LFBV Calibration

S

Thread Starter

SI1405

Liquid fuel bypass valve calibration.

The liquid fuel bypass valve on one of our machines has been identified as over travelling and the recommendation from the Controls TA on site was to adjust this out in the calibration. ACALIB.DAT has been updated so that the Negative saturation current is now greater than 100%. Now when the valve is stroked the correct stroke lenght is achieved for 0-100%.

My concern with this is that 100% is the closed position and now by reducing the stoke in this manner there is a possiblity that the full fuel flow may not be acheived. If the adjustment was made at the oposite end at the fully open position, it would have little or no effect on operation.

The TA has so far rejected any concerns raised and has stated that the LFBV shouldn't close fully to protect the pump. If the this is the case what is the purpose of the relief valve (VR4). I can't help but think i am being fobbed off.

I have data the indicates this valve was running at 99% for base load operation.

any comments regarding this would be appreciated.

Thanks
 
"Fobbed off": to be put off by [(evasion)], from http://www.wiktionary.org. (Had to look that one up! Thanks for the Internet, right?!)

You say the LFBV has been identified as "over traveling". Exactly what does that mean? Does it mean the valve is more closed than it "should" be? Does it mean that more fuel is flowing than "should" be?

We really need to know what frame size machine you are talking about (and we're presuming this is a Frame and not an aero-derivative). If this is a Frame machine, you should refer to the Control Specification, Sect. 05.01, to see what the 100% stroke is supposed to be for the LFBV; it should be stated there, and that is what should be used.

Years ago, when the LFBVs of Frame 7E/EAs and 9Es were fitted with LVDTs, many times they were "short-stroked", meaning they were calibrated such that when the valve was fully closed it would read less than 100%. If I recall correctly, Frame 7E/EAs were to be calibrated such that 100% stroke would occur at 1.125 inches, but the valves only had a 1.000 inch mechanical travel. I've never given that much thought until just now, so I'm not exactly sure why that was done.

You need to consider and remember this: The servo regulator used for control of the LFBV is a *flow-control* loop, meaning that regardless of the calibration of the LFBV LVDTs the valve will go to the position that it needs to go to in order to make the flow divider feedback equal to the liq fuel flow-rate reference. It's just that simple; the LFBV will be operated at the position that makes the actual fuel flow-rate, as measured by the flow divider, equal to the liquid fuel flow-rate reference (usually signal name FQR or something similar).

The LFBV is *not* like a Gas Control Valve (GCV); most units don't have a gas fuel flow-rate reference like there is for liquid fuel. Rather, it is presumed that flow through the GCV is proportional to stroke (position) (at rated speed) and that for a given stroke (position) there is a particular flow through the GCV. Gas fuel flow-rate is not usually fed back to the GCV servo regulator loop. The GCV position *should be* reasonably proportional to flow through the valve.

Liquid Fuel Flow-rate is measured, using the Liquid Fuel Flow Divider. FSR1, the Liq. Fuel FSR, is converted to a Liq. Fuel Flow-rate Reference, FQR, which is compared to the Liquid Fuel Flow Divider Feedback, which is supposed to be scaled to reflect expected Liq. Fuel Flow (meaning, in terms of calorific value, the heat content of the fuel).

The flow divider feedback has a scaling constant, which sounds like it may not be exactly correct. Has anyone compared the fuel analysis of the fuel being used to that of the expected fuel analysis supplied to GE or the packager during the requisition phase of the purchase? That may affect or change the flow divider feedback scaling constant.

Also, you haven't told us if you have observed the LFBV position; there is usually a scale provided to be able to monitor valve position during operation (if we're talking about something other than a Frame 5 or Frame 61B). What is the physical valve position when the unit is at Base Load? Is it really 99%? If so, I'd say the flow divider feedback ain't scaled properly to begin with. Also, at Base Load, what is the Liq Fuel Flow-rate Reference, and what is the Liq Fuel Flow Divider Feedback?

To my mind, the LVDT calibration isn't critical. I would also suggest that if VR4 were to open, that it would relieve so much fuel flow that the LFBV would not be able to control fuel flow very well. It really is just a safety relief valve, and if it opened, the unit would likely trip on loss of flame.

If the LFBV is truly at 99% stroke as indicated on the scale when the unit is at Base Load, something other than LVDT feedback calibration is much more likely wrong. If this is occurring on one machine out of several "identical" machines on the same site, then something is causing the LFBV of that unit to close more than the others in order to make Base Load, something like plugged Liquid Fuel Filters (either Low Pressure or High Pressure, or both), or a slipping Liquid Fuel Pump Clutch, or a worn High-Pressure Liquid Fuel Pump, or clogged Liq Fuel Forwarding Pump suction strainer. Because the LFBV should not be anywhere near full closed at Base Load.

