Headloss Calculation for Parallel Pumps with Common Suction Piping

[WastewaterM]

Hi everyone!

New guy here though I’ve lurked quite a bit. I have a couple of questions that have been bothering me a bit regarding pump/system curves in a parallel pumping set-up and came across this old thread (

http://www.eng-tips.com/viewthread.cfm?qid=245347

- numbered as thread407-245347) in my search for answers. JJPellin did a great job on directly answering one of them—how to appropriately calculate the NPSH of a system when you have variable flows/number of pumps running in a set-up, but I do have one thing that I’d like to clear up and would appreciate if someone is able to provide an answer. Unfortunately it doesn’t appear that I’m able to reply directly to that thread anymore, so I’ve started a new one.

Say you have the same set-up as in the old linked thread--parallel two-pump set-up with common suction and discharge piping lengths and short individual suction/discharge suctions leading into/out of the pumps. How do you treat the losses in the common suction section? Let’s also say that at our 2 pump flow we have 1,000 gpm through the system as our design point with each pump contributing 500 gpm. I’ve have a spreadsheet that created a system curve based on just the common discharge headlosses that was based on the total flow through the system, and the same spreadsheet calculated the minor/individual losses by adding the losses experienced at 500 gpm through the remaining sections of pipe. This makes sense to me to do for the individual sections leading into/out of the pump, but what about for the common suction section? Should a flow rate of 1,000 gpm been used instead and the losses added to the system curve instead of being treated as individual losses? Is it possible to work this either way?

Thanks for any input!

 

[BigInch]

split the flow to each pump's suction piping according to what ever each pump is taking. If you have 501 gpm in one side of the header and 499 in the other, the head loss is calculated in the first side using the 501 gpm, and on the other using 499 gpm. The trick is knowing the flowrate in going to each pump.

If the piping is sized properly it shouldn't be much more than a few psi in the short pieces of suction pipe that we're talking about, if that much.

I hate Windowz 8!!!!

 

[WastewaterM]

Thanks for the response. I had previously designed several pumps with short/non-existent/purely individual suction runs as you stated (using flow to each pump for the suction side) where it really wouldn't have changed things by more than a few psi and never gave too much extra thought about the situation because of this. My primary motivation for asking this now was I have a pipe run of ~150’ on the suction side so there are major losses to be considered that could swing things if it was one way or the other.

As a side note the NPSH is fine as we’re pumping from an extremely high head condition that helps give 35-45 feet available in most scenarios and for this particular set-up the flow split will be pretty dang equal despite the draw length.

 

[LittleInch]

I think you've answered your own question. Each part of your system you divide into sections with the flow that each section has until it reaches a branch. That point will be common for both individual systems in terms of pressure. Then each inlet and discharge line will have a different flow and pressure drop, but at the point they meet up will have the same pressure. Hence there may, as BI notes, be a difference between the pumps, say 55/45 if "identical", but the increase in pressure shouldn't be that significant in the whole system. If it is do one system curve at 50/50 and one at 55/45 to get a range.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

Length theoretically makes no difference in terms of the calculations. The pressure drop is what it is, for 1 ft or 150 x that (as long as the diameter doesn't change).

I hate Windowz 8!!!!