Hi all,

My gas turbine (GE Frame 9E) CPD dropped 0.5bar within 2-3 hours and load dropped at about 10MW. Prior to the incident, no work was carried out and no abnormal alarms were annunciated. The significant drop was partially due to higher ambient temperature. On a day to day operation comparison, the load supposed to dropped about 3-4 MW and not 10MW. The followings are inspections that we have carried out but to no avail:

1. Inspected bearing #2 venting and sealing piping and flange.

2. Checked waterwash isolation valves

3. Shut isolation valves of compressor bleed valves during baseload (no improvement of CPD or load)

4. Inspected all porous stone filters (FA-3,-4 and -6)

5. Performed boroscope inspection (no findings)

6. Inspected CPD transmitters

7. Inspected atomizing air line in accessories compartment

8. Inspected AD1

9. Inspected VIGV (measured at site)

10. Listened for noise during cruising down and cranking

11. Replaced VA-18 (found minor leak at piston)

12. Flame scanner gasket replacement

13. Atomizing air cooler - gasket replaced due to flange minor leak.

14. 20CB union joint tightened due to minor leak found

15. Air filters physical conditions and manometer readings.

16. Checked all the constants on HMI and compared with other GT. All normal.

17. False start drain

18. Bearing temperature (same before and after incident)

19. Inspected AE5

Ever since the incident, the gas turbine efficiency has dropped. We have also noticed that during FSNL before the incident, the CPD was at 6.4bars. After the incident, the CPD was at 6.9bars during FSNL. Troubleshooting is still being carried out. At baseload, we are only getting 9.8 bars instead of 10.3bars before. Although, the CPD was higher than before during FSNL. We appreciate any feedback.

 

> My gas turbine (GE Frame 9E) CPD dropped 0.5bar within 2-3 hours and load dropped at about 10MW.

Inspected CPD transmitters:

Does this mean you have verified the instrument accuracy, as seen by the turbine controller?

Have you blown these lines out, just to make sure no water is trapped any where?

Are the lines free of any leaks?

first thoughts.

 

> My gas turbine (GE Frame 9E) CPD dropped 0.5bar within 2-3 hours and load dropped at about 10MW.

2nd thoughts

Have you looked at your IGV reference; this maybe changing based on a temperature issue. CPD/exhaust.

L83TVON_CMD
L3TCI

 

Have you looked at your control spec. to see how much output you should drop / Deg. C inlet Temp?

Was it particularly humid or raining that day? this can give a big temporary drop across the Air Inlet Filters.

 

Dear all,

We took the opportunity during shutdown over the weekend to inspect the IGV. It was found that at site angle was 7 degrees higher than what it should be. Hopefully this is the culprit.

That will explain why we got higher CPD during FSNL because instead of 57 degrees, the actual opening was 64 degrees.

Upon baseload, the CPD was at 9.8 bar. We suspect that the actual opening during baseload was 86 + 7 which is 93. At this angle, the IGV was actually closing a little and thus affecting our CPD and Load.

We will be starting up today. I shall keep you guys posted on what happens.

 

RY,

Thanks very much for the feedback.

However, There is normally a mechanical stop on the ring gear that prevents travel at the open end beyond approximately 89 degrees. And, most of the IGV actuators I've seen also have a mechanical stop that prevents the actuator from traveling too far. In fact, it's a collar that's machined to prevent excessive travel of the actuator rod (which usually moves down to open the IGVs).

So, it would seem that something<b>S</b> (plural) are amiss mechanically. Not to mention an improper LVDT calibration.

This is an overly simple explanation for a very serious problem with the IGV actuator/mechanism.

 

Dear Friend,

Kindly check following points.

1. Cleanliness of Turbine air Inlet filters, lower the DP, better the load

2. Compressor fowling - regularly clean compressor, offline compressor water wash produces better results

3. Change in ambient temperature, humidity factors directly affects the load

4. Check any kind of compressor air leakage.

5. check CTDA temperature before and after any maintenance carried

6. check exhaust ducts, increase in backpressure, check exhaust thermocouples for some reading higher temperature

7. Check IGV angle perfectly asp per guidelines

Hope some of this points will further help in your diagnosis

Mambo

 

Dear all,

We have managed to solve the problem. It was due to the excessive IGV angle opening. This was due to the lock nut not tack welded to the linkage causing it to drift slightly.

