The day after a major winter storm swept across the US, hundreds of thousands of homes and businesses remained without el
ctricity and the prospect of more outages loomed amid frigid temperatures. On
e early estimate suggested the storm could result in $24 billion in total economic losses.
But a new analysis by RMI, a clean energy nonprofit, suggests that the current methods for counting the damages from po
wer outages mean we are likely undercounting its impact — and those of similar storms — by a significant margin.
The report finds that current insurance metrics focus too narrowly on physical property damage while ignoring the “n
on-linear compounding losses” that occur when the grid stays down, such as food and medicine spoilage, as well as transportati
on disruptions that can extend well beyond the outage area. Traditional estimating tools like the Value of Lost Load (VoLL) fail to capture
the reality of a multi-hour blackout beyond a given time window or narrow geography, the authors say.
“The methodologies that go into power outage data in particular are often based on survey data, taken after the storms at a specific
point in time, and often, as a result, underestimate the kind of systems-wide impacts that occurred,” said co-author Elizabeth Harnett,
a research and impact expert at RMI’s Center for Climate-Aligned Finance.
See more beautiful photo albums Here >>>
This gap may be massive: In the aftermath of Hurricanes Sandy and Harvey, one study found that business inter
ruption losses were 800% to 900% higher than actual property damages. Even if business interruptions added just 30% to 50% to dire
ct totals, the RMI authors noted, it would imply at least an additional $35 billion per year that are not captured in US disaster-loss calculations.
The RMI authors suggest two other tools that are based on VoLL but with refinements. One, developed by the Lawrence Berke
ley National Laboratory (LBNL), looks at outage costs by customer class — such as residential, commercial and industrial — and i
s well suited for estimating localized and shorter-term events. The other, developed by LBNL and the University of Southern California, combines behavioral survey data with a computer m
odel to assess economy-wide “disequilibrium” effects such as evacuation, supply-chain disruption, and recovery costs.
The authors also point to a need to develop new, forward-looking methods less reliant on historical data and survey responses.