Summing CTs for load flow/metering

[Mbrooke]

Anyone have a general idea how to sum the currents of breaker CTs from many lines (say 10 138kv lines) for the purpose of calculating total MVA for load flow?

 

[davidbeach]

Basically the same way you'd sum them for a high impedance bus differential. Bring the individual CT circuits to a common point and wire them together. I've also heard of, but never seen, summing CTs that have multiple primary windings and a single secondary winding that connects to the meter.

 

[Mbrooke]

Thanks, crystal clear now! This must be a dedicated CT winding, right? And since this is technically a protection CT, can I get away with standard protection ratios instead of higher accuracy ratios provided I only have a small error in readings?

 

[davidbeach]

You could parallel them after they've gone through the relays. Certainly shouldn't plan of revenue uses of protection CTs, but for a general estimate of how much power they're perfectly fine.

 

[ScottyUK]

Mbrooke,

I can get you a scan of the summation CT type of scheme from work tomorrow if its any use / interest? I think the benefit of the multiple wound primaries is that it allows for correction of differing CT primaries to a common base. Direct paralleling only works if all the CTs are of equal ratio.

 

[stevenal]

The summing CT would serve to keep secondary current to your meter reasonable. Probably wouldn't want 50 A going to your meter. Not a problem for differential where you expect them to sum to zero, but might be a problem elsewhere.

 

[Mbrooke]

@ScottyUK, I would certainly be interested!

@stevenal, had not thought of that, I do need the current down the meter-able levels. My prior concern was fault current however. The primary L-G fault currents are around the 40,000 amp range.

@davidbeach, makes sense and thank you.

 

[davidbeach]

Fault currents are a good reason to use metering CTs that are intended to saturate at far less than 20 times rated. Keeps those high currents out of the meters.

 

[Mbrooke]

Can metering CTs be incorporated into dead tank breakers? Here we are planning on taking one winding of the normal protection CT for MVA metering.

 

[ScottyUK]

I've managed to get a copy from a remote login. Here's one of the summation schemes - apologies for the quality of the scans, most drawings I have to work with are much worse than these! Hopefully you can piece together how the scheme is connected from the two drawings, I'll try to explain anything that's not clear.

I don't have precise details of the summation CTs themselves unfortunately but my understanding is that each summation CT primary is specific to the CT connected to it, such that the primary for a 400A CT creates four times the amount of flux in the summation core that the primary for a 100A CT does for equal primary current.

http://files.engineering.com/getfil...a-17ae-4f9a-8314-f498cc1cfb33&file=g1899p.tif

http://files.engineering.com/getfil...e-1b27-4087-a071-496b3209c011&file=G1897P.TIF

 

[Mbrooke]

Awesome. I will study these and see if any questions come up. BTW, those might be old, but they are still gold.

 

[FreddyNurk]

Personally, if I had the opportunity, I'd be using whatever data capture is available on the relay, and aggregate the loads that way.

One of my former employers used to use summation CTs for metering applications, although as the equipment was LV, changing them was nowhere near as challenging as dealing with dead tank equipment. They had 4 primary inputs, all at the same ratio, and as a result all the primary side CTs had to be the same ratio. They also restricted the number of primaries thus decisions on number of feeders and so on were (partially) influenced by the summation CTs. I've heard of just paralleling CTs into equipment without using summation CTs, but never seen it done.

I'd not be feeding protection class CTs into them, rather I'd use metering CTs as inputs. Custom primaries like ScottyUK has asserted would work, but make any changes or upgrades more painful.

Its much easier to either use dedicated metering (which was done on one or two sites), or use whatever data is available from the relay, and then collate that data for total load flow. Of course, if the relays aren't capable of providing the data, then that's a different story. Whether it would be cheaper to upgrade the relays or change the CTs is debatable.

 

[davidbeach]

What size breakers? If you can get three bushing CTs, you can get two protection CTs and one metering CT. But if all you need is good enough and not revenue, just read the values from each relay over DNP or IEC 61850 and sum those values in your SCADA system.

 

[Mbrooke]

These are 145kv dead tank breakers:

https://library.e.abb.com/public/ec0a90a6b482e53d85257554005348c1/121-169PMI_2GNM110055_new.pdf

The other option is a live tank with separate CT but I am unfamiliar with my options on that side.

 

[davidbeach]

We use dead tank breakers from 57kV to 500kV. Bushing size may or may not allow enough CTs per bushing. At 500kV 3 CTs per bushing is no big deal; but at 57kV I'd doubt that we can get three on a bushing without running into clearance issues.

 

[Mbrooke]

Can protection and metering even fit on the same bushing?

 

[davidbeach]

No reason that they couldn't that I'm aware of. Not that that's a conclusive conclusion...

 

[scottf]

The number of CTs that can fit on a breaker bushing is a function of the space available on the breaker and the CT size, which is based on the specs. In general, the higher the ratio CT (with all else held constant) the smaller the size of the CT.

A few points to note:
- in the IEEE world, it is not uncommon for bushing CTs to have both a protection and metering accuracy rating.
- if the application is indicating metering and not revenue metering, a protection class CT will work just fine. In reality, it would probably work fine for revenue metering, you just wouldn't have the ratings and test results to back it up like you would need for revenue metering. If using for metering and if you have a multi-ratio design, try to pick the most appropriate ratio for the expected current level. Use as low of a ratio as you can and let the rating factor be used to cover the maximum continuous current expected.
- if you need to order a breaker with more CTs, take a look at lowering the protection class rating of the protection CTs. Just about everyone specifies C800 ratings still, but that is normally over-kill for most applications that use electronic relays. With everything else held constant, a C400 class CT is roughly half to 3/5 the height of a C800 class CT.

 

[LionelHutz]

If you're using almost any decent digital protection relay, then communicate with the relay and sum the power that way. A panel PLC has got to be cheaper then trying to install new CT's for this metering and a communications link has got to be cheaper and easier to run to a common point then CT wiring. You could run fiber optics if there is any concern about grounding differences messing with the communications.

But if you need to do it via the CT;s and if the CT's are all the same ratio, then simply pass the secondary of each through the window of a 50:5 CT and use that to feed the power meter.

 

[Mbrooke]

Are C400s more likely to saturate? I mean you would want that for metering, but if it also did protective relaying saturation is a no-no.