Peru's reserve margin treats hydro and gas as independent risks. This year they aren't
Hydro supplied 50.4% of Peru's electricity from January through September 2026, per COES. ENFEN's latest communiqué puts the probability that this year's El Niño Costero reaches extraordinary magnitude at 88% for October, and a Cenepred risk assessment has already flagged more than 110 hydroelectric plants nationally as exposed. Peru's adequacy planning is built on the assumption that a hydro shortfall and a gas shortfall arrive separately, on their own schedules. This year they are arriving together.
That assumption usually goes unnoticed because most years it holds well enough. A dry month here, a maintenance outage there, each priced as its own independent probability and summed into a reserve margin that looks comfortable on paper. The problem shows up only when one event drives both inputs at once, and an extraordinary-magnitude El Niño is exactly that kind of event.
Where the independence assumption breaks
Reserve margin and capacity adequacy studies work by weighting each generator's nameplate capacity against its own historical availability, then summing across technologies. Implicit in that arithmetic is that a hydro unit being short of water and a gas unit being short of fuel are unrelated draws. Fine, most of the time, because Peru's hydrology and its gas supply chain don't normally share a cause.
El Niño removes that separation. It cuts inflows to the hydro fleet concentrated in the center and south of the system, and it does so in the same window the system would normally lean harder on thermal to cover the gap. The problem is that gas generation in Peru is already running close to its own ceiling. A former deputy energy minister said as much this year: the expected drop in hydro output can't simply be offset with more gas-fired generation, because gas output has nowhere left to go. The two legs of the reserve margin that are supposed to back each other up are both constrained by the same season.
A megawatt of reserve margin is not one number. It is a probability, and this is the year the probabilities stopped being independent.
This isn't a forecast anymore, it's a trend
In May 2026 alone, hydro generation fell 6.21% against May 2025, while thermal generation rose 27.61% over the same comparison, per INEI. That is the system already substituting at the margin, months before the rainy season ENFEN is watching most closely even arrives. COES itself, speaking at its Smart Energy Summit 2026, warned that the exhaustion of low-cost generation will push system costs up sharply, and tied that warning to the same three causes at once: a shortfall of new renewable and hydro capacity expected before 2029, persistent transmission delays, and rising climate exposure from El Niño. Those are not three separate risks stacking by coincidence. They compound.
What we think follows
If hydro and thermal shortfalls correlate under El Niño and non-hydro renewables don't share that correlation, then wind and solar capacity should be worth more to system adequacy in exactly the years the reserve margin math is weakest, and Peru's capacity framework doesn't currently credit that. Colombia ran into a related version of this problem when its reliability charge derated solar statistically and priced it close to a thermal firm-energy obligation, without weighting for hydrology correlation either way. Peru hasn't built its own capacity remuneration mechanism yet. Whoever designs it should start from the correlation, not retrofit it later the way Colombia is now having to.
There's a real counterargument, and it deserves a line rather than a footnote. El Niño in Peru isn't purely a drought signal for the north coast, where much of our own solar pipeline sits. It also brings heavier rainfall, flooding risk to access roads and substations, and its own construction and O&M headaches. Decorrelation from hydro inflow is not decorrelation from El Niño altogether. It's a different exposure channel, which is a weaker claim than "independent," and worth underwriting as such.
None of ZFC Energy Group's pipeline carries hydrology risk, by design rather than by luck of geography. Olmos Wind, approximately 135 MW in Lambayeque with battery storage integrated into the project design, and Frontera, approximately 135 MW also in Lambayeque, are both wind assets targeting Ready-to-Build in 2028. Aurelion, 300 MW of bifacial solar in Moquegua, targets RtB in 2028, and Solar II, 300 MW of bifacial solar also in Moquegua, targets RtB in 2029. All four sit inside an 870+ MW pipeline toward a 2 GW+ target across LATAM by 2030, and none of it depends on a reservoir.
The open question
Does Osinergmin, once it finally designs a capacity remuneration mechanism for Peru, build in a correlation-aware credit for technologies that don't draw on the same water the hydro fleet needs, or does it default to the same statistical derating Colombia already applied to solar and is now revisiting? I don't know which way that goes, and the agency hasn't said.
See the full active portfolio or read more Insights.