If generator breaker opens at full load, (due to a disturbance on the grid, for example) will the GT typically avoid an over speed trip and instead reduce fuel fast enough to operate at full speed no load?

Specifically am interested in people's experience with an F7A operating on natural gas.

Have been discussing the scenario described above with colleagues. People have different opinions and experiences regarding the scenario when an F7A is at full load and the generator breaker opens, say due to a disturbance on the grid. Discussion has centered around whether the controls would reduce fuel fast enough to avoid an overspeed trip and allow the unit to continue operating at FSNL. And if experience indicates different responses with a Mark VI controller and control algorithms from other vendors?

Thought I would open the discussion to a larger and more experienced audience.

Thanks.

 

That depends on a lot of factors, and some tuning and testing. What you're referring to is commonly called 'load rejection', and when it occurs at Base (full) Load it's commonly called 'full load rejection'. The unit is going to overspeed (meaning, the speed is going to be above 100%) but the trick is to try to avoid an overspeed trip (above 110%).

When fuel is cut back very quickly to try to prevent overspeed, the other danger is that it's cut back so quickly and by so much that the flame might be extinguished and the unit would trip on loss of flame. So, Minimum FSR is typically set to try to avoid flame-out on load rejection to keep the unit running so that it can be quickly re-synchronized if the fault that opened the generator breaker has been cleared.

The length and size (volume) of fuel piping between the control valves and the turbine fuel nozzles has a lot to do with avoiding overspeeding as well. That's why GE wants to keep the fuel valves close to the turbine, so that there's not a large volume of fuel in the piping when load is suddenly lost and the fuel valves are cut back and the fuel that's already in the piping has to continue to flow into the turbine until the pressure drops as a result of the valves closing.

This is where the overspeeding comes in, the amount of speed above 100% but below 110%. That's a function of the the fuel trapped in the pipe when load is suddenly lost and the fuel valves are cut back that contributes to the overspeed that's hopefully less than 110%.

And F-class turbines have another set of "issues" altogether because of the way the Inlet Bleed Heat control valves operate. And, there's the issue of changing combustion modes (transferring flame to different nozzles) very quickly as fuel is cut back and the IGVs and IBH are changing positions. And, F-class turbines can have different DLN combustors and different compressors and different IGVs.

I think GE is the best entity to answer this question for your particular unit(s) because we don't know what kind of combustion system you have on your turbine, how old the unit is, etc. I believe they have, for some combustion systems, developed sequencing to "stage" combustion mode changes during load rejection events while trying to prevent an overspeed trip <b>and</b> trying to prevent losing flame.

It's all a very delicate balancing act, and usually, the only way to be sure that it works is to test it. And let me tell you, the Pucker Factor is <b>MIGHTY high</b> when you open that breaker with the unit at or near Base Load! And if the unit flames out, then some Control Constants usually have to be adjusted and another full load rejection has to be performed. Usually, most grid operators and plant supervision personnel don't have the stomach for this testing, at least not more than once anyway. Losing 150 MW of generation in a split second is tough for some grid operators at some times of the day or year. So, a lot of planning and explanation and buy-in has to be done before one can test this ability.

And don't forget, the exciter has to cut back very quickly, too. There's just a lot of things that all have to work correctly for a unit to not trip on overspeed or loss of flame or exciter problem or ??? on a full load rejection.

And testing is not for the faint of heart!

Hope this helps!

 

The gas turbine control is designed to be able to handle any load rejection. It should not go to overspeed on a breaker trip. It is usually tested during commissioning for ability to withstand a load rejection from base load (they would usually manually trip the breaker rather than induce an electrical fault, but the effect is the same). On DLN machines, the bigger risk is flameout rather than overspeed.

 

I don't think a trip will cause the unit to go to FSNL. It will trip the fuel valves and require a restart. You can restart at some point, not sure where, but you will be well below FSNL. On a steam turbine a trip will close all valves and due to the speed difference between a breaker opening and valves closing I have seen some minor overspeed but as was stated above the systems are designed to not overspeed. Good luck.

 

Mike Borelli, you didn't read the post. The originator was talking about a load rejection, which is an event which causes the generator breaker to open but does not cause a trip of the turbine. The originator wasn't referring to a turbine trip, just a generator breaker trip that leaves the turbine running. Not all units have these types of "events" but many do.

A lot of people think that a full load rejection or a load rejection test is necessary when commissioning any turbine and generator. They say, "It proves the turbine control system." And when you ask them, "What does it prove? What constitutes a successful test, or a failure?" they have no response, other than, "It's always done, and it proves the turbine control system."

For steam turbines, it's very necessary since a steam turbine can overspeed <b>very</b> quickly if the generator breaker opens and steam isn't quickly shut off and/or diverted. <b>VERY</b> quickly.

Gas turbines, on the other hand, can still overspeed, but, they have this huge "brake" on the other end of the shaft, called an axial compressor, which is a large load. In fact, when a gas turbine is operating at rated power output as much as two of every three horsepower produced by the gas turbine is required to drive the axial compressor. And even when it's running at Full Speed-No load, it still takes a lot of power just to drive the compressor (and spin the generator rotor).

So, gas turbines, while they can reach overspeed trip speeds on load rejections, have the compressor acting as something of a brake to slow the rate of acceleration. Steam turbines have nothing at all to slow them down when the generator breaker opens; they just start accelerating very quickly and can reach overspeed trip speeds very quickly if the valves don't operate correctly.

