Technical Methodology for Energy and Charge Dimensioning of Metal-Oxide Surge Arresters in Modern Power Systems
- Florent Giraudet
- Jan 19
- 9 min read
1.Introduction and Theoretical Foundations
The protection of electrical power systems against transient overvoltages is a discipline that sits at the intersection of electromagnetic field theory, material science, and thermodynamic engineering. At the core of this discipline lies the Metal-Oxide Surge Arrester (MOSA), a device whose non-linear voltage-current (U - I) characteristics allow it to act as a dynamic insulation barrier. However, the correct application of these devices requires more than merely selecting a voltage rating; it necessitates a rigorous dimensioning methodology for energy absorption and charge transfer.
Recent evolutions in international standards, specifically the transition within IEC 60099-4, the introduction of IEC 60099-9 for HVDC, and the harmonization of line surge arrester requirements in IEC/IEEE 60099-11, have fundamentally altered the landscape of arrester sizing. The historic reliance on "Line Discharge Classes" (LDC) has been superseded by a more precise, physics-based classification system relying on Repetitive Charge Transfer Rating (Qrs), Thermal Charge Transfer Rating (Qth), and Thermal Energy Rating (Wth).
This report articulates a comprehensive methodology for selecting these ratings based on transient studies. It addresses the critical engineering challenge of distinguishing between the single-event withstand capability (mechanical integrity of the varistor microstructure) and the multi-event thermal stability (thermodynamic equilibrium of the arrester assembly). A central theme of this analysis is the identification and rectification of a pervasive industry error: the direct comparison of single-event simulation results to thermal energy ratings (Wth), a practice that can lead to problematic under-dimensioning of protective equipment.
1.1 The Microstructural Physics of Energy Absorption
To understand the necessity of the modern charge-based ratings, one must first appreciate the physical limitations of the Zinc Oxide (ZnO) varistor or often called Metal-Oxide Varistor (MOV). The MOV block is a ceramic semiconductor composed of conductive ZnO grains surrounded by highly resistive intergranular layers (Schottky barriers).
When a surge occurs, the current flows through these grain boundaries, generating heat.
This energy absorption manifests in two distinct failure modes:
Impulse Degradation (Cracking/Puncture): If the energy or charge is injected too rapidly (a single high-current impulse), the adiabatic heating of the localized current paths creates differential thermal expansion. This mechanical stress can fracture the ceramic block or cause dielectric puncture before the heat has time to distribute to the surrounding housing. This limit is characterized by Qrs.
Thermal Instability (Runaway): If the energy is injected over a longer period or in multiple accumulation events (e.g., two switching surges), the bulk temperature of the block rises. If the temperature exceeds a critical threshold (typically around 170°C - 200°C), the leakage current at the continuous operating voltage (Uc) increases exponentially, generating more heat than the housing can dissipate to the ambient air. This leads to thermal runaway and eventual failure. This limit is characterized by Wth and Qth.
The fundamental error in legacy dimensioning methods was treating "Energy" as a homogenous scalar quantity, ignoring the rate of injection and the cooling intervals. The modern standards rectify this by decoupling the single-shot mechanical limit from the multi-shot thermal limit.
2.Methodology for Distribution Class Arresters (IEC 60099-4)
Distribution systems, typically operating at voltages Us > 52 kV, present a unique challenge due to their high exposure to direct and induced lightning currents. The dimensioning of arresters for these applications (Pole-mounted transformers, riser poles, cable terminations...) is strictly governed by the charge content of the lightning flash.
2.1 The Transition to Charge Ratings (Qrs and Qth)
Under the current edition of IEC 60099-4, distribution arresters are classified into three distinct categories: Distribution Low (DL), Distribution Medium (DM), and Distribution High (DH). These classifications are no longer solely defined by the nominal discharge current (In) but are rigidly linked to their charge handling capabilities.
Class | Designation | Nominal Discharge Current (In) | Minimum Qrs Rating (Single Impulse) | Minimum Qth Rating (Thermal Stability) | Typical Application |
Distribution Low | DL | 2.5 kA or 5 kA | 0.1 C | 0.45 C | Low lightning activity, shielded cable risers |
Distribution Medium | DM | 5 kA or 10 kA | 0.2 C | 0.70 C | Standard overhead distribution lines |
Distribution High | DH | 10 kA | 0.4 C | 1.10 C | High lightning density, unshielded lines, high reliability requirements |
Table 1: Classification of Distribution Class Arresters based on Charge Ratings.
