Doc #67 — Transpower Grid Operations Under Isolation

---

RECOMMENDED ACTIONS (BY URGENCY)

First week

1. Confirm operational status of all Transpower substations and the HVDC link.
2. Classify all Transpower and lines company operational staff as essential personnel (Doc #1).
3. Secure spare parts at Transpower stores, lines company warehouses, and any electrical wholesale suppliers. Particular priorities: protection relay units, transformer bushings, SF6 gas cylinders, transformer oil, SCADA RTU spares.
4. Establish direct communication between System Operator and government emergency management.
5. Begin printing protection relay settings for all circuits.

First month

6. Complete national inventory of transmission and distribution spare parts (Transpower, all 29 EDBs, electrical wholesalers, generation companies).
7. Begin structured knowledge capture interviews with System Operator staff, HVDC specialists, and key substation maintenance engineers.
8. Conduct condition assessment (DGA, oil testing) of all major transmission transformers.
9. Review and reconfigure AUFLS scheme to reflect revised priority loads.
10. Establish the priority load framework (Section 3.4) in consultation with regional Civil Defence.
11. Make and implement the Tiwai Point smelter decision (Section 3.3).
12. Inventory national SF6 stocks and begin leak detection program.

First year

13. Complete knowledge capture for all critical operational and maintenance functions.
14. Print and distribute all critical documentation to substations and regional archives.
15. Establish transformer oil reconditioning program covering the entire transmission fleet.
16. Begin cross-training program between Transpower and lines companies.
17. Develop and document manual operating procedures for all major substations (for when SCADA is lost).
18. Develop black start procedures in print and distribute to all relevant stations.
19. Identify and source electromechanical relays from anywhere in the system (warehouses, decommissioned sites, retired equipment).
20. Begin planning for long-term transformer rewinding capability (Doc #69) and copper wire supply (Doc #70).

Years 2–5

21. Recruit and begin training the next generation of grid operators and maintenance staff.
22. Transition substations to manual operation as SCADA connectivity is lost.
23. Begin replacing failed digital relays with electromechanical units.
24. Commence transformer rewinding for units identified as highest risk through the DGA program.
25. Begin circuit breaker fleet transition planning (SF6 to oil circuit breakers).

Ongoing

26. Continuous transformer monitoring (DGA, oil quality, bushing condition).
27. Maintain training pipeline for all grid functions.
28. Update documentation as conditions and procedures change.
29. Regular black start exercises.
30. Coordination with hydro maintenance (Doc #65), geothermal maintenance (Doc #66), wind turbine maintenance (Doc #71), and distribution networks (Doc #68).

---

Economic Justification

Person-years required for grid maintenance

The national transmission grid cannot maintain itself. It requires a continuous workforce of specialist personnel whose combination of training and experience takes years to assemble and cannot be improvised after the fact. The workforce categories and approximate sizing are:

Role	Estimated headcount	Notes
System Operator — control room operators	20–30	Shift-based, 24/7 coverage of national control centre
High-voltage transmission engineers	40–60	Protection, asset management, planning functions
Substation maintenance engineers and technicians	80–120	Based across the network at regional maintenance centres
HVDC specialists	8–15	Highly concentrated — converter stations at Benmore and Haywards
Transformer specialists	10–20	Oil testing, DGA, condition assessment, overhaul
Lineworkers (transmission)	150–200	Line patrols, conductor repairs, tower maintenance
SCADA and communications technicians	20–30	Maintaining RTUs, control centre hardware, microwave and fibre links
Protection relay specialists	15–25	Testing, calibration, settings management
Total (Transpower core)	~350–500 FTE	Excludes EDB distribution workforce

These are not approximations that can be rounded down. A substation without a qualified maintenance engineer is not a slightly less well-maintained substation — it is a substation where the next equipment failure is not caught early and becomes a major outage. A System Operator team below critical staffing cannot manage frequency events safely. The HVDC specialist team is not a team that can absorb attrition without loss of critical capability — there is no redundancy at the margins.

Over a decade-long isolation horizon, the workforce investment in person-years is substantial: approximately 350–500 FTE × 10 years = 3,500–5,000 person-years of skilled labour dedicated to grid maintenance. This is a large number. It is also the correct number. There is no lower-cost alternative that maintains the grid.

Grid maintenance investment vs. grid failure consequences

The appropriate comparison is not “grid maintenance cost vs. doing nothing.” The comparison is “grid maintenance cost vs. cost of grid failure,” because grid failure is the consequence of not maintaining the grid.

Every function that uses electricity stops when the grid fails:

• Approximately 12,000 dairy farms lose milking capability immediately;¹ milk cannot be collected; within days, herd welfare forces culling decisions (Doc #76)
• Water treatment stops — electric pumps drive both drinking water supply and sewage treatment; public health consequences begin within 12–48 hours in major centres depending on reservoir storage capacity
• Hospitals exhaust backup generator fuel within 1–5 days (depending on generator capacity, on-site fuel storage, and whether full or partial hospital load is supported); intensive care and surgical capability is then lost
• Telecommunications infrastructure — cell towers, exchanges, repeaters — loses power; SCADA, financial systems, and emergency communications all depend on continuous power
• Cold storage fails; refrigerated food stocks (meat, dairy, pharmaceuticals) begin spoiling within 48–72 hours
• Every manufacturing workshop, repair facility, and processing plant ceases operation
• Electric rail (Wellington commuter rail, NIMT electrified sections) stops

The grid is not one infrastructure element competing for maintenance resources against other infrastructure elements. It is the infrastructure that makes most other infrastructure functional. Every hour of continued grid operation has a multiplier effect across the entire recovery. Every hour of grid failure has a multiplier effect in the opposite direction.

There is no reasonable scenario in which redirecting skilled electricians, protection engineers, or lineworkers away from grid maintenance to other recovery tasks produces a net benefit. The grid workforce is not available for reallocation. It is the load-bearing structure on which the rest of the recovery stands.

Breakeven analysis

The question of whether grid maintenance investment “breaks even” is somewhat misconceived — the grid will be maintained because the alternative is loss of most modern economic and social functions — but the arithmetic nonetheless illustrates the scale of value at stake.

NZ’s pre-event electricity consumption is approximately 40,000 GWh/year, at a retail value of roughly $8–10 billion NZD per year.²³ Under isolation, market pricing loses its meaning, but the physical productivity enabled by that electricity — dairy production (approximately $9 billion export value pre-event),⁴ food processing, water supply, manufacturing — remains real. A conservative estimate of the economic productivity dependent on grid electricity is $20–30 billion NZD equivalent per year, much of it irreplaceable with any non-electric alternative on any reasonable timeframe.

The annual cost of the grid maintenance workforce — at an imputed value of $80,000–$120,000 per skilled FTE (reflecting the genuine scarcity value of these skills under isolation; pre-event median earnings for electrical engineers and line mechanics in NZ are in the $70,000–$100,000 range, and isolation premiums would increase this) — is approximately:⁵

• 350–500 FTE × $80k–$120k = $28–60 million per year

Against $20–30 billion per year in enabled productivity, the workforce cost is approximately 0.1–0.3% of the value it protects. The breakeven threshold is crossed in the first week of continued operation.

