We have 2 Frame 9E Gas turbines. One of the gas turbines is not reaching base load on liquid fuel. Both FSR1 and FSL are going to 100% position. We have changed the LFBV, fuel pump relief valve, HP filters and skid filters. Liquid fuel pressure before VS1 (stop valve) is normal 3.0 bar. Can anyone suggest how to solve the problem?

 

Any ideas what your Fuel Flows are when you are at 100% FSR and what is your HP fuel pressure. If you have changed your LFBV already that should be OK, it obviously looks like you are not getting enough fuel. Any recent maintenance, recal work on the fuel system? What %age load are you getting at 100% FSR?

 

First, is this something that just started on units which have been running without problems for some time, or is this a problem discovered on new units during or shortly after commissioning? If the units have been running for some time, what has changed recently? Has a maintenance outage just been completed on either unit? Do the units usually run on liquid fuel, or just occasionally?

Second, if you're certain the relief valve isn't passing, and you're certain the LFBV is closing, then something is either restricting the fuel flow, or the clutch is slipping and not allowing the full torque to be transmitted to the fuel pump, or the internal clearances of the pump are not to specification, or there's some "internal leakage" in the pump.

Third, what is the FSR1 of the other unit when it's running on Liquid Fuel at Base Load? If the other unit is very close to 100% FSR1 and the LFBV is very nearly closed, then it might just be that you've hit the limit of the Liq. Fuel System for the fuel that's being burned.

Fourth, have you checked the Liq. Fuel Forwarding Pump suction strainers (presuming you have separate Liq. Fuel Fwd. Skids, one for each unit)?

Fifth, have you observed the position of the LFBV when it's at 100%? Is the LFBV fully closed?

Fifth, does the unit have Low-pressure Liquid Fuel Filters upstream of the High-pressure Liquid Fuel Pump? If so, have you checked/replaced the elements? Water can swell the paper filter elements and restrict the flow through the filters; they might not appear to be filled with dirt, but they could be "blocked" by swelling caused by water in the fuel.

Sixth, have you checked to make sure the Liquid Fuel Pump Clutch is not slipping? One way to do so is to remove the coupling guard between the clutch and the pump and use a bright strobe tach to observe the rotation. Slipping will usually be characterized by "jumping", but if the shafts are rotating steadily you should be able to determine what the clutch output shaft speed is from the nameplate of the Accessory Gear, and use the tach to observe and measure the shaft speed to determine if the clutch is continually slipping or not.

Seventh, have you checked the clearances inside the High-pressure Liquid Fuel Pump?

Have you changed Liquid Fuel suppliers or fuel spec's?

Do both turbines draw from the same Liq. Fuel Storage Tank? If not, are you certain the fuel in the two tanks is the same?

 

Thank you very much for the response.
Unit1 is giving about 7 MW less than unit2.Both units are drawing Naphtha from same tank.
At 100% FSR, LFBV is full close. We have replaced the skid filters and HP filters.
Warren pump discharge pressure is about 37 bar almost same as other unit.

We have not checked the pump clearances.Infact we are facing the problem since commissioning.
Unit2 FSR at base load is about 70% and 1to 2 % margin is there on LFBV.

The speed of pump in both units measured to be same.

thanks and regards

 

You haven't told us if the units operate on any other fuel (natural gas) or distillate, and if they both have relatively similar outputs when operating on the other fuel(s).

Also, we don't know anything about when the two units were first commissioned, and what control system each unit has. In fact, we're presuming that the two units have exactly the same exhaust temperature control curves.

I don't understand what you mean when you say that the second unit has 1-2% "margin" on LFBV. What does that mean? That the LFBV is at 98-99% stroke (nearly closed)?

We're also presuming the fuel nozzles are the same in both machines.

What is the CPD of each unit when Base Load is selected and both units are trying to make rated output? What is the exhaust temperature of each unit at this point? What are the exhaust temperature spreads of each unit at this point?