So, steam turbines should be tested for load rejection. And re-synchronizing a steam turbine very quickly is not a possibility for many steam turbines.

Gas turbines, on the other hand, really only require testing if there's some contractual requirement to be able to resynchronize the generator in some very short period of time on a breaker open event. Not all sites have this requirement, so load rejection testing isn't necessary for these sites and isn't always done. (Except for those Customers, usually in a certain region of the world between Europe and Asia, that insist that a load rejection test be performed without any idea why or what constitutes success or failure. These same people also frequently want to perform "load throw-on" tests with the unit operating independently of other generators and in Droop Speed Control mode, and they get quite upset when the generator frequency drops so far that the under-frequency relay trips the generator breaker and the grid goes black. The turbine doesn't usually trip, but the generator breaker does. And they don't understand why, and declare the load throw-on test a failure!)

Now, some bean-counters, er, ... uh, ... Plant Managers are eyeing the dollars they can be paid if they can re-close the breaker very quickly after a breaker open event. And those bean counters, er, ... uh, ... Plant Managers need to understand that testing is necessary to prove this capability and it usually requires some amount of tuning, which means multiple tests.

And, as was said before, in this example losing 150 MW at one whack is not usually very well tolerated by some grid operators, even with advance warning.

And, not only F-class turbines, but any gas turbine also has the opposite condition, loss of flame, to consider. Cutting the fuel back very quickly to avoid an overspeed trip can result in loss of flame, and F-class turbines with DLN combustors in which fuel is split and transferred between fuel nozzles have their own problems, not to mention the IBH control valve and the IGVs.

So, it's best to talk to "the GE" to get some idea of the capability of your unit if you're not sure. And if anyone wants to get paid for this capability, be prepared to do some testing and tuning. And repeated testing. Because it's likely going to be necessary, one to prove the capability, and two, to ensure it not only does not result in an overspeed trip but that it doesn't result in a loss of flame trip.

And, that something else doesn't trip the unit (like excessive gas fuel supply pressure) or exciter overvoltage or excessive volt-per-Hertz.

There's a lot of things that can cause the turbine to trip, not just excess fuel or insufficient fuel!

 

I might add our own experience with this test, and illustrates some of the points CSA and otisad made.

Ours is a MS9001E unit, dual fuel, with the added twist of the gas being Low BTU. The unit starts up on liquid fuel (HSD), and then changes over to gas fuel when a certain minimum load is achieved on the turbine. This is because the unit cannot accelerate on our low BTU(low heating value) gas during startup.

So what happens if the unit is on baseload, and then the breaker opens. Ideally, like otised mentions, it should not overspeed. There control logic/control constants built into the speedtronic just to ensure that this doesnt happen.

But ours is a peculiar case. A smooth fuel chnagover take about 2 minutes to accomplish. If our breaker opens, and since the GT cannot run on Gas fuel at any load less than 30MW, this means that within a second or two, the turbine has to

A- Cut back on gas fuel
B- Changeover to HSD
C- Cut back on HSD
D- bring it to FSNL

And this is NOT going to happen in seconds! So there is no Baseload rejection on our gas turbine when on GAS fuel. Its just not possible. The unit WILL trip each time. We tried once during commissioning anyway

Then we tried doing baseload rejection when on liquid fuel. This is a more manageable situation because we didn't have to deal with the change over whilst we were rejecting base load. We tried this a couple of times (we had enough heart for a few tries), and on the last try it was successful. Although the inside scoop on this one was that someone tweaked (reduced) the Gain on the HSD servo regulator from Toolbox just for the test. Its an example of the tuning CSA has talked about. But whats interesting was that the gain was returned to its old value after the test Test successful, everyone goes home happy, but lets return that gain to what it was so the normal behavior of VS-1 stays unaffected.

The only way to be sure if your unit has baseload rejection capability for sure is to try one Or perhaps you can look in commissioning reports, perhaps the test was attempted then? Ask your plant's planning /document controller to fish those out for you.

 

Yes this helps. The background of the various processes taking place is a big help.

Thanks!

 

CSA, Thanks for the detail. I will have to review the operating procedures again to see how I missed this. It could be the result of many 'bad' trips.

 

I am glad that Mike gave the steam turbine's perspective even though the original question was about GT.

I was a witness to 3 FSNL tests of 3 x GE 7EA units. The commissioning contract required the tests because the units were being commissioned on a small island grid where re-synchronising quickly was a priority.

The GE commissioning engineers were totally against the tests but after some haggling with the owners they eventually agreed to do the tests. All three tests "passed" on their first attempt. "passed" - meaning the units went to FSNL after opening the 52G with the GT at full load. They did one unit at a time and the grid frequency was run a bit higher than 60Hz to cater for the test.

I hope my experience will help here.

 

Dear experts,

How about the excitation CB? Is this CB will tripped once GCB tripped during load rejection event?

If yes, is there any specific reason to trip excitation CB? my understanding is that if load rejection occurred, GCB (52G) will open, and AVR will regulate the voltage terminal to rated voltage to avoid overvoltage and governor control will prevent the turbine from overspeed.

 

mwadym,

Your understanding is correct for a load rejection event.