Hi!

I want to understand how IGV settings for 9FA Gas Turbine are decided:

a. What is the designed value for Open IGV position (maximum angle up to which IGV can open) i.e. CSKGVMAX for 9 FA GT. In several posts it is mentioned as 84 degree. Control setting in GE's manual mentions it as 84 degree however now contractor want to change it before Contract Performance Guarantee Test?

b. What will be impact of increasing maximum IGV opening beyond 84 degree? Will it affect life of the gas turbine? Is it advisable to allow setting it beyond 84 degree for contract performance guarantee test?

c. GEK 110899 say that correction curves are dependent on IGV angle. Will increase in maximum IGV open angle result in increase in GT output and increased exhaust mass flow? If yes, will it lead to change in correction curve?

d. For Contract Performance Guarantee Test, should CSKGVMAX be kept at 84 degree as correction curves for PG Test had been submitted at the time of award or should GE be allowed to submit revised correction curve based on change in maximum open IGV angle?

Since, this is my first post in the forum, kindly excuse me if I have posted my query in wrong forum.

In case additional information is required, please let me know.

Thanks

 

You haven't provided enough information.

When a turbine is new and being commissioned, the manufacturer/packager have the right to determine during commissioning what final parameters will be used for control and protection of the unit. And, this only makes sense--they own the warranty and they decide how the turbine operates (control and protection) for that period (warranty period) at least. If they "over-fire" the unit, or push too much air through the axial compressor there is a likelihood of parts failure, premature and/or catastrophic.

Tuning during commissioning usually includes any performance guarantees. F-class gas turbines are very highly technically advanced machines. Unlike other GE-design heavy duty gas turbines they are literally running on the very edge of the limits of the materials used to build them with very little margin particularly when it comes to performance. And, one of the most limiting factors of almost any gas turbine is the ability of the axial compressor to push more air through the machine--and that's certainly true of F-class machines. There are so many designs of IGVs and rotating and stationary compressor blading that it's virtually to keep track of how each one is configured to operate. Compound that with the fact that so many axial compressors have been replaced and/or upgraded over the years, and, well, it's very confusing.

And when big machines like this experience failures when running, it's just ugly--and expensive. And potentially deadly. All of which add up to lots of money, time and lost production. Most OEMs/packagers just don't want to have to deal with these issues just to get a few hundred kW more to make a performance guarantee.

And, of course, when it comes to any written documentation about just about any GE-design heavy duty gas turbine, well, they just don't produce very good documentation and a lot of what is around is dated and incomplete. And, even the turbine-specific documentation (the Control Specification document and the P&IDs) don't always get updated to reflect as-built and as-commissioned conditions. So, what do you trust?

We don't know enough about the turbine you are referring to. Is it new? Refurbished? Upgraded?

About the only thing you can trust is the written Purchase Contract documents--<b>and</b> that the OEM/packager has their own best interests at heart when it comes to ensuring that the unit does not require warranty service or repair or replacement--because that comes out of their pocket (if they can't prove the unit was improperly operated!). Manuals? Not to be trusted. Remember--most manuals are produced long before the turbines. And things can--and do--change, and in today's world they change very quickly.

And, have you read the Performance Guarantee Uncertainties section of the Contract, yet? (I'm referring to new units, now.) That's sure to cause a huge disagreement, and many a legal battle. But, someone signed that Contract--whether they understood it or not. There's a good deal of wiggle room in that document.

Again, if it's a new unit--it's the OEM's/packager's machine until a letter of acceptance is granted by the purchaser. And, they have a right--really an obligation--to tune it to run optimally (which includes performance and parts life). And, while it's under warranty (and if properly operated!) the OEM/packager is obligated to repair or replace defective items--so they're not likely to try to squeeze a few hundred kW out of the machine at the expense of parts life (and the service required to replace them). No. Most turbines can be "tuned" to make performance--but that sometimes would come at the expense of parts life, and possibly catastrophic failure.

I've personally been on jobs where the units barely made performance and wouldn't have without the Uncertainties clause--but that's how the contract was written and signed, by both parties. Even if the purchaser didn't know exactly what was being signed (though it should have been thoroughly reviewed and vetted by technically knowledgeable and savvy people before being signed).

I've also been on jobs where the unit didn't make performance, and in those cases "negotiations" occur--usually to the purchaser's great favor.

But, I've never been on a job where the OEM/packager tuned the unit control and protection parameters to make performance at the expense of parts life and potential catastrophic failure. Never. Ever. They took their lumps and learned not to do that again. (They lose some, too--just not too many.) They wouldn't remain in business too long if they did that on any kind of regular basis.

Now, I do know of other OEMs/packagers that have tweaked things during commissioning--but lived to regret it, and very, very dearly. (Not all of them are still in business, either.)

