Hey Guys
Could anyone help me finding the standard document on how to perform a load rejection test in a 25 MW gas turbine driving a pump in oil industry?

thanks

[email protected]

 

Hey Guy,

What is the purpose of the load rejection test <b>at your site</b>?

What is it that you're trying to demonstrate or prove by rejecting load from the turbine <b>at your site</b>?

Is it a turbine test <b>at your site</b>?

Or a compressor/compressor auxiliary/compressor control/compressor control valve test <b>at your site</b>?

When you define what it is that you're trying to prove or demonstrate <b>at your site</b>, then you can develop your site-specific load rejection test. Because all load rejection tests are not equal. It all depends on what it is that's being tested when load is lost (rejected for testing purposes).

For example, on a load rejection are you trying to prove that the LP turbine and compressor will not overspeed? Or that the valving associated with the compressor will function as necessary (go into bypass, or block reverse flow, etc.)?

Or that the turbine HP shaft does not overspeed?

What?

What is it that you're trying to prove or demonstrate <b>at your site</b>?

Again, when you define what it is you're trying to prove <b>at your site</b>, then you can device the test.

Unfortunately, most of the time I hear this "requirement" the reason given for performing a load rejection test is, "It's always done." Or my personal favorite (NOT), "It proves the control system." And when asked to provide details about what is being proven, there are none. Just, "It has to be done. It's always done. It will be done."

Someone mentioned to someone else that a load rejection test is always done on a turbine (gas or steam or hydro or wind or whatever). That is absolutely untrue--that a load rejection test is always done. It is done when there is a specific function that needs to be proven or demonstrated, such as avoiding an overspeed trip, or valve sequencing, or some other definite test. A load rejection test (or any test) is done to prove that some function works or that some sequence works or that some undesired effect does not occur. Tests, even a spelling test, are meant to demonstrate something, to prove something.

But, just to perform a load rejection test because "it's always done" without criteria for success (or failure) is just extremely poor project management.

Define what it is that is being proven or demonstrated when load is lost <b>at your site</b>, and you can define the criteria for the load rejection test <b>at your site</b>.

It's as simple as that. Full stop.