Frequency dependant efficiency of vfd and e-motor wanted for analysis

[Yobbo]

Hi,
I have been reading a interesting thread about THE usefullnes of THE application of vfd to save energy. What I picked up mainly from this thread is that à vfd is not à miracle cure. THe efficiency of a vfd seems to decrease dramatically at about 40% of THE nominal frequency. Additional to this e-motors show à similar trend with its efficiency. I have been rooming THE internet for typical curves that show these trends. I am not an electrical engineer and i do not know whether THE shown trend is à combination of vfd and e-motor or just of à vfd. I find Some leading books on pumps very limited in their clarification of speed controlled pumps. I would like to get à hold of the complete picture and understand how I can asses and calculate myself what controle method would be best in à given situation. What I see in my company is that vfd are still regarded às à miracle cure in all cases. Suppliers of vfd find à fertile playground for their products. I would appreciate if somebody in this forum could give me some tips about where to gather my information in order to be able to become à good opponent to just blindly assuming that vfd's are just THE thing to apply.

Karel Postulart, The Netherlands
Nuon Power Generation

 

[GPRnD]

Sorry I don't have any more specific information than what you probably have already found.

I suppose my question back to you is "does it really matter ?"

If you are running your VFD at 40% of nominal frequency, the pump is producing only 16% of its rated head and consuming only 6% of its rated power. I suppose there might be some applications that require that degree of turndown capability but I can't think of too many.

So even if the VFD is horribly inefficient, the power consumed by the pump is tiny.

If you were achieving that reduction in head the conventional way (using a throttle valve), your efficiencies are going to be far worse than with a VFD.

I agree that VFDs are not a universal solution (for example they are not so useful in systems where the required flow varies a lot but the head does not), but careful analysis of the application should determine whether they are a good solution.

The Hydraulic Institute has a decent book on the subject of energy saving called "Optimizing Pumping Systems".

 

[Pumpsonly]

Yobbo
I am glad that you have not been taken in by all those sales talks of miracle cure energy saving wonders of VFD on pump application by VFD vendors.They only show you the reduction in pump kW when you reduce the pump RPM. They conveniently hide the kW lost in the VFD in the form of heat dissipated to the surrounding and the fan kW required to provide the cooling.The maintenance cost of the VFD itself are seldom disclosed.

As pointed out by bradshsi, every case need careful analysis to determine if there is advantage in using VFD.

 

[BigInch]

Not really every cases needs such a carefull analysis. Some generalities can be drawn.

1.) That the process must have variable flow and that the flowrate cannot be economically "smoothed" to be closer to a constant average rate, for example by use of tanks to store overproduction, or to provide more product during times of underproduction.

2.) That the piping system must have little or no static head requirements, ie. making required head very much proportional to k * Qⁿ, where k is a constant, Q is flowrate.

3.) That the process must not only require a considreable amount of time running very near to one constant average design flowrate, but also that the any variable flowrate deviations needed by the process also remain within a range of 70% to 85% of the same "design flowrate". If too much time is spent lower than 70%, compounded efficiency losses become too severe to run economically, ie. a smaller pump might be considered, and too much time above 85%, but less than 95% would suggest that the design flowrate be lowered slightly, such that the design flowrate fall within 95 to 110% of the new design rate.

AS you can see from the above conditions required to ensure that pumped systems remain economical when using VFD controls, VFD controls will not be very useful for typical pipeline, boiler feedwater and many types of chemical processing applications where minimum reactor pressures must be maintained. VFDs would generally be much more suitable to hydraulic power apps where wide drive speeds and consequential flowrate ranges are required.

Let your acquaintances be many, but your advisors one in a thousand’ ... Book of Ecclesiasticus

 

[Yobbo]

Dear gentlemen (assuming that no woman is hidden behind one of your acronyms).
First of all my thanks for your input. Especially Biginch with his elaborate answer. Point is that I have to substantiate my "defence" with Pure facts. Therefore I think I need chracteristics about how motorefficiency ( Total effect of powerfactor cos fi as well as heat losses) as à function of THE frequency of THE voltage supplied by à vfd as well às the internal efficiency of à vfd ( including energy loss through heat loss or extra auxiliary power to cool THE vfd. Are these data that I should be able to get from à supplier? Apart from this I can calculate THE feasibility to install à smaller pump parallel to bigger ones in order to cover low demand periods instead of letting a Huge pump rotate with a negligable speed. Any hints on how to get THE data I mentioned before would be gladly accepted.

Karel Postulart, The Netherlands
Nuon Power Generation

 

[electricpete]

Why do you capitalize the word "THE" every time you write it?

=====================================
(2B)+(2B)' ?

 

[BigInch]

If you find that information anywhere at all it is more likely to be expressed by the VFD's overall efficiency as a function of its power load, where power load is expressed as a percent of the VFD's rated load.

Let your acquaintances be many, but your advisors one in a thousand’ ... Book of Ecclesiasticus

 

[Pumpsonly]

If you need pure facts, the only way is to do a detail analysis of the actual application.

 

[SNORGY]

For centrifugal pumps, I think it also depends on the shape of the pump characteristic curve near its operating point / over its operating range. If the characteristic is somewhat flat, then in my mind there is little to be gained from a VFD because a large swing in flow can be achieved with a small variation in pump speed. At that point, your system head curve is probably doing more controlling of the pump than is the VFD.

For PD pumps, VFDs make sense to me.

Regards,

SNORGY.

 

[BigInch]

The system head flow curve is always controlling "one side" of the dynamics while the pump head flow curve is always controlling the "other side" of the dynamics. A VFD on the pump doesn't change that, it only allows change of the operating point by making a change in the pump speed.

If the system curve also happens to be flat, or more precisely stated, if the system curve is parallel to the pump curve at the point of interest, it is unlikely that anything is going to control it.

Let your acquaintances be many, but your advisors one in a thousand’ ... Book of Ecclesiasticus

 

[Yobbo]

Just for the onces that are a bit irritated by the capitalized THE in my messages I want to clarify that I typed those messages on an I-Pad with some peculiar language setting. It automatically changes the into THE for some weird reason. At first I tried to correct it, but than it got so time consuming that I stopped doing that. Bear with me a bit. Thank you.

Karel Postulart, The Netherlands
Nuon Power Generation

 

[BigInch]

That could work nicely with "THE" Netherlands.

Let your acquaintances be many, but your advisors one in a thousand’ ... Book of Ecclesiasticus

 

[Yobbo]

Yes it would!

Karel Postulart, The Netherlands
Nuon Power Generation