© Crown copyright 2019
This publication is licensed under the terms of the Open Government Licence v3.0 except where otherwise stated. To view this licence, visit nationalarchives.gov.uk/doc/open-government-licence/version/3 or write to the Information Policy Team, The National Archives, Kew, London TW9 4DU, or email: firstname.lastname@example.org.
Where we have identified any third party copyright information you will need to obtain permission from the copyright holders concerned.
This publication is available at https://www.gov.uk/government/publications/flood-and-coastal-risk-management-in-england-long-term-investment/long-term-investment-scenarios-ltis-2019
LTIS 2019 brings together our understanding of long-term investment scenarios for flood and coastal erosion risk management (FCERM). It’s based on the LTIS 2014 report and new analysis carried out in 2018.
LTIS 2019 has expanded the range of climate change scenarios. It now includes a plausible extreme scenario (high++) in addition to the high, medium and no climate change scenarios of LTIS 2014.
The Met Office published UK Climate Projections 2018 (UKCP18) in November 2018. LTIS covers the full range of climate futures from no climate change through to high++. The UKCP18 projections only affect the middle 2 scenarios, medium and high. UKCP18 does not affect the high++ scenario. As a result, the range of climate futures which LTIS describes is unchanged.
The differences between the LTIS medium and high scenarios and the new UKCP18 projections are not large. They are not likely to:
- change a significant number of investment decisions within the analysis
- make a material difference to the LTIS 2019 outcomes
1. Main findings
LTIS sets out the challenges of managing flood and coastal erosion risk in the face of:
- asset deterioration
- climate change
- a growing population
This report describes:
- future risks
- future opportunities
- the implications for long-term investment levels
In the high climate change scenario and with current effective planning outcomes it’s possible to prevent risk from increasing. This is if we invest in conventional FCERM activities wherever the benefits are greater than the costs. This is because with high climate change, investment in conventional FCERM activities reduces risk by 4%. However, this would be cancelled out by the increased damages from a continuation of the current outcomes of planning policy and its implementation.
Under a plausible extreme climate change scenario, there are more places where new investment is not cost effective. Using strict economic criteria alone and not considering the social impacts, assets would be left to degrade. This would increase risk.
In combination with conventional FCERM activities, property flood resilience (PFR) and natural flood management (NFM) help to manage future risks. Large scale PFR could reduce risk by about 16%. NFM has a complementary and beneficial role alongside investment in conventional FCERM assets. But it alone cannot reverse national increases in risk due to climate change. With higher climate change, raising defences becomes technically and socially challenging. This may force society to think differently about the balance of FCERM activities. LTIS describes the potential of these activities at a national level, but the best combination of activities will vary from place to place.
LTIS 2019 has new scenarios which quantify the significant benefits of investing in very high levels of protection. We recognise that social and technical limitations make these difficult or impossible to achieve in many places.
Evidence from the new scenarios shows the overall economic optimum level of investment is higher than previously estimated. The new optimum is a long-term annual average of over £1 billion. This compares to £860 million for the LTIS 2014 baseline. The new estimate is higher because it uses the range across medium and high climate change scenarios. It also includes a better understanding of infrastructure impacts and the investment needed to manage surface water.
LTIS 2019 has had an independent peer review by:
- Jaap Flikweert of Royal HaskoningDHV
- Paul Sayers of Sayers and Partners
- Edmund Penning-Rowsell of Flood Hazard Research Centre
- Richard Dawson of Newcastle University
2. The scope of the long-term investment scenarios
LTIS 2019 is based on the LTIS 2014 report and new analysis carried out in 2018. Together they describe future investment choices for FCERM and include risk from:
- the sea
- surface water
- coastal erosion
Some of the new analysis is limited to risk from rivers and the sea. Where this is the case we’ve made it clear in the appropriate places. The LTIS 2019 Technical Summary contains the technical detail of the new analysis. Contact the Environment Agency for a copy of the LTIS 2019 Technical Summary.
The new analysis includes:
- a plausible extreme climate change scenario
- more realistic scenarios for development on the flood plain
- a better understanding of the impacts of flooding, particularly on infrastructure
- the effectiveness of different ways to manage flood risk
- scenarios with specific levels of protection
In this report:
- sections 3, 4 and 5 outline the principles behind LTIS
- sections 6 to 8 describe future risks
- sections 9 and 10 describe opportunities to manage those risks
- section 11 concludes what this means for investment in FCERM
- section 12 considers further work
3. The economic optimum level of investment
LTIS sets out the total national level of investment if we invest in all the places where the benefits are greater than the costs. This is called the economic optimum level of investment for FCERM. All the LTIS analysis uses current government guidance. LTIS identifies an approach to investing in FCERM activities over the long-term to achieve the greatest reduction in flood and coastal damage per pound invested. In some places this means that the ratio of benefits to costs is only slightly higher than 1. Across the whole baseline long-term investment profile the overall ratio is about 5:1. LTIS does not suggest who should pay. It includes investment from central and local government alongside partnership funding contributions.
