Ready to Start Your Project?

Contact us today to discuss your renewable energy engineering needs.

Get in Touch
Fault Ride-Through Under S5.2.5.5 and S5.2.5.5A: Reactive Current Injection Explained

Fault ride-through is a generating system's ability to stay connected through a network fault and to help hold the voltage up while the fault is on. In the National Electricity Rules it now sits in two clauses. S5.2.5.5 sets which disturbances the plant must remain in continuous uninterrupted operation for, including sequences of faults. S5.2.5.5A, introduced by the AEMC's 2025 access standards rule (Package 1), sets how the plant must respond during and after them: how much reactive current it injects or absorbs, how fast, and how quickly its active power recovers.

For inverter-based plant the response is set by control software, so the clause is really a specification for firmware. That is why it is negotiated more often than almost any other.

What the Plant Must Survive (S5.2.5.5)

The automatic access standard requires continuous uninterrupted operation for:

  • any credible contingency event
  • a three-phase fault in the transmission system cleared by primary protection
  • unbalanced transmission faults cleared in the breaker fail time (or a default of 430 ms where no breaker fail protection exists, if longer than primary clearance)
  • faults in the distribution network cleared in the breaker fail time, or 430 ms, whichever applies

The clause also covers sequences of faults, which we deal with separately in multiple fault ride-through. The minimum access standard narrows the event list and allows the NSP and AEMO to exclude faults whose total system impact is small.

A disturbance is taken to end when the connection point voltage is back within 90% to 110% of nominal and stays there for at least 20 ms. That definition matters, because the recovery time requirements run from it.

How the Plant Must Respond (S5.2.5.5A)

For asynchronous plant (inverters), the automatic access standard has four parts.

RequirementAutomatic access standardMinimum access standard
Capacitive reactive current for undervoltageAt least 4% of maximum continuous current for each 1% drop in positive sequence voltage below the initiating pointGreater than 0% per 1%
Inductive reactive current for overvoltageAt least 6% of maximum continuous current for each 1% rise above the initiating pointGreater than 0% per 1%
Initiating conditionsResponse commences at or before 85% (under) and 115% (over) of nominal80% and 120%, or other agreed values
Commencement and rise timeResponse starts within 10 ms; positive sequence rise time no more than 40 msStarts within 40 ms; rise time no more than 80 ms, or longer if agreed
Active power recoveryAt least 95% of pre-disturbance level within 100 ms after the disturbance endsWithin a period agreed with the NSP and AEMO
Negative sequenceNegative sequence current, or equivalent, to oppose unbalanced voltageResponse must not contribute excessively to voltage rise on unfaulted phases

The required response levels, within those capability ranges, are agreed with the NSP and AEMO at whatever level best achieves the control objective: minimising the deviation of each phase voltage from its pre-disturbance value while maintaining stable control. The response must also be adequately controlled, which the Rules now define. In short, the plant must reach the agreed reactive current within the disturbance, transition into and out of current limits stably, and not cause voltages outside the system standards or oscillations that could trip other plant.

Two details are easy to miss:

  • Reactive current may be limited to the plant's maximum continuous current. A plant at full active output cannot simply add reactive current on top. The performance standard must record how the plant prioritises when it hits its current limit: active versus reactive, and positive versus negative sequence.
  • The measurement location can be agreed. Commencement and rise time can be measured inside the plant, for example at the inverter terminals, where the connection point is remote or the signals are cleaner. The required injection is then recalculated for that location.

Why the Numbers Are Hard for Inverters

On paper, 4% per 1% is a simple gain. In practice four things get in the way.

Current headroom. An inverter at rated output is already at, or near, its current limit. During a deep fault the reactive current demanded by a 4% gradient soon exceeds the maximum continuous current, so the plant must reduce active current to make room. How fast it does that, and how fast it gives the active current back afterwards, drives both the reactive current achieved and the 100 ms recovery time.

Weak grids. At low short-circuit ratio, a fast reactive current injection raises the voltage quickly, which reduces the demanded current, which lowers the voltage again. Aggressive gains that look excellent at a strong connection point can oscillate at a weak one. The clause requires the response to be adequately controlled across the range of system impedance the NSP nominates.

Unbalanced faults. The voltage on the faulted phase falls while the unfaulted phases can rise. Injecting positive sequence current alone can push the healthy phases higher. Negative sequence injection helps, but not every inverter supports it, and the RMS (PSS/E) model often cannot represent it, which leaves the EMT model as the only evidence.

Grid-forming plant. Grid-forming inverters usually respond to faults through a virtual impedance rather than a reactive current gain. Their natural response is fast and voltage-supporting, but it does not map neatly to a gradient, it can decay during long faults, and the percentage achieved depends on when you measure it. Projects with grid-forming inverters frequently land on a negotiated standard expressed in terms that reflect how the plant actually behaves.

Distribution networks add another twist. Some DNSPs prefer embedded inverters to hold their pre-fault current during a fault, with no reactive boost at all, to limit voltage spikes at fault inception and clearance and to avoid desensitising feeder protection. See our Chapter 5A guide.

