In a single day, Hurricane Ian became one of the most expensive storms in American history, estimated to have caused over $67 billion in insured damages alone (and extensive uninsured damages). Even a month after the storm, communities across Florida and beyond are only beginning to rebuild and reopen businesses. As enterprise leaders adapt to a rapidly changing world and the dollar impact of climate volatility, they need to prepare for extreme weather events and climate change by predicting losses before they occur, so they can be mitigated, or at least managed.
One Concern makes this future-proofing possible with advanced resilience analytics, designed to predict the financial impact of business interruption from disasters on property locations and their dependencies. We tested our digital twin technology and downtime insights, against the storm’s impact across 8 Florida counties hit hardest by the hurricane. Our simulated predictions were accurate within 0.6% median error (-3% to +4% range) during the storm, calculating the peak number of customers without power and for how long within hours.
One Concern DNA helps you understand not just direct property damage, but predicted property downtime, emanating from the failure of critical dependencies. Learn about our predictions and how they held up against the reality of the storm below.
Fig.1 Peak number of customers losing power comparison between the ground truth & 1C simulation
Hurricane Ian caused widespread power damage from high wind speeds across the state. Our power model predicted power downtime at building-level, and aggregated to county/regional level for validation purposes. As illustrated above (Fig. 1), our model accurately estimated the peak number of customers in each county without power during the storm, with Lee and Collier Counties the most and least affected, respectively.
Hurricane Ian: power loss simulation vs ground truth
Our model is able to estimate the peak number of customers without power, CAIDI and SAIDI (within a range) for the storm. Ground truth and simulation are both calculated using the recovery curves for this specific major event.
*CAIDI: Customer Average Interruption Duration Index is the average total downtime (reported in days) per customer-interruption.
*SAIDI: System Average Interruption Duration Index is the average total downtime (reported in days) per customer during this event.
Hurricane Ian: power recovery simulation vs ground truth
Our model also closely predicted the recovery curve for the 8 most wind-impacted counties from the storm. The ground truth aligns well with the 50%-95% quantile of our prediction in the early days after the storm made landfall and then tightly matches the 5%-50% quantiles during later stages.
Mapping Outages by County
Our resilience analytics also enabled us to predict the storm’s impact at building level. We aggregate building-level results to county-level to accommodate the resolution of the ground-truth data.
One Concern prediction validated by ground truth data
As above, One Concern correctly estimated SAIDI, CAIDI, and peak number of customers without power across nearly every metric. For the 24 metrics predicted across 8 counties, our model was missing the prediction range for only four, in 3 counties. Those 4 were within a margin of error of 2000 customers (peak) and 1 day (SAIDI/CAIDI). Through parameter training, model calibration from historical events and data, our technology can estimate the impacts of even an extreme storm like Hurricane Ian, enabling enterprises to understand their long-term dependency risks via our One Concern DNA product.
Extreme weather and natural disasters will continue to wreak havoc on communities and businesses across the country. Through comprehensive downtime analysis – leaders in finance, real estate, and insurance can be prepared before extreme weather events, armed with better insights and stronger data. Measuring the full scope of physical climate risk, including dependency and infrastructure risk, helps enterprise leaders select and mitigate risks more effectively, in increasing the resilience of our economy as a whole.
About One Concern
One Concern, a climate resilience technology company, enables organizations to focus on adaptation and resilience strategies by using newly developed resilience analytics for supporting risk selection, mitigation, pricing, scenario analysis and risk management. Applying machine learning and state-of-the-art resilience modeling, One Concern helps organizations better understand and prepare for physical climate risks with the mission of making disasters less disastrous. A 2019 Technology Pioneer, One Concern is part of the World Economic Forum’s Global Innovators community. oneconcern.com