The technical definition of the optimum level of investment is based on net present value (NPV). NPV shows how much the long-term benefits are greater than the long-term costs (in today’s prices). This is calculated over 100 years. The optimum level of investment has the highest NPV. It’s enough to fund all work where benefits exceed costs. It’s the tipping point between investing enough to achieve all the cost-effective benefits but not too much that we get diminishing returns.
The Environment Agency calculates the economic optimum at ‘system’ level rather than for individual assets or whole catchments. A system is a collection of assets that protects a discrete area of a river catchment or the coast. The economic principles underpinning the analysis are explained in more detail in Appendix B of LTIS 2014.
To describe the optimum level of investment, we use the long-term annual average level of investment. This is calculated over the 50 year period 2014 to 2063, and makes it easier to compare between scenarios. This provides a simple overall level of investment. Choices remain about how investment is distributed over a 50 year period.
4. The baseline scenario
LTIS is a set of national scenarios which describe possible ‘What if …?’ futures, but does not predict which of these futures will happen.The Environment Agency now has a broad range of investment scenarios. Each scenario explores one aspect of future investment. This allows us to explore the drivers of future risk, such as climate change or development on the flood plain. We can also explore the effectiveness of particular combinations of FCERM activities.
To bring the analysis together, we compare all the scenarios to the same baseline. Other scenarios vary from the baseline by changing one variable at a time. We use the LTIS 2014 baseline and 2015 prices.
The baseline scenario has the following set of variables:
- climate change – medium climate change, based on UK Climate Projections 2009 (UKCP09) medium projection
- development on the flood plain – controlled development (no allowance for new development)
- cost – reductions of 10% achieved between 2015 and 2021 and then held at that level
- investment in FCERM activities – conventional activities include building and maintaining assets (defences), modelling, mapping, development advice, engagement and flood incident management
Sections 5, 6, 7 and 9 address the baseline variables. They provide new scenarios which are more realistic and useful for future investment choices. Further details of the baseline and its inputs, models and assumptions are in the LTIS 2014 report.
Table 1: long-term investment need for the baseline scenario by period (based on LTIS 2014)
|Period||Annual average investment need for the baseline scenario|
|First decade: 2014 to 2023||£750 to £800 million|
|Middle decades: 2024 to 2053||£850 to £900 million|
|Last decade: 2054 to 2063||£900 to £950 million|
|Long-term: whole 50 year period 2014 to 2063||£860 million|
Table 1 shows the optimum investment and outcomes for the baseline scenario as reported in LTIS 2014. In summary:
- the annual average cost of funding FCERM activities is expected to increase over 50 years - by £100 million to £200 million
- the optimum level of investment is a long-term annual average of £860 million (range: £790 million to £920 million)
- over 100 years the present value cost (the long-term cost in today’s prices) is £25 billion and the NPV is £101 billion
- the overall benefit to cost ratio is about 5 to 1, so for every £1 invested we avoid about £5 of property damages
- we estimate this level of investment would reduce risk (the property damages we would expect to see in a typical year) by 12% over 50 years
- for comparison, with no investment, overall economic damages would increase by approximately 250% over 50 years
We usually compare other scenarios to the baseline by comparing:
- the optimum level of investment (the long-term annual average level of investment)
- the change in risk (in terms of the property damages we would expect to see in a typical year) over 50 years
5. Risk and cost information
The Environment Agency reviewed the risk and cost information used by LTIS to understand the changes since 2014.
5.1 Risk information
LTIS uses the National Flood Risk Assessment (NaFRA) for the risk of flooding from rivers and the sea. NaFRA was updated in 2018 and provides a new estimate of present day expected annual damages (EAD). LTIS 2014 noted that results were relatively insensitive to this amount of change in EAD.
For coastal erosion and surface water risk, LTIS uses the:
- National Coastal Erosion Risk Mapping
- Risk of Flooding from Surface Water map
These have not changed much since 2014.
We know less about the costs and benefits of managing flood risk from surface water compared to that from rivers and the sea. New evidence on surface water flood risk suggests that this component of the annual average optimum level of investment should increase by about £44 million. This investment would continue to maintain and improve the drainage systems described in LTIS 2014. This is based on several factors including evidence about rainfall intensity under climate change from United Kingdom Water Industry Research (UKWIR).
5.2 Cost information
The Environment Agency reviewed the long-term average construction costs of FCERM walls between 2011 and 2016. This did not show a strong pattern of increase or decrease in wall costs over time. This was despite construction prices rising by 3.2% from August 2015 to January 2018. The cost scenarios in LTIS 2014 show this will have little impact on results.
The baseline scenario assumes a 10% reduction in costs. In practice, we’re achieving 10% efficiencies which have enabled us to absorb inflation in the current programme. It’s useful to know how outcomes would change without the 10% cost reduction. This would result in risk (property damages) reducing by 10%, rather than 12% in the baseline scenario. To achieve a risk reduction of 12% the annual average level of investment would need to increase from £860 million to £900 million.
5.3 Investment levels and outcomes
Analysis of recent data shows that investment in the FCERM investment programme 2015 to 2021 is similar to the optimum level identified in LTIS 2014. We’re also achieving the expected reduction in damages.
