Validation of gas turbine performance test measurements


Thread Starter



I am curious as to how people in the field validate the data that they collect (pressure, temp, etc) during a performance test of a gas turbine.

If there is anyone that has any experience of this or perhaps has some knowledge that they wouldn't mind sharing, it would be greatly appreciated.



Usually, highly accurate instruments are used for gas turbine performance tests. They are meticulously maintained, and the accuracy of their calibration is frequently checked (by periodic "calibrations"--during which any as-found errors are immediately reported to the previous user of the instrumentation).

Most gas turbine performance tests are conducted by people sent to site expressly for this one, sole purpose: to conduct gas turbine performance tests. They have experience with the machines, and the data being gathered.

If during the data-gathering any values appear to be unusual, then instrumentation is either replaced or the connections are checked and corrected if necessary. For example, if the gas turbine exhaust temperatures seem a little off, then it's very common for the thermocouples to be spot-checked after the test. If the exhaust temperature spreads are higher than they should be, then usually the test is stopped until the problem can be found and corrected. If the inlet pressures/differential pressures don't seem to match previous test results for similar unit, then the instrumentation is inspected and/or replaced as necessary

In general, should someone have reason to question an upcoming performance test, they usually contract with a knowledgeable third-party firm or -consultant to oversee the performance test, and to review the results and make comments or suggest corrections.

It's very difficult, after a performance test has been conducted and documented (the results presented in a report) to question the instrumentation used to gather the data, or the methods used to gather, analyze and report the data. Especially if some time has elapsed.

You could try hiring a third-party firm or -consultant to review the reported results of a performance test. Many performance tests include calibration data sheets (from calibration checks/corrections done before the gas turbine performance test).

I sense, from your prior post, some question regarding the exhaust heat flowing into a waste heat recovery steam generator ("boiler") being used to produce steam or provide heat for some process. (I hope I'm wrong, because, as previously said in another thread: it's very difficult to accurately measure.)

You are always free to order a new gas turbine performance test, at your expense, and conducted by a firm or consultant of your choosing. It is suggested that you engage the services of some firm or consultant that will help you in crafting the terms and configuration of the test to aid in determining not only if there is a performance issue, but what might be the possible cause(s) of the performance issue.

Finally, the most "aggravating" thing about gas turbine performance tests I have encountered over the years are the corrections applied to the data gathered during the performance test. Such corrections are necessary because many times the performance test is conducted when the ambient conditions (primarily ambient temperature and barometric pressure) are not at the machine's rating. There are other correction factors that can be applied for inlet duct pressure drop and exhaust duct back pressure, depending on the site and the configuration. It is these "corrections," which are clearly stated in the contract, that most people have frustration over when the gas turbine performance is only marginally above, or even slightly less than, rated. But, they are the best--and only--method of adjusting for ambient- and site conditions that are other than that used to determine the rating of the machine.

In short, many people agree to gas turbine performance tests without understanding how they are conducted and don't understand or question the corrections which will be applied to the results. When they have a problem and start reading the performance test, the contract and the corrections they start screaming--loudly. But, a contract is between two parties, at a minimum, and if one party signs without understanding every portion of the contract they have little recourse. Except to get informed, and order their own test with their own criteria and independent verification.

In the three cases I have seen independent gas turbine performance tests conducted after initial contractual performance tests were concluded, the results were better than the initial gas turbine performance test results. And, the corrections that had to be applied because of the number of hours of operation since initial commissioning the results were viewed with even more skepticism. Further, the lost revenue that occurred because of the shutdowns (plural) to install and remove the instrumentation were very hard to justify given the results of the second tests--and the cost of them.

I suggest you obtain the services of a knowledgeable firm or consultant to review the contract, the testing methods, and the data before you continue too much further. This is not a simple process to understand, review or conduct, and best left to a professional.
CSA has it exactly right.

The first document you should review when designing or reviewing a gas turbine performance test is ASME PTC-22 Gas Turbines.

This Performance Test Code includes guidelines on how to setup a test, as well as outline the measurements and uncertainty levels required for the most accurate test results.

The code is written by experts in the industry, and can be purchased from ASME at

If exhaust energy is the desired value, you'll need to do a heat balance around the gas turbine (and HRSG, if you can). This does require a lot of measurements... and a lot of data validation to get a good result.
in the acceptance tests we relay on our contract with the test conductor (the contractor or the OEM).

this conductor supplies a calibrated temporary instrumentation for the test (this instrumentation installed for testing and then removed, we also gather the control system measurements (permanent instrumentation) to make some comparison with precision instruments.

with those instruments a calibration report is supplied for each instrument, The calibration report provide a tabular comparison between the reading from the instrument and the reading from a calibration standard across a range of readings typical of the expected performance test conditions. you can review those reports to prove that the ASME PTC 22 requirements are met (these requirements are tabulated in this code, every parameter and its Maximum Allowable Measurement Uncertainty.
in one test, i see the performance engineer validate his reading by himself or by his experience - the result of the corrected heat rate indicate a wrong value of fuel flow. although the flowmeter is calibrated and has its calib sheet, he did a further investigation through energy balance (ISO 2314) and introduce a further correction factor for the heat rate.