Transformer Overcurrent Protection: Why Inverse-Time (51) Plus Definite-Time and Instantaneous (50) Stages Matter — A US Engineering Guide

November 12, 2026 | Blog

Part A — Staged Transformer Overcurrent Protection at a Glance

Every distribution and substation transformer in the United States carries some form of overcurrent protection on its high-voltage (HV) side. On a small pad-mount unit that may be a fuse. On a substation transformer it is almost always a microprocessor relay driving a breaker. Modern relays give the protection engineer several independent overcurrent stages: an inverse-time element (ANSI device 51), one or more definite-time elements (device 50 with an intentional delay) and a high-set instantaneous element (device 50 with no delay).

Most relays name these stages 50P1, 50P2, 50P3 and so on. The engineer chooses which stages to enable, how high to set each one and how long each should wait. Done well, staging lets a transformer:

• Carry its load and short-time overloads without nuisance trips.
• Ride through magnetizing inrush every time it is energized.
• Back up the downstream main breaker with a clean coordination margin.
• Clear a fault inside the transformer or on its terminals in a few cycles, before mechanical damage occurs.
• Cut arc-flash incident energy at the secondary terminals, where the hazard is often worst and hardest to reduce.

Done poorly, the same settings trip the transformer every time it is energized, fail to see a real fault, or let a fault on the 480 V terminals burn for two or three seconds.

This guide is Keentel Engineering's practical walk-through of staged transformer overcurrent protection in US practice. It explains:

• What each stage does and the ANSI device numbers behind it.
• The equations behind inverse-time curves.
• How to read the IEEE C57.109 through-fault damage curve.
• How to check inrush, NEC 450.3 limits and coordination time intervals.
• Two complete worked settings: a 1,000 kVA, 12.47 kV–480Y/277 V unit-substation transformer and a 20/27/33 MVA, 115–13.8 kV substation transformer.

In plain terms: the inverse-time stage (51) protects against overloads and slow faults, and waits longer for small currents than for large ones. A definite-time stage (50 with delay) waits a fixed time once current exceeds its pickup, so faults in a known current range clear in a known, short time. The instantaneous stage (50) trips with no intentional delay for currents so large they can only mean a fault on the HV side of the transformer or inside it.

The protection stages side by side

Six findings from Keentel's analysis

• Staging is about time, not only current. The same relay can protect a 1,000 kVA transformer far better with three stages than with one inverse curve. In our first worked example, a single 51 curve clears a 480 V terminal arcing fault in about 2.3 seconds. Adding one definite-time stage brings that down to about 0.7 seconds, a 70% reduction in incident energy.
• The instantaneous stage must never see through the transformer. A pickup that sits just above the symmetrical through-fault current can still trip on the DC offset in the first cycle. The 1.6 factor used in US practice covers that asymmetry.
• Inrush sets the floor for every fast stage. Any stage with less than a few tenths of a second of delay must have a pickup above the inrush point (8 times full-load current for units up to 3 MVA and 12 times for larger units, at 0.1 second), or must be blocked by second-harmonic restraint.
• A pickup below full-load current is a setting error, not a conservative choice. It trips the transformer on normal load and gives no added protection.
• The damage curve is the ceiling. Every stage, taken together, must clear before the IEEE C57.109 through-fault damage curve. For delta–wye units, the curve must be shifted to 58% to cover single-line-to-ground faults on the secondary.
• CT selection is part of the setting. A CT ratio chosen only from full-load current can saturate badly for a primary bus fault, and the fast stages are the ones that suffer.

How Keentel helps here: Keentel Engineering performs protective device coordination studies for transformers from 4 kV through 765 kV. We plot every relay, fuse and low-voltage breaker on one time–current coordination (TCC) chart, using industry-standard coordination software. We check each stage against damage curves, inrush and NEC limits, and issue relay setting files ready to load and test.

Engineering review of common claims

The table below reviews settings and statements often seen in transformer protection discussions. Example values have been converted to a 60 Hz US unit with the same per-unit ratios.

The illustrative systems used in this guide



This guide uses two transformers, both on 60 Hz US systems with standard voltages:

Figure 1 shows the complete coordination for Example 1. The rest of the guide builds it one stage at a time.

Figure 1. Time–current coordination for a 1,000 kVA, 12.47 kV–480Y/277 V transformer, plotted at 12.47 kV. The relay (navy) combines an extremely inverse 51 curve, a definite-time 50P2 stage at 350 A and 0.6 s, and an instantaneous 50P1 at 1,400 A. It stays below the C57.109 damage curve (red, solid and 58%-shifted dashed) and above the inrush point (star). It keeps a 0.25 s margin over the 480 V main breaker's short-time band (green). Illustrative settings only.

