When the power goes out for an hour at one of Sweden’s electricity-intensive factories, the true bill is far higher than standard economic statistics suggest — and some businesses never fully recover the lost production. That’s the finding of a new study from CERE researchers Lars Persson, Tommy Lundgren and Runar Brännlund, recently published in Utilities Policy.
The team surveyed 359 of Sweden’s largest electricity-consuming industrial plants — across the forest, chemical, mining and basic iron and steel sectors — and matched the answers against official production statistics, combining survey-based and production-based methods in a single sample. That dual approach turns out to matter for policy: it reveals that the cost of an outage varies substantially both by sector and by which method is used to estimate it, with the cost companies say a blackout actually causes them running, on average, about 60 times higher than the loss implied by conventional value-added figures.
For a one-hour outage, average lost value added works out to roughly 68,000 SEK — but companies themselves report an average direct cost closer to 977,000 SEK, equivalent to a jump from about 26 SEK to 1,500 SEK per kilowatt-hour.
“Official statistics only capture what is lost during the outage itself,” the researchers explain. “They miss the restart costs, spoiled materials and knock-on disruption a power cut actually triggers — and our data show those consequences can be severe and long-lived.”
Short outages cause disproportionately large production and revenue losses, the study finds: fixed shutdown and restart costs mean a brief interruption can cost nearly as much as one lasting hours, with stated costs ranging from around 6,900 SEK/kWh for a one-minute cut to about 100 SEK/kWh for a twelve-hour one. Recovery is far from guaranteed either — roughly 70% of plants say recovering a one-hour outage within 24 hours is difficult or impossible, and about 15% still hadn’t recovered a year later.
Given this variation, the authors call for a more nuanced use of cost estimates in regulatory decision-making rather than simplified, flat-rate assumptions — with direct implications, they argue, for how reliability standards are set and how compensation schemes for affected industries are designed.
Persson, L., Lundgren, T., & Brännlund, R. (2027). The value of lost load in electricity intensive industry – combining and comparing stated and production value approaches. Utilities Policy, 104, 102318. https://doi.org/10.1016/j.jup.2026.102318