How the Clause Is Studied

The core study is an EMT (PSCAD) simulation on a single-machine infinite-bus (SMIB) test system, repeated in PSS/E for consistency, and then checked in wide-area cases.

S5.2.5.5 and S5.2.5.5A Study Matrix

  • Operating points at maximum export and import (for a BESS), with reactive power at zero and at the capability limits
  • Grid strength at the maximum SCR and at the agreed minimum SCR, with the site X/R
  • Three-phase faults with fault impedance stepped so the retained connection point voltage covers zero to about 0.9 pu
  • Overvoltage events (for example switching a shunt capacitor) covering the 115% to 130% range
  • Unbalanced faults: phase to ground, phase to phase and two phase to ground, bolted and through impedance
  • Fault durations matching the longest relevant clearance times, including breaker fail times
  • Wide-area cases with the NSP's contingency list, using current network snapshots and committed projects

A practical way to set the retained voltage on the SMIB is to make the fault impedance a multiple of the source impedance. Stepping that multiple from zero upwards gives a family of faults from bolted to shallow.

What is measured

For each case the study extracts, at the agreed location:

  • positive sequence reactive current, on a clearly stated base (normally the maximum continuous current at the connection point)
  • the time from the initiating condition to the start of the response, and the rise time
  • the positive and negative sequence current split for unbalanced faults
  • active power recovery time to 95% after the disturbance ends
  • overvoltage on unfaulted phases
  • whether the plant's fault ride-through mode actually triggered

Then the results are plotted as achieved reactive current against voltage deviation, with the proposed standard drawn on the same chart. Reviewers will check that points from faults that did not trigger the fault ride-through mode are included too. A response defined only for faults that trip the mode leaves a gap at shallow faults.

What reviewers commonly ask

  • State the base and the sampling instant. If reactive current decays through a long fault, the percentage depends on when it is read. Agree whether it is read at typical primary clearance or at the end of the fault.
  • Explain spikes. Current or power spikes at fault inception and clearance are usually PLL re-synchronisation or control mode transitions. They need explaining, not hiding with axis scaling.
  • Explain the control sequence. Entry and exit thresholds, delays, the handshake between plant controller and inverters, and whether the plant changes mode or freezes its control loops. Functional block diagrams from the OEM save a review round.
  • Show maximum SCR as well as minimum. The weak-grid case shows stability, but the strong-grid case is closer to normal operation.
  • Keep the documents consistent. Thresholds in the GPS, the model, the user guide and the protection settings must match.

Negotiating a Lower Standard

A negotiated standard must sit as close to the automatic standard as practicable, and the NSP and AEMO must be satisfied it will not cause other plant to trip that would not otherwise have tripped. In practice a credible proposal contains:

  • the specific parameter that falls short, and in which cases, rather than a blanket reduction
  • evidence of the tuning steps taken to improve it, and why further improvement is not possible
  • evidence that the minimum standard is met and the response is adequately controlled
  • wide-area results showing no adverse effect on nearby plant or voltage recovery

Common outcomes for inverter plant are a reactive current gradient below 4% per 1%, wider initiating thresholds, and active power recovery slightly longer than 100 ms. Each needs a reason that the reviewers can follow from the plots.

Our article on automatic, minimum and negotiated access standards explains the negotiation framework, and EMT vs RMS simulation explains why PSCAD usually governs this clause.

Get Fault Ride-Through Right

Fault ride-through is where inverter settings, model quality and negotiation all meet. Settling the control philosophy with the OEM before studies start is the single biggest time saver.

ElectraGlobe runs fault ride-through studies in PSCAD and PSS/E, tunes plant and inverter settings with OEMs, and prepares the S5.2.5.5 and S5.2.5.5A sections of GPS proposals. See our steady-state and dynamic studies service, or contact us.

Frequently Asked Questions

What is the reactive current requirement for fault ride-through?

Under the automatic access standard in S5.2.5.5A, inverter-based plant must be capable of injecting capacitive reactive current of at least 4% of its maximum continuous current for each 1% drop in positive sequence voltage below the initiating point, and absorbing inductive reactive current of at least 6% for each 1% rise above it. The response must start within 10 ms and have a rise time of no more than 40 ms.

What changed with S5.2.5.5A?

The 2025 access standards rule moved the response requirements out of S5.2.5.5 into a new clause S5.2.5.5A. S5.2.5.5 now deals with which disturbances the plant must ride through, and S5.2.5.5A with how it responds, including a defined control objective and a definition of "adequately controlled".

How fast must active power recover after a fault?

Under the automatic access standard, to at least 95% of the pre-disturbance level within 100 ms after the disturbance ends. Under the minimum access standard the period is agreed with the NSP and AEMO.

Is PSCAD or PSS/E used to assess fault ride-through?

Both are submitted, but the EMT (PSCAD) model usually governs the assessment of fast quantities such as reactive current rise time and active power recovery, because it represents the inverter controls and unbalanced behaviour in more detail.

Power System Studies

Steady-state and dynamic studies in PSS/E and PSCAD, giving AEMO and your network service provider the technical evidence they need to approve connection, GPS applications and registration.

Learn more