Turbine Control Valves

[marada93]

I'm seeking information on operation of turbine control valves wide open. Currently we have sequential valve opening, yet at full load we end up with two valves atapprox 50% open, and a third at 15-20%. This equates to a 300 psig throttle pressure loss across the valves.

My feeling is that the turbine efficiency would be increased by opening the valves and taking the pressure loss in the turbine, and not through the valves.

Any supporting documetation out there?

 

[byrdj]

The answer envolves a lot of considerations.

To simply go valves wide open and carry the same steam flow, you will need to lower the throttle pressure. the simplified approximation would be CV%rated/CV%reduced = throttle pressure rated/throttle pressure reduced. If the MWs increased, you improved the HP efficency.

please note, I am only familar with impulse design turbines and following is NOT TRUE for REACTION design.

For the info provided, you are assumed that the throttling losses at the 2 full open CVs is minimun, but for #3 at 15%, #3 is causing a loss at the valve.

Instead of valve throttling losses alone, Another way to look at HP efficeincy is the steam velocity at the first stage. If your turbine has "Governing Control", which means each CVs feed a segment of the first stage, Then maximun efficiency is acchieved if the steam velocioty out of the nozzle block is twice the 1st stage bucket rotational velocity. The efficieny curve of bucket/nozzle velocity is a bell shape curve with 0.5 the peak (and 0.2 & 0.8 being ~50%). Thus in the current operating condition, the velocity for CV#3 could be less than peak and actually acting as a braking force. Assuming you have 4 CVs, going CVWO MAY reduce the steam velocity below peak (poor efficency). Therefore reducing throttle pressure and allowing the CVs to open till the next valve is just about to open could be the best by having 3 nozzle segments at near peak velocity. (I have read that actually opening the next valve a few thousands is the best for efficency, but I think the wear on the seat would become a problem.) If you have 4 Cvs(or more), and 3 full open (reduced throttle) is determine to provide best eficiency, the 4th valve can be mis adjusted to allow CV3 to open slighlty more with out opening 4th

To determine actual velocities, knowing 1st stage nozzle area would be benificial. Having the nozzle rebuilt to match your operating conditions would allow operation at near rated throttle pressure with CVWO improving efficeincy. Is it possible that the first stage nozzle has opened (worn)?

 

[marada93]

Heres what I've got.

I have four CVs feeding a nozzle block admiting steam to the 1st stage.

To prove a point (mainly to myself), I lowered inlet steam pressure from 1450 psig until the four valves were wide open. (Approx 1200 psig)So it wasn't quite a 300 psig throttle pressure loss, but you see my point.

Still maintaining rated load, and still maintaining all extraction pressures as they were prior to dropping pressure. This proved to me that the pressure downstream of the CVs must have been 1250 psig with the three valves throttled. Therefor, the turbine didn't realize there had been any change at all. Follow me?

My thinking is this, if I can make full load at 1250 psig, and the turbine is well within design admission steam conditions, wouldn't this be better from the following point of view:

1- No throttle pressure loss across the control valves may increase the life of the control valve seats and plugs. Less wear and tear. Thermodynamics should impove with equal steam loading through the for admission points.

2-Lower admission steam pressure equates proportionally to lower boiler pressure. As such, tube life may be improved.

3- Lower boiler pressure also means that drum pressure is lower, therefor the feedpumps wouldn't have to work as hard to deliver water to the drum. Possible efficiency improvement due to lower motor current on operating feed pumps. As such, auxiliary power requirements would be lowere making more MWs availlable for production. Lower firing rates should be realized.

Theres another really important point I had as well, which is quite significant in spite of the fact that I can't remember what it is right now. I'll follow up later.

I recall as a control room operator years back on an 820 MW machine, that at full load we were alway required to be VWO.
Thats just the way it was. It made sense to me then. I just can't seem to find a way to quantify the improvement in efficiency.

Anybody familiar with this?

 

[byrdj]

If lower steam pressure from 1450 to 1200 open the CVWO (100%), the original total CV opening should have been about 83% (1200/1450)= (X% / 100%)

In response to your items
(1) Absolutley, with a CV just open, the velocitoty accross the seat will allow SPE damage (usally called drawing on CV seat). For the turbine, steam flow will be over the 360 degrees of the nozzle and allow uniform thermal.

(2) & (3) I don't KNOW the effects on the boiler, but like you, I feel it must help. Also the lower pressure should reduce the power needed to pump feedwater. Farther improving the plant's efficency. My question would be the boiler's efficiency, ie transferance of combustion heat into the steam.

