The Scottish Government

Contents |

Listen

CHAPTER 11: FINDINGS AND RECOMMENDATIONS

Information on Flood Prevention Schemes

11.175 One of the key findings during the project has been that the level of information on each scheme promoted under the Flood Prevention (Scotland) Act 1961 varies widely. Information is held by the Scottish Executive and the local authorities. Of the 72 schemes registered by the Scottish Executive, and still in existence, file information is available for only 52.

11.176 Some local authorities were initially unaware of FPS within their area e.g. the Yoker Burn FPS lies within Glasgow and Sauchie Burn lies within Clackmannanshire. This tends to be for older schemes where information has not been passed forward by predecessor bodies.

11.177 Other local authorities have information on Flood Prevention Schemes for which there is no information with the Scottish Executive files, e.g. Dumfries and Galloway have existing schemes in Carsphairn, Minigaff, Glenluce, Monreith, Moniaive and Creetown that are not known to the Scottish Executive.

11.178 Additional flood protection measures were submitted by some local authorities and their location and status has been recorded within the database. As they fall outside the 1961 Act none have been studied in detail, however this suggests that more properties may be protected and this could be reviewed.

Biennial Report

11.179 The Biennial reports required by the Flood Prevention and Land Drainage (Scotland) Act 1997 provide information on flood risk in each local authority area. It is recommended that Scottish Executive and local authorities agree to their list of formal schemes to assist in biennial reporting and ensure clarity for the purposes of SPP7.

11.180 Faults identified at each scheme during the condition assessment were identified to the relevant local authority. Faults amended by the local authority should be identified to Scottish Executive so that they can be updated or removed from the SFDAD as necessary. This process could be integrated within the biennial reports. Defences particularly soft defences should be periodically surveyed to ensure that the design defence level is maintained.

11.181 Standardisation of the structure and content of the reports would be beneficial. A template biennial report should be produced and made available to the local authorities to enable the easy and consistent production of the biennial reports. The biennial report template could include a blank datasheet for the authority to return updates to the database at the Scottish Executive.

11.182 The reports also hold information on the frequency of flooding which if held on GIS could become a useful tool in determining overall levels of flood risk and providing information to support planning and development control issues and scheme development. The availability of a central asset database offers an opportunity to help standardise information for these reports, possibly generating diagrams and annexes, and act as a depository for them. This information could readily be integrated within the database and it is recommended that the inclusion of this data is investigated.

11.183 The reports indicate that over 48 potential schemes are either under appraisal or consideration. Given the previous rate of scheme submission, this suggests an increased demand on resources for flood management.

Coastal Schemes

11.184 The SFDAD currently holds information on Flood Prevention Schemes promoted under the Flood Prevention (Scotland) Act 1961. Information on the coastal Flood Prevention Schemes within Scotland could be gathered and included in the SFDAD. This will provide a clear picture of flood risk and areas benefiting from defences in tidal areas. The number of properties at risk within tidal areas is likely to drop ²⁵ once the new Flood Maps are used in preference to the current assessment using the 5m contour.

Asset condition

11.185 The condition of each flood prevention asset was appraised and assigned a score between 1 and 5. In addition an assessment of the residual life of each asset was made. Most flood defence structures are designed to provide service over long life spans i.e. 50 or 100 years. Most of the formal Flood Prevention Schemes are less than 30 years old and therefore most assets still have considerable life. Well maintained flood embankments can be considered to last for at least 100 years. Well constructed flood walls are also likely to last considerable periods. The exception to this is where assets are continuously impacted or eroded, i.e. sea walls or erosion protection measures. Sea walls at Prestonpans are in need of maintenance for instance.

11.186 In total the schemes provide over 90km of assets including 35km of embankments, 21km of walls, 16km of culverts and 18km of channel improvements. Of assets surveyed 87% are in good and very good condition and a number of the rest could readily be improved through maintenance improvements.

11.187 Notably no mention has been found in any scheme's proposals for the renewal or replacement of short life assets. Other assets such as pumps can be considered to have a design life of 20-25 years after which they are usually replaced. This can lead to a significant maintenance cost and needs to be included within economic appraisals.