No other single infrastructure maintenance investment has a comparable return ratio. This is a consequence of the grid’s role as enabling infrastructure rather than productive infrastructure — it does not produce value directly, but it enables nearly all other value production.

Opportunity cost

The opportunity cost of deploying 350–500 skilled specialists on grid maintenance rather than other tasks is real but limited in practice:

The relevant alternative deployment for most of these roles is not available. System Operators cannot be usefully redeployed to, say, food production — their skills are highly specific. High-voltage engineers cannot perform the work of a mechanical engineer or a farmer. The opportunity cost of grid maintenance is largely the opportunity cost of activities these individuals could not competently perform in any case.

The genuine opportunity cost arises in two areas. First, lineworkers have some transferable skills (rigging, mechanical work, working at height) that could be applied elsewhere. Second, electrical technicians and engineers could potentially contribute to manufacturing or workshop capacity. The honest assessment is that this transferable capacity is small relative to the value of its primary deployment, and that the grid cannot tolerate reduction below current staffing levels without increasing the risk of failures that would erase any gains from reallocation many times over.

The grid maintenance workforce should be classified as non-reallocatable. Its members are in the highest-priority essential personnel category (Doc #1), and their continued deployment in grid maintenance functions should be treated as a fixed constraint in workforce planning, not as a variable to be optimised.

---

1. GRID ARCHITECTURE

1.1 The transmission network

Transpower operates approximately 12,000 km of high-voltage transmission lines and over 170 substations linking generation to the distribution networks that serve end users.⁶ The network is structured in voltage tiers:

HVDC inter-island link: A high-voltage direct current link across Cook Strait, connecting the South Island (predominantly hydro generation, surplus capacity) with the North Island (majority of population and demand). The link consists of submarine cables across Cook Strait and converter stations at Benmore (South Island) and Haywards (North Island). Rated capacity is approximately 1,200 MW.⁷ This is the single most critical link in the national grid — without it, the two islands operate as separate electrical systems, and the North Island loses access to South Island hydro surplus.

220 kV backbone: The main transmission backbone, running the length of both islands. Carries bulk power over long distances between major substations. In the South Island, this connects the Waitaki hydro scheme and other major generation to the HVDC link and to Christchurch/Canterbury load centres. In the North Island, it runs from Haywards (Wellington) northward through the Waikato to Auckland and Northland, connecting the Waikato hydro chain, Huntly thermal plant, and geothermal generation in the Taupo Volcanic Zone.

110 kV regional networks: Regional transmission connecting the 220 kV backbone to major load centres and smaller generation sources. Serves as the interface between bulk transmission and local distribution.

66 kV and lower sub-transmission: Connects transmission substations to distribution substations, particularly in smaller centres and rural areas. Operated partly by Transpower and partly by local lines companies (Electricity Distribution Businesses, or EDBs).

1.2 Key substations

Several substations serve as critical nodes. Loss of any one would significantly disrupt power delivery to a region:

Substation	Voltage	Function
Benmore	220 kV / HVDC	South Island HVDC terminal. Loss isolates South from North.
Haywards (Upper Hutt)	220 kV / HVDC	North Island HVDC terminal. Loss isolates Wellington from HVDC supply.
Islington (Christchurch)	220 kV	Major South Island load centre. Serves Canterbury.
Otahuhu (Auckland)	220 kV	Major North Island load centre. Serves Auckland.
Hamilton	220 kV	Waikato hub. Connects Waikato hydro to backbone.
Whakamaru	220 kV	Waikato generation hub. Multiple hydro stations connect here.
Bunnythorpe (Manawatu)	220 kV	Central North Island junction.
Twizel	220 kV	Waitaki hydro scheme hub.
Manapouri	220 kV	Connects Manapouri station. Currently serves Tiwai Point smelter.

1.3 The HVDC link in detail

The Cook Strait HVDC link deserves special attention because its loss would be the single most consequential grid failure.

Physical components:⁸

• Converter stations at Benmore and Haywards, containing thyristor valve halls, converter transformers, harmonic filters, and control systems
• Approximately 40 km of submarine cable across Cook Strait (three cables, with some redundancy)
• Overhead DC lines connecting the converter stations to the cable terminal stations

Vulnerability assessment:

• The submarine cables are physically robust and protected by burial or armouring. Cable failure is possible (anchor strike, seismic movement) but historically rare. Repair requires a specialist cable-laying vessel — NZ does not have one, and under isolation, obtaining one would be extremely difficult. Cable failure is therefore a low-probability but very high-consequence risk.
• The converter stations contain the most complex and least replaceable equipment: thyristor valves, converter transformers (different from ordinary AC transformers — they handle combined AC/DC stresses), and sophisticated control systems. Thyristor valves have finite life (typically 20–30 years for the thyristor elements themselves, though supporting components may fail earlier) and NZ cannot manufacture replacements — thyristor fabrication requires semiconductor-grade silicon wafer processing, diffusion furnaces, and clean-room facilities that NZ does not possess and could not develop under isolation on any reasonable timeline. The control systems are highly specialised digital equipment dependent on imported microprocessors and custom firmware.
• Cooling systems for the converter stations require maintenance — pumps, heat exchangers, coolant.

If the HVDC link fails:

• The South Island has surplus generation capacity (particularly hydro) that it cannot export north
• The North Island loses up to 1,200 MW of transfer capacity and must rely entirely on North Island generation (Waikato hydro, geothermal, wind, and Huntly thermal — though Huntly depends on imported coal or gas)
• Auckland’s power supply becomes significantly more constrained
• Load shedding in the North Island becomes more likely
• The two islands must manage frequency independently

The HVDC link should be treated as a national strategic asset. Maintenance of the converter stations — particularly the thyristor valves and control systems — should receive priority allocation of whatever electronic spare parts are available. Transpower’s HVDC maintenance team represents a concentration of specialized knowledge that should be protected and its expertise documented (see Section 6).

---

2. CRITICAL EQUIPMENT AND FAILURE MODES

2.1 Power transformers

Power transformers are the grid’s most important and least replaceable active components. Transpower operates hundreds of transformers across the network, ranging from small distribution-level units to large 220 kV units weighing over 100 tonnes.⁹

How transformers fail:

• Insulation degradation: Transformer insulation consists of cellulose paper and pressboard immersed in mineral oil. Over time, heat and moisture cause the cellulose to degrade (depolymerisation), eventually losing mechanical strength. Once the paper is sufficiently degraded, the transformer is at risk of winding failure from mechanical forces during through-faults (short circuits on the network that cause surge currents through the transformer). This process is slow — decades under good operating conditions — but irreversible. The rate depends heavily on operating temperature and moisture content.
• Oil contamination: Transformer oil serves as both insulation and coolant. It degrades through oxidation, accumulates moisture, and develops dissolved gases as internal components age. Dissolved gas analysis (DGA) is the primary diagnostic tool — specific gas ratios indicate specific internal problems (overheating, arcing, partial discharge). Oil can be reconditioned (degassed, dried, filtered) to extend useful life, but the paper insulation cannot be restored once degraded.
• Bushing failure: Bushings — the insulated conductors that bring power leads through the transformer tank — can fail from moisture ingress, oil leakage, or internal electrical stress. Bushing failure can be catastrophic (explosive failure, tank rupture, fire).
• Tap changer failure: On-load tap changers (OLTCs) adjust transformer voltage ratio under load. They contain moving contacts in oil and are the most maintenance-intensive part of the transformer. Contact wear, oil contamination, and mechanism failure are common failure modes.