While all 4 CVs wide open will be the Best from a mantance standpoint for the valves and the turbine. It may not be the best for efficency. For maximun efficency, the steam velocity at the nozzle most be peak. For an impulse design turbine, the wheel speed/steam speed needs to be 0.5 with equal losses occuring whether faster or slower, for a reaction design, the peak efficency occurs at 0.7 with efficency losses for faster steam, but the losses are not as great for slower steam (80% at 1.5) For a impulse, the efficency may be improved by using 3 valves wide open 4th closed. for a reaction, the efficency loss would be less at CVWO with less than peak velocity. Do you know what type of turbine design you have?

To quantify the results, the HP turbine efficeincy is the best method. HP efficency is easily calculated by determing the "h" enthalpy drop across the HP turbine. All you need is steam press / temp at throttle and at HP exhaust.

For the plant total efficency, measuring fuel compsumtion is the bottom line.

If you take data at varous conditions, I would be interesteing in reviewing.

I think you are definetly on the right track and this type of questions is becoming asked by more utilities. The next step in your search for efficency and less severy thermal cylcling would be to start making load changes with pressure and leaving the CVs are far open as possible.

 

[marada93]

Thanks byrdj.

Your deeper into the valve curves than I am. I can tell by your responses.

I'm just looking at it more from an operator perspective, and trying to squeeze a bit more efficiency out of a unit that runs great already. High capacity and availability.

However, when your already running in the upper 90 % on the unit, making changes (even improvements) is a tough sell to upper mgmnt, if you know what I mean. But this just seems like something there for the taking. No major capital expeditures, in fact...no expeditures at all. We just do it.

So I'm trying to get enough ammunition to make my case. I think the answer is to try it over the course of several months, then compare heat rates before and after. I'll post back if/when I get results.

By the way, this turbine is mostly impulse, with several stages of reactionary blading on the low pressure turbine.

Thanks for your help on this! Keep looking though. I may need you again

 

[byrdj]

Sounds good

While heat rate comparisions are the "last word", a simple enthalphy drop effieciency comparison would give an ideal if improvements for a different steam condition occurred. If you could supply Press. and Temp. for Main steam entering and high pressure turbine exhaust exiting for rated (cv throttling) and reduced (CVWO) pressure. I will calculate the two efficiencies. You would need to operate at each condition any where from 10 minutes to an hour untill "steady state" is noted before recording data.

I am not familar with the $$ side of the calculation (cost of fuel, etc.), but a quick calculation (done by another engineer earlier this year)for a unit to reduce pressure (to a specific value) at night during minimun load hold at 60MW would easily save $80K/year for only a 0.7% efficency improvement.

From some "expected" correction charts, the HP efficeincy varies 3% from worst throttling lift to best for CV 3 and 4.

 

[poetix99]

marada93:

Steam OVERALL turbine efficiency CANNOT BE IMPROVED BY THROTTLING AT THE TURBINE INLET. This is true if you account for the available energy from upstream of the trip & throttle valve, as you must.

The Carnot efficiency of the cycle will be reduced if the boiler is operated at a lower pressure, with the same (or lower) temperature as the original (higher) pressure. Boiler life, etc. should be factored into these considerations, but not in an anecdotal way; ignore the boiler life in figuring out the maximum turbine efficiency, OR INCLUDE IT WITH A RIGOROUS ANALYSIS OF THE ENTIRE CYCLE.

Boiler feedpump work is relatively small; that is one of the advantages of the Rankine cycle; the fluid (liquid water) that is being pumped has a very low specific volume, and the specific work required to raise the pressure is very small.

It seems unlikely (but possible) that the first two valves are "50% open", and that the third valve (of four) is "15-20% opened". Multi-valve steam turbines are typically sequenced so that as one valve if EFFECTIVELY fully opened, the next valve in the opening sequence begins to open. This is the means by which part-load performance is kept at a relatively high level; the throttling losses are confined to only one group of nozzles. What valve lift dimension is the basis for determining the percentage openings that you have indicated?

Regardless of the actual effective openings of the first three valves, it seems clear that if you are operating on three of the four control valves, the turbine is substantially oversized for the "full-load" that you mention. The best way to improve the turbine efficiency, is, as "byrdj" as obliquely hinted at, to resize the first stage nozzles (and control valves) so that all four valves are fully opened at full-load. Perhaps there will be some margin of nozzle area in allowance for variations of steam conditions, or in allowance for fouling - clogging - of the nozzles, but it seems clear that you have more than a normal amount of "margin".

 

[poetix99]

I made two typos.

#1
From the first sentence:
"OVERALL steam turbine efficiency..."

#2
The word "if" should be "is" in the passage from the 4th paragraph, requoted below:
...so that as one valve is EFFECTIVELY fully opened...

