Benchmarks
Where incidents cluster, how severe they are, what fails and how a portfolio mix compares with the register. Computed live from every published entry and, for subscribers, from the verified records behind them.
Loading...
Where incidents cluster
Published entries by technology and incident category. Darker means more entries.
Trend by technology
Entries by technology and year of incident, last ten years. Recent years fill in as the register catches up, so treat the newest column as a floor.
What the register holds, by technology
Entries, share of the register, how often a site capacity is recorded, and the typical size of the sites where one is. Read the coverage column first: capacity is quoted in public reports far more often for battery storage and offshore wind than for substations, so a technology with low coverage is thinly described here rather than safer.
Why there is no incidents per gigawatt figure here. A frequency needs the installed fleet as its denominator. The register knows the capacity of the sites where an incident happened, not the capacity of every site at risk. Dividing one by the other returns the reciprocal of average site size: it moves when sites get bigger and stays still when incidents get more common. That figure was published until 8 September 2026 and has been withdrawn rather than caveated, because it reads like a rate and is not one. A real rate needs an independent installed-capacity series by technology, country and year, which is the next thing being built. Until then the register publishes counts and says so.
Severity mix
Share of entries by severity within each technology.
Verified records: who, where, how old
From the verified records. These grow daily as the register builds and confirms records. The names, counts and the country map are public. What failed, the age, downtime and loss figures are for subscribers.
Technology by country
Manufacturers named in the records
A count of incidents where the source names the manufacturer, with what failed. It is not a failure rate: the register does not know how much of each manufacturer's fleet is installed, so a larger supplier will appear more often for that reason alone. Use it to find the incidents behind a name, not to rank suppliers.
Integrators named in the records
The same counting rule applies. Integrators are named far less often in public reports than manufacturers.
Age at incident
Downtime and loss
Failed elements
Portfolio comparison
Enter the capacity you insure, finance or operate by technology. This compares the shape of your book with the shape of the register: which technologies carry more of your capacity than they carry of recorded incidents, and which carry less. It answers where to look first, not how often anything will happen. The register cannot give you a frequency, because it does not know the installed fleet behind each technology, and a model that quietly assumed one would be worse than no model.
Included with every paid plan.
See plans