If the units have Mark IV, or Mark V, or Mark VIs, can you tell us what the values of FQR (the liquid fuel flow-rate reference; it might also be FQROUT or something similar), and FQL or FQL1 (the value of the Liquid Fuel Flow Divider feedback signal)? Because the Speedtronic is trying to make the flow divider feedback equal to the flow reference, so it's crucial for us to know what these values are for each unit.

We're also presuming the LFBVs are identical.

We need much more information.

 

Both units are giving same output on natural gas.

Control system is Mark V for both units.

Liquid fuel data of both units as follows:
T2 T1
Load MW 113.4 107.9
Freq Hz 49.10 49.39
ctim DegC 24 24
TTXM DegC 566 561
FSR1 % 77.8 100
FQR % 76.3 98.78
CPD Bar 10.7 10.28
FQLV1 % 76.73 71.98
LFBV % 96.85 100.02

Regards,
ankaro

 

This is good information.

[It's also interesting because operation at 49.10/49.39 Hz demonstrates just how bad the grid frequency control is in some parts of the world. That's approximately 98.2%/98.78% of rated frequency, and I'll bet this information was taken from each unit at slightly different times, meaning that the frequency is fluctuating quite a lot (because two units operating in parallel with each other and on a grid with other units should not have different frequencies at the same instant in time). But, we digress.]

I still believe there is some problem with the liquid fuel system not being able to deliver the same amount of fuel for each unit. Because the LFBV is "closed" on Unit T1 (as indicated by the LVDT feedback; 100% should be fully closed on the bypass valve) and yet the indicated fuel flow, FQLV1, is lower than T2's. Presuming the Liq. Fuel Flow Divider feedback scaling is the same on both units and the scaling constants for the calculation of FQLV1 are the same on both units, T1 is not getting the same amount of fuel as T2.

Most units have some type of fuel meter on the Liq. Fuel Forwarding Skid. If your unit has these meters, have you compared the fuel consumption at Base Load for both units to see if roughly the same difference (percentage) exists as indicated by FQLV1? This would be kind of a 'double-check' on the accuracy of the Liq. Fuel Flow Divider feedback and the FQLV1 calculation.

Now, we need to know the following information for both units:

TTKn_I
TTKn_C
TTKn_S
TTKn_K
TTKn_M

where "n" is 0 through 7 (defining an array). We want the actual running values from the Control Constants display, not the Control Specification. We need the values for each of the arrays (if any of the values are 0s, we don't need to know them).

TTRXP
TTRXS
TTXM
TTXSP1
TTXSP2

The above values should be recorded for both units while running with Base Load selected. TTRXP is the primary exhaust temperature control
reference, and TTRXS is the secondary exhaust temperature control reference. I'm wondering if one or both of the units are operating on secondary "back-up" exhaust temperature control instead of primary (CPD-biased) exhaust temperature control. The TTXSPn values are the
magnitudes of the highest and second highest exhaust temperature spreads.

Also, please check the scaling of the Liq. Fuel Flow Divider feedback for both units in the I/O Configurator; it should be on TCQA Screen 17/21, and there should be one of the Pulse Rate Inputs defined as 'fuel'; report the entire line for that input, please. And give us the information for *both* units.

There's another "limiting" factor which is probably coming into play, but it's caused by some other problem. FSR is usually limited to the
value FSKMAX, which is usually set to 100%. We need to find out why FSR is being driven to 100% in one unit, and not in the other unit. I'm trying to help eliminate the possible control system problems, but do not discount the fuel system. As happens so often, the control system gets all of the focus and attention (because it's not usually very well understood or understandable), when most problems can be traced to I/O problems, or, in some case, I/O scaling values. The Mark V is only as good as its inputs and outputs, and usually (if configured properly) has far fewer problems than the devices which provide inputs or serve as outputs, including interconnecting wiring.

One thing we haven't asked about: Diagnostic Alarms. What Diagnostic Alarms are being annunciated by the unit which is "load limited"?
(Please provide *ALL* Diag. Alarms, even if you think they might not be related to the problem; let us be the judge of relevance.)

Also, what Process Alarms are being annunciated by the unit which is "load limited"?

 

Thanks for your reply.

