Vertical Turbine Pumps with Variable Frequency Drives 1

[Overbudget]

I vaguely recall that there is something to watch for when using a VFD with a VTP, that is usually not an issue with a volute type pump.

In this application, we want to maintain a constant flow at a variable head. The operating rate is 95 gpm while the BEP for this is at 105 gpm at the rated speed. In our system the VFD will be at 55 hz to develop the max head needed, and 45 hz at the min head needed. I calculated 12 BHP at the max head using the peak effieciency at 60 hz. The motor is 15 HP and it's a 6-stage pump.

As you reduce speed, does the efficiency drop off faster than you expect? Does the BEP shift left? Would we be better off trimming the impellers? Trimming impellers may not be an option on small pumps like this.

We'll be contacting the manufacturer and asking for efficiency curves at lower speeds. They may or may not have this. We also want to verify that there wont be harmonics at lower speeds. I wanted to see if I'm correctly remembering that an issue even exists.

Thanks!

 

[BigInch]

The efficiency will track closely with the affinity relationship for flow with changes in speed.
If you have for example a BEP with 70% efficiency at a flow rate of 2000 gpm on the 1760 rpm curve, at 1500 rpm your flow rate (and new BEP) will be 1500/1760 * 2000 gpm = 1278 gpm, which will also have roughly the same efficiency of 70%. BEP Head reduces according to the affinity relationship for head, ie. by the speeds squared. All efficiency points on the pump curve will track down and to the left in a similar manner. At least that's the idealistic theory of it.

 

[BigInch]

See this conveniently recent post on VT issues with VFDs.
thread407-381065

 

[bimr]

Don't think the OP has stated enough about the application for someone to reach a conclusion.

Is there a minimum flow? What is the application?

Yaskawa makes a quality VFD:

https://www.yaskawa.com/pycprd/look...jmu0KhDpCi3M3i3AXb5LTbm50D8Ksg87LNCzJj6p23n8=

 

[LittleInch]

Err, BI - your calculator seems to be malfunctioning - (1500/1760) X 2000 = 1704 gpm?

If you want a constant flow at variable head, why not use a PD pump of some sort - screw, piston, disc etc?

Any turbine or centrifugal system is really set up for stable head and variable flow....

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

 

[TenPenny]

Check for critical frequencies, that's usually the biggest thing with variable speed VT pumps. Some manufacturers make sure that the pump always runs below 1st critical speed, sometimes you'll find it runs between 1st and 2nd, so you have to know if there are speeds to avoid.

 

[Overbudget]

This is a water supply well, and the pump is actually an 8-stage submersible vertical turbine. For the pump that is proposed, the minimum flow is 45 gpm, we are designing a constant flow of 90 gpm, the BEP is at 105 gpm, and the max flow is 160 gpm.

If the BEP shifts left and down, then at some lower speed, we will be at the BEP for that speed.

I was doing a Q/A review on this project and suggested to the designer to follow up with the manufacturer to verify this applicaiton with the VFD. I suggested that he question the mfr about:

1. using 7 stages instead of 8
2. trimming the impellers (probably not an option on a small pump like this)
3. whether efficiency curves are available for lower speeds.
4. vibration at lower speeds

Thanks for all the replies.

 

[Artisi]

I see no problem with VFD other than critical speed as pointed out by TenPenny, but if you make the manufacturer aware of your intentions they can design / select the correct shaft dia. bearing spacing etc to accommodate your application.

I can never understand the hang-up on pump efficiency when faced with a difficult application - if the only way to achieve the various duties is running into inefficiencies then that has to be accepted. Ok - you need to select equipment that gives you the best eff. that can be achieved but certainly not at the expense of an incorrect pump selection just because it was / is / might be more efficient.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[bimr]

If this pump is transferring to a water storage tank, there is no need for a VFD. The operators tend to operate these pumps with a single continuous daily fill cycle. The last plant operation that I observed, the operator ran the pump for 8 hours per day which equates to his total daily water usage.

One problem that you have not addressed is the variability of the water elevation. The water elevation can only be determined through a long term pumping test. It is not unusual to find that the water elevation is different than when the well was installed.

There is very limited opportunity for power saving with this application. Since you are pumping into a water distribution system, the head variability will only be the working volume in your storage tank.

The reason for installing a VFD would be that you may want to vary the flow out of the well. Perhaps the storage tank size is inadequate.

I don't see any problems if you want a VFD. We installed a Yaskawa VFD on a 1,000 gpm well with 1,100 feet of head without any problems.

 

[DubMac]

As per several of the above, crossing through a rotor critical speed during operation is the issue you are thinking of. If you are determined to keep the VFD just make sure to lock out operation close to any critical speeds. You must get these speeds from the manufacturer.