Subsequently, the tack weld was done and the IGV was recalibrated to its original design (32 - 86). I hope this would be a good learning lesson for others as well.

 

Oh yeah, i forgot to mention that the we have also noticed on the other unit that the CPD increased to more than usual (about 0.5 bar) and the load dropped by 3-4MW.

Mech took out the atomizing air booster for overhaul. So the unit was running without atomizing air booster. The piping connection to atomizing air booster inlet was blanked with a rubber gasket. During start up of that unit on gas, there was loud hissing sound due to air leaking out from VA22. VA22 wasnt fully shut. Atomizing air which came out out it heated up the 96CD and caused the CPD read in the hmi to be 0.5bars higher and the load to dropped by 3-4 MW.

Can you explain how the rise in CPD can cause the load to drop?

 

Hi CSA,

I am still waiting for your feedback. Appreciate it.

> Can you explain how the rise in CPD can cause the load to drop?

 

RY,

I didn't respond because <b><i>based on the information provided</i></b> I don't understand how a rise in CPD could cause a drop in output. And, it's not clear what effect replacing the gasket had (made the output return to near normal, or possibly even increase). Further, it's not clear if the CPD remained high and load low after the Main AA Compressor were replaced; so there's just not enough information to hazard a guess--which is all any response <b><i>based on the information provided</i></b> would be: a guess. Actionable data is necessary to understand problems and to provide meaningful responses, and there just wasn't enough information.

In my personal opinion, it was not advisable to operate the unit (even temporarily) without the Main AA Compressor--which provides cooling and purging air to prevent the backflow of hot combustion gases into the AA passages, and if I recall correctly, also to the liquid fuel nozzles.

So, if the Main AA compressor was re-installed and everything returned to normal (CPD and load), then it was an aberration caused by operation without all components installed and working correctly, and therefore is not a continuing issue worth spending a lot of effort on trying to understand. Or, if replacing the gasket and continuing operation resulted in some change to the conditions which was not reported then the problem description is not complete.

Hence, the reason for not responding (and also a possible reason other readers didn't respond???). GE-design heavy duty gas turbines are not complicated machines but there are knock-on effects on other systems when operated abnormally, or when abnormal operation occurs. And, lastly, this is (hopefully) a temporary condition which will be resolved when the refurbished Main AA Compressor is installed. Expecting normal operation with abnormal component configuration just isn't a reasonable expectation.

 

Hi CSA,

Sorry for the lack of information. You may have overlooked the reply above. I will explain again. The mechanical team took out the atomizing air booster (not Main atomizing air) for overhaul purpose. They installed a blank on the inlet piping to the atomizing air booster. During running, VA22 would be shut. On that day, the blank became loose and the VA22 was slightly passing.

Thus, the atomizing air was leaking out through the blank. Subsequently, it heated up the 96CD (CPD transmitter to mark v) and caused the CPD to be higher by 0.5bars as read on the HMI. This inadvertent increase caused lower MW output on the machine.

We rectified the problem by manually shutting by tightening VA22. After we did that, the CPD decreased and load increased to like it was before. The VA22 has since been replaced during the shutdown opportunity.

My question is, how did the inadvertent heating of 96CD (and thus higher CPD reading on HMI) reduce the MW output?

 

RY,

Yes, I did miss the word booster; my bad. But I still fail to understand how heating of the 96CD transmitter and higher than expected CPD readings could lower MW. AND, you've <i>never said</i> if the unit was operating at Part Load, or Base Load.... The only time CPD feedback affects power output is when the unit is operating at Base Load. (Is that part of the problem here? Does someone think CPD <b>feedback</b> always affects power output? I'm being very specific when I say 'CPD feedback'--because, actual CPD does, in fact, affect power output at all loads, but the only time the control system biases fuel flow with CPD feedback is when CPD- (or CPR-)biased exhaust temperature control is active--under most normal conditions.)

