Inari Library · Independent analysis · New Zealand
The Opportunity Party wants New Zealand to build a mountain of clean electricity. That's the right instinct. But a mountain of power is worthless until something is plugged into it. Here's what our model says about the missing half — and what it means for New Zealand.
In plain terms
A political party (TOP) has a big energy plan. We build a model of exactly this problem — matching spare electricity to the industries that could use it. This explains their plan, our model, where they meet, and what it means for regular people, businesses, regulators and researchers. We're not endorsing anyone — we're checking the maths, and every figure is built to be checked in turn.
The 30-second version
New Zealand could productively use about three-quarters of a 30-gigawatt clean-energy build at home — roughly 77 of 108 spare terawatt-hours — with nothing exported. But only if the plan says what plugs in. The Opportunity Party's Abundant Energy policy gets the hard half right — build the power — and leaves the decisive half, which industries absorb it, as a single paragraph. That paragraph is the whole game, and it can be engineered. Below: the missing half, what it means for you, and where we'd push back.
The policy, plainly
New Zealand, TOP says, is an energy-rich country that behaves as though it's energy-poor — world-class sun, wind, hydro and geothermal, yet we still ship over $20 billion offshore every year to burn fossil fuels, and Kiwis still get cold homes and volatile bills. Their fix has three pillars: build generation (triple to 30 GW by 2050 via an Australian-style capacity investment scheme), rewire the market (one regulator not four, 6–8 lines companies not 29, ring-fenced power dividends), and electrify homes, communities and transport. And buried in pillar one is a single paragraph — section 1.3 of TOP's policy paper, which we'll call the demand paragraph: pay large industrial users, by tender, to switch to electricity, timed to soak up the new generation.
In plain terms
TOP's plan is 95% about making more clean power and 5% about using it. That last 5% — one paragraph — is the most important part, and it's the part we can engineer.
The model, plainly
Inari is an energy-quant engine. Its job is the question TOP leaves as a paragraph: when a country has spare clean electricity, what should plug into it? We call the answer an anchor — a large, patient customer (a factory, data centre, processing plant) that signs a long contract to soak up surplus power and turn it into something durable: fuel we stop importing, food, materials, jobs, resilience. The engine searches every combination of anchors for the strongest portfolios, caps any single industry, and throws out any answer a coin-flip could match. It publishes its failures — it currently marks its Australian result an "honest-fail" — a result it labels a failure rather than hiding — left visible on purpose.
In plain terms
TOP's 30 GW is a huge new water supply. Inari is the plumber working out which taps and tanks to connect so the water does something useful instead of running down the drain at midday. An anchor (錨, ikari) is just a big reliable tap promised to stay open.
Same diagnosis, complementary halves
TOP's opening line and Inari's founding thesis are the same sentence. They independently found the same disease and the same vicious cycle: high, volatile prices scare off the big industrial users, whose absent demand is exactly what would justify the generation that would lower prices. TOP's demand paragraph (§1.3) is the anchor model written as government policy. TOP has the politics; Inari has the engineering — which industries, how much each can take, what flexibility to demand, how to avoid betting the country on one sector, and how to time it.
“TOP has the politics. Inari has the engineering. One paragraph, meet one engine.”
Physical facts the model has to respect
An islanded nation can't "export" electricity
New Zealand has no international electricity interconnector and never has — the grid is physically cut off, so you cannot ship electrons to Australia. Surplus power can only become one of three real things: embodied in exportable products (fuels, ammonia — molecules, not electrons), stored (pumped hydro — the "NZ Battery" idea), or honestly curtailed (spilled — simply thrown away). Never raw electron-export. Any plan that counts on New Zealand "selling surplus electricity" abroad is counting on something that can't physically happen — so the surplus has to find a home at home.
The country can build far more than five industries
New Zealand's real menu is 15+ viable industries that could together absorb ~62 TWh at sensible sizes, up to ~119 TWh at full stretch — against 108 TWh of surplus. A plan that tries to place abundance through only a handful of anchors manufactures a leftover it then has to "export" or spill. Let the portfolio grow to the industries New Zealand actually has, and that leftover shrinks toward zero. An islanded nation that can't sell electrons doesn't need to — New Zealand has enough of its own industry to soak up nearly all its abundance at home. That is the sovereignty story, and the numbers below bear it out.
Exhibit A — Where New Zealand's 108 TWh of surplus actually goes
Sidebar — does this count the industry New Zealand already has?