6. Climate change scenarios
The Environment Agency has expanded the range of climate change scenarios. We now include a plausible extreme scenario, high++. Existing LTIS 2014 scenarios include:
- medium (the baseline)
- no climate change
We have based all LTIS scenarios on UKCP09 emissions projections. See LTIS 2014 and the LTIS 2019 Technical Summary for details. Contact the Environment Agency for a copy of the LTIS 2019 Technical Summary.
The Met Office published UK Climate Projections 2018 (UKCP18) in November 2018. We need to review guidance for risk management authorities. This will provide new evidence for the LTIS medium and high scenarios. The differences are not large. They are unlikely to make a material difference to the LTIS 2019 outcomes.
UKCP18 is not updating projections for the high++ scenario. With the high++ scenario, LTIS now covers the full range of possible climate futures.
6.1 Medium and high climate change scenarios
Headline 1: medium and high climate change scenarios show the annual average cost of funding FCERM activities increases over 50 years - by £100 million to £260 million.
LTIS 2014 reported that for a high climate change scenario, the optimum level of investment is a long-term annual average of £920 million (range: £850 million to £980 million). This optimum investment would only reduce risk (the property damages we would expect to see in a typical year) by 4%. This is less than the 12% reduction achieved by the optimum level of investment in the medium (baseline) climate change scenario.
Table 2: optimum levels of investment for medium and high climate change scenarios
|Climate change scenario||Optimum level of investment (long-term annual average 2014 to 2063)|
The 25 Year Environment Plan states that current global commitments under the Paris Agreement are insufficient to limit average temperature rise to well below 2°C. LTIS has taken both medium and high climate change scenarios into account in assessing the optimum investment level.
We can consider investment needed across both the medium and high scenarios:
- the cost of funding FCERM activities increases over 50 years - the annual average increases by £100 million to £260 million
- the optimum level of investment across medium and high scenarios is a long-term annual average of £860 million to £920 million
Table 3: changes in property damages after 50 years that can be achieved with optimum investment for each climate change scenario
|Climate change scenario||Peak flow change factors (range for 2060s, relative to 2014)||Sea level rise (range for 2060s, since 2014)||Net present value (NPV, over 100 years)||Change in property damages after 50 years (negative values show a reduction)|
|No change||0%||0 cm||£59 billion||-24%|
|Medium||9 to 16%||27 cm to 34 cm||£101 billion||-12%|
|High||15 to 39%||32 cm to 40 cm||£115 billion||-4%|
|High++||29 to 57%||70 cm||£140 billion||+7%|
Others have looked at FCERM and climate change. The Climate Change Risk Assessment (2017) used 2˚C and 4˚C scenarios with change factors similar to the medium and high LTIS 2014 scenarios. The 2˚C scenario used changes in peak flow for the 2050s of 3% to 18%, and the 4˚C scenario used changes of 21% to 33%.
Headline 2: with high climate change, it’s possible to prevent risk from increasing if we invest in conventional FCERM activities at the optimum level.
The changes in property damages after 50 years also depend on assumptions about development on the flood plain. In the baseline scenario with controlled development (no allowance for new development on the flood plain):
- the medium climate change scenario shows optimum investment in conventional FCERM activities can reduce property damages in the long-term by 12%
- the high climate change scenario shows optimum investment in conventional FCERM activities can reduce property damages in the long-term by 4%
Section 7 considers development on the flood plain. One scenario describes the current outcomes of planning policy and its implementation. If these stay the same, there will be 4% more property damages than under the controlled development scenario. This means that:
- in the medium climate change scenario the reduction in property damages is 8%
- in the high climate change scenario the reduction in property damages is likely to be cancelled out by the 4% more damages from future development in the flood plain
To summarise, with high climate change, there is a 4% reduction in risk due to optimum investment in conventional FCERM activities. This would be cancelled out by the damages from a continuation of current planning outcomes. We conclude that it’s possible to prevent risk from increasing in the long-term in a high climate change scenario.
These climate scenarios assume investment in conventional FCERM only. There are opportunities to reduce risk in the long-term if we consider different ways to manage risk and very high levels of protection. We explore this in sections 9 and 10.
6.2 High++ climate change scenario
LTIS 2019 adds a high++ climate change scenario. In this scenario, damages increase in the long-term despite optimum investment. To reduce risk by 12% (in line with the baseline) we would need to invest in places where the costs are greater than the benefits. This would be contrary to government guidance. This would need an average annual investment level in the order of £1.5 billion to £2.0 billion.
Headline 3: with higher climate change, raising defences becomes technically and socially challenging. It may force society to think differently about the balance of FCERM activities.
Climate change raises questions about the long-term sustainability of FCERM assets. We can often refurbish embankments and walls with an increase in height of around 0.6 metres. Above this height they may need full re-engineering. We expect a 0.6 metres increase in sea level by the:
- 2060s under the high++ climate change scenario
- 2110s under the medium climate change scenario
6.3 Impact of climate change on asset deterioration
Levels of investment under climate change are uncertain. We know that climate change will cause assets to deteriorate faster - particularly on the coast. This will increase costs.