Important note: Qrs and Qth ratings for distribution-class arresters are tested using standardized 8/20 µs lightning impulse waveforms.
2.2 Dimensioning Methodology: The Single Flash Criterion
The primary objective of a distribution transient study is to quantify the charge transferred through the arrester during a lightning event. This process requires a sophisticated integration of the current waveform rather than a simple peak current measurement.
2.2.1 Step 1: Characterizing the Lightning Event
A comprehensive study must utilize statistical distributions of lightning parameters (e.g., CIGRE or IEEE 1410 data). A negative downward lightning flash is complex, consisting of:
First Return Stroke: High amplitude, relatively short duration.
Subsequent Strokes: Fast rise times, intermediate amplitudes.
Continuing Current: Lower amplitude (hundreds of Amperes) but very long duration (milliseconds).
The simulation must capture the total cumulative charge of this composite event. The arrester on a distribution pole sees this entire sequence as a single thermal/mechanical stress event because the time interval between strokes (60 ms on average) is insufficient for any meaningful cooling.
2.2.2 Step 2: Selection Against Qrs
The critical dimensioning rule is to compare the Total Charge of a Single Flash (Qflash) against the Repetitive Charge Transfer Rating (Qrs).
Rationale: The Qrs rating represents the maximum charge the MOV blocks can withstand in a single injection (or group of closely spaced injections treated as one event) without sustaining mechanical damage or unacceptable electrical degradation. Although the IEC 60099-4 test procedure involves 20 impulses to verify statistical reliability, the rating itself is the limit for a single operational event.
Common Dimensioning Error: A frequent mistake observed in utility specifications is comparing the single flash charge to the Thermal Charge Rating (Qth). As shown in Table 1, Qth is typically 2 to 4 times higher than Qrs (e.g., for DH class, Qrs=0.4 C vs Qth=1.1 C).
Consequence: If a simulation predicts a lightning flash of 0.8 Coulombs, and the engineer compares this to the Qth of a DM arrester (0.7 C), they might mistakenly accept a DH arrester (Qth=1.1 C). However, the single-shot limit of the DH arrester is only 0.4 C. An injection of 0.8 C would likely cause immediate fragmentation of the varistor blocks, leading to a permanent fault, despite technically being below the "thermal" limit.
2.2.3 Step 3: Verification Against Qth
The Qth rating is reserved for verifying thermal stability. This rating is established by a test involving two impulses injected within a specified timeframe (typically to simulate a multi-event scenario or extreme duty), followed by a thermal recovery verification under applied voltages (Ur and Uc).
In dimensioning, Qth should be used to verify scenarios involving:
Two independent lightning flashes occurring within a short interval (e.g., < 1 minute).
A lightning flash followed immediately by a temporary overvoltage (TOV) event or a reclosing operation.
2.3 The Fallacy of Kilojoule Ratings in Distribution
Historically, distribution arresters were sometimes specified with energy ratings in kJ/kV Ur. This approach is fundamentally flawed and has been removed from modern IEC recommendations for distribution classes.
The energy dissipated (Eflash) is the integral of voltage times current.
Because the residual voltage is a property of the arrester design, a "bad" arrester with a high residual voltage will dissipate more energy for the same lightning current than a "good" arrester with a low residual voltage. If arresters were rated in Joules, manufacturers would be incentivized to produce arresters with poor protection levels to achieve higher "Energy Ratings."
By standardizing on Charge (Coulombs), the rating becomes independent of the device's voltage characteristic, focusing purely on the external stress imposed by the system (the lightning current).
3.Methodology for Station Class Arresters (IEC 60099-4)
For Station Class arresters (designated SL, SM, SH), typically applied at voltages Us ≥ 72.5 kV, the dimensioning logic shifts from lightning charge to switching surge energy. While insulation coordination (Upl/Ures) is still driven by lightning performance, the volumetric sizing of the MOV blocks is dictated by the energy associated with switching operations.
3.1 Classification and Energy Definitions
Station class arresters also utilize a dual-rating system that distinguishes between the single-event withstand capability (Qrs and its associated energy) and the thermal limit (Wth).