Yes. The 108 TWh is what's left after the model meets a projected 2050 demand of ~56 TWh (163.9 TWh generated − 108 surplus) — and that ~56 already contains today's ~40 TWh of demand plus its electrification growth, current heavy industry included. The anchors are all net-new load on top. One honest caveat: the country's single biggest existing load — the aluminium primary smelter (~5 TWh, ~13% of demand) — is assumed to stay in that baseline. If it left, the surplus would be larger; if the country wanted, keeping it could itself be one of the anchors.
And the twist that surprises people: replacing that primary smelter with aluminium recycling wouldn't soak up the power — it would free it. Recycled aluminium takes ~0.6–0.9 MWh per tonne against ~14.5 for primary — roughly twenty times less energy. Swap primary for recycling and you release ~4.6 of those 5 TWh back into the surplus pile for other industries. Recycling is wonderful for a circular economy — but because it barely sips power, it is the opposite of an anchor. It shrinks the "where does the power go" problem instead of answering it.
The potential anchors — flexibility, size, confidence, and where they sit
Every anchor is graded on four things a plan needs to know: how flexible it is (can it ramp with wind and sun?), how big it plausibly is (a range, not a false point), how bankable it is today, and whether it's centralised (tied to a place or plant) or distributable (can be sited anywhere). Flexibility is gold — a flexible anchor eats the cheap midday surplus and eases off when power is tight. Bankable means: could you finance and build it today? — from investment-grade (banks will fund it now), to project-finance (fundable once a long contract is signed), to emerging (still needs pilots).
Exhibit B — Flexibility vs bankability (bubble = central size). Top-right is the sweet spot.
Exhibit C — How big is each, really? (low · central · high — the honest uncertainty; ~ = estimate-grade)
Exhibit D — The full table
| Anchor class | Flex | Size TWh (lo–hi) | Bankability | Central / distributed | Load shape | Sign by | Confidence |
|---|---|---|---|---|---|---|---|
| e-fuels | 0.60 | 4–60 | emerging | can be sited anywhere / distributable | flexible | 2033 | SOURCED |
| green hydrogen (industrial) | 0.60 | 2–10 | emerging | can be sited anywhere / distributable | flexible | 2031 | estimate |
| dairy process heat | 0.45 | 1–5 | investment-grade | tied to existing plant load | flexible | 2027 | SOURCED |
| cold chain | 0.35 | 1–4 | investment-grade | tied to existing plant load | flexible | 2028 | estimate |
| controlled-env. greenhouses | 0.35 | 0.2–1.5 | project-finance | can be sited anywhere / distributable | seasonal | 2029 | estimate |
| mineral refining | 0.30 | 0.7–7 | investment-grade | tied to a resource | baseload | 2030 | SOURCED |
| ammonia / fertiliser | 0.30 | 2.1–4.4 | project-finance | tied to a resource | baseload | 2028 | estimate |
| red-meat processing | 0.30 | 0.5–3 | investment-grade | tied to existing plant load | flexible | 2029 | estimate |
| wood processing heat | 0.30 | 0.2–1 | project-finance | tied to a resource | flexible | 2030 | estimate |
| aquaculture (RAS — recirculating tanks) | 0.25 | 0.5–3 | project-finance | tied to a resource | baseload | 2030 | estimate |
| grid-storage cells | 0.15 | 1–5 | project-finance | can be sited anywhere / distributable | baseload | 2031 | estimate |
| green steel (EAF — electric arc furnace) | 0.15 | 0.3–1.5 | investment-grade | tied to existing plant load | baseload | 2026 | SOURCED |
| data centres | 0.10 | 0.5–10 | investment-grade | can be sited anywhere / distributable | baseload | 2030 | SOURCED |
| silicon / polysilicon | 0.10 | 0.8–3 | emerging | tied to a resource | baseload | 2032 | estimate |
| cement process heat | 0.10 | 0.2–0.6 | emerging | tied to a resource | baseload | 2034 | estimate |
A lens the aluminium point opens up
Every material has a virgin (high-energy) path and a circular (usually low-energy) one — and whether a circular industry is an anchor depends entirely on how much power it draws. That splits the circular economy into three strategic classes:
In plain terms
One subtlety decides the class: the power a plant buys from the grid, not its total energy. Paper mills burning their own black liquor and pyrolysis plants burning their own gas both scramble the naive comparison — and an all-electric glass furnace saves energy versus virgin melting while still adding grid demand, because it replaces a fossil furnace. The classification below is sourced on that purchased-electric basis, and several industries land in a different class than intuition suggests.
