we have a GE gas turbine of frame 9E. if we start the unit at distillate fuel the exhaust spread reaches allowable at fsnl.

The exhaust temperatures shows a cold spot in TC. from 11 to 17 with 250c as average at 10mw. The fqlm1 gives a value of 2.7 kg/s at 10mw which is a high value and a yellow smoke from the stack. the fuel command and feedback is maching.

The LFBV LVDT has been calibrated and magnetic speed pickups at flow divider also have been calibrated. atomizing air pressure at firing moment is 0.5bar. flow divider pressures are similar.

Is there are possibility of un-flamed combustion chambers in the respective cold spot area, and how to check that?

kindly note that starting at gas fuel is normal with reasonable exhaust spreads.

 

I'm going skip right past the LVDT calibration business and ask first, how does one calibrate a passive magnetic speed pickup?

Next, what makes you believe that the fuel flow-rate at 10 MW is "high"?

Last, <b>WHAT ARE THE FUEL PRESSURES OF EACH COMBUSTOR AS MEASURED USING THE MANUAL SELECTOR VALVE DOWNSTREAM OF THE FLOW DIVIDER WHEN THE UNIT IS AT FSNL AND WHEN IT'S AT 10 MW?</b>

 

Dear CSA

regarding mangnetic speed pickup they have been calibrated by I&c staff.

regarding high fuel consumption at 10 MW, it is already known the specific fuel consumption of the unit at an average rate of 0.3kg/kwh

regarding flow divider pressures at 10MW it was 16.0barg.

your quick response would be highly appreciated.

 

If your I&C department can calibrate passive magnetic speed pick-up, you need a new I&C department.

Okay; so how much fuel does it take to maintain Full Speed-No Load? Do you think a combustion turbine doesn't require any fuel just to get up to and maintain rated speed? So, if there's no kwh when the turbine is running and the generator breaker is open then there's no fuel flowing?

Lastly, I find it very difficult to believe (impossible to believe, actually) that all 14 combustors were at 16.0 barg with the exhaust temperature spread you described.

Please have someone knowledgeable come to site to assist with troubleshooting your problem.

This topic has been covered MANY times on control.com. Please use the search feature on this site to find tens of threads related to high exhaust temperature spread causes and troubleshooting. None of them will recommend re-calibrating LVDT feedback or "calibrating" speed pick-ups. Most of them will advise the original poster to review P&IDs to understand the systems and components involved.

Best of luck! Please write back to let us know what you found.

 

Passive speed picks calibration !!

"If your I&C department can calibrate passive magnetic speed pick-up, you need a new I&C department."

ha ha ha, hurray for tongue and cheek humour CSA. well i have heard this phrase many times. what they mean by calibration is "we have fixed the speed pick ups in the flow divider in the required position", 0.04 inches from the tooth i believe.

another aspect of field "calibration" is to check to see if the magnetic pickups and the upstream mass flow meter match. before startup of the machine during manual purging. the FQL which is in kg/s compared to the mass flow (t/hr) or volumetric flow (m3/hr) meter. this is a quick field check mostly done if the flow divider has undergone a replacement. mass flow meters are rugged and are usually though to give correct values.

both the above though not in a sense "calibration" by any definition. it is usually a slang used by i/C and operations to refer to any field instrument.

Lion's site might have meant something like that, but it us just me taking a guess.

coming back to the original question. i have seen and heard about this issue many times. in all duel fuel machines Gas seems to give a much better combustion. i do not know why though. asked may people and no one has given any satisfactory answer. this is especially true if you do not run in liquid for a long time. The practice to reduce the spread in distillate operation is to run the machine in full base load or open the IGV in part load if base load operation is not possible. after a few hours it is seen that the spread does come down. I do not know why this works. i cannot explain any relation between spread and excess air. But this operation works out most of the time. open the IGV to max at part load and let the machine settle down for some time (min half an hour). you will most probably see a reduction in spread. run it for two to three days and then you will see that the spread remains normal for most loads. give it a shot and write back.

(note - opening the IGV at part load is a inefficient operation. if you have a downstream HRSG there will be a slight reduction in steam output)

"LFBV LVDT", Lion usually LFBV has the magnetic speed pickup as the feedback only the GCV/SRV had the LVDT arrangement. are you sure your LFBV has a LVDT??

 

>....but it [is] just me taking a guess.

Guessing yet <b>again</b>, Process Value.

The original poster, like so many others with dual fuel machines, has not taken the time to understand how simple the gas fuel delivery system is, nor how complicated the liquid fuel system is,including the Atomizing Air-, Purge-, and False Start Drain Systems.

They are looking for simple answers to questions they haven't contemplated from a free Internet forum.

And, they either don't try to understand what their I&C department is telling them or the I&C department isn't explaining it properly, or some combination of the two.

Most of the originators don't understand GE-design heavy duty gas turbine operation and don't feel the need to understand it. The Speedtronic takes care of everything, and the I&C department is all-knowing and takes care of the rest. After all, they calibrated the LFBV LVDT feedback and that certainly causes high exhaust temperature spreads....

