Climate Adaptation of Roads to Flooding Hazards—A Review

Abstract

Flooding hazards due to climate change are increasingly becoming a frequent global occurrence. The aim of this study is to provide a comprehensive review of the various structural mitigation and adaptation strategies available to engineers and designers at various stages of road construction and rehabilitation to increase the resilience of roads to flooding damage. The criteria for categorising the various strategies available were the time of intervention with respect to the occurrence of the hazard. Thus, all studied strategies were separated into pre-construction design changes, post-construction mitigation and adaptation options like Sustainable Urban Drainage Systems (SuDS). The main findings were that changing the specifications of commonly used materials can provide increased flood resilience, and a preliminary design for flooding can reduce post-flooding rehabilitation. The study can be used as a guide for the different options available to deliver a design that takes flooding into consideration.

1. Introduction

Flooding events are an emerging effect of climate change that cause disruptions to transport services, damage to infrastructure and significant life and economic losses. Depending on the severity, floods can either reduce the capacity and increase congestion or damage transport infrastructure, severely causing loss of connectivity and incurring increased repair, maintenance and operational costs [1]. This affects the entire economy as transport infrastructure connects the more central areas of a country to the most remote [2]. Among the tangible effects of flooding, the cost of traffic disruption of any main road in the UK is estimated to be £ 100,000/h at peak times [3] and the expected annual damage caused by surface water flooding is expected to increase by 130–140% by 2080 due to future climate scenarios [4]. Concurrently, the worldwide investment in infrastructure is estimated to increase to £ 4.5 trillion by 2030 [5]. These statistics highlight the importance of well-planned decision making in reducing damage to transport infrastructure [6]. In urban areas, the intensity and frequency of flooding damage is expected to increase in the coming decades due to the global phenomenon of urbanisation and the consequent rise in dense population centres and the expected high frequency of extreme weather events.

Disasters has been defined by the United Nations International Strategy for Disaster Reduction (UNISDR) as events that break down social structures and disrupt social functions. They arise from the complex interactions between natural, social and built environments like infrastructure systems [7]. While more frequent and intense extreme weather events as well as rising temperature and sea levels affect the transport infrastructure directly, an indirect effect is the climate effect on infrastructure need due to changes in the distribution of transportation, traffic flows, human behaviour and societal changes [8].

With regard to road infrastructure, the main effects of climate change on roads are included in

Table 1. Flooding effects on road infrastructure [9].

Hazard	Damage/Impact	Typical Mitigation Measure
Fluvial/river flood due to extreme precipitation	Inundation, washout, deterioration and loss of skid resistance due to excess water	Improve/maintain drainage (increase ditch and culvert capacity)
Pluvial/surface flood due to extreme precipitation	Inundation, washout/cracking	Improve/maintain drainage
Landslides (rainfall induced, including sliding, debris flow, mudflow)	Closure by debris flows or mudflows	Warning signs; protection measures (debris shelters, barriers, fences, ditches, tunnels)

. The hazards listed in this table do not have the same return time. For example, in the UK, flooding due to extreme precipitation was recorded several times in the last two years but landslides were less frequent.

Flash floods have severe effects on transport infrastructure, vehicle circulation and accident rates. Such extreme climate hazards reduce the structural strength of the infrastructure elements causing damage, an increase in accidents and high repair costs. Other effects of flooding include the increase in travel time, speed drops and road capacity reduction [10]. Figure 1 illustrates damage caused by flooding in roads.

The response to climate change challenges includes mitigation and adaptation. While mitigation measures are those that are taken to reduce the underlying reasons for causing and enhancing climate change, for example, greenhouse gas emissions, adaptation is used to refer to interventions that reduce the vulnerability of the infrastructure. Mitigation and adaptation complement each other with the common aim of a reduction in losses due to climate change phenomena [13].