This probably isn't what you wanted, but, this is my experience. And, as a controls engineer I've been privy to a lot of performance testing, data-gathering, analysis and tuning. Again, my experience is that this particular OEM/packager isn't likely to tweak things at the expense of a catastrophic failure and all the costs that go with that.

And the bad public relations.

 

Also, most OEMs/packagers don't have machines they can run for testing and data-gathering. Instead, they take massive amounts of data during commissioning ("performance testing") which they then feed into large, powerful computers to predict performance of present and future machines (it's called "empirical data" because it's gathered over time, and from machines that are being commissioned).

They use this data to determine how to get more and better performance and parts life, and so sometimes they do a little "experimentation" during "performance testing" to get data they otherwise can't get (from a complete running unit). And, they use this data when building and configuring and controlling and protecting new--and existing--machines.

You see, testing isn't always perfect. New methods and new instrumentation are always being developed. And, uncertainties and errors in existing or past test methods and test instrumentation are always being found. So, the opportunity to get running data from a new and clean machine during "performance testing" is something they will always want--and need--to do to keep improving their product.

So, their testing protocol and procedures might not just be about "making the performance guarantee" or squeezing a few hundred kW out of a machine to meet contract requirements.

A lot of this data can even be used to sell upgrades to the Purchaser a few years down the road--if enough margin can be detected.

And, lastly, I've seen the OEM/packager exceed the Contract guarantee--by a lot (not on an F-class machine, of course), and just have to give that away. But, you can bet they used the data to increase the output guarantee of future machines--and charge more money for the increased output!

It cuts both ways, my friend!

 

Hi CSA

Thanks for your response. While I would tend to agree with your view regarding OEM not doing anything which will harm equipment during warranty period however, some of my queries still remain. I am putting them below. Regards machine, its new machine which is being commissioned and contract performance guarantee needs to be proved.

My queries are:

a. As per documentation provided by GE (GEK 110899), performance correction curve are dependent on IGV Angle among other parameters. It implies that any correction curve is for a particular IGV Angle and if IGV Angle changes, correction curve should undergo change. Please let me know if my understanding is correct.

b. If my understanding is correct then when IGV angle is changed as part of IDLN Tuning, performance correction curve submitted earlier as part of contract will be required to undergo change. Should OEM be allowed to submit revised correction curve?

c. If revised correction curve is allowed to be provided will it not lead to a situation where OEM knows the performance of the machine and also gap between actual and guaranteed performance curve and submits a performance correction curve which will help him to bridge the gap.

Shall appreciate if you can help in understanding and resolving above issue.

Regards
Shree Narayan

 

narayanshree,

> a. As per documentation provided by GE (GEK 110899), performance correction curve are dependent on IGV Angle
> among other parameters. It implies that any correction curve is for a particular IGV Angle and if IGV Angle changes,
> correction curve should undergo change. Please let me know if my understanding is correct.

Base Load is defined to occur at a specific IGV angle and only at that angle. Performance Guarantees are generally for Base Load and Performance Testing is done at Base Load. So, if the OEM/packager changes the Base Load IGV angle then a new Performance Correction Curve should be issued. But Correction Curves are typically only for trying to correct for a difference in ambient temperature at the time the unit is running versus the rated ambient temperature of the machine on the machine nameplate. So, if the nameplate ambient was 20 deg C. but the temperature on the day a Performance Test was being done was 31 deg c the Correction Curve is used to correct actual performance back to 20 deg C to determine if the actual machine performance is near rated.

> b. If my understanding is correct then when IGV angle is changed as part of IDLN Tuning, performance correction curve
> submitted earlier as part of contract will be required to undergo change. Should OEM be allowed to submit revised correction curve.

Performance Testing and DLN tuning are mutually exclusive. And, IGV angles aren't usually changed during DLN tuning--only fuel splits.

> c. If revised correction curve is allowed to be provided will it not lead to a situation where OEM knows the
> performance of the machine and also gap between actual and guaranteed performance curve and submits a performance
> correction curve which will help him to bridge the gap.

Machine performance is what it is; I can't think of a situation where tweaking the Correction Curve will allow the OEM/packager to make a certain output (performance). You don't seem to understand what the purpose of a Correction Curve is.

