ocean · energy · infrastructure · family: it worked in the lab
the cable breaks, thewind farm goes dark
Offshore wind submarine power cable failure detection
Problem statement
Submarine power cables are the critical link between offshore wind turbines and the onshore grid, and cable failures are the largest source of offshore wind insurance losses: GCube's "Uncharted Waters" claims report found subsea cables accounted for 30% of claims incurred and over 50% of total claims spend across 2010–2020, and IMCA reports insurers attributing between 50% and 70–80% of their offshore wind claims to subsea cable damage. Locating a fault in a buried submarine cable requires shutting down the affected circuit (losing tens to hundreds of MW of generation), deploying specialized cable-laying vessels (typical day rate around £150,000; BVG Associates), and excavating the cable from the seabed for repair. Fault location from the cable terminations (e.g., time-domain reflectometry) is coarse over long buried circuits, so extensive seabed survey is needed to find the actual damage point. Average downtime is around 40 days for an inter-array cable repair and 60 days for an export cable repair, with export cable damage costing $10–30 million per incident (University of Strathclyde figures cited by IMCA).
Why this matters
Global offshore wind capacity reached 83 GW by the end of 2024, and GWEC forecasts about 350 GW of additions over 2025–2034 (441 GW total by 2034). Installed offshore wind cable has grown six-fold from 9,000 km (2015) to 55,500 km (2025), with a further ~117,640 km forecast for 2026–2040 (TGS 4C). The average power cable failure rate for European wind farms is approximately 0.003 failures per km per year (University of Strathclyde study cited by IMCA) — low by onshore standards but devastating at offshore repair costs. Damage incurred during installation contributed to 46% of all power cable failures (ORE Catapult, cited by IMCA), and 44% of cable claims spend traces to contractor error in transit and cable laying (GCube). Faster, cheaper cable fault detection and repair would significantly reduce the levelized cost of offshore wind energy.
What’s been tried and why it hasn’t worked
TDR and frequency-domain reflectometry (FDR) provide fault location from cable terminations but with accuracy limited by cable impedance variations, joints, and branching points. Distributed temperature sensing (DTS) using fiber optic cables co-installed with power cables can detect hotspots indicative of insulation degradation, but most existing cables lack fiber optic elements. Partial discharge monitoring can detect incipient insulation faults but is unreliable over the long cable runs typical of modern offshore wind export circuits due to signal attenuation. ROV-based visual inspection can find external damage (anchor strikes, abrasion) but cannot detect internal insulation degradation. The fundamental challenge is that the cable is buried in the seabed, in an electrically noisy marine environment, and the failure modes (water treeing, mechanical fatigue, thermal cycling) develop internally and invisibly over years.
What would unlock progress
Embedded distributed sensing (acoustic, thermal, strain) integrated into cable design at manufacture rather than retrofitted. Real-time cable health monitoring systems that can detect degradation years before failure, enabling preventive repair during planned maintenance. Improved fault-location accuracy would shorten repairs by cutting the seabed search phase. Novel cable designs with modular repair sections or redundant conductors could reduce repair scope. On the vessel side, faster cable repair techniques that don't require full cable recovery to the surface would dramatically reduce repair duration and weather sensitivity.
Entry points for student teams
A team could design a cable health monitoring system concept, specifying sensor types, data processing architecture, and degradation detection algorithms for a representative offshore wind cable. Test data from cable manufacturers' accelerated aging tests (some published in CIGRE reports) provides a starting point. Alternatively, a team could model fault location accuracy improvement using distributed acoustic sensing (DAS) fiber integrated into cable design, simulating signal propagation and detection sensitivity.
Genes
Structural cousins: stuck the same way, in other fields
Sources
IMCA (International Marine Contractors Association), "Unravelling subsea cable failure in offshore wind" (commentary citing University of Strathclyde 2019 study and ORE Catapult), accessed 2026-08-20; GCube Insurance, "Uncharted Waters" offshore wind claims report (2010–2020 data), via offshoreWIND.biz, "Subsea Cables Account for Over 50 Pct of Total Claims Spend in Offshore Wind – GCube Insurance," 2021-11-09, accessed 2026-08-20; CIGRE Working Group B1.57, Technical Brochure 815, "Update of service experience of HV underground and submarine cable systems," 2020, accessed 2026-08-20; GWEC, "Offshore wind installed capacity reaches 83 GW as new report finds 2024 a record year for construction and auctions," accessed 2026-08-20; TGS 4C, "4C Analysis Flags Rising Subsea Cable Demand and Vessel Bottlenecks Driven by Offshore Wind Growth," accessed 2026-08-20; BVG Associates et al., "Guide to a Floating Offshore Wind Farm," § I.2.3 Cable-laying vessel, accessed 2026-08-20 go to source 1 ↗ go to source 2 ↗ go to source 3 ↗ go to source 4 ↗ go to source 5 ↗ go to source 6 ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
Feeds C1 (lab-to-field sensor gap): cable testing in factory conditions doesn't replicate the combined mechanical (wave loading, tidal burial/exposure cycles), thermal (seasonal, operational), and chemical (seawater ingress) stresses that drive in-service failures. The `temporal:worsening` tag passes: offshore wind capacity is scaling faster than cable reliability improvement (specific mechanism), cable failure rates are not decreasing despite design improvements (trajectory evidence), and longer export cables for deeper/farther sites increase exposure per circuit (barrier worsening). Related to `energy-floating-offshore-wind-structural-mass` (another offshore wind infrastructure challenge). Distinct from existing ocean-domain briefs, which focus on biological/environmental ocean science rather than marine energy infrastructure.