Examples of mitigation measures in road construction include changes in construction materials that reduce the impacts of high temperatures on roads like asphalt rutting, flushing, and bleeding of bituminous surface and cracking. Some examples of such changes are the use of binders with a higher softening point, the polymer modification of bitumen, the selection of strong aggregate material, the higher use of concrete due to its higher temperature resistance, changing concrete mix design to reduce the requirement of water, increasing the reflectance of pavements, etc. [14,15]. Examples of adaptation measures include Sustainable Urban Drainage Systems (SuDS) like swales, filter drains, filter strips, bioretention systems, etc., that increase the capacity of a built-up area to tolerate high stormwater loads.

While examining the existing literature on the topic of adaptation to flooding for road infrastructure, it was noted that a review study focused on discussing the various mitigation and adaptation strategies in the context of construction materials and methods for the purpose of reducing the effect of flooding on road infrastructure is not available.

For example, de Abreu et al. [16] conducted a systematic review of 280 studies on the impact of climate change on road transport infrastructure and the identification of hard and soft adaptation measures. However, while the hard adaptation options for various climate hazards were identified, these were not discussed in detail in their study and a deeper analysis of the various action strategies for the climate adaptation of road infrastructure was not provided.

Lu et al. [17] conducted a review of the adaptation strategies used to reduce flood risk to pavement infrastructure. However, while the pavement adaptation strategies of different countries were described, very few engineering strategies were discussed, and the focus remained on adaptation framework models. Most studies on studying or improving pavement resilience to flooding focus on determining the structural damage to pavement by flooding [18]. Qiao et al. [19] conducted a comprehensive review on the effects of climate change on pavements. However, the study was split between different hazards, and while many measures to mitigate the effects of flooding were included, these were not differentiated on any basis and received less attention in the study. Swarna and Hossain [20] conducted a comprehensive study on the climate impact on and adaptation of highway asphalt pavements. However, while they captured most elements of the damage caused by climate change to pavements, the solutions to combat these were not included in much detail. Based on these studies, a gap in the literature for a comprehensive study on hard or structural adaptation measures that can be implemented to make pavements more resilient to flooding damage was identified. Many researchers have also pointed out the importance of using innovative construction materials and methods in climate mitigation and adaptation [21]. While reviews on the use of innovative and flood resilient materials are present in the literature [22], they were not differentiated based on the suitability of application in relation to the pavement’s lifecycle. In this study, the application of various innovative materials as well as construction techniques has been included, and these have been differentiated into when and where they can provide the maximum benefit.

The aim of this study is to provide a comprehensive review of the various structural mitigation and adaptation strategies available to engineers and designers at various stages of road construction and rehabilitation to increase the resilience of roads to flooding damage.

2. Methodology

This study is a scoping review that was conducted in the initial phase of a research programme focused on the climate adaptation of transport infrastructure. The broad purpose of the review was to focus on the descriptive characteristics of the included sources to find viable research problems. The scoping review allowed the authors to broaden the objective from an initial objective of identifying existing structural mitigation and adaptation strategies to increase the flooding resilience of road infrastructure to the wider objectives of identifying structural mitigation and adaptation strategies and determining when these strategies can be implemented to reduce the effect of flooding events. The framework Population Concept Context (PCC) was used to identify the key elements of the scoping review.

The primary questions were the ‘structural mitigation, adaptation strategies, flooding resilience of roads’. The secondary questions were the ‘existing resilience, increasing the flooding resilience’. The inclusion criteria involved the concepts of the studies which included ‘changes in design of roads and associated transport infrastructure, structural changes in existing flood-affected or flooding-prone roads’. Another inclusion criterion was the context of the study which was ‘changes in construction materials and methods’.

The alternative keywords to the concepts and their synonyms were identified as ‘structural mitigation of roads to flooding damage, adaptation strategies, resilience of roads to flooding damage, Sustainable Urban Drainage Systems, infrastructure flooding resilience, transportation flooding resilience, transport infrastructure adaptation, road adaptation to flooding, road mitigation for flooding’. Using the Knovel database, the following relevant subject headings were identified for the identified concepts, ‘Pavement Resilience, Post-Flood Damage, The flood hazard, Natural Hazard Protection, Mitigation, Climate Adaptation Strategies, Transport Infrastructure Systems, Flood and Infrastructure, Resilience, Post Disaster Reconstruction, Hazard Mitigation Planning, Asphalt and Bituminous Materials’. The search was streamlined using truncation to find plurals and different word endings. The search was joined together using the connectors ‘And’ to narrow the search and ‘Or’ to broaden the search.

