When will Parallel Pumps Fight? 1

[gte682n]

Hello, This was inspired by the recent thread on series vs parallel pumps.

I would like to understand in general terms when to worry about 2 pumps discharging into a common header fighting with each other?

If the two pumps and associated piping are identical, then they back up to the same point on their individual pump curves at the intersection of the system curve, and pump flows are equal at equal heads. No issues with this mode of operation.

On the other hand, if one pump generates double the head at the same flow rate as the other pump, the higher head pump will win and the actual total flow will be at the intersection of the system curve and combined pump curve. The higher head pump will be pumping more flow than the smaller pump. To get the actual flow rate of each pump, you would go to the individual pump curves, locate the observed system head, and read the flow for each pump at this value for head.

Is the main concern with having dissimilar pumps in parallel service the potential to dead head the smaller pump? What are other items to worry about besides inefficient operation?

Thanks,

 

[MFJewell]

I have seen first hand what happens when you have two dissimilar pumps in service. I had one large pump in parallel with three smaller pumps. Whenever the system flow requirements were low, whatever small pump that was running would deadhead (pressure spike), then recover, then deadhead, then recover. It would cycle like that over and over. I suspect an undersized recirc line to the tank was partially to blame. If the system had been designed with dissimilar pumps, it may not have been an issue. But the larger pump was added after the fact and the recirc was not changed. I have since replaced all the small pumps with larger units that have curves that are much closer to the one bastard pump (the 3 small pumps were original from the late 1960s). I haven't seen any issues since then, even with the undersized recirc (and system piping that is most likely too small for what operations wants to run through it).

 

[JJPellin]

The worst problems I have had with pumps running in parallel occurred when the pumps were identicle with very flat curves. When one pump became weaker than the other as a result of increased internal clearances or reduced speed, the weaker pump could drop below minimum stable continuous flow. On the other hand, I have a number of sets of dissimilar pumps that run very well together. In all cases, the smaller pump has a very steep curve with a shutoff head much higher than the larger pump.

Johnny Pellin

 

[LittleInch]

Yes dead heading the lower head pump / smaller pump is a concern, but you can also get into surging and oscillations in flow between the two depending on the amount of difference and the pump curve.

"No issues with this mode of operation."
Even two "identical" pumps have issues if you're operating in a flat area of the pump curve or you have high static head and many threads show large variance in flow between two so called "identical" pumps.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[e43u8]

gte682n,

To minimize the concerns with having dissimilar pumps in parallel service it's required, at least, they have similar shut off pressure and steep curves...

 

[gerhardl]

And in addition:

1. Correctly sized and placed check-valves, preferably soft-closing will help against gulping actions, along with correctly sized piping for all production variations will help to avoid problems.

2. In addition to visible problems ono pump (of two or more in parallell) might supply less than others against higher pressure and give more wear. Solution: check or measure flow from each pump.

3. Pumping up and then downwards piping to open tank may give suction with uneven results for the pumps.

4. Construction details as larger piping size or vessels (manifolds) for evening out feeding and reception pressure could help.

5. Air pockets in the system could give (for all pumping solutions) give problems if not correctly vented.

 

[JJPellin]

Actually, in my experience, it works better if the smaller pump in the set has a higher shutoff head than the larger pumps. I have attached two sets of curves from dissimilar pumps currently operating in parallel in our refinery. With a higher shutoff head, the smaller pump is protected against being backed out by the larger pump. And, the larger pump is protected by its size. The total required flow could never be satisfied by the smaller pump, so the larger pump cannot be backed out. It took me some time to come to believe that these sets of pumps can run well in parallel. But, history has proven that they do.

Johnny Pellin

 

[JJPellin]

Both curves were not attached to my last post. Here is the other set.

Johnny Pellin

 

[AlaskanME]

I attached a double pumping guideline for your reference.

 

[Meshry]

Dear Johnny Pellin,
Currently I am working on some project to run two dissimilar pumps inn parallel,
I have seen your recommendations about the pump shut off head and I fully agree with that.

in my case the two pumps are not sharing a common suction. each pump has its own suction vessel however the operating pressure for the two vessels is the same; off course the two pumps share a common discharge header.

my concern here is the friction losses in both suction and discharge sides. what if there is a pressure drop at some side which is different at the same side of the other pump??