RESILIENCE & GRID FLEXIBILITY
Two blackouts. A better set of investment questions.
Chile and Iberia show why installed capacity, market prices and a good location score do not explain the full resilience of an energy investment.
15 October 2026 · Retrospective analysis
The summary
The 2025 blackouts in Chile and Iberia had different sequences, but both exposed the importance of how equipment, controls and system conditions interact. For investors, the useful lesson is to examine what an asset can deliver during stress, which services it depends on, and how it recovers.
What the investigations actually found
Chile’s blackout on 25 February 2025 followed the loss of a major transmission connection and the separation of the system into electrical islands that could not remain stable. EPRI’s independent review, published by the national coordinator, found that pre-event transmission transfers respected the applicable N−1 criterion. It also identified the need to strengthen contingency controls, protection testing, communications redundancy and restoration arrangements. This was more complex than a simple shortage of installed generation. CEN / EPRI, August 2025
For the Iberian blackout on 28 April 2025, ENTSO-E’s final investigation identified interacting oscillations, weaknesses in voltage and reactive-power control, output reductions and generator disconnections. Its recommendations cover operational practice, monitoring, coordination and data exchange. Reducing the event to one technology or one headline cause loses the practical lesson: individual components and their collective behaviour both matter. ENTSO-E, March 2026
Our reading: test the investment beyond normal operation
The SSI analysis uses these events to ask where dependencies and constraints concentrate. This is a retrospective interpretation to guide screening; it does not establish that SSI predicted either blackout. A credible forward-looking assessment needs dated inputs, explicit assumptions, comparison with a baseline and validation against operator evidence.
That distinction is useful commercially. A screening tool can identify questions worth investigating before a purchase or investment commitment without claiming to replace a protection study or dynamic network model. The decision is whether the current evidence supports the investment case, or whether additional engineering, operating commitments or contractual protections are needed.
Three checks that change the investment discussion
Capability: distinguish rated MW and MWh from services available at the required moment. A BESS intended to provide resilience needs an agreed reserve policy, suitable controls and a validated operating mode. Backup, islanding and grid-support functions should be specified and tested, rather than inferred from the presence of a battery.
Dependencies: review power alongside communications, cooling or water, physical access and restoration support. Two nominally separate arrangements may share a dependency. An outage scenario should identify the operational consequence and the evidence needed to confirm that dependency, without treating proximity on a map as proof.
Economics: include the cost of reserving capacity, interrupting operations and restarting. Compare practical alternatives against the same service requirement. Cheap electricity or strong historic arbitrage returns alone cannot establish resilience; equally, a theoretical resilience benefit is not automatically a contracted revenue stream.
What an assessment should deliver
A decision pack can combine a dependency register, a small set of stress scenarios, the effect on availability and cash flow, and a prioritised action list. Each conclusion should identify its source, uncertainty and owner for follow-up. That gives an investment committee something more useful than a single resilience score: conditions under which the project remains acceptable.
Takeaways
- Ask which services remain available during stress and how normal operation resumes.
- Separate verified event findings from model interpretation and prospective claims.
- Use screening to focus due diligence, then validate material technical assumptions.
Sources & reading
Report background: SSI Index, “Two 2025 blackouts, two failure modes, one lens”, 15 October 2026. xIntegra interpretation and investment questions above. Official event evidence: CEN’s publication of EPRI’s independent review, 20 August 2025; CEN’s publication of the Chilean university analysis, 6 August 2025; ENTSO-E Expert Panel final-report announcement, 20 March 2026. The source report’s proposed advance-warning interpretation is not presented here as demonstrated predictive performance.