La Palma outage: critical-load BESS energy-adequacy screen#

Dated event: 2 November 2026; source: Red Electrica, Electricity supply restored in La Palma. The operator reports an island-system blackout beginning at 06:38 and restoration of all supply by 10:26 local Canary Islands time. The 228-minute restoration interval is not a verified per-customer interruption duration or a sampled power series. The operator states that causes were still under investigation at publication. No power, energy-not-served, or wind/PV generation measurements were taken from this news item.

Business case#

Screen the usable energy capacity (MWh) of a correctly designed, grid-independent, islandable BESS serving a hypothetical critical facility during an outage as long as the observed restoration interval. Compare no local generation (conservative) with an explicitly synthetic conditional solar-and-wind scenario, and use a zero-storage baseline.

Inputs and reproducibility#

  • Observed: 2026-11-02 local start/end times only, from the Red Electrica primary press release. Access checked 2026-11-06; time difference 228 minutes.

  • Modeled, not observed: constant 0.50 MW critical load, 0.65 MW discharge rating, 80% starting SOC, 90.25% round-trip efficiency (95% each direction), 3-minute resolution, assumed conditional PV/wind output.

  • Method: energy_methods.outage_resilience.minimum_islanded_capacity reuses simulate_hybrid, including inverter power, state-of-charge, charging/discharging efficiency, and direct renewable supply. A bounded bisection finds the least modeled usable battery capacity meeting all energy demand within 1e-9 MWh. It raises an error if inverter power or the imposed capacity bound makes full service impossible.

  • Baseline: no solar, no wind, no storage. The conservative no-generation required energy is calculated independently as 0.50 MW * (228 / 60) hours / (0.80 * 0.95) = 2.50 usable MWh.

  • Outputs: comparison plot of critical-load energy served versus battery usable energy capacity, plus calculated capacity thresholds.

Run from an installed repository checkout (no credentials or live requests):

uv sync --locked --all-packages
uv run --no-sync python examples/analysis/plot_la_palma_outage_resilience.py
uv run --no-sync pytest methods/tests/test_outage_resilience.py

Physical, scientific and business limitations#

This is energy-only, deterministic, synthetic screening, not a design certification, a grid restoration simulation, or an estimate of avoided customer outages on La Palma. Grid-following solar or wind plants do not automatically operate in an outage. Achieving the displayed coverage requires independent islanding/transfer equipment, suitable grid-forming and black-start capability, protection coordination, inverter surge performance, load prioritization, availability and operating reserve. None is validated by this code. The assumed 80% SOC is not a guaranteed pre-outage reserve. The modeled capacity is usable DC-equivalent energy, not battery nameplate, lifetime or warranted discharge energy. Weather during the storm, renewable resource correlation, degradation, minimum SOC, cycling limits, capex and downtime economics remain unmodeled. The published timeline is a restoration interval, not measured energy not supplied. Do not use this as an asset investment or safety decision without site measurements and qualified engineering design.

Next step: connect licensed per-facility critical-load data and outage telemetry, calibrated PV/wind observations with storm derating, and formal island-mode feasibility and reliability risk models.