upstream of the SRV. Because something made the P2 pressure feedback go low causing the SRV to go open more than 95% for some period of time.

And unless there's something very amiss with the tubing and piping of the gas fuel supply system, something was restricting the flow, and hence the pressure, upstream of the SRV of this particular unit.

The originator has never said if the unit was re-started on gas fuel and runs at Base Load without SRV problems.

A Mark VI has a Trip History buffer, which could provide some useful information sometimes. Like P2 pressure prior to the events. Or SRV position prior to the events. Or gas fuel supply pressure upstream of the events prior to the event. (If it's configured to do so. Even if it's not, it should always be consulted.)

 

>But, I don't believe that the position
>of a vent valve (unless it's also
>venting P2 pressure--and can vent a
><b>LOT</b> of P2 pressure!) can affect
>the position of the SRV.

I agree if the original poster supplied a PID it would solve a lot of work it out your self nevertheless ....

Our 9E units have a double block and bleed arrangement. The first block valve FSV351(20VS) is just after the gas filtering skid. The second block is the stop ratio valve VSR-1 in the DLN skid. The connecting line is 6inch. Between these valves (recommended physically close as possible to the DLN skid by the OEM) is the junction for the vent valve FSV352 (20PS). This is a 4inch line.
VSR-1 has a T4 certification, which along with the auxiliary stop valve will satisfy the double block and bleed local statutes. The VGC-x do not have this. The P2 cavity vent controlled by 20VG-1 is a separate vent (0.5inch).

Believe me this valve can vent a lot of gas. The venting of a single valve has previously tripped off all units at our site due to low header pressure. We have disabled fuel gas warming during start up to prevent both valves opening at the same time. Currently fuel gas warming is a manual exercise via 1 inch maintenance valves. Of course the EPC/OEM can give us a final solution whenever they want.

 

Have you even dismantled the cylinder of Gas Vent Valve to check the seal ring? It is possible that had some leaking inside between the both sides. Sometimes, the pressure of instrument air too high will make this happened.