The information sources used were Google Scholar and the University of Huddersfield Library. The databases that were used included Scopus, ASCE Civil Engineering, ScienceDirect, TRID-Transportation Research Information and Knovel Search. The search was first conducted on the two main information sources, Google Scholar and Huddersfield Library, using the keywords, subject headings and their combinations. The text words in the titles and abstracts of the research papers and reports were analysed to run another search across all the databases identified. The reference list of identified reports, research papers and articles were then searched for additional studies.

The exclusion criteria were studies on climate prediction, the effect of landscape on flooding, flow models, flooding progress through the landscape, vulnerability assessment, risk assessments, measurement of structural strength loss of pavements, qualitative and quantitative measures of flooding damage. While these are themes connected to the topic of this paper and affect and measure the damage caused by flooding, they are excluded because they do not answer the objective of the study, which focuses on changes in road materials and construction methods that can increase the resilience of roads to flooding.

Several approaches to classifying the knowledge obtained from the sources were studied. However, a novel approach introduced by Mouratidis [23] was applied. Mouratidis pointed out the lack of accuracy in the existing models of risk identification and assessment and has shared several examples of failures of road infrastructures in the face of climate hazards in recent decades. The reasons enumerated by the author for these failures include the reduced engagement of road authorities and stakeholders in climate change adaptation policy and the difficulties in applying existing models of infrastructure adaptation prior to road network planning stages. In response, the author proposed a different adaptation approach to extreme climatic events which includes assessing the road design and its exposure level to climate hazards, making a distinction between rainstorms that cause events like flash floods and long rainfall events and restoring the initial balance of a site to natural pre-construction runoff routes and methods. The adaptation measures have been divided into preliminary, preventative and balance restoring, protective and auxiliary measures. This new approach was defined by the author but not applied to their study. This approach also does not provide clarity about the time of application of the mitigation or adaptation strategy.

In the current research study, the categories outlined in the approach by Mouratidis [23] are broadly used, but the criterion adopted for categorising the mitigation and adaptation measures is the time of intervention with respect to the occurrence of the hazard, which separates them into pre-flooding and post-flooding mitigation techniques.

The pre-flooding mitigation techniques can be pre-construction design changes or post-construction mitigation measures. The post-flooding interventions involve repair and rehabilitation as well as increasing the resilience of pavements against further hazards. The adaptation of roads has been covered separately because it involves changing the entire environment in which the road exists to improve the response of the road to the climate hazard. This involves changing the approach to the design of public spaces, drainage systems, involvement of multiple stakeholders and agencies.

3. Flooding Damage and Its Mitigation

Flooding damage in roads can have several mechanisms. While the damage to road infrastructure like roads being washed out and the presence of flood debris can be visually observed, the damage to the pavement layers is often unseen. This damage can be assessed by non-destructive testing, hydrological surveys, comparison with pavement design standards, comparison with pavement management databases, historical damages, comparison with projected traffic volumes and an analysis considering future heavy load traffic. The entry of water into the pavement layers reduces the strength and stiffness of these layers causing a reduction in the load-carrying capacity of the saturated road. When this structurally weakened road segment is subjected to traffic load, the result can be rutting, cracking, loss of pavement life and local structural failures, as shown in Figure 2.

The draining out of water can result in the restoration of strength. However, the timeline for this restoration is dependent on the permeability of the pavement layers, the underlying soil and the hydraulic condition on site. Due to this reason, a relationship between the degree of saturation and the resilient modulus of the pavement layers has been explored by several researchers to determine the loss of strength due to moisture ingress [24]. Thinner pavements have been observed to sustain more damage than thicker pavements, and flexible pavements have been found to be more vulnerable to flooding damage as compared to rigid and composite pavements [25]. The measures to prevent flooding damage to roads have been differentiated into pre-flooding and post-flooding mitigation and adaptation measures.

