CCVT Ratio Verification 1

[thermionic1]

From time to time we test HV CCVT's (115kV-345kV). We have found that when using a TTR or similar, the measured ratio is nowhere near the calculated (5-15%). We have used TTR kits that claim to be able to accurately measure the ratio and have varied the voltage up to 100V on the primary.

What we have found that works is using a Power Factor test set as a source (10kV) and then measure the secondary voltage with a DMM. The measured ratios are then within an acceptable range. Unfortunately with this method, we can't verify the polarity in one step, like a TTR does, nor does the calculation / documentation get done electronically.

I am thinking there is either some voltage dependant issue or that the TTR does not have enough "uumph". We've used the same TTR to ratio test 138kV Iron/Copper VT's with no issue, but have issues with a 138kV CCVT. We have used a rental (Omicron) test set on 345kV CCVT's at 2kV and the ratios were spot on. I've experimented with a variac and a 120:480V VT (backfed) and that didn't seem to help.

I'm wondering what others do and if there is a reason for the large deviations at lower voltages.

Thanks.

 

[scottf]

The intermediate transfer of CCVTs is typically heavily compensated, meaning that the true turns ratio is not exactly the rated ratio. As manufacturers, we vary the turns ratio to compensate for expected and measured inaccuracies in the capacitor divider and compensating reaction.

When you try to measure the ratio with a TTR, the voltage applied to the primary terminal is typically in the range of 100V. This voltage is not high enough to fully excite the core, which is why the results aren't as expected.

Most utilities perform insulation power factor measurements (Doble test) at 10 kV. They normally use that voltage source to check the CVT ratio. 10 KV will almost always be high enough to work (I have run into issues with 10kV not being high enough to test an 800kV unit).

On that same point, I'm surprised you have gotten such a TTR to work on 138kV inductive VTs. It might have just been a manufacturer that didn't use turns compensation. We (Ritz) start using turns compensation at 36kV and above and it's common to field calls from customers trying to use TTRs testing at 30V to test 36kV VTs and not getting good results.

Keep in mind VTs (and CVTs) are only required to be within the stated accuracy class from 90% to 110% rated voltage. So when you put 100V on a 138kV VT, that is only 0.1% rated voltage.

 

[thermionic1]

Hello ScottF-

Thank you for the prompt response. Thank you also for the description of the problem, using conventional TTRs. I will share your information with others who have had the same question. I guess we will continue using the 10 kV Doble as our source - we have to test the capacitance and PF/DF/TD as part of the required tests anyway.

The 138kV units we tested were done with a small battery operated, Megger branded unit (electronic) unit. It takes several minutes for the unit to display a result. This is very tedious when there are 155/67 & X/Y/Z windings to test.

As you are probably aware with NERC requirements, the documentation comes under scrutiny, which is why the TTR with automatic reporting would be an attractive option. Using the Doble as a source, we can paste in an excel sheet that
provides the results and calculations.

Thanks again.