What NZ can and cannot do:

• Oil reconditioning: Feasible with existing NZ capability. Requires oil processing equipment (vacuum degassing, filtration, drying). This capability exists within Transpower and some lines companies, and should be maintained and expanded. Oil reconditioning is the single most effective action for extending transformer life.
• DGA testing: Requires laboratory capability — gas chromatography or equivalent. Transpower has this capability; it should be maintained as a national priority. DGA enables condition-based management rather than operating blind.
• Bushing replacement: Some standard bushings may be interchangeable between transformers. Spare bushings should be inventoried nationally. Bushing repair is possible in some cases (re-oiling, seal replacement) but depends on the failure mode.
• Tap changer overhaul: Mechanical overhaul of tap changers is within NZ workshop capability. Contact replacement requires suitable contact materials (copper alloys) — NZ has limited stock.
• Transformer rewinding: Feasible but a major undertaking requiring copper wire, insulation materials, and skilled labour. See Doc #69 (Transformer Rewinding and Fabrication) for detail. Rewinding restores a transformer to near-new condition and is how NZ extends the fleet life beyond the original insulation life.
• New transformer construction: Significantly harder than rewinding. Requires precision winding, laminated silicon steel cores (NZ does not produce transformer-grade electrical steel), and testing capability. This is a Phase 3–4 capability at earliest — see Doc #69.

Strategic implication: Every transformer in the fleet must be treated as a depreciating national asset. Oil reconditioning and DGA monitoring programs are the highest-return maintenance investments Transpower can make under isolation.

2.2 Circuit breakers

Circuit breakers interrupt fault currents — thousands of amperes in milliseconds. At transmission voltages, this is an extreme engineering task. NZ’s grid uses several types:¹⁰

• SF6 (sulphur hexafluoride) breakers: Modern, compact, reliable. The insulating and arc-quenching medium is SF6 gas, manufactured by reacting elemental fluorine with sulphur under controlled conditions. Fluorine production requires electrolysis of anhydrous hydrogen fluoride (itself produced from fluorspar — calcium fluoride ore — and sulphuric acid), using specialised nickel or Monel alloy electrolysis cells resistant to fluorine corrosion. NZ has no fluorspar mining, no fluorine production capability, and no facility for handling elemental fluorine (which is extremely reactive and hazardous). SF6 synthesis is therefore not feasible under isolation. Existing SF6 stock is finite. SF6 leaks slowly from equipment through seals (typical leak rates of 0.5–1% per year per unit). When the SF6 is exhausted, these breakers cannot operate. Leak detection and repair extends the useful life.
• Oil circuit breakers (OCBs): Older technology, still present in the fleet. Use mineral oil as insulation and arc-quenching medium. More maintainable than SF6 breakers because the oil is reclaimable and the mechanisms are entirely mechanical/hydraulic. Contacts wear from arcing and require periodic replacement or resurfacing.
• Vacuum circuit breakers: Used at lower voltages. The vacuum interrupter is a sealed unit with limited life (typically rated for 10,000–30,000 switching operations, depending on the fault current level).¹¹ Not repairable or refillable — when the vacuum is lost, the unit must be replaced.
• Air-blast breakers: Older technology using compressed air. Mechanically complex but all components are maintainable with NZ workshop capability. Compressor maintenance is required.

Management strategy:

• Maintain SF6 breakers for as long as SF6 stocks last. Detect and repair leaks aggressively. A national SF6 inventory — combining Transpower stock, lines company stock, and any industrial supply — should be established.
• Retain and maintain oil circuit breakers. As SF6 units fail, replacement with reconditioned OCBs may be necessary. OCBs are significantly larger (a 220 kV OCB may weigh 2–4 times as much as an equivalent SF6 unit), require more frequent maintenance (oil testing, contact inspection, and oil replacement after fault interruptions), and have slower fault-clearing times. However, they are indefinitely maintainable with NZ workshop capability, which SF6 units are not.
• For vacuum breakers at distribution level, stockpile replacement interrupters.
• Air-blast breakers, where they still exist, should be retained rather than scrapped.

2.3 Protection relays

Protection relays detect faults (short circuits, earth faults, over-currents, under-frequency) and trip circuit breakers to isolate the faulted section. Without functioning protection, a single fault can cascade through the network and cause widespread damage.

Modern digital relays: Most of Transpower’s protection system has been progressively upgraded to digital (microprocessor-based) relays. These are more capable than their predecessors — programmable, self-monitoring, capable of multiple protection functions in a single unit. But they are electronic devices with finite life, dependent on components NZ cannot manufacture (microprocessors, power supply capacitors, optical interfaces). They fail from capacitor aging, component drift, power supply failure, and firmware faults.

Electromechanical relays: The previous generation of protection relays — mechanical moving-coil instruments that detect electrical quantities and trip contacts. Still installed in some older parts of the network. Less capable than digital relays (single function per relay, no self-monitoring) but far more maintainable. They use copper coils, spring mechanisms, and contact materials — all within NZ’s repair capability indefinitely. Their calibration drifts over time and requires periodic testing and adjustment, but the failure mode is gradual rather than sudden.

Management strategy:

• Maintain digital relays as long as possible. They provide better protection and more flexible operation. Stockpile spare relay units and circuit boards.
• Do not discard electromechanical relays from anywhere in the system. As digital relays fail, the fleet will need to transition back to electromechanical protection. The performance gap is substantial: a single digital relay can perform 5–8 protection functions simultaneously (overcurrent, earth fault, distance, directional, breaker failure), whereas each electromechanical relay performs one function — meaning a circuit that required one digital relay may need 3–6 electromechanical units. Electromechanical relays also lack self-monitoring (faults are only detected during periodic testing), have wider measurement tolerances (typically 5–10% vs. 1–2% for digital), and cannot implement adaptive protection algorithms. However, the network operated safely with electromechanical relays for decades before digitalisation, accepting reduced discrimination and slower fault clearance as a trade-off for maintainability.
• Document all protection settings in print. Every protection relay in the network has specific settings (current pickup levels, time delays, directional elements, zone reaches for distance relays) that are tailored to the circuit it protects. These settings currently exist in Transpower’s electronic databases. They must be printed and stored at each substation and in a national archive. Without correct settings, reinstalling protection after a relay replacement is guesswork — dangerous guesswork.
• Maintain relay testing capability. Secondary injection test sets (which simulate fault currents to verify relay operation) are essential. These are robust analogue instruments that NZ can maintain.

