Current Transformer Saturation Relay Misoperation Case Study

Current transformers make high-current power-system faults measurable by relays. They also create one of the most important protection-security risks: if a CT saturates during an external fault, the relay may receive unequal secondary currents and interpret the resulting spill current as an internal fault.

This case study follows a medium-voltage bus differential trip during a feeder fault outside the protected zone. The case is simplified for engineering learning. It is not a substitute for a project-specific protection study, relay manual, CT excitation curve, transient simulation, manufacturer data, commissioning procedure, or qualified protection-engineering judgement.

The central question is:

Did the bus differential relay correctly trip for an internal bus fault, or did it misoperate because one CT saturated during an external through fault?

The answer depends on evidence from the fault location, breaker targets, oscillography, CT burden, CT excitation capability, relay characteristic and post-event inspection.

Case Context

An industrial 13.8 kV switchgear lineup uses low-impedance bus differential protection. A downstream cable fault occurs on Feeder F3. The F3 feeder breaker opens, but the bus differential relay also trips and locks out the main bus. Site inspection finds no damage on the bus, but production is interrupted because the differential lockout requires a controlled restoration review.

Item	Value
switchgear bus voltage	13.8\ \text{kV}
external fault current through F3 CT	18.0\ \text{kA RMS}
CT ratio	1200:5
CT ratio factor	N=240
CT knee-point voltage, all zone CTs	V_k=110\ \text{V}
CT winding resistance	R_{ct}=0.18\ \Omega
source CT external burden	R_{ext,A}=0.32\ \Omega
F3 CT external burden	R_{ext,B}=0.92\ \Omega
screening transient factor for DC offset and remanence	K_t=1.8
relay differential pickup	I_{min}=0.5\ \text{A secondary}
relay percentage slope in the event region	m=25\%
measured source-side CT current during first cycle	75\ \text{A secondary}
measured F3 CT fundamental-equivalent current during first cycle	55\ \text{A secondary}

The values are representative. Real CT performance must be checked using the actual CT class, excitation curve, remanence assumption, X/R ratio, relay input burden, lead resistance, terminal resistance, saturation detector, differential algorithm and fault-current study.

Field Evidence

The event evidence points away from an internal bus fault and toward CT saturation during an external fault.

Evidence	Engineering interpretation
downstream F3 breaker target asserted	external feeder fault is credible
bus differential target also asserted	relay saw spill current inside the differential zone
bus inspection found no flash marks, pressure damage or insulation tracking	no independent internal bus-fault evidence
oscillography shows the F3 CT current flattened after fault inception	CT saturation is plausible
source CT current remains sinusoidal for longer	CTs did not saturate equally
F3 CT secondary leads are longer and routed through an added test switch	burden asymmetry is plausible
voltage recovers after the feeder breaker opens	fault was cleared outside the bus zone

A differential relay can be correct when the primary system is wrong, or wrong when the primary system is healthy. The event record must therefore be reviewed as a measurement-system problem as well as a protection problem.

Step 1: Calculate CT Secondary Fault Current

For a CT ratio of 1200:5, the ratio factor is:

The ideal secondary current for an 18.0\ \text{kA} primary fault is:

Engineering Comment

The secondary current is much higher than the nominal 5\ \text{A} rating because the event is a fault, not a load condition. A protection CT is expected to reproduce high multiples of rated current for a limited time, but only if the required secondary voltage and flux do not drive the core into severe saturation.

Step 2: Estimate Required Secondary Voltage

A first burden screen uses:

For the source-side CT:

For the F3 CT:

Both are below the nominal knee-point voltage if the current is symmetrical and the CT has little remanence.

Engineering Comment

This symmetrical-current check is necessary but not sufficient. A high X/R fault can include a large decaying DC component. DC offset and remanent flux increase the volt-second demand on the CT core, so a CT that appears acceptable under symmetrical RMS current may still saturate in the first cycles of a real fault.

Step 3: Include a Transient Saturation Screen

Use a simplified transient multiplier:

For the source-side CT:

For the F3 CT:

Compare with the CT knee point:

The source-side CT screen is:

The F3 CT screen is:

Engineering Comment

This result explains the asymmetry in the event record. The source CT has margin, while the F3 CT is expected to saturate under the selected transient assumption. The exact saturation time requires the CT excitation curve and waveform simulation, but the screening calculation already shows that the F3 measurement is not trustworthy during the first fault cycles.

Step 4: Calculate Differential Spill Current

For an external fault, ideal differential current should be near zero because the current entering and leaving the differential zone balances:

The event record shows that the source-side CT reports:

The saturated F3 CT reports a fundamental-equivalent current of:

The simplified spill current is:

The restraint current for this simplified two-current case is:

Engineering Comment

The relay did not need a real internal bus fault to see operate current. It only needed unequal CT reproduction during an external fault. This is why differential protection depends on CT selection, burden control, saturation logic and commissioning evidence, not only on relay settings.

Step 5: Check the Relay Trip Characteristic

Use a simplified percentage-differential operating threshold:

With:

Then:

The measured operate current is:

Since:

the relay characteristic allows a trip.

Engineering Comment

The relay trip is explainable by the settings and the measured spill current. That does not prove an internal fault. It proves that the protection system did not remain secure for this external fault and CT-burden condition. Corrective action must restore security without making the relay blind to real internal bus faults.