Constants information

TTK0_c TTK0_s TTK0_k TTK0_I TTK0_M
Unit-1 8.57 13.8 50.99 593 2.22
Unit-2 8.67 13.78 50.99 593 2.22

TTK1_c TTK1_s TTK1_k TTK1_I TTK1_M
Unit-1 9.71 13.31 61.98 615 2.07
Unit-2 9.83 13.29 61.98 615 2.07

TTK2_c TTK2_s TTK2_k TTK2_I TTK2_M
Unit-1 8.72 13.8 51.89 593 2.22
Unit-2 8.82 13.78 51.89 593 2.22

TTK3_c TTK3_s TTK3_k TTK3_I TTK3_M
Unit-1 9.86 13.31 62.86 615 2.07
Unit-2 9.97 13.29 62.86 615 2.07

rest of the constants are zeros.
FSKmax is 100%.
FQLV1 is matching qith skid flow meter.
There are no process alarms during load limt.Diag alarms, I did not notiec.

regards
ankarao

 

And the running values of primary- and secondary exhaust temperature control (TTRXP and TTRXS) are????

It is noted that the two units do NOT have similar exhaust temperature control curves (from the Control Constant Values), which was kind of presumed (at least on my part).

When was each unit first installed (in other words, how old are the two units or how many years between the installations of each unit)? When were the control systems installed (were the units originally installed with Mark Vs, or were they upgraded to Mark Vs)?

What other differences between the two units are there?

Please check the Diag. Alarms and let us know exactly what they are.

 

In our case, we are having two Frame 6 GT. Recently we have commissioned inlet air chilling facility. Now since last one year, our GT base load have been reduced by approx 4 MW.

The inlet air diff pressure is hardly 2" WC, so what should be the reason for drop in base load? Also the benefits achieved in terms of heat rate improvement with inlet air chilling are much less than expected. What should be the reason?

 

As far as I know, inlet air chilling is not used to improve heat rate but to improve power output when ambient temperatures are on the high side with respect to GT design. In fact inlet air chilling lowers slightly heat rate due to the increased auxiliary loads associated with the system.

You did not indicate whether:

1. Your loss of 4MW is with inlet air chilling active or not.
2. The GT control system has been reprogrammed with new parameters which are limiting base load.
3. What type of inlet air chilling system has been adopted, i.e. fogging, compression chilling or absorption chilling.
4. The inlet air cooler has been fitted as per GT/Inlet air filters' OEM instructions or not.
5. The 4MW lost were from day 1 of starting up the inlet chilling system, or after some time.
6. The inlet air differential pressure includes all components of inlet air system, namely filters, heat exchanger, etc.

 

Jojo,

I think there is a heat rate improvement when inlet air cooling is used and there is some older thread which makes some kind of reference to a white paper on the fact, perhaps something on the Mee Fogger website.

Haven't we had a similar question recently (http://www.control.com/thread/1026246054#1026246088, read the entire thread for all the details, because just like this question, it was tacked on to another question and there wasn't much response to the requests for more data)? And wasn't there some other work that was done at roughly the same time (HRSG duct burners, something like that...)?

What does the inlet chiller supplier say?

What is the compressor inlet temperature (usually signal name CTIM, or something similar, on a Speedtronic turbine control panel)?

Where are you measuring the inlet filter differential? Just across the inlet air filters? Or across the entire filter/chiller assembly, including any trash screens?

Did you take any static inlet pressure data before the chiller installation? Have you taken any since the installation?

What is the compressor discharge pressure at Base Load before and after the installation?

What is the median exhaust temperature before and after the installation?

What are the exhaust temperature spreads?

What are the fuel flow-rates at Base Load before and after the installation?

Was there any other work done at the time the chiller was installed (combustion inspection, hot gas path inspection, major inspection)?

Was there any other work done on the unit, such as modifications to the inlet filter house or inlet duct work, or exhaust duct work, or something similar?

It would certainly seem that a chiller vendor would be very interested in having their product perform properly and would be intimately involved in troubleshooting any problems. This might be a case of someone designing their own inlet chiller, but we don't know yet. There's just a lot we don't know, and are interested to find out.