Design ZnO/MOV surge arresters for transmission systems per IEC 60099. Sweep system voltage, neutral grounding, lightning current and install distance to see, in real time, the arrester rated voltage, MCOV adequacy, residual voltage, absorbed energy and the insulation coordination margin against the transformer BIL.
Parameters
System voltage Um
kV
Neutral grounding
Earthing factor k (TOV factor)
MCOV
kV
Lightning current I_l
kA
8/20μs nominal discharge current
Line propagation delay
μs
Arrester type
Transformer BIL (LIWV)
kV
Arrester-to-transformer distance d
m
Results
—
Phase voltage Um/√3 (kV)
—
Arrester rating Ur (kV)
—
Residual voltage V_res (kV)
—
Protection margin (kV)
—
Energy rating (kJ)
—
Failure mode
—
Power line + arrester + transformer — surge propagation
The path overhead line → arrester (ZnO disc stack) → transformer, with the lightning surge being clipped by the arrester.
Residual voltage Vres is roughly 2.5·Ur at the 8/20μs nominal discharge current. MCOV keeps a 5 % margin over the steady-state phase voltage.
$$E = \tfrac{1}{2}\,I_l\,V_{res}\,\tau \quad[\text{kJ}],\qquad \Delta V = 2\,d\,S\quad[\text{kV}]$$
Absorbed energy E and distance-effect rise ΔV. I_l is the lightning current, d the arrester-to-transformer distance, S the wavefront steepness. Insulation coordination requires BIL > 1.2·(V_res + ΔV).
High-voltage arrester rating & MOV energy — IEC 60099
🙋
A surge arrester is the thing that dumps lightning to earth, right? But why is choosing the "rated voltage" such a big deal? Isn't the only spec that matters how much lightning it can take?
🎓
Good question. The hard part of arrester design is not the lightning, it's making sure the device does absolutely nothing the rest of the time. Inside the housing is a stack of ZnO discs that look like ceramic pucks — an almost perfect insulator below a knee voltage, almost a short circuit above it. Pick Ur too low and the phase voltage already drives a small current through the discs; they heat up and self-destruct within months. Pick Ur too high and the residual voltage V_res at a lightning strike rises past the transformer's BIL. The whole game is the trade-off between temporary-overvoltage (TOV) endurance and protection level.
🙋
I see the required Ur changes a lot when I switch the grounding selector. Why is "solidly grounded" so different from "ungrounded"?
🎓
Because the two systems behave very differently during a single-phase-to-ground fault. In a solidly grounded network the healthy phases barely rise — maybe 1.4·Um/√3 — because the neutral stays near ground. In an ungrounded network the healthy phases jump all the way to the line-to-line voltage, so √3 ≈ 1.73·Um/√3. The arrester has to survive that for several seconds, so you need roughly 25 % more Ur for the same Um. Japan's 6.6 kV distribution is mostly ungrounded, and people who copy IEC specifications blindly often end up over-spec'd or under-spec'd here.
🙋
The residual voltage V_res is ~2.5 times the rated voltage? "A 100 kV-rated device that puts out 250 kV under a strike" feels counter-intuitive…
🎓
That is the magic — and the trap — of ZnO. The V-I characteristic is so nonlinear that pushing the current up five orders of magnitude only raises the voltage by a factor of two-and-a-half. Look at the V-I chart on the right: even on a log scale, the curve is almost flat. That's how the arrester "clamps". So protection design never compares Ur to the transformer rating; it compares V_res to BIL, with a comfortable factor of 1.2–1.4. In rigorous studies you also add the distance effect and any cable contribution before deciding the margin is real.
🙋
The "Failure mode" stat says "MCOV insufficient" right now. What does that actually mean physically, and why does raising the MCOV slider fix it?
