Combined Cycle Power Plant Start Up

Our Plant consists of Two GE Frame-9E GTs, Two Vertical HRSGs (Manufacturer: HBG, China) one Steam Turbine (HTC, China).
GT Rated Capacity: 115MW @ site conditions; 1013 mbarg, 32 degC, 85%RH
HRSG Rating: HP Steam -75.6 barg/505 degC/182.9 t/hr
LP Steam-6.3 barg/259.4 degC/36.8 t/hr
Steam Turbine Rating: 123 MW.

As per HRSG manufacturer guideline during cold start of the plant we usually open diverter damper step by step starting from 20% to 100%/ to control the HRSG HP drum metal temperature difference (between upper half & lower half) within the limit (<41 degC). It also takes 1~2 hours to start HRSG from cold condition, and during start up, GT load hold at 6~8MW or at FSNL when GT Exhaust Flue gas temperature is 300~310 degC.

With this type of start up method we have faced problem in our Diverter Damper. i.e. generated crack on diverter damper shaft, deformation of diverter damper casing, etc. So we called our HRSG manufacturer for investigating the problem. After investigation they have suggested to operate diverter damper full open or full close for any kind of operation. (At full open or full close position diverter damper cooling & sealing system is active & useful. At D/D intermediate position it is useless.)

Our plant need to start 7~12 times in cold condition due to off taker load demand.

Noted that during start up of HRSG, we maintained min drum level as per OEM recommendation. We tried to start HRSG in D/D full open condition, but the upper & lower metal temperature difference of HP drum could not maintain within limit. Now its a great problem for us to start the plant in cold condition.

My query is that how we can overcome the problem of cold start up? Anybody have any experience please share.

My experience--which continues to be correct--is that most sites with diverter dampers find they can't use them as designed. They develop all manner of problems, including with the actuators. They bind up, they stick, they won't open or close fully, and usually then just end up being welded in the position that allows full flow to the HRSG. (Yes; there are exceptions--but they are few and far between (that is: not many).)

GE offers something called "Temperature Matching" which has been somewhat successfully used for HRSG warm-up, but is primarily used for ST warm-up. You could talk to them to see if your control system has the option or it can be modified to help with the HRSG issue.

But, mostly, you are going to have to work with HRSG designer to find a way to eliminate the use of the diverter damper when starting the GT. I have heard of manufacturers providing new, improved superheater sections and some baffling (which does somewhat reduce efficiency, but improves HRSG life and protects HRSG materials/parts which can't be "upgraded."

The GT exhaust is what it is--about the only way to minimize exhaust temperature "easily" would be to disable Combined Cycle mode, also called IGV Exhaust Temperature Control, which most non-DLN combustor equipped GTs may have. (DLN combustor-equipped units almost must have IGV Exhaust Temperature Control for achieving and maintaining combustion stability; so if your unit has DLN-I combustors, this probably isn't an option.)

In the early days of combined cycle plants, many GE-design heavy duty gas turbines had switches to select SIMPLE CYCLE mode or COMBINED CYCLE mode (IGV Exhaust Temperature Control mode). COMBINED CYCLE mode utilizes the IGVs to maximize exhaust temperature at part load--which somewhat reduces the efficiency of the GT, BUT increases the steam production, improving the overall plant thermal efficiency when the GT is at part load. If your unit does not DLN-I combustors, this may be an option. The IGVs will not close until the GT gets to approximately 60-70% of rated load, and that could be helpful in your case (if the unit does not have DLN-I combustors).

If the unit does have DLN-I combustors, you might also talk to GE about a possible modification of the Primary Combustion Mode operation to allow the IGVs to remain open longer than normal for a short period of time which may be long enough to help with warming up the HRSG. But, I'm not thinking this will be a good option.

One thing I've seen done at a couple of plants with very good success is using roll-up door (just like on warehouses!) to seal up the inlet (downstream of the inlet air filters) to prevent air from flowing through the unit when it is on cooldown to help keep the HRSG warmer longer. It isn't difficult, and it does require wiring up a small 3-phase AC motor, and some limit switches. And there needs to be a motor starter and some controls (to prevent closing the door when the unit is running!)--but it's NOT complicated, nor expensive, in the scheme of things. And, it really helps "bottle up" the heat in the HRSG for a couple of days, sometimes longer depending on the ambient. Both installations were "home-grown" and the sites were very happy with the results.

The door closes when the unit goes on cooldown, and can remain closed until it is time to start the unit again. The natural draft caused by the heat of the HRSG and exhaust stack that would normally cool the turbine and HRSG is interrupted by the roll-up door. One thing one site learned is that there may be a pretty good vacuum on the roll-up door if they tried to open the door not too long after closing it, and they ended up needing a stronger motor--and on the first unit, a stronger door...!

Hope this helps! Please write back to let us know what you find!
Dear CSA,

During commissioning of our GE frame 9E dual fuel gas turbines with HSD GE T/A was commissioned the Gas Turbine with Fire Mode of Start in which the machine go upto 1200 rpm after ignition and unless & until select auto mode it was stayed at 1200 rpm.

My questions are- Do we can start gas turbine in Fire Mode and keep the machine at 1200 rpm for half hour or twenty minutes or more? Is there any impact on the machine for holding at 1200 rpm?Do we need to perform this on Natural Gas Fuel?

Waiting for your valuable suggestions.


1200 RPM out of 3000 RPM is 40% of rated speed. The electric motor is typically still providing torque up to 60% of rated speed--AND it is usually drawing 150-160% of rated current during starting and acceleration. This includes during FIRE mode (which just holds FSR at the Warm-up value instead of ramping it up to assist with accelerating the unit.