Part B — Staged Overcurrent Protection: An Engineering Guide

A transformer's HV overcurrent relay has to satisfy several requirements that pull in different directions:

• It must not trip on load. That includes the top cooling-stage rating, emergency overloads allowed by IEEE C57.91 and cold-load pickup after an outage.
• It must not trip on inrush. Energizing a transformer draws magnetizing current that can reach 8–12 times full-load current for the first few cycles.
• It must coordinate with the secondary main. A fault on a 480 V feeder or a 13.8 kV feeder should be cleared by the downstream device, with the transformer relay waiting as backup.
• It must clear faults before damage. Large through-faults stress the windings mechanically and thermally. The IEEE C57.109 damage curve sets the time limit.
• It should clear terminal faults fast. A fault between the transformer secondary and the secondary main breaker is seen only by the HV relay. On a delta–wye unit-substation transformer this zone often carries the highest arc-flash incident energy in the building.

A single inverse-time curve is a compromise among all of these. To coordinate with the secondary main, it must be slow at the currents a secondary fault produces. That same slowness lets a fault in the secondary throat (the zone the secondary main cannot clear) burn for seconds.

Staging breaks the compromise. The inverse curve handles overload and coordination at lower currents. A definite-time stage takes over at fault levels and holds a fixed, short delay. The instantaneous stage handles the very high currents that can only come from a fault on the HV side.

2. Device functions and terminology

US practice names protective functions with the ANSI/IEEE C37.2 device numbers. International literature often uses different terms for the same characteristics. The mapping below helps when reading relay manuals or settings sheets prepared to another convention.

A microprocessor relay typically offers four or more phase overcurrent stages. Each stage can be set as instantaneous, definite-time or (on some relays) inverse-time. Each stage can also be torque-controlled by logic: blocked by second-harmonic restraint, enabled only in a maintenance mode, or supervised by a directional element.

In plain terms: "IDMT" and "51" describe the same thing. So do "DMT" and "50 with a time delay". What matters is the characteristic, the pickup and the time, not the label.

3. Inverse-time curves in US practice

3.1 The IEEE C37.112 equation

IEEE Std C37.112 defines the inverse-time characteristic used by US microprocessor relays:

where t is the operating time in seconds, TD is the time dial, and M is the multiple of pickup current:

A, B and p are constants that set the curve's shape. The higher the exponent p, the more steeply the curve falls as current rises.

The U-curve labels and constants above are used by many US relays and follow the electromechanical relay shapes. IEEE C37.112 publishes its own standard moderately inverse, very inverse and extremely inverse constants (for example A = 28.2, B = 0.1217, p = 2 for extremely inverse). Always confirm the constants in the relay manual before modeling, because the same name can carry different constants.

3.2 How the curves compare

The table gives operating times at time dial 1 for each U-curve:

The extremely inverse U4 curve is about 20 times faster at 10 times pickup than at 2 times pickup. That shape matches the I²t behavior of fuses, low-voltage breaker long-time elements and the transformer's own thermal damage curve. It is why U4 (or the C37.112 extremely inverse curve) is the most common choice for transformer HV relays that must coordinate with a 480 V main breaker.

3.3 International curves sometimes found in US relays

Many relays also offer the IEC 60255-151 curves. The IEC extremely inverse curve is:

At a time multiplier (TMS) of 2.74 and 10 times pickup, it gives 2.74 × 80/99 = 2.21 seconds. That is slow for a transformer relay in a US unit substation. If a settings file uses IEC curves, convert them carefully: TMS and TD are not interchangeable, and neither are the curve constants.

3.4 Reset characteristics

Electromechanical induction-disk relays reset slowly after current drops, because the disk must rotate back. Microprocessor relays can emulate that with the C37.112 reset equation:

4. The transformer through-fault damage curve

4.1 IEEE C57.109 categories

IEEE Std C57.109 defines through-fault capability curves for liquid-immersed transformers. It sorts transformers into four categories by minimum nameplate rating:

"Frequent faults" means more than about 10 (Category II) or 5 (Category III) through-faults over the transformer's life. Transformers serving overhead lines or industrial loads should normally be treated as frequent-fault units. Transformers feeding cable systems that rarely fault can use the infrequent-fault (thermal-only) curve.

4.2 The thermal curveA

The thermal portion follows a constant I²t, with current in per unit of base-rated current:

It is anchored at its low-current end by tabulated points: 2 pu for 1,800 seconds, 3 pu for 300 seconds and 4.75 pu for 60 seconds. Above 4.75 pu the I²t = 1250 line applies.

