Power transformer vibration limits

[PrebenJakobsen]

We have an application with an offshore wind farm substation designed with a monopile giving quite some vibrations in the structure due to wind and waves.
How vulnerable will an e.g. 250MW 33/150kV 50Hz transformer be for such vibrations?
More steel means less requirements to transformer vendor and vice versa. What would be reasonable limits?
Best Regards.

 

[TurbineGen]

What type of transformer is this? Shell form or core form? What kind of vibrations? High speed small vibrations or lower frequency large scale oscillations?

The amount the unit can take will vary depending on manufacturer and type of unit. We have several mobile substations which are bsically a transformer that is mounted to a trailer and is transported every couple of months for it's entire lifetime. The vibrations from running down the less than perfectly paved roads do not seem to affect the units at all, but then they were designed for it.

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

 

[WindTurbinesAreFine]

I'm not familiar with offshore wind farm designs, so please forgive my ignorance.

Are you saying that the substation is going to be installed on an offshore platform? Or is the substation going to be on the shoreline, and you are concerned about the vibrations from the waves crashing into the shore?

 

[PrebenJakobsen]

The Substation is placed at the windfarm at app 15m water 30km from shore.
The design is ongoing, but data on similar projects shows 0,37-0,38 Hz with RMS displacement of 0,053m and peak displacement of 0,462m (100 years).
I expect a ordinary core form transformer, and it has to be powered when it vibrates.
Best regards

 

[jghrist]

I would expect that core and coil bracing for through faults would prevent any problems there. You may want to specify some extra supports for internal leads to prevent fatigue failure from the constant vibration. Operation of Buckholz relays may be a problem during large displacements if the oil sloshes around too much. I would avoid electromechanical auxiliary relays such as the G.E. HAA target relay that can be sensitive to vibration.

 

[electricuwe]

If your transformer is equipped with a tap-changer (on-load as well as off-load) you should also ensure that this device will be able to cope with the vibrations.

 

[PrebenJakobsen]

Are the values "normal" or will it add large cost to the transformer design?
Are there any IEC or other standards dealing with vibrations?

 

[electricpete]

Here is one way to look at it:

Make a big assumption that the vibration is roughly sinusoidal ... sinusoidal amplitude may change over time but on a time scale much longer than the period of the sinusoid.

sinusoidal vibration at 0.38hz which has an rms displacement of 0.462m rms would have a peak/0 acceleration of 0.4 g’s.

I have seen large GSU’s shipped on a truck – the limit for acceleration for the impact recorder was 5 g’s... agreed by the OEM.

On that basis, I would think it is not a problem. But that’s just a quick look, and need to check the assumption.


=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I guess there are two categories of damage:
1 - damage due to infrequent high-magnitude accleration
2 - fatigue or loosening due to long-term low-level vibrations.

My comments above addressed only the first.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[TurbineGen]

For most transformers I wouldn't think the low level vibration would be much of a problem. The securing blocks inside the transformer are blocked in tight and all threaded parts should be coated with a thread locker. I have seen these oocasionally break, but I have yet to see one come loose.

It might be worth the time to get someone inside the transformer and add a second nut to all of the threads. This is what we do with our mobile units to ensure nothing comes loose.

If it is a concern, you could mount the transformer on an air lift platform similar to a truck trailer suspension system. truck trailer air bags would do the trick. You would need to make a solid base supported in several places by airbags. I would think you'd need about 20 bags for a transformer of that size. These would help absorb some of the impacts. I would work similar to the suspensions system in your car. This would be more for large impacts, though it could help with the vibration issue.

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

 

[Pam60]

Hi electricpete

How did u do this mathematics of accelration in g's.

Thanks

 

[electricpete]

Converting 0.462 m rms sinusoidal displacement at 0.38hz into g's

f = frequency = 0.38hz
w = radian frequency = 2*pi*0.38 sec^-1

Write expression for displacement d(t)
d(t)=sqrt(2)* 0.462 * sin(w*t) meters

Compute velocity v(t) as derivative of displacement d(t)
v(t) = d/dt{d(t)} = w * sqrt(2)*0.462*cos(w*t) meters

Compute acceleration a(t) as derivative of velocity v(t)
a(t) = d/dt{v(t)} = w^2 * sqrt(2)*0.462*cos(w*t) meters

Substitute w = 2*pi*0.38 sec^-1
a(t) = (2*pi*0.38 sec^-1)^2 * sqrt(2)*0.462*sin(w*t) meters

Move units to the end
a(t) = (2*pi*0.38)^2 * sqrt(2)*0.462*sin(w*t) meters/sec^2

Multiply by g*sec^2/(9.8*m^2)=1
a(t) = -(2*pi*0.38)^2 * sqrt(2)*0.462*sin(w*t) meters/sec^2 * g*sec^2/(9.8*meter)

Cancel units
a(t) = -(2*pi*0.38)^2 * sqrt(2)*0.462*sin(w*t) g / (9.8)

Rearrange all numerical factors to the front
a(t) = -[(2*pi*0.38)^2 * sqrt(2)*0.462 /9.8] * sin(w*t) g's

Solve [ ]
a(t) = -[0.38] * sin(w*t) g's

peak value of a(t) is 0.38 g's

Here is an application note on converting vibration units (handy for rotating equipment vibration conversions)

http://www.reliabilitydirect.com/appnotes/cvr.html

Also there are vsiour programs to perform these conversions Here is one for free:

http://www.wilcoxon.com/knowdesk_VibCalc.cfm

Attached is a spreadsheet I wrote for the same purpose

=====================================
Eng-tips forums: The best place on the web for engineering discussions.