Frame 6 GT automatic loading rate Mark V


Thread Starter


Dear Markvguy,

There are total 7 frame six GE Gas turbines are connected in our grid and we were done a load through off test in one GT (Load 30MW). While doing this test other GTs behavior is as shown below.

14:43:26.875 18.8 10.92 50.12 10.88 50.09 5175
14:43:26.906 19.9 10.92 50.12 10.88 50.05 5167
14:43:26.937 19.9 10.92 50.12 10.88 50.05 5167
14:43:26.968 22.0 10.92 50.04 10.88 50.04 5167
14:43:27.000 22.0 10.92 50.04 10.88 50.04 5167
14:43:27.031 23.9 10.92 50.04 10.88 49.98 5160
14:43:27.062 23.9 10.92 50.04 10.88 49.98 5160
14:43:27.093 24.8 10.91 49.97 10.88 49.97 5160

To increase load from 18.8 MW to 24.8 MW GT took 0.218 sec.

1) My question is how we can calculate loading rate of the GT while doing load through off test.

2) Can we use the following formula for calculating Droop?

Droop = FSRMAX(100%)-FSKRN1 / FSKRN2
= 100-18.3/18.266

GT set points are:

TNKR1_0 -Startup Accel Rate <9 %/min>
TNKR1_1 - Speed SetpointRaise-LowerRates<22%/min>
TNKR1_2 - Manual Loading <4 %/min>
TNKR1_3 - Manual Speed/Fast Load Rate <6 %/min>
TNKR1_4 - Auto Load Rate <0.75 %/min>
TNKR1_5 - Startup Accel Rate <8 %/min>
TNKR1_6 - Startup Accel <15 %/min>
TNKR1_7 - Startup Accel Rate <18 %/min>
TNKR1_8 - Speed Matching Rate <6 %/min>
TNKR1_9 - Speed/Load Setpoint Rate #9 <0 %/min>
TNKR1_10 - Speed/Load Setpoint Rate <0 %/min>
TNKR1_11 - Speed/Load Setpoint Rate <0 %/min>
TNKR2-- Breaker Opens Setpoint <100.3 %>
TNKR3 - Setpoint High Speed Stop <107 %>
TNKR4 -- Setpoint Low Speed Stop <95 %>
TNKR5 -- Setpoint Start/Stop Low Speed Stop <0 %>
TNKR7 -- Setpoint Startup Preset <100.2 %>
FSKRN1 -- Full Speed No Load FSR Constant 18.3 %
FSKRN2 -- FSR Droop Correction Factor 18.266 %/%

Thanks in advance,

You have already asked this question, and it's not clear what you are trying to demonstrate or prove with the Load rejection ("throw-off") test, or how all the remaining units are being controlled. Are you expecting the units to maintain frequency AND load? What EXACTLY are you expecting the load rejection test to tell you???

When units are being operated in Droop Speed Control mode and experience a frequency decrease (such as would happen on the loss of a generator due to a load rejection test or breaker open event or turbine-generator trip), they are NOT loaded by means of the normal load raise logic (L70R). They are loaded by the increase in the error between the actual turbine speed and the turbine speed reference. The faster the error increases, the faster the load will increase.

It is presumed your units are conventional combustor-equipped and NOT DLN-I units; DLN units CANNOT be loaded very fast and are not good for frequency control for that reason. This discussion presumes conventional combustors.

It appears from the data that the units are all being operated in Droop Speed Control mode, with no other power islanding control scheme. This is being said because the frequency is dropping as the units pick up the load.

0.218 sec is fairly fast--do you want it to be faster? If the total load is not changing when one unit is being "tripped" off line, and all the other units are in Droop Speed Control mode, then the frequency will NOT increase back to normal.

For example, if the total load BEFORE one unit was removed from the local grid was 210 MW and the total load AFTER one unit was removed from the grid was 210 MW, then the remaining units accepted the load which was "thrown off" from the unit which was removed from the grid.

If the remaining six units are all being operated in Droop Speed Control--and Preselected Load is NOT enabled on any of the remaining units--the frequency will decreases, and it is the frequency decrease (because of the increased error between the actual speed and the turbine speed reference) which will cause the units to increase power output. If the frequency returned to normal, the error would decrease to zero and the frequency would decrease.