4.3 The mechanical curve

For Categories II and III with frequent faults, a mechanical curve applies at high currents. It also follows constant I²t, but is anchored at 2 seconds at the maximum through-fault current:

Here Z_t is the per-unit impedance (transformer plus source for Categories III and IV; transformer only for Category II in the strict reading, although many engineers include the source for a more accurate curve) and I is the current in per unit. At I = 1/Z_t, the maximum through-fault current, t = 2 seconds.

The mechanical curve begins at 70% of maximum fault current for Category II and at 50% for Category III. At that point, the curve steps down from the thermal curve, which is why damage curves show a vertical jump.

4.4 The 58% shift for delta–wye transformers

On a delta–wye transformer, a single-line-to-ground fault on the wye secondary produces a primary line current of only 1/√3, or 0.577, of the per-unit winding current. The HV relay sees less current than the winding actually carries.

To make sure the HV relay still protects the winding, the damage curve plotted on the primary side is shifted left to 58% of its current values for this case. Figure 1 shows both curves.

Worked Example 1 — Damage curve for the 1,000 kVA transformer

Full-load current on the 12.47 kV side is:

On the 480 V side, it is 1,202.8 A.

Including the 250 MVA source (0.004 pu on a 1 MVA base), the total impedance is:

so the maximum through-fault current is 1/0.0615 = 16.26 pu, or 752.8 A at 12.47 kV.

For a secondary ground fault, every current in the table is multiplied by 0.58. For example, the 2-second point moves from 752.8 A to about 437 A. The HV relay must clear below both curves.

5. Magnetizing inrush and cold-load pickup

5.1 The inrush point

When a transformer is energized, its core can be driven into saturation, drawing a magnetizing current many times rated current. The current is highly asymmetric, rich in second harmonic and decays over a few tenths of a second to several seconds.

For coordination, US practice represents inrush by one point on the TCC:

• 8 times full-load current at 0.1 second for transformers up to about 3 MVA.
• 12 times full-load current at 0.1 second for larger transformers.

Some engineers add a second point at 25 times full-load current at 0.01 second. Every overcurrent stage must plot above and to the right of the inrush point.

For the 1,000 kVA example:

5.2 Second-harmonic blocking

Inrush current typically contains 15–60% second harmonic, while fault current contains very little. Most transformer relays can block or restrain fast overcurrent stages when second harmonic exceeds a threshold (commonly 15–20% of fundamental).

Second-harmonic blocking lets a definite-time or instantaneous stage be set below the inrush point. Modern low-loss cores with high-permeability steel can produce inrush with less than 15% second harmonic, so blocking thresholds should be checked against energization records or EMT studies.

5.3 Sympathetic inrush and cold-load pickup

Two related events also need attention:

• Sympathetic inrush: energizing one transformer can drive a parallel, already-energized transformer into saturation. Both relays see inrush-like currents that decay slowly.
• Cold-load pickup: after an outage, thermostatically controlled loads all restart at once. Current can stay at 2–3 times normal load for several seconds to minutes.

A 51 pickup at 125–200% of rated current, with an extremely inverse curve, usually rides through cold load. A definite-time stage set just above full-load current would not.

How Keentel helps here: where inrush or sympathetic interaction is uncertain, we run EMT energization studies with realistic residual flux and point-on-wave closing. The results set harmonic-blocking thresholds and definite-time pickups from data, not from rules of thumb.

6. NEC limits for transformer overcurrent protection

For transformers over 1000 V, NEC Section 450.3(A) and Table 450.3(A) (2023 edition; earlier editions used 600 V as the threshold) set the maximum rating or setting of overcurrent protection, as a percentage of transformer rated current.

In supervised locations (where conditions of maintenance and supervision ensure only qualified persons service the installation), primary-only protection is allowed at 300% (breaker) or 250% (fuse). Notes to the table allow the next higher standard rating or setting in some cases and permit a higher secondary setting where a differential relay protects the transformer.

For relay-operated breakers, "setting" means the pickup of the time-overcurrent element. For Example 1:

Two other NEC requirements shape the 480 V side:

• Section 240.87 (arc energy reduction): a breaker rated or adjustable to 1,200 A or more must have a method to reduce clearing time, such as zone-selective interlocking, differential relaying, an energy-reducing maintenance switch or an active arc-flash mitigation system.
• Sections 230.95 and 215.10 (ground-fault protection of equipment): a solidly grounded wye service or feeder above 150 V to ground and not over 1000 V phase-to-phase, with a disconnect rated 1,000 A or more, needs ground-fault protection.

Neither requirement covers the zone between the transformer secondary bushings and the secondary main. That zone is protected only by the HV relay, which is why its staging matters so much.