And, you are correct--<b><i>if the unit was operating at Base Load AND with all other aspects of operation being normal</b></i> a higher CPD should have resulted in a higher output--especially 0.5 barg. But, you've never told us if the unit was operating at Part Load, or Base Load, when this aberration occurred. (You actually said, "...During start-up of that unit..." which, to me, doesn't mean the unit reached Base Load. This is a technical business we're in, and words do mean something (a lot, actually--like me missing "booster".).

Again, aberrations are difficult to explain, and throw in the the lack of information and it gets even tougher. Especially when all of the facts aren't known. I dislike guessing--a lot. Too many people do it--especially in power plants (think servos and "calibrations" for starters). And just because so many do guess doesn't mean everyone (or anyone) should. The human Lemming population is too great as it is--and being a Lemming is highly over-rated.

If you want me to say it, I will: I don't have a clue about why a "heated" CPD transmitter would cause an elevated CPD and a decrease in load. Frankly, I'm surprised the CPD transmitter was in a location where a leak of nearby piping could impinge on it and cause it to get over-heated. Usually, they are mounted on the side of the Accessory Gage Cabinet, or on the base plate underneath the Starting Means. I've even seen them mounted on the roof of the turbine compartment (an EXTREMELY poor location from a maintenance perspective) but never near the AA Booster Compressor. And certainly not near enough to allow a leak to impinge on it.

So, without being able to see a picture of the mounting location and piping in the surrounding location, and without a LOT Of other information, I won't hazard a guess--for all of these reasons, and the 'don't-wanna-be-a-Lemming' reason above. (Not even with more information dribbling in.) And, again, relating something in one's mind doesn't mean they are, in fact, related. (Doesn't mean they're not--but without a much deeper understanding of the facts and configuration, well, you get the idea.)

Add this to your list of lessons learned: Aberrations can be difficult, if not impossible, to explain.

 

Sorry CSA for the lack of info. The unit was running on baseload. The 96 CD is situated below the platform where the gearbox is. I suspect that the heat from the leak caused the air in the tubing to expand. This caused the cpd to increase by 0.5 barg, although in actual case the cpd was not that high.

 

Dear guys

Just share my though on Alstom GT13E2 in term of baseload close loop control.

CPD & Exhaust temp are factors of the TIT formula. In case of higher CPD than design it'll causing TIT limit (i.e 1100 C) reach earlier than normal and result to fuel reduction. Therefore end up with load drop.

I also facing the same problem (high CPD than normal baseload) causing load drop. This load drop wasn't appeared in the range low load up to TIT control. Issue still under investigating and appreciate expert inputs to be shared.

 

Sobri,

<b>When operating at Base Load (CPD-biased exhaust temperature control)</b> when CPD <b><i>increases</b></i> load should <b><i>increase,</b></i> and when CPD <b><i>decreases</b></i> load should <b><i>decrease.</b></i> So, it's very difficult to understand how load increases when CPD increases. Very difficult.

You seem to be confusing TIT (Turbine Inlet Temperature) and exhaust temperature. <b>When operating at Base Load (CPD-biased exhaust temperature control)</b> as CPD increases the turbine control system increases fuel flow-rate to maintain a <b>constant TIT</b> (also called "firing temperature"), <b>BUT</b> the exhaust temperature goes down. Yes--<b>exhaust temperature decreases as CPD increases <i>for the same firing temperature (TIT)</i></b>. Look at the CPD-biased <b>exhaust temperature</b> control curve--it has a negative slope (downward and to the right). This means that as CPD increases the exhaust temperature <b>for a constant firing temperature (TIT)</b> will <b>DECREASE</b>. It's counter-intuitive--but that's the way it works.

CPD-biased exhaust temperature control holds (maintains) a constant firing temperature (TIT). Not a constant exhaust temperature--exhaust temperature varies as CPD varies <b>for a constant firing temperature (TIT) when operating at Base Load</b>. Yes--it's called "exhaust temperature control"; but exhaust temperature is being varied as CPD varies in order to maintain a constant firing temperature (TIT). And, firing temperature doesn't change when operating at Base Load--exhaust temperature changes. And, it changes opposite to what one would expect--it decreases as CPD, and load, increases. And, it increases as CPD, and load, decrease. That's just how gas turbines work--<b>when operating at Base Load (CPD-biased exhaust temperature control)</b>.

