We are operating 2 gas turbine in synchronisation with national grid.
Both are Frame 6B with DLN-1 system with only gas firing. We run in part load 20 MW only @ atmospheric temp 45deg C. Recently control system was replaced with MarkVIe system and tuning was done. It was working perfect. For the last 3-4 weeks we are finding the CO emissions in the exhaust of gas turbine is increasing during the noon time (being in middle east atm temp reaches 45Deg C). HRSG firing also was stopped to verify the CO level. Still CO remains high. Analyzers were calibrated. Can someone guide how to proceed?

 

aksunil,

What is the NOx at 20 MW during the high ambient condition?

Do the units have any kind of inlet cooling (evaporative cooling; fogging; etc.)?

Does the unit have Inlet Bleed Heat, and if so, is it in operation at 20 MW?

Is the unit in Lean-Lean combustion mode at 20 MW or in Premix combustion mode at 20 MW? I don't believe emissions are guaranteed when the unit is not operating in Premix combustion mode.

What happens to the CO if the load is increased?

Why is the load limited to 20 MW when the ambient temperature is so high? Is it an Operations decision? Is it because of Exhaust Temperature Control (decreased CPD)?

When was the last off-line compressor water wash?

It has been common even in more temperate climates that DLN tuning has to be done during the coldest times of the year and again during the warmest times of the year. (This doesn't make a lot of owners/operators very happy....)

In the DLN-I combustion system, CO is burned by the diffusion flame in the secondary combustion zone to reduce the amount of CO in the exhaust. (It's a quandary--that to reduce NOx the temperature of combustion must be low, but that to reduce CO the temperature of the combustion must be high. Part of the joy of Dry Low NOx emissions reduction methods.) If the unit is operating in Premix combustion mode at 20 MW that would mean the fuel flow to the diffusion flame in the secondary combustion zone is not as high as it would be when the load was higher.

Lastly, I also believe that even if emissions are guaranteed in Premix combustion mode sometimes there is a lower limit to the load range where emissions are guaranteed--just because of this reason (because fuel flow to the diffusion flame in the secondary combustion zone is reduced causing CO, and even NOX, to be out of range).

Have you asked the packager of the turbine about this issue? If so, what has been their response?

 

> What is the NOx at 20 MW during the high ambient condition?

NOx level reaches up to 60ppmV during 14.00 hrs to 18.00 hrs when ambient is about 45 deg C.

> Do the units have any kind of inlet cooling (evaporative cooling; fogging; etc.)?

No inlet cooling system is Provided

> Does the unit have Inlet Bleed Heat, and if so, is it in operation at 20 MW?
Yes IBH is in line and machine is in pre mix steady state (Low NOx mode)

> Is the unit in Lean-Lean combustion mode at 20 MW or in Premix
> combustion mode at 20 MW? I don't believe emissions are guaranteed
> when the unit is not operating in Premix combustion mode.

Yes Machine is in pre mix mode

> What happens to the CO if the load is increased?

Remains on higher side, but varies from 45 to 60 ppm V during noon time 14.00 hrs to about 18.00 hrs. (Before 6 months it was 0 to 2 ppm V)

> Why is the load limited to 20 MW when the ambient temperature is so
> high? Is it an Operations decision? Is it because of Exhaust
> Temperature Control (decreased CPD)?

The GT is for captive power requirement of process plant . Our load required is 46 MW. So we are operating in part load .

> When was the last off-line compressor water wash?

4 months before

> It has been common even in more temperate climates that DLNtuning
---- snipped by moderator ----

> Lastly, I also believe that even if emissions are guaranteed in
> Premix combustion mode sometimes there is a lower limit to the load
> range where emissions are guaranteed--just because of this reason
> (because fuel flow to the diffusion flame in the secondary
> combustion zone is reduced causing CO, and even NOX, to be out of range).

We will study based on this in put. But machine is about 11 years old and machine used to run from 15 MW up tp 34 MW with no change in emission.

One more additional information I would like to share. The combustion hard ware of GT was changed from 16K to 32 K. This will increase the outage between 2 inspection. But after this machine is in line only for 5 months.

> Have you asked the packager of the turbine about this issue? If so,
> what has been their response?

No we are yet to communicate. Initially we were misguided and we thought it to be because of HRSG supplementary firing. After cutting off supplementary firing and analyzing from GT and HRSG stack it is now concluded, emission is from GT.

 

aksunil,

THANK YOU VERY MUCH FOR TAKING THE TIME TO ANSWER ALL OF THE QUESTIONS!!! That rarely happens, and it is sincerely appreciated.

Unless you aware of a change in the fuel being burned in the last few months I would have to presume that something has changed with one or more of the combustors. I neglected to ask if the exhaust temperature spreads have also increased during this time, in general, since ambients have hit their highest levels.

As you said, it was working fine before the "extender" hardware was installed, and I presume it was installed before the ambient reached its highest levels. I would again suggest that DLN tuning is likely necessary, at a minimum, and may be necessary again when the ambient dips to its lowest in a few months.

Another thing I might suggest is that if IBH is in service (I presume that's what "in line" means) and is extracting air from the compressor discharge casing and re-injecting it back into the axial compressor inlet that this may be contributing to the problem. This essentially "thins" the air being drawn into the axial compressor (done for protecting the axial compressor because the IGVs are closed below approximately 57 DGA). I'm no combustion expert but if the IGV LVDTs weren't calibrated properly and/or the positioner on the IBH control valve wasn't working properly.