3.1. Pre-Flooding Mitigation Measures for Asphalt Pavements

Pre-flooding mitigation measures for roads include selecting flood-resilient construction materials. For roadways subjected to flooding, an experimental study has shown that the subgrade California Bearing Ratio (CBR) decreases with a higher inundation time and quick and high settlement can also occur in such a situation [26]. For a typical asphalt pavement consisting of a hot mix asphalt surface layer over an unbound aggregate base, subbase and subgrade courses, water enters through the surface layer due to its coarse grade and high permeability and increases the moisture content of the surface and base layer. For high inundation times and depths of ponding, the full saturation of these layers causes a drastic decrease in the load-bearing capacity of the unbound aggregate layers while the surface layer continues to support traffic due to the strength of its constituent materials, even after the base has been weakened. Thus, it is essential that pavement design should include materials resistant to damage at high levels of saturation and that the time of saturation is higher than the expected flooding duration. The saturation time can be increased by using thicker surface and base layers and using asphalt mixes with lower deterioration rates of tensile strength with time [27].

The saturation time also depends on the material and layer properties as well as the distresses in the pavement, for example, the cracking potential, predicted from the ageing potential of the mix, with the presence of cracks increasing permeability and decreasing the saturation time. Sealing cracks and resurfacing of old pavements is required to prevent the saturation of the underlying layers. The cracking potential is also dependent on the mean annual air temperature of the road location and can be reduced by using low-viscosity asphalt and a lower initial void ratio [27]. The lower initial void ratio is significant for a newly constructed pavement where compaction due to traffic has not yet occurred, and a higher void ratio in a flooding scenario means high permeability, leading to lower saturation times and a higher risk of damage. The lower void ratio is ensured by providing a very dense layer by using finer gradations and binder-rich, specialised seals, using the appropriate numbers of rollers and with the possible use of geosynthetic layers. However, a void ratio below 3% or above 8% is not recommended for dense-graded mixes [28] because a lower void ratio can cause rutting and shoving, while a high void ratio increases the risk of water damage, oxidation, ravelling, and cracking [29].

Other measures to decrease permeability include providing a surface layer with higher thickness to protect the underlying granular base course, using materials on the surface to prevent cracking and to seal existing seal cracks. Permeability has been found to be dependent on particle size distribution, shape and density (air void or percent compaction), the nominal maximum aggregate size and the thickness of the layer. Relationships between these factors have been obtained by testing, and these can be used for design purposes [29].

The important parameters in increasing the saturation time for the asphalt layer are air voids, gradation, thickness, cracking and tensile strength loss, listed in decreasing order of effect. For the base layer, the relevant factors are the thickness and matric suction in order of decreasing effect, while for embankments, the presence of riprap can increase saturation time. The effects of these factors were determined by system dynamic simulations carried out on a mathematical model [30] with equations derived from various sources [31,32,33]. While these studies cover the standard parameters that predict pavement saturation time, a factor that is not considered is the use of innovative materials or additives designed to specifically address the saturation time and moisture susceptibility which can be a disrupter to these calculations and yield higher or different results.

While the previous section discussed practical measures that can reduce the effect of flooding on road surfaces, embankments are also an integral part of roadways, and therefore, this discussion cannot be complete without examining the possible measures that could be used to prevent flooding damage in roadway embankments. Established techniques for the erosion control of slopes include vegetative turfing, use of netting, asphalt mulch, slopes according to soil type, drainage solutions, structures like diversions, berms, waterways, pipe outlets, retarding structures like spillways, grade stabilisation structures, various types of mechanical structures like drop, drop inlet and chute spillways, etc. [34] These techniques fall on a spectrum varying from low-cost measures to high-cost structural measures like spillways and concrete paving, where the decision to apply a particular solution depends on the requirements of the site. However, to extend the variety of solutions available to designers and engineers beyond the traditional choice between slope paving, riprap, concrete and vegetation [35], three innovative and cost-effective materials are examined in this section to examine their effectiveness in embankment flood protection.

