Worked example · Device 87T · SEL-787

Proving an 87T Scheme on the SEL-787 — Through-Load, Internal, External, Inrush, Over-Excitation, REF

One scheme, proven on the actual relay. The full SEL-787 proving walk on the default scheme — a worked demonstration of the workspace, not a specific site installation.

Published 2026-08-03 · EI Portal

One 87T scheme. Proven on the actual relay — four relay families modelled. MiCOM P63x, ABB RET/REG, Siemens 7UT86 and the SEL-787 are all vendor-faithful backends. Here is the full SEL-787 proving walk on the default scheme: through-load, internal fault, external fault, inrush, over-excitation and restricted earth fault.

• 90 MVA, 132/33 kV, HV YN clock 0, LV D clock 9
• HV CT 400/1, LV CT 2000/1
• Relay: SEL-787 — percentage-restraint 87R (IOP / IRT), unrestrained 87U, 2nd/4th-harmonic restraint (HRSTR) and cross-phase blocking (HBLK), independent 5th-harmonic over-excitation blocking, plus zero-sequence directional REF
• Factory defaults (SEL-787-4 Group Settings): O87P 0.3, SLP1 25%, SLP2 70%, IRS1 6, U87P 10, PCT2/PCT4 15%, PCT5 35%, HRSTR = Y, HBLK = N

The dual-slope characteristic

The operate threshold is a function of the restraint IRT (the scalar sum of winding magnitudes):

• IRT ≤ IRS1 (6): threshold = max(O87P, SLP1 × IRT) — the shallow 25% slope
• IRT > IRS1 (6): threshold = SLP1 × IRS1 + SLP2 × (IRT − IRS1) — the steep 70% slope

Operate when IOP ≥ threshold and not harmonic-blocked. The two slopes are what let the scheme trip a small internal fault yet stay stable under heavy through-current and CT mismatch.

Through-load → internal → external

• Through-load (stable): IRT 2.0, threshold 0.5, IOP ≈ 0 → STABLE (compensated cancellation)
• Internal fault (operate): IRT 4.38, threshold 1.096 (= 0.25 × 4.382) → IOP > threshold → OPERATE (trip) on the 25% SLP1
• External fault (restrained): IRT 16.26, threshold 8.68 (= 0.25 × 6 + 0.70 × 10.26) → RESTRAINED — the compensated currents cancel (Idiff 0.000); even a real CT error would have to beat 8.68

Inrush and over-excitation held back

• Magnetising inrush: 45% 2nd harmonic > PCT2 15% → restraint raises the threshold to 4.380 and the element rides through (Idiff 1.348 stays below it) → warn
• Over-excitation: 40% 5th harmonic > PCT5 35% → 5H block → restrained (warn)

The differential is held back by the harmonic content, not by the operating point — exactly where a raw biased characteristic would misfire.

The REF catches what the phase 87 misses

A low-level internal earth fault: the main 87 sits below pickup (differential current 0.102 pu < O87P 0.300, so the phase differential does not operate) — but the zero-sequence directional REF operates on the neutral CT (operate 0.800 pu vs its 0.100 threshold) → trip. The restricted earth fault element is what protects the winding near the star point.

Six cases, one story

Six proving cases, one relay, one set of factory defaults: stable on load, operate on the internal, restrained on the external, held on inrush by the restraint-raised threshold, blocked on over-excitation, and tripped by REF where the phase 87 can't see. The dual slope (25% / 70% about IRS1 = 6) is the whole story. Then prove the same scheme on MiCOM P63x, ABB RET/REG, or Siemens 7UT86 — vendor-faithful, manual-backed, function name for function name.