We looked at this issue in more detail in the high++ scenario. In this scenario, long-term asset maintenance and replacement costs can increase by factors of up to 3 and 5. While this has little impact on the overall optimum level of investment, it has a big impact on the type of investment for individual places. For example whether the model chooses to make no new investment or to maintain or improve the levels of protection.
Headline 4: under a plausible extreme climate change scenario, there are more places where new investment is not cost effective. Under strict economic criteria alone and not considering the social impacts, assets would be left to degrade. This would increase risk.
In the high++ climate scenario we included the increased costs due to the impact of climate change on asset deterioration. This results in many more places where the model chooses not to invest, or simply to maintain the height of existing assets. The number of properties in systems with these investment outcomes increases by 50%. In practical terms this means assets would be left to degrade, and risk would increase. The model’s decision to effectively stop investing in these assets is made on strict economic criteria alone. It does not consider the associated legal costs. Similarly, in fewer places the model chooses to improve levels of protection. The places with little or no new investment are likely to be sparsely populated. Places where levels of protection are improved are likely to be densely populated.
6.4 Information for future analysis
Future analysis will benefit from a new national assessment of flood risk. This will enable a better analysis across all sources of flooding. It will also use UKCP18 to identify places which will become at risk due to future climate change, in particular future sea level rise.
7. Development on the flood plain
The government’s policy on planning and flood risk states that:
inappropriate development in areas at risk of flooding should be avoided by directing development away from areas at highest risk (whether existing or future). Where development is necessary in such areas, the development should be made safe for its lifetime without increasing flood risk elsewhere.
7.1 Planning policy and its implementation
Development scenarios are based on 2 variables at a local authority scale:
- population growth
- the proportion of new development on the flood plain
The baseline scenario includes a simple assumption that development on the flood plain is controlled (no allowance for new development). LTIS 2019 adds more realistic development scenarios. It uses Office for National Statistics population projections. One scenario assumes a continuation of the current planning outcomes. These outcomes are the result of both planning policy and its implementation. This is based on data quantifying development on the flood plain from 2008 to 2014.
Another scenario explores how outcomes may change should planning policy or its implementation become weaker. In this scenario, the proportion of new development on the flood plain increases by 45%, with no mitigation measures applied. This represents what could happen if flood risk ceased to be taken into account in planning decisions.
In these development scenarios, the Environment Agency used the same assumptions as in the baseline scenario. This includes medium climate change. This analysis is limited to flood risk from rivers and the sea. The damages reported in this section are damages to properties.
Headline 5: with the number of properties in the flood plain set to almost double over the next 50 years, planning policy and its implementation effectively mitigates flood damages in the long-term
If current planning outcomes continue for the next 50 years the total number of properties at risk would almost double. This is mainly due to development in areas at low risk of flooding. The increase in damages compared to the controlled development scenario would be limited to 4%. This would be achieved by planning policy mitigation measures which we have represented as raised floor levels. This mitigation also contributes to flood recovery which is an extra benefit not quantified here. The analysis represents the fundamentals of planning policy and its implementation. Changes to planning policy came into effect in July 2018. Current planning outcomes are based on data from before 2018, so cannot capture the effects of the 2018 policy changes.
Compared to the baseline, the current planning outcomes scenario increases NPV to about £107 billion, but does not change the optimum level of investment.
Table 4: summary of development on the flood plain scenarios
|Scenario (applied for 50 years)||Change in property damages after 50 years (negative values show a reduction)||Properties in areas with medium and high flood likelihood, without flood risk mitigation||New properties in areas with medium and high flood likelihood , with flood risk mitigation||Total number of properties on the flood plain|
|Controlled development (baseline)||-12%||766,000 existing||-||2.4 million|
|Current planning outcomes||-8%||766,000 existing||358,000 new||4.6 million|
|Weakened outcomes||+38%||766,000 existing and 506,000 new||None||5.0 million|
The implementation of planning policy is currently effective overall. The Committee on Climate Change (CCC) recently reviewed a small sample of planning applications. They identified examples of non-compliance with planning conditions to mitigate flood risk. There’s a risk that if planning policy is not adequately implemented and enforced, non-compliance with planning conditions could increase. This could lead to billions of pounds of extra flood damages in the long-term.
7.2 Creating resilient places
Headline 6: the challenge of creating resilient places, despite long-term flood risk and population growth, is concentrated in a few areas.
We need sustainable housing and infrastructure development to create resilient places. LTIS 2019 uses data at a local authority scale which we combine to give national results. The local data shows complex variations in:
- population growth (lower in the north and west, higher in the south and east)
- the proportions of new and existing properties on the flood plain
It’s clear that these local variations are important to future flood risk. For example, 3% of local authorities contribute over 20% of new development on the flood plain. These areas have the greatest challenges for creating resilient places in the long-term. In these places we expect:
- above average population increases
- there is limited non-flood plain land
- or a combination of both
Many new properties are located in areas with low to very low likelihood of flooding due to the presence of tidal and fluvial defences. In these areas there’s still a risk of flooding because defences and property flood resilience products can fail or be overtopped. More development in these areas will lead to an increase in potential damages from flooding in the longer term. An example of this is development in areas of London protected by the Thames Barrier. Surface water is often an extra risk to properties benefiting from tidal and fluvial defences.