2.4 SCADA and communications

Transpower’s System Operator manages the grid from a national control centre in Wellington, using SCADA systems that monitor and control equipment at substations across the country via a telecommunications network.¹²

What SCADA provides:

• Real-time monitoring of voltages, currents, power flows, frequency, temperatures, and equipment status across the entire grid
• Remote control of circuit breakers, tap changers, and other switchgear
• Automatic load shedding schemes (under-frequency load shedding)
• Alarms and event recording
• Generation dispatch coordination with generators

What happens without SCADA:

• The System Operator loses visibility of the grid beyond what can be reported by telephone
• Remote control of equipment is lost — all switching requires personnel at each substation
• Automatic protection still operates (protection relays are independent of SCADA) but the operator cannot see what has happened until someone reports from site
• Coordinated generation dispatch becomes much slower and less precise
• Frequency management becomes harder because the operator cannot see real-time system frequency at multiple points

SCADA degradation path: The SCADA system depends on: (a) computers and software at the control centre, (b) Remote Terminal Units (RTUs) at each substation, (c) a telecommunications network connecting them (microwave radio, fibre optic, or a combination). Each of these has different failure timelines:

• Control centre computers: standard server hardware with finite life. Can be replaced from available computer stocks for some years. Software may require specialized knowledge to reinstall or migrate.
• RTUs: industrial computers installed in substations. More robust than office computers but still electronic with finite life. Spare units may be available but are specific to the SCADA system installed.
• Communications: Transpower operates its own telecommunications network. Microwave links require electronic equipment at each end. Fibre optic links require terminal equipment. Both degrade as electronics fail.

Management strategy:

• Maintain the full SCADA system as long as possible — it provides enormous operational efficiency.
• As components fail, accept graceful degradation: substations that lose SCADA connectivity revert to local manual operation with telephone reporting.
• Develop procedures for “semi-blind” operation — managing the grid with partial visibility.
• Ultimately, if SCADA is entirely lost, the grid reverts to a mode similar to grid operation in the 1950s–1960s: staffed substations, telephone coordination, manual switching, local frequency meters. This works — NZ operated this way for decades — but the performance gap is significant: switching operations that take seconds via SCADA take 20–60 minutes when a phone call must be made and an operator dispatched to a substation; fault identification that is instantaneous with SCADA telemetry may take hours with telephone-based status reporting; and staffing requirements roughly double or triple because every major substation requires on-site operators around the clock.
• The national control centre function itself can revert to a manual dispatch board if necessary, provided the telephone network (landline or HF radio) functions.

2.5 Conductors, towers, and insulators

The passive components of the transmission network are its most robust:

Conductors: Mostly aluminium conductor steel reinforced (ACSR) — aluminium strands wrapped around a steel core. These last 50–80+ years under normal conditions.¹³ Failure from corrosion, fatigue, or mechanical damage is possible but slow and localised. Damaged spans can be replaced from NZ stock or by recycling conductor from decommissioned or lower-priority lines. NZ’s Tiwai Point aluminium smelter (if it continues operating) produces primary aluminium; otherwise, recycled aluminium is available.

Towers and poles: Steel lattice towers (galvanised steel) last 50–100+ years with corrosion management. Wooden poles have shorter life (30–50 years for treated timber) and NZ produces treated timber. Concrete poles are very long-lived.¹⁴

Insulators: Porcelain and toughened glass insulators are extremely robust — many in service for 50+ years.¹⁵ Polymer (composite) insulators are newer, lighter, and have shorter expected life (20–40 years). Porcelain and glass are the preferred type for long-term resilience. Local production of porcelain insulators would require: high-purity ball clay or kaolin (available from NZ deposits in Northland and Southland), feldspar, silica sand, kiln capability reaching 1,200–1,300°C, glaze formulation to achieve the required surface resistance and hydrophobicity, and quality testing for dielectric strength (requiring a high-voltage test set). NZ has some ceramic production capability and the raw materials are domestic, but achieving the electrical-grade consistency required for transmission insulators would require development — this is a Phase 3–4 capability. Volumes are low (hundreds per year rather than thousands), which makes small-batch production more feasible than many industrial processes.

These components require periodic inspection (corrosion, cracking, vegetation encroachment, foundation condition) but not replacement on any urgent timeline. Transmission corridor vegetation management — approximately 12,000 km — will become more labour-intensive as herbicide stocks deplete and aerial inspection capacity degrades; supplementing engineering inspections with structured ground-level monitoring by local community observers (including kaitiaki with environmental knowledge of specific areas) can extend effective surveillance coverage at low cost.

---

3. DEMAND MANAGEMENT UNDER CHANGED CONDITIONS

3.1 How demand changes

Pre-event NZ electricity demand is approximately 40,000 GWh per year, with peak demand around 6,500–7,000 MW.¹⁶ Under isolation, demand patterns change significantly:

Demand reductions:

• Heavy industry that depended on exports ceases or contracts (dairy processing for export, some manufacturing)
• Commercial activity contracts (offices, retail)
• Electric vehicle charging may decrease as vehicles are reserved for essential use
• Air conditioning (already small in NZ) becomes negligible

Demand increases or shifts:

• Electric heating demand may increase if other heating fuels (gas, wood pellets from imported machinery) become scarce and if nuclear winter increases heating requirements
• Water pumping for adapted agricultural practices
• Workshops, foundries, and manufacturing operations that replace imported goods require power
• Aluminium smelting at Tiwai Point: consumes approximately 570 MW (roughly 13% of NZ’s generation capacity).¹⁷ Whether to continue smelting is one of the most consequential single decisions for grid management (see Section 3.3)

Net effect: Total demand likely decreases in the short term as the economy contracts. In the medium term, demand may increase as local manufacturing replaces imports. The demand profile changes — less commercial daytime peak, more industrial base-load, potentially more residential heating.

3.2 Demand forecasting without data

Under normal conditions, Transpower uses sophisticated forecasting models fed by weather data, historical demand patterns, and economic indicators. Under isolation, much of this data pipeline breaks down.

Practical approach:

• Direct communication with major industrial loads (smelter, dairy processing, water utilities, hospitals) to understand their expected demand
• Regional demand estimates based on population and known industrial activity
• Frequency as the real-time indicator of supply-demand balance (see Section 5)
• Accept that forecasting accuracy will be lower and maintain larger operational margins

3.3 The Tiwai Point decision

The New Zealand Aluminium Smelter at Tiwai Point (Bluff) consumes approximately 570 MW — roughly 13% of NZ’s total generation and approximately 15% of South Island hydro output.¹⁸ Under normal conditions, this electricity produces aluminium for export. Under isolation, there is no export market.