Now, if the Liquid Fuel characteristics have changed, this could cause a problem like this. But, it would seem it would happen on all machines (presuming that all machines have the same Liq. Fuel Flow Divider feedback scaling!). So, why is this one different? That's what enquiring minds want to know.

We really want to know how this turns out, so feel obligated to let us know! (And I'm not talking about the situation with the Controls T.A.; some of them are better at explaining things than others, and sometimes they just don't have time, and sometimes they just don't take the time. He or she might be saying to calibrate the valve per the Control Spec. (if it says to "short-stroke" the valve) and just isn't explaining it very well, if at all.) I'm talking about what else might be wrong which might be causing the valve to nearly close (we don't really know that yet, because you haven't told us the position of the valve from the scale on the valve) to make Base Load.

So, first let us know the position of the valve from the scale when the unit is at Base Load, and then, let us know how the problem is ultimately resolved.

Thanks!
 
Hi CSA,

Thanks for the response.

Over travelling, the valve should be stoking approx 28mm (sorry it’s in metric) and is actually 29-30mm (sorry I don’t have the exact figures to hand) and it is used on a 9E.

The adjustments that have been made appear to be the opposite of the short stroking. The valve will now reach 100% at approximately 96% (28mm) from the fully open position and when pushed to current saturation will show in excess of 100%.

The LFBV is flow control, but will the MKV output more than 100% valve position to achieve this flow? I thought it probably wouldn’t, but I cannot see any clamps in the Big Block. (FQROUTV2)

The data I quoted originally of 99% at base load I have since checked again was probably bad information. It was from a time we were experiencing problems and the HP fuel oil pump has been replaced since and the last data I have was from approximately 1 year ago where the valve was operating at approximately 91%. This was taken during testing after a planned shutdown, normal fuel oil operation would be at lower ambient temperatures during the winter so I am assuming would be high fuel flow. I have seen data from another unit, at colder ambients, that was approximately 95%. Unfortunately we don’t often run on fuel oil. It has been over a year since we actually ran on fuel oil. My concern still is that if the valve is restricted to approximately 96% travel and we usually run 90-95% it doesn’t leave much room, if any. It isn’t the best place to be in if TNR winds right up, you don’t get any control back until it has worked its way back down again. (this happened on one occasion with the HP FO pump issue).

Discussions have taken place with the TA and our engineers on day shift and this has now been accepted. I guess we will have to wait and see the outcome now.

I don’t think I will be able to get any further details from the scale of the valve as it unlikely to be run on FO for another six months.

Thanks.
 
You didn't address the issue of why this unit is different from the other units; you have said there is more than one unit on the site. How do the other units respond?

Have you lookd at the Control Spec, Sect. 05.01; in addition to the 100% stroke details, it usually has a table of expected fuel flows and will list the expected fuel flow, and sometimes valve position, at full load.

Yes; fuel flow will be higher at low compressor inlet temperature.

Yes; the Mark V will move the valve to whatever position is required to make the actual fuel flow equal to the liquid fuel flow-rate reference, whether it's 43.7% or 107.8% or 99.9% or 72.3%.

CHANGING THE LVDT FEEDBACK CALIBRATION ISN'T GOING TO CHANGE THE FUEL FLOW-RATE THROUGH THE VALVE/PUMP WHETHER OR NOT THE MARK V THINKS THE VALVE IS AT 32.4% OR 101.6%. SO, IF THE LVDT FEEDBACK IS CALIBRATED SUCH THAT 27mm IS 98% AND THE ACTUAL FUEL FLOW-RATE IS EQUAL TO THE REQUIRED FUEL FLOW-RATE AT 27mm, AND THEN THE LVDT FEEDBACK CALIBRATION IS CHANGED SO THAT THE INDICATED POSITION IS 94% AT 27mm, THE MARK V IS GOING TO MAKE THE VALVE GO TO 27mm (94% STROKE) TO MAKE THE ACTUAL FUEL FLOW-RATE EQUAL TO THE REQUIRED FUEL FLOW-RATE. PERIOD. THE LFBV SERVO REGULATOR IS A FLOW CONTROL LOOP; THE INDICATED POSITION OF THE VALVE IS IMMATERIAL.

It's very common for many Gas Control Valves and Stop-Ratio Valves to have a physical stroke greater than the effective (100%) stroke value; so this isn't a problem.

It's still a mystery why this unit is operating differently from other units, presumably the same units as this one.

I wouldn't expect a properly sized valve (for the fuel bein run through it) to be operating at or near fully closed to make rated load (remember this is a bypass valve in a recirculation line around a positive displacement pump to the fuel nozzles), and I wouldn't expect that from any properly sized control valve. Even flow divider feedback calibration isn't going to solve this problem; either the valve is the wrong valve for the fuel that's being burned (as opposed to the expected fuel characteristics supplied when the unit was purchased) or there is a problem in the liquid fuel system.

And again, if this unit is operating differently from other units, calibration isn't the problem.

Period.
 
Top