VSD Pumps performance Curves

[Sawsan311]

Dear All,

I have noticed that various pump vendors provide their performance curve for VSD pumps based on fixing the rated head and flow at THE MAXIMUM SPEED and without considering any further speed curve for further modulation beyond the rated speed.

If we compare the case to centrifugal compressors, the rated guaranteed point can be at 90% speed while compressor is capable of absorbing the variations in system's head and flow with the variable speed ramping up to 105% speed.

Can I understand why pumps tend to be designed differently and is there a possibility to specify the vendor to provide a higher speed range within the selected impeller design of the pump and the selected gearbox.

Regards,

 

[LittleInch]

Write your data sheet and specification in the right way and you'll get what you asked for. Usually.

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

 

[Sawsan311]

Dear Mr. LittleInch,my post is not enquiring about any specific design or data sheet. Hence, I believe there is absolutely no need to share such information.
I am simply enquiring about the maximum speed of a VSD pump and whether it is usually dictated by the rated conditions.

Regards,

 

[LittleInch]

Yes.

I don't know why different vendors do it differently, but if you state a maximum speed then it would seem reasonable to base your design on that maximum speed, not some mysterious other speed.

if you want something different then specify it via the data sheet and the specification.

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

 

[EdStainless]

We built motor/pump systems that ran on 60Hz nominal. We would test up to 120Hz for engineering work. Our units shipped with the limits of 45-70Hz and our published curves covered this range. This could be adjusted if needed. If a customer asked we would test outside of this range

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[georgeverghese]

For centrifugal compressors, usually the design operating point is to be achieved at 100% speed, not at 90%.

The risk of a machine going overspeed is higher on gas turbine or steam turbine or gas engine drives or a hydraulic coupling like Voith. While speed control is much better with electric motors with frequency modulation (VFD), with DC drives being even better than AC drives. So the requirement for having to assess the impact of the driven machine (compressor or pump) going to 105% speed is clearly necessary on GT, GE, ST drives. While this may be optional for electric - VFD drives due to better speed control at the VFD. However, would agree that generally, ALL VSD units should assess the process side impact of overspeed at 105% speed, regardless of the speed control mechanism or driver.

 

[LittleInch]

I agree with george and the original OP is very broad and vague.

Various pump vendors?? Electric drive, size range etc?
Does this compare to "centrifugal compressors"??

Who knows?

Larger compressors are often driven with turbines of different sorts. Pumps tend to be lower power usage so tend to be electric motor driven.

But the opposite can apply so you need to compare apples with apples.

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

 

[Sawsan311]

Dear Mr. Little Inch, it seems you misunderstood the original post. In fact your second reply was fulfilling my query about why vendors set the rated conditions at maximum speed and don't provide speed curves beyond.

However. I Thank you all for your great feedback.

I would like to ask about the typical range of the variable speed drive speed range for centrifugal pump, what actually dictates the minimum speed limit which can practically cover the backpressure and capacity requirements. Is there a typical rough range of speed.

Additionally, do you foresee that for larger pumps with larger impellers , the variable speed range would be narrower in comparison to smaller pumps.

Regards,

 

[LittleInch]

Typical I've seen is whatever the speed range is for 45 to 70 Htz. So depends on the number of poles on the motor. A 2pole motor would then be 2,700 rpm to 4,200 rpm. I've seen as low as 40htz.

You need to remember than head is proportional to speed^2, so this actually gives you a large change in head of between 40% to 100%. Actual flow will then depend on your system curve.

If you need a bigger range than that then you really should be looking at different sized pumps.

No reason to change it for larger pumps.

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

 

[danschwind]

Regarding speed, one must pay attention to the peripheral speed of the impeller and the mechanical limits of the impeller material before attempting to increase the speed. To lower the speed, theoretically there is a cooling limit for the motors, even though I've never seen a set-on-stone limit. There are motors that are more suited for this speed reduction than others too, and some motors may also be supplied with an external independent cooling system that is not coupled to the motor rotation, so you could theoretically lower the speed indefinitely.

I'll try to give my design approach regarding your original question (if I understood it correctly). For VSD-controlled pumps I 'usually' (though not a rule) define my impeller diameter for 90% of the rated speed of the pump (here in my country, usually 1750 RPM or 3500 RPM for "mundane" pumps). I've never, to this day, expected to control my pump to a speed higher than the rated one.
I personally tend to select my own pumps due to a problem with the local market - If I wait a pump vendor to give me a solution for my rated conditions I will never be able to issue a project in the correct deadline. So, I go directly to the pump curve catalogue and select it directly from a couple of manufacturers. This is only valid however for "mundane" pumps.