How Keentel helps here: our coordination studies include an NEC 450.3, 240.87 and 230.95 compliance table for every transformer, so the authority having jurisdiction and the owner can see each limit and the setting that meets it.

7. Fault currents that set the stages

Worked Example 2 — Fault currents for the 1,000 kVA transformer

The source contributes 0.004 pu on the 1 MVA base (250 MVA available). A bolted three-phase fault on the 480 V terminals gives:

With an infinite source, the same fault gives 20,918 A. The current the HV relay sees for that fault is reflected through the turns ratio:

The full set of currents used for the settings:

Using the transformer's nameplate impedance minus its tolerance (typically 7.5% per IEEE C57.12.00 for two-winding units) gives a slightly higher through-fault current. A careful 50P1 setting uses that lower bound.

7.1 What the arcing fault means for staging

The arcing current reflected to the HV side (about 414 A) is barely more than half the bolted through-fault current. On a single extremely inverse curve that is coordinated with the secondary main, 414 A falls in the slow part of the curve. That is the gap a definite-time stage closes.

8. Current transformer selection

The CT ratio decides whether the relay can see the currents it is set for. IEEE C37.110 gives the rule for avoiding saturation on symmetrical fault current:

and, to avoid saturation even with full DC offset:

where I_f,sec is the secondary fault current, Z_b the total burden (leads, relay and CT winding) and V_x the CT's C-class knee-point rating.

Worked Example 3 — Choosing a CT ratio for the 12.47 kV relay

A 100:5 CT looks natural for a 46 A full-load current. But a 12.47 kV bus fault of 11,575 A would drive 579 A secondary, 116 times the CT rating. The CT would saturate heavily, and the fast stages are the ones that suffer.

A 600:5 multi-ratio CT, C200, on its full tap gives:

With 0.5 Ω total burden, the symmetrical check gives 96.5 × 0.5 = 48 V, well below 200 V. The full-offset check (X/R of 10) gives about 530 V, so DC saturation is possible in the first cycles. That is normally acceptable for instantaneous elements in microprocessor relays, which measure fundamental current and operate before deep saturation. It should be confirmed against the relay manufacturer's CT requirements.

The 0.58 A pickup must fall within the relay's setting range for a 5 A nominal input. Most US relays accept it, but confirm.

In plain terms: pick the CT from the fault current, then check the pickup range. Picking it from load current alone is a common cause of slow or failed instantaneous trips.

9. Setting each stage

9.1 The 51 inverse-time stage

The 51 stage is set in three steps:

• Pickup: 125–200% of the transformer's maximum rated current (top cooling stage). This must also meet NEC 450.3, sit above cold-load current and leave room for permitted overloads.
• Curve: usually extremely inverse, to track the secondary main's long-time band and the damage curve.
• Time dial: the lowest value that keeps the required coordination time interval (CTI) over the downstream device at every current where both operate.

The CTI between a relay-operated breaker and a downstream low-voltage breaker is usually 0.2–0.3 second. It covers breaker clearing time, relay overtravel (zero in microprocessor relays) and setting and CT errors. Between two relay-operated breakers, 0.2–0.3 second is common; with electromechanical relays, 0.3–0.4 second.

9.2 The 50P2 definite-time stage

The definite-time stage is the one that most improves protection on small and mid-size transformers. It is set by:

• Pickup: above the inrush point (or below it, if second-harmonic blocking is enabled and verified) and above cold-load current. It should also be below the minimum reflected arcing fault current at the secondary terminals.
• Delay: the secondary main's short-time clearing time plus the CTI.

9.3 The 50P1 instantaneous stage

The instantaneous stage is set by one rule: it must never operate for a fault on the secondary side. US practice sets its pickup at least 1.6 times the maximum symmetrical through-fault current (some engineers use 1.7–2.0, or 1.25 with relays that filter DC offset). It must also exceed the inrush current at 0.01–0.1 second and still pick up reliably for an HV terminal fault at minimum source strength.

Worked Example 4 — Complete settings for the 1,000 kVA transformer

The downstream device. The 480 V main breaker has a trip unit with long-time pickup 1,250 A, long-time delay 4 s at 6 times pickup (I²t, +15% tolerance band) and short-time pickup 4,000 A at 0.3 s (clearing by 0.35 s). Its instantaneous function is off so the feeders below can coordinate.

Step 1 — 51 stage.

