Surges & arresters · Arrester duty

How hard does an arrester work?

An arrester that prevents a flashover still has to survive the stroke. Once its series gap sparks over, the metal-oxide column clamps the voltage at about 300 kV and carries part of the lightning current to the tower. What it absorbs is measured in current, charge and energy, each against a rating. This page follows the arresters on the 150 kV line through direct strikes and shield-wire strikes, single strokes and whole flashes.

Rated voltage
132 kV
EGLA, gap 650 ± 50 mm
Nominal discharge
10 kA
residual 306 kV
Repetitive charge
1.2 C
per single transfer
Thermal charge
6.4 C
total, then cool down
Housing withstand
428 kV
across the arrester
Counter
≥ 100 A
max 5 counts per second

Arrester current

kA

Charge passed

mC, running total

The most loaded arrester against its ratings

Every arrester that conducted

ArresterIpkVresQWCounter

Counter: the paper's screen, a pulse registers if it reaches 100 A and carries at least 2.3 mC.

The whole flash

Several strokes down the same channel

A negative flash usually has 3–5 strokes, tens of milliseconds apart, down the same channel to the same point. Each stroke loads the S arresters again (direct strike, position above). Each single transfer is compared with the repetitive charge rating, and the total with the thermal rating. The counter can register at most one pulse every 0.2 s.

charge per stroke (more loaded S arrester)running totalcountednot counted
What to take away

Six lessons

Within ratings, with margin

For direct strikes up to 15 kA the paper found at most 11.8 kA (at T21), 327 mC and 314 kV: about 1.2 times the nominal discharge current, 27% of the repetitive charge rating, and below the 428 kV housing withstand. Exceeding 10 kA is not a failure; it is the class current used for testing.

The nearer arrester works hardest

At mid-span the two S arresters share the stroke. Move the strike to u = 7/8 and the T21 arrester takes nearly twice the current, and more charge, before the T20 arrester's share arrives.

The tail sets the charge

Charge is the area under the current. A first stroke with the same peak but a long tail carries about three times the charge of a subsequent stroke (paper: 114 against 327 mC at 15 kA). The peak current hardly changes.

Silent counters do not mean no current

A counter registers only pulses above its threshold, and at most five per second. Several strokes of one flash can give one count. On 2 May 2026 the R/T counters at T20 and T21 were unchanged. That fits a direct strike on S, in which the R/T arresters stay silent up to 12.5 kA.

Shield-wire strikes load R/T only lightly

In a backflashover the tower, the footing and both shield wires carry most of the current. Each R/T arrester takes a few kiloamperes and some tens of millicoulombs, even at −96 kA. Their job there is to tie the phase to the tower, not to drain the stroke.

The gap keeps it cool

The series gap holds off the 87 kV phase-to-earth voltage, so the metal oxide carries no continuous current and cannot run away thermally. After the stroke the gap clears, usually within the first half-cycle, without tripping the line.

Model behind the page

Direct strike on S

The Relief Wave Lab model with S arresters at T20 and T21: single conductor, Zc = 535 Ω, 1.5 km distributed loss, Δx = 5 m, θ = 300°. Subsequent stroke: Heidler, peak at 0.7 µs, 14 µs tail, 150 µs window. First stroke: peak at 5.5 µs, 45 µs tail, 300 µs window; the tail is chosen so that 15 kA at mid-span gives about the paper's 327 mC and 93 kJ.

Shield-wire strike

The Backflashover Lab model with R/T arresters at the struck tower, 100 µs window, Δt = 10 ns.

Arrester

Gap closes at 570 kV, then u = 306 kV·(i/10 kA)^(1/14). Charge Q = ∫|i|dt, energy W = ∫|u·i|dt.

Against the paper

Mid-span 10 and 12.5 kA: 4.2–5.4 kA and 60–80 mC per S arrester (paper 4.7–6.1 kA, 68–91 mC). u = 3/4, 10 kA: 6.2 kA, 96 mC (paper 6.6–8.4 kA, 86–115 mC). 15 kA at u = 7/8: 11.4 kA (paper 11.11 kA). Residual voltage 288–309 kV (paper up to 314 kV). Shield-wire strike −96 kA: R/T arresters 4.2 kA (paper 4.7–5.8 kA).

Counter logic and multiple-stroke timing are illustrations. Real counters respond differently to short non-standard pulses, which the paper flags as uncertain.