8. The impacts of flooding
LTIS analysis includes damage to properties as well as wider impacts of flooding. This is in line with FCERM options appraisal guidance and practice. LTIS 2014 estimates the wider impacts of flooding using the approach established in the costs of the summer 2007 floods in England report. LTIS 2019 improves our confidence in these wider impacts. It assesses risk to life and infrastructure in greater detail. It also uses evidence from the costs of flooding in the winter 2013 to 2014 and the costs of flooding in the winter 2015 to 2016.
The Environment Agency calculated risk to life at a local scale by using different factors in areas with different levels of risk. For example, catchments where water levels rise rapidly in response to rainfall have a higher risk to life. At a national scale this did not significantly alter the optimum level of investment. In other countries, for example the Netherlands, risk to life plays a greater role in investment decisions.
New work has been carried out on the mental health impacts of people who have experienced flooding. It shows that these impacts are greater than previously thought. This will increase benefits and present a case for further investment. For example, we used a potential new method to calculate the benefit cost ratio for a flood risk management scheme. The new method included the impacts on mental health. In this example, the benefit cost ratio increased from 16:1 to 24:1.
Headline 7: over 60% of properties in England are served by infrastructure sites and networks located in (or dependent on others in) areas at risk of flooding.
The Environment Agency worked with the University of Oxford to explore the impacts of flooding from rivers and the sea on transport and utilities infrastructure including:
The analysis included:
- the interdependencies between networks
This is ground-breaking work using a unique set of national infrastructure modelling tools. The analysis did not have access to data about the resilience of individual sites. Consequently it was limited to describing the extent to which properties in England are served by infrastructure located in (or dependent on others in) areas at risk of flooding. We were unable to include telecommunications because data was not available. The main limitations to this work are the availability of data on:
- infrastructure location
- site resilience
- redundancy in networks
- disruption duration
The results highlight the importance of infrastructure resilience. Forty-one per cent of transport and utilities infrastructure assets are in areas at risk of flooding, comprising:
- 36% at direct risk of flooding
- 5% at risk due to dependencies on electricity
The analysis shows that impacts are sensitive to the resilience of infrastructure – in particular the upper estimate of impacts. Many infrastructure owners have invested to improve infrastructure resilience, as outlined in the National Flood Resilience Review. As a result, some infrastructure on the flood plain is resilient to a very high level of protection.
LTIS 2019 significantly improves our understanding of:
- the potential impacts to infrastructure without resilience
- how these impacts can be used to improve investment modelling
Confidence in our assessment of these wider impacts has improved. This allows them to be better represented within the LTIS optimisation calculations. The effect of this on optimum investment in the baseline scenario is:
- an increase in NPV from £101 billion to about £220 billion
- the overall benefit to cost ratio increases from about 5 to 1, to about 9 to 1
- a revision of the optimum level of investment upwards from a long-term annual average of £860 million to £933 million, which would increase risk reduction from 12% to about 15%
9. Different ways to manage risk
LTIS 2019 analysis looks at a broader range of ways to manage flood risk. It goes beyond the activities included in the baseline scenario, but does not include everything. Insurance, coastal realignment and property relocation are outside the scope of LTIS. The activities included are:
- conventional FCERM activities used in baseline scenario
- different ways to manage risk – property flood resilience, temporary flood barriers and natural flood management
The Environment Agency analysed the national potential of these different ways to manage risk so investment has not been optimised at a local scale. We did this using a separate piece of analysis, because we’re not able to integrate them into the core LTIS model. They were analysed individually and in combination. We can now better reflect the investment outcomes using a range of FCERM activities, both in isolation and together.
This section considers the management of the residual flood risk from rivers and the sea. The residual flood risk is the risk that remains after FCERM investment has taken place. In this case it is the risk:
- after the optimum level of investment in the baseline scenario
- with investment in conventional FCERM activities
- with medium climate change
It does not:
- assess what mix of activities best manage risk in a particular location
- include how the extra activities may reduce flood risk from surface water - so in that respect it understates their potential
This analysis has high uncertainty around individual values. We can describe the potential of these different approaches at a national level. We can comment on their combined effect on NPV and the optimum level of investment.
9.1 Property flood resilience (PFR)
PFR aims to make people and their property less vulnerable to the physical and mental impacts of flooding. It prevents water entering a property or minimises the impact if water does enter. We explore the potential effectiveness of PFR by considering a range of different packages. These include:
- resistance (water exclusion) and resilience (water entry) measures
- active (manual) and passive (automatic) components
PFR is limited to reducing the risk from shallower flooding. It’s often used alongside conventional assets, or where conventional approaches are not feasible. PFR can:
- reduce the risk of flooding
- help recovery
- enable people to take responsibility for their own risk
To date, PFR has had relatively small scale application but as our understanding of PFR options improves there is potential for larger scale application.
