Control a pumping system off pressure or flow?

[KernOily]

Looking for opinions here... I'll try to explain this as best I can. I have a pumping system consisting of two identical 150 hp 3600 rpm pumps in parallel that provide water to a variable load. The pumps take common suction from a tank. Sometimes there is no demand throughout the day, sometimes both pumps are required to meet the demand. The service is well kill water and the pumpage is 150° F oilfield produced water.

Sometimes the load is below one of the pumps' rated minimum flow so there is a variable FCV bypass to bypass some of the discharge back to the tank. The system must remain pressure charged at all times so that there is adequate flow when the need arises. This is sort of like a firewater system, in that the system has to stay pressured up all the time in case there is a demand. The system is controlled by a local PLC. I am stuck with making these two big pumps work, else I would recommend one big pump and a smaller jockey pump.

I am in a debate with the owner's engineer as to whether this system is better/more easily controlled using a discharge pressure signal or a flow signal to stage the pumps off and on line. I claim the load is better met, and the controls will be more stable, i.e less stops and starts, controlling the pumps off flow; he claims pressure.

My plan is to set the pumps to come on-line each day at 5 am since there is no demand at night. If there is intially no demand upon startup, i.e. if the discharge FT says zero, then open the bypass valve and flow the minimum bypass rate for 5 minutes. If the discharge rate does not increase in this time, i.e. if there's no load, shut off the pump after the 5 minutes. The bypass valve and the second pump will be controlled off the discharge FT.

I'm thinking he may have a point in that it might be a good idea to stage the second pump on-line on a pressure signal. Opinions? Thanks!
Pete
pjchandl@prou.com

 

[MortenA]

There seems to be two problems:

1) Controlling the flow (by pass flow)
2) To determine when to turn the 2. pump on

I can only come up with something for 2. (for now):

I beleive that using P for turning the 2. pump on will be the most reliable thing. But you have to think about when to turn it off not to get a situation where the pumps turn on anf off constantly.

Best Reagards

Morten

 

[quark]

Elsinore!

A pressure switch after the end user pointin the header can be better option. Try to control the flow with a variable speed drive. This will give you better result.

Regards,

 

[stressriser]

Elsinore,
Quark is right. Use a pressure control sensor (not switch) to maintain a constant pressure in the system. Write a program into the PLC that monitors the pressure sensor. Then you speed up or slow down the motor (hopefully, you have a variable drive motor) to adjust the flow.

In essence, you will be doing both. You monitor the pressure and adjust the flow to maintain a constant pressure.

There is an alternative also. There is a valve made by "Cycle Stop Valves" that will vary the flow without variable speed. There literature says they are at

http://www.cyclestopvalves.com

this valve is definately worth looking at.

Stressriser

 

[KernOily]

Thanks for your replies.

Unfortunately I do not have variable-speed drives avaialable for this, and I'm not sure the project will justify the use of VFDs, although I still have to run through the economics. I should have mentioned that in my first post. Thanks!
Pete
pjchandl@prou.com

 

With hydraulics of this type of system, the pumps are alike and designed to meet a certain discharge pressure, which both can do. Your increasing demands will drop pressures when one pump is in service to a point where it can no longer meet the demand, thus, using a pressure signal to turn on the second pump is the thing to do. But the problem is when both pumps are running and the bypass is operating, they will both strive to maintain the setpoint pressure. How would you know to turn one off? Using shut off head pressures is not a good thing to do since you dont want your pumps operating close to them and off their curve. Thats where flow comes in. Use the reduction in flow, taking into account bypass flow, to shut the second pump off.

 

[Drazen]

There is no better thing than variable speed drive. However, it is so expensive that can't be justified easily.
What do you think about TWO pressure switches, one for each pump. One pump, call it base pump, will have setting eqivalent to pressure that fully opens bypass so it will run as long as you have any flow demand.
The other (peak pump) will have setting bellow fully closed bypas setting.
Of course, you can alter the pumps every week to have same working hours.

 

[boilerguy]

You could stage the pumps and bypass in the following manner using a end-user pressure reading tied into your PLC:

High-high pressure -- only one pump running, open FCV bypass
High pressure -- one pump running, close bypass
Low pressure -- two pumps running

Of course, you would have to design deadbands around the pressure stages to keep the pumps from switching on and off all the time when close to one of the transition points. Are you planning on having at least one pump run continuously during the day?

 

[vitrom]

Let me put my two cents in the discussion.
I assume that either the pumps are connected to a common manifold, souce the water from the same reservoir and use the same return diverter pipe.
This diverter pipe has some sort of porportional valve that can be controlled by an input signal.
The pumps are either on or off and the they maintain a constant output pressure/flow.
One can monitor both pressure and flow either before and after the diverter pipe.
The requirements are that upstream pressure is maintained at all times during operations so a pressure sensor is needed upstream to monitor and to control the proportional valve. A flow sensor after the proportional valve will determine the requirements of the system and enable starting of the second pump.
Depending on the starting charcteristics of the pumps you may need a pressure relieve circuit if the proportional valve is not able to react fast enough. However if you have soft start on the pumps and proper settings, you may not need it.
This type of system will also allow rotation of pump usage so that you don't have one pump working all the time while the other idles most of the time. This is important to the life of the pumps.
Hope this helps.

Antonio Reis
01 209-834-1900

http://www.vitrom.com

 

[ChasBean1]

Hi Pete. 150 HP pumps? Use speed drives! they will pay you back in seconds (well, not literally).

Control the pressure, not flow. You can't control the system by flow - if you think about it for a minute, you'll see why. What would a sensed reduction in flow cause the pumps to do? Ramp up? So when the system is not demanding any water, the pumps will be at max capacity...

My sticking point & purpose of this post is a VFD. I think this is the perfect (and essential) application for it's use. -Chas