🎓
That's the most insidious mistake in the business — the "the arrester exploded after five years" failure. If the MCOV is below the phase voltage Um/√3 the ZnO discs leak a tiny continuous current. It's only microamps at first, so nobody notices; but the leakage heats the disc, the disc gets more conductive at higher temperature, and the loop closes into a thermal runaway. The 154 kV / MCOV 87 kV combination in this default is exactly that pathological case — phase voltage is 89 kV. The IEC rule of thumb MCOV ≥ Um/√3 × 1.05 keeps you safe.
🙋
And the install distance d also matters quite a lot — push it out to 30 m and the margin collapses. What's the physics there?
🎓
A lightning surge is a traveling wave moving down the line at roughly two-thirds the speed of light. When it hits the transformer terminal it reflects, and the reflected wave superposes on the incident wave — the transformer-end voltage almost doubles. If the arrester is far from the transformer it clamps too late: the local voltage at the transformer has already risen by ΔV ≈ 2·d·S (with S the wavefront steepness, 1–2 kV/m). That's why IEEE and IEC guides recommend keeping arresters within 10 m — or just bolting them onto GIS busbars 1–2 m away.
Frequently asked questions
Per IEC 60099, Ur is selected to withstand the temporary overvoltage (TOV) imposed by the system's highest voltage and its grounding configuration. The rule of thumb is Ur = (Um/√3)·k, where k is the earthing factor (≈1.4 for solidly grounded, √3 ≈ 1.732 for ungrounded, ≈1.5 for high impedance). For a 154 kV solidly grounded system, Ur = 88.9·1.4 ≈ 124.5 kV. The same Um with ungrounded neutral demands a higher Ur.
MCOV is the upper limit of the steady-state voltage the ZnO discs can carry indefinitely. If MCOV falls below the system phase voltage Um/√3, the arrester leaks a small but continuous current; the ZnO heats up locally and thermal runaway destroys it within years. A safe rule is MCOV ≥ Um/√3 × 1.05; designs that violate this fail the IEC 60099-4 selection requirement. This tool flags the case as "MCOV insufficient".
Vres is the protective level — the voltage that appears across the arrester during the lightning impulse discharge — and is roughly 2.3 to 2.7 times Ur. The transformer's lightning impulse withstand level (BIL) must exceed Vres plus the distance-effect rise by at least 20% (insulation coordination margin). For BIL = 750 kV, Vres = 311 kV and ΔV = 30 kV, the combined 341 kV gives 750/341 ≈ 2.20 of margin — clearly safe.
Lightning surges propagate as traveling waves and are reflected at the transformer terminal, almost doubling the local voltage. The greater the distance d between arrester and transformer, the larger the extra rise ΔV ≈ 2·d·S (with S the wavefront steepness, roughly 1–2 kV/m). Best practice keeps arresters within 10 m, ideally 5 m, of the transformer terminal; beyond that, BIL margin erodes very quickly.
Real-world applications
Substation main-transformer protection: Every 66–500 kV substation places ZnO arresters at both primary and secondary transformer bushings. During new construction or upgrades, engineers pick Ur, MCOV, V_res and energy class (2–5) per IEC 60099-4 / IEEE C62.11, install the arrester within 10 m of the transformer bushing, and in GIS designs they integrate the arrester directly into the busbar branch. This tool covers the early sizing pass before consulting the manufacturer's catalogue.
Overhead distribution (6.6 kV / 22 kV): Japanese MV distribution is predominantly ungrounded; the TOV during a single-line-to-ground fault is large, and choosing Ur properly matters more than at HV. Arresters protect pole-mounted transformers, sectionalisers and PAS (pole-air switches) from induced surges. Polymer-housed designs dominate today thanks to their light weight and safe pressure-relief behaviour, with leakage-current monitoring increasingly used to flag end-of-life units.
Renewable plant grid interconnection: Wind and solar plants sit on hilltops, off-shore platforms or wide outdoor fields where lightning exposure is intense and the PCS (power-conversion) semiconductors are extremely fragile. A typical design uses staged arresters: grid-side at the step-up transformer, string-side on the DC arrays, plus communications-line SPDs. The DC side requires its own IEC 61643-31 design.
