Hi 9FA users.

Gas leak test B failure, I have doubt on Gas leak test sequence in frame 9FA DLN 2.6+.

After SRV opens for 1 sec to trap gas pressure in P2 cavity maximum pressure in the P2 cavity is latched, and the latched value is reduced by 150psi, and a logic is generated V86GLTB which is the reference for comparing the value with FPG2 in GT block. THIS FUNCTION TESTS THE LEAK OF GCVs or SRV??

If FPG2 > V86GLTB then it is testing SRV passing, can any explain this in detail to understand easily.

 

Hi, JP,

Many people have been surprised to learn the gas leak test tests both GCV <b>and</b> SRV.

The "S" in SRV stands for STOP, and in fuel stop valve. The gas valve leak test tests both the SRV and the GCV downstream of it (or at least it used to).

A leaking SRV can cause problems with firing, and since the primary purpose of the SRV is to act as a fuel stop, or shutoff, valve it's very important to the integrity of the machine that it seat and stop fuel flow when it's closed.

Some (many, if not most) F-class units seem to have been fitted with what were called Auxiliary Stop Valves upstream of the SRV. Sometimes this was done because of local, regional or national regulations, and it seems GE may be standardizing on the use of Aux. Stop Valves on many heavy duty gas turbines these days, with only some installations using a solenoid-operated vent valve between the Aux. Stop Valve and the SRV (consistently inconsistent, some would argue).

A properly functioning SRV is more important than a properly functioning GCV as, again, it's the primary fuel shut-off/stop valve in addition to being the speed ratio valve (another widely used definition for SRV, but the STOP function is more important than the SPEED function).

Hope this helps!

 

Dear CSA,

Thanks for replying to my post, I understand the concept and use of SRV in GTs. I am not able to clearly understand leak test logics, that is L86GLTB fail logic is for SRV passing which I understood from the logics, but actually machine tripped when the pressure in the P2 cavity dropped below 17 bar and we found GCVs were passing heavily.

Can you please explain in details to understand, L86GLTB is for SRV passing and if P2 exceeds above 150 psi in 30 seconds comparing with maximum latched pressure in greater than block.

Thanks in advance.

 

Hello,

Send me an email to [email protected] and I will forward some nice info.

Regards.

 

JP,

Without being able to see the exact sequencing running in your Speedtronic panel, I can't be much more specific.

Also, I can't really understand what you're asking about. It seems you understand the logic, which is usually pretty straightforward, but have a problem with the timing or the setpoints.?.?.?

I suggest you work with the packager or OEM to sort out the details. If valves are passing, it's a problem, right?

I've seen people complain about the length of the test; I've seen people complain about the setpoints of the test; I've seen people complain about the need/requirement to perform the test; people like to pick depending on what their point of reference is (it takes too long; why is THAT pressure chosen as the alarm/trip setpoint?; why do we have to do this test at all?; etc.; etc.; etc.).

I don't really understand what your complaint or point of reference is, and I can't see the sequencing running in your machine so I can't be more help.

There are usually two trip logic signals, each driving an individual alarm for clarity, which drive a single logic which goes to the L4PRE or L4PST which goes to the L4T logic to drop out L4. That's what I'm accustomed to seeing, but it sounds like your logic might be slightly difference.

 

Dear JP,

Not all controller code is the same so this may not be 100% applicable to your unit. There are typically two tests that are ran as part of the GasLeakTest task.

One test is ran to check that the Stop/Ratio valve is not leaking by, and or the Aux. stop valve if equipped. The vent valve for the interstage is closed and the P2 pressure monitored to look for any rise in pressure that would indicate the SRV leaking by.

The next test is ran to check the leakage of the gas control valves and vent valve. The vent valve is closed, then the SRV is pulsed open momentarily to charge the P2 cavity, then closed. The pressure in the P2 cavity(FPG2) is monitored for a period of time and compared to the constant to check for leakage from any of the GCV's or vent valve that is above the allowed amount. Depending on the leakage rate of the any of the valves the machine may be inhibited from starting, or only an alarm may be generated.

I believe that GasLeakTest A tests the integrity of the SRV to make sure it is not leaking by into the P2 cavity.

GasLeakTest B tests the integrity of the GCV's and vent valve.

There may also be a third test test that is ran, testing of the vent valve. When the P2 cavity is vented the time for pressure to decay is measured to verify the vent is opening and not restricted.

But the caveat is that your code may not have all these tests. I hope this helps some. The testing logic is somewhat complicated and does take a bit to digest.

 

Dear CSA,

Thanks for responding again to my post,

I have no complaints...My question is if L86GLTB is for SRV leak, pressure should increase more than latched pressure (25 Bar).
In our case P2 maximum pressure was latched to 25 bar and machine tripped when the pressure dropped to around 12 Bar within 30 sec, and the result was GCV3 and GCV4 passing. All passing valves were replaced and machine was put back in grid.

Alarms received
L30GLTSU_ALM
L86GLTB_ALM
L4PRET
L4T.

thanks and regards
JP

 

JP,

MIKEVI is correct. L86GLTB is the "B" part of the the gas leak test. In the "B" test, the SRV is momentarily opened and then closed to pressurize the P2 volume. The pressure is then monitored and if it drops too much, L86GLTB will cause a pre-ignition trip. As MIKEVI said, this is a check of the gas control valves and the P2 vent valve only. The SRV is checked during the part "A" test, which closes the vent valve and checks for excessive rise in P2 pressure. The part "B" check does not check the SRV, the part "A" check does.

