Base load temp limit calculations

R

Thread Starter

Rob

Can anyone shed some light on base load limit temp calculations for Frame 6B gas turbines running with mark 4 speedtronic control systems? We have two identical machines both running with HRSGs and both running with TX control on. At the same loading (31MW) with IGVs at 84deg, one machine has reached the base load limit while the other still has 20deg C to go. The machine with some to go has an exhaust temp median (TTXM1) roughly 5 degrees higher than the one which has reached base load. The CPD for the machine with some to go is 800KPa and the one which has reached base load is up around 970KPa.

Going off previous experience, a lower CPD normally indicates a dirty compressr or fouling inlet filters and results in a decrease in base load, but this is not the case this time.

Doing some research I found the base temp exhaust limit calculation from CPD is as follows:

A=(CPD/1380)-0.5434
B=0.5298-(0.5298-(A/0.1613))
IF B<0 then C = 0.5298
IF B>0 then C = 0.5298-(A*0.1613)
Base load temp limit = (1138*C)-18

This seems to be following pretty closely to what is actually happening.

My question is, how and at what stage does the exhaust temperature bias start effecting the limit and are there any other factors which can come into it?
 
B

Bob Johnston

From our MS6000 control spec (Sorry about the non-metric units), the Isothermal Ref is 1100 Deg F, the CPD Ref. Corner is 106.55 psi and the Tx decrease above this CPD is 1.92160 Deg F/psi. If you work this out for your two units, the 800KPa unit should come on temp. control at 582 Deg C and the 970 KPa unit at 557 Deg C. You seem to have a very big difference in CPD, I don't think that this is only compressor fouling unless it is very bad, maybe you have a leak or bad bleed valve or the CPD transmitter needs to be calibrated. Are both units on Temp Control and what is the Exhaust Temp on both and the Mw output at Temp Control?
 
Thanks for the reply Bob Johnston

Both units use IGV exhaust temp control to 555degC. The following data is taken at the same time and conditions for both machines

Load 31.6 31.7
IGV 84 84
TTXM1 560 564
CPD 963 780
Base temp 556 581

The CPD transmitter was checked and found to be OK. It is just about impossible for us to check bleed valves on the run but I gather a leaking bleed valve would cause a drop in machine capability which we are not seeing. There are no obvious external leaks on the machine, do you think some type of internal leak could cause this??
 
One thing you haven't told us is: Did this problem just start happening, rather suddenly, or has it been happening for some time? If it happened suddenly, what changed at that time?

It's not uncommon for two identical units sitting side-by-side to have different power outputs, especially if they have different maintenance outage schedules. When and what were the last maintenance outages performed on each of the units? When hot gas part parts were replaced during the last outages, were new or refurbished parts installed? What manufacturer were the parts purchased from? What are the fired hours for each machine?

You say, a dirty compressor is not the case this time. When was the last off-line compressor water wash done on each unit? When you do off-line compressor water washes, do you close the isolation valve to the CPD transmitter?

What is the inlet filter differential pressure of each unit? Do you have an exhaust duct back-pressure gauge on the unit--especially if the unit exhaust into a heat recovery steam generator (HRSG) or boiler?

Do the units have Water- or Steam Injection for NOx control?

You say, "The CPD transmitter was checked..." The CPD transmitters of both units were checked, or the CPD transmitter of one of the units? If only one unit, which one? A CPD transmitter problem could be on either unit. A CPD transmitter can indicate a higher-than-actual CPD pressure and cause output to be higher than it should be, which could eventually lead to hot gas path parts issues if left unchecked.

A leaking compressor bleed valve can usually be determined by using an infrared heat detector on the piping both upstream and downstream of the valve. The temperature of the piping should be much less than the compressor discharge temperature, usually signal name CTDA, when the valve is closed. If the temperature is at or near comp discharge temp, that indicates a leaking compressor bleed valve.

One could also put a T/C on the piping to measure temperature. Many Frame 6Bs have the compressor bleed valves outside the turbine compartment on the roof, so the piping temperature should be much less even than turbine compartment temperature if not leaking. If the piping and valves are outside the turbine compartment, you could put a welding glove on and touch the piping to see of one or the other is hotter (they should be fairly cool and about the same temperature).
 
I'm still having some problem understanding this problem--you say both units have "TX control" on. Do you mean both units have Base Load selected and Active--meaning that both units have the LED lit next to the Base Load selection softswitch? Does the display indicate 'Base Load' for both machines, or 'Base Load' for one and 'Part Load' for the other?

This may mean that one of the units is on secondary, or backup, exhaust temp control--not primary (CPD) exhaust temp control. Most older Mk IVs used FSR as the backup exhaust temp control. What are the FSRs for both units if they both are indicating Base Load?

Primary FSR is usually signal name TTRXP; backup FSR is usually signal name TTRXS. The lower of the two values is usually written to TTRX. What are the values of TTRXP, TTRXS, and TTRXB, and TTRX for both units?

If you have a GE Control Specification drawing, usually the one called 'Control Specification-Control System Settings' has some good information on the exhaust temp control curves in Section 3. You can usually find the formula there, if you can't derive it from looking at the Mk IV Speedtronic elementary's representation of the algorithm for exhaust temp control.

There is usually an eight-value array of Control Constants, several of them, for eight different possible operating conditions (most of the eight weren't used for Frame 6Bs; only a couple for Base Load, and a couple for Peak Load if so equipped, for each fuel if the unit was a dual fuel unit). In Mark IV and early Mark Vs, the array is usually named TTKn_x, where n is the array number (0-7, for a total of eight), and x is the single character designation for corners, and slopes and offsets and biases, etc.

There are some logic values which tell you which array set is being used, usually signal name L83JTn, where n=1 through 8, and corresponds to the array set to be used. Later Frame 6Bs might use something like L83BPn. But you should be able to find the proper algorithm around Sh. 28B for exhaust temp control and work backwards from there.
 
Top