What EIA Reported for 2025
EIA reported that Puerto Rico electricity customers averaged 36 hours of interruptions excluding major event days in 2025. That was 19% more than in 2024 and above the territory’s 2021–2025 average of 29 hours per year. Major event days added another 23 hours in 2025.
EIA said no major hurricane directly affected Puerto Rico in 2025, while heavy rain, flooding, and indirect effects from Hurricane Erin contributed to outage time. EIA reports duration with the System Average Interruption Duration Index and frequency with the System Average Interruption Frequency Index.
| Reliability Metric | 2025 Result | How to Read It |
|---|---|---|
| Non-major-event outage duration | 36 hours | Up 19% from 2024; Puerto Rico averaged 29 hours annually from 2021 through 2025. |
| Major-event outage duration | 23 hours | Reported separately from routine interruptions; heavy rain, flooding, and indirect Hurricane Erin effects contributed. |
| Non-major-event interruption frequency | Nearly 16 | Average interruptions per customer, up 13% from 2024 and about double the 2021 frequency. |
| All-event interruption frequency | Nearly 17 | Includes interruptions EIA attributes to major event days. |
The Mainland Comparison Is a Planning Signal
EIA says mainland U.S. customers generally experienced about two hours of non-major-event interruptions per year over the same 2021–2025 period. Using EIA’s 36-hour Puerto Rico result and roughly two-hour mainland reference, KilowattLogic calculates an approximate 18-to-1 duration comparison for 2025.
That comparison is useful for setting continuity-planning priority, but it is not a forecast for a specific meter. Utility territory, feeder, facility location, event exposure, restoration priority, and on-site systems can make an individual site’s experience materially different from the annual average.
Commercial Resilience Questions to Answer
| Planning Question | Evidence to Build |
|---|---|
| Which loads cannot stop? | Name the critical circuits, peak kW, starting loads, and acceptable transfer time instead of backing up the entire nameplate load by default. |
| How long must continuity last? | Use interval data and event history to model hours of required autonomy. An annual utility average is not a battery-runtime specification. |
| What already provides resilience? | Inventory UPS systems, generators, fuel logistics, transfer switches, controls, maintenance status, and safe islanding capability before adding assets. |
| Does solar or storage improve the business case? | Model resilience value separately from tariff savings, demand-charge management, incentives, export compensation, degradation, and financing. |
What Not To Infer
- The annual average does not mean every customer lost power for exactly 36 routine hours or 23 major-event hours.
- Solar panels alone do not necessarily supply power during an outage; safe islanding, controls, and an appropriate storage or generator design are separate requirements.
- The reliability data does not prove that a battery, generator, microgrid, or solar project will save money. Load shape, tariff, critical-load duration, incentives, fuel, maintenance, degradation, and financing still determine the case.
Sources: U.S. Energy Information Administration Today in Energy analysis dated August 10, 2026 and early-release 2025 Form EIA-861 annual electric power industry reliability data.