CT sizing for Reverse Power Application 2

[Elec_Engr]

Hello,

I am working on designing a reverse power protection scheme for a 40 kW PV System at 480 V. The utility is requiring a reverse power setting to be set at 10 kW in an SEL relay. I am struggling with identifying a CT that would work for this application.

System details
Switchboard Rating: 1200A @ 480V
System Voltage: 480V (line-to-line)
Design Load: 1100A (there are a lot of redundant loads on this switchboard and the maximum load is not expected to exceed 600A)
Maximum Measured Load Current on the switchbaord:350A
PV System Rating: 40 kVA (AC) --> 48A
Reverse Power Setpoint: 10 kVA --> 12A

If I select a 600:5 CT with a 2x thermal rating, the CT has to measure 2% (12A on the primary) current in the reverse direction. The typical relay class CTs start to measure accurately at 5%. So my questions are:

1. Can I use a metering class CT rated at 600:5 that can measure with 0.5% accuracy between 1% and 300% of the rating?
2. I am not planning on using any other protection function in the relay, so is a metering class CT ok in this case?
3. Is it required by any code or standard to use a relay class CT for all the protection elements? In this case, it makes more sense to use a metering class CT as these are accurate in the low current ranges.
4. If there is a fault on the switchboard or downstream, I understand that these CTs would saturate. But I am not planning on using these CTs for any other protection. So once the fault is cleared and the system is back to normal, would these CTs have any issues measuring the low currents or reverse power?

Looking forward to responses from the experienced engineers in this field. Thanks in advance.

Thanks

 

[scottf]

The first question is how accurate does the measurement need to be at that low level?

You mentioned that the typical protection class CT starts to measure accurately at 5% Inom. Unless you're speaking to revenue metering accuracy, I would not agree with that statement.

Some things to consider:
- What do you define as "accurate" for this application?
- What would be the protection class of the CT you're planning on using and what would be the actual applied burden? (I think you'll find that the rated burden of the CT will be well above the actual connected burden, meaning that the performance at the low-end current will be "better" than you expect)
- If you can get an excitation curve of the protection class CT you're wanting to use, you can get a pretty good calculation of what the accuracy will be at the low-end at your applied burden. Trick might be getting an excitation curve down to that low of secondary voltage.

 

[bacon4life]

If this is a typical blanket utility requirement, they are probably expecting folks to use typical relay class equipment, and to pick the relay setting closest to to 10 kW. It seems unreasonable to require the relay setting to be exactly 10 kW on a 1000 KVA panelboard since relays have limited resolution(i.e. number of decimal points in the setting).

One bug/feature to be aware of in SEL relays is that they suppress low current values. Depending on the relay/firmware version and which part of the relay (LCD vs DNP vs SEL fast messaging) the relay may report values less than 2% of Inominal as 0.00 amps. In later firmware versions for certain relays they have added a user configurable setting to disable the the low current suppression. Your setting would be right at the 2% cutoff, so the relay might jump straight from zero to 12 amps.

Do consider a 100 ms time delay on the reverse power setting. During transients a load might momentarily export power to the grid.

 

[jghrist]

The reverse power setting is to ensure that there is no significant power export to the utility, without risking a false trip because of measurement error or current feeding transformer losses only. The accuracy is not important.

 

[EnzoAus]

If the utilities in your jurisdiction are similar to those in Australia, they often set unrealistic requirements on PV installations without thinking through the ramifications. A good example has been a requirement here to maintain minimum power factor at all times - impossible to achieve if export is possible. At import/export transition, P = 0 and even the smallest amount of Q will ensure that the power factor drops to 0 without any consequence.

If your site is a commercial/industrial load with a measured load of 350A (290kVA), then it will probably consume all the PV power being added, except during weekends or public holidays. Even then, given the variability of the weather, it is quite possible that standing loads will leave little for export.

AS/IEC standard 61869.2:2021 for instrument transformers quotes standard metering CT accuracy in terms of a ratio error from 5-120% of rated current. Higher accuracy class CTs go down to 1% of rated current. Of course, this is with an appropriate burden and power factor. By contrast, protection CTs only seem to quote the ratio error at the rated primary current. My understanding is that they are designed primarily to handle short circuit conditions without saturating.

Given the difficulty of accurately measuring such a low export power in an area of low CT accuracy, I would consider the ramifications. With such a large load, the utility fault level must be significant and it is difficult to see how a 10kW 3 phase export will materially impact voltage rise. Even 40kW of 3 phase export is only equivalent to 6.7kW (40/3x2) single phase export allowing for the factor of 3 and the additional effect of the neutral voltage rise in a single phase system. In Australia, hundreds of thousands of residential customers have been approved for 5kW inverters (on 400V line to line systems) without export restrictions, although V-Q and V-P restrictions have been steadily introduced.

If you do not have to test/prove the system and it is unlikely to impact on the voltage, I would suggest making a pragmatic CT selection that complies with all the proper requirements for burden, output voltages not exceeding relay inputs etc and not worry too much about the accuracy. The system will comply in intent.