And, again, I know of a few machines where the emissions wasn't guaranteed when the IGVs were less than 57 DGA (so, when IBH was in service at Part Load) even though the unit was in Premix combustion mode. Many times the expected emissions levels at Part Load (with IBH in service) is still less than the permitted emissions levels and the plant sells the credits when emissions are much less than permitted levels (so when the unit is operating at approximately 80% of rated load or above). Do you know what the emissions permit level is for your site? I presume the design DLN emissions level is around 9 ppm (corrected) for NOx and somewhere between 5 and 15 ppm for CO (and that would be at or near Base Load). There can be so many permutations of various levels and periods it can be dizzying.

And, I guess that begs the question--are you in violation of your permit during these times?

If you're not in violation of the permit is the concern that there is some undue wear on the combustion hardware which is occurring when the emissions levels are high? If this is the case, or is at least of some concern, I think you would have to ask the packager of the turbine and/or the extender hardware.

 

Thankyou Sir for the guidence. Point wise observations are mentioned below.

> I would again suggest that DLN tuning is likely necessary, at a
> minimum, and may be necessary again when the ambient dips to its
> lowest in a few months.

We have done the calibration before restarting the machine 5 months before .(IBH/ IGV/ GCV & SRV)

> Another thing I might suggest is that if IBH is in service
> ......IGV LVDTs weren't calibrated properly and/or the positioner
> on the IBH control valve wasn't working properly.

IGV single coil calibration was done including the physical verification with industrial protractor, as mentioned before IBH was also calibrated.

> And, again,......I presume the design DLN emissions level is around
> 9 ppm (corrected) for NOx and somewhere between 5 and 15 ppm for CO
> (and that would be at or near Base Load). There can be so many
> permutations of various levels and periods it can be dizzying.

Emission parameters you had stated is correct. Regarding the turndown or operating range, we were operating and tested the machine with new 32K hardware and environmental parameters were excellent for about 4 months from load 16 MW up to 33 MW.

> And, I guess that begs the question--are you in violation of your
> permit during these times?

It is matter of concern we are in the border line

> If you're not in violation of the permit is the concern that there
> is some undue wear on the combustion hardware which is occurring
> when the emissions levels are high? If this is the case, or is at
> least of some concern, I think you would have to ask the packager
> of the turbine and/or the extender hardware.

We had initiated the communication with GE.

I welcome any additional suggestions.

 

aksunil,

DLN tuning and LVDT calibration are related, but not the same thing. DLN tuning has been found to drift, particularly with ambient temperature swings. I can't explain why it didn't do so before the extender hardware was installed, and would suspect that "improvements" to the combustion components included with the extender hardware are likely at least partially to blame.

LVDTs don't drift very much, if at all. The LVDTS used on GE-design heavy duty gas turbines are the same ones used on many aircraft and rockets--and the reason they are used (as opposed to many other types of position sensors) is precisely that they are not prone to drift. (Older analog circuits used with LVDTs were prone to drift, but not newer digital circuits; and, again, the output from the LVDTs is proven over decades to be very stable in many applications--not just GE-design heavy duty gas turbines).

I have seen a lot of problems with IBH control valve positioner calibration and feedback over the years. Because they don't use the same regulator outputs as servo-operated devices, and because they don't usually get regular maintenance and calibration checks they have been found to drift. It's something many gas turbine power plant Instrument & Control technicians are not usually very familiar with (the I/P positioners used on the IBH control valves), as well.

Again, I'm at a loss to explain why the unit worked so well in the same environmental (ambient) conditions for so many years, but doesn't after the installation of the extender hardware. Were any other machine upgrades done when the extender hardware was installed?

So, it seems the emissions are marginal but still acceptable in your region of the world at the elevated levels you cited. But, you should be concerned about the change in emissions--and probably GE is the best entity to answer the question, after they ask for lots and lots of data. DLN-I technology is often described by the OEM as "mature" when in point of fact it is continually evolving (they are continuing to make tweaks to all aspects of hardware and even software over the years). And, those tweaks can sometimes prove to need even more tweaking when they are implemented in the field. And, since the OEM doesn't operate machines with this new hardware to fully test it before selling and installing it, there can be--and have been--issues. And they can take quite some time to resolve.

Finally, if I understand your comment about "single coil LVDT calibration" correctly I don't understand how reliable calibration and verification could be done using only a single servo-valve coil--without mucking with the null bias current value. The only reason to check servo operation with a single coil is to ensure the polarity of the servo current being applied is correct since the servo-valves used on GE-design heavy duty gas turbines are bipolar DC devices. Calibration and/or verification using a single coil would be problematic at best, and near impossible at worst--again without setting null bias current values to extremely unusual values. And, then they would have to changed back after all three coils were again re-connected. So, there must be something here I'm not understanding, and since its probably not a part of this issue I suggest we not take this discussion any further. (The only reason I'm mentioning this is because many other people read these posts, and some might misunderstand what was being said, and try to calibrate LVDTs using only a single servo coil--and that would be disastrous. There is too much tribal knowledge and mythology surrounding LVDT calibration as it is; I don't want to add to the mystique and misinformation.)

It would be most helpful if you would remember to update us on the progress of this problem resolution--a lot of people will benefit from the information you will provide!

 

aksunil,

One more thing came to mind--entrained liquids in the gas fuel. Not likely, but stranger things have happened. If they are hydrocarbon-based (compressor lubricating oil; gasoline; diesel; etc.) or condensate from the gas temperature being too close to the dew point this could be at least some of the issue. Again, it's not very likely that it would only show up when the ambient temperature is high, but it's worth checking out.

 

We had studied the CO emissions based on the on the discussions we had. It is observed that IBH valve is first opening, and after about 2 minutes the exhaust temperature slightly comes down (flame cooling), and CO emission is increasing. Based on this preliminary observations we are doubting the IBH valve positioner. We have on line trend in DCS. I can share the trend record in control.com, but I am not finding a way for attachment in control.com. Being a process plant we are looking for an opportunity to address the issue. Thank you for the support.

Will update once the job is done along with results.