CT & VT "online" calibration

[vuan]

An electrical authority requires for HV and MV energy meters that CT and VT are tested with primary current and voltage injection each four years.

Does any authority accept “online” or “live” calibration for CT and VT, and if such standards are developed by some certified body or lab?

This is because as it can be seen, these tests, particularly in HV, require big dismantling work and also big voltage and current sources that have to be moved from lab to field. And of course in some cases with no backup power, some days blackouts that everybody wants to avoid.

 

[rcw retired EE]

I recall an IEEE article in IAS or PES Transactions describing a prototype system of clamp-on optical CT's & VT's for metering system verificaiton in high voltage substations. I think it was being used to "calibrate" existing relaying CT's now used for metering.

This method only compares meter readings using the normal load currents. It does not inject currents to verify full range accuracy. But it would prove the metering accuracy at the actual load operating point, which is the critical point.

I don't have time to search for that article right now. Good luck.

 

[Zogzog]

Not a big deal at all, I test these CT's and PT's all the time. The test equipment is small (not the same as a primary injection test on a CB)and the disassembly is usually easy. Not sure what the big concern is.

Cant help you on the blackout problem.

 

[oldfieldguy]

As Zogzog says, testing of CT's And PT's via low energy techniques as opposed to high currents and voltages is an established method as listed in many specifications.

This includes testing of CT's via the voltage ratio method and testing PT's at much less than rated primary voltage. These methods provide accurate results and sidestep a lot of safety and logistics requirements for using full voltage and high current.

Like Zogzog, I've tested a huge number of these devices in industrial and utility environments and never have had a question about accuracy, adherence to specification or indication of abnormal condition. The times where we did find problems, we knew what questionable results or indications would look like.

I've located CT's with problems in core magnetization, shorted turns affecting saturation voltage, incorrect ratio and polarity, and PT's with shorted turns and other winding failures, among other things.

I'm thinking that it's time for somebody to take another look at their testing specification.

old field guy

 

[rcw retired EE]

Zog & old field guy - I have also tested a lot of CT's using secondary injection and voltage methods. But I was never able to come up with a good way of verifying the phase angle and ratio correction curves that are sometimes applied to the meter readings in high accuracy metering applications. We even had trouble doing it at KEMA labs in Arnhem using null mehods that compared the CT results to an ultra precision CT.

Is there a secondary injection or voltage method to verify the accuarcy curves and phase angle shifts??

 

[DTR2011]

as far as I knom, Omicron makes a CT Analyzer that can test any parameter of a CT with extreem accuracy.

 

[scottf]

In my opinion, there is no way to do an accurate field accuracy test unless you can inject primary current across the range needed (for CTs) or rated primary voltage (for VTs).

There are other kinds of tests that can be done in the field while units are on-line or using low current/low voltage sources while off-line...however, none of these are true accuracy tests. They can be used to pick up other problems that could contribute to inaccuracy, but that's not the same thing.

In addition to a true source, one also needs calibrated burdens, and standard CTs / VTs that have been calibrated to a degree of accuracy many time better than the result being measured (in other words, you can't measure a VT to a 0.3% class using a standard VT only calibrated to a 0.3% class).

There is no hand-held or readily portable device that does a true accuracy test. Many claim to and will even give results, but they are not valid as a true indication of metering accuracy devices. Most of these hand-held devices aren't accurate past 1%, not to mention they can't generate suitable currents and voltages.

 

[stevenal]

I believe the system rcwilson spoke of is a service provided by Square D, formerly Power Measurement. Correction factors are obtained and dialed into an ION meter.

 

[scottf]

Stevenal-

The PML/SqD calibration uses an optical type CT that is placed on the same line as a reference.

That system is valid to a point. The limitation is that it can only measure/test at whatever current is on the line at the time of the test. I have heard of some sites leaving the reference sensor on for a period of time to study the accuracy at different primary current, however, often just snap-shot is done at whatever current is available.

Using the transformer correction function of the meter comes with a certain risk. First, the ION meters come with a number of correction points as a function of %current. The concept is geared around having the factory test data available, which would typically be at 10%, 100%, and RF at both zero and full burden rating (in the IEEE/CSA markets). To correctly compensate the meter, the actual burden must be calculated and the ratio and phase angle errors derived from the factory test reports (linear approximation between the zero and full burden points). Once the ION meter has this data, it compensates for the CT error based on the percent rated current present at that time.

If just one error reading is taken at one current level, the question becomes what correct factors do you use in the meter and what do you use for the other points to allow it to use approximation as the primary current level changes.

Field testing in this manner certainly has validity in that it can verify the CT is working correctly and giving expected readings. However, accuracy performance of the CT across its current range is not available, unless the station is off-line and current injection is used and/or the test is run at varying current levels.

Of course, this system is only for CTs...still doesn't address VTs/CCVTs.