New Zealand's Dry-Year Problem Needs a Different Kind of Battery

 Every conversation about New Zealand's energy future eventually circles back to the same uncomfortable question. When the rain stops, what runs the country?

Eighty per cent of our electricity comes from renewables. We are, by most measures, in a position the rest of the developed world envies. But the asterisk on that headline is now well understood: in a dry year, our hydro reservoirs cannot do the heavy lifting we ask of them, and the only tool we have left is to burn gas at Huntly. Through the 2024 winter, that asterisk became a daily news story. Wholesale prices spiked. Industrial users were paid to switch off. Manapōuri ran lower than anyone wanted. And the country was reminded, again, that 80 per cent renewable is not the same as 100 per cent secure.

The standard answer to dry-year vulnerability has been some combination of hope, gas, and a longer-term debate about whether to build pumped hydro at Lake Onslow. The answer that has actually started arriving — quietly, in megawatt and megawatt-hour increments — is grid-scale battery storage. Contact Energy's NZ$235 million expansion of the Glenbrook battery announced in February 2026 will deliver 200 MW for two hours from early 2028. Genesis Energy's 100 MW / 200 MWh battery at Huntly is scheduled to come online later this year. Transpower's project pipeline now sits at over 6 GW of storage and renewables enquiries.

This is good news. It is also the wrong shape of news to solve a dry year.

Two hours is not eight days

Lithium-ion batteries — the chemistry behind almost every grid-scale battery being commissioned in New Zealand right now — are extraordinary at what they do. Sub-second response. Compact footprint. Mature supply chain. And, crucially, they are economic at two-hour discharge durations. Stretch a lithium system to four, eight, or twenty-four hours of storage, and the maths breaks. You do not just need more batteries — you need three to four times the cells, three to four times the fire-suppression load, and three to four times the replacement burden when the cells reach end of life around year twelve.

Two-hour batteries are excellent at smoothing an evening peak. They cannot smooth a dry week.

The honest scale of the dry-year problem is measured in days, not hours. A typical Manapōuri shortfall during a dry winter sees the country short perhaps a hundred or two hundred gigawatt-hours of expected generation across several months. Solar and wind cannot fill that gap reliably without somewhere to store the surplus from windy or sunny days for the calm, cloudy ones that follow. And that is what long-duration energy storage — eight hours and up, sometimes much more — is designed to do.

The chemistry that does the other shape of the problem

This is where flow batteries enter the picture. Vanadium redox flow batteries store energy not in stacked cells but in two large tanks of liquid electrolyte. Power and energy are decoupled — the cell stack determines how fast you can discharge, the tanks determine how long. Doubling the storage duration means buying more electrolyte, not more cells. The cost curve goes the other way to lithium.

The technology is not new. Sumitomo Electric's 60 MWh installation at Hokkaido's Minami-Hayakita Substation has been running since 2015. The same company's 2 MW / 8 MWh battery at a San Diego Gas & Electric substation completed a seven-year independent performance evaluation in 2022, with reported availability above 99 per cent. A peer-reviewed study published last year in the Journal of Energy Storage documented a commercial flow battery operating for over twelve years with very little capacity loss and no leakages since commissioning.

What flow batteries cannot do is matter quickly in small spaces — they have lower energy density than lithium and a slightly larger physical footprint. What they can do, that lithium cannot, is store useful energy for many hours per day, every day, for two to three decades, without degrading and without burning. That is the shape of the dry-year problem.

What this means for the next ten years

The grid we are building between now and 2035 will be a hybrid one. Lithium will continue to win the short-duration market — frequency response, two-hour arbitrage, peaker replacement. Long-duration, dry-week, multi-day storage is a different market, and it will need different chemistry. Some of that need will be filled by pumped hydro, if Onslow or its successors get built. Most of it will not.

The question for utility planners, lines companies, and large industrial users is whether they wait for that storage market to arrive, or help build it. The first NZ project to deploy a multi-hour vanadium flow battery alongside an existing solar farm or wind site will not have to be enormous to be valuable. It will, however, have to start. Real-world deployments paired with renewable generation are how the rest of the world has tested this technology, and that is the work New Zealand needs to begin in 2026 rather than 2030.

We have spent fifteen years building a renewables-heavy grid we are rightly proud of. The next fifteen will be defined by how well we close the gap that grid still has — the gap that opens whenever the rain doesn't fall.