• Pickup: 70 A, which is 151% of full-load current and well under the NEC 600% limit of 277.8 A.
• Curve: U4 extremely inverse.
• Time dial: the main breaker's long-time band ends at its short-time pickup (4,000 A, which reflects to 154 A at 12.47 kV). The closest approach is near that corner. A time dial of 10.9 keeps the 51 curve above the main's upper long-time band everywhere: 59.8 s against 38.3 s at 100 A and 17.6 s against 17.0 s at 150 A (the tightest point, where margin is also set by the steeper slope of the relay curve).

Step 2 — 50P2 stage.



• Pickup: 350 A (7.6 times full-load current). It sits below the reflected arcing current of 414 A and essentially at the 85% arcing variation of 352 A.
• Delay: 0.6 s. The main breaker's short-time band clears by 0.35 s, so the CTI is 0.25 s.
• Inrush check: at 370 A, the relay waits 0.6 s, well above the 0.1 s inrush point. Inrush current falls quickly below 350 A for a transformer this size, but if energization records show a slow decay, enable second-harmonic blocking on 50P2.

Step 3 — 50P1 stage.

• Pickup: 1,400 A. That is 1.86 times the 752.8 A through-fault and above the 1,288 A worst-case asymmetric through-fault with an infinite source.
• Delay: none (relay operating time about 0.02 s, plus breaker).
• Dependability: the 12.47 kV bus fault of 11,575 A is 8.3 times pickup.

The resulting operating times:

Damage check.

At the maximum through-fault of 752.8 A, the relay clears in 0.6 s plus breaker time, against 2.0 s on the damage curve. At the 58%-shifted 2-second point (437 A), it clears in the same 0.6 s. Across the 51 range, the relay curve stays below the shifted thermal curve.

What the definite-time stage buys.

 With 51 alone, the relay would take 0.85 s at the maximum through-fault and 2.67 s at the inrush current. Adding 50P2 leaves the overload and coordination behavior unchanged and cuts the time for every fault between 350 A and 1,400 A to 0.6 s.

How Keentel helps here:



we deliver settings like these as a complete package: the coordination study report, annotated TCC plots, the relay setting file in the manufacturer's format, a settings summary table and a test plan for the field technician.

10. Arc-flash incident energy and staging

Under IEEE 1584-2018, incident energy at a given working distance is close to proportional to arcing time. For a fault on the 480 V transformer terminals or the line side of the secondary main, the arcing time is the HV relay's operating time plus the HV breaker's interrupting time.

Worked Example 5 — Arc-flash benefit of the 50P2 stage

For the 1,000 kVA example, the 480 V arcing current is about 10.76 kA, which reflects to about 414 A at 12.47 kV. With a 5-cycle (0.083 s) HV breaker:

The definite-time stage cuts arcing time, and therefore incident energy, by about 70%. IEEE 1584-2018 caps the arc duration used in calculations at 2 seconds where a worker can reasonably escape, but that cap is a judgment call, not a protection scheme.

The margin between the 85% arcing current (352 A) and the 50P2 pickup (350 A) is thin. Several options can widen it:

• Lower 50P2 with second-harmonic blocking, after confirming inrush harmonic content.
• Add a maintenance-mode stage, such as a 50P3 at 250 A with 0.05 s delay, enabled only while workers are present, through a selector switch or SCADA bit with status indication.
• Add transformer differential (87T) covering the secondary throat, or an arc-flash detection relay using light plus current.
• Use a low-impedance secondary main with zone-selective interlocking, which helps downstream equipment but not the throat itself.

How Keentel helps here: Keentel performs arc-flash studies to IEEE 1584-2018 and NFPA 70E, with labels, a mitigation report and before-and-after incident energy for each option. We often find that one added relay stage removes the highest-energy labels in a facility at almost no hardware cost.

11. Staging on a larger substation transformer



Worked Example 6 — 20/27/33 MVA, 115–13.8 kV transformer

Ratings and currents.

Primary protection. At this size the transformer has 87T differential and a 63 sudden-pressure relay, both tripping an 86 lockout. The 50/51 relay on the 115 kV breaker is backup.



Downstream device. The 13.8 kV main breaker relay uses a 51 at 1,600 A (116% of the 33 MVA rating) on a 2000:5 CT, U3 very inverse, time dial 2.0. It operates in 0.52 s for a 13.8 kV bus fault.

HV relay settings (CT 600:5).

• 51 pickup: 1.5 × 165.7 A = 248.5 A (2.07 A secondary), U3 very inverse to match the 13.8 kV main.
• Time dial: at the 956 A through-fault, the main operates at 0.52 s, so the HV relay needs at least 0.82 s. U3 at TD 1 gives 0.377 s at M = 3.85, so TD = 2.2 gives 0.83 s.
• 50P1 pickup: at least 1.6 × 956 A = 1,530 A. Choose 1,800 A, which is also above the 1,205 A inrush point.
  

