Gas turbine in parallel operation

[nimesh9]

I want to operate two frame VI machines in parallel. consider rating of 33 MW each with 4% droop.Our total plant load is 28 MW & not connected with grid.My requirement is, on tripping of one machine other should survive.Should I kept both machines at 14 MW.What can be operation philosophy?
I want back up calculation also.

 

[davidbeach]

I'm pretty sure it can be done, but proving that is far more than anybody is going to provide for free over the internet. You need an Electrical Engineer on the project who is familiar with that type of project.

 

[ScottyUK]

You might want to post a link to this question in Forum666.

A step load change of nearly 50% of the unit rating is certainly going to cause some disturbance to the engine, perhaps severe enough to initiate a compressor stall or surge. It will almost certainly exceed the OEM's recommended loading ramp rate and will likely have a detrimental effect on hot parts life. Your governor and IGV control loops will need to be tight, which means that the hardware needs to be well maintained and accurately calibrated plus the software tuning must be optimised. Don't assume the latter to be the case!

You should engage the engine OEM's assistance if possible. They will have the tools to dynamically model the engine-generator train with sufficient accuracy to predict the response.

 

[crshears]

Continuously operating two gas turbines at <50% of their capability would surely place both in a very poor place on their efficiency graphs...economically unsound unless your fuel is dirt cheap - and environmentally unfriendly in any case. I've always been told that gas turbines operate most efficiently when loaded between 80% and 100% of their rating; is there any reason you cannot use a greater number of lower-rated machines?

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[ScottyUK]

"is there any reason you cannot use a greater number of lower-rated machines?"

Apart from the capital cost of buying new machines and scrapping those he has? Probably not... ;-)

 

[GTTofAK]

I've done some analysis (masters project) and could never find an n-1 solution that was economical. Bottom line IMHO its simply better to use the grid in event of unit failure and pay the ratchet than to endure the extra cost of installing an n-1. You would be better off simply investing the money.

 

[davidbeach]

But, he states there is no grid.

 

[GTTofAK]

But, he states there is no grid.

I interpreted it as he does not want to connect, not that there is no grid. I have yet to find an economical mix that yields n-1 contingency especially when taking time value of money analysis into account. Its simply cheaper to use the grid in the event of failure and pay the cost than it is to pay the upfront cost for the contingency.

What makes engineering sense does not always make economic sense.