Pressure control with parallel pumps 1

[xJAyx]

I have a client that wishes to run a pressure control loop using two identical pumps in parallel. The pumps will run on VFDs and the system will not have automatic isolation valves for the pumps.

The pumps will feed a cooling loop to several pieces of equipment that will be switched on and off throughout the day and will have their own independent isolation valves controlled by their independent control systems. All of these pieces of equipment require a constant inlet pressure on the cooling water. The equipment can tolerate some fluctuations in the pressure so an extremely fast response time is not critical. There will be a pressure relief bypass to prevent an overpressure situation.

The temperature control of the cooling water is separate and isolated from these two pumps.

I have never worked with parallel pumps in a pressure control loop like this. has anyone else done this and how would you suggest controlling it? I am thinking there are two main avenues to the control, one is to run both pumps in parallel at the same speed at all times, varying their speed through the VFDs. I would program in a pump cut off if the two pumps in parallel can not bring the pressure down low enough. The second idea would be to use the first pump until it can not maintain the system pressure, and then bring the second pump online as a kind of booster pump. I don't know how the pumps will behave in either of these scenarios, but I suspect running the two pumps at different positions on their pump curves will not work properly.

Any help is appreciated.

Thanks,
JAy

 

[itsmoked]

This has been discussed several times. You might want to use the search function to hunt them down. None of the discussions were in this forum. They were in either the forum237 or the forum407 forums.

Generally you do kind of both methods. Ramp one until it can't muster then bring ON the other. Keep in mind you would not likely be running any of the pumps at less than about 60% since their performance tanks down there. Also
when you bring ON the second pump, you often need to bring back the first pump so they match. Then bring them both up as needed for further flow.

Other things to note. Some VFDs understand this and can help with this specific application.

Often it's very beneficial to have the two pumps trade-off the lead so one doesn't get 'used up' years before the other.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

A stellar reply Keith.
I would add again that when two pumps are running they should both be developing the same head.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[djs]

Do the pumps have check valves? If not you will probably have to run both pumps all of the time.

 

[xJAyx]

Keith, Thanks for the great reply and my apologies for not performing a thorough search before posting this.

Bill, Yes, that was my concern when running the lead/lag configuration, that if they were not running at the same speed, he system might become unstable. I will run them at the same speed with a lead/lag alternating configuration.

djs, yes, they will have check valves.

 

[roydm]

If they are identical centrifugal pumps they will share the load just by virtue of the pump curve.
I have often seen multiple pumps in parallel where they come on as required, I'm thinking of a water supply system, as the demand goes up another pump is added to maintain the pressure, there is no point in ramping the speed up from a low setting the oncoming pump needs to be nearly full speed to reach the head of the one already on line.

Positive displacement pumps like a gear or vane type would be a different ball of wax of course.

Roy