Arguments for continued smelting:

• Aluminium is a valuable industrial material for the recovery (conductor wire, structural material, castings, cookware, roofing)
• NZ has no other primary aluminium source — once smelting stops, NZ depends entirely on recycled aluminium
• Restarting an aluminium smelter after shutdown is extremely difficult and expensive — the smelting cells (pots) freeze and must be rebuilt, a process that takes many months and significant resources¹⁹
• The smelter workforce represents skilled industrial labour

Arguments for diversion of power:

• 570 MW freed up provides enormous headroom for the rest of the grid
• Under reduced demand conditions, diverting smelter power may allow Transpower to take other generation or transmission equipment offline for maintenance while maintaining supply
• NZ’s near-term aluminium needs can be met from recycled stock for many years
• The smelter depends on imported alumina (refined from Australian bauxite) — if alumina stocks run out, the smelter stops regardless

The alumina constraint is probably decisive. The smelter imports approximately 1 million tonnes of alumina per year from Queensland.²⁰ On-site alumina stocks are typically measured in weeks of production, not years. Unless Tasman sail trade can deliver bulk alumina from Australia (plausible but not immediate — see Doc #106), the smelter will exhaust its alumina supply within months regardless of the power decision. The practical question may be whether to smelt remaining alumina stocks into aluminium ingot for national stockpiling before the smelter is wound down.

This decision should be made by government in consultation with Transpower and the smelter operator, informed by: (a) current alumina stocks, (b) prospect of Australian alumina trade, (c) national aluminium demand estimates, and (d) grid maintenance requirements.

3.4 Priority load categories

Under constrained conditions, not all loads can be served simultaneously. A priority framework is needed:

Priority 1 — Non-deferrable critical loads:

• Hospitals and medical facilities
• Water treatment and supply pumping
• Sewage treatment (public health consequence of failure)
• Transpower and lines company operational facilities
• Telecommunications infrastructure (exchanges, cell towers, repeaters)
• Emergency services
• Government operations (national and regional)

Priority 2 — Essential economic loads:

• Dairy processing (food production — but note seasonal variation)
• Meat processing and freezing works
• Cold storage facilities
• Flour mills and bakeries
• Fertiliser production (when established)
• Essential manufacturing workshops
• Rail traction (electrified sections — primarily Wellington commuter rail and potentially freight corridors)

Priority 3 — Important but deferrable loads:

• Residential supply (can be time-managed through scheduled supply windows)
• Schools and educational institutions
• Commercial premises
• Street lighting (reduced schedules acceptable)
• Non-essential manufacturing

Priority 4 — Discretionary loads:

• Sports facilities, entertainment venues
• Non-essential commercial (retail that can operate during daylight)
• Any load that can be shifted to off-peak periods

This framework should be developed in consultation with regional Civil Defence and community leaders. The ratings above are illustrative — the actual priority of specific loads depends on local circumstances and seasonal conditions.

---

4. LOAD SHEDDING

4.1 What load shedding is

Load shedding is the deliberate disconnection of electricity customers to maintain overall system stability. It is a last resort — it means some customers lose power so that the rest do not. Without load shedding, a generation shortfall causes frequency to collapse (see Section 5) and the entire grid may black out, which is far worse than controlled disconnection of some customers.

4.2 Automatic Under-Frequency Load Shedding (AUFLS)

NZ’s grid has an existing automatic under-frequency load shedding scheme.²¹ When system frequency drops below defined thresholds (indicating generation is insufficient to meet demand), relays at distribution substations automatically disconnect blocks of load in stages:

Stage	Frequency threshold	Load disconnected
1	~47.5 Hz	~16% of load
2	~47.0 Hz	Additional ~16%
3	~46.5 Hz	Additional ~16%

(Approximate figures — exact settings are specified by Transpower’s System Operator and are periodically revised.)²²

The AUFLS scheme is implemented through relays at distribution substations, operated by local lines companies under Transpower’s coordination. These relays are themselves electronic devices subject to the same degradation concerns as other protection relays, and their continued functioning is essential.

Under isolation, the AUFLS scheme must be reviewed and potentially reconfigured:

• The load blocks assigned to each stage may need to change to reflect priority categories (Section 3.4)
• Some loads currently in low-priority shedding blocks may become critical (e.g., a workshop now producing essential manufactured goods)
• The frequency thresholds themselves may need adjustment if generation diversity changes (see Section 5)

4.3 Manual (planned) load shedding

If generation capacity is insufficient to meet all demand simultaneously — a likely condition during peak demand periods under isolation, or when major generation or transmission equipment is out of service for maintenance — planned rotational load shedding may be necessary.

Rotational supply schedules:

• Divide each distribution network area into blocks
• Each block receives power for defined hours (e.g., 6 hours on, 6 hours off, or 8 on / 4 off, depending on the supply-demand gap)
• Schedules should be published and predictable so that households and businesses can plan
• Priority loads (hospitals, water supply, etc.) should be on dedicated feeders that are not shed, or on feeders that are shed last

Implementation requirements:

• Lines companies (EDBs) execute load shedding at the distribution level. Transpower coordinates but does not directly control distribution switching.
• Communication between Transpower and EDBs must be reliable — if SCADA is degraded, this requires telephone or radio communication.
• Public communication of schedules is essential for compliance and social stability. Unpredictable or unexplained blackouts erode public trust far faster than scheduled reductions.

4.4 Demand response as an alternative to shedding

Before resorting to shedding, demand response — voluntary or directed reduction by large consumers — can close smaller supply-demand gaps:

• Large industrial consumers reduce load on request (pre-arranged agreements)
• Ripple control of hot water heating (NZ’s existing ripple control infrastructure, used to manage hot water cylinder heating, is one of the world’s most extensive demand-response systems and can defer significant load)²³
• Public appeals for voluntary conservation during specific periods
• Time-of-use signalling (even without sophisticated pricing, simple schedules — “please avoid high-demand activities between 5pm and 8pm” — can shift load)

---

5. FREQUENCY MANAGEMENT

5.1 Why frequency matters

NZ’s grid operates at a nominal frequency of 50 Hz. System frequency is determined by the balance between generation and demand — when they are exactly balanced, frequency is stable. When demand exceeds generation, frequency drops. When generation exceeds demand, frequency rises.

Frequency must be maintained within tight limits (normally 49.8–50.2 Hz).²⁴ Outside these limits:

• Generators can be damaged by sustained off-frequency operation
• Some industrial processes are frequency-sensitive
• Clocks and timing devices that rely on mains frequency lose accuracy
• Below approximately 47–47.5 Hz, generators begin tripping offline on their own under-frequency protection (exact settings vary by unit) — causing cascading loss of generation and potential system blackout

5.2 How frequency is managed

Primary frequency control (governor response): Each generator has a governor that automatically adjusts output in response to frequency deviations. When frequency drops, the governor opens the turbine gate to admit more water (in hydro) or steam, increasing power output. This is fast (seconds) and automatic. Hydro generators are particularly good at frequency response because water turbine governors can respond rapidly.

Secondary frequency control (generation dispatch): The System Operator adjusts generation levels to restore frequency to 50 Hz after a disturbance. Under normal conditions, this is coordinated through SCADA and the electricity market. Under isolation, it requires communication with each generator (telephone, radio) and takes longer.

Tertiary response (reserve management): Maintaining spinning reserve — generation capacity that is synchronised to the grid but not fully loaded, ready to increase output quickly. The amount of reserve required depends on the size of the largest single contingency (the largest single generator or transmission element that could be lost instantaneously).

5.3 Frequency management challenges under isolation

Reduced generation diversity: If some generation sources are lost (thermal plant without fuel, wind turbines failing from maintenance issues), the remaining generation fleet has less diversity and potentially less reserve margin. Any individual generator’s trip has a larger proportional impact.