Reconciliation 2026-08-20: Two of the three original citations could not be verified as real documents: 4C Offshore publishes offshore wind and subsea cable market intelligence (transmission/export cable databases, now under TGS) but no "Submarine Cable Failure Database, 2023" was found, and CIGRE WG B1.57's actual publication is Technical Brochure 815, "Update of service experience of HV underground and submarine cable systems" (2020) — not "Guidelines for Submarine Cable Reliability, 2022," which appears to be a constructed title. Source line replaced with verified sources, and every hard number re-anchored or trimmed: the "75–80% of insurance claims" figure corrected to GCube's verified 2010–2020 findings (30% of claims incurred, >50% of claims spend) plus IMCA's report of insurers citing 50% to 70–80%; vessel day rate changed from unsourced "$150,000–300,000/day" to ~£150,000/day (BVG Associates floating wind guide); repair duration changed from "50–150 days" to Strathclyde-via-IMCA averages (40 days inter-array, 60 days export downtime); the $10–30M export cable repair cost kept, now anchored to the same Strathclyde/IMCA source; failure rate corrected from "~0.1 per 100 km/yr" to ~0.003 failures/km/yr (Strathclyde via IMCA); capacity trajectory updated from "~75 GW (2024) to 300+ GW by 2030" to GWEC-verified 83 GW end-2024 and ~350 GW added 2025–2034, with cable-km figures from TGS 4C. Removed as unsourceable: the ±500m TDR accuracy figure (and the ±500m→±50m unlock framing), the "20–40% probability of a major failure over a 30-year life" estimate, the "300% insurance premium increase since 2020" claim, the "50–200 km" cable-length and "buried 1–3 m" specifics, and the WindEurope citation (not needed for any remaining claim). All cited URLs fetched and confirmed 2026-08-20.
Reconciliation 2026-09-29: Fact check of the whole write-up against its sources. Five of the six cited pages were fetched on 2026-09-29 from the addresses in the Source line and read in full: the IMCA commentary, the offshoreWIND.biz article on the GCube report, the GWEC news release, the TGS press release, and section I.2.3 of the floating wind guide. The sixth source, CIGRE Technical Brochure 815, is sold to non-members for 120 euros, so only its catalogue page was read. (1) The text said the average downtimes of 40 and 60 days were "including vessel mobilization and weather windows." The IMCA commentary gives the two averages and does not say what they include, so the phrase was removed. (2) The text gave the failure rate of 0.003 failures per km per year with no region attached. IMCA gives that rate for European wind farms, and the sentence now says so. Confirmed against the fetched pages and left as written: subsea cables at 30% of claims incurred and over 50% of claims spend for 2010–2020, and 44% of cable claims spend traced to contractor error in transit and cable laying (GCube, as reported by offshoreWIND.biz); about 50% of claims, and 70% to 80% for other insurers, put down to subsea cable damage (IMCA); the 46% estimate from ORE Catapult (IMCA); export cable damage costing $10 million to $30 million (IMCA); a typical cable-laying vessel day rate of about £150,000, and BVG Associates as the guide's author (the guide's page and its About page); 83 GW installed at the end of 2024, 350 GW more over 2025–2034 and 441 GW in total by the end of 2034 (GWEC); installed cable growing six-fold from 9,000 km in 2015 to 55,500 km in 2025, with 117,640 km more forecast for 2026–2040 (TGS); and the titles, dates and publishers in the Source line, including the number, title, year and working group of the CIGRE brochure. One thing to know about the downtime and cost figures: IMCA gives them in the same paragraph as the University of Strathclyde study and does not say separately where they come from. The 2019 Strathclyde paper that matches the failure rate (Warnock, McMillan, Pilgrim and Shenton, "Failure Rates of Offshore Wind Transmission Systems," Energies 2019, 12, 2682) was fetched from the University of Strathclyde repository on 2026-09-29 and searched. It gives a mean rate of 0.00299 failures per km per year for European AC wind farm connections, covers transmission cables and not array cables, and does not give the downtime or cost figures, so those figures may come from somewhere else. The attribution in the text is left as IMCA presents it. NOT checked, because they are not in the five pages read and might rest on the CIGRE brochure, which could not be read: that a fault means shutting down the circuit and losing tens to hundreds of MW; that the cable must be dug out of the seabed for repair; that fault location from the cable ends is coarse; that the failure rate is low by onshore standards; that better fault detection would lower the cost of offshore wind energy; everything in "What's Been Tried and Why It Hasn't Worked"; the statement that some accelerated aging test data is published in CIGRE reports; and the statement in the working notes above that cable failure rates are not decreasing. These are left as written. Genes untouched.