Risk to life calculations

The risk to life calculation uses the methods defined in the following key references (it is recommended you refer to these references to understand the method):

1. DEFRA flood and coastal defence appraisal guidance - social appraisal - supplementary note to operating authorities - Assessing and valuing the risk to life from flooding for use in appraisal of risk management measures, May 2008. (Referenced below as: DEFRA 2008)

2. DEFRA research contract: reservoir safety advice - supplement no. 1 to interim guide to quantitative risk assessment for UK reservoirs, May 2006 (Referenced below as: DEFRA 2006)

Input Data (ASCII grids)

Flood hazard rating (HR)

Flood hazard rating (HR) is calculated using the following equation (from DEFRA 2008):

Where:
d = depth of flooding (m)
v = velocity of floodwaters (m/s)
DF = debris factor
The debris factor varies as follows:
DF = 0 if depth ≤ 0.25m
DF = 1 if depth > 0.25m or if velocity > 2m/s

These data can be output directly from the 2D Solver (select additional output option on Domains > Outputs tab of 2D model interface).

Damage Calculator requires flood hazard data to be defined in ASCII grid format (.asc). Multiple options are available to generate these data:

• Use the 2D Flood Map tool in Flood Modeller to extract a selected timestep from the 2D hazard parameter file (2dm/dat format) as a new ASCII grid.

• For larger file sizes you may need to utilise the “2dm_to_ascii.exe” program. This is located in the “C:\Program Files\Flood Modeller\gui” folder and is run from the Command Prompt window)

• Use one of the above two options to extract velocity and depth datasets as separate ASCII grids then use the Grid Calculator tool (in the Grid Tools section of the Flood Modeller Toolbox) to calculate a hazard grid.

• Use the 2D flood map tool option to combine velocity and depth data to calculate a flood hazard output. Note that this tool will overestimate the hazard values for the maximum timestep (t=9999) as the maximum depth and maximum velocity values will likely not occur at the same time during the simulation.

Time of inundation

Time of inundation for each cell in a 2D model is the time (in hours) when the cell first becomes wet (this is used to determine the Speed of Onset score). This parameter is calculated by the 2D Solver and written to the 9999.1 timestep of the velocity parameter output. However, Damage Calculator requires these data to be defined in ASCII grid format (.asc). Thus, you will need to use the 2D Flood Map tool in Flood Modeller to extract this timestep from the 2D results file as a new ASCII grid. For larger file sizes you may need to utilise the “2dm_to_ascii.exe” program. This is located in the “C:\Program Files\Flood Modeller\gui” folder and is run from the Command Prompt window).

Input Data (Property Attributes)

Property data

Risk to life calculations are undertaken based on property details. For each property the following data are required:

• Easting and northing - location information (i.e. X coordinate and Y coordinate of the property)

• MCM Code - property type code referenced in your specified depth damage data folder (in file; “MCMcode.csv”)

• Nature of Area - Multi-storey apartments; typical residential/commercial/industrial properties; bungalows, mobile homes, campsites, schools, etc. (score)

• Flood Warning Score - includes % of at risk properties covered by the flood warning system; % of warnings meeting the two-hour target; and % of people taking effective action (score).

• Floor Area - the area of the building 'foot print' in m². If an area value is missing then the tool will look in the data file “DefaultFloorArea.csv” in your specified depth damage folder. The version of this file supplied in the example depth damage folder (in the Flood Modeller example data zip file) contains default floor areas for most UK property types (i.e. all UK MCM codes are referenced). These data have been obtained from documentation available from MCM online (www.mcm-online.co.uk). In the event of a property having no floor area specified and its type having no default area defined in “DefaultFloorArea.csv” the property will not be included in the risk to life analysis. A comment will be added to the analysis log file that states “Warning: Floor area not found ” plus the associated property MCM code (you could then use this information to edit the “DefaultFloorArea.csv” and re-run).

• % infirm - the %residents suffering from long-term illness.

• % > 75 years - the %residents older than 75 years.

Calculations (using steps from DEFRA 2008)

The 'speed of onset' score is calculated in Damage Calculator using the following algorithm:

Nature of Area and Flood Warning Score is provided by user as part of the property dataset.

DEFRA 2008 calculates X, the % of people exposed to risk, as the product of HR and AV and caps this to a maximum of 100%. The same calculation in Damage Calculator is capped to 100%.

1. Calculate flood hazard: HR. The 2D Solver can directly calculate these values.

2. Calculate area vulnerability (AV): DEFRA 2008 defines this as the sum of scores for: flood warning, speed of onset and nature of area. In general, each score has a minimum of 1 and a maximum of 3 and hence the AV is in the range 3 to 9. The Damage Calculator calculates the AV as the sum of these scores.

