Estimation variable speed. Savings 1

[PalomaP]

I would estimate savings by installing a variable speed in a centrifugal pump of 355 kW, 1480 rpm. Could anyone give me an idea about how much a variable speed regulator would cost?
Thanks in advance

 

[BigInch]

I don't know the answer to the cost question, but I will ask two other questions first.

Do you have a reasonably constant flowrate (within 60% to 110% of BEP flowrate) that isn't practical to smooth to a constant flow by filling and withdrawing from a storage tank?

Do you have a high static head at low flows (Head >= 50% of the head needed at the BEP flowrate)?

If you answered yes to either one, I doubt you need a VSD.

http://virtualpipeline.spaces.msn.com

 

[PalomaP]

The flow rate changes from 2200 m3/h @ 105 BEP, to 1680 m3/h @ 90% BEP
The total head at 2200 m3/h is 50 m (Hst= 40m)
The total head at 168 m3/h is 46 m (Hst=40 m)

Is not worth it to use the vsd? In which cases it would be a good way of flow regulation?
thanks again

 

[BigInch]

Very quickly and very roughly, it doesn't look like a great prospect for VSD, since the reduction in head to 40 m is a very small % of the 50 m needed at max flow. If you turned down the pump speed proportionally from 2200 reach 1680 m3/h that would be an RPM of 76% of BEP RPM, so discharge head would fall to 50 * 0.76^2 = 30 meters, so you won't be able to reach your 1680 m3/hr target flow with a system curve that requires 46 meters.

To generate the 40 m head, you'd have to run at about 90% speed, giving a minimum flowrate of around 2000 m3/h from the pump. If that matches your system curve and if you could plan on doing that most all of the time, maybe you could reach a 10% on your power bill.

Looks like good old fashioned flow control valves, but 10% is 10%, so if that saving potential looks attractive, do a more in depth study that includes estimating the running time at various flowrates to get a better idea of the actual expected power cost and decide based on that result.

http://virtualpipeline.spaces.msn.com

 

[ozmosis]

To show savings with a VSD you need to compare it to something otherwise your idea of "savings" are meaningless.
Taking the basic cost of a VSD is only part of the calcualtion to determine a return on your investment (assuming you have answered the first part of my post). You need to take into consideration the following:
1) Purchase cost of VSD
2) Installation costs including labour and additional materials e.g. cabling (shielded from VSD to motor), maybe specific drive enclosure for the environment, EMC filters, maybe harmonic filters or reactors, control strategy (you will need some sort of intelligence to 'tell' the VSD what sort of speed the pump should be running at), plant downtime if the pump is existing etc.
3) If it is an existing pump, then you need to check the motor is suitable for inverter use. If not, then the cost of a suitable motor may be required or some sort of sinewave filter on the output to ensure the PWM waveform is not detrimental to the motor life.
4) I've already mentioned the issue of harmonics but 355kW is quite a large motor/drive. If your power system is not large enough then the harmonic distortion could be an issue. It needs investigating and maybe analysing. .I know some water companies in the UK would demand a 12-pulse configured drive or other similar type to ensure harmonic distortion is kept to a minimum. This is expensive.
All these cost money and need to be included in any calculation you do to determine any potential savings.
But as I say at the start, you will only see 'savings' when you compare it to another method of flow control.

 

[BigInch]

Sed2,

For estimates, I have the habit of using about 10% of pump cost for providing a VSD option. Do you think that's a reasonable number?

http://virtualpipeline.spaces.msn.com

 

[quark]

Neglecting (for the sake of calculation) efficiency at the reduced speed, if you require 46meters at 1680m[sup]3[/sup]/hr, then you need to run the pump at (46/50)[sup]1/2[/sup] times the original speed, i.e 96% and your flowrate will be not less than 2110m[sup]3[/sup]/hr. You can manage power savings upto 12% by this. Then you have the efficiency at the reduced speed that reduces the power savings.

Check the power at partial closed condition of the valve corresponding to 1680m[sup]3[/sup]/hr flowrate and see if you have any significant difference in the savings.

I, generally, consider 2 years as payback period for viability.

 

[ozmosis]

BigInch
To be honest, I couldn't be sure on that figure. It wouldn't be linear is my guess due to the 'additional' items needed as drives get bigger.
There's an interesting article published here concerning Pump life cycle costs and it includes some estimates on different control strategies:

http://www1.eere.energy.gov/industry/bestpractices/pdfs/pumplcc_1001.pdf

 

[BigInch]

Ya not much is really ever linear when you get down to it ...

pdf looks usefull.

http://virtualpipeline.spaces.msn.com

 

[Valvecrazy]

I agree you have to have something to compare with. If that pump has any natural drop in horse power when the flow is reduced, I doubt the VSD will save any energy over simply throttling with a valve.

 

[ozmosis]

Valvecrazy
You will save energy compared to throttling a valve but the key is whether it will pay back your investment in a reasonable time.
2 years ROI seems to be the norm now. I remember when 5 years was acceptable....

 

[Valvecrazy]

I agree the variable speed should save a little energy but, sometimes the difference between VSD and throttling is very small, and makes a payback take decades.

 

[ccfowler]

In addition to the above well stated considerations, it is also wise to take a close look at the inherent parasitic losses of the adjustable speed drives that you are considering. (There is a reason that the cooling is needed for the drive and additional cooling may be needed for the motor.) Since you are already looking are fairly modest potential savings, these parasitic losses will further erode the marginal savings that you are hoping to realize. In this case, these parasitic losses may be a disturbingly large percentage of the potential energy savings.

 

[BigInch]

Normally (lacking specific information) I throw in a 4% efficiency loss when evaluating VFDs options.

http://virtualpipeline.spaces.msn.com