Irish Electricity Outages

Power cuts on the ESB Networks distribution system. Pick your county for details.

Independent. Not affiliated with ESB Networks or the CRU.

Loading…

Every area with an outage - a full directory of the Census areas this site has recorded outages near, the named ones linking to their outage history.

Storm days are not excluded here, though ESB excludes them from its own published figures, so winter months will read worse than ESB's.

What this measures

The grade is ESB's own published aim: supply restored within 4 hours in 95% of cases. Each county is scored on the share of its fault-interrupted customers back within that window - A meets the 95% aim, then B 90%, C 80%, D 70%, E 60%, F below.

How these numbers are worked out

The aim comes from the Customer Charter the CRU approves: our aim is to restore supply within less than 4 hours in 95% of cases. The same guarantee pays compensation after 24 hours without power; outages past that mark are counted separately on each county page. A month with fewer than five faults is left ungraded, because one long outage in a quiet county would swing the letter two bands. So is a month less than five days old: there is barely a month there to judge, which is why grades appear a few days into each one. So is one in which no fault both started and was restored, because there is nothing to score. Where a county carries no letter, its own page says which of the three it is. Planned works are excluded just as the CRU's incentive excludes them: they are announced in advance and never report a restore time. Restored by ESB's first estimate is the share of faults back no later than five minutes after the first restore time ESB named for them. It is held to the first because ESB mostly revises an estimate after that time has passed. A month with fewer than five to count is left blank.

The figure the CRU actually regulates is Customer Minutes Lost - the minutes an average customer spends without supply over a year. For 2024 the target was 78.7 and ESB Networks reported 117.47 (Distribution Annual Performance Report 2024; targets set under the CRU's PR5 framework). The figure shown against a month on this page is the same measure over that month, so it is not the one to hold against those two; the comparable figure is this site's rate across everything collected, about – a year. It does not set the grade: the feed counts about – more interrupted customers than ESB finally settles on, though the durations agree - an implied – minutes per interrupted customer against ESB's –. A share of customers is immune to that overcount; a total of them is not.

How to read an outage

Each outage shows when it began and when supply was restored, both times as ESB reports them; the rows between are the customer count falling as crews restore sections. ESB back-dates the start to when the fault began rather than when it was published, so the durations here measure the outage, not the notice. Where ESB never published a restore time the end is marked as an estimate or a last sighting, never presented as a measurement.

What it cannot tell you

ESB removes storm days from its own published figures to compare with other utilities (DAPR 2024), and the Customer Charter exempts exceptional cases such as storms. Nothing in the public feed identifies a storm day, so this site excludes nothing.

Counties come from the nearest Census Small Area to each outage's coordinates - the feed carries no county field. A "customer" is a meter (a home, farm or business), and ESB publishes no per-county count, so the county figures split its roughly 2.5 million meters by each county's share of the Census 2022 population.

Source code · not affiliated with ESB Networks or the CRU.