Vertical Turbine Pump Min-Flow Recirculation: On/Off or Modulating 2

[MattNCSU03]

I am designing a pumping system that consists of 3 x 50% vertical turbine pumps designed for 12,800 gpm at 910 ft TDH. Based on preliminary data from a pump manufacturer, the minimum hydraulic flow is about 6,700 gpm.

The Pump Handbook (Amazon Reader) says "Modulated bypass control is required when the minimum flowrate specification exceeds approx. 40% of the rated pump capacity. If On-Off bypass control is used above this point, the pump may exceed its capacity rating with resultant pressure pulsation and unsteady flow"

Based on that and knowing that 6700/12800 > 0.4, I presented the client a P&ID showing a modulating recirculation valve system and received a comment that essentially says "here in France we use on/off control in this application and haven't had a bad experience" and to convince us otherwise or do it our way.

I believe using an on/off recirculation valve with a pump this size is a bad idea based on my limited experience and reviewing the pump handbook but this has not been enough to convince the client otherwise. Are there any codes or guidance out there I can use to reinforce the use of a modulating recirculation line or should I just put in on/off control based on their experience with it?

 

[MattNCSU03]

I should add that the pump isn't likely to spend much time on recirculation.

Normal Sequence:

Pump 1 Start
Flow at 7% until water chemically conditioned (~48h)
Ramp up to 50% flow
Pump 2 Start
Ramp up to 100% flow
Run at 100% for 24 months
Ramp down to 50%
Pump 2 Stop
Ramp down to 0%
Pump 1 Stop

Flow may be held at various levels for a few hours on the way up and down for testing but generally the pumps will be run stead state at 100%.

 

[Artisi]

Make your recommendation and if not acceptable do it their way with the over-rider that any problems "belong them" as the by-pass is their design and recommendation.

 

[BigInch]

It would help if you gave a page number of the Pump Handbook from where that quotation is made. Without seeing that, I'm having a hard time realizing exactly what the operational intent is when they say that modulation is the preferred method to use.

Personally, without knowing the specifics of your system curve, I would suggest starting the system by opening the bypass and keeping it open, or slightly closing it, until you have reached (at least) minimum flow actually going into the discharge pipeline. At that point you could choose to close it, or if it will still be a long time until you reach normal pipeline flows near BEP, continue slowly closing until you get say 60% BEP flow being discharged into the pipeline. That assumes that you can get the necessary head from the pumps at 60% BEP flow to move that flow (by sufficient pressure to open the discharge check valve) into your discharge pipeline. If your static head is really high where it takes, let's say 80% of BEP flow to generate the head you need to open that check and begin flowing into your pipeine, then you'll probably have to keep the bypass open until then. I would try to do pretty much the same thing when stopping, but in reverse sequence. In these transitional cases, all you should normally require is an ON/OFF valve with an appropriately selected travel time set for its actuator.

Your question and the Pump Handbook quotation seem to be suggesting some scenario where things were operating at normal flows, but some variation in that flowrate is required. Perhaps you need to bring on another pump, or change the pipeline's flowrate just a little. In that case, it could be beneficial to have a modulating bypass, recirculating any excess flow, (Pump BEP flow - pipeline target flow) over your pipeline target flowrate. That way you could closely control things such that you could always have near-BEP flow circulating through the pump(s), but still have a slightly different flowrate running in the pipeline. If you had a rough modulating action with the valve, such as being able to set the bypass valve at 10%, 20%, 30% etc, then you could control pump flows to be within 10%, 20% etc. of BEP and control pipeline flows to within the same margin of your target pipeline flowrate. If you had ON/OFF bypass, then you would effectively be agreeing that you don't need to control pipeline flowrates to such a close tolerance, or perhaps you could do that by another method, such as adjusting a pipeline backpressure valve somewhere in, or at the end of the pipeline, or otherwise control flowrates, with a discharge flow control valve. If such things can be done with your system, or a tight margin of flowrate control is not needed in your pipeline, then you wouldn't have any need for a modulating bypass valve at all... as I see it.

**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[JJPellin]

We almost never use a simple on/off minimum flow scheme. I agree with the comment from The Pump Handbook. With a required minimum flow of 6,700 gpm, we would size the spill-back for 6,700 gpm. The pump would be started up in full spill-back and gradually ramp up the outgoing flow to the desired set-point. Once the outgoing flow increased above 6,700 you could close the spill-back. But, at that point, the total flow would be 13,400 gpm, well beyond rated flow for the pump. The same problem would occur on the way down. When the outgoing flow dropped below 6,700 gpm, the spill-back would open and the pump total flow would exceed the rated flow again. Admittedly, the exceedance would not be by much (about 105% of rated). But those are awfully big swings to be taking with a pump this large.

My comments are based on my experience in an oil refinery. In our situation, almost all flows must be tightly controlled in order to hold a constant level or pressure in the upstream or downstream equipment or to optimize the operation of a fractionating column, fired heater or reactor. We could never tolerate the sudden swings when an on/off spill-back opened or closed.


