UPS vs transformer

[Fc1908]

I'm working on a project in which a 400kVA 400V UPS feeds a cable line (included two 500kVA 0.4 / 20kV transformers) in order to reduce problems of voltage dips.
Would I know if this is possible?
Is it possible to supply transformers via UPS? If yes, there is a criterion in order to define the size limit? How long does it take a UPsS to switch from normal network to battery? The UPS need to magnetize the transformers or the voltage dip is minimal?

 

[ScottyUK]

What kind of UPS? A rotary could possibly handle this configuration, most electronic types will have problems unless they are massively oversized (and 400kVA is already fairly large for a solid-state type). Time to start speaking to manufacturers I think.

 

[waross]

Do you want this to provide power if the normal supply fails or are you trying to support voltage dips with the normal supply still online?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Fc1908]

Thanks to eveeyone for the answers.
Scotty, it is not a rotary type, it is an electronic type.
Waross, the idea is to support normal network. An EDG will start in case of longest lost of network.

 

[iop95]

For such power, UPS is online topology and tranfer time is near zero.
Depending of distorsion of UPS output voltage may be same additional coils and core heating.
As output voltage is composed by PWM with a frequency not so high for this power, I think main problem will be risk of core saturation because may be small difference in voltage x time product.

 

[davidbeach]

It's very common for a UPS to feed a transformer, look up PDU (Power Distribution Unit). In the US, most larger UPS installations run at 480V and support 208Y/120V loads via PDUs. The UPS static switch may momentarily go into bypass when the transformer is energized.

 

[FreddyNurk]

One of the problems, that ScottyUK has already alluded to, is that there's no spare capacity for extra current for energising transformers, or providing fault current, or other issues that a 'normal' network would be able to do. Rotary UPS units would likely have a different topology, although I'd expect that inverter coupled units would have the same issue.

In this case 400kVA is 400kVA, there's no additional capacity, hence ScottyUK's statement about having to oversize the unit. Of course, this isn't any different to the larger scale battery energy storage that is starting to be discussed, although some units do advertise additional fault current capacity.

It would also be worthwhile finding out about the MV connected inverter units for solar farms and so on, and ask the manufacturer what capability they have for energising their own transformers without a network present. While some manufacturers only provide for solar installs, battery units are becoming more common.

EDMS Australia

 

[ScottyUK]

Thanks Freddy,

I should have expanded - the true rotaries are plain electromagnetic machines, loats of rotaing iron and copper. The types with a flywheel energy store which use an inverter for the output seem to combine the worst features of both technologies in my opinion. Some well-built products, based on a flawed idea.

David,

You're right, but it's the relative sizes of the transformers which gives rise to the problem in this application. The local PDU type are, as individual loads, usually small compared to the capability of most central UPS used in datacentre applications. This is a pair of back-to-back 500kVA transformers in series connected to a 400kVA UPS; the UPS doesn't stand a chance of energising that. Perhaps if it was pre-connected during the soft ramp-up of the inverter output, but that's not a normal operating state and would be a manufacturer-specific question.

Assuming this could be energised, how easy would it be to protect that 20kV line given such a weak source?

 

[itsmoked]

how easy would it be to protect that 20kV line given such a weak source

Scotty; I'm betting all inverters (UPS) of those sizes are massively instrumented and know to the erg, how much load they're supplying. I'd expect installations to have a shunt-trip breaker that the UPS could dump at a specific amp overload. Rather like VFDs that protect themselves and their loads rather dependably.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Fc1908]

Application is not a solar plats but it is an oil&gas installation. It is an existing installation with some problems of voltage network.
The UPS is intended as a buffer for voltage dips that produce the lost of well substation. In case of longest problems on network an emergency diesel will start and will energize the line and the well substation.
In case of first start up the transformers will be normaly energized via the emergency line that bypass UPS.
I know UPS will not be able to energy such transformer if not live but I don't know the behavior in case of voltage dips.
The transformer produce an inrush again? How much is reactive power absorbtion?
I know this is not the best solution and I am not sure about its correct application, but I beleave you could understand because we have always the same problems of money saving.
Attached one line diagram.

About electrical protection, in case of fault
there is no way to improve selectvity but I think UPS will disconnect for overload due to reactive power absorpion.
MV side is earthed via resistance.

 

[davidbeach]

I guess I skipped over that 20kV part of the original question; the UPS would be in bypass a lot probably. Some sort of power conditioning equipment at the load would work better.

 

[waross]

1. Install the UPS downstream of the two transformers.
2. If you want the UPS to run in parallel with the normal supply, you should be talking to the manufacturer.

Bill
--------------------
"Why not the best?"
Jimmy Carter