Headline 8: property flood resilience is cost effective for a large number of homes and could reduce risk by about 16%.
LTIS 2019 shows we could reduce damages by up to a quarter if PFR was adopted by all residential properties at risk. This is not a realistic scenario but it shows the potential of PFR to manage flood risk.
It’s cost effective to use PFR to manage residual risk. The Environment Agency found it would be cost effective to invest in PFR for large numbers (over 200,000) of residential properties. This would reduce the risk (property damages) in the baseline scenario by a further 15% to 18%. A package of manual resistance measures (such as flood guards) will generally achieve the best value for money. This depends on the property owner putting them in place when they receive a flood warning. It’s a good example of how combinations of FCERM activities interact to reduce risk.
Table 5: relationship between PFR and flood warning service level
|Flood warning service level||Percentage effective action for this service level||Percentage of best value NPV that can be achieved with active (manual) resistance||Percentage of best value NPV that can be achieved with passive (automatic) resistance|
|No external service||41%||0%||77%|
|Present day warning service||84%||23%||59%|
|Enhanced warning service||90%||66%||21%|
|Hypothetical perfect service||100%||100%||0%|
9.2 Temporary flood barriers
Temporary barriers are a useful tool for risk management authorities. They can provide flood protection for communities that are not, or cannot be, protected by permanent defences.
Headline 9: the potential of temporary flood barriers to reduce risk at a national scale is limited by the number of practically viable sites.
We estimate there are about 150 to 300 viable sites which could reduce risk (property damages) by 1% to 4%. Local use is dependent on a number of factors which are hard to assess at a national scale.
Temporary barriers have much less potential than PFR to reduce residual risk at a national level. Where they are viable they can be more cost effective than PFR.
9.3 Natural flood management (NFM)
LTIS 2019 is the first time the Environment Agency has analysed the economic potential of NFM at a national scale. NFM aims to work with natural processes, features and characteristics to reduce flood risk. It achieves this by protecting, restoring, and emulating the natural function of catchments, rivers, estuaries and coasts.
We based the analysis on the Working with Natural Processes (WWNP) evidence directory. The WWNP sets out the best existing knowledge. It uses case studies to estimate the cost of reducing peak river flow. We applied these reductions to areas on maps showing potential for WWNP. We found that NFM alone is less likely to be able to offset the impacts of climate change for later epochs and for higher emissions scenarios.
Headline 10: natural flood management has a complementary and beneficial role alongside investment in conventional FCERM assets. But it alone cannot reverse national increases in risk due to climate change.
We can manage catchment-scale flood risk more cost effectively. We can do this by investing in NFM alongside conventional FCERM activities rather than by investing in conventional FCERM alone.
We found a moderate amount of NFM combined with conventional FCERM investment gave the highest NPV for a given level of investment. This combination increases NPV from £101 billion in the baseline scenario, to £104 billion. This suggests NFM can complement conventional FCERM activities. Note that NFM is not appropriate everywhere. It’s most effective where the best opportunities exist. Competing land-use pressures may constrain its use. It also needs the cooperation and support of landowners.
The LTIS 2019 analysis only considers the flood risk reduction benefits of NFM so has undervalued its potential. It does not include other benefits of NFM such as:
- improved resilience to droughts
- better water quality
- improved biodiversity
- improved amenity values for people
Significant uncertainties in the evidence for NFM and our method limit the accuracy of the results. Most NFM case studies cover small catchments. We know less about its effectiveness to manage large floods and large catchments. Future improvements to the analysis depend on more evidence. Better evidence on the costs and benefits of NFM would benefit policy and funding decisions for both FCERM and land management. Government gave £15 million of funding for NFM projects which will help provide more evidence.
We can compare different ways to manage risk. Residual risk (property damages) can be reduced to a different extent by different FCERM activities:
- a reduction of 15% to 18% by large scale property flood resilience (up to a maximum of 26%)
- a reduction of 1% to 4% (up to a maximum of 7%) by temporary barriers
- in some catchments NFM can offset medium climate change, but this proportion decreases in later epochs
9.4 Combined effect of a range of risk management activities
LTIS 2019 shows the potential effectiveness of combining a number of different ways to manage risk. The results show that by combining approaches it may help to tackle increases in risk in higher climate change scenarios. This finding is confirmed by evidence from:
- portfolios of responses in Foresight Future Flooding
- the enhanced whole system adaptation scenario within evidence for the Climate Change Risk Assessment 2017
Headline 11: in combination with conventional FCERM activities, investment in a wider range of activities to reduce risk gives better value.
A wider range of activities has the potential to increase NPV by £6 billion over the baseline scenario. The extra activities need an extra £50 million annual average investment over the baseline optimum. With these activities, the benefit cost ratio remains about 5:1. The best range of activities is likely to include conventional FCERM at its current level combined with investment in a wider range of activities.
