Transformer Explosion

[SidiropoulosM]

There was an explosion of a 500/230 kv transformer in southern California last Friday. The explosion escalated to a fire through the entire substation. Sabotage is not suspected.

All transformer experts out there, I would like to know what are the various causes of catastrophic transformer failures and explosions and how they can be prevented.

Thanks. Michael Sidiropoulos

 

[dpc]

One possible cause is low oil level exposing the windings. The oil provides insulation and the transformer can fail internally if the windings are not flooded with oil. These failures can be pretty catastrophic and may not cause protective relays to actuate in the initial seconds of the fault.

Sounds like someone had a bad day.

 

[cuky2000]

It is hard to suggest a deterministic answer. The root-cause of catastrophic failure in power transformer may happen as a series of cumulative stresses causing degradation in time of the transformer insulation to withstand the a an overvoltage event.

Perhaps one way to narrow down the ramdom possibilities is to study failure statistic and apply that to the specific site. Forensic data suggest the fallow frequency of failures:

- Overvoltage stresses: (Lightning, switching surge, ferroresonance, SC,………….) ~ 45% to 50%
- Poor & infrequent maintenance (Including lost connection, dirty insulators, oil moisture, vibration...) ~ 20% to 25%
- Aging (Insulation degradation, oil degradation, ……) ~ 10% to 15%
- Overloading or LTC over-operation (if applicable)… ~ 5% to 0%

For the root-cause analysis of the catastrophic failure it is recommend gathering fault recorder/SCADA relevant system data. In term of priority, Insulation coordination assessment including surge arrester degree of protection should be one of the first areas to evaluate. Secondly, maintenance record and practices should be investigated among other causes.

 

[jstickley]

Although I don't claim to be an expert on transformers (I know just about enough to be dangerous), my thought is that there may have been a buildup of combustible gasses inside the unit, which ignited during an internal fault. Various transformer problems can result in buildup of a different gasses over time (hydrogen, methane, ethane, acetylene, ethylene, carbon monoxide, carbone dioxide, and others), which is why dissolved gas analysis is such a useful tool (you can trend the buildup of these gasses and assess possible causes and fix the unit before it fails catastrophically). My recollection is that overheating can cause a buildup of methane and ethane, and that arcing problems can cause a buildup of acetylene and ethylene.

Just a thought.

 

[rlwhite]

cuky2000 has recommended some good areas to look. Is is a dry type or oil filled. Did the transformer blow when it was turned on?

 

[prc]

As someone who knows a little about trfs and trf fires ,I can give some ideas.Before that give some more details about the failure mode of this trf-this must be a pretty big interconnecting auto trf.What are the types of bushings used?when these were last tested for C &tan delta?Where the fire was first noticed?starting point/component of fire?Is there a tank rupture ?What was the type of fire protection provided for this unit?

 

[ausphil]

Check last DGA results for the unit, and history of results. Are there any hints as to problems?

Are the bushings in tact? Is the tank in tact?

If there is an on-load tapchanger, check the position of the contacts to check if it got caught between taps. This sort of falut can be fatal.

Was protection operating properly at the time, (or is there no way to check the settings cause you now have molten relays)? Look for maloperations of protection (as cuky2k suggested, fault recorders will indicate if a fault held in for longer than it should).

Are bushings oil filled? When was the levels last checked?

Our main transformer failures that have resulted in fires have been cable endbox or linkbox failures. We have not only looked at this source of the problem, but also mitigation techniques for if a fire happens to start (ie blocking cable entry points through floors, fire extinguisher systems, smoke detection systems, even down to building design features such as having roofing eves flush with the external walls, to eliminate any overhang which would channel the fire inside the building.)

 

[SidiropoulosM]

This transformer does not belong to us, so I don't have all the details to answer some of the questions here. We will soon know more when the investigation is completed. Thank you all for your valuable comments. Michael Sidiropoulos