Pre-flood mitigation measures to prevent damage to embankment slopes include the use of vegetation and its substitutions like Erosion Control Blanket (ECB) and bluegrass sod, geogrids, metal chain link fence, Turf Reinforcement Mat (TRM) paired with Fibre Reinforced Matrix and flexible concrete geogrid mat. One technique to install the ECB is the ‘armoured sod’, with a biaxial plastic geogrid underlain with the ECB to model vegetation. The plastic grid is placed over existing grass as a retrofit or pinned over seeded slopes with anchoring systems, e.g., wire U-shaped staples. This solution, illustrated in Figure 3 and Figure 4, helps stabilise the root system against washing away by stormwater.

TRMs, shown in Figure 5a, are non-degradable, synthetic, three-dimensional mats that are pinned to the slope with U-shaped wire staples and provide permanent support to the vegetation roots. Fibre Reinforced Matrix (FRM) is then sprayed over the TRM as a water-based slurry of fibres (wood or cellulose), tackifiers, seed, and fertiliser [35].

Flexible concrete geogrid mats, shown in Figure 5b, consist of rough-surface concrete blocks cast on flexible and tough polypropylene geogrids that hold the concrete blocks together. This configuration is placed over a TRM. This is the heaviest among the previously mentioned configurations, with the weight of the blocks holding the TRM in place while the geogrid’s flexibility helps to adjust it to the shape of the ground underneath.

Armoured sod is suitable for cases where the vegetation is newly planted, and this solution stabilises the young and weak root system against washing off in chunks. TRMs are suitable for long-standing vegetations with root structures that require permanent support. The sizes of the mat openings are also large to allow for large seeds to grow through the surface. When tested experimentally, the armoured sod showed better erosion protection than bare soil, and the flow per unit width at the time of erosion was 10 times the flow it took to erode bare soil. However, it was found to be only marginally effective if laid on bare soil and flooded before the vegetation takes root, and it was found to be suitable for low slopes. Erosion occurred after 12 h. The blue-grass sod option showed slightly better performance on the higher slope. The TRM showed better resistance to erosion, but still demonstrated slope failure at the toe of the slope, while the flexible concrete geogrid mat performed the best out of the three types, with no erosion after 30 h of maximum flume flow [30]. When compared to riprap, the armoured sod and the TRM have a much lower purchase price, while flexible geogrid is more expensive (but still one-third the cost of fully grouted riprap). However, coverage is partial for all three, while grouted riprap ensures full coverage against erosion and scour in comparison. It is concluded that the three options considered provide partial protection to embankments against an overtopping flow, and while the flexible geogrid shows better results among the three, the product needs further development to show results comparable to grouted riprap and other slope protection measures.

Design Changes to Infrastructure

Design changes to road infrastructure to prevent flooding damage include raising embankment height and adding hardstone pitching, improving embankment side slopes and enhancing natural drainage systems. Stone pitching is shown in Figure 6. When the difference in water levels between the upstream and downstream sides is significant, hardening the downstream shoulder and armouring the downstream embankment can protect them from scour. When overtopping is expected, hardening the entire roadway section, including the slopes, is recommended. Embankment protection measures include vegetative turfing, geosynthetic netting, stone pitching with graded rocks, flexible mats, flexible and rigid pump-up revetment mattresses, grouted rock and gabions.

Scouring in culverts can be prevented by adding a layer of grouted rock adjacent to the upstream and downstream ends of the wing walls and head walls. Construction of concrete downturn walls at the inlet and outlet ends of apron slabs prevents the erosion of bedding materials. Floodways or causeways can also be designed at a shallow location of a road that is expected to flood, and they can be designed to overtop but resist damage due to flooding [37]. The suitable flood protection measure for any location can be determined by conducting a hydrological survey. The cost and design of the measures should be justified by the requirements of the location.

