Loud Transformer and Harmonic Mitigation

[BigJohn1]

We've got an unusual problem: 75kVA 480Δ/208Y transformer with a "noise issue". It's within NEMA decibel specs for a transformer that size, but the customer still doesn't like the sound because they work around it.

We actually did a load test on a purely resistive load, and it was nice and quiet (so it's none of the usual suspects like vibration dampeners or mounting) but it does hum a bit under normal building load.

When on building load the phase currents are reasonably balanced, but the THD is high at about 17% with pronounced 3rd, 5th, and 7th order harmonics. This is to be expected because it's running a mostly discharge lighting with electronic ballasts. There is not excessive heating. It appears the only side effect is making the transformer slightly louder than the customer wants to hear.

What would you do to mitigate this?

The talk is of putting line reactors ahead of the secondary. I've never seen or heard of this being done for a power transformer, and while it may actually take care of some of the harmonics, it just seems a little odd to use a big inductor to try and control harmonics in a big inductor.

Thoughts?

 

[Kiribanda]

BigJohn1,
Is it not a K rated transformer?
Have you taken a harmonic spectrum of the sec load currents & voltages?

 

[dpc]

17% Voltage distortion is quite a bit. I'm not optimistic about reducing the noise. The fact that it is quieter on resistive load doesn't mean it wouldn't benefit from better noise isolation. It just means the transformer makes more noise with non-linear loads. Improved noise-dampening mounting is worth trying, IMO. Is is sitting on the floor?

 

[BigJohn1]

No, it's not K-rated. I brought it up, but nobody seems sure that would make a difference (myself included): Would we actually expect less sound production from K-rated? I thought they were mostly designed to lessen heating.

DPC, I wasn't clear: 17% current THD. There is actually almost no voltage distortion. This transformer is ceiling suspended. I suppose we could try spring-hangers on the rods, but I was wondering if trying to reduce harmonics wouldn't be the better solution, and if so, how?

 

[dpc]

I don't think trying to reduce the harmonics will be cost-effective, especially since you don't know what level must be achieved to get the noise to an acceptable level. These things are highly subjective.

I've been down this road a couple of times in the past. Once the occupants get fixated on the transformer hum, they have a hard time not hearing it. In the end, the transformers were re-located. You might want to start thinking along those lines. You also might try switching loads off to see what the effect is. If the lighting ballasts are causing a lot of the noise, you could consider replacing those.

 

[MikeHalloran]

I used to build medical lab equipment. The customers tend to be fussy.

We had some success with immobilizing very small ( ~100VA ) transformers' laminations by pouring on a low viscosity self wicking Loctite normally used for sealing castings, and then spraying the surface with a Loctite catalyst to kick it. Eventually we talked the transformer manufacturer into vacuum impregnating his product with something similar.

For much larger transformers, that may not be effective, or practical, and for a quantity of one used unit it may not be possible.

-----

Spring isolators in the hanger rods will only reduce structureborne noise. At powerline frequencies, rubber washers might be as effective, and easier to try for fun.

If the noise is radiating from the shell, then covering the shell with a mass-loaded foam would help, but would interfere with cooling, because the stuff is excellent thermal insulation, and not fire-resistant, which would probably upset various regulatory folks.

----

Is there room to put in a false ceiling and still maintain adequate spacing and ventilation?



Mike Halloran
Pembroke Pines, FL, USA

 

[magoo2]

Have you checked the voltage at the transformer? If it's higher than normal, you may be adding additional harmonics due to overexcitation.

Admittedly, you have nonlinear loads to contend with also. Changing a tap could be a cheap fix. You might get lucky.

 

[prc]

John says there is no voltage distortion ie exciting flux is sinusoidal with out harmonics.Then the harmonics will be in currents. Eventhough load current do create noise it may not be a major component to my knowledge.3 rd harmonics will be circulating in delta and will not be transferred to primary.
The solution mentioned my Mike is a good solution to reduce vibration from laminations. This is some times used in large transformers too.Brush a low viscosity epoxy resin on the core edges which will be sucked in between laminations.Contact a transformer manufacturer who may be able to help you.Avoid over voltages. Overexcitation will substantially increase transformer noise.