 

[marada93]

You are corect in that the turbine is oversized for our normal "dispatched" full load. Approx 15% over. However these are the conditions we run at when we are at our full dispatched load...not the turbines rated full load. I should have been more specific.

The valve positions are sequential, and the valve positions stated are approximate...give or take varying condenser backpressure and the like. But the curves for these valves indicate that once the valves are approaching 60% open, they are effectively passing 100% steam flow..or close to it. I should note that the 1st two valves open as one. Then the 3rd one opens at NEAR the effective full open, followed by the 4th.

The dispatched load is where we run virtually all year round. And, with a very high capacity factor. While I agree that there may well be significant changes (and expenses) with changing the nozzle block size, or valve dimensions, would it be safe to assume that simply lowering the steam pressure until the valves are full open would increase the turbine and/or unit efficiency with zero capital expense involved? It seems logically that this would be much better than throttling on three control valves.

 

[byrdj]

Clarifications to my comments for "impulse, governing nozzle" and sperical disk / ventura seat CVs. The only type i work on.

Effictive full open stem lift can be determine knowing the seat diameter and stem to disk crack point lift. Basically when the area of the seat equals the opening area (seat diameter X disk lift). for example a CV with a 0.03" crack point lift and a 4" valve seat diameter will be 100% open at 1.03" stem lift.

Governing nozzle area is several times valve seat area. The flow per lift curve looks something like a cubed root. Like 90% flow at 70% open.

To linearlize steam flow with governor demand (the CV operating cylender for rack operating cams) the design CV opening sequence will have intercept points (controling valve lift when next valve is at crackpoint lift) at about 60%. The measurement of flow linearity is reffered to as increamental regulation. (another topic of discussion) For example, if the unit speed regualtion is 5%, the valve sequencing could have a loading flat spot where the regulation "at that increament" is only 10%

Besides controling mass flow, the CV opening determines to steam nozzle velocity. For example, when #3 just opens, mass flow through the turbine increases, but the 3rd quadrant nozzle flow is less than the rotation speed of the bucket. This "throttling loss" will reduce the HP turbine efficeincy. A plot of HP turbine efficency verses flow (for rated throttle pressure) will be a scallopped like curve, with best efficiency occuring near the CV intercept lifts. The deviation from an HP efficency curve drawn from "best valve" points and true efficency will vary with the number of CVs. that is a 4 cv unit will have a larger decrease in efficeincy than an 8 valve unit. The decrease in efficency will also be graeter with less flow. The "expected", worst throttling, decrease in HP efficency for a 4 valve at 85% flow is 1 to 1.2% (2.5% at 35 % flow).

A 1% change in HP effiecency will only result in a 0.2 to 0.1% change in heat rate. As you can see the expected, if any, improvement is minimum (but a $ is a $). "Hp efficencies" (not total plant) at different operation condtions of valve openings and throttle pressure can be compared with an enthalpy drop. All that is needed is steam in and out, press and temp.

"to Measure is to Know"

The "changes" to the nozzle my be needed repairs. Operators complaining about lost maximun MWs are common when HP nozzle area (and efficency) is restored.

Hope this makes sense, i've been up for 36 hours (problems with returning from outage and airline delays)

And a Happy Thanksgiving to All!!!

 

[poetix99]

I understand (up to a point) your inordinate concern with “throttling losses”, etc. but you are ignoring the “throttling loss” that will occur upstream of the control valves if you try to force all of the control valves opened at your typical operating point.

A multi-valve steam turbine is designed exactly for the purpose of minimizing the throttling losses at “part-load” (by which I mean any flow at less than valves wide open - VWO). Give up the idea that reducing the inlet pressure further upstream is going to help the plant heat rate.

If the valves are properly sequenced, only the third valve is throttling in the case that you’ve cited. As “byrdj” described, the valve opening is overlapped a little bit. This is for some economy of total accumulated valve lift within the valve chest, and (as byrdj said) to keep the flow-lift characteristics more nearly linear for controllability (sp?).

The first two valves are effectively wide opened. You ARE paying a performance penalty for the fact that the turbine is effectively oversized. You will not recover that, or even reduce that by reducing the turbine throttle pressure; in fact you will make it worse.

Indulge yourself in an experiment: if you have enough instrumentation to determine the heat rate over a relatively short period of time, you could throttle as you wish to and see for yourself what is the difference in performance. HOwever, your evaluation of each case must be based on the ORIGINAL steam conditions upstream of the trip valve; that is what you have "paid for" out of the boiler. If you do this, you must realize that you will have lost your high capacity factor for that machine. It might be necessary to inform the regional dispatcher of your experiment (in which case, it would no longer be a quick and simple thing to do...).