IN Brief:
- Two geothermal units tripped while the Tauhara–Wairakei circuit was being returned from a planned outage.
- North Island frequency fell to 49.2Hz before automatic reserves restored normal operation within approximately one minute.
- The regulator’s draft finding attributes the event to omissions in outage assessment and consideration of switching voltage angles.
New Zealand’s Electricity Authority has opened consultation on a draft determination that identifies Transpower in its role as system operator as the causer of an October 2025 North Island under-frequency event.
The event occurred at approximately 5.42pm on 23 October 2025 while the Tauhara–Wairakei-1 circuit was being returned to service after a planned outage. Within seconds of switching, the Tauhara and Te Huka 3 geothermal generating units tripped and system frequency fell to 49.2Hz.
Automatic reserves responded and restored frequency to the normal operating band by 5.43pm. The Authority said the event was resolved without widespread impact, although Contact Energy indicated that its generating units may have been damaged; the regulator said it did not have analysis confirming whether damage occurred.
The consultation remains a draft finding rather than a final allocation of responsibility. Submissions from substantially affected participants are due by 5pm on 31 August 2026, after which the Authority will consider the evidence and issue its final determination.
Switching exposed a large voltage angle
Immediately before the event, the TAB-WRK-1 circuit was out of service. The draft paper states that impedance between Tauhara and Wairakei through the parallel Hawke’s Bay loop was approximately 78 ohms in that condition, compared with about 2 ohms when the direct circuit was in service.
Tauhara was generating approximately 171MW and Te Huka 3 about 52MW during the outage. When the line was ready to return, the breaker at the Wairakei end was closed first to energise the circuit, followed by circuit breaker 662 at the Tauhara end.
The voltage angle across the open contacts of CB662 was approximately 23 degrees immediately before closure. Both generating units tripped within seconds, removing enough power for North Island frequency to cross the 49.25Hz threshold used to define an under-frequency event.
New Zealand’s normal system-frequency band is 49.8Hz to 50.2Hz. Frequency falls when generation becomes insufficient for demand, and reserves must respond rapidly to arrest the decline before automatic load shedding or wider instability develops.
The one-minute recovery demonstrates that reserve arrangements responded after the loss. It does not settle why the generators tripped or which participant should bear responsibility under the Electricity Industry Participation Code, which is the purpose of the determination process.
The system operator’s investigation had identified Contact Energy as the causer. The Authority disagrees in its draft finding, focusing instead on the assessment of the planned outage and the voltage-angle conditions that could arise during routine switching.
Draft finding focuses on prior engineering assessment
The Authority’s case is that the system operator omitted to assess the October outage against its obligation to dispatch available assets in a manner that avoids cascading asset failures and consumer supply losses arising from frequency or voltage excursions.
It also says the system operator did not adequately consider the voltage-angle conditions that could reasonably arise when forming its opinion on the design and configuration of the grid connection to the Tauhara generating unit during commissioning.
According to the draft, a compliant assessment would have identified the significant angle across CB662 when the geothermal units were operating at high output. The system operator could then have required measures to manage the risk before the breaker was closed.
The operating changes made after the event indicate the type of intervention available. The grid owner reduced the synchronism-check setting for automatic reclosing from 20 degrees to 10 degrees and required the Tauhara unit to be ramped to zero before a manual reclose.
The regulator says those changes are materially similar to the operating policy that would have resulted from the assessments it believes should have been completed before the incident.
The case illustrates why planned-outage studies must examine the restoration state as well as the period when equipment is unavailable. A network may remain secure while a circuit is open but experience different electrical conditions as breakers close and power flows redistribute.
Commissioning assessments also have a long operating life. Assumptions made when a generator is connected can influence switching instructions, outage planning, and protection coordination years later, particularly where output patterns and surrounding network conditions change.
The regulator has invited feedback on the proposed causer and on the system operator’s calculation of the megawatts lost, which is used to determine the event charge. Affected participants may therefore challenge both the technical chain of causation and the financial consequences attached to it.
A final determination could confirm the draft, identify another participant, or revise the reasoning after submissions. Until that process is complete, it would be premature to describe Transpower as finally responsible or Contact Energy as formally cleared.
The operational sequence lasted little more than a minute, but the investigation reaches back into outage assessment, commissioning decisions, protection behaviour, and the obligations assigned under the Code. Automatic reserves corrected the frequency. The regulatory process is deciding whether the conditions that made their response necessary should have been identified before the breaker was closed.