"Recycle everything" and "use up our spare power" are not the same goal, and sometimes they fight. Recycling aluminium makes the country more efficient — which shrinks the very surplus TOP wants to put to work. The circular industries worth chasing for an abundance plan are the hungry ones — electric steel, electric cement, chemical recycling — where being circular and soaking up cheap power line up.
“Recycling makes a country more efficient — which shrinks the very surplus an abundance plan wants to put to work.”
Four values-driven configurations
"What to plug in" is a values choice — but a smaller one than it first appears. Because New Zealand is islanded and its portfolio is free to grow to the industries it has (12 anchors here), every posture commits 72–78 productive TWh with zero export — the country has enough real industry to soak up most of its own abundance regardless of which value you optimise. Each configuration below survives the falsification test (the coin-flip check from section 2) and the no-single-industry-over-40% cap.
Exhibit E — Four ways to spend the surplus (TWh, screening-grade — first-pass estimates, not final numbers)
| Configuration | Optimises for | Committed | The anchors it picks |
|---|---|---|---|
| Balanced / Revenue | value + spread | 77.6 | 12 anchors · e-fuels, data centres, minerals, dairy heat, cold chain, ammonia, storage cells, steel… · 0 exported |
| Resilience-first | robustness | 71.8 | 12 anchors, flexibility-weighted · 0 exported |
| Sovereignty-first | self-reliance | 77.6 | 12 anchors, all domestic · 0 exported |
In plain terms
The headline is simple and strong: New Zealand can productively plug in roughly three-quarters of its own surplus at home — about 77 of 108 TWh — with nothing exported, across every values lens. The choice isn't "revenue vs sovereignty"; it's which flexible industries fill the last few slots. Abundance, used.
Households, community, industry — the full stack
Fair question — and it matters. The surplus these anchors soak up sits on top of two layers that come first:
In plain terms
Nobody's power gets "used up" by a factory before your house is warm. The order is: homes and community first, essential services second, then industry soaks up what's spare. The anchors are tier three — they only ever touch surplus.
“The real choice isn’t revenue versus sovereignty. It’s which flexible industries fill the last few slots.”
Cheaper bills only arrive if the extra power gets used — an idle 30 GW strands investment and prices climb again. The anchors make the savings stick. The "$2,700 household saving" is a midpoint of a wide range ($1,500–$4,500) leaning on rooftop solar and an EV; renters and lower-income households see the smaller number. Ask any party promising abundance: who's plugged in, and by when?
The demand paragraph (§1.3) needs an engine: a tender that names anchor classes, demands flexibility as a qualifier, enforces a concentration cap, and times the batches (tranches) to generation. And — as section 4 above showed — for an islanded grid, don't design around "export"; design around domestic absorption, storage, and embodied-product exports. Let the anchor count scale with surplus, not sit at an arbitrary five. Consent reform flips the data-centre anchor from ≈0 to ~10 TWh; the single-regulator merge unblocks connections. Design the demand tender in lock-step with the generation scheme.
This signals where cheap, long-dated power contracts are coming. The bankable, ready-now anchors — dairy heat, cold chain, mineral refining, red-meat, data centres, green steel — are existing NZ industries, not moonshots. If a demand tender arrives, the firms that have already mapped their flexible load win the cheapest contracts. The edge is being ready to say "we can shift this much load to these hours" before it opens.
The honest gaps are the research agenda. There is no official NZ study of rooftop solar's system value or its hydro-spill interaction — the regulator's own intermittency work excludes household solar. The dry-year problem is open. The emerging-bankability anchors (e-fuels, hydrogen, silicon, cement heat) need the domestic techno-economic and demonstration work TOP wants to fund. A decade of fundable, publishable questions — and Inari's method is built to be checked and torn apart, which is what research is for.
Inari is not a political party and not for hire to one. This analysis is public-realm and party-neutral: the same engine run, anchor menu and falsification test are available to any party, official, business or researcher who wants to pressure-test an abundance claim — TOP's or anyone else's. We sell method and evidence, not endorsement. If the demand side of an abundance plan is worth engineering — and it is — we'd rather it be engineered in the open, with the failures visible, than promised in a paragraph.
So when any party promises you abundant energy, ask the one question that decides whether your power bill actually falls: who's plugged in — and by when?