11.188 Gabions are widely used in Flood Prevention Schemes often as erosion protection but also to construct or reinforce banks or walls. There are differences in opinion as to the life span of gabion structures. For this project their life has been assessed as being in the order of 20-30 years in line with research undertaken by Scottish Natural Heritage ²⁶. This means that a number of schemes will incur recurring costs and this has been reflected in the revised benefit cost appraisals. Given the nature of their construction more work on assessing gabions suitability in sustainable flood management is recommended.

11.189 As part of this study, model reports were issued for each scheme studied in detail. Information on action taken by the local authorities following their review is recorded in Table 11-1.

Table 11-1: work undertaken since draft reports

Reassessment of Peak Flood Flows

11.190 As part of this study new flood estimates were derived for fluvial sites and tidal levels for coastal sites. It would appear that a number of schemes were designed following a damaging flood and simply to prevent the flooding associated with a flood of similar magnitude.

11.191 In the majority of cases, updated flow estimates are higher than those used in the original scheme design. An increase in flow estimates has lead to a decrease in the standard of protection offered by the Flood Prevention Schemes. This is not uncommon and relates to the introduction of new methods (i.e. Flood Estimation Handbook) and increased flood information.

11.192 As this report and maps will be used in conjunction with the Indicative River and Coastal Flood Maps (Scotland) the flows used by SEPA in the generation of the flood map have been included in the database. Flows have been taken from the SEPA flow grid which is predominantly Auto FEH values derived by CEH with some changes using single site analysis of individual SEPA gauging stations extrapolated to cover an associated catchment. Flows estimated for this work may differ from the Auto FEH as we have used additional datasets made available by SEPA that are not used in the derivation of Auto FEH flows.

11.193 For the sites assessed in detail we requested information on gauging stations within 40 km of the site. This yielded an additional 73 stations appropriate for use in the flow estimation procedure. This suggests that there is a large amount of valuable data not readily available. This information could be made more readily available, ensuring more refined flood estimates for studies and flood risk assessment and prevent the need to request the data from SEPA. This would both improve routine flood estimates and free resources within SEPA from data requests. The HiFlows- UK website went part of the way to achieving this but not all sites that are suitable were put forward for the HiFlows- UK project.

11.194 It was found that flood flow estimates were often determined early on in a scheme's promotion. As FPS's tend to take some time to come to construction, this means that even relatively new schemes rely on hydrology that was derived before the Flood Estimation Handbook was widely used. This has led to differences in flow estimates associated with using the Flood Estimation Handbook rather than the Flood Studies report which was widely used in FPS design. Flow estimates should be checked close to the final construction stage of an FPS and scheme designs checked to ensure they still meet the required design standard.

11.195 Flow estimation methods for small catchments have not been updated or improved in recent years. Uncertainty in such estimates is therefore higher. Small catchment hydrology is an area that could benefit from further research to assist in defining flows on ungauged catchments. Small catchment flooding in urban areas is a particular area of current focus e.g. east end of Glasgow and Renfrewshire.

11.196 Research and guidance into the appropriate approach for flow estimation and probability with disparate sized catchments would be beneficial. Many flood problems are associated with small catchments discharging into larger rivers with defences required on both.

11.197 Scottish and Southern Electricity controls large catchments and heavily influences flows in some Scottish catchments. Flow estimation in hydro regulated catchments is difficult and this could be improved if models used in the design and operation of hydro-electric schemes was made available for flood management purposes.

11.198 The revised rainfall runoff procedure has been launched in the UK but requires further validation in Scotland. Confirmation that this method is appropriate and that it can be extended to the 200 year flood estimates required for SPP7 and scheme design would be a useful part of the validation process.

Hydraulic Modelling

11.199 Hydraulic modelling is widely used in flood management. Design events are often well beyond the magnitude of recently experienced floods and models are used to extend the understanding of flooding and provide information for economic analysis. Ideally they are calibrated against previous flood events. There were very few hydraulic models readily available and surprisingly little information with which to calibrate models. Four models of existing schemes were made available for this study (Collier Street, Crosslee, Turfford Burn, and Brock Burn and Levern Water); a fifth for Fettercairn, upstream of an existing scheme and was extended as part of this study.

11.200 New river cross sections were used to provide accurate information on the channel which is important where channel improvements were part of the scheme. The sections were then extended using NEXTMapDTM data. The NEXTMapDTM was found to be variable and often needed to be used to provide indication of landform or to be adjusted using threshold levels or levels taken in the floodplain. LIDAR, where available, was more accurate. The NEXTMapDTM is sufficient for feasibility but it is recommended that the detailed design of new schemes should be based on topographically surveyed data.