Largest single contingency: Under normal conditions, the largest single contingency may be the loss of the HVDC link (up to 1,200 MW) or a large generating unit. The system must carry enough reserve to ride through this loss. If reserve margins narrow, the risk of cascading failure from a single contingency increases.

Reduced system inertia: System inertia — the kinetic energy stored in the rotating masses of synchronised generators — determines how quickly frequency changes after a disturbance. More generators online means more inertia and slower frequency changes (giving governors more time to respond). If generators are taken offline (reduced demand, maintenance), inertia decreases and frequency becomes more volatile. This is not a new problem — NZ already manages this issue as wind and solar (which provide no synchronous inertia) have increased — but it may become more acute under isolation if fewer generators are running.²⁵

Governor maintenance: Governor systems (hydraulic actuators, speed-sensing mechanisms, control logic) require maintenance. Hydraulic governors use oil that must be maintained. Electronic governor controls face the same degradation as other electronics. Mechanical/hydraulic governors are more maintainable long-term than digital ones.

5.4 Operating with wider frequency bands

Under isolation, it may become necessary to accept wider normal frequency variations — for example, 49.5–50.5 Hz rather than 49.8–50.2 Hz. This reflects:

• Less precise generation-demand matching (slower dispatch, less visibility)
• Lower reserve margins
• Fewer generators available for frequency response

Most loads are tolerant of modest frequency variations. Motors run slightly faster or slower. Resistive loads (heating, lighting) are essentially unaffected. The main concerns with wider frequency bands are:

• Increased mechanical stress on generators operating at off-nominal frequency
• Timing applications that depend on mains frequency (relatively few in modern equipment)
• Potential issues with older industrial equipment designed for tight frequency tolerance

This is a pragmatic operational adjustment, not a sign of system failure. It extends the operating envelope to match reduced control capability.

---

6. KNOWLEDGE PRESERVATION AND WORKFORCE

6.1 Transpower’s knowledge base

Transpower employs approximately 700–800 staff,²⁶ including system operators, transmission engineers, maintenance engineers, project engineers, protection engineers, and specialist roles (HVDC, SCADA, asset management). The System Operator function — the people who actually manage the grid minute-to-minute from the control centre — is a relatively small team (perhaps 20–30 operators on rotating shifts, plus support staff). This represents a critical knowledge concentration.

Knowledge capture priorities (immediate — first weeks):

• System Operator procedures and practices, including abnormal situation management
• HVDC converter station operation and maintenance — extremely specialized knowledge
• Protection relay settings for all circuits in the network (print from electronic databases)
• Transformer condition data and maintenance histories (print from asset management systems)
• Substation-specific operating procedures and switching schedules
• Emergency procedures (black start, islanding, AUFLS coordination)

Institutional risk: Transpower is a state-owned enterprise. Under isolation, government direction of Transpower’s operations is straightforward legally (the Commerce Act and Electricity Industry Act provide mechanisms, and emergency powers under the CDEM Act are available). The risk is not legal authority but operational continuity — ensuring that the people who know how to run the grid continue doing so, and that their knowledge is captured and transmitted to successors.

6.2 Black start capability

Black start is the ability to restart the grid from a complete shutdown without external power. If the entire grid collapses (a very unlikely but possible event following a cascading failure), generation must restart in a coordinated sequence.

Hydro stations with black start capability — those that can start their own auxiliaries from stored energy (batteries, small diesel generators) and then energize transmission lines — are the foundation of NZ’s black start plan. Key black start stations and the energization sequence should be documented in print and kept at each relevant station.²⁷

Black start is a practiced procedure under normal conditions, typically tested through simulation. Under isolation, the consequences of needing black start and failing are severe. The procedure should be documented, tested (as far as practicable), and known to multiple people.

6.3 Workforce development

As the existing Transpower and lines company workforce ages, replacement staff must be trained. Grid operation and maintenance is not something that can be learned quickly — it involves:

• High-voltage safety (lethal voltages, arc flash risk, working procedures)
• Protection system theory and practice
• Power system analysis (fault levels, load flows, stability)
• Transformer maintenance and testing
• Switchgear operation and maintenance
• SCADA/communications systems
• Regulatory and procedural compliance

Training pathway:

• Apprenticeship under experienced staff (the single most important training mode)
• Polytechnic and university programs reoriented to emphasise power systems (see Doc #162)
• Printed training materials (textbooks on power systems engineering — NZ universities have these)
• Cross-training between Transpower and lines companies
• Regular exercises and drills for abnormal situations

---

7. PHASED OPERATIONAL TRANSITION

7.1 Phase 1 (Months 0–12): Full modern capability

Status: All modern systems functional. SCADA, digital protection, SF6 breakers, full communications. Grid operates essentially as before, with modified demand profile.

Key actions:

• Secure all spare parts (electronic boards, SF6 gas, transformer oil, bushing spares, relay units) — conduct national inventory
• Begin knowledge capture program (Section 6.1)
• Print all critical documentation (protection settings, operating procedures, transformer data, switching schedules, as-built drawings)
• Make the Tiwai Point decision (Section 3.3) in consultation with government
• Establish priority load framework (Section 3.4)
• Review and if necessary reconfigure AUFLS scheme for new demand patterns
• Begin condition assessment of transformers network-wide (DGA, oil quality, bushing condition)

7.2 Phase 2–3 (Years 1–7): Managed degradation

Status: SCADA begins losing substations as RTUs fail. Some digital relays fail and are replaced from stockpile or substituted with electromechanical units. SF6 leaks accumulate. Transformers are aged but manageable with oil reconditioning.

Key actions:

• Transition substations that lose SCADA to staffed manual operation
• Maintain oil reconditioning program — this is the primary defence against transformer failure
• Begin replacing failed digital protection with electromechanical relays where stockpiled units are available
• Develop manual switching procedures for substations losing remote control
• Recruit and train replacement operators and maintenance staff
• Establish regional coordination centres (if national SCADA connectivity degrades, regional centres can manage local grid sections)
• Monitor SF6 stocks and plan for circuit breaker type transitions

7.3 Phase 4–5 (Years 7–30): Progressive simplification

Status: SCADA substantially degraded or lost. Digital protection largely replaced by electromechanical. Some transformers reaching end of insulation life — rewinding program necessary (Doc #69). SF6 stocks exhausted or nearly so — transition to oil or air-blast breakers.

Key actions:

• Grid operation resembles 1950s–1960s practice: staffed substations, telephone/radio coordination, manual switching, local instruments
• Transformer rewinding program at scale — this becomes one of the most important industrial activities in the country
• Training pipeline must be fully established — the original Transpower workforce is retiring or ageing
• Copper wire production (Doc #70) needed to support rewinding and general electrical maintenance
• Possible simplification of the grid — closing less-used circuits, consolidating load onto fewer, more maintainable routes
• Protection becomes simpler (fewer relay functions, less discrimination) — accepted as a tradeoff for maintainability

7.4 Phase 6+ (Years 30+): Steady-state operation

Status: Grid operates on locally maintainable technology. Electromechanical protection. Rewound transformers. Oil circuit breakers. Staffed substations. Generation from hydro and geothermal is indefinite; wind turbines may be declining. Grid is smaller and simpler but functional.