A Performance Guarantee is made at a particular IGV angle and at a particular ambient temperature and at a particular ambient humidity and at a particular ambient pressure and at a particular exhaust duct back pressure (the latter is usually implied but may also be stated in the Contract). Since all these factors affect actual performance on since they can vary on any given day--<b>EXCEPT IGV angle</b> since by definition Base Load occurs at a specific IGV angle--then there must be a way to determine if the output (performance) on any day is comparable to the output on an ISO day (at the stated conditions of temperature and pressure and humidity). It's virtually impossible to test on an ISO day when ambient conditions are exactly equal to guarantee conditions. So, Performance Correction Curves allow one to "compare" performance on a 32 deg C day to that of a 20 deg C day. That's what Correction Curves allow--to correct for a difference in actual operating condition(s) back to ISO guarantee conditions. So, it's hard to see how a Correction Curve can be be tweaked to make a machine meet performance.

And since the only entity qualified to develop the performance curve--<b>per industry standards</b>, <i>which, by the way is <b>very important</b> to understand</i> one has to accept the Performance Correction Curves provided by the OEM/packager.

Performance Correction Curves are developed using industry standards, not just some arbitrary numbers. if the IGV angle is changed as part of commissioning the way the Performance Correction Curves are derived <b>does NOT</b> change. I venture to say that if you compared a Correction Curve at 84 DGA to one at 85 DGA the slope of the curve would be exactly the same, only the stated IGV angle would change. Everything else would be the same, except for the IGV angle.

Hope this helps! DLN tuning is not (typically) done by changing IGV angles, and by definition Base Load is at a particular IGV angle. DLN tuning is done by changing fuel splits--the amount of fuel that goes to each set of nozzles, not by changing IGV angles.

 

Dear CSA

Thank a lot for comprehensive and detailed response. I would like you to consider the following and clarify my doubts:

a. In our case CC output is guaranteed at site condition which is defined in the contract. Guaranteed site condition includes ambient temperature, ambient pressure, relative humidity, cooling water inlet temperature, frequency, power factor and natural gas analysis (gas composition)

b. Following performance correction curves have been provided as part of contract based on which measured Combined Cycle performance shall be corrected: ambient temperature, ambient pressure, ambient relative humidity, degradation, Cooling Water temperature, power factor, frequency and Fuel LHV & composition.

c. You have confirmed two things in your response: (a) base load is at a particular IGV (b) correction curve is for a particular IGV

d. In view of above measured combined cycle performance at base load has been achieved at a particular IGV (say X).

e. Correction curve which has been submitted as part of contract is for a specific IGV. During IDLN Tuning, IGV Angle has been changed (say Y) and may not be same IGV at which correction curve was generated. Hence, can we apply correction factor using correction curve which has been developed using IGV Angle X when measured output is at IGV angle Y.

f. If correction curve needs to be changed for IGV Angle Y before applying correction factor to measured output so that correction factor and measured output are at IGV Y then we have following dilemma:

While I appreciate that OEM is the best person to develop correction curve, is he not in a position to take advantage of the fact that he knows what will be measured output (since he has just completed IDLN Tuning and has measured the CC output/heat rate at guaranteed site conditions) and he also knows that correction factor he needs to apply to measured output/heat rate to arrive at guaranteed output/heat rate? How can one be sure that revised correction curve submitted by him is not cooked-up to suit his interest. You will appreciate that site conditions will not vary much when IDLN Tuning is being done and when PG Test is being done since they shall be carried out within short interval. Id there any way to ensure that revised correction curve is proper and not to suit OEM's financial interest.

Once again thanks you a lot for clearing several of my doubts and showing patience to read my lengthy mail.

Beast regards
Shree Narayan

 

narayanshree,

In your first post you inquired about the <b>GT</b> maximum operating angle and Contract Performance Guarantee Testing. Now, you're talking about Combined Cycle Performance Guarantee. The two are not the same. The information I have provided refers exclusively to the <b>GT</b>--not the combined cycle.

1) Are you saying that the OEM/packager is guaranteeing <b>GT</b> output at any load less than Base Load? In other words, is the OEM/packager guaranteeing output at 100 MW? Or 125 MW?

In my experience, <b>GT</b> performance (power output; heat rate) is only guaranteed at one point: Base Load. And, again, by definition Base Load occurs at the maximum IGV operating angle defined for the unit.

And, all the <b>GT</b> Performance Correction curves I've seen are for Base Load operation.

Think about it like this. You want to compare <b>GT</b> performance (power output; heat rate) at some set of operating conditions to another set of operating conditions. That needs to be done with as many of the variables (IGV angle; ambient temperature; ambient pressure; ambient humidity; frequency; power factor) that can be stabilized as possible. And in the list, the one that's easiest to stabilize when any of the others are different than nameplate rating is: IGV angle. With the exception of power factor, all of the other variables can't be controlled.

You don't seem to be saying that the OEM/packager is providing a set of correction curves for anything other than a single IGV angle--so that implies, by definition, that <b>GT</b> performance is guaranteed at Base Load, which only occurs at one IGV angle.