Damage check. At 478 A, the mechanical curve allows 8 s and the relay clears in about 3.4 s. At 956 A, the curve allows 2 s and the relay clears in 0.83 s. With the 58% shift for a 13.8 kV ground fault, the 2-second point moves to about 555 A, where the 51 operates in about 2.4 s. That is the one place the backup overcurrent relay does not fully protect the transformer on its own, and it is covered by 87T and the 13.8 kV ground relays.

Would a definite-time stage help? Here, very little. The secondary-side fault current range is narrow (the through-fault is only 3.85 times the 51 pickup), the 13.8 kV main's own operating time sets the floor, and 87T already clears internal and throat faults in about 1.5 cycles. The definite-time stage earns its place mainly on transformers without differential protection, which in US practice usually means units below about 10 MVA.

In plain terms: on a small unit substation, the HV overcurrent relay is the transformer's main protection, and staging is the cheapest way to make it fast. On a large substation transformer, differential relaying does that job, and overcurrent staging is mostly about clean backup.

12. Ground-fault protection and winding connections

The winding connection decides what a ground relay on each side can see:

• Delta primary: a ground fault on the grounded-wye secondary produces no zero-sequence current on the delta side. The HV phase relays see it only as reduced phase current (the 58% case). An HV 51N on a delta winding sees only ground faults on the HV system and the HV leads.
• Grounded-wye secondary: a neutral CT (51G) on the transformer neutral sees every secondary ground fault, including faults in the throat between the transformer and the secondary main. On a 480 V system it should coordinate with the main's ground-fault (G) function.
• Resistance-grounded secondary: on low-resistance grounded 4.16 kV or 13.8 kV systems, the neutral resistor limits ground-fault current (often to 200–400 A at 4.16 kV and up to about 2,000 A at 13.8 kV). The 51G must be set below the resistor's let-through current, coordinated with feeder ground relays and fast enough to protect the resistor's time rating.
• Grounded-wye to grounded-wye: zero-sequence current passes through, so ground relays on both sides interact and must be coordinated together.



A neutral 51G with a definite-time stage set at 20–30% of the resistor or available ground-fault current, with a delay of 0.3–0.5 s above the secondary main's ground function, is a common and effective way to protect the throat for ground faults.

How Keentel helps here: we review grounding and ground-fault protection as one system, from transformer neutral to feeder ground relays and NEC 230.95 ground-fault protection of equipment, including resistor ratings and alarm-versus-trip decisions for high-resistance grounded systems.

13. Where NERC requirements apply

For transformers that are part of the Bulk Electric System (BES) or that serve BES generation, several NERC Reliability Standards touch overcurrent staging:

Whether a particular 115–13.8 kV distribution transformer falls in scope depends on the BES definition and the registered entity's documented decisions. A distribution transformer is usually excluded, but its relays can still be in PRC-005 scope if they trip BES elements. Confirm applicability with the transmission owner and the compliance group.

How Keentel helps here: Keentel supports PRC-027 coordination reviews, PRC-005 program documentation, PRC-023 loadability checks and PRC-019/PRC-024 generator protection studies, with evidence packages organized the way auditors expect.

14. Testing and commissioning the stages

Good settings are only as good as their implementation. A site acceptance test (SAT) for a staged overcurrent relay should include:

• Settings verification: compare the loaded settings file against the approved settings, line by line, including CT ratios, curve types, time dials and logic.
• Pickup tests: inject current at each stage's pickup and confirm operation within the stated tolerance (typically ±5%).
• Timing tests: inject at least three multiples on the 51 curve (for example 2, 5 and 10 times pickup) and confirm operating time against the C37.112 equation. Inject above the 50P2 and 50P1 pickups and confirm their fixed times.
• Stage interaction: confirm that 50P2 overrides the slower 51 curve above its pickup and that 50P1 overrides both.
• Harmonic blocking: inject fundamental plus second harmonic above and below the blocking threshold and confirm the expected behavior.
• CT checks: ratio, polarity, insulation resistance, excitation (saturation) curve and secondary burden.
• Trip path: functional trip of the breaker from each stage, through any lockout, with breaker clearing time recorded.
• In-service check: once energized, verify phase currents and angles at the relay against expected load.

The first energization is also a free inrush test. Capture the event record and confirm the measured inrush magnitude, decay and second-harmonic content against the assumptions used in the settings.

How Keentel helps here: our field engineers perform relay SAT and commissioning for MV switchgear and substation transformers, with test reports, as-left settings files and PRC-005 baseline records.