There are lots of uncertainties in the analysis of individual activities. In the future, a more integrated analysis could optimise the balance of activities at a catchment scale. Investment in conventional FCERM activities is already closely aligned to the optimum. Table 6 suggests where further investment could be allocated. These are inexact national proportions, and will vary from place to place.
Table 6: indicative proportion of investment in the different FCERM activities
|FCERM activities||Indicative proportion of cost|
|Conventional FCERM (as per baseline)||84%|
|Property flood resilience||8%|
|Natural flood management||7%|
You can use our beta (prototype) tool to show the effects of changing national investment in different FCERM activities. You get different outcomes by choosing different levels of national investment in different activities.
10. Scenarios with specific levels of protection
The purpose of LTIS analysis is to optimise investment in FCERM. We have limited flexibility to address other policy questions.
The Natural Capital Committee and National Infrastructure Commission (NIC) recommended some goals for future FCERM. LTIS 2019 includes new scenarios to explore the investment and outcomes associated with these recommendations. These scenarios reflect the specific levels of protection that conventional FCERM assets can provide to reduce flood risk from rivers and the sea. In the baseline scenario, conventional FCERM assets can provide levels of protection up to 0.5% annual likelihood of flooding.
10.1 Optimising investment with the option of a very high level of protection
We added a new scenario to LTIS analysis. It includes an option to invest in FCERM assets that provide a very high level of protection. A very high level of protection has a 0.1% annual likelihood of flooding from rivers and the sea. This is an option, not a guaranteed minimum level of protection. It has only been adopted in places where the benefits of a very high level of protection are greater than the costs.
In theory this could make a significant difference to the number of properties at risk. It could halve the number of properties at high and medium risk. In practice there are social and technical limitations which make these very high levels of protection difficult or impossible to achieve in many places.
For example, to contain an extreme flood through an urban area, very high flood walls or a lot more space for the river is required. To make this space, it may be necessary to remove buildings from the centre of towns and cities. This may not be socially acceptable and is likely to be expensive. This would have an impact on value for money.
We could invest limited funds in very high levels of protection. However, this may also mean investing less in other places where value for money would be better.
Headline 12: theoretically very high levels of protection could make a big positive difference to long-term risk outcomes. There are social and technical limitations which make these very high levels of protection difficult or impossible to achieve in many places.
Analysis of these very high levels of protection is more uncertain. This is because of limited cost data. A practical way to improve our understanding of costs would be to appraise a very high level of protection option wherever feasible.
Table 7: comparison of the baseline scenario to a new scenario with the option of a very high level of protection
|Metric||Baseline scenario (negative values show a reduction)||With the option of a very high level of protection (negative values show a reduction)|
|Properties at high and medium risk from rivers and sea in 2060s||765,000||288,000|
|Change in property damages after 50 years||-12%||-42%|
|Annual average investment need over 50 years (range)||£790 million to £920 million||£900 million to £1100 million|
|Present value cost||£25 billion||£28 billion|
|Net present value (NPV)||£101 billion||£112 billion|
A higher estimate of the costs of very high levels of protection gives a present value cost of £32 billion and NPV of £100 billion (both over 100 years).
10.2 Estimating the costs and benefits of national protection standards
In its National Infrastructure Assessment, the NIC recommended:
that government should set out a strategy to deliver a nationwide standard of resilience to flooding with an annual likelihood of 0.5 per cent by 2050 where this is feasible. A higher standard of 0.1 per cent should be provided for densely populated areas where the costs per household are lower.
We created a comparable scenario in LTIS. We adjusted the baseline scenario, applying these resilience standards to FCERM assets only. We made the model apply these standards, so investment has not been optimised in this scenario.
Table 8: comparison of the baseline scenario to a new scenario representing the NIC recommendation for national standards
|Metric||Baseline scenario (negative values show a reduction)||NIC recommendation (modelled in LTIS) (negative values show a reduction)|
|Properties at high and medium risk from rivers and sea in 2060s||765,000||598,000|
|Change in property damages after 50 years||-12%||-23%|
|Annual average investment need over 50 years (range)||£790 million to £920 million||£1000 million to £1340 million|
|Present value cost||£25 billion||£30 billion|
|Net present value||£101 billion||£101 billion|
A higher estimate of the costs of very high levels of protection gives a present value cost of £40 billion and NPV of £90 billion (both over 100 years).
These results compare with the NIC estimate of £1.1 billion average annual capital investment (in 2017 prices). Removing revenue costs from the LTIS results gives a figure slightly lower than the NIC estimate. Both analyses use similar climate change assumptions, but the NIC assumes low population growth. The LTIS baseline effectively assumes zero population growth, so the NIC estimate should be higher. Given the uncertainty in the methods and cost data used in these analyses, the results are well aligned. The NIC also estimated costs for high population growth and 4˚C climate change at over £1.8 billion average annual capital investment.
10.3 Comparing scenarios with specific levels of protection
These new scenarios show that better FCERM outcomes (numbers of properties and expected damages) can be achieved by an economic optimum. This would give some areas very high levels of protection compared to consistently applying national protection standards. The optimised scenario reduces property damages almost twice as much as the national protection standards. It showed a reduction of 42% compared to 23%.