So, the issue is if you repeatedly and regularly operate the starting/cranking motor at 150-160% of rated current for an extra 20-30 minutes during every start the motor windings will be degraded faster than normal. Also, it's very possible that the starting/cranking motor protective relay (which uses RTDs in the motor stator windings to monitor temperature) may also result in high temperature alarms or even possibly trips.

My recommendation would be to try using FIRE mode during a START (you can do it with natural gas or distillate; doesn't make a difference), and closely monitor the starting/cranking motor by recording the current drawn by the motor during starting, the motor stator temperatures (via the embedded RTDs), and the speed of the turbine-generator shaft during the trial.

When you are finished with the trial, just select AUTO mode and the unit will begin to accelerate by ramping up the fuel. Or, at least it should; sometimes if the speed is too high when you "release" the clamp on FSR by selecting AUTO the Mark* will immediately reduce the fuel to FSRMIN and sometimes that can result in loss of flame in one or more combustors, and even a loss of flame trip. That can be rectified by tinkering with the FSRMIN Control Constants, but because of the way GE uses FSRMIN for start-up, shutdown and maintaining flame on a load rejection it can take a few attempts to get "tuned."

You could then look at the data you have, and may even possibly consider ducting cool air to the motor to help with cooling during starting. However, if you have a failed start and need to attempt a re-start the motor protective relay may block you for a longer period because of the high temperatures caused by the high currents drawn during FIRE operation.

If that's the only choice you have at this point, it's worth a try. But be aware of the consequences--short term and long term. It may take a little fine-tuning, also, which can be time-consuming and frustrating.

And before anyone freaks out about the current drawn by the electric starting/cranking motor during purging and acceleration (including during FIRE mode operation) the motor has Class F insulation, and it does not run continuously at the higher current draw. It draws the higher current when purging, and when acceleration (after flame is established). And when the motor is shut down, it is usually run unloaded (approximately 40% of rated current) for a short time as a means of cooling down the motor (instead of just stopping it without letting it cool at speed when the rotor fans would be circulating air).

Adding an extra 20-30 minutes at the higher current draw is definitely going to add to the heat build-up in the motor (stator and rotor). The long-term effects if this turns out to be a viable option just need to be considered in the operation and maintenance planning.

I think the torque converter may also be subject to some additional heating--because the oil for the torque converter is drawn from the L.O. tank/reservoir, which is not cooled. And, the oil discharge from the torque converter is returned to the L.O tank/reservoir un-cooled, also. (Only the oil that goes to the bearing header is cooled.) So, if the L.O. tank/reservoir temperature is elevated to begin with (such as after a trip, or from cooldown operation), extended operation at the higher current draws (which means higher torque production which is transmitted through the torque converter) will not cool the torque converter as much, and the extra heat added to the oil passing through the torque converter will just be "dumped" back in to the L.O. tank/reservoir without any cooling.

Because the cooling water passing through the coolers does so through a tube bundle there really is nothing adding to the cooling of the L.O. tank/reservoir, other than radiant heat transfer through the walls of the L.O. tank/reservoir.

You should add the L.O. tank/reservoir temperature to the data you gather during the FIRE trial. You can easily create a Trend Recorder file to gather the data during the START attempt. I think GE Belfort has included starting/cranking motor amperage input readings and possibly some stator winding RTD input readings in the Mark* turbine control system so you can add them to the Trend file, as well. If not, you need to have someone recording amperages and stator temps (probably two people, one to toggle through the readings on the motor protective relay and one to record the values) and times at least once per minute.

The more data you have, the better. You could add exhaust GT exhaust temperature to the Trend file, as well, to have a record of exhaust temperature during the FIRE trial, and if you let the Trend run until a short time after synchronization you will have more data about exhaust temperature during starting and acceleration, also. More data is more better.
Dear CSA

After long discussion with HRSG manufacturer they were recommended us to operate diverter damper in throttling condition(25%-70%-100%) less than one hour for HRSG warm up during start up at cold mode provided that warming the diverter damper at 370degC for about 15 minutes before operating the damper.So we need to increase the GT load 30~40 MW for warming the diverter damper and then again reduce the load to 10 MW for HRSG start up.A little change we did at our site increase the HRSG drum level for start up to avoid uneven heating of drum metal and now its working fine(slight increase in water consumption since we need to drain out water to maintain drum level <High Alarm level).

Thanks for your valuable support & time.


Thanks for the feedback, but this is from your original post to this thread:

>As per HRSG manufacturer guideline during cold start of the
>plant we usually open diverter damper step by step starting
>from 20% to 100%/ to control the HRSG HP drum metal
>temperature difference (between upper half & lower half)
>within the limit (<41 degC). It also takes 1~2 hours to
>start HRSG from cold condition, and during start up, GT load
>hold at 6~8MW or at FSNL when GT Exhaust Flue gas
>temperature is 300~310 degC.
>With this type of start up method we have faced problem in
>our Diverter Damper. i.e. generated crack on diverter damper
>shaft, deformation of diverter damper casing, etc.

I'm not quite clear exactly what has changed with the HRSG manufacturer's recommendations, and the real question is:

Is this going to solve the problems with diverter damper shaft cracking and deformation of diverter damper casing, etc.?

You were already stepping the diverter damper open, but if the shaft cracks and the diverter casing is deforming what is going to prevent that?

Sorry; I don't mean to increase any confusion. I don't understand what will change to prevent the cracking of the diverter damper shaft and deformation of the diverter damper casing, etc. I can envision that increasing the exhaust temperature will decrease the warm-up time, but how does that prevent the damage to the diverter damper components and increase the reliability of the diverter damper operation and diverter parts life?