15. Settings checklist

Before issuing settings for a transformer HV overcurrent relay, confirm:

• Ratings: top cooling-stage rating, impedance and tolerance, winding connection and C57.109 category.
• Fault study: maximum and minimum source strength, secondary three-phase and ground faults, HV bus faults, arcing currents per IEEE 1584-2018.
• CTs: ratio, accuracy class and burden checked per IEEE C37.110 against maximum fault current.
• 51: pickup within 125–200% of maximum rated current and within NEC 450.3; curve and time dial set from CTI against the secondary main across its range.
• 50P2: pickup above inrush (or harmonic blocking verified) and below the minimum reflected arcing current; delay set from the secondary main's short-time band plus CTI.
• 50P1: pickup at least 1.6 times the maximum through-fault current and above inrush; still dependable for an HV terminal fault at minimum source.
• Damage curve: all stages below the C57.109 curve, including the 58% shift for delta–wye units.
• Ground: 51N/51G settings match the winding and grounding arrangement.
• Arc flash: incident energy calculated with the final settings, labels updated, NEC 240.87 method documented.
• Compliance: NERC applicability checked and records prepared.
• Testing: test plan issued with each stage's pickup and timing points.
  

Part C — Case Studies

The three case studies below are drawn from typical project experience. Names, locations and identifying details have been removed, and values have been rounded.

Case Study 1 — A unit substation that tripped on every energization

Situation. A manufacturing plant in the Southeast had a 2,000 kVA, 13.2 kV–480Y/277 V unit-substation transformer that tripped its primary breaker roughly one time in three on energization. Each trip meant a delayed restart and a call to the electrical contractor.

Findings

The relay had been set by copying a template from another site. Its definite-time stage was set at 3 times full-load current with 0.1 s delay, well below the 8-times inrush point. The 51 pickup was 90% of full-load current, so the relay was also timing on heavy production loads.

Solution

 Keentel rebuilt the coordination from a new fault study. The 51 pickup moved to 150% of full-load current with an extremely inverse curve. The definite-time stage moved above inrush with a delay coordinated to the 480 V main's short-time band, and second-harmonic blocking was enabled with a threshold confirmed from an event record. The instantaneous stage was raised to 1.7 times the through-fault current.

Result

 The transformer has energized without a trip since the change. The arc-flash study, rerun with the new settings, confirmed that incident energy at the 480 V switchgear line side stayed within the existing PPE category, because the corrected definite-time stage still clears terminal arcing faults in well under a second.

Case Study 2 — Reducing arc-flash energy at a data center electrical room

Situation

 A colocation data center in the Mid-Atlantic had eight 2,500 kVA, 12.47 kV–480Y/277 V transformers feeding switchboards. The arc-flash study showed incident energy above 40 cal/cm² on the line side of every 480 V main, driven by the HV relays' extremely inverse curves, which took more than 2 seconds to clear a terminal arcing fault.

Findings

The HV relays had only 51 and 50P1 enabled. The instantaneous stage could not see the secondary arcing fault, and the 51 curve had been set slow to coordinate with the 480 V mains' long-time bands.

Solution

Keentel added a definite-time stage to each relay, set between inrush and the minimum reflected arcing current, with a delay coordinated to the mains' short-time bands. Because the margin at 85% arcing current was tight, a maintenance-mode stage with near-instantaneous operation was also added, enabled by a labeled selector switch with status shown on the facility's monitoring system. The NEC 240.87 documentation was updated to describe both methods.

Result

 Normal-mode incident energy at the 480 V line side dropped by roughly two-thirds, and maintenance-mode energy fell below 8 cal/cm². No hardware other than selector switches and wiring was required.

Case Study 3 — Backup protection review for a 138 kV substation transformer

Situation

A municipal utility in the Midwest owned a 30/40/50 MVA, 138–13.8 kV transformer whose 138 kV breaker relay was flagged in a PRC-027 coordination review after a nearby generation addition raised fault levels.

Findings

The higher source strength raised the maximum through-fault current enough that the instantaneous stage's margin had dropped to 1.2 times the symmetrical through-fault current. A 13.8 kV bus fault with high DC offset could have tripped the 138 kV breaker instead of the 13.8 kV main. The 51 curve was also above the 58%-shifted damage curve over part of its range.

Solution

 Keentel updated the short-circuit model, raised the 50P1 pickup to 1.7 times the new through-fault current, changed the 51 curve and time dial to restore the CTI with the 13.8 kV main, and confirmed that 87T covered the remaining damage-curve gap. The study package included the PRC-027 documentation and updated PRC-005 test records.