 

JP,

It would seem from the information that you provided--and the information from MIKEVI--that L86GLTB is not for SRV leak, but for GCV leak. Again, without being able to see the sequencing/application code running in the Speedtronic panel at your site it's impossible to say for sure based on the information provided.

As MIKEVI said, it seems some gas leak test logic tests both the SRV and the GCVs, separately, which is something I haven't encountered myself.

You have correctly determined that if the SRV leaked during a test the pressure in the intervalve volume would INCREASE, and that if the GCVs leaked the pressure in the intervalve volume would DECREASE.

I wonder if your dilemma stems from the "longnames" given to signal names. Sometimes they are just flat WRONG. People change code and don't change the longnames--the descriptions of what the signal is or does. It happens. It happens a LOT more often than it should. The longname <b>does NOT</b> dictate what the signal does; it's only meant to be an indication, a "nice-to-have" if you will. The problem comes from when they are wrong, they can be very misleading.

But you have done exactly what you should have done: You have thought through the various conditions and determined how it really works. Are the longnames incorrect? Is the Process Alarm text message incorrect? (I've seen that happen, too; you haven't provided the texts of the Process Alarms....)

>Alarms received
>L30GLTSU_ALM
>L86GLTB_ALM

From the signals above, I can <b>guess</b> what they are <b>intended</b> to do--but that doesn't mean what they <b>actually</b> do.

L30GLTSU_ALM
This is the logic that annunciates an alarm for 'Gas Leak Test - Start-up'. Probably a single alarm message used for multiple conditions. What exactly is the Alarm Text for this logic signal?

L86GLTB_ALM
This is a lock-out (meaning it likely requires a Master Reset to clear/reset the condition once it has been detected) alarm message; the GLT portion is clear, the B is not, but from the information you have provided it would seem it's for the GCV leak test.

The other two signals are not alarms; they are EVENTs. And it would seem it's as supposed: the gas leak test energizes the Pre-Ignition Trip (L4PRET) logic which de-energizes the L4 (Master Protective) logic which results in a trip.

(Please remember that many people read these posts, so I try to provide as much information for them as possible. You, JP, likely know what many or most of the signals do but there are others who are freshers to this.)

 

Dear CSA / MIKEVI

Thanks very much for responding to my post. Can you please through some light and correct my understanding if I am wrong.

Test A L86GLTA - gas leak test A failed
L86GLTA = FPG2 > K86GLTA (100 PSI) within 30 seconds
If FPG2 > K86GLTA in more than 30 seconds Test A is successful. This indicates at the beginning of leak test SRV is closed FPG2 is zero if tight closed, in case if there is passing in SRV it should not exceed 100 psi in 30 seconds (pressure up to 99 psi is allowed within stipulated time 30 seconds).

After TEST A is completed, SRV opens for ONE second to trap pressure in P2 (FPG2) and latch the maximum recorded pressure (FPG2LATCH)
TEST B - L86GLTB Gas leak test B failed
L86GLTB = V86GLTB > FPG2 within 30 Seconds
V86GLT is calculated from FPG2LATCH - K86GLTB (150 psi) for example 400 – 150 =250 psi (V86GLTB) . IF GCVs are leaking heavily how V86GLTB can be greater than FPG2?, FPG2 is interstage gas pressure between SRV and GCVs.

 

Dear JP,

I will be reviewing code I have for a newer 7FA. According to the code I have, I believe your understanding of the logic is not quite correct, but I want to take the time to put together a correct interpretation of the logic as I understand it. Unless someone beats me too it, please standby!

 

I think you are missing the fact that if the gas control valves are leaking, FPG2 will be dropping during this test. Therefore, if V86GLTB becomes greater than the falling FPG2, you fail test B.

 

Thank you, be_anon,

Although the Mark VI COMPare blocks can do greater than and less than and greater than or equal to and less than or equal to, because early COMPare functions in Mark IV and Mark V could only do greater than, GE did a lot of "inverting" of comparisons in Mark VI--just to be "consistent" with what they'd done in the past.

Reading the equation from right to left it says, when FPG2 is less than V86GLTB then the logic is "1". Reading when from left to right it says the inverse: When V68GLTb is greater than FPG2 the logic is "1". They are both correct, and the inverse of each other.

Even if it meant that being consistent was more confusing.

Many people fail to recognize that when the term on the right of a greater than symbol decreases below the term on the left of the symbol that it still evaluates to true (logic "1"). It doesn't always mean the term on the left of the greater than symbol has to increase to be greater than the term on the right. GE turbine control sequencing/application coding quite frequently makes the term on the left of the compare symbol a "constant" and the term on the right a "variable".

It can be confusing, though.

 

Firstly it should be K86GLTB and it is either set at a %age of the checked P2 or an absolute value.

I have a complete write up on the function of the Gas Leak Test for a Frame 6, but it is pretty much the same for all machines. Immediately, if anyone wants a copy, drop me an Email to [email protected], I will try to post it to a file share in the next day or so and update.

 

Thanks anon,

You are right sometimes when we are troubleshooting we get freeze at some place and don't think out of box.