Step 6: Review Corrective Options

The engineering correction should address the physical and logical causes of the misoperation. Reasonable options include:

1. reduce F3 CT secondary burden by shortening leads, removing unnecessary test-switch resistance, improving terminals or moving relay inputs;
2. replace the F3 CT with a class and knee-point voltage suitable for the fault current, X/R ratio and burden;
3. configure relay CT-saturation detection or external-fault blocking, limited to the interval where saturation evidence is present;
4. increase differential slope in the high-restraint region if the relay philosophy permits it;
5. verify CT polarity, ratio, grounding and phase compensation;
6. replay event records and secondary-injection cases before release;
7. retain fast differential operation for credible internal bus faults.

The weakest correction is simply raising pickup until the nuisance trip disappears. That may hide an internal bus fault. A good correction explains why the previous event is secure and why true internal faults still trip.

Step 7: Check the Burden Correction

Suppose the F3 CT external burden is reduced from:

to:

The new symmetrical voltage demand is:

The transient screen becomes:

Compare with the knee point:

The corrected burden gives a screening margin of:

Engineering Comment

The corrected burden does not make the CT ideal. It makes the first-cycle saturation risk much less severe for the studied external fault. The release package should still include waveform replay, relay event simulation, high-current injection where feasible, and confirmation that the maximum credible fault current has not been understated.

Step 8: Relay Replay After Correction

After burden correction and relay event replay, the expected F3 CT current during the first cycle improves to:

The spill current becomes:

The restraint current becomes:

The threshold is:

Since:

the differential element should remain secure for the replayed external fault.

Engineering Comment

This is the key security check. The corrected system no longer turns an external fault into a differential trip. The engineer must also run the dependability check: simulated internal bus faults must still exceed the operating characteristic and trip within the required time.

Step 9: Internal-Fault Dependability Check

A secondary-injection test for an internal bus fault gives:

Test quantity	Value
operate current	I_{op,test}=8.0\ \text{A}
restraint current	I_{res,test}=10.0\ \text{A}

The threshold is:

Since:

the internal-fault test still trips.

Engineering Comment

Protection validation must show both security and dependability. Security means no trip for the external through fault. Dependability means trip for the internal fault. A correction that proves only one side of that pair is not a complete release basis.

Engineering Decision

The bus should not be returned to unrestricted service using the original CT-burden and relay-security basis. The evidence supports a protection misoperation caused by unequal CT saturation during an external through fault.

The release decision is:

Keep the bus differential lockout under engineering control, verify that the primary bus is healthy, correct the high-burden CT circuit, replay the external-fault event, test internal-fault dependability, document the remaining CT saturation margin, and release only after the relay remains secure for the external fault while still tripping for internal faults.

This is not a reason to distrust differential protection. It is a reason to treat CT circuits as part of the protection system, not as passive wiring.

Failure Modes and Controls

Failure mode	Technical effect	Control
excessive CT lead burden	core saturates during high-current faults	measure loop resistance and include terminal/test-switch resistance
CT class too weak for fault current and X/R ratio	secondary current is clipped or delayed	select CT using fault study, burden and transient requirement
relay slope too low for high-restraint external faults	spill current crosses operate threshold	validate slope regions with external-fault replay
saturation detector disabled or misapplied	relay trips before blocking/security logic works	test relay logic with recorded and simulated waveforms
pickup raised without internal-fault test	true bus fault may be delayed or missed	run internal-fault secondary-injection cases
event reviewed only as a primary fault	measurement failure is missed	inspect oscillography, CT circuits and relay targets together

Risk Review

Risk item	Severity	Occurrence	Detection	RPN
bus differential misoperation for external feeder fault	8	3	5	120
internal bus fault desensitized by excessive setting increase	10	2	5	100
CT burden change undocumented after maintenance	7	4	6	168
event replay omitted before release	8	3	4	96

The highest-priority control is configuration discipline: CT ratios, burden, relay settings, wiring changes and test-switch states must be treated as one protection system.

Release Criteria

Release should require evidence that both the primary equipment and the protection measurement chain are fit for service.

Criterion	Required evidence
primary bus health	inspection, insulation checks and absence of internal-fault indicators
external-fault classification	feeder target, breaker status, fault location and voltage recovery agree
CT circuit configuration	ratio, polarity, grounding, lead resistance and test-switch state verified
burden margin	corrected V_{req,tr} remains below the accepted CT capability for the studied fault
relay security	external-fault replay does not cross the operate characteristic
relay dependability	internal-fault injection crosses the operate characteristic and trips on time
setting governance	changes are recorded, reviewed and locked under the protection-setting process
operational handover	operators understand lockout reset, event capture and escalation criteria

Transferable Lessons

Current-transformer saturation is a measurement failure that can become a protection failure. The practical workflow is:

1. classify the primary event before changing settings;
2. compare fault location, breaker targets and oscillography;
3. calculate CT secondary current from the ratio;
4. calculate burden voltage for each CT circuit, not only the nominal CT;
5. include a transient saturation screen for DC offset and remanence;
6. calculate differential spill current and restraint current;
7. test security for external faults and dependability for internal faults;
8. release only when CT circuits, relay settings and evidence agree.

The case is distinct from transformer energization inrush. Inrush is a normal energization transient that can mimic an internal transformer fault. CT saturation during an external through fault is a measurement-chain failure that can create false differential current even when the protected zone is healthy. Both are protection-security problems, but they require different evidence and different controls.