Calculation Process
The One Concern Domino AI APIs provide business interruption information for the closest building location to the requested specified latitude and longitude.
Input Fields
For either the Business Interruption Return Period request or Business Interruption Average Annual Request, a user specifies the following input details:
| Input | Description | Data Type | Units | Valid Values and Definitions |
|---|---|---|---|---|
| latitude | Latitude of place of interest | Number | N/A | Any number -90 to 90 |
| longitude | Longitude of place of interest | Number | N/A | Any number -180 to 180 |
| max_distance_m | The maximum distance tolerance from the provided (lat,long) pair to use | Number | Meters | Any number between 1.0 and 500.0 |
| return_period_yrs* | Return period of interest | Number | Years | 50 : 50-year return period 100 : 100-year return period 250 : 250-year return period 500 : 500-year return period 1000 : 1000-year return period |
| peril | Cause of downtime | String | N/A | flood : inland and coastal flooding seismic : earthquake ground-shaking wind : tropical cyclone wind integrated : all perils combined |
| climate_change | Climate change scenario | String | N/A | ccbaseline : baseline climate view cc2050_45 : RCP 4.5 or equivalent SSP view for 2050 projection year |
| interruption_type | Component(s) of infrastructure essential to recovery | String | N/A | ingress_egress : ports, airports, roads, bridges community : residential community only utility : power distribution network only repair_time: building structural downtime integrated : all of the above |
*This only applies to the Business Interruption Return Period API request. Note return periods are commonly used to estimate the likelihood and frequency of extreme events. For example, a 100-year return period flood is an event that is probabilistically equivalent to a 1% chance of occurring in any given year, whereas a 250-year return period earthquake is an event that is probabilistically equivalent to a 0.4% chance of occurring in any given year.
Output Fields
The following data is output for each
| Output | Description | Data Type | Units | Valid Values and Definitions |
|---|---|---|---|---|
| id | One Concern unique building ID | String | N/A | UUID |
| downtime_avg_days | Average (mean) downtime | Number | Days | Any number >= 0 |
| downtime_stdev_days | Standard deviation of downtime | Number | Days | Any number >= 0 |
| score | One Concern business interruption risk score | String | N/A | low : Low BI risk med : Medium BI risk high: High BI risk |
Business Interruption Risk Score Methodology
For a given building, One Concern generates downtime statistics following these steps below:
1. Calculate the recovery curves for the building and all individual components of each building related lifeline/dependency for all return periods and each peril.
2. Aggregate the recovery curves to create a single recovery curve for each dependency, return period, and peril.
3. Calculate the downtime statistic values as the conditional mean and standard deviation downtime from the recovery curve distribution, based on a threshold of 0 days for each dependency, return period, and peril.
4. The mean and standard deviation of downtime are reported for selected dependencies from the interruption type specified in the API request. For interruption types other than integrated, downtime statistics are determined as the maximum mean downtime of dependencies included in the scope of the selected interruption type. For interruption type integrated, downtime statistics are calculated as the weighted sum of each dependency’s downtime statistic. This generates the Business Interruption Return Period output. Risk bins are then applied to the downtime_avg_days to determine the score.
5. For the average annual downtime calculation, recovery curves at all return periods are used to generate the downtime statistic based on a threshold of 0 days and annualized by converting all return period curves to annual rate of return curves. This annualized value mean and standard deviation are the Average Annual Business Interruption output. Risk bins below are applied to the (downtime_avg_days + (2 * downtime_stdev_days)) to determine the score*
The BI Risk Score is divided into the following categories according to the length of downtime:
- High risk: downtime >= 7 days
- Medium risk: downtime < 7 days and downtime >= 2 days
- Low risk: downtime < 2 days
*One Concern has performed validation of the Average Annual Downtime statistic. As such, the One Concern default view of average annual downtime adds two standard deviations to the mean downtime and then applying the risk bins. Customers may consume the downtime values and generate their own view of risk as they see fit.