Switchgear Vibration 1

[KenAlmon]

I am at a new construction 500 MW power plant where the various pumps, fans and other rotating equipment are going through their initial runs.
The 13.8 kV and 6.9 kV switchgear is located on the 9 m elevation on an 8" (I believe) reinforced concrete slab.
Several (but not all) switchgear units have a distinct audible vibration that occurs when certain pieces of rotating equipment operate.
We have not taken vibration readings off of the gear as yet, but that is scheduled in the near term.

I have several questions:
1. Where should measurements be taken (presumably 3 planes) but on the outside of the gear or on the CB or contactor itself? All gear is live at present.
2. Is acceleration, velocity or displacement the key measurement (or some combination)?
3. What value might be considered to be significant?
4. What might be the areas of concern if vibration is excessive? Hard connections like busbar joints versus spring connections etc.
It is believed that this may go away once all the fluids are in the pipes, vessels etc. However, if it persists, I need to know whether this requires additional investigation. If the values of vibration are excessive in the gear, I suppose one may have to be concerned with loosening bolts/nuts on the busses, on the CT circuits etc. Spring connections like where the breaker tulips connect with the bus may be less of an issue, but then again perhaps fatigue may present an issue.
Any comments/suggestions welcome.
Thanks

 

[Zogzog]

Have all of the shipping accesories been removed?

Have you done acceptance testing of the gear?

 

[KenAlmon]

Zogzog, yes all shipping appurtances have been removed. Vendor was on site to accept installation and agree with power-up.

 

[Zogzog]

Vendor being on site and doing acceptance testing are not necessarily the same thing.

 

[TurbineGen]

KenAlmon - My main concern would be the interrupters. Depending on the amount of vibration, this could wear out the contacts and lead to premature failure. Bus bars can be monitored for contact resistance by taking a temperature reading from a thermal camera or even a distance thermometer. Connecting with braided straps might reduce vibrations as well. You might also want to consider mounting the switchgear on rubber dampers. We do this on our mobile substations and it works well.

Are these vacuum switches? Solenoid or motor operated?



------------------------------------------------------------------------
If it is broken, fix it. If it isn't broken, I'll soon fix that.

 

[rasevskii]

KenAlmon:

What are the currents involved? 100's or 1000-2000A? It can well be a magnetic vibration due the outgoing conductors or busbars inducing eddy currents in the steel cladding. Not an uncommon problem. There was a thread here about steel gland plates and single-phase cables recently, a related situation.
The three conductors (or bus bars) being in flat formation naturally causes induction in the steel as the magnetic field is unsymmetrical. If there are loose rivets or bolts there will be vibration as well as heating. Is there any heating? Any loose or badly fitted steel plates or panels may well vibrate.

Usually these sort of problems occur at, say, 500A or around 1000A and upwards. The switchgear vendor has to investigate thís and decide on a solution. Using non-magnetic plates or panels can be necessary if no other solution can be found

rasevskii

 

[KenAlmon]

Thanks for the responses. The 13.8 kV gear is fed by a 72 MVA tx so the currents are significant. Three of the loads are 11 MW pump motors. Again big currents. The problem goes away when the various pumps are shut off. All cables are 3 phase cables. Vacuum bottles. motor operated, spring release.

 

[electricpete]

If you monitor the frequency of the vibration you can get a clue as to the source. Vibration from currents in the switchgear would normally be 120hz and harmonics thereof. Vibration from running machinery will often be running speed and harmonics thereof.

My vote for the parameter of interest would be velocity. Maximum stresses in equipment excited at a frequency far above resonance would often be related to acceleration, excited at a frequency far below resonance often related to displacement, at resonance often related to velocity. (In fact if you consider a simply supported steel beam vibrating at resonance, I can tell you the maximum stress knowing only the material properties and the velocity, irrespective of the beam dimensions and mode number...non-intuitive but true and proveable). The at-resonance tend to be the most destructive, so my vote is velocity. There are similar reaons that velocity has evolved as a general first-cut indicator of severity for rotating equipment housing vibration (although acceleration may be better for detecting bearing defects and impacting, and displacement is clearly better when examining relative shaft position by prox probes).

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

 

[KenAlmon]

Thanks Pete for the info on vibration. Simple info, but very valuable. I will provide data on the vibration measurements as soon as we get them.

 

[PHovnanian]

Measure the vibration frequency as well as velocity. Actually, you should plot v vs f. But once you've got that, you can calculate displacement and acceleration.

As far as a remedy goes, I'd take this data and look at fixing the fans/pumps/motors. There are lots of things in your plant that vibration can damage. And while your concern might be for the switchgear, all systems will benefit from eliminating the source.

If the motors are equipped with VFDs, it might be possible to exclude resonant frequencies of surrounding structures by programming them to pass through speeds that excite them.

 

[KenAlmon]

Thanks PHovnanian for your suggestion. There are no VFDs in the plant save some small (less than 50 hp ones).