Multiple fault ride-through (MFRT) is the requirement that a generating system stays connected through a rapid series of network faults, not just one. In the National Electricity Rules it sits in clause S5.2.5.5, and the automatic access standard is demanding: up to 15 disturbances within any five-minute period, with no minimum gap between them.
The requirement came out of real events. Storms that cause repeated faults on the same lines have tripped plants whose protection or control logic counted faults, overheated a component, or simply ran out of the ability to recover between them.
What the Clause Requires
| Parameter | Automatic access standard | Minimum access standard |
|---|---|---|
| Disturbances in five minutes | Up to 15 | Up to 6 |
| Disturbances below 50% voltage | Up to 6 | Up to 3 |
| Three-phase faults | Up to 2 where three-phase reclosing is permitted, otherwise 1 | None |
| Breaker fail cleared faults | Up to 1 | Not specified |
| Temporary overvoltage events | Up to 1 in the S5.2.5.4 low-voltage ranges | Up to 1 in the agreed ranges |
| Minimum time between disturbances | Zero | More than 200 ms, and no more than 3 in 30 seconds |
| Cumulative time below 90% voltage | Excluded if over 1,800 ms in five minutes | Excluded if over 1,000 ms |
| Integral of voltage deficit below 90% | Excluded if over 1 pu-second | Excluded if over 0.5 pu-second |
| Rest after a five-minute sequence | Not specified | At least 30 minutes with no disturbance |
Sequences that would island the plant, remove network elements that materially reduce transfer capability, or take the system impedance beyond the highest value the NSP nominates are excluded. A fault that re-strikes after an automatic reclose counts as a separate disturbance.
A negotiated standard can record specific plant limitations, for combinations of disturbances the plant cannot survive, with the required response for each. Those limitations cannot reduce the overall number of disturbances below the minimum access standard.
Why Plants Fail MFRT
A plant that rides through any single fault can still fail a sequence. The common causes are:
- Fault counters. Some inverter firmware counts ride-through events and trips or locks out after a set number in a set time. It is the first thing to ask the OEM about. Many applications include an OEM statement that no such logic is active.
- Energy and thermal limits. During a fault a solar inverter cannot export its DC power, and the energy goes into the DC link and is dissipated, typically in a chopper or braking resistor. Repeated faults can exceed its thermal rating. Battery systems handle this better, but their DC side and cooling still need checking.
- Incomplete recovery between faults. If active power ramps back slowly after each fault, a second fault arriving 200 ms later finds the plant in a partly recovered state. With a zero minimum gap in the automatic standard, the plant must handle a fault that starts the moment the previous one ends.
- Control wind-up. Integrators in the plant controller or inverter that wind up during one fault and have not unwound by the next can produce a large overshoot or an unstable response.
- Protection interaction. Voltage or frequency protection elements with accumulating timers can operate on the combined duration of several short excursions.
How MFRT Is Simulated
MFRT is normally studied in PSCAD on a single-machine infinite-bus system, because the fast inverter behaviour and the DC-side dynamics matter.
MFRT Study Set-Up
- Build sequences that meet the clause: up to 15 faults, the permitted number below 50% voltage, the permitted three-phase faults, one breaker fail duration, and one overvoltage event where required
- Mix fault types (three-phase, phase to phase, phase to ground, two phase to ground) and fault impedances so retained voltages vary
- Use realistic durations, typically drawn from primary clearance, delayed clearance and breaker fail times
- Vary the gaps between faults, including a near-zero gap case and longer gaps up to several seconds
- Run at full output and, for a BESS, at full charge, at the agreed minimum SCR
- Check each sequence against the cumulative time below 90% and the volt-second integral limits, and show the calculation
AEMO's dynamic model acceptance tests also include MFRT sequences, typically one specified sequence and several randomised ones. Those tests check the model behaves credibly. The GPS study checks the plant meets its standard, usually with sequences designed to be onerous within the clause limits. See our guide to DMAT.
What to show
For each sequence the report should include:
- connection point voltage, active power, reactive current and the plant's fault ride-through status flags across the whole sequence
- a table of each fault: type, retained voltage, duration and gap before the next
- the cumulative time below 90% and the volt-second integral, compared with the limits
- confirmation that the plant returned to its pre-disturbance operating point at the end
- any internal quantity that approached its limit, such as DC link voltage or chopper energy, where the model exposes it
Reviewers check the sequences against the clause carefully. A sequence that accidentally breaks a limit, for example by including too many deep faults, is either invalid or overly onerous, and either way it needs fixing.
Reading the Results
A pass looks like fifteen fault responses that each resemble the single-fault response, with active power recovering between faults and no drift in the operating point. Things to look for:
- active power that recovers more slowly with each successive fault, which points to a thermal or energy limit
- reactive current that falls away in later faults
- an overvoltage at clearance that grows through the sequence
- a fault ride-through mode that fails to exit, or re-triggers on recovery
If a weakness only appears for a particular combination, such as a three-phase fault immediately after a breaker fail event, that combination is a candidate for a negotiated plant limitation rather than a reason to reduce the whole standard.
Get MFRT Right Early
MFRT is cheap to check in simulation and expensive to discover at commissioning. Ask the OEM early about fault counters, chopper ratings and recovery logic, and run a representative sequence as soon as the PSCAD model arrives.
ElectraGlobe designs and runs MFRT studies as part of GPS packages for solar, wind and BESS projects. See our steady-state and dynamic studies service, and our article on fault ride-through and reactive current injection for the single-fault response.
Frequently Asked Questions
How many faults must a generator ride through under MFRT?
Under the automatic access standard in S5.2.5.5, up to 15 disturbances within any five-minute period, with up to six taking the connection point voltage below 50%. The minimum access standard is six disturbances, with more than 200 ms between them.
Why is MFRT difficult for solar inverters?
During each fault a solar inverter cannot export its DC power, so energy is absorbed in the DC link and dissipated, often in a braking chopper. Repeated faults can exceed its thermal rating. Fault-counting firmware and slow active power recovery between faults are other common causes of failure.
Is MFRT studied in PSCAD or PSS/E?
Usually PSCAD, because EMT models represent the inverter controls, DC-side behaviour and unbalanced faults more accurately. Both models are tested with MFRT sequences in AEMO's dynamic model acceptance tests.