Yes, as the machines is loaded from zero load to Base Load, the exhaust temperature, firing temperature (TIT) AND CPD increase. But once the unit reaches Base Load (primary exhaust temperature control; CPD-biased exhaust temperature control), firing temperature (TIT) is held constant as CPD and exhaust temperature vary. And, as CPD increases exhaust temperature will decrease for the same (constant) firing temperature (TIT). Again, it's counter-intuitive--but that's the way it works.

I wonder if the turbine control system is operating on the back-up exhaust temperature control curve instead of the primary (CPD) exhaust temperature control curve. This has caused similar problems in the past. Older machines used FSR (Fuel Stroke Reference) as the back-up parameter for exhaust temperature control

Hope this helps. It's very difficult to understand how load can decrease when CPD is higher than normal. I suppose if the CPD transmitter (sensor) isn't calibrated properly (or if the CPD transmitters--there are sometimes multiple, redundant CPD transmitters) that could cause a problem. But, in general, <b>when operating at Base Load (CPD-biased exhaust temperature control)</b> as CPD increases load increases, and exhaust temperature decreases <b>for a constant firing temperature (TIT)</b>. It's very difficult to imagine otherwise for a gas turbine operating on primary (CPD-biased) exhaust temperature control (Base Load).

 

I am working in GE 7fa, DLN 2.6 machine. When i was looking over trip diagnostic, there is a protection named L3TFLT (Compressor discharge pressure low), and the explanation given for this as follows: Trip when compressor discharge pressure drops below CAKCPD psig(35psig)while in operation.

When in operation what are all the causes which can lead cpd to became too low up to trip limit.

 

CPD is SO critical to DLN control that the unit cannot be operated safely and without damage that the protection trips on low CPD readings.

Possible causes for low CPD readings are plugged sensing lines, manual isolation valves not fully opened after off-line compressor water washes and maintenance outages, loose tubing fittings, failed or failing transmitters, loose wiring terminations, etc. All of these causes are not related to actual low compressor discharge pressure.

If you're asking about actual low compressor discharge pressure, it's extremely unlikely that the actual flow through the compressor would ever be so low as to cause the CPD to be that low--under normal running conditions. The compressor would likely be severely damaged and the unit tripped "long" before the CPD ever actually got that low at or near rated speed.

The value of that Control Constant has always been the source of some discussion, but since there should be three redundant CPD transmitters as this is a critical operating parameter the likelihood of two failing successively is low, and there would be Diagnostic-and Process Alarms to warn the operators/technicians of a single failure so that it can be resolved before a second fails.

That trip logic usually causes trips only when the CPD transmitter manual isolation valves are not opened after washing or a maintenance outage, and prevents operation from possibly damaging the unit. It's more of a safety "check" (sometimes called "sanity check") than anything, as, again, the likelihood of the actual CPD ever getting that low under normal running conditions is basically zero. Most operations managers are very upset when they learn the control system tripped the unit simply because someone forgot to open the manual isolation valves--not that someone forgot to open them but that the control system tripped the unit without warning because the valves weren't open. They usually want that protection disabled, instead of performing proper checks of valve positions before re-starting.

Hope this helps!!!

 

I work in a combined cycle plant with two 7 fa's Nooter Erikson HRSGs and a Toshiba Steam Turbine. The plant was commissioned in 2002. I have regular issues with the units going into backup temperature active while loaded or loading unless in base load or temperature control. The isotherm temperature setting is 1200 degrees f and the TTRx2 is 30 degrees f.

We hover around 1200+ a couple degrees most of the time running. Is the back up curve active because of issues with the CPD or any other points?

Richard

 

careeroperator/Richard,

Is there a reason you are specifically asking about CPD? Are there CPD-related Diagnostic Alarms?

Does CPD seem to be abnormally high or low for the current ambient and machine conditions?

Have the emissions changed recently, requiring a re-tune, or a liner dilution hole size change?

Personally, I've never understood why the OEM continued to use Back-up Exhaust Temperature Control for DLN units. Typically, back-up exhaust temperature control comes into effect when TTRXS (Turbine Temperature Reference-wXhaust, Secondary) is lower than TTRXP (Turbine Temperature Reference-eXhaust, Primary). And, the secondary, or back-up, exhaust temperature reference for a lot of machines is based on FSR. If FSR is "high" and the back-up exhaust temperature control curve Control Constants are not set correctly, then back-up exhaust temperature control (TTRXS) can become lower than primary exhaust temperature control (CPD- or in the case of most F-class units and MANY newer GE-design heavy duty gas turbines, CPR-biased exhaust temperature control).

