Fuel pump noise

[malz]

I work in an office building, where we have a 2000KW power plant on the roof. The fuel pumps and cells are in the basement. Pumps are of a positive displacement variety. For a long time, whenever the pump set is on, one can walk literally anywhere in the building and hear the "whine" of the pump being on line. This is annoying to say the least, and this allows the staff to operate the plant only in the late evening. We had changed out the flexible connections from the pump set to the riser piping in hopes of eliminating the problem, which it hasn't. Now someone suggested to put in a VFD drive to eliminate the noise (in my opinion, the noise is a harmonic resonance, and has more to do with the anchoring arrangement with no sound attenuation, but thats only me). 1- Will this idea work, or is this guy just trying to sell me a VFD? 2-I think adding a VFD to a setpoint below the spec of the system will effect capacity of fuel, no? 3- Is there something I am overlooking? I did check out the springs on the inertia block (steel pan) to make sure there were no problems with suspension. Any help would be appreciated. Thanks- malz

 

[GregLocock]

It sounds like you are on the right track. The possibilities are: conduction via the motor mounts, conduction via the pipes and other connections, electrical noise , acoustic excitation of the building or conduction via the liquid itself.

You've looked at 1&2, 3&4 are less likely, so you could try an air reservoir type damper in the fluid line (anti water-hammer). This may need to be tuned to the whine.

A VFD may allow you to find a better operating speed, but it may just change the pitch of the whine.






Cheers

Greg Locock

 

[butelja]

Have you taken a look at how the pipes are supported in the pipe chase that runs from the basement to the roof? Ideally, these should have rubber insert isolated clamps rather than rigid metal to metal contact. Also, does the conduit for the motor wiring have a flexible section adjacent to the motor?

 

[malz]

The motor connection is greenfield, so it is flexible. My feeling about the rubber boots around the anchors (sound attenuation on the pipe riser) probably does not exist, & I think this is the problem. I have my doubts about a VFD, although it may work. I just think I am missing something (probably obvious). I have checked the inetia assembly for bad springs, and looked for loose components, but have found all within spec.
I really appreciate the help.

 

[Nige]

Yes it would benefit from VFD control. Depending on the model, a low noise setting is available that will change the switching pattern to 4kHz making the motor run considerably quieter. The resonance would also be reduced as the unit would be 'softstarted. Make sure you have an inverter which also skips 'critical frequencies'or a certain band when resonance occurs, i.e 40-45Hz on ramp up.
Hope it works for you!

 

[malz]

We have contracted a vendor, who will install a VFD module on the pump set in question temporarily, to see if this is a feasable idea. I will post result after testing (if interested.)

 

[ANTOINE]

Have you tried to add a quarterwavelenght resonator at the pump discharge ? For a single pure frequency this can be very efficient.
rgds.

 

[butelja]

Antoine has a good suggestion for a constant speed pump. However, if a VFD is installed then forget it.

 

[malz]

I will pose this suggestion to my team. Thank you very much.

 

[malz]

Message to butelja:
Please provide details regarding the quarterwavelength resonator. Is this an elecrically powered device? Where would one install it for greatest impact? The name indicates it limits frequency resonance, but we have no indication as to how this is accomplished. Would this be a cost effective answer as it relates to a VFD installation?
Please advise, if you can. Again, I appreciate the assistance, and all the info!!
malz

 

[butelja]

Here is a technical paper I found by doing a google web search using the terms "side branch resonator".

http://fluid.power.net/techbriefs/papers/proc_mikota.pdf

. It gives a brief description of the theory of several types of devices, of which the 1/4 wavelength resonator is the simplest.

Bear in mind that all the devices outlined in this paper work over a fairly narrow range of frequencies, so some trial and error will likely be involved in implementation. The less accurate your knowledge of the frequency(ies), the more error you will have with your trials.

 

[GregLocock]

If you are game to try passive silencing techniques like a quarter wave tuner, then you might want to consider its more complex and interesting brother, the Helmholtz resonator. Don't bother doing a web search (well, you could) as you won't find much useful info.

The advantage of the Helmholtz resonator is that the attenuation is pretty reliably tunable, and you can get a lot of damping from a small device


fr=c/2/pi*sqrt(Ac/lc/V)

Y=f/fr-fr/f

X=(sqrt(Ac*V/lc))/2/Ap/Y


Transmission Loss (dB)=10*log10(mod(1+X^2))


Ac = Cross-Sectional Area of Connector
Ap = Cross-Sectional Area of Main Tube
lc = Length of Connector
V = Resonator Chamber Volume
fr = Resonant Frequency
c = Speed of Sound

Note that the chamber's linear dimensions should be fairly small, less than wavelength/10.

Beranek (Noise and Vibration Control) has a good discussion of these things.

Cheers

Greg Locock