Grid outages and the connections due by 2030

About 8,000 a year is from NESO’s Clean Power 2030 advice (5 November 2024, page 39). The outages taken each year are from its Clean Power 2030 data workbook (sheet NA1), drawn from its outage database. The workbook also has rows 5% to 20% above the 2023/24 figure. These are flat uplifts, not a forecast, so we leave them out. The split of 2023/24 is of “the overall system access activity”, which NESO took from unavailability data in its system performance report for 2023/24, not of outages; it is Figure 9 of NESO’s Transmission Acceleration consultation (February 2025, pages 53 and 54). That NESO doesn’t forecast outage numbers is its reply to an Environmental Information Regulations request (FOI/25/094, 10 September 2025). NESO’s own figures differ in places: its Clean Power 2030 advice says 40% of outages are agreed 12 months ahead (page 39), its Clean Power 2030 next steps (June 2025) 50% (page 25); and its web page “What happens when there’s a request for an outage?” puts the final year-ahead plan at around 2,500 outages, its February 2025 consultation the plan for 2025/26 at over 4,000 (page 55). NESO’s Grid Code proposal GC0191 (October 2026) gives 8,100 planned outages for 2025/26 (page 17).

Each owner reports under the GB Network Access Policy (Appendix A, KPI 1): the outages in its year-ahead plan submitted at week 49, how many of these it carried out in the regulatory year, and the share. The reports call these outages “delivered”; this page says “carried out”. Either way, they are the plan’s outages that went ahead, and all three count only those, not every outage done in the year: NGET reports its “YA outages delivered” against its plan; SHET counts “only the … planned outages (KPI 1b) included in the week 49 … baseline plan”; and SPT “delivered 563 of the 984 planned outages in the year ahead plan” in 2025/26, its KPI 9 table giving 349 on time, 128 early, 86 late and 421 that didn’t start. Outages added during the year come on top. The reports differ: NGET’s and SHET’s plans count in-service bookings too, works done without taking the equipment out of service, such as telecoms works or works that risk tripping a circuit, while SPT’s report doesn’t say whether its plan does. None says how close to plan counts as on time. Each figure is the report’s own, with its page, in the reviewed record the page is built from.

A stage is a row of the register still to connect. Its year is the year of its own connection date (the register’s MW Effective From), and it connects at its project’s connection site. “Already past its connection date” means before the latest register. Stages still to connect with connection dates before 2026 (50 stages, 5.3 GW) and those due after 2030 are left out. A substation with stages due in more than one year counts once in each of those years, and once in the total. How often connection dates were pushed back comes from Which connection dates slip least?, over the full years 2021 to 2025: how often a connection date that far ahead at the start of a year was pushed back within that year. That is history, not a forecast. The page doesn’t count outages per connection: one outage can serve the works of several connections, and others serve none.