11.201 There is a need to check hydraulic models once schemes are constructed. This is particularly important when in channel variations have been made over the assumed sections used in the modelling exercise. Ideally this requires 'as-built' drawings to be remodelled to check capacity of the constructed FPS.

11.202 To facilitate future flood management, copies of hydraulic models should be kept by the local authorities and also be submitted as part of the scheme record to Scottish Executive. While the Intellectual Property Rights would need to reside with the organisation undertaking the work, a licence for the Scottish Executive and its appointed agents to use the model data files should be included in contract conditions and made a condition of grant aid.

11.203 Better collection of flood event data would assist in calibrating models for the design and assessment of extreme flood events. There are current research proposals to increase knowledge in this area, the information could be integrated into the database.

Flood Mapping

11.204 There were very few flood outlines available and those that were, were predominantly historical maps. As part of this exercise all historical maps were collated and recorded. SEPA kindly provided their flood map data and a revised outline and table were provided to SEPA. Scottish Water is understood to have records of flooded properties which could assist greatly in flood management, particularly in understanding urban flooding and overland flow. Although this data was requested it was not received within the project time frame.

11.205 Better collection of flood event data would assist in flood management, planning and scheme design. The data should ideally be held centrally and made available to planners, regulators, local authorities and the public. Historical flood outlines should be recorded and stored ideally with levels of the flood discharge and details of the data and circumstance of the flood event. Historical outlines can provide real data for calibrating modelled outlines.

11.206 There were no differences in survey datum's picked up during the work. Detailed surveys can show the line of the watercourse to be different to that on OS landline maps. This is an issue with the accuracy of maps and can lead to defence assets being located within watercourses.

11.207 The NEXTMapDTM is very variable, it provides data for broad scale floodplain modelling but is weaker for detailed studies and very weak in urban areas. In this study, levels were checked and supplemented with local survey. The DTM does provide good indication of landform. LIDAR was used where available and is significantly better for detailed flood mapping and associated flood studies. LIDAR data would improve flood mapping in urban and areas of extensive floodplain.

11.208 New schemes should provide probabilistic flood maps at the time of scheme submission, and should show the situation before and after the scheme is in place and include an assessment of the 1000 year flood event.

11.209 The database has layers for these return periods, historical maps and for the area benefiting from a scheme. Additional outlines are likely to be derived during the production of benefit cost appraisals and these should ideally be recorded.

Standard of Protection

11.210 New schemes are designed to meet at least a 100 year flood and then a freeboard is normally added to allow for uncertainties in modelling and hydrological or tidal estimates. This research has established how the scheme will perform with no freeboard and recorded the standard as the "onset of flooding", in addition the "standard of protection" has been determined using a freeboard allowance. The freeboard was calculated for each defence element using the Fluvial Freeboard Guidance Note ²⁷. The standard of protection can be seen as a robust estimate of the return period to which a scheme protects. The onset of flooding is the easiest standard to understand and the one most local authorities are most comfortable with, generally because it provides the best standard. It is recommended that the standard of protection is standardised and this should comprise of the standard of protection with a freeboard allowance.

11.211 The results of this study indicate that few schemes currently protect to the designed standard. Whilst this may initially appear alarming this is to be expected, return periods may change when flood estimation is reassessed and these can change substantially with new methods i.e. the use of the Flood Estimation Handbook or more recent high flows may alter the return period of previous estimates. In addition many projects are promoted over a considerable time and the hydrology is often determined early on in the project and not revisited at the time of detailed design or promotion. Some of the earlier FPS were promoted to defend against a flood that had recently occurred and the return period assessed for that event was used in the design.

11.212 This indicates that current and future scheme design flow estimates will be further refined and updated, and may indicate the need for a flexible and/or a precautionary approach to future scheme design.

11.213 Throughout this analysis the standard of protection has been shown as both a return period and a flow or level. The performance of a scheme in relation to a defined flow or level is more robust and less likely to vary (although channel changes may affect this). Any changes in the performance in relation to the design standard cannot be accredited directly to climate change.

11.214 The database also makes an assessment of the impact of climate change at 2080; new schemes may choose a different horizon. It is recommended that one design horizon is used for climate change analysis.

