Australia’s CSIRO has published the final version of its annual GenCost 2025-26 cost report, and the storage numbers now move in opposite directions: pumped hydro is revised up, grid-scale batteries down. Because GenCost is the cost reference baked into AEMO’s planning scenarios, state storage tenders, and utility project valuations, this is not a minor data revision – it changes the default answer to how Australia’s electricity market will firm the gap left by retiring coal plants.
The immediate consequence is practical: for the 2-8 hour firming window that most planning models treat as the critical need of the next decade, batteries get cheaper on paper while their main competitor gets more expensive. That shift will show up in the next round of procurement decisions long before it shows up in any spreadsheet.
What GenCost is – and why Australia builds policy around it
GenCost is the CSIRO’s annual cost reconciliation for energy generation and storage technologies – generation capital costs, operating costs, and storage cost curves presented in a standardised, consistent format. It is produced each year through an industry consultation process involving AEMO, federal agencies, and sector stakeholders, and it is among the most heavily cited documents in Australian energy policy. The final iteration of the 2025-26 report has now been released, which means the adjustment affecting pumped hydro and batteries is the settled version that planners and investors will mention in their modelling.
That is not the usual abstract engineering relevance. AEMO’s Integrated System Plan, which determines what gives Australia’s grid over the next couple of decades, relies on GenCost-style assumptions as its planning foundation. State and federal procurement mechanisms – the Capacity Investment Scheme is the largest – use those same costs to set tenders, buy electricity and structure project support. Even private PPAs and financial bank models that are rock-based on their costs use their own risk or methodology; it takes GenCost as a point of reference.
So the phrase “storage costs are finished” is meaningful. It is a whole-of-market price for the storage asset itself, applied by all parties in the comparison. Under the firming premium valuations that most storage stacks rely on, the new inputs directly influence what type of asset is considered viable in 2030 at the same spread, regardless of plant operation.
Battery deflation and civil-construction inflation: why the curves are splitting
There are actually two separate global underlying trends behind this divergence. Battery costs have been declining for years as lithium-ion production capacity continues to exceed measured demand, especially with China’s cell manufacturing. Which means solar modules and more amenities come down globally, but at much larger scale, and the system-level cost of a large-scale battery bank has fallen in line with that broader trend – in many, on the order of a half-dozen next-generation programs have gone through. In Australia, cell prices and system integrators are more competitive than they have been in previous cycles, so the cost per kWh to the storage is lower than pasta years. That reflects not just the price curve but the sector’s scale.
Pumped hydro sits on a completely different stage: civil engineering. A pumped storage plant is a construction project with a turbine hall, man-made reservoirs or existing dams, water management, sensors, and geotech risk. Those costs are far less tied to the cheap manufacturing line and far more exposed to typical conditions: concrete supply, steel, skilled civil labor, and risk premiums already inflated by cost overruns on long durations. The upward revision in GenCost’s pumped hydro cost estimates points to those factors now being sharpened in Australian conditions, where the cost of large infrastructure, and insurance against, have increased at a rate that exceeds general inflation.
The cross-sector context matters here. Another way to see the split is that the cost curves have changed for exactly two different things: the “commodity” part of storage (batteries) follows the global manufacturers’ supply, and the “site asset” part of storage (pumped hydro) is, instead, locked to the location and spent civil project costs. The longer the duration of storage means you are buying more “site asset” hours, so the cost in the curve is not linear. This confirms what GenCost’s data direction since: for daily cycling and multi-hour overnight firms, batteries are – but for week-long, multi-season storage, the ordinary economics are still likely to prefer a site that has the natural water structure.
The technical shift in Australian planning is not necessarily total disappearance; it is more a freight. Large-scale pumped hydro no longer makes sense for its hours by default. The 4-8h battery has cost less, is faster to deploy and has no geographic issue, so it becomes the margin. Pumped hydro can survive as a store for multi-day purposes, where the fraction of cycles is lower, and with built-in issues that energy storage costs still reach for. But without spread to multi-day, the case is debatable.
From a network perspective there is also an ancillary cost that favors batteries: location. Pumped hydro is determined by the terrain, so it often requires creating a new transmission corridor and a separate transmission investment before it even supplies. Battery cost and time produced.
Who the GenCost 2025-26 storage revision affects
- Utility and system planners: Re-run the integrated system plan and firming portfolios with a lower battery capex and a higher hydro capex; expect the model to install fewer hours of pumped hydro, finance them later, and emphasize 4-hour batteries in the near-term replacement sequence.
- Storage developers: The battery business model gets stronger across most states, but a lower capex curve also means intensifying competition drives down revenue-share and margins – the key is to secure land, grid connection and offtake before the price compression fully resets. Pumped-hydro CBDs will need to pass a stricter test: show a specific 12-hour-plus role that batteries cannot fill, or the cost curve will not discipline the “compatible” decision.
- Policy analysts and regulators: Review the tender threshold based on the stated storage costs – firms using the previous pumped hydro number may be stalling. The federal and state storage capacity targets, which were set when the two techs were closer on levelized cost, should be rechecked and tested for building efficiency.
- Investors and lenders: In a portfolio of 2-8 hour storage projects, the revised levels increase the attractiveness of battery-backed contracted schemes, and the proposed more aggressive returns require more of the pipeline; the capital allocation for pumped hydro should be expensive, but the most valuable long-story infrastructure with VRE – will require higher return.
What to watch next: signals from the storage cost curve
- AEMO’s next ISP response: Watch the capex inputs and storage portfolios in the next program – if the change holds, the system plan should build more hours of battery firm but keep it short, and commence any asset that models over-provisioning of long-duration pumped capacity.
- Cost and schedule updates on large hydro projects “for extant assets”: Snowy 2 and yesterday, whether the next any of those projects reference updated capital cost, and in schedule. They are public data points either confirming or countering new cost schedules, including tests with projected GenCost.
- Battery energy storage duration under risk in spot markets: 2026 price and curtailment once the new spot: If average daily prices become routinely elongated (5-hours periods), the cost decline and matching arises without an EV – and the year will be critical for the new cheaper economics to actually find marginal revenue.
- The draft GenCost 2026-27: Watch how large battery contracts are placed into the new forecast (still sharper change) and whether any storage site continues upward – and if the gap keeps widening, it captures the longer-term tradeoff between mass-produced and site-scale storage.
Bottom line
The final GenCost 2025-26 defaults place Australia profitably clearly in beyond-the-real climate: cheaper batteries own the front end of the firming timeline, and pumped hydro must survive in less than the sharpened cost of its civil works. The real effect is not a sudden cancellation of the long-duration storage future; instead, it matters much more how long the capacity goes without already – the default at 8-hour storage is now even more of a stretch being built, and hydro’s only case is the weather-based month mode. The planner whose work benefits from this report has a single new bottom line: treat storage duration as a variable, not an option, and update the formula accordingly.
Read the full report at Energy Storage News
Note: facts and figures attributed above to Energy Storage News reflect that outlet's original reporting. Broader context, cross-sector connections, and forward-looking scenarios reflect independent analysis by our editorial team.
About this article: Drafted by Energy Ai with AI-assisted research and writing based on public reporting, then reviewed under our editorial process before publication.
Leave a Reply