Key uncertainty: The HVDC link. If the converter stations’ thyristor valves and control systems can be maintained this long (uncertain — depends on component life and stockpile), the inter-island link continues. If not, the two islands operate independently. This is manageable but significantly reduces system flexibility.

---

8. COORDINATION WITH OTHER ENTITIES

8.1 Lines companies (EDBs)

NZ has 29 Electricity Distribution Businesses that operate the local distribution networks connecting Transpower’s grid to end users.²⁸ Under isolation, Transpower and the EDBs must coordinate closely on:

• Load shedding execution (EDBs implement at the distribution level)
• Distribution transformer maintenance (the EDB fleet of distribution transformers faces the same degradation issues as transmission transformers, at smaller individual scale but larger total numbers)
• Rural distribution network maintenance — particularly Single Wire Earth Return (SWER) lines that serve remote farms (see Doc #68)
• Workforce cross-training and mutual aid

8.2 Generation companies

Transpower does not generate electricity — it transmits it. The generation companies (Meridian, Mercury, Genesis, Contact, and smaller generators) operate the power stations. Coordination between Transpower and generators on dispatch, maintenance scheduling, and contingency planning is essential. Under isolation, the commercial electricity market mechanism becomes less relevant — dispatch based on system need rather than price bids is the practical approach, whether or not the formal market structure is maintained.²⁹

8.3 Government

Transpower is a state-owned enterprise. The government’s role under isolation includes:

• Setting national energy policy (priority loads, smelter decision, rationing framework)
• Directing Transpower’s operations if necessary under emergency powers
• Allocating resources (spare parts, skilled personnel, materials) between grid maintenance and other national needs
• Ensuring grid maintenance is resourced adequately — the grid is the enabler of everything else

---

9. CRITICAL UNCERTAINTIES

Uncertainty	Impact	Mitigation
HVDC converter station component life	Determines inter-island link longevity. Loss of HVDC is the single most consequential grid failure.	Prioritize HVDC maintenance. Stockpile thyristor modules and control components. Document everything. Plan for two-island operation.
Transformer insulation degradation rate	Determines when the rewinding program must reach scale.	DGA monitoring. Oil reconditioning. Begin developing rewinding capability early (Doc #69).
SF6 depletion rate	Determines when circuit breaker fleet transitions from SF6 to oil/air-blast.	National SF6 inventory. Leak detection program. Retain and recondition oil circuit breakers.
SCADA and digital relay lifespan	Determines pace of transition to manual operation and electromechanical protection.	Stockpile electronic spares. Document all configurations. Prepare manual procedures. Retain electromechanical relays.
Nuclear winter effects on demand	Changes the supply-demand balance, particularly heating demand.	Monitor actual demand. Maintain reserve margins. Prepare load shedding schedules.
Workforce continuity	Knowledge loss if key staff unavailable.	Knowledge capture (Section 6.1). Apprenticeship training. Printed documentation. Cross-training.

---

10. RELATIONSHIP TO OTHER DOCUMENTS

This document is part of a cluster of electrical grid documents:

• Doc #65 — Hydroelectric Maintenance: Generation side — maintaining the power stations that feed the grid. This document (Doc #67) covers the transmission system that connects them to consumers.
• Doc #66 — Geothermal Maintenance: Maintaining geothermal generation — NZ’s second-largest renewable source.
• Doc #68 — Rural Distribution and SWER: The last-mile distribution networks, particularly Single Wire Earth Return lines serving remote farms.
• Doc #69 — Transformer Rewinding and Fabrication: The technical detail of how to rewind and eventually fabricate transformers — the grid’s longest-term vulnerability.
• Doc #70 — Copper Wire Production: Essential input for transformer rewinding, generator rewinding, and all electrical maintenance.
• Doc #71 — Wind Turbine Maintenance: Maintaining wind generation capacity.
• Doc #72 — Micro-Hydro Design and Construction: Expanding distributed generation to reduce dependence on the transmission grid.
• Doc #73 — Solar Panel and Inverter Maintenance: Maintaining solar generation and its grid interface.
• Doc #1 — National Emergency Stockpile Strategy: Framework for securing spare parts, classifying essential personnel.
• Doc #8 — Skills Census: Identifying and locating grid specialists in the national workforce.

---

11. A NOTE ON THE GRID’S STRATEGIC IMPORTANCE

The grid is not one infrastructure asset among many. It is the infrastructure that makes most other infrastructure work. Without the grid:

• Water treatment fails (electric pumps)
• Sewage treatment fails
• Milking sheds cannot operate (electric milking machines serve approximately 12,000 dairy farms)³⁰
• Hospitals lose power (backup generators run for hours, not months)
• Telecommunications infrastructure loses power
• Workshops and manufacturing facilities cannot operate
• Food processing and cold storage fail
• Street lighting, traffic signals, and public safety systems fail

Every hour of grid operation is an hour in which NZ functions as a coordinated economy. Every hour of grid failure is an hour in which regions operate in isolation, without access to the generation, water treatment, and cold storage infrastructure that serves them through the grid. Maintaining the grid is the single highest-priority infrastructure maintenance task in the entire recovery.

This does not mean the grid is fragile or likely to fail. NZ’s grid is well-built, predominantly renewable, and maintained by competent professionals. The baseline expectation, consistent with the Recovery Library’s shared scenario assumptions, is that the grid continues operating throughout the recovery period. But that continued operation depends on maintenance, and maintenance depends on knowledge, spare parts, and institutional continuity — all of which require deliberate effort to preserve under isolation.

---

---

1. DairyNZ. Dairy farm numbers and electrification are well-documented. Virtually all NZ dairy farms use electric milking machines, electric fencing, water pumping, and refrigerated milk vats. Loss of electricity to dairy farms means loss of milk production capacity — a food security issue.↩︎

2. MBIE Energy in New Zealand and Electricity Authority Market Performance reports. Annual demand varies with economic conditions and weather. The figures cited are approximate pre-event values.↩︎

3. Retail electricity value estimate based on average retail electricity prices of approximately 20–25 cents/kWh (residential) applied to total consumption. Actual retail value varies significantly by customer class — industrial rates are lower (8–12 cents/kWh), residential rates are higher. The $8–10 billion figure is an order-of-magnitude estimate consistent with MBIE Energy in New Zealand data and Electricity Authority market statistics.↩︎

4. Dairy export value from DCANZ (Dairy Companies Association of New Zealand) and Stats NZ trade data. NZ dairy exports were valued at approximately $18–22 billion NZD per year in the years before the event, of which roughly half represents farmgate/production value dependent on electricity for milking, refrigeration, and processing.↩︎

5. FTE cost estimate is an imputed value, not a market wage. Under isolation, the concept of “cost” shifts from monetary to opportunity cost. The dollar figure is used here as a proxy for comparison purposes. Pre-event median earnings data from Stats NZ Household Labour Force Survey and Hays Salary Guide for engineering and trades occupations in NZ.↩︎