There are large uncertainties associated with analysis of very high levels of protection. However, if we appraise their costs and benefits at a local level they can be good value for money in some places. The economically optimised scenario provides better value for money than applying national protection standards.
Both scenarios are subject to social and technical limitations which make them difficult or impossible to achieve in many places. They would also require an increase (compared to the baseline) to annual average investment of over £100 million.
11. New evidence about long-term investment levels
The new evidence shows that the overall economic optimum level of investment is higher than was estimated by the LTIS 2014 baseline. The evidence is from separate pieces of analysis. Together they can help us manage the greater risk in higher climate scenarios. The challenge for future analysis will be to examine integrated sources of, and solutions to, flood and coastal erosion risk.
Headline 13: new scenarios show that the overall economic optimum level of investment is higher than previously estimated. The new estimate is a long-term annual average of over £1 billion.
The Environment Agency has improved the analysis of the wider impacts of flooding – in particular the impacts on infrastructure. It shows the true optimum level of investment is greater than the baseline suggests. There’s also evidence that the baseline underestimates the amount of investment needed to manage risk from surface water. If we revise the baseline scenario to include these 2 improvements we expect the optimum level of investment to increase by an annual average of over £100 million.
Policy choices also affect the optimum level of investment. These include:
- preparing for a high rather than a medium climate change scenario would increase the optimum from a long-term annual average of £860 million to £920 million
- high climate change both increases costs and limits risk reduction: property damages reduce by 4% rather than 12%
- investing across a wide range of activities could increase NPV by £6 billion and justify further annual average investment in the order of £50 million
- investing in very high levels of protection can be effective at reducing the numbers of properties at risk - it could require an annual average increase of over £100 million
The new analysis has highlighted a limitation of the current LTIS model. The limitation is that it’s difficult to explore options around the timing of investment. If we postpone or bring forward investment it could have a significant impact on short-term outcomes. The new analysis has also highlighted the impact of high++ climate change on asset deterioration. In some places this could increase costs so much that assets are left to degrade.
11.1 The overall optimum level of investment
The evidence above shows that the overall economic optimum level of investment:
- is higher than estimated in LTIS 2014
- has a long-term annual average of over £1 billion
The level depends on policy choices and could range from £1.0 billion to £1.2 billion. This follows current government guidance for investment.
The long-term annual average of over £1 billion is based on the optimum level of investment across medium and high climate scenarios. We then add an annual average of over £100 million due to a better understanding of risk from surface water and infrastructure impacts. Other policy choices, such as a wide range of activities or very high levels of protection, could increase it further, and achieve further benefits.
Investment profiles can vary over time. To reduce risk at a greater rate than the long-term average, short-term levels of investment must be greater than the long-term average. This overall estimate is in 2015 prices, and is based on recent costs. If costs go up, investment will need to increase. Cost scenarios in LTIS 2014 show how increases in costs push up the optimum level of investment and bring down the risk reduction achieved.
12. Developing science to improve the analysis
The Environment Agency will continue to improve our understanding of future investment choices for FCERM.
12.1 Priorities and recommendations
We’ll be able to improve the scope and accuracy of the LTIS analysis using the new:
- national assessment of flood risk from all sources
- climate change projections (UKCP18)
The surface water management action plan sets out steps to improve our understanding of the risks from surface water. Other emerging research, for example on mental health impacts, will also contribute to better scenarios.
We want future LTIS analysis and capabilities to include:
- an integrated assessment of investment across all sources of flood risk, all FCERM activities and a wider set of metrics to support place-based decision making and the 25 Year Environment Plan
- robust decision making so that uncertainty informs the results
- alternative, flexible ways to model future investment in order to explore timings and sequences of investments and different targets and services provided by both authorities and individuals
There are specific areas of relative weakness which future analysis should improve:
- how development and land use change can alter run off rates
- assessment of the benefits of FCERM so that the natural capital and local infrastructure benefits can inform investment choices
- the costs and benefits of managing surface water flood risk
12.2 Open and collaborative approach
We take an open and collaborative approach to LTIS development. This has:
- enabled more experts to contribute
- provided more scrutiny of new work
- provided advice on how to meet future aims
Jacobs developed the technical evidence summarised in this report, with support from:
- University of Oxford
- Horritt Consulting
- Centre for Ecology and Hydrology
- the Flood Hazard Research Centre
We established a LTIS Development Group of leading academics to steer strategic thinking. We have a comprehensive review process to give us confidence in the LTIS 2019 findings. This involves working with:
- independent expert reviewers (Jaap Flikweert of Royal HaskoningDHV, Paul Sayers of Sayers and Partners, Edmund Penning-Rowsell of Flood Hazard Research Centre and Richard Dawson of Newcastle University)
- the National Infrastructure Commission and Climate Change Committee to share data and compare approaches and results
- experts from the Environment Agency, government departments and other organisations (such as Flood Re and Natural England)
- risk managers and long-term investment analysts in the Netherlands (Rijkswaterstaat and Deltares) and the United States Army Corps of Engineers.