National and regional levee systems analysis in the UK and relevance to USA

Transcription

1 National and regional levee systems analysis in the UK and relevance to USA Jonathan Simm Presentation for ASFPM Conference, Grand Rapids MI Overview UK background Overview of method and description of model components (including comparison with FEMA RiskMAP approach) Model outputs, uses and case study applications Page 2 1

2 UK background Page 3 Where does England fit in conceptually Land area USA Michigan England 3,806,000 97,000 50,000 miles 2 miles 2 miles 2 100% 4.3% 1.3% Coastline length USA Michigan England Population Michigan 10 million Population England 53 million i.e. 5 x population in 50% land area! 95,500 miles 3,200 miles. 5,600 miles 100% 3.6% 6% Page 4 2

3 England has a problem with flooding % of land in flood plain % of development in flood plain Page 5 Background UK has a long history of flooding, especially over the last 20 years, with major flood events in / /16 e.g. Storms Desmond December 5 th, 2015 Eva December 24 th, 2015 Frank December 29 th, 2015 Page 6 3

4 Storm Desmond air temperatures Page 7 England south coast Page 8 4

5 Calder after Storm Desmond Page 9 Calder after Storm Desmond Page 10 5

6 Tadcaster Bridge after Storm Desmond Page 11 Carlisle after Storm Desmond Page 12 6

7 Today is Brexit vote day and guess what? Page 13 Background UK has a long history of flooding, especially over the last 20 years, with major flood events in / /16 In 2002, UK government moved from flood defence towards flood risk management: To manage flood risk there is a need to quantify it. Page 14 7

8 National Flood Risk Analysis (NaFRA) National model run NaFRA 2004 NaFRA 2006 NaFRA 2008 SoN NaFRA 2005 NaFRA 2007 NaFRA 2011 Method/Software development Page 15 Flood risk analysis underpins Environment Agency s public flood maps Requirement under EU Floods Directive to produce flood risk maps and make them available to public For this purpose, information is expressed in high, medium and low flood risk bands Page 16 8

9 Analysis supports UK planning and policy Environment Agency s Long Term Investment Strategy informing 2 national comprehensive spending reviews Government contingency planning and advice to Association of British Insurers Futures work: Foresight Future Flooding analysis First UK Climate Change Risk Assessment Ongoing work for Adaptation Committee on Climate Change Municipality planning depts Page 17 Overview Overview of method and description of model components (including comparison with FEMA RiskMAP approach) Page 18 9

10 Water level (m) UK risk analysis approach (addresses both probability components) NFIP omits Flood risk = Probability Load Probability (Breach and overtopping) Consequence $ or Page 19 Sources: Extreme value analysis NFIP focus: return periods of 100 yr (plus 10, 50 and 500yr) UK approach: 40 return periods between 1 and 1000yrs Return period (years) Data requirements: River flows / levels; coastal water levels and wave conditions Note: All data and models are owned by the Environment Agency (EA) and publicly available. However, EA analysis outcomes are rarely appealed Insurance private impacts land-use. Page 20 10

11 Probability of failure Sources: 2015 upgraded estimate of coastal forcing 1) Multivariate extreme value analysis 2) SWAN wave model emulator 3) Surfzone model 4) BAYONET overtopping model Page 21 Pathways: Assessment of levee failure probability Defence Fragility Curve: Rasp type 8, Condition grade 2 Probability Not required under FEMA Structural-based Inundation Procedure (SIP) Lower Bound Upper Bound UK approach uses fragility curves to assess X(m) Freeboard Water level load Breach simulation Page 22 11

12 Developing fragility curves: the RELIABLE tool Structure-specific parameters, probability distribution functions and ranges Numerical Integration Structure-specific Fault tree Limit state equation Failure mode 1 Limit state equation Failure mode 2 Limit state equation Failure mode 3 Structure-specific fragility curve Limit state equation Failure mode..n Tool available at Page 23 Site specific fragility curve Page 24 12

13 Reliability of levees generic fragility curves Over 60 generic HLM+ Fragility levee types curve Generic - Condition curves Grades provide 1 an to economic 5 (Central approach estimates) to implementing BW12/NAS P (breaching feeboard) Condition grade1 Condition grade2 Condition grade 3 Condition grade4 Condition grade Overflow head (Water level - crest level) Page 25 Breach simulation defines [typical] breach size Erodible top layer Resistant top layer Breach simulation defines breach dimensions: i.e. levee doesn t vanish if failure occurs Page 26 13

14 All levees but with variable influence LAMP is applied only to non-accredited levees, whereas UK process is applied to all levees using a whole system approach NFIP process involves a binary (in/out) influence of levees.uk process is graduated Page 27 Pathways: Flood inundation analysis Current FEMA process: independent reach-specific inundation UK process: integrated across all reaches Analysis includes: (1) 40 return periods between 1 and 1000 years return period. (2) For defended flood plains for each return period, a Monte Carlo analysis is required with multiple realisations of system state (levee segments failed or not failed) (3) Analysis also covers each return period for undefended floodplains Page 28 14

15 Pathways: Flood inundation Data requirements: Digital terrain model Range of hydrodynamic models: Volume based Rapid Flood Spreading Model (RFSM) Diffusion wave (Dynamic RFSM) Hybrid models (local acceleration with sub-grid topography, RFSM EDA) Full Shallow Water Equations e.g. InfoWorks RS & CS, TELEMAC, Tuflow, etc For repeated analyses in Monte Carlo simulations, models must run very fast Page 29 Receptors: Consequences Both UK and FEMA (HAZUS) approaches are similar. (More individual calculations in UK case) Data requirements: Property types and locations; Depth/Damage ($) relationships Depth Metres Depth Damage Curve Damage /m2 High Susceptibility Band Low Susceptibility Band Indicative Susceptibility Page 30 15

16 Model outputs, and case study applications Page 31 Likelihood of tidal flooding for the present day June

17 Present value damage if we do nothing 5B 14 June 2016 Contribution to damage from individual levees Breach Overtopping 14 June

18 Risk attribution to levee segments supports risk mitigation by structural interventions Page 35 Managing flood risk Levees only one solution: All changes and interventions can be expressed in probabilityconsequence space (Adapted from Sayers et al, 2003) Page 36 18

19 Managing changing flood risk system (Evans, et al, 2008) Page 37 Likelihood of tidal flooding for the present day June

20 Likelihood of tidal flooding if we do nothing June 2016 Present value damage if we do nothing 5B 14 June

21 Present value damage if we do minimum maintenance 2.5B 14 June 2016 Summary and conclusions UK and US face similar flooding problems and analysis methods use the same underlying principles The ability to communicate risk and potential solutions to decision-makers is critical Full system flood risk analysis enables spatial variation in risk to be evaluated, including residual risk attributable to levees A wide range of flood risk management actions/decisions can be supported by the UK style of risk modelling presented Page 42 21

22 National and regional levee systems analysis in the UK and relevance to USA Jonathan Simm Presentation for ASFPM Conference, Grand Rapids MI 22