3.2. Post-Flooding Adaptation Measures and Pavement Rehabilitation

Post-flooding adaptation measures for roads include rehabilitation and reconstruction depending on the level of damage. Pavement failures are often linked to low strength and saturated subgrade, even more so in the case of reactive subgrade materials that have a low CBR and a high swelling potential, which could cause longitudinal cracks due to changes in the moisture content [38]. The rehabilitation design depends on the extent and type of damage incurred on a road or road network.

An example of extensive post-flooding rehabilitation work is the Queensland Transport Network Reconstruction Program (TNRP) that was developed to reconstruct and repair the road transport network after a cyclone and major flooding damage in the region. In this programme, tests were conducted on flood-affected pavements and the results were used to classify pavement rehabilitation work into pavement patching, part-width, and full width pavement rehabilitation [39]. Figure 7 shows a flowchart of the methodology adopted by the TNRP for determining the appropriate rehabilitation measure for a flood-affected pavement.

A key learning point from the TNRP was that post-flooding recovery and rehabilitation are often site-specific, and planning of rehabilitation works depends on information about the subgrade type, slope conditions, period of inundation, extent of damage, as well as the strategic importance of the road. It was found that pavement failures mainly occurred in regions with weak and reactive subgrades, and that the flat terrains caused the roads to be inundated for longer periods of time. In response to such site-specific weaknesses, sealing the full width of the pavement, raising the level of the road, providing longitudinal drains and treating the subgrade were found to be useful interventions. Strategically important roads were often given a more resilient treatment in the rehabilitation process, while second-order roads were given less robust treatments. The most common treatments provided were a cement-modified base, the granular re-sheeting of unsealed roads and granular overlay [38]. It was also observed that post-flooding rehabilitation funding and eligibility rules for funding can change the scope, pavement treatment selection and nominated design life of different roads in the network. Local knowledge with regard to locally sourced construction materials and construction practices while handling the material were also found to be relevant and useful in the rehabilitation process. The methods of increasing flood resilience included (a) improving road drainage by repairing or deepening table drains, clearing underground drains and culverts, repairing surface cracks, adding geotextile seals and a drainage layer at the subgrade level, (b) reducing shoulder and embankment erosion by increasing flood protection on downstream sides of floodways and increasing the drainage capacity of waterways [38]. The pavement treatments applied used 1% general blended cement in the base (addressed moisture sensitivity of some granular materials), in situ modification of unbound granular materials, improving the strength of white rock (CBR < 80) through cementitious modification, lime stabilisation of subgrades, cement as secondary agent in foamed bitumen stabilisation work, geotextile/fabric layers and geogrids (over low-strength subgrade areas), slag bends as stabilising agents (increases working time and controls shrinkage cracking), bitumen-treated base, single-pass mixing during cement modification and 3% cementitious stabilising agent. It can be observed that in flooding-affected regions, rehabilitation measures include the stabilisation and strengthening of existing materials.

Another example of the post-flooding rehabilitation of pavements is given by Saniga et al.’s study of the Kerela 2018 floods [40], where several combinations of materials were examined, as shown in Figure 8, and it was found that the flood resilience index of cement-treated bases was superior to that of emulsified asphalt-treated bases. Cement-treated base layers reduced the required pavement thickness, thus proving more economical while also increasing flood resilience and service life. The cement stabilisation of the base course builds a stronger bond between the base and the surface course and prevents the stripping of the surface layer. It also builds a stronger bond between the base and surface courses and prevents stripping of the surface layer. The cement-treated base or subbase roads were also found to have a longer service life than alternatives like the emulsified asphalt-treated base combined with wet mix macadam or reclaimed asphalt pavement combined with a granular subbase.

For granular base flexible pavements, a thicker surface course protects the granular base, for example, increasing the surface course thickness from 30 to 200 mm can create a six-fold increases in critical flooding time [37]. For the surface course, the flood resilience of bituminous concrete was found to be higher than that of premix carpet with a seal coat [40]. By reducing the air voids in the surface course, the permeability of the surface course is reduced [35]. For resurfacing, the flood resilience of bituminous concrete was higher than that of premix carpet.

