False Ground Trip Caused by CT Saturation on a 13.8 kV Feeder
A 13.8 kV feeder is protected by a microprocessor relay using 600:5 multi-ratio CTs rated C100. The measured secondary loop resistance, including the CT winding, leads, and relay, is 1.3 Ω. The feeder's sensitive ground element is set at 20 A primary. During a bolted three-phase fault of 20,000 A on an adjacent feeder, this feeder's ground element operated and tripped the bus, blacking out two healthy feeders. Oscillography shows the phase currents were heavily distorted and the residual current peaked at about 0.9 A secondary. There was no ground fault on this feeder, and a later insulation resistance check of the cable showed no defects. The plant wants the relay ground pickup raised or the ground element disabled to prevent a repeat.
1. Using the CT and burden data, show why the ground element operated. Support the conclusion with numbers.
Read the worked answer
The CTs saturated. The residual current is a measurement artifact, not a ground fault.
Secondary current at the 20,000 A through-fault:
- Ratio = 600/5 = 120, so I_sec = 20,000/120 = 167 A = 33× rated secondary current
- Voltage the CT must produce = 167 A × 1.3 Ω = 217 V
What a C100 CT can actually do:
- A C100 rating means the CT will deliver 100 V into a 1.0 Ω standard burden at 20× rated secondary current (100 A) with no more than 10% error.
- At the installed 1.3 Ω burden, the accuracy limit is I_sec ≈ 100 V / 1.3 Ω = 77 A secondary, which is about 9,200 A primary.
The through-fault was more than twice that limit. Two of the three CTs saturate at different points because of residual flux and DC offset, so the three secondary currents no longer sum to zero. The resulting spurious residual of about 0.9 A is more than five times the 0.167 A secondary pickup, and the ground element trips. The distorted phase currents in the oscillography are the confirmation.
2. The plant proposes raising or disabling the ground pickup. Evaluate that, and give the remedies you would actually apply.
Read the worked answer
Raising the pickup is a partial fix; disabling the element is not acceptable.
The 20 A primary setting is a deliberate sensitivity choice for a 13.8 kV feeder, and disabling ground protection to solve a measurement problem trades a nuisance trip for an ungrounded-fault exposure. Ranked remedies:
- Dedicated ground sensing. 20 A primary on a 600:5 CT is 0.167 A secondary - near the bottom of any CT's dependable range and well below where ratio error is controlled. Use a core-balance CT or a separate low-ratio ground CT for this element. This is the correct fix for the sensitivity requirement.
- Reduce secondary burden. Shorten or enlarge the CT leads, remove auxiliary devices from the loop, or move to a lower-burden relay input. The accuracy limit scales directly with burden, so dropping 1.3 Ω to 0.7 Ω raises the accuracy limit to roughly 17 kA.
- Increase CT capability. Larger ratio and higher class (for example 1200:5 C400) takes the 20 kA through-fault to about 17× rated and provides a far higher knee. Note that this also cuts the ground element's secondary current in half, which pushes remedy 1 to the front.
- Coordinate the ground element. Add a short definite-time delay so a saturating through-fault cannot produce an instantaneous trip. This buys time but does not restore accuracy.
Apply 1 and 2 together and keep the sensitive setting.
3. What field tests and records do you require before the feeder is returned to service?
Read the worked answer
Prove the saturation with measurements and put the limit in the study.
Field tests:
- CT excitation (saturation) test on all three CTs, with the knee point and excitation curve recorded and compared phase to phase. An individual CT with a degraded knee explains why one phase contributed more residual than the others.
- Burden measurement of the actual secondary loop, including the CT winding, leads, and relay input, at the final installed configuration - not a nameplate assumption. Confirm it against the number used in the calculation.
- Ratio and polarity check of each CT, since a reversed polarity would also produce a standing residual current.
Records to leave behind:
- The saturation calculation: 77 A secondary / about 9,200 A primary as the accuracy limit at the measured burden
- The new setting, with the through-fault current above which the element can no longer be trusted, stated in the coordination study so the next person does not repeat this analysis
- Oscillography from the event, the excitation curves, and the burden measurement, filed with the relay settings
- Relay targets and trip log cleared only after the record is made, and the event reported as a misoperation, not as an unexplained trip