Phase overcurrent (50/51) and ground overcurrent (50N/51N) are set differently because the phase element sees all the load current and the ground element sees only the residual current 3I0. Since balanced load produces no 3I0, the 51N can pick up at a fraction of rated current and detect high-resistance ground faults.
What 50/51 and 50N/51N overcurrent protection is
Overcurrent is the oldest and most widespread protection: it trips when the current exceeds a set value. In IEEE C37.2, 50 is the instantaneous overcurrent relay and 51 the AC time overcurrent relay, inverse or definite time. With the suffix N (or G, depending on the connection) they apply to ground current. In IEC 61850, time overcurrent is modelled with logical node PTOC and instantaneous overcurrent with PIOC.
- 50, instantaneous: trips with no intentional delay above a high threshold. It clears close-in faults quickly.
- 51, time overcurrent: trips with a delay that decreases as the current grows, or with a fixed time, to coordinate with downstream protection.
- 50N/51N: the same, applied to the residual current 3I0 = Ia + Ib + Ic. It detects ground faults, the most frequent faults on overhead lines.
Where it is applied
Medium-voltage feeders, industrial switchgear, motors, transformers as backup to the differential, subtransmission lines as backup to distance protection, distribution reclosers and solar plants at their point of interconnection. In high and extra-high voltage networks it is almost always backup; in medium voltage and in industry it is usually the main protection.
Phase and ground: why they are set differently
Phase protection sees all the load current, so its pickup must be above maximum load. It also sees ground faults, as long as their current exceeds that pickup. The reason for a separate 51N is sensitivity: balanced load produces no 3I0, so the 51N can pick up at a fraction of rated current and detect high-resistance ground faults that phase protection would never see.
3I0 is obtained from the residual connection of the three current transformers (CTs), by calculation in the relay, or with a core-balance (zero-sequence) CT around the three phases. The core-balance CT is more sensitive and is the usual choice with an isolated or resistance-grounded neutral.
How it works: inverse-time curves
The standard curves are in IEC 60255-151, which covers both the IEC and the IEEE curves, and in IEEE C37.112. With M = I / I_pickup:
- IEC: t = TMS · k / (M^α − 1)
- IEEE: t = TD · [A / (M^p − 1) + B]
| Curve | Constants |
|---|---|
| IEC standard inverse | k = 0.14; α = 0.02 |
| IEC very inverse | k = 13.5; α = 1 |
| IEC extremely inverse | k = 80; α = 2 |
| IEC long-time inverse | k = 120; α = 1 |
| IEEE moderately inverse | A = 0.0515; B = 0.1140; p = 0.02 |
| IEEE very inverse | A = 19.61; B = 0.491; p = 2 |
| IEEE extremely inverse | A = 28.2; B = 0.1217; p = 2 |
Very and extremely inverse curves look more like the fuse characteristic and coordinate better with fuses; standard inverse gives more even times when the fault current varies little along the feeder. Definite time is useful when the fault current depends heavily on the generation connected.
Setting criteria
The values come from the coordination study, not from universal rules. The general criteria are:
- Phase pickup: above maximum load, including contingencies and allowable overload, and below the minimum fault in the zone it backs up.
- Ground pickup: above normal unbalance and the residual CT error, and below the minimum resistive ground fault.
- Curve and time dial: with a coordination margin against the downstream relay. That margin covers the breaker time and the relay tolerances; with numerical relays it is typically in the order of 0.2 to 0.3 s and somewhat larger with electromechanical relays, according to each company’s policy.
- Instantaneous: above the maximum fault at the next bus and above the inrush current of the transformers fed.
- Damage curve: the time overcurrent element must stay below the damage curve of the transformer or cable it protects.
When directionality is needed
In a radial network with a single source, non-directional overcurrent is enough. In loops, parallel lines and networks with several sources, including solar plants and on-site generation, fault current can come from either side and directional 67/67N is needed. For ground faults it is also needed with an isolated or resonant-grounded neutral, where the healthy feeders contribute capacitive current. We cover it in zero-sequence directional elements (67N).
How each manufacturer implements it
Every overcurrent relay offers the IEC and IEEE curves, but the number of stages, the setting names, the time dial scale and the reset type (instantaneous, or delayed to emulate the disk of an electromechanical relay) vary. Siemens SIPROTEC is configured with DIGSI, SEL with AcSELerator QuickSet, Relion (670/650: Hitachi Energy; 615/620: ABB) with PCM600, GE Vernova (Multilin) with EnerVista, and Schneider MiCOM and Easergy with their own software. Some manufacturers scale the time dial differently from the standard formula, so a setting is never copied from one brand to another without checking the manual.
Why it trips or fails to trip
| Symptom | Typical cause | What to check first |
|---|---|---|
| Trip when energizing a transformer or starting a motor | Instantaneous set below inrush or starting current | Oscillography, 50 setting, harmonic blocking |
| Ground trip with no ground fault | False 3I0 from unequal CT saturation in a residual connection | Record of the three phase currents |
| Main feeder relay trips before the branch relay | Insufficient coordination margin or curves from different families | Coordination study and actual settings |
| Resistive ground fault not cleared | Ground pickup too high, or set with phase criteria | 51N sensitivity |
| Trip for a fault on another feeder | No directionality after a new source was connected | Fault flow with the new generation |
How it is tested
Secondary injection checks the pickup of each stage, several points of the time curve against the standard formula, the instantaneous time, the reset and the associated logic, such as blocking or acceleration. The test is done per phase and for ground, and compared with the approved settings and the manufacturer’s tolerances.
We review your overcurrent settings
Did your relay trip on overcurrent and the reason is not clear, or do you need to review coordination after connecting a solar plant? At Tiza Group we review the events and settings remotely, for substations and industry at medium, high and extra-high voltage. See remote relay configuration or write to us.
Related guides
References
- IEEE Std C37.2, Standard for Electrical Power System Device Function Numbers, Acronyms, and Contact Designations.
- IEC 60255-151, Measuring relays and protection equipment – Part 151: Functional requirements for over/under current protection.
- IEEE Std C37.112, Standard for Inverse-Time Characteristics Equations for Overcurrent Relays.
- IEC 61850-7-4, logical nodes PTOC and PIOC.
Frequently asked questions
Why is a 51N needed if the phase 51 also sees ground faults?
For sensitivity. The phase 51 must pick up above maximum load; the 51N can be set much lower and detect high-resistance ground faults.
Can time dial settings be copied between relay brands?
Not without checking the formula in the manual, because some manufacturers scale the time dial differently.
When is directional 67/67N protection needed?
In loops, parallel lines and networks with several sources, including solar plants. For ground faults, also with an isolated or resonant-grounded neutral, where the healthy feeders contribute capacitive current.