Havana's grid falls below 60 MW as capital posts second daily blackout
Havana woke on 11 July 2026 with just 51.91 MW of service from a system designed for more than ten times that, after a second disconnection of Cuba's national grid.

At 16:46 UTC on 11 July 2026, the Empresa Eléctrica de La Habana (EELH) reported that the Cuban capital was receiving 51.91 megawatts of electricity against a demand profile the province's own infrastructure was built to serve at more than ten times that figure. The update, distributed through the state-aligned CubaDebate channel on Telegram, was the second such bulletin of the day. A 16:06 UTC dispatch had already confirmed partial restoration after the Sistema Electroenergético Nacional (SEN) disconnected the previous afternoon; twelve substations were back online, twenty circuits re-energised. By late afternoon the city had been put back together just enough to fail again.
The framing matters more than the megawatt. Cuba's central-grid crisis is no longer a discrete weather event or a fuel-shipment delay; it has hardened into a recurring rhythm of partial restorations followed by re-disconnections, each bulletin narrower in scope than the last. The capital's chronic shortfall is the visible edge of a national generation deficit that has been running, in various forms, for years.
A province sized for a country
Havana's installed capacity is not the binding constraint on its service level. The province's thermal plants, peaking units and distributed generation were sized for an industrial and residential load that no longer exists at the same intensity. The problem sits upstream, in the SEN's national balancing logic: when central generation falls below a threshold, the dispatcher sheds load geographically, and the capital, politically privileged but electrically fungible, takes its share. A 51.91 MW envelope for a city of more than two million people is the system telling residents, in operational language, that there is not enough fuel, gas or diesel in the pipeline to run a normal rotation.
The Cuban Electric Union (Unión Eléctrica, UNE) has attributed previous collapses to a combination of aging thermal units, deferred maintenance on Russian-built blocks, and gaps in imported fuel deliveries. The bulletin format, restoration counts measured in substations and circuits rather than megawatts, is itself a clue: the dispatcher is managing an emergency, not a normal schedule.
What the bulletins are not saying
EELH's messages do not name a cause. They do not cite a specific unit trip, a transmission line fault, or a fuel shipment that failed to dock. The technical reason for the 11 July disconnections has not been published in the Telegram thread; readers in Havana are told what is back online, not what came off. That silence is not unique to this bulletin, it is the standard register of UNE communications, which treat generation dispatch as an operational matter to be managed through the official press and Cubadebate rather than as a public dataset.
The result is a reporting environment in which residents, diaspora watchers and outside analysts are reduced to counting restored circuits as a proxy for system health. The absence of a published demand figure, an installed-capacity table, or a fuel inventory makes independent verification difficult. What can be verified is the pattern: multiple bulletins per day, a downward drift in served MW, and a restorer count that resets each cycle.
A structural frame, in plain language
What is happening in Cuba's grid is the electrical signature of a balance-of-payments problem. Foreign exchange for fuel imports competes with foreign exchange for food, medicine and intermediate goods; thermal generation, which runs on imported diesel and fuel oil, is the most exposed line item when reserves tighten. Solar build-out on the island has added daytime megawatts but does not yet supply the firm evening capacity that a heavily air-conditioned Caribbean capital needs after sundown. The grid, in other words, is reflecting the country's external account, with the load-shedding schedule as the transmission belt.
A second structural factor is the age of the fleet. Cuba's largest thermal plants entered service in the 1980s and 1990s, and several of the units that drove the post-Soviet recovery are now past their design life. Replacement parts have been sourced intermittently from Russian and Chinese suppliers, but the pace of overhaul has not matched the pace of attrition. The pattern of disconnections reported on 11 July, successful afternoon reconnection, followed by evening re-disconnection, is consistent with a unit coming back online and then tripping again under evening load.
What to watch
Two near-term markers will tell whether 11 July was an ordinary bad day or the start of a worse run. The first is whether EELH publishes a megawatt figure above 250 in its next morning bulletin; that threshold, roughly, marks the difference between managed rotation and emergency-only service. The second is whether the official press names a specific plant or fuel cause; sustained silence across a week of bulletins would suggest the dispatcher has lost optionality, not just capacity.
Over a longer horizon, the question is whether the solar build-out the government has touted in official media begins to register in the evening load curve. Daytime peaks are increasingly being met by distributed photovoltaic; the constraint is the four-hour window after sunset, when batteries (where installed) discharge and the thermal fleet has to ramp. Until that window closes, the bulletins will keep reading the way they read on 11 July: partial restoration, narrow margin, and a capital sized for a country that the grid can no longer fully serve.
Desk note: Wire coverage of Cuba's grid crisis has largely tracked UNE press releases, which use restoration counts as the primary metric. Monexus is reporting the megawatt figures published by EELH through Cubadebate as the most granular verified indicator available, while flagging that cause-of-outage detail is not being disclosed in the public thread.
Wire provenance
This editorial synthesis draws on the following public wire/social posts:
- https://t.me/cubadebate/1
- https://t.me/cubadebate/2