6. Transpower. “Our Network.” https://www.transpower.co.nz/our-network — Approximate figures. Transpower’s annual reports and asset management plans provide detailed breakdowns.↩︎

7. The HVDC link was commissioned in stages from 1965. The current system includes upgrades completed in the 1990s and 2000s. Rated transfer capacity has varied with upgrades; approximately 1,200 MW is the current effective capacity. See Transpower, “HVDC Inter-Island Link.” https://www.transpower.co.nz/HVDC↩︎

8. HVDC link technical detail from Transpower public documentation and the Electricity Authority’s information disclosure. The system uses thyristor-based line-commutated converter technology — older but well-proven. More modern voltage-source converter (VSC) technology would offer advantages but is not what NZ has installed.↩︎

9. Transpower’s transformer fleet includes units from multiple manufacturers installed over several decades. The largest transmission transformers are custom-built, heavy (80–200+ tonnes), and have lead times of 12–18 months even under normal international procurement conditions. Under isolation, replacement is not possible — hence the emphasis on life extension and eventually local rewinding (Doc #69).↩︎

10. Circuit breaker types based on general NZ power industry practice. The specific mix of SF6, oil, vacuum, and air-blast breakers in Transpower’s fleet would need to be confirmed from Transpower’s asset register. The trend over recent decades has been toward SF6, which creates the vulnerability described.↩︎

11. Vacuum interrupter rated life from manufacturer specifications (ABB, Siemens, Eaton). The rated number of operations varies by voltage class and rated fault current. At distribution voltages (11–33 kV), 10,000–30,000 mechanical operations is typical, but rated fault-current interruptions may be limited to 30–100 full fault-current operations. Once the vacuum deteriorates, dielectric withstand capability is lost and the interrupter must be replaced as a sealed unit.↩︎

12. Transpower’s System Operator function is described in the Electricity Industry Participation Code administered by the Electricity Authority. https://www.ea.govt.nz/ — The System Operator role includes real-time dispatch, frequency management, and security of supply coordination.↩︎

13. ACSR conductor service life estimates based on international utility experience. Conductor life depends on corrosion environment (coastal exposure reduces life), fatigue from aeolian vibration, and thermal cycling. Transpower’s asset management plans assess conductor condition through sampling and testing. The 50–80+ year range reflects inland conductors in good condition; coastal conductors may require earlier replacement.↩︎

14. Tower and pole lifespans based on general transmission industry practice. Steel lattice tower life is primarily limited by galvanising degradation and foundation condition. Wooden pole life depends on species, treatment method (CCA or creosote), and exposure conditions. NZ uses primarily radiata pine treated with CCA for distribution poles. See NZS 3605:2005 for timber pole specifications.↩︎

15. Porcelain and glass insulator service life based on general power industry experience. CIGRE Technical Brochure 481 (2011), “Guide for the Assessment of Composite Insulators in the Laboratory after Their Removal from Service,” provides data on polymer insulator ageing. Porcelain and glass units in service for 50+ years are common worldwide; NZ’s temperate maritime climate is relatively benign for ceramic insulators.↩︎

16. MBIE Energy in New Zealand and Electricity Authority Market Performance reports. Annual demand varies with economic conditions and weather. The figures cited are approximate pre-event values.↩︎

17. New Zealand Aluminium Smelters Limited (NZAS). The smelter’s electricity consumption is widely reported at approximately 570–600 MW, or roughly 5,000 GWh per year. Exact figures vary by source and operational conditions. The smelter’s share of national demand is large enough that its operational status significantly affects grid management.↩︎

18. New Zealand Aluminium Smelters Limited (NZAS). The smelter’s electricity consumption is widely reported at approximately 570–600 MW, or roughly 5,000 GWh per year. Exact figures vary by source and operational conditions. The smelter’s share of national demand is large enough that its operational status significantly affects grid management.↩︎

19. Aluminium smelting fundamentals: the Hall-Heroult process dissolves alumina (Al2O3) in molten cryolite at approximately 960°C. The smelting cells (pots) operate continuously — shutdown and restart is a major industrial operation requiring months and significant resources. The Tiwai Point smelter imports alumina from Rio Tinto’s operations in Queensland, Australia. On-site alumina stocks are a commercial matter but are understood to represent weeks rather than years of supply.↩︎

20. Aluminium smelting fundamentals: the Hall-Heroult process dissolves alumina (Al2O3) in molten cryolite at approximately 960°C. The smelting cells (pots) operate continuously — shutdown and restart is a major industrial operation requiring months and significant resources. The Tiwai Point smelter imports alumina from Rio Tinto’s operations in Queensland, Australia. On-site alumina stocks are a commercial matter but are understood to represent weeks rather than years of supply.↩︎

21. Transpower’s Automatic Under-Frequency Load Shedding (AUFLS) scheme is defined in the Electricity Industry Participation Code. The exact settings (frequency thresholds, percentage of load in each block, relay locations) are periodically reviewed and updated. The figures in the table are illustrative of the general scheme structure.↩︎

22. Transpower’s Automatic Under-Frequency Load Shedding (AUFLS) scheme is defined in the Electricity Industry Participation Code. The exact settings (frequency thresholds, percentage of load in each block, relay locations) are periodically reviewed and updated. The figures in the table are illustrative of the general scheme structure.↩︎

23. NZ’s ripple control system is one of the most extensive demand-response systems in the world, controlling approximately 800,000 hot water cylinders and some other loads. It uses an audio-frequency signal injected onto the power lines to switch load on or off. The technology is robust and locally maintainable. See Electricity Authority publications on demand-side participation.↩︎

24. Normal operating frequency band from the Electricity Industry Participation Code — Principal Performance Obligations. Wider bands are specified for contingency conditions.↩︎

25. System inertia concerns in NZ are documented in Transpower’s system security reports and the Electricity Authority’s work on frequency management. The integration of inverter-based generation (wind, solar, battery) reduces synchronous inertia — a challenge NZ shares with other high-renewable grids worldwide.↩︎

26. Transpower workforce figures from annual reports. The exact number varies by year. The figure cited is approximate. The number of staff specifically in the System Operator function is smaller.↩︎

27. Black start procedures are a Transpower planning responsibility. Specific stations designated for black start and the energization sequence are not publicly disclosed in detail but are understood to rely primarily on hydro stations with self-start capability.↩︎

28. The 29 EDBs (as of 2024) range from large urban networks (Vector in Auckland, Orion in Christchurch) to small rural networks. They are regulated by the Commerce Commission under Part 4 of the Commerce Act.↩︎

29. The Electricity Authority administers the wholesale electricity market. Under isolation, the commercial rationale for market-based dispatch (economic efficiency in normal conditions) is less relevant than system-need-based dispatch. The market structure may be maintained for administrative convenience but the practical dispatch decisions should be driven by security and priority considerations.↩︎

30. DairyNZ. Dairy farm numbers and electrification are well-documented. Virtually all NZ dairy farms use electric milking machines, electric fencing, water pumping, and refrigerated milk vats. Loss of electricity to dairy farms means loss of milk production capacity — a food security issue.↩︎