Operating Pumps with Variable Speed 1

[MarineElect]

Hi,

I am looking at an application from an energy saving perspective, where there are three circulating pumps in a manufacturing cooling process working in parallel with one another to provide the required amount of flow to the cooling application. All pumps are on/off control, and in most cases the flow required is just above the pumping capacity of one pump, so two pumps need to be left on at all times.

I am looking at the idea of operating the second pump using a Variable speed drive to control the speed of it, and the VSD controlled by a differential pressure input transducer to cause the VSD to operate at the required speed to maintain the required differential pressure across the pumping operation, and thus the load application.
thus only one pump would be run flat out, with considerable energy sagings on the other.

Has anyone here ever tried this out, or can anyone see a reason why it might not work. Looking at the pump power curves I would see this as giving a constant head pressure across both pumps and the relative power input determining the flow through each pump, with the first pump pumping close to max capacity all the time.

Am i missing something that might cause this to operate incorrectly.

Thanks for any assistance rendered.

Marineelect.

 

[kcj]

Are these centrifugal pumps?
Head pressure is highly dependant on pump rpm, need to verify where the pump would be at on it's curve.

kcj

 

[MarineElect]

Thanks for your response.
This point is what I am really unsure of.
My understanding of this is that the differential pressure transducer input will regulate the flow to maintain the required pressure head and so as long as the system head is set at a point between the run out point and the shut off head the system will operate correctly.

Am i correct in saying that because the pump is not operating at a constant speed, one pump curve is not adequate to explain the behaviour of the pump. My understanding of this is that the operating point is at the intersection between the system curve and the pressure head line set by the differential transducer. Thus for variations (increase) in system resistance to flow, the pressure head will tend to rise and the pump speed will reduce flow proportionally to bring pressure back to equilibrium, the opposite happening in the event of a fall in resistance to flow.

Am I correct in my assumptions as I am by no means familiar with pump performance curves.

 

[kcj]

I am rushed for time before leaving on a business trip.

Correct, one pump curve is not enough, there will be a different curve for each rpm.

With dead head psi on Y axis, full flow max output on X axis, the pump curve is a slope from deadhead psi, 0 flow,at upper left, to the lower right full flow at 0 output pressure. Pump psi is dependant on rpm, I think to a square term, so lowering the rpm drops the entire curv down, and changes it shape.

Two pump curves in parallel would I think put the first pump way on lower right, low pressure at high output, and the second pump at upper left, low output, high pressure, until they match the system demand curve. Sum of flows = system demand at that pressre, etc. I would watch that the VFD pump curve maynot have enough psi at the lowered rpm.

I think it needs some detailed examination by the pump mfr or reps. Hpefully, all pumps are by same mfr, else no one wants to get involved in the design.

need to run. later. kcj

 

[ScottyUK]

Probably better off posting in forum407. Ask one of the moderators via the red flag if they could move the thread as-is.


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[meintsi]

This may help ....

http://www.gambica.org.uk/pdfs/VSD_Pumps.pdf#search="pump efficiency speed variation"

Remember...
[navy]"If you don't use your head,[/navy]
[navy]your going to have to use your feet."[/navy]

 

[MarineElect]

Meintsi,

This looks great.

Thanks for all the assistance.


Marineelect.

 

[whieber]

I have setup several Goulds PumpSmart devices and they do exactly what your looking for(Chaining multiple pumps to VFD's). One of the biggest advantages is cost cutting in energy consuption and having complete control of the application as well as a bunch of pump protection logics to prevent catastrophic failures. These can be chained together or run independantly. You can equip them with various transmitters such as pressure, flow, temp, vibration,etc. If you want.

Here is a link to the brochure on thier web site. If you have a local Goulds representative in your area you should give them a call they would be your best bet at obtaining one.

http://www.pumpsmart.com/itemfiles/bpsfamily.pdf

Walter Hieber
PIB, INC. -

http://www.thepumpbox.com