   • if 'time of onset' ≤ 1 hour then score is 3

   • if 'time of onset' ≥ 3 hours then score is 1

   • if 'time of onset' is between 1 and 3 hours then score is 2

3. Calculate those exposed to the flood : The first part of this calculation is to estimate the number of people likely to be located in the flooded area.

   The method uses the DEFRA 2006 method which was specifically derived from estimating the average number of people likely to be present in inundation areas following a reservoir dam break

   For residential properties (MCM codes starting with a “1” or residential property sector average = 0), a national average household size of 2.3 is used (from DEFRA 2006) together with an assumed occupancy factor of 50%. Thus, each residential property is given a likely population present value, N(Z), of 1.15.

   For non-residential properties (MCMcodes other than 1), the occupancy estimates are taken from DEFRA 2006. For farm buildings (MCM code 810) the occupancy value is 1 per building with an occupancy factor of 30%. For other non-residential properties, the values are a function of floor area as shown in the table below. Thus, if the floor area for a retail property (MCM code 21) is 100m², then the estimated occupancy is 3.33, which together with an occupancy factor of 30%, gives an average number of people within the building of 1.

   The occupancy data are held in a file called “Occupancy.csv” which must exist in your specified depth damage data folder. The version of this file supplied in the example depth damage folder (in the Flood Modeller example data zip file) contains the data shown in the table below. If the MCM Code for a property does not exist in the occupancy lookup file (or if the MCM Code field for the property is blank) then the associated property will be excluded from the analysis. A comment will be added to the analysis log file that states “Warning: Occupancy not matched ” plus the associated property MCM code (you could then use this information to edit the “Occupancy.csv” and re-run).

   strMCMCode	blnPerBuilding	dblOccupancyRate	dblOccupancyFactor
   1	TRUE	2.3	50
   2	FALSE	40	25
   21	FALSE	30	30
   214	FALSE	160	21
   22	FALSE	160	21
   23	FALSE	40	21
   234	FALSE	10	15
   235	FALSE	8	10
   3	FALSE	40	21
   4	FALSE	40	25
   410	FALSE	200	30
   5	FALSE	40	25
   511	FALSE	100	50
   6	FALSE	40	25
   610	FALSE	7	20
   630	FALSE	5	5
   8	FALSE	60	21
   810	TRUE	1	30
   9	FALSE	40	25

   Data fields in this file represent the following:
   • strMCMCode – building type defined by numeric code, i.e. MCM code (as referenced in the MCMCode.csv data file).
   • blnPerBuilding – variable must be set to true or false. If true, then the subsequent occupancy rate represents the expected number of people in this building type. If false, then the associated occupancy rate represents a floor area per person (in same units as property floor area, e.g. m2). In this case, Damage Calculator will combine this value with each property floor area to determine the number of people in a building.
   • dblOccupancyRate – the occupancy rate value either represents number of people per building or floor area per person depending on the blnPerBuilding setting. The units of the latter variable will be the same units as your property floor areas, e.g. m2 or ft2.
   • dblOccupancyFactor – percentage factor that is combined with the occupancy rate to determine the number of people per property for the purposes of the risk to life calculation. This is detailed in DEFRA 2006 (see explanation above).

4. The number of people exposed, N(ZE) : Defined as the product of X and the average number of people likely to be present at the receptor (Pop), ie:

5. Calculate people vulnerability (PV) : DEFRA 2008 defines PV as the sum of the % > 75yrs and the % infirm.

6. Calculate number of possible fatalities : DEFRA 2008 defines this as:
   N(ZE) x {2 x PV x 2 x HR}
   where the value in brackets is expressed as a percent according to DEFRA 2008.
   Within the Damage Calculator software this is implemented as:
   LLOL (likely loss of life) = (min[(HR x AV), 100] x Pop) x 2 x (%infirm + %>75yrs) x 2 x HR/1,000,000
   The division by 1,000,000 is made up of 1/100 for converting the HRxAV percent value to a decimal, 1/100 for converting the (%infirm + %>75yrs) from a percent to a decimal, and a further 1/100 for converting the {2 x PV x 2 x HR] value from a percent to a decimal.

Example of calculations

As an example of the calculations, the data contained in Annex 1 of the DEFRA 2008 Note have been run through the Damage Calculator. The table below contains the values from the DEFRA Note compared with those calculated using the Damage Calculator. The values from the Damage Calculator are consistent with the values from DEFRA 2008 (which are rounded to show no decimal places).