FINDING A CORRECT BUSHING CT RATIO FOR A GSU TRANSFORMER

[subash148]

Hello,

I am in the process of reviewing specification for a GSU transformer for a hydropower plant. As calculated, the primary current on the low voltage side(13.8 kv)is equal to 6521 amp and the secondary current on the high voltage side (241.5 kv) is equal to 373 amp. What ratio bushing CTs are required on the primary as well as the secondary size? How the CT ratios are determined based on GSU transformer primary and secondary current? What accuracy class CT should be used and how it is determined for this particular application?

Below is the information about the GSU transformer.

1 phase transformer
90 kVA
13.8 kV/241.5 kV

 

[davidbeach]

Just looking at it, I'd go with 6000:5 and 400:5.

 

[randyman]

It is often better to select the CT ratio to be higher than the max rated current it is measuring. in this case there is nothing wrong with specifying 7000:5 or higher ratio. The Rating Factor (RF) of the CT should be coordinated with the rated continuous current of the bushing, not exceeding RF 2. For example, let's say the bushing is rated for 10,000A and you select 7000:5 CT with RF 1.5. This should cover your normal full load of 6521A and any short-time overload you may encounter without exceeding the bushing rating.

Specifying the BCT for the application is better than over-specifying it and making the CT difficult to produce. This is especially true in higher current CTs and this is coming from a CT designer. But please, whatever you do, make the continuous current and bushing ratings part of your specification.

 

[scottf]

I agree with Randy's comments as long as the BCTs are to be used for protection purposes.

If they're to be used for metering (specifically revenue metering) then the sizing should be done differently.

 

[randyman]

Hey Scott, just curious ... why would metering making it any different?? From a technical perspective, metering and/or protection class wouldn't change a thing except how I would test it. Unless you are looking for a more "preferred" ratio??

 

[scottf]

If I was recommending a ratio for revenue metering...especially on the low side, I would recommend an RF4.0 (assuming IEEE market with Cl 20 meters). So 6521A max would be a 2000:5A or so (unless they want to go with an 1800:5A which is a bit of an oddball). For the best metering performance guarantee, always best to size the ratio as low as possible and use the RF to cover the max current. Goal is to operate above the nominal current rating as much as possible for the best metering performance.

For the HV side with a max current of 373A, it gets a little more tricky. As you know, tough to get good metering performance when you get down to that current level on a bushing CT, depending on how much physical room is available. To do it correctly for revenue metering, would need to get the burden rating as low as possible and possibly use some kind of extended-range or 0.15 class rating with a slightly higher ratio, but that is able to cover down to the lower currents. As we all know, the current may not be at or near the max most of the time.

 

[randyman]

OK Scott, so you are playing the RF game. If this were a dry-type CT I would tend to agree with you. However, this is a GSU so the BCT will be under hot transformer oil at 80C average and possibly higher. So with class A insulation and trying to limit a BCTs rise to 20-25C under those conditions could be challenging, BUT not impossible. I guess with the right design you could go either way but I tend to go for lower RF for high ratios under hot oil. I would apply the same concept on the high side to. It is not difficult to obtain good metering performance in 0.15 class with a 400:5 down to 5%, but you lower that ratio you can get into big trouble fast.

To the OP, with regards to accuracy that depends on the protection scheme and what's up and down stream of it. The higher ratio will have no trouble delivering 0.3B1.8/C800 class. And if you have IPB protection a much higher ratio CT may also be needed to match upstream generator CTs. As for the lower ratio CT, well depending on what you are needing its size can get out of hand. If for metering as I mentioned earlier it shouldn't be a problem obtaining 0.15 class with a low or medium burden, and on the protection side stay as low as you can possibly get away with. Requesting C800 for this ratio will make it quite large in size. If that becomes a problem in size then perhaps increase the ratio to 600:5. Again this is all generally speaking, you still have to coordinate with other devices. These are best answered by a systems guy.

 

[davidbeach]

A bushing CT is outside the tank, how would it be in 80C oil?

 

[ScottyUK]

David,

They're frequently mounted on the oil side of the bushing in a GSU transformer, at least over here in IEC-land. There are exceptions, but not many.

 

[randyman]

They are mounted inside the tank in IEEE-land as well. BCTs in general, are indoor type construction and the environment depends on the device they are going on. For example, in a power breaker they are dry-type mounted external of the interrupter but always under a protective shell/cover. In transformers they are oil-type construction mounted inside the tank either in the hot air pocket above oil or totally submerged in oil. Sometimes molded CTs are mounted externally about transformer bushings and called "slipover CTs". In all cases they are bushing CTs and mounted on the bushing. Can get rather confusing at times if one doesn't know where they are to be mounted!!

 

[scottf]

Randy-

The issue of temp rise is a valid concern, but normally there is enough physical space in a power transformer install to use a secondary wire of sufficient size to keep the Cu losses low enough at RF4 or RF3. Given how transformers are so often over-sized and the load on the transformer can vary so much, if you really want to have good metering performance across the entire operating range, going with a lower ratio and higher RF is definitely preferred from a metering perspective when possible.