Number of Defended Properties

11.215 The total number of properties currently protected by flood prevention schemes in Scotland is 4580. A significant number of properties are protected by the Perth scheme. The number of properties still to be protected to the 100 year return period, the current preferred design standard, based on the schemes with sufficient information and assessed in detail is 1570. There is a significant benefit that could be derived from improving the condition of the defences to the 100 year standard; however this is likely to require increases in defence heights that may go outside original deviations shown in scheme drawings.

11.216 The ratio of commercial/industrial and residential properties protected by the schemes in Scotland is approximately 1:4. The number of properties protected by flood prevention schemes in Scotland has followed a trend of gradual increase until the mid to late 1990's, at which point several schemes have significantly increased the total number of properties protected.

11.217 The remaining FPS's tend to be small and old and are less likely to increase the overall number of properties dramatically. A notable exception would be Yoker Burn which flows through an urban area and would protect a number of properties.

Economic Assessment

11.218 Until recently flood prevention schemes were not rigorously appraised for economic benefit. Many of the schemes in Scotland show a positive benefit-cost ratio; however they may only offer a low standard of protection. The more recent schemes often have benefit cost assessments and where available these have been reviewed. The analysis undertaken for this project suggests a number of key areas for review.

Information supplied at submission

11.219 Few schemes were found to have economic assessments, which is largely a function of the relatively recent requirement to complete an assessment. Of those completed few made estimate of maintenance or recurring cost within the calculations. Most appraisals however, did use recurring benefits and this has lead to a significant reduction in benefit cost ratios when maintenance is included in the analysis. It is recommended that all benefit cost analysis includes future maintenance within the appraisal.

11.220 When overall costs for a scheme escalate they reduce the benefit cost ratio, this may lead to the local authority being required to undertake maintenance on a scheme under the Flood Prevention and Land Drainage (Scotland) Act 1997 whilst the activities may not be economically sustainable. It is recommended that guidance is drafted for this situation.

11.221 Flood extents at a number of return periods, levels, depths, property type and threshold levels should ideally be provided as part of the detailed submission. This will allow subsequent review of appraisals. Future prioritisation of schemes will become increasingly important should the demand on funds available for flood prevention schemes become constrained.

11.222 The occasional review of schemes may be required in the future and having a range of flood outlines would assist in this and any nationwide appraisals of flood management.

Planning

11.223 The damages have been assessed up to the 1000 year flood. The analysis shows there is considerable residual damage within schemes, which has implications for planning and development within areas protected by flood prevention schemes. At present SPP7 allows development in areas benefiting from flood prevention schemes. This still represents a risk and a potential cost to the nation. An alternative would be to consider development but to assess the impact up to the 1000 year flood. This would encourage any development behind the defences to be further protected.

Future

11.224 The economic assessment has shown that many of the schemes are cost beneficial even though they only provide a low standard of protection. Thus benefits from low return period events are important in making a scheme economically viable. At present, schemes have to meet a minimum standard of 100 years before they are eligible for grant aid. Relaxation of this standard would remove a key parameter in establishing schemes that are sustainable and will provide genuine benefit to residents protected by a scheme.

11.225 Sustainable flood management is currently under investigation; it would appear that few schemes considered options in any real detail. In addition most schemes have left little space for rivers and tended instead to maximise developable areas. This can result in damage to the environment and leaves little opportunity to improve existing flood management. The new guidelines on sustainable flood management may improve the awareness but need to be carefully integrated with the legislation to encourage environmentally sensitive schemes.

11.226 The construction of new property behind existing defences will increase the residual damages for above design floods. Any new development in areas benefiting from flood prevention schemes should minimise the increase in residual damages.

11.227 The removal of the 100 year design standard may enable an increase in damage avoided and provide schemes that are economically sustainable.

11.228 At present there are at least 48 potential flood prevention schemes under consideration. This represents a potential challenge in prioritisation of resources.

Management of information

11.229 The Scottish Flood Defence Asset Database provides a valuable asset that provides Scotland with a concise repository of information collected using standard methods. Having all this information in one place is both unusual and highly beneficial in determining future flood management. It also presents an opportunity to continue to build on the information collected and consider the future management of flood asset information.

11.230 Flood outlines derived from this study are to be included within the Indicative River and Coastal Flood Map (Scotland) and both the flood maps and the database are to be published on the web. There could be benefit in merging both flood risk information and flood asset information in one application.

Contents |