IGV angle does change as load changes, but we're not talking about anything other than <b>GT</b> Base Load.

2) In your original post you didn't mention "<b><i>I</i></b>DLN" tuning. I've been away from new unit installations for a few years, and have not encountered "<b><i>I</i></b>DLN" tuning; can you please provide more information about this?

In my experience, DLN tuning involves operating at various loads in various combustion modes and measuring emissions output, and making adjustments to the fuel splits--<b>not</b> the IGV angles. DLN tuning takes place at stable operating conditions, and if one were to change the IGV angle at any load (Base Load; 96.7% load; 63.2% load; etc.) <b>and didn't change the fuel flow-rate</b> then the load will change.

So, our communication problem must be that I don't understand "<b><i>I</i></b>DLN" tuning and what it involves and what it does. But, I don't think that's the problem here. The problem is that the question keeps changing. First, you're asking about <b>GT</b> IGV angle versus Performance Guarantee. It's slowly changed to include CC Performance Test and "<b><i>I</i></b>DLN" tuning. I don't think the OEM/packager is going to change IGV angles during <b>GT</b> Performance Testing--it's going to remain at CSKGVMAX (whatever the OEM/packager determines that should be for the unit). And, I don't think the OEM/packager is going to guarantee <b>GT</b> performance at anything other than Base Load--which means IGV angle is going to be stable at one value, CSKGVMAX.

When tuning the <b>GT</b> emissions at Base Load, the IGV angle needs to remain at the defined maximum operating angle--again, that's part of the definition of Base Load. Then, the unit load is lowered to different loads, and the load and fuel flow are allowed to stabilize at each load point for a few minutes--and that includes the IGV angle, which if varying, would affect load and operating conditions, including emissions--and then emissions data is gathered and fuel splits are adjusted to make the actual emissions equal to or just slightly below the emissions guarantee value(s). I'm not aware of any <b>GT</b> whose performance is guaranteed at any load other than Base Load. Emissions--yes. Performance--no.

I understand that it seems possible that if the OEM/packager is in charge of <b>GT</b> performance testing AND for issuing the <b>GT</b> performance correction curves that they could tweak the curves in their favour. In my experience, that's just not the way this OEM operates. And, again, <b>GT</b> performance correction curves are calculated based on industry standards which most GT OEMs/packagers subscribe and adhere to.

I do know of Customers who have used third-party companies to perform <b>GT</b> performance testing, using <b>GT</b> performance correction curves provided by the GT OEM/packager. That's because everyone believes industry standards dictate how performance correction curves are created, and so they take the data at the machine operating conditions and the apply the manufacturer's corrections and that's the Performance.

Lastly, at the overwhelming majority of sites I've been involved in <b>GT</b> performance testing the <b>GT</b> performance has been in excess of the guarantee. The combined cycle performance, however, has usually been marginal--but not the <b>GT</b> performance.

So, if the <b>GT</b> performance guarantee is only at Base Load, then, by definition, Base Load occurs at the manufacturer-defined maximum IGV operating angle (CSKGVMAX for most digital Speedtronic turbine control panels). So, <b>I</b>DLN tuning--whatever it is--<b>at Part Load</b> isn't applicable.

You seem to be in the position that a lot of people find themselves in during the commissioning of a GT--someone signed a GT purchase contract without considering the performance testing procedure and now many people have taken the time to read that contract and are questioning the performance testing procedure. (I've even known Customers who actually signed the purchase contract question this procedure--<b>well AFTER</b> they signed the contract--and have sent minions to try to "correct" this error. But, it's too late, obviously.)

It's a tough position to be in. But, the best you can do is what you're doing--trying to understand the conditions and the procedure(s). But don't confuse things. And do try to understand how <b>GT</b> performance correction curves are developed--I don't know the answer to that; just that they are developed in accordance with accepted industry standards.

Many times sites try to hire third-party consultants to oversee and/or review the performance testing, but the <b>GT</b> performance is usually not in question--but the combined cycle plant performance is. Usually the two are done simultaneously (testing <b>GT</b> performance and combined cycle performance), but I can tell from painful experience the combined cycle plant designers--when their plant performance is in question--try pointing the finger of blame at the <b>GT</b> OEM/packager. And the GT OEM/packager wants to be certain the <b>GT</b> performance is not in question. And, the combined cycle plant designers usually beg the <b>GT</b> OEM/packager to change the <b>GT</b> output to help the combined cycle plant make guarantee.

I wish I could be more help--but this is the extent of my knowledge on <b>GT</b> performance testing and correction curves. I have seen revised <b>GT</b> performance correction curves when CSKGVMAX has been changed during commissioning, but the slope of the curves don't change--they just slide to the left or right, and very slightly.