The originators of posts like this have never, and will never, take the time to study P&IDs, or read and follow the GE recommendations for reliable operation of dual fuel units. They just want a simple, one sentence answer to their problem after they've been told by the I&C department that everything that can be done has been done because the LVDTs have been calibrated, and the passive magnetic speed pick-ups have been "calibrated."

Worse, we never hear back from these people.

And, yes, there are many Frame 9E LFBVs with LVDTs as many were built with them before just a few years ago.

 

Dear all

regarding flow divider pressures -16.0barg- it is well known that the pressure is controlled by the actuating of LFBV and affected by the condition of liquid fuel line,nozzles, and related check valves. If there is not any variations of these factors; then the pressures would be similar

Regarding LFBV it is definitely has an LVDT.

The main question is how to ensure that all cans have a flame.

 

> The main question is how to ensure that all cans have a flame.

I'm going to defer to the I&C Department at your site, sir. They have you most well informed.

Best of luck with the problem.

 

Lion,

With a cold spread you need to look at the fact that one or more of your fuel nozzles are not allowing the same amount of oil flow through it compared to the other nozzles. This will also be seen at the Selector valve since the can not flowing the same as the others will have a higher pressure. That's the can you need to investigate. If the pressures are all the same and you have water injection you can also stop the H2O injection to see if the spread goes away. If that happens then look at the H2O injectors based on swirl patterns for the can it might be originating in.

 

> And, yes, there are many Frame 9E LFBVs
> with LVDTs as many were built with them
> before just a few years ago.

Just as a curious offshoot, the LFBV can have either a LVDT or a magnetic speed pickup. what is the necessity of both?? i have not seen such a design can you elaborate CSA??

 

It's not an "either/or" proposition. An LFBV will always have flow divider feedback, just as an SRV will always have P2 pressure feedback.

As has been said in other recent posts on control.com, LVDT feedback was thought to be necessary on LFBVs when GE went from Frame 5s to Frame 7s, where the expected fuel flow-rate increased nearly three-fold.

In recent years, GE has determined (through cost-reduction efforts) that the LVDT is not necessary for LFBV stability on Frame 7E/EAs and Frame 9Es. So, they were eliminated on Frame 7E/EAs and Frame 9Es in recent years.

It should be noted that when this change was made, other changes to regulator gain and servo part numbers were also made. So, it's not advisable to change regulator types without reviewing all other parameters.

 

SalemGeek,

Please note that the main reason for exhaust temperature spreads caused by liquid fuel system components is low pressures, not high pressures. A high pressure would indicate some kind of blockage. This usually occurs when there is coking of the liquid fuel nozzle because of poor purge air flows or reverse flows through the liquid fuel check valve.

A low pressure can occur when the spring of the liquid fuel check valve fails, or when something obstructs the liquid fuel check valve preventing it from operating properly, or when internal components of the liquid fuel nozzle are not assembled properly or become loose, or when the liquid fuel purge check valve leaks in the reverse direction (which occurs fairly frequently on dual-fuel machines which are operated primarily on gas fuel and can also be detected by monitoring the Tell-Tale Leakoff (see the Liquid Fuel Purge P&ID)).

Most exhaust temperature spreads related to fuel flow are the result of low fuel flow, and most low fuel flows are the result of failed check valves (primarily Purge Air Check valves), or coking of the liquid fuel nozzles. Coking causes a blockage which can increase the pressure, but primarily it causes poor atomization of the liquid fuel which can also cause cold spot(s).

But, you are <b>absolutely</b> correct when you say that one of the best ways to find the cause of an exhaust temperature spread when operating on liquid fuel is by using the manual selector valve at the liquid fuel flow divider. The guideline is that the fuel pressures should be within approximately 10% of each other for a machine that has a low exhaust temperature spread. And when the pressure of any can/fuel nozzle differs from the average by more than approximately 10% and the exhaust temperature spread is high, then that can/nozzle is suspect.

The original poster said the "pressure" was 16.0 barg for all 14 cans/nozzles. That would mean the "allowable" pressure difference would be approximately 1.6 barg (23.2 psig) out of 16.0 barg (232 psig).

The original poster also said there was "...a yellow smoke from the stack...." Liquid fuel operation usually does result in a light brown, or yellow, smoking of the exhaust. If there is poor atomization of the liquid fuel (such as caused by coking of the liquid fuel nozzles) then the smoking could be darker than normal.

 

oh so it had both. wish i could work with bigger machines but unfortunately no opportunity as yet. anyways thanks for the info CSA

 

I mis-spoke, again. The sentence below is incorrect.

>If there is poor
>atomization of the liquid fuel (such as
>caused by coking of the liquid fuel
>nozzles) then the smoking could be
>darker than normal.

If the fuel is poorly atomized, usually it will result in white smoke, not dark smoke. Dark smoke is a sign of heavy load; white smoke is a sign of unburnt liquid fuel.

Sorry for any confusion.