Chile priced storage capacity for a battery almost nobody is building

Chile's new capacity-payment mechanism for storage pays full credit only to a battery that can discharge for five hours straight. A one-hour system earns about 36% of that payment, a two-hour system near 50%, per market analysis of the regulation published by AmericasMI. Chile had roughly 2,291 MW of battery storage operating by mid-2026, close to 35% of the government's 6 GW-by-2030 target, with another 4,705 MW under construction, per CNE project tracking cited by Energy Storage News. Almost none of that fleet is built to run five hours flat.

For two years the complaint about Chilean storage was that nobody would say what a megawatt of flexibility was actually worth. Decreto Supremo N°1 of 2026 answered that, folding storage into the country's power-sufficiency framework through a modified D.S. 125, with a permanent remuneration methodology and a ten-year transitional one running alongside it. That is real progress, and I don't want to undersell it. It is also not the number most of the pipeline was built to earn.

What the duration curve actually prices

A capacity payment is a proxy for one question: can this resource still be there during the hour the system comes up short. A battery that empties after ninety minutes answers that question worse than one that can hold through an evening ramp or a multi-day dry stretch on the hydro side of the system, and the sliding scale, 36% at one hour, roughly 50% at two, 75 to 85% through the three-to-four-hour band, full credit only past five, prices that difference honestly enough.

The problem is what actually got financed. Most of the storage built in Chile over the last three years sits behind a solar interconnection point, sized to absorb the midday glut and release it into the evening peak, typically two to four hours of duration. That design solves curtailment. It does not solve five-hour duration, and retrofitting duration onto an operating asset is not a software update. It is new cells, a bigger footprint, and a capex number nobody underwrote.

A capacity payment calibrated for five-hour duration tells the market what to build next. It does very little for the gigawatts already financed on a two-hour clock.

The revenue line is thinner than the headline suggests

Even where the duration qualifies, capacity payments aren't carrying the project. Analysts tracking the mechanism put the payment near $90 per kW per year, about $8 per kW-month, which works out to roughly 15% of a stand-alone storage asset's expected revenue. The other 85% is still arbitrage, the spread between the hour storage charges and the hour it discharges.

That spread is the same one this market has been quietly compressing. More solar built behind the same congested corridors pushes midday prices toward zero and widens the gap to the evening peak, which is good for arbitrage today and a crowded trade tomorrow as every other battery chases the identical hours. A lender treating the new capacity line as a hedge against that compression is overstating what 15% of revenue, haircut again by duration, can do for a DSCR.

What we think follows

Duration is now a financing decision as much as a technical one. Pushing an asset from two hours to five adds real capex per MW to capture full capacity credit, and that trade only pencils if the asset's actual job, firm capacity during scarcity hours versus intraday arbitrage against a solar glut, matches the duration being paid for. Building five hours onto a battery whose real function is shaving a midday curve spends capital on a credit the asset was never designed to earn.

There's a fair counter to this. Chile's duck curve is steep enough that even a compressed arbitrage spread may carry debt on a two-hour asset without the capacity payment doing any work at all, and treating it as a bonus rather than a load-bearing revenue line is exactly the conservative posture that makes the duration skew irrelevant to underwriting. That's a reasonable position. It just isn't the one the current wave of financing announcements is implicitly taking.

Our own storage exposure sits in Peru, not Chile. Olmos Wind, our most advanced project at roughly 135 MW in Lambayeque, carries battery storage built into the design rather than added after Ready-to-Build, targeted for the first quarter of 2028. Whatever Peru's own storage tariff eventually pays, the discipline of treating storage as part of the asset rather than a later retrofit is the one choice that travels across both markets.

The open question

The ten-year transitional methodology sitting alongside the permanent one is the piece I haven't been able to pin down. If it grandfathers the two-to-four-hour fleet already under construction onto softer terms, the duration mismatch mostly resolves itself and this becomes a story about the next investment wave, not the current one. If the transitional schedule applies the same haircut from day one, a lot of recently financed Chilean storage is earning less capacity credit than its sponsors modeled, and the market hasn't fully priced that in yet. I don't know which it is, and I would want to before sizing debt against this revenue line.

See the full active portfolio or read more Insights.