So, if you want your rated conditions to be selected @ 90% of the pump nominal speed, and you are dealing directly with the pump vendors, I think just specifying this in the datasheet shall be enough for them to comply with your requirements. Remember - most pump vendors have no idea of the overall process this pump is inserted in.



Daniel
Rio de Janeiro - Brazil

 

[EdStainless]

I have VFDs that will drive from 0.1Hz to 120Hz, but I don't have pumps that will work over this range.
Bearings, seals, shaft and flow harmonics (critical freq), peripheral velocity, and few other factors limit the practical range of centrifugal pumps.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[georgeverghese]

From what I've gathered from electrical engineers here on E-Tips, pd pumps with electric motors (constant torque machines) with VFD can only turn down to about 50% of rated speed; below that an auxiliary cooling fan for the EM is required. There is no practical turndown limit on centrifugal pumps (variable torque) with EM-VFD drives.

 

[Sawsan311]

Thank you Mr. georgeverghese & All for your great support.

We can agree that

- rated conditions are usually set at 100% speed and any requirement to set it at 90% of the speed needs to be spelled out in project's specifications.
- the limitation in minimum speed for VFD is dictated by the requirement for additional cooling of the EM. and that:Minimum VSD electric motor speed limit is 50%. is this for a PD pump ? what about centrifugal?

I also would like to qnuire:

- if we specify the rated conditions at 90% speed, we need to be super cautious to ensure the design is covering the end of curve operation at maximum speed with respect to shut off as well as motor power.
- I would like to understand the rational behind the need for cooling when going to low speeds, is it attributed to the potential run out conditions at lower speeds i.e. shorter pump curves.

Thanks

Regards,

 

[TiCl4]

FYI, inverter rated motors typically come with a design to withstand a 10:1 turndown ratio. I have a couple of PD pumps that operate most often at a 15-20% nominal motor speed - there are no issues with overheating.

 

[Sawsan311]

Thanks TiCl4, I believe that the high turndown ratio is ok from motor perspective, but it may not be suitable from VFD control application as well as pump impeller design, would there be any limitation preventing from achieve such large turndown ratio. Minimum continuous stable flow at minimum speed is also a limitation i.e. lower MCSF at lower speed in comparison with high speeds.

kindly please confirm.

 

[TiCl4]

Sawsan311,

I believe VFD application to centrifugal pumps is pump-specific. The pump curves I have seen that include various motor speeds stop somewhere around 30-50% speed. So pump stability is definitely more restricting than motor limitations. However, perhaps not all pumps can even go that low - it’s always a conversation with the OEM or distributor if you intend to lower pump speed and multiple pumps at different speeds are not available.

 

[EdStainless]

Many of our units that required wide range had specific speed bands that were locked out because of critical freq issues. You could run above or below them.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Sawsan311]

Dear All,

Thanks for all your replies,

can you elaborate more why the motor tends to experience limitations when the speed is ramping down and if a cooling system is not in place. Referring to MR.
danschwind 's post:
''Regarding speed, one must pay attention to the peripheral speed of the impeller and the mechanical limits of the impeller material before attempting to increase the speed. To lower the speed, theoretically there is a cooling limit for the motors, even though I've never seen a set-on-stone limit. There are motors that are more suited for this speed reduction than others too, and some motors may also be supplied with an external independent cooling system that is not coupled to the motor rotation, so you could theoretically lower the speed indefinitely.''

Can I understand from electrical perspective what happens when the motor slows down, is it attributed also to having a shorter pump curve with a narrower stonewall limit and motor overheating or is coupled with electrical aspects although any gearbox as stated above can deliver sufficiently low frequency level?

Regards,

 

[danschwind]

Sawsan,

What I described is something that I personally find to be kind of confusing, so don't take my words for granted.

I've seen multiple different people claiming that there should be a limit to reduce a motor speed due to cooling (the fan rotates at the same speed as the motor usually), both for PD pumps and also for centrifugal pumps. But, there is also multiple different people that claimed this limitation is not applicable to centrifugal pumps due to the variable torque.

Consulting local motor supplier engineers was the same - some say it has limits, others don't.

Hence, due to my own ignorance on the subject of electric motors, I tend not to reduce the motor speed too much. This is something that do bother me but fortunately most pumps I design I can survive with a tight control range (if I use VFD).

But please, do not take that my opinion or what I do to be the truth here. As some others have pointed out, there are EE here in eng-tips that may have a way stronger and more correct opinion than I do. Let's hope they can contribute to the topic and maybe I can also learn a thing or two regarding this.

Daniel
Rio de Janeiro - Brazil

 

[EdStainless]

We built submersible motors and pumps. The lower speed was limited by flow past the motor for cooling.
We did a lot of work on this because it is a serious issue.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed