GE Frame 9E Increasing Exhaust Spread after Fuel Change Over

Gas turbine GE Frame 9E undergone fuel change over from Gas to Hsd occured due to Gas field tripping and GT MW went form 120 to 150MW. GT MW fluctuating for 2min before get stable.

After 24 hours of event, GT exhaust spread is on increasing trend and has increased from Avg 33 to Avg 36 in one day. As per ACT swirl software, 2, 3 and 4 can are showing.

Kindly suggest what can be the reason behind increasing Exhaust spread keeping in view HSD leakage event and GT MW reaching 150MW peak.
 
Use the selector valve handle on the discharge of the flow divider to record the first 14 positions (two people are best, one to turn the handle, the second to write down the pressures). Any pressure which is 10% higher or 10% lower (or more!) means there is some problem with either the liquid fuel check valve at the fuel nozzle, OR the liquid fuel purge check valve at the fuel nozzle. If the swirl chart is good, then the pressures for 2, 3 & 4 will also be higher or lower than the average of all the other pressures. In any case, the problem will be found in the can(s) with the higher- or lower pressure(s).

If the liquid fuel purge check valve is leaking, you will see liquid fuel flowing out of the False Start Drain (look at the Liquid Fuel/Liquid Fuel Purge and AA P&IDs; the drain is usually on the left side of the turbine compartment, under the walkway, and usually flows into an open scupper). If the liquid fuel purge check valves are working properly, there should be NO flow out of the False Start Drain into the scupper.

If the liquid fuel purge check valves are good, then either the liquid fuel check valve at the fuel nozzle has failed, or there is a problem with the internals of the liquid fuel portion of the fuel nozzle (loosening; blockage; etc.). A higher than normal pressure measured at the discharge of the flow divider means there is some blockage in the flow of liquid fuel to that fuel nozzle/combustor. A lower than normal pressure measured at the flow divider means the spring in the liquid fuel check valve has failed, or the liquid fuel purge check valve at the fuel nozzle has failed, and the fuel is not being atomized properly or the fuel is leaking out through the liquid fuel purge check valve.

Spreads are caused by uneven combustion in one or more combustors. That results most often when fuel flows to the combustors are uneven or the liquid fuel isn't being atomized properly. The whole function of the liquid fuel flow divider is to take to flow from the Liquid Fuel Bypass Valve/High-pressure Liquid Fuel Pump and divide it evenly into 14 individual streams to the fourteen fuel nozzles of the fourteen combustors. When something happens to cause unequal flows (which is <b>VERY</b> evident by looking at the pressures on the selector valve/gauge assembly on the discharge of the liquid fuel flow divider) then spreads will be higher than normal. The fact that the spreads weren't a problem BEFORE the transfer to liquid fuel and are increasing AFTER the transfer to liquid fuel says quite conclusively that one or more combustors is either getting too much liquid fuel, or too little liquid fuel, or is not being atomized properly (which is a function of the BOTH the liquid fuel check valve AND the atomizing air flow--and it's HIGHLY unlikely, though not impossible, that the AA flow to one or more combustors is uneven).

Please write back to let us know what you find. This is NOT a control system problem, and a leak at the high pressure liquid fuel pump--unless it a SIGNIFICANT leak--isn't going to cause a high spread. And, if it's a SIGNIFICANT leak, then the turbine isn't going to maintain load, and the Accessory Compartment is going to be a VERY dangerous place to be or to be near.
 
<b>CORRECTION</b>
>If the liquid fuel purge check valve is leaking, you will
>see liquid fuel flowing out of the False Start Drain (look
>at the Liquid Fuel/Liquid Fuel Purge and AA P&IDs; the drain
>is usually on the left side of the turbine compartment,
>under the walkway, and usually flows into an open scupper).
>If the liquid fuel purge check valves are working properly,
>there should be NO flow out of the False Start Drain into
>the scupper.

It&#8217;s the <b>Tell-Tale Leakoff</b>, not the False Start Drain. Still located in the same area, and both should be connected to the Gas Turbine Drains Tank.

Apologies for any confusion.
 
Thank you for the quick response CSA

We have carried out the inspection of GT combustion can 2, 3 and 4 and we have found a damaged male Cross fire tube (Collar damage) between 2 and 3.

So cans identification is Ok. Also after startup, Spread 1 reduced from 37 to 27.

Any checklist you can provide to us for detail inspection will be very fruitful.
 
Moreover, we are experiencing leakage from atomizing booster air compressor suction NRV despite being SOVs is closed.

Will it contribute in abnormal atomizing of HSD?
 
Thank you for the feedback. The leaking will not help with atomization of the liquid fuel, but unless it&#8217;s a significant leak it shouldn&#8217;™t affect the liquid fuel operation at rated speed.

It should, however, be repaired.

It&#8217;™s interesting that the spread was not worse on gas fuel prior to the trip, and it&#8217;s not clear if the reduction in speed was observed while running on liquid fuel or gas fuel. And it&#8217;s also telling that the spread is still high (in my personal opinion) even after the cross-fire tube problem was found and resolved.