Johnny Pellin

 

[BigInch]

JJ, while that is quite true, I don't think comparisons to refineries are valid in this case. Water systems are built to accomodate daily variations of flowrates of 50% or more. An oil refinery would probably be shutting down and trying to figure out what happened after 15%.

**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[BigInch]

JJPellin, please don't take that the wrong way, as critizing your comment. It is totally correct. I was just trying to point out that there could still be a reason that it might not be accepted by a water pipeline client. Thanks for understanding.

**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[MattNCSU03]

BigInch: The best way to find the section is to use the "Search Inside This Book" link and look for "minimum flowrate specification exceeds" It is on Page 11.128 in the book or 1170 on Amazon but much easier to find by searching the text (unless you own the hardcopy)

I appreciate the oil industry perspective. This particular application is for a base loaded power plant so flow variations will be relatively minor under normal circumstances. A control valve station is located downstream of the pumps that will throttle the flow to maintain level in the receiving tank.

According to a model I created in AFT Fathom, the system curve has a rather gentle slope over the plant's operating range (most losses due to elevation change).

Under abnormal conditions, the plant could experience a rapid reduction in flow demand but this would be buffered somewhat by the receiving tank so a fast response isn't necessarily required.

 

[BigInch]

OK. My mistake again. I should have guessed it wasn't a pipeline and certainly should have known it was not an oil refinery, right? I guess my tunnel vision is getting worse. What the hell is a powder plant? Cement powder? My point is basicly that the quality of the comments you receive are directly related to the quality of information supplied with the question. How is it that we've gone so far along and you still haven't mentioned actually what is being pumped. Instead of allowing us jump to a lot of erroneous conclusions, wouldn't it be better if you could manage to some good relevant info like product and application. Thanks for explaining the search function. Do you think I have time to use it? Anyway, the next time I want to search for an answer to a question I have, I'll know how to do it.

Sorry for the rant, but it would be nice if you made it a little easier for folks that are really just trying to be nice and help you. So, OK. Did you get an answer to your question, or is there something else?


**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[MattNCSU03]

I wasn't referring to the search function to be rude, it is literally the only easy way to get to the page I was talking about (unless you want to hit next page 1,170 times). I spent a few minutes trying to determine a way to directly link to the page but with the Amazon Reader, there is no way to do so. I could have taken a screen shot, saved it to a file and attached it to the post but then I didn't want to post copyrighted material to the site and there would be no way to look at other pages for additional context.

I am not sure why you took so much offense to my follow up post. Just like you and the other contributors in this thread have been nice to take time from your busy days to provide insight, I tried to provide additional information as best I could. I apologize if it came off wrong. You are right, I should have should have been more clear about the material pumped. I forgot that people on this site come from a wide variety of backgrounds. I've been in Power my entire career and the only thing we pump in mass quantities is water. I didn't think about refineries and chemical plants. I’m pumping water at 115°F.

My question has been answered but I am always interested to hear other opinions.

I don’t know what a powder plant is either or why we are talking about one…

 

[BigInch]

No worries. I'm probably just moody today. And I tend to be "short" anyway. Likewise I meant no offense, just trying to tell you how you can get the best help from us in the fastest manner possible.

Yes I do truely believe you were just trying to be nice and show me how to use the search function. My point, which you do understand now, was just that I don't have time to do it, nor did I really want to do that. Copyrights? yes, but you wouldn't have to copy and post that at all, just the link and a page number would have been perfect. It took a long time to load that link, then I said, "I've gotta find it myself?", not realizing it would have been easy to do it, if I knew how.

I'm frustrated with myself too, for jumping to unnecessary conclusions then beating JJ up, etc., etc. But really nobody to blame for that, but myself. Sorry if I took that out on you. Then I went and misread "power" for "powder". I am seriously not batting 100 today.

OK, let's both do better in the future. No hard feelings.

So then this thing is like a hydraulic storage power plant and maybe you pump up water during the night and make electricity by day??? OK, so what little I do know about hydraulic power plants is what you finally confirmed, that you don't want much variation in flows at all then. That pretty much leaves only start and stop considerations as a primary concern, for which ON/OFF will probably do just fine, as your times at low flows will be transitory only, probably only for a couple of minutes, so ON/OFF will probably do just fine in most cases. I'd just keep it simple and go with the client's recommendation, but then again, I don't have to change the PIDs.

I really AM happy you know what you want to do about this now. Good luck with your client... and sorry for the mess up

**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[MattNCSU03]

No worries, there is too much blood in my caffeine vessels right now so I probably didn't word things the best.

The pumps are actually the condensate pumps for a traditional power plant. They pump condensate through several feedwater heaters into a dearator/feedwater storage tank. I probably should have said that from the beginning instead of being vague.

Artisi's post reminded me of a process we have in place to declare something vendor or client input which takes us off the hook and your posts gave me the piece of mind that on/off recirculation can be managed to prevent damage to the pump.

 

[BigInch]

Excellent. OK, watch that caffeine pump. That one doesn't work so well in ON/OFF mode. There I do agree that modulation is by far the better method.

**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch

http://virtualpipeline.spaces.live.com/