5000kVA 480V Unit Substation Transformer

[rockman7892]

We are considering using two 5000kVA 13.8 V-480V transformers for a M-T-M 480V unit substation to feed 6000A 480V Switchgear. The normal load calculations on each transformer with the tie breaker open is about 3000kVA and 2500kVA which would be well within the transformers rating. During a situation where one transformer out of service and the tie breaker closed the calculated load on each transformer would be a little more than 5000kVA.

I've never seen a transformer this large be used for this type of application so I was curious to hear the thoughts of others for using a 5000kVA transformer on a LV Substation?

The two downsides that come to mind are the large amount of available secondary fault current and high Arc Flash Hazard, although the Arc Flash hazard would be high on the secondary for most any transformer used in this application.

I supposed these transformers could be sized to 3750kVA or maybe even lower since this would be adequate for normal conditions and utilized the transformers increased forced air cooling during a tie situation. Is this utilization of the forced air cooling under a tie condition a standard practice?

 

[davidbeach]

To me, that set up is asking for trouble. Unless you're dealing with a three or four transformer spot-network or a large generator, anytime you get above 4,000A you really need to take a step back and consider what you're doing. To me, you're either at too low a voltage or your transformers are too large. 5MVA at 4.16kV wouldn't be an issue nor would 2.5MVA transformers at 480V.

Maybe two sets of M-T-M gear and four 2500kVA transformers would be a better choice. You could even run a ring like system where there's always an open between any two transformers and the remaining transformers. Kinda all depends on exactly what the load requirements are and how much outage can be tolerated when switching from a source that just died to a still good source.

 

[rockman7892]

Davidbeach

In terms of "asking for trouble" is this mainly due to the large available fault current with these transformers?

The loads being served are all 480V motor loads divided up between 4 MCC's. The original concept was to have 4 2500kVA transformers with two sets of M-T-M MCC's for these loads. The 480V switchgear option is being considered in order to provide future flexibility for additional MCC's as well as provide an extra level of reliability ( M-T-M at upstream 13.8kv switchgear) This option may also realize a cost savings with eliminating two transformers, two upstream 13.8kV breakers and two 13.8kV feeder runs.

Is sounds like however tbt 5MVA transforms may be a bad idea and the original concept of 4 2500kva transformers feeding 2500A MCC's directly is a better fit.

 

[davidbeach]

My experience suggests that anything over 4000A is too much load at too low of a voltage. Spot networks represent an odd exception, as do large generators. Too much fault current, horribly expensive and low production circuit breakers, too many eggs in a single basket. One can get backed into a corner like that and have to deal with it, but better to avoid it in the first place. Long term I'm sure you'll be much happier with 4 x 2500kVA than with 2 x 5000kVA, even if the accountants aren't in regards to this project. What you might save now in first costs with the 2 x 5000kVA option you will spend over and over and over again dealing with the flexibility that you threw away if you take that route.

 

[itsmoked]

I don't know but transporting 2.5MVAs may be considerably cheaper than 5MVAs. Needing a replacement quasi-quickly may also be better to have the 2.5MVA.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[unclebob]

Arc Flash issue can be easily mitigated with remote operation. Another issue is the kA rating needed on the 480V gear.

 

[Mbrooke]

I'm with others, break this up into smaller units and sections if you can. Fault current at 480 volts can not even touch MV at 2.5MVA let along 5MVA. Around here buildings that need that level of power often have many smaller 1 to 1.5MVA substations (or exterior padmounts) distributed through out. In fact, its often cheaper to run MV cable then buss duct. Another benefit is selective coordination. The higher the fault current at the source the more breakers will trip for a fault on the load end.

 

[GroovyGuy]

Hi RockMan,
I would never recommend 5MVA xfmrs with a 480V secondary. The AF levels would be way too high. We never go above 2MVA with a LV secondary, that tends to keep the AF levels a tad more reasonable. I know of at least one local refinery that never goes above 1.5MVA for the same reason.

What is your reasoning to go secondary-selective (ie M-T-M) with the 480V system? I don't understand why anyone would do this unless your Facility has two independant power sources. I have noticed that in my 40 years of experience that there are more issues with secondary switchgear than with any other type of electrical equipment, certainly more so than a liquid-filled transformer.

Of course with smaller transformers, you do require additional primary switchgear. But keep in mind that this switchgear does not need to be one (1) 15kV metal-clad circuit-breaker per transformer. You can get away with a much less expensive load-interruptor-switch (LISw) per transformer. Many times I have uses one (1) metal-clad CB to feed a lineup of six (or more) LISw's, with one (1) LISw each [tt][/tt]feeding a transformer primary. Each LISw is equipped with a set of CTs and a PR, which in turn trip the upstream metal-clad CB. Of course every Client is different wrt their expectations or reliability, as well the true cost of down-time.

Regards,
GG

ps I would like to hear from other E-Engineers wrt their experience with what they would consider the most unreliable type of electrical equipment.

"I have not failed. I've just found 10,000 ways that won't work." Thomas Alva Edison (1847-1931)