How often is the unit operated on liquid fuel, or is liquid fuel just a back-up in the event of loss of gas fuel? GE recommends switching to liquid fuel at least once per week to exercise all the liquid fuel components and keep the liquid fuel supply lines free of air, which is probably the reason the load spiked so high during the transfer.
 
Yet another<b>CORRECTION</b>....

>It's) interesting that the spread was not worse on gas
>fuel prior to the trip, and it's not clear if the
>reduction in <b>spread</b> was observed while running on liquid fuel
>or gas fuel.

Apologies, again, for any confusion.
 
Incomplete combustion will be most evident as whitish or white smoke coming out of the stack. Usually, when burning liquid fuel (distillate or HSD (High-Speed Diesel)) the exhaust is light tan or light brown (depending on the amount of water- or steam injection, and I'm referring to coventional combustors primarily, though it's similar with DLN combustors). The more incomplete the combustion, the whiter the haze coming out of the stack. So, usually during starting there can be some white smoke as flame flickers and combustion cans lose flame during starting and acceleration, but it usually clears up before reaching FSNL (Full Speed-No Load).

A lot of people mistakenly believe that incomplete combustion causes dark smoke from the exhaust, but that's not true. It's not true of lorries (diesel trucks) and it's not true of combustion turbines. Dark black smoke when burning liquid fuel is indicate of high lead--and is VERY bad for combustion (gas) turbines. White smoke is not good either, and will usually be accompanied by high exhaust temperature spreads.

<i>Based on the information provided,</i> it does not seem like the high exhaust temperature spread was entirely related to the cross-fire tube issue that was found and resolved. If the cross-fire tube was entirely responsible for the high exhaust temperature spread the magnitude of the spread would surely have come down more than a few degrees Celsius.

And, it's still not known what the magnitude of the exhaust temperature spread was while running on natural gas....
 
OK.

I will gather more data and will share the details of GT performance.

Moreover, is it possible that the air ingress in the diesel system causes the HSD leakage from Diesel pump and similar elevated MW?
 
Air in the liquid fuel supply piping is likely the cause of the large MW swings you saw during the transfer, because of large swings in the liquid fuel supply pressure. And, once the air got out of the lines--and the liquid fuel supply pressure stabilized--then the load stabilized.

An instantaneous jump in MW isn't going to hurt the turbine or the generator (it's not good for it, but a short duration (of milliseconds) isn't going to hurt it either). And, neither will the short duration drops in power output that occurred as the air in the liquid fuel supply piping was being passed through the system. What goes up must come down, and what goes down must go up. And, the liquid fuel bypass (control) valve is sensing the air going through the liquid fuel flow divider and is trying (in vain) to respond and maintain a stable fuel flow-rate. But until the air gets out of the piping and into the combustors and the liquid fuel supply pressure stabilizes, the load is going to swing, sometimes very violently.

You're probably very lucky the unit didn't trip on exhaust overtemperature--but the fact that the liquid fuel flow-rate was swinging wildly (violently) because of the air in piping and the LFBV trying to respond to the air in the piping/flow divider, it probably didn't exceed the overtemperature trip setpoint.

Again, short durations of load spikes, and load drops, aren't going to hurt the turbine or the generator. And the control system should be protecting against long duration spikes by tripping on exhaust overtemperature. Generators can produce MUCH more power than they are rated for--but not for long periods of time. As the load increases, and the amperes flowing in the stator increase, the heat increases. And it's the ability to cool the generator that sets the limit on <b>continuous</b> operation. But, when the fuel flow-rate (liquid fuel or gas fuel) is unstable then the load produced by the generator will also be unstable, but for short periods of time which won't cause excessive heat to build up in the generator.

You haven't answered the question: How often does the gas turbine operate on liquid fuel? When was the last time the unit operated on liquid fuel before this most gas field trip occurred? Because, periods or even a week or two can cause air to get into the liquid fuel supply piping. AND, GE-design Frame 9E machines have been known to have liquid fuel check valves leak in the reverse flow direction and blow combustion gases back into the liquid fuel lines up to the flow divider. If the leaks are bad enough, they have even been known to lift the liquid fuel stop valve. That's why the OEM/packager of these units recommends--VERY strongly--that the liquid fuel system be exercised at LEAST once per week for a couple of hours, to keep air out of the system and keep it "primed" for when it's critical to have a smooth transfer from gas- to liquid fuel. They even sell VERY EXPENSIVE liquid fuel recirculation systems to keep air from getting into the liquid fuel piping if you don't want to do weekly fuel transfers.

YES. Air in the system can cause load swings when switching fuels while on-line. VERY LARGE load swings (both positive and negative).

As for air in the liquid fuel supply piping causing a leak at the high-pressure liquid fuel pump, well, that's more likely caused by dry seals because the liquid fuel system hasn't been run for a long time. The details of the leak are non-existent--where did the leak come from? How long did it leak? How bad was the leak? Did it stop "by itself"?

The load swings were caused by the air in the liquid fuel piping. Where the air came from is anybody's guess--but it DOES get into the piping if the piping isn't kept liquid-tight. Air is insidious, and if the liquid fuel system hasn't been run in a couple of weeks, or months, then air is going to make its way into the piping. From several possible points. Including leaking liquid fuel check valves.
 
Top