Concrete or rigid pavements are preferred where the inundation time as well as the traffic load in such conditions is high. Other concrete pavement types for such conditions include Interlocking Concrete Block Pavement (ICBP), cell-filled or geogrid, short-panelled and Roller-Compacted Concrete (RCC) pavements [37]. Figure 9 illustrates some of these concrete pavement solutions.

An important lesson learned from these studies are the design changes carried out to increase the flooding resilience of roads, which include raising the road level, sealing the road surface, providing longitudinal drains and treating the subgrade when the quality of the subgrade is poor. These changes increase flood resilience and reduce the damage caused by flooding events. The superior performance of cement-treated bases is also seen in both the pre-flooding and post-flooding studies and highlights the importance of improving material properties to resist water damage.

4. Climate Adaptation and Retrofitting of Road Infrastructure for Flooding

Another approach to preventing climate change damage is the climate adaptation of the entire infrastructure system. This includes increasing climate resilience in newly built infrastructure as well as making existing infrastructure more resilient through the appropriate retrofitting and maintenance regimes. Climate adaptation planning includes risk assessment, vulnerability assessment and multi-criteria analysis for the development of a robust adaptation policy. However, as the scope of this paper is limited to structural adaptation options, this topic is not covered in this study.

While in the previous sections, the pre-flooding, design and post-flooding scenarios with respect to the individual road structure were examined, another way of reducing flood damage is changing the landscape to minimise the surface water or optimise the treatment and disposal of surface water. This can be achieved by implementing Sustainable Urban Drainage Systems (SuDS), which is a broad, overarching term including all the design measures that can be adopted to maximise the benefits and opportunities of surface water management [41].

One difference between the execution of the pre-flooding, post-flooding and design changes that were discussed in previous sections and SuDS adaptations is that the previous measures could be implemented by a single engineering or road construction agency, whereas the successful SuDS design and implementation requires liaising between multiple agencies like the local planning authority, the environmental regulator, planners, developers, the drainage and flooding authority, the local council, water and sewerage undertakers, open-space management, the municipal authority, etc., and inputs from drainage engineers, urban designers, highway engineers, architects, landscape architects, ecologists, local community representatives, resident organisations, etc., are required for its successful implementation [41].

In urban areas, impermeable surfaces create large volumes of runoff from rainfall, and if the path that the water takes while reaching the drainage system is short, the drainage system can become overwhelmed by stormwater. Sustainable Urban Drainage Systems (SuDS) seek to remedy this problem by reducing the runoff coefficient of surfaces receiving rainfall and by increasing the time that water takes to reach the drainage system. By reducing the runoff coefficient, surfaces like roofs, pavements, paved and unpaved surfaces retain some of the rainfall received by them and by prolonging the time that the water spends in the engineered system before reaching the drainage pipes, the opportunity for water to evaporate, transpire or infiltrate is created, which further reduces the volume of water. Stormwater runoff also has the effect of mobilising sediment, litter and pollutants as it washes over a developed urban catchment, and this water can pollute the water bodies that receive this water if allowed to move untreated. SuDS create the opportunity to capture and treat this water prior to discharge in natural water bodies. If allowed to mix with sewage, it causes an increase in the volume of contaminated water that must be treated prior to discharge. SuDS are also adaptable for increasing the capacity at a reasonable cost compared to subsurface drainage systems if water quantity increases beyond the design values in the future. SuDS provide urban cooling, shade and insulation to buildings and increase natural ventilation. A limitation of SuDS is that they are effective for short- and medium-duration, high-intensity events that cause local flood risks but have less impact on flood risks caused by longer-duration events [41].

