Storm Floris, negative prices
Mon 4 Aug · Tue 5 Aug
Research notes
Figures are for London unless they say otherwise.
What happened
- Mon 4 Aug: London min -1.5p at 23:30, avg 11.3p, max 24.2p. Below zero for 1h in total: a half-hour at 13:00 and the 23:30 slot. The 23:30 slot belongs to the next day's auction, which runs 23:00 to 23:00, so it is really the start of Tuesday's event.
- Tue 5 Aug: London min -4.2p at 12:00, avg 4.5p (the lowest daily average of the summer in this list), max 25.6p. Below zero for 13h, longest spell 6.5h from 09:30. Lowest anywhere -5.0p in Northern Scotland. The longest negative day of the summer.
Likely causes
- Storm Floris. Bloomberg, 4 Aug 2025, headline "UK Power Prices Go Negative as Storm Floris Batters Scotland": prices for Tuesday went below zero, negative "for as many as eight hours", and the day-ahead rate fell to £16.79/MWh, "the lowest level since late May". I could only read the headline and the search-engine summary because the article is paywalled, so check the wording before quoting it. (confirmed by source, with that caveat) Bloomberg, 4 Aug 2025
- The Met Office named Floris on 1 Aug and issued an amber wind warning for much of Scotland for 4 Aug. The peak gust was 82mph at Wick Airport, "provisionally the joint-highest wind gust reported in Scotland in August"; Northern Ireland equalled its August record with 66mph. The Met Office called it the sixth named storm of the year. (confirmed by source) Met Office, 4 Aug 2025, Met Office naming announcement, Met Office August 2025 stats
- Wind output after curtailment averaged 10.7GW on 4 Aug (peak 15.2GW) and 11.7GW on 5 Aug (peak 14.8GW, fourth windiest day of the summer). At 23:30 on 4 Aug, when the first negative price of the Tuesday auction landed, wind was 12.2GW and Britain was exporting. (confirmed by source) Elexon FUELHH 4 to 5 Aug
- Tuesday was the bigger price event even though Monday was the stormier day, and the data suggests why. The Met Office forecast for Tuesday was still breezy but "drier and sunnier conditions further south"; solar peaked at about 10.5GW at 13:00 on 5 Aug against about 6.9GW on the stormy Monday, and transmission demand averaged only 19.9GW on Tuesday against 21.8GW on Monday. So Tuesday had Floris's wind plus sunshine plus lower demand. (confirmed by source for the figures, inference for the reading) PV_Live 4 to 5 Aug, INDO 4 to 5 Aug, Met Office, 4 Aug 2025
- A lot of wind was switched off because the grid could not carry it south. Montel says wind curtailment peaked at 6.7GW on 4 Aug, the highest of the quarter. The Telegraph, citing the Wasted Wind website run by Robin Hawkes of Octopus Energy, put the cost of curtailment at £14.4m on the Monday and £19.3m on the Tuesday, £33m in all. NESO's own daily balancing totals were £18.84m and £21.98m (those cover all balancing actions, not only wind). (confirmed by source) Montel Q3 2025 summary, Telegraph via Yahoo, 6 Aug 2025, NESO daily balancing costs, Aug 2025
- NESO's commentary for 5 Aug: market "predominantly long until 09:30", with thermal constraints "predominantly in Scotland/Northern England". On 4 Aug it needed trades for both "Downward Regulation and Margin". (confirmed by source) NESO daily balancing costs, Aug 2025
Wider context that week
- Curtailment does not show up in Agile. The day-ahead price is a single national price set before NESO deals with grid bottlenecks, so Scottish wind that was later paid to switch off still helped push the national price negative. That is the issue the zonal pricing debate was about; the government had rejected zonal pricing on 10 Jul. (inference for the first part) Norton Rose Fulbright on the REMA summer update
- Nuclear output was low all month at about 3.0 to 3.6GW (it had been about 4.8GW in late June). EDF later said 2025 nuclear output was 12% down on 2024, mainly because of an extended Hartlepool outage. Less nuclear means fewer negative prices than there would otherwise have been, so it was not a cause here. Elexon FUELHH 5 Aug, EDF fleet update
- An academic piece on why a storm this strong is unusual, and more damaging, in August when trees are in leaf: The Conversation.
- An independent blog (not an authoritative source) argues NESO switched interconnectors from import to export during the storm and that the biggest potential surpluses came on 5, 8 and 9 Aug when the sun came out: Chris Bond, 20 Aug 2025.
Confidence
High. This is the one event of the summer with direct press coverage linking the GB price to a named cause, backed by the Met Office, Montel, NESO and the generation data.
Gaps
- Sources conflict on how long prices were negative on 5 Aug. Bloomberg says "as many as eight hours"; the site's Agile data shows 13 hours at or below zero in London. I could not establish why. Bloomberg may be counting a different auction (hourly rather than half-hourly) or only hours strictly below zero. Treat the 13h as the Agile fact and do not attribute it to Bloomberg.
- I could not read the Bloomberg article in full, so the lowest hourly price in £/MWh and any wind forecast figures it quotes are not captured here.
- I found no NESO statement of a wind generation record during Floris.