A recent question I received from an engineer in the Middle East had to do with a “double control” of boiler feed pumps (BFPs) at his power plant. The system was controlled both by set position of the valves as well as variable speed drives. I was intrigued. Was this control method unique to his region? Or varied from one country to another? Or was it simply another method of control, which perhaps depends on the pump’s power level, main versus standby status and start up units, etc.?
Question from Reader
I am a mechanical engineer working in a power station in the Middle East. In our plant we have many heat recovery steam generators (HRSGs) as part of a combined cycle. My question is regarding feed water pump operation. I have noticed that most of our HRSG’s feed control valves are in the range of 50 to 60 percent open position. My understanding is that this means our feed pumps are operating at 50 to 60 percent of the best efficiency point (BEP). Is it correct? I presume that HRSG feed water pumps are normally sized for maximum HRSG load conditions, which would be the case if the control valve is wide open. Could you please confirm the relationship between control valve opening and pump operating point? Also, how can I find out the actual pump operating point for pumps running in the field?
I asked two of my colleagues working in U.S. power generation utilities to comment. Below are their replies to Mr. Salman’s question.
Answer from Colleague 1
I checked with the control room regarding Ammar’s question. Since our BFPs are variable speed the discharge control valves (or drum level control valves) are at 100 percent open and we control the flow by pump rotations per minute (RPM).
Thinking back to my former job as a control room operator, where the BFPs were constant speed pumps, the discharge control valves ran in the 50 to 60 percent open range. We had to run two pumps for full load, so each pump was 50 percent.
From my understanding of control valve sizing, the maximum valve flow would probably be in the 70 to 80 percent open valve position. I would guess you would not want the max flow at 100 percent valve opening. You would want to leave a little margin, in case the sizing wasn’t exactly right (which it never is). There are a lot of valve trim designs that have very different positions versus flow relationships.
Regarding the control valve in Ammar’s case—it would depend. But I guess that the assumption that since the control valve (CV) is at 50 to 60 percent open, it means that the BFP at 50 to 60 percent BEP is incorrect. I would speculate that there is no relationship between the CV position and pump’s position on the pump curve.
Barry Brown, P.E.
Sr. Engineer - Spurlock Station
East Kentucky Power Cooperative
Answer from Colleague 2
A few principles apply here:
1. The feed control valve position is entirely irrelevant with respect to BFP as a percent BEP. The feed water control valve position is determined by a so-called three-element controller:
- steam flow rate (feed-forward)
- feed water flow rate
- drum level
2. In general, the power plant boiler feed pump is sized to provide full feed water flow, plus some amount of spray water, all at maximum continuous rating (MCR).
In theory, this point should be as close as possible to 90 percent BEP (leaving a little extra to recover from low drum levels at full load, and during transients).
3. Flow rate is far from linear with respect to valve position.
4. Determine the pump operating point on curve by application of well-known pump performance formulas, and compensate for feed water density changes at elevated temperatures. (Use steam tables or public domain Excel add-ins).
If you have any more questions, please do not hesitate to let me know.
Mr. Salman replied with his thanks, also adding that BFPs with variable frequency drive (VFD) motors are unusual in the part of the world where his plant operates.
In closing, I submit my own thoughts to Mr. Salman and Pumps & Systems readers.
Answer from Dr. Nelik
I would be interested to hear from other readers how common such control techniques may be at their plants, and if this is unique to a specific region or particular operating practice. It would be interesting to obtain a curve showing the relationship between the valve opening and the flow. In my experience, this relationship is very nonlinear. There is hardly any change in flow when the handle of a valve is turned even for many turns (even 10 to 20 percent of the max turns).
Then, suddenly, the flow (and backpressure) begin to change. I am unclear about the relationship between flow and turns after that happens toward a full closure, but I suspect it is also nonlinear.
Also, I am still not fully clear with your closing comment about power plant boiler feed pumps with motor VFDs being unpopular in your plant.
From what we heard from Barry and Kristo, it appears such control is not bad, and in fact does provide good flexibility, provided, as they noted, that the valves are positioned properly as well.