4.5 Overcurrent Device Testing: Primary Injection, Secondary Injection, and Time-Current Curves
Key Takeaways
- Primary injection drives real current through the whole protection chain including the CT or sensor; secondary injection bypasses the sensing element and tests the trip unit electronics only.
- Primary injection is the only method that proves the complete current path and is required for acceptance testing of the assembled system.
- Long-time delay is verified at 300 percent of setting; instantaneous pickup is verified with a pulsed high-current shot to avoid heating the device.
- Time-current curves are plotted on log-log axes because both current and time span several orders of magnitude.
- NETA tolerance for a thermal-magnetic breaker long-time element is typically minus 0 to plus 100 percent of the manufacturer curve, while electronic trip units are held to the much tighter published band.
Overcurrent Device Testing: Primary Injection, Secondary Injection, and Time-Current Curves
Quick Answer: Level II task 2.1b.3 requires performing "contact resistance, insulation resistance, overpotential, and overcurrent protective tests" and operating "a high current circuit breaker test set." Primary injection forces real current through the entire protection chain — bus stab, current sensor, trip unit, trip actuator, mechanism, contacts. Secondary injection feeds a simulated signal directly into the trip unit, testing the electronics but proving nothing about the sensor or the current path.
1. The two methods, and when each is acceptable
| Primary injection | Secondary injection | |
|---|---|---|
| Current path proved | Entire chain, sensor included | Trip unit only |
| Equipment needed | High-current test set, hundreds to thousands of amperes | Small portable trip-unit test kit |
| Setup time | Long — heavy cable, connection to the primary terminals | Short — plug into the test port |
| Finds a shorted CT or open sensor lead | Yes | No |
| Finds a wrong CT ratio | Yes | No |
| Finds a mechanical trip failure | Yes | Only if the trip actuator is included |
| Typical use | Acceptance testing, commissioning, post-repair | Routine maintenance verification of trip unit calibration |
The governing principle: only primary injection proves the system. A secondary-injection test can pass on a breaker whose current sensor lead is broken, whose CT ratio was set wrong at installation, or whose rating plug does not match the frame. Those are exactly the defects that acceptance testing exists to find, which is why NETA acceptance testing of an assembled system calls for primary injection while maintenance intervals often accept secondary injection on the trip unit plus a separate mechanical trip verification.
The mechanical half must not be forgotten. Whichever injection method is used, the test is not complete until the breaker has actually been made to open its main contacts through the trip actuator at least once. A trip unit that correctly issues a trip signal into a shunt trip coil with a broken plunger has still failed to protect anything.
2. The verification points
Electronic trip units are conventionally described by the LSIG letters, and each element has a standard way it is verified.
| Element | What it protects | Standard verification |
|---|---|---|
| L — Long time | Conductor and equipment thermal overload | Inject 300 % of the long-time pickup setting and measure the time to trip against the published curve |
| S — Short time | Coordinated clearing of downstream faults | Inject just above the short-time pickup and verify the intentional delay band |
| I — Instantaneous | High-magnitude close-in faults | Verify pickup with a pulsed high-current shot, ramping in steps |
| G — Ground fault | Arcing ground faults below phase pickup | Inject through one pole and return outside the sensor, or use the neutral, to simulate residual current |
Why 300 % for long time. Long-time delay is an inverse function — the greater the current, the shorter the trip time — and at low multiples the time becomes impractically long and highly sensitive to small errors in the applied current. Three times pickup places the test point on a well-defined, repeatable part of the curve and keeps the test duration reasonable while producing enough separation to distinguish a correct device from a defective one.
Why the instantaneous test is pulsed. Instantaneous currents are large. Applying them continuously heats the current path, the sensor, and the test set. The technique is to apply a short pulse, step the magnitude up, and repeat until the breaker trips, recording the level at which it operates. Continuous application at instantaneous levels can damage the device you are testing.
Cooling between long-time shots. A thermal-magnetic breaker's long-time element is a bimetal that responds to heat. Two long-time tests in quick succession will produce a shorter second trip time simply because the bimetal has not cooled. Allow the manufacturer's specified cooling interval, and record ambient temperature — thermal-magnetic curves are published at a reference ambient, commonly 40 °C, and require correction outside it.
3. Tolerances — and why they differ so much
This distinction decides exam questions:
- Thermal-magnetic (bimetal and magnetic) breakers have wide published bands, commonly stated as minus 0 percent to plus 100 percent of the manufacturer's curve for the long-time element. A device that trips anywhere within a broad envelope is conforming. This looks alarmingly loose until you realise the curve itself is drawn as a band, not a line, and the device is a thermal analogue.
- Electronic trip units are held to the manufacturer's much tighter published band, because a microprocessor measuring true RMS current against a programmed setpoint has no thermal analogue in it.
The universal instruction is the same as everywhere else in NETA: compare against the manufacturer's published time-current curve. The tolerance is a property of the device, not a number the technician chooses.
4. Reading a time-current characteristic curve
TCC curves are plotted on log-log axes — current on the horizontal axis, time on the vertical — because both quantities span several orders of magnitude, from a few amperes and hundreds of seconds down to tens of kiloamperes and a few milliseconds. Log-log also has the useful property that shifting a device's pickup setting translates the curve horizontally without changing its shape.
Reading the shape:
- The upper left, steeply sloping region is the long-time element: high time, low current, inverse relationship.
- A flat horizontal shelf is a definite-time element — the short-time delay band.
- The vertical drop at the right is the instantaneous element: above that current, trip time collapses to the mechanical clearing time of the breaker.
- Total clearing time always exceeds trip unit response time, because the mechanism and arc-extinction take time. For coordination work, the downstream device's total clearing curve must fall below the upstream device's minimum melting or pickup curve.
The curve is a band, not a line. The left edge is the minimum operating characteristic and the right edge the maximum. Coordination is achieved when bands do not overlap.
5. Practical setup issues that corrupt results
- Test lead resistance. Injecting 5,000 A through undersized or loosely connected leads drops the current actually reaching the breaker. Measure the current at the device, not at the test set output.
- Connection resistance heating. A loose primary injection connection heats rapidly at test currents and can weld or arc. Torque connections properly and inspect them between shots.
- Rating plug and setting verification. Before injecting anything, record the frame rating, the rating plug or sensor size, and every dial or programmed setting, and confirm they match the coordination study. Discovering the wrong rating plug is a more valuable finding than any trip time you measure.
- Return the settings. A breaker left on test settings after commissioning is a serious latent defect. As-left settings are recorded and verified against the study.
- Ground fault interaction. On a system with ground-fault protection, verify that testing one element does not falsely operate another, and that zone interlocking behaves as designed.
Exam trap: A question asks which test method would detect a current transformer whose secondary lead has been left disconnected inside a low-voltage power circuit breaker. Secondary injection will pass, because it feeds the trip unit directly and never uses the CT. Only primary injection drives current through the sensor and exposes the fault.
A low-voltage power circuit breaker has a current sensor lead that was left disconnected during installation. Which test will detect this defect?
At what multiple of the long-time pickup setting is the long-time delay element conventionally verified?
Why is instantaneous pickup verified using a pulsed rather than a continuously applied current?