CPR bias is a fancy CPD bias; it attempts to squeeze a few extra hundred kW out of a gas turbine--BUT the ambient pressure transmitters (96AP-1, -2 & -3) MUST be working properly, as well as the the compressor discharge pressure transmitters (96CD-1, -2, & -3) for the CPR calculation to work properly.

The ambient pressure transmitters are usually very touchy devices. And, they are usually mounted in a NEMA 4 junction box, and connected to ambient air pressure via small (1/8" or 1/4" tubing), and to prevent insects from making their nests in the open end of the tubing there is usually a small amount of a plastic scouring-pad like material. This usually gets dirty over time, or usually falls out of the tubing, and then insects make their nest inside the tubing.

Or, the isolation valves of the tubing arrangement for the Performance Monitoring Cabinet (the NEMA 4 enclosure which houses the AP pressure transmitters and other pressure transmitters) end up in the wrong position.

Or, the calibration of the transmitters can't be done with plant equipment, or the calibration company hired to do the job doesn't have the proper equipment.

Or, someone changes the range of the transmitter during calibration without changing the range in the Mark*.

If the fuel supply has changed significantly over time since commissioning without any change in the Control Constants, it can affect FSR--and it can also affect emissions tuning.

You get the drift--a LOT of things can go wrong.

If you have operating data from years past, you should be able to check CPD, AFPAP (the name of the ambient pressure input, if I recall correctly), and FSR for the same ambient temperatures as you're currently experiencing, and see if things are similar or very different.

Back-up exhaust temperature control, using FSR, is intended to allow the unit to continue to operate if the compressor discharge pressure transmitter feedback goes low. HOWEVER, on most DLN units, if the CPD drops below some setpoint, the unit will trip--again, because CPD is SO important to the protection and proper operation of the machine.

But, when the gas system (including the gas fuel supply (heating value; BTU content; etc.) gets out of whack AND the back-up exhaust temperature control curve Control Constants are marginal to begin with, then TTRXS can become lower than TTRXP--which will make TTRX equal to TTRXS, which will be an "artificial" limit on the power which can be made.

The back-up exhaust temperature control curve Control Constants are supposed to be calculated such that the back-up exhaust temperature control curve mirrors the primary exhaust temperature control curve. BUT, it's very difficult to compare curves which use two different biases--CPR (or CPD) and FSR. More often than not, one of two scenarios happens; either the TTRXS Control Constants are such that the back-up exhaust temperature control curve is too close to the primary exhaust temperature control curve, OR the TTRXS Control Constants are such that the back-up exhaust temperature control curve is too far away from the primary exhaust temperature control curve.

If TTRXS is too close to TTRXP then the unit can go into and out of primary- and back-up exhaust temperature control, or remain in exhaust temperature control. This can limit the power output of the unit. If something happens to cause TTRXP to rise and makes TTRXS lower than TTRXP then the power output of the unit could be higher than design--resulting in over-firing of the unit. Neither condition is really good, though the latter is potentially worse than the other.

So, the whole concept of some kind of back-up to a loss of CPD is kind of foolish with the current set-up. If the CPD fails (which would require multiple transmitter failures), the unit will trip. And, if the TTRXS Control Constants aren't calculated properly, then problems can occur. It's really not the best set-up--in my personal opinion.

If the machine is worn out (worn compressor; dirty inlet air filters; worn turbine section (nozzles and buckets)), then it could be possible (though not very likely) that problems with CPD could be real and causing back-up exhaust temperature control to become active.

Having said all of the above, it would be best to have someone knowledgeable about GE turbine control systems to come out and have a look at past and current data. It's NOT a good idea to run a DLN unit on back-up exhaust temperature control for extended periods of time as the unit could be over-firing (under extreme circumstances) and that could cause increase wear on the hot gas path parts.

Please write back to provide the answers to the above questions, and to let us know what you discover as you work through restoring the unit to primary exhaust temperature control.