GenCost 2025-26 – What it does (and does not) say

On 17 December 2025, CSIRO and AEMO released the draft GenCost 2025–26 Report (GenCost), an annual report about projected capital costs for electricity generation and storage.

It updates costs using two metrics:

• Levelised cost of electricity (LCOE): the unit price a generator must recover to cover capital, operating and financing costs; and

• System levelised cost of electricity (SLCOE): total system-wide costs spread across electricity supplied in a given year, reflecting how different technologies interact within a defined system boundary.

GenCost has been the subject of significant debate, reflecting broader questions about how reliability, flexibility and risk should be valued in a decarbonising electricity system. The dispute is less about the accuracy of GenCost’s figures than about what they actually measure.

What does GenCost measure?

GenCost is sometimes cited in support of the proposition that solar, wind and batteries represent the cheapest forms of new electricity generation, and that Australia should therefore deploy only those technologies. That conclusion, however, reflects a misunderstanding of what the GenCost report says.

The GenCost report estimates the cost per megawatt of constructing individual generation or storage facilities. It does not identify the combination of technologies that would deliver the optimal pathway for reliably transitioning the electricity system to net zero.

GenCost’s emphasis on LCOE has attracted criticism: LCOE is suited to comparing broadly similar generation assets, but less effective when applied to technologies that make different contributions to system reliability and security.

In particular, LCOE understates additional system costs associated with an electricity grid with a high penetration of variable renewable energy. Cost comparisons that focus on averaged metrics fail to reflect the value of dispatchability, certainty and resilience.

The 2025-26 draft of the GenCost report responded to these criticisms by introducing SLCOE, a material refinement of the GenCost framework, and an acknowledgement that generation technologies cannot be assessed meaningfully in isolation from their system impacts.

However, SLCOE does not fully addresses the underlying limitations. It still relies on averaged assumptions, is sensitive to system boundary choices, and may not fully capture the operational value of dispatchable capacity.

The draft report itself states that LCOE and SLCOE estimates both indicate “the minimum price needed for investors to enter the market”. That is, they are directed to marginal investment decisions, rather than identifying the mix of technologies best suited to Australia’s ongoing electricity needs.

The draft report does not explicitly model or impose the National Electricity Market’s formal Reliability Standard (which expects to have enough supply to meet demand 99.998% of the time, in every region every financial year).

Instead, it approximates system “reliability” indirectly by modelling technology mixes that include storage, transmission and firming capacity, and then deriving average costs from those mixes, rather than testing whether they would formally meet the NEM Reliability Standard in all credible stress scenarios.

If solar had the lowest LCOE or SLCOE, it would not follow that a 100 per cent solar electricity grid would be sensible or reliable.

GenCost's key findings at a glance

Capital cost trends

For specific technologies, the draft GenCost 2025-26 Report states that in 2025-26:

batteries will continue to show significant double-digit cost reductions;

large-scale solar costs are projected to decline, after three consecutive years of cost increases;

onshore wind costs show tentative signs of stabilising;

offshore wind is still uncertain, with high capital costs but possible upside if deployment increases; and

nuclear, coal and gas open cycle cost trends will increase (due to higher steam and gas turbine technology costs).

Significant capital cost shifts for key technologies

Capital costs have been volatile over the past four years.

Table 1 shows year-on-year real cost changes for selected technologies, highlighting stabilising onshore wind costs, renewed declines in solar, and significant recent cost reductions for batteries.

Longer-term upward pressure on capital costs is being driven by persistent increases in land costs (typically 2%–9% of total capital costs) and above-inflation growth in construction labour costs.

Current capital costs are an unreliable guide to future costs

Renewables capital cost projections have repeatedly failed to anticipate secular cost shifts driven by technological innovation, manufacturing scale and learning-curve effects, particularly for modular technologies that can be mass-produced and deployed incrementally.

This pattern is clear in solar PV, lithium-ion batteries and onshore wind, where costs fell far faster than expected as global manufacturing scaled, performance improved and supply chains matured.

GenCost reports current costs and historical trends; it does not forecast future cost trajectories. Given the rapid and often unexpected cost shifts seen across major renewable technologies, GenCost figures should not be used to definitively rule out particular technologies.

What's next?

The consultation period for the Draft 2025-26 Report closed on 2 February 2026 and the final report is expected to follow later in 2026.