In your case, since the grid did not trip on underfrequency, the load rejection test was successful from that perspective, as well.

But it's just NOT clear what you are expecting the units to do during this test. If all the units are being operated in Droop Speed Control mode--and Preselected Load is NOT enabled on any of the remaining units--the frequency will decrease.

If one unit was lightly loaded before the load rejection was performed and was switched into Isochronous Mode when the unit was tripped off the grid, it would pick up the load VERY quickly AND maintain the frequency--up to Exhaust Temperature Control limit. It might even pick up the load in less than 0.218 seconds WHILE maintaing frequency.

To determine what the droop setting is for your units it would be necessary to look at the sequencing (CSP, Control Sequence Program since it appears at least one of the units is a Mk V from the data you have provided) to determine what type of Droop Speed control mode the units have--"normal" Droop or Constant Settable Droop or ??? The formula you provided is not familiar for any type of Droop this author is familiar with.

In any case, to have any idea of how to respond to this query, it is necessary to understand EXACTLY what you are trying to demonstrate, prove, or measure. It is also necessary to understand how each unit is being operated (Droop Speed Control mode, Pre-Selected Load, External Load Setpoint, etc.). This forum is just not really the place where it is easy to present all the information and the data and to understand what the intent of the testing is.

Dear Markvguy
Thank you for your reply
One new frame 6 GT was added to our grid recently, Load rejection test was carried out as a part of GT commissioning. has very good search facility and I have got good idea about Droop control from previous threads, from your answers.

Thanks & Best Regards
This is EXACTLY the kind of response expected to the question of what is trying to be proven or demonstrated with the "load throw-off" test: "We ALWAYS perform a load rejection test when we commission a new unit." So, the reason a load rejection test is necessary is: Because we've always done it. No reason; no proof of any feature or demonstration of any ability; just, that one has always been done in the past.

How can one require a test when one doesn't have any criteria for success or failure? "Let's have a subjective test--just do the test and we'll tell you afterwards if it was a success or a failure!"

But, this author's FAVORITE reason for performing a load rejection test is, when really pressed for a reason or for a set of criteria for success or failure of the "required" test:

"It proves the control system."

Period. No requirement for taking data to "prove the control system." No pass or fail criteria for "proving the control system." Nothing. Nada. Zilch. Zippo. Just: "to prove the control system."

It's a potentially destructive test if the fuel valves don't reduce fuel fast enough and the unit overspeeds. But--don't worry about that--we need to "prove the control system!" Reducing fuel extremely quickly can also cause combustion liners to collapse--but, hey, the unit's under warranty! The packager will pay for any damage!

When commissioning a heavy-duty gas turbine, the ONLY reason to perform a load rejection test is to prove that the control system will not reduce fuel so quickly--or by so much--to avoid an overspeed condition that the flame is lost--in other words, the purpose is to demonstrate the unit will not flame out when the breaker opens from a loaded condition. This is a requirement for some contracts--that on a breaker open event (not an emergency unit shutdown--just a breaker open event), that the unit will remain at FSNL (Full Speed-No Load) so that it can quickly be re-synchronized, and one won't have to wait for the unit to coast down to a speed at which it can be re-started and then wait for it to purge, and fire, and accelerate back to FSNL so it can be resynchronized. That's the ONLY reason to perform a load rejection test on a heavy-duty gas turbine.

So, it's pretty obvious that there was no criteria for performing the load rejection test and that the results weren't understood (since the original post was asking how to determine the loading rate for the remaining gas turbines when 0.218 seconds didn't seem fast enough or too fast--still no response to that question, either!). Fishing for reasons to pass or fail a test which was "required" but not understood.

Now, when commissioning a steam turbine, especially a large steam turbine, it's VERY important to perform a load rejection test. Steam turbines have no axial compressor to drive--which in effect, is a "brake" on the turbine--and it is CRITICAL to prove the unit will not overspeed. (Of course this is an even more potentially destructive test, but it's warranted in the case of a steam turbine.) There are "trip anticipation" circuits in a large steam turbine control system and there is SO much energy in a small amount of steam and there is no "braking" action of any kind on a steam turbine, that a load rejection test is very important to "prove the control system."

But not so on a heavy-duty gas turbine....