CT Accuracy result 2

[SooryaShrestha]

I have a 200/1 CT ratio tested with primary injection kit. I have given 100 amp on the primary and got 0.52 amp on the secondary. I have further given 200 amp in the primary and got 1.02 amp in the secondary. Can these values be used to determine the accuracy class of the CT?
I have been informed that since the burden needs to be between 25% to 100% of the rated burden, which is 30 VA, the above values can not determine the accuracy. Is it true?
My third question is: does the accuracy class need to be tested only in the lab? Can the accuracy class be tested in the field? If it can be tested at the field, how sophisticated are the instruments of testing?

 

[peterb]

Soorya -
A couple of points -
1. What is the rated accuracy class of the CT you are testing? Is it a relaying or metering CT?
2. As regards the burden, this will be defined in the standard to which the CT has been built. Suggest that you consult this (sounds like an IEC CT?)
3. In theory, you COULD test the accuracy in the field - IF you have lab grade test equipment of the required accuracy; this is NOT normal primary injection test equipment and clip on ammeters. The more usual approach might be something like this -
- Specify and witness a lab test of the CT in the manufacturer's factory or an independent lab as part of the supply contract
- Specify & obtain certified test results of these tests
- Perform field tests to confirm that the CTs have not been damaged in shipment or during installation on site. These tests would normally include insulation resistance and excitation curve checks. If these results conform with factory test results, there is very little possibility that the accuracy would have been altered from the lab test value.

 

[electricpete]

It sounds like peterb was right on the money.

Regarding the effect of circuit burden upon saturation/accuracy - it plays just as big a role as does the current magnitude. If I'm not mistaken, we read the saturation curves for a given CT using secondary voltage which is naturally the produce of secondary current and secondary impedance. The further the CT is into saturation, the larger the error.

 

[electricpete]

Another comment about testing... I think you can get a pretty good feel for the accuracy by applying varying secondary voltage over a range and measuring the resulting current (primary equipment arranged to prevent loop on primary).

This gives you the excitation curve. Accuracy can be calculated for any current and burden (complex) from the excitation curve. If I could remember how, I'd tell you, but it has been awhile.

 

[SooryaShrestha]

Thanks both of you.
There are two cores one for measuring with 0.5 class accuracy and the other for protection with 5P10 accuracy.
Yes the specification is 185 IEC-1987.
The problem is like this:
The material has already arrived at site. The equipment has been charged and supplying power to the consumers. The CT can not be taken out from the switchgear. At the time of witnessing the test at factory, the person sent for inspection is a junior person and somehow he overlooked and did not make the accuracy test. When the material arrived at site. we made a check to the ratio test, which i have given above.
My question is can those data be used for accuracy test result?
Second is can such test be performed at site?
Regarding calculation with excitation curve Mason's book says it can be done but not so sure.
Since this might go into litigation, I do not want to make any weakness on the part of testing.
Thanks both of you again. You are all worth STAR.

 

[peterb]

One way to test under the situation that you are in would be to temporarily install a post-type CT of known accuracy in the external circuit and do a comparison reading against the disputed CT. Depends on the availability and accessibility of the circuit to be taken out of service and modified (assuming open wire connections that would allow temporary CT installation). Otherwise you need to find an organization that has portable equipment of the required accuracy for field calibration.
As far as calculating the accuracy from the mag curve is concerned, this could be done if the data is available. If you draw the equivalent circuit of the CT, you will see that the major source of inaccuracy is the magnetizing current flowing in the magnetizing impedance that is connected across the secondary of the ideal CT. The external rated burden impedance connected in series with the CT secondary resistance, multiplied by rated load current gives the CT secondary voltage developed. The magnetizing current corresponding to this value of voltage will then be the error current of the CT. Sounds simple in principle, but the accurate calculation is more involved.
If you are going to litigation, perhaps you should consider getting a good consulting engineer to review the situation in detail on your behalf.

Hope the junior engineer involved has learnt a valuable lesson here. Perhaps the contract document should also be reviewed to ensure that all required tests are spelled out and agreed on ahead of time - for future projects.

 

[jbartos]

Suggestions:
1. Some posting above sound really serious, citing "litigation." In this case, it may be better to test it by the book or go by the book.
2. ANSI/IEEE C57.13-1978 "An American National Standard IEEE Standard Requirements for Instrument Transformers" includes various approaches to CT testing, e.g. you may need a "Standar Current Transformer" which serves as the ratio standard. This may be hard to come by.

http://www.ieee.org

Else, there are also other industry standards, e.g. IEC, etc.

http://www.iec.ch