Class | Designation | Nominal Discharge Current (In) | Minimum Qrs Rating (Single Impulse) | Minimum Wth Rating (Thermal Stability) | Typical Applications |
Station Low | SL | 10 kA | 1.0 C | 4 kJ/kV Ur | Min. Us ≤ 300 kV System specific |
Station Medium | SM | 10 kA | 1.6 C | 7 kJ/kV Ur | Min. Us ≤ 420 kV System specific |
Station High | SH | 20 kA | 2.4 C | 10 kJ/kV Ur | Min. Us ≤ 800 kV System specific |
Table 2: Station Arrester Ratings. Note that Qrs/Wth values are minimums; specific products often exceed these.
Important note: Qrs and Wth ratings for station-class arresters are tested using standardized 2 ms to 4 ms switching impulses.
3.2 The Physics of Switching Surges
Switching surges, such as those caused by line energization or high-speed auto-reclosing, result in long-duration current impulses (typically 2 ms to 4 ms). Unlike the 8/20 µs lightning impulse, these events allow heat to diffuse from the grain boundaries into the bulk of the ZnO grain, making the energy (Joules) a more relevant metric than for lightning. However, the distinction between single-shot mechanics and thermal accumulation remains paramount.
3.3 Dimensioning Methodology: Qrs vs. Wth
The most critical nuance in station arrester dimensioning is the correct application of the Qrs and Wth ratings to simulation results.
3.3.1 Defining the Simulation Events
Line Energization: Closing a breaker onto a transmission line. This typically results in a single energy pulse through the line-end arrester.
Reclosing (O-C-O): A fault occurs, the breaker opens, and then recloses (often onto a trapped charge). This results in a sequence of energy injections.
3.3.2 The Single Event Criterion (Qrs)
When a simulation calculates the energy of a single switching event (e.g., one auto-reclose attempt), this value must be compared to the Single Impulse Energy Capability.
In IEC 60099-4, this capability is represented by Qrs or roughly (not exactly) 50% of Wth
As per IEC requirements, manufacturers shall provide a specific datasheet value (Qrs rating) for "Single Impulse Energy Rating" corresponding to the Qrs test and relevant type test reports.
The Engineering Trap: A widespread error in transient studies is comparing the energy of a single switching event directly to the Thermal Energy Rating (Wth)
Why this is wrong: The Wth rating is defined by a test procedure that involves two impulses injected roughly three minutes (max.) apart. Therefore, Wth represents the total heat capacity of the unit for a multi-shot event.
The "Factor of 2" Rule: As a heuristic, the single-shot capability is approximately half of the Wth rating (simplification, not exactly true).
Consequence: If a study indicates a single switching surge of 6.0 kJ/kV, and the engineer selects an arrester with Wth = 7.0 kJ/kV (Class SM), they have undersized the arrester. The single-shot limit of the SM arrester is likely around 3.0–3.5 kJ/kV. An injection of 6.0 kJ/kV in one shot will certainly cause mechanical failure (cracking) of the MOV blocks or a thermal runaway, even though it is below the "thermal" limit of 7.0 kJ/kV.
3.3.3 The Thermal Stability Criterion (Wth)
The Wth rating is correctly used only when evaluating the cumulative energy of a sequence of events occurring within a short timeframe (typically 3 minutes).
Scenario: Two auto-reclosing cycles (Open - 0.3s - Close - Open) following each others.
Calculation: Sum the energy of both events.
E_total = E_event1 + E_event2
Verification: Compare E_total < Wth. This ensures that the total heat generated does not drive the arrester into thermal runaway during the post-event recovery phase.
3.4 Summary of Station Class Dimensioning
The methodology can be condensed into a strict logical flow:
Run Simulation: Obtain energy for the single worst-case switching event Emax_single.
Check Single Shot Wth: Ensure Emax_single < EQrs (approx. 0.5 x Wth). Often adequate values must be found in type test reports, they can't be found in datasheets.
Check Single Shot Qrs: In any case, always verify that the total charge Qmax_single for that single worst-case switching event remains below the declared Qrs rating. Your EMT tool should give you both charge and energy values.
Consider Cumulative Events: Sum total energy for a sequence of events (e.g. two auto-reclosing cycles within 3 minutes) Emax_cumulative
Check Thermal Stability: Ensure Emax_cumulative < Wth
Iterate: If either check fails, move to the next higher class (e.g., from SM to SH) or increase the number of parallel columns. Please note that higher class or parallel columns will influence the residual voltage and therefore the simulated energy values.
Comments