Result

The utility closed the coordination finding with a complete evidence package and avoided a misoperation risk that would have dropped the entire 13.8 kV substation for a single feeder fault.

Part D — 20-Question Technical FAQ

How Keentel Engineering Helps

Keentel Engineering is a US power system engineering firm that designs, studies and commissions protection for utilities, independent power producers, data centers, industrial plants and campuses, from 4 kV distribution to 765 kV transmission. For transformer protection, our services include:

• Protective device coordination studies: fault studies and TCC coordination using industry-standard coordination software, covering every relay, fuse and low-voltage trip unit from the utility source to the largest downstream feeder.
• Relay settings development: complete settings for 50/51, 50N/51N, 51G, 87T, 63 and lockout logic, delivered as manufacturer-format files with a settings summary and rationale.
• Arc-flash studies: IEEE 1584-2018 calculations, NFPA 70E labels, NEC 240.87 documentation and mitigation options with before-and-after incident energy.
• Short-circuit and EMT studies: maximum and minimum fault current, DC offset, CT saturation and transformer energization studies for inrush and sympathetic interaction.
• NERC compliance support: PRC-027 coordination reviews, PRC-005 program and test records, PRC-023 loadability checks and PRC-019/PRC-024 generator protection studies.
• SAT and commissioning: relay and CT testing, trip path verification and as-left records for MV switchgear and substation transformers.
• Owner's engineer services: settings review for utility interconnections, design review of protection and control packages and independent verification of contractor settings.
• Substation and switchgear design: protection and control design, one-lines, three-lines and relay panel layouts for AIS, GIS and metal-clad switchgear.

Why Clients Choose Keentel

• Study, settings and field work under one roof: the engineers who run the coordination study also write the settings and support the testing.
• US standards fluency: daily work with IEEE, NEC, NFPA 70E, NESC and NERC requirements.
• Independent advice: we are not tied to any relay or equipment supplier, so recommendations follow the engineering.
• Practical deliverables: clear TCCs, auditable calculations, ready-to-load settings files and test plans.
• Full voltage range: from 4.16 kV plant distribution to 765 kV transmission substations.

Talk to an Engineer

Is a transformer tripping on energization, carrying high arc-flash labels or facing a PRC-027 review? Keentel Engineering can review your settings and deliver a coordinated, documented solution.

• Phone: (813) 389-7871
• Email: contact@keentelengineering.com
• Book a 15-minute call: calendly.com/keentel-engineering/15min
• 
  

References

• IEEE Std C37.112, IEEE Standard for Inverse-Time Characteristics Equations for Overcurrent Relays.
• IEEE Std C37.91, IEEE Guide for Protecting Power Transformers.
• IEEE Std C57.109, IEEE Guide for Liquid-Immersed Transformers Through-Fault-Current Duration.
• IEEE Std C57.12.00, IEEE Standard for General Requirements for Liquid-Immersed Distribution, Power, and Regulating Transformers.
• IEEE Std C57.91, IEEE Guide for Loading Mineral-Oil-Immersed Transformers and Step-Voltage Regulators.
• IEEE Std C37.110, IEEE Guide for the Application of Current Transformers Used for Protective Relaying Purposes.
• IEEE Std C37.2, IEEE Standard for Electrical Power System Device Function Numbers, Acronyms, and Contact Designations.
• IEEE Std C37.230, IEEE Guide for Protective Relay Applications to Distribution Lines.
• IEEE Std 242 (Buff Book), IEEE Recommended Practice for Protection and Coordination of Industrial and Commercial Power Systems.
• IEEE Std 3004.5, IEEE Recommended Practice for the Application of Low-Voltage Circuit Breakers in Industrial and Commercial Power Systems.
• IEEE Std 1584-2018, IEEE Guide for Performing Arc-Flash Hazard Calculations.
• NFPA 70, National Electrical Code, 2023 edition: Sections 215.10, 230.95, 240.87 and 450.3.
• NFPA 70E, Standard for Electrical Safety in the Workplace.
• NERC Reliability Standards PRC-005, PRC-019, PRC-023, PRC-024, PRC-027, MOD-032 and MOD-033.
• Keentel Engineering, companion technical blogs on transformer differential protection (87T), 50N/51N ground overcurrent protection and protection relay SAT for MV switchgear.

Disclaimer

This article is for general technical information only. Worked examples use illustrative values and simplified assumptions; they are not settings for any specific installation. Curve constants, CT requirements and setting ranges vary by relay model, so always confirm them in the manufacturer's documentation. Standards are cited as of publication and may be revised; confirm the current edition and its adoption by the authority having jurisdiction. Protection settings must be developed and approved by qualified engineers using site-specific studies.

© Keentel Engineering