HVDC and gas-insulated switchgear: In HVDC converter stations or 550 kV GIS, the absorbed energy (kJ per kV of Ur) becomes the limiting factor. With kA-class currents lasting hundreds of microseconds — as the "Energy rating" stat shows — total energy can reach thousands of kJ, forcing Class 4 or 5 arresters or parallel banks plus dedicated thermal balance studies.
Common misconceptions & cautions
The biggest trap is treating Ur and MCOV as if they were the same thing. Ur is a short (few-second) TOV-withstand voltage; MCOV is the steady-state limit. The two were not separated in the old SiC era, but IEC 60099-4 makes them distinct for ZnO arresters. A design that satisfies Ur but violates MCOV ≥ Um/√3 × 1.05 looks operational on day one and turns into a slow-motion thermal-runaway bomb. Whenever this tool reports "Failure mode: MCOV insufficient", pick the next MCOV step in the manufacturer's catalogue.
The second is assuming V_res is a single fixed number. The 2.5·Ur figure used here corresponds to the 8/20μs nominal discharge current (about 10 kA). Real datasheets list several V_res values for different waveshapes — 1/20μs steep front, 30/60μs switching impulse and so on — under the labels LIPL (lightning protection level) and SIPL (switching protection level). Detailed design must use the manufacturer's protective-characteristic curve and the V_res that matches the surge waveform of interest.
Finally, "the shorter the install distance, the better" is only half the story. Reducing d lowers the reflection-induced overvoltage, but the inductance of the arrester ground lead (≈1 μH/m) is not negligible. A long ground lead introduces an L·di/dt drop that increases the effective protective level. The real metric is the protective margin that includes both the distance effect and the lead inductance — typically checked in an IEC 60071-2 insulation-coordination study using EMTP. Treat this tool as the first-pass sizing screen, not the final verification.
How to Use
Enter three-phase system voltage (e.g., 230 kV) and maximum continuous operating voltage for the arrester (typically 0.8–0.9 × Um/√3)
Input lightning impulse current magnitude in kiloamperes (e.g., 10 kA) and line propagation delay in microseconds (e.g., 2.5 μs) to model traveling wave effects
The simulator calculates phase voltage Um/√3, selects appropriate arrester rating per IEC 60099-1, computes residual voltage via V–I characteristic, determines protection margin against equipment BIL, and evaluates energy absorption capacity and failure mode
Worked Example
For a 132 kV transmission system: system voltage = 132 kV, maximum continuous voltage = 91 kV (0.87 × 76.2 kV nominal phase), lightning current = 15 kA, line delay = 1.8 μs. Phase voltage Um/√3 = 76.2 kV. Arrester rating selected: 96 kV (IEC class). At 15 kA impulse, residual voltage = 312 kV (metal-oxide disk stack). Protection margin = 450 kV (typical transformer BIL) − 312 kV = 138 kV. Energy rating = 18 kJ. Failure mode: none (margin exceeds 20% threshold).
Practical Notes
ZnO arresters exhibit nonlinear V–I curves; residual voltage scales with log(current). At 1 kA reference, typical ZnO residual ≈ 1.45 × Ur; at 10 kA, slope increases 5–8% per decade
For substations with long transmission lines (>50 km), propagation delay shifts wavefront arrival; use EMTP or ATP to verify switching surge coordination before arrester selection
Neutral-to-ground impedance affects zero-sequence voltage during single-phase faults; ungrounded or high-impedance neutrals can induce 1.7–1.9 × phase voltage transients requiring higher-rated arresters
Thermal stability: verify energy rating against rated short-circuit current duration per IEC 60099-5; MOV disks rated <20 kJ/inch³ risk thermal runaway in high-frequency repetitive surge scenarios (>50 Hz)