Urban flooding adaptation strategies involve measures classified as followed with respect to water: (a) harvesting (urban greenery and rooftop retention for, e.g., green walls, green roofs, tree alignments); (b) storage (retention and bioretention basins and reservoirs for, e.g., cisterns and underground reservoirs); (c) infiltration like permeable pavements and infiltration techniques for, e.g., infiltration trenches, leaky wells; (d) conveyance (stream recovery by rehabilitation and restoration); (e) tolerance (ability of system to tolerate excess water, e.g., floating systems) and (f) avoidance (impeding the presence of stormwater by automatic floodgates, waterfront embankments, glass flood walls, storm surge barriers, removable metal plaques, breakwaters, etc.) [42,43]. Permeable pavements, green roofs, swales, raingardens, underground retention basins and other such features are part of the blue city, green city or the blue–green city concepts which are all a part of the overarching term of SuDS. Several SuDS solutions used for urban flood adaptation have been categorised in

Table 2. Urban flood adaptation categories and measures.

Category	Measure	Category	Measure
Urban greenery	Green walls	Levees	Gentle slope levees
Urban furniture	Inverted umbrellas Art installations	Floodwalls	Sculptured walls Glass walls
Rooftop detention	Green roofs	Raised structures	Cantilevered pathways
	Blue roofs		Elevated promenades
Wet proof	Submergible parks	Coastal defences	Breakwaters
	Submergible pathways		Embankments
			Multifunctional defences
Infiltration techniques	Infiltration trenches	Barriers	Demountable barriers
Stream recovery	Stream rehabilitation Stream restoration Daylighting streams	Floating structures	Floating pathway Floating platform Floating islands
Open drainage system	Street channels Extended channels Enlarged canals Check dams	Permeable paving	Open cell pavers Interlocking pavers Porous paving
Reservoirs	Artificial detention basins	Bioretention	Wet bioretention basins
	Water plazas		Dry bioretention basins
	Underground reservoirs		Bioswales
	Cisterns		Bioretention planters
			Raingardens

. SuDS reduce the volume of water that will be drained by the urban drainage system, and this reduces the required size of the built or ‘grey’ infrastructure capacity (pipes, culverts, treatment facility, outlets, etc.) needed to manage high storm runoff.

SuDS retrofitting methods aimed at making roads more resilient to flooding include low-impact development technologies like permeable pavements, vegetated swales and bio-retention [44]. When designing a SuDS solution for a site, the four pillars of SuDS, i.e., water quantity, water quality, biodiversity and amenity should be taken into consideration [45]. Some of the SuDS solutions are explained in detail in the following sections.

4.1. Swales

Swales are suitable for draining long stretches of a roadway when the road and ground are at the same level. They are not suitable for limited spaces or where roads exist on embankments (unless lined) because the pressure of the water may cause destabilisation. Figure 10a shows an example of a swale with a check dam that limits water velocity. In Figure 10b, the swale and the wetland system are designed as such so that, at some spots, the water flows directly into the swales, while at other locations, the water is collected in edge channels (when embankment height is low) or kerb drains (when kerbs are required) before its discharge into the swales [45,46,47].

4.2. Filter Strips

These are uniformly graded strips of grass or other vegetation that have a gentle slope (approximately 1 to 5%) and drain off water from impervious areas. These are often placed prior to swales to promote filtration, sedimentation and limited infiltration. The gradient should be low and uniform as the water should ideally move as a sheet at a low velocity (maximum recommended 1.5 m/s) to deliver the required filtration. Clogging and improper landscaping can affect performance; length is an important factor (>2.5 m is ideal and >5 m is highly effective) and their primary function is water quality improvement [45,47].

4.3. Filter Drains

These are shallow trenches filled with aggregate like stones and gravel. They may be lined with an impermeable liner like geotextiles or by concrete troughs. Alternatively, they may allow infiltration depending on the suitability of the ground. Their purpose is attenuation, conveyance and filtration and they ideally receive pre-treated water in the lateral direction from an impervious surface. Though their primary purpose is not sediment trapping, if the water received is completely untreated, a shallow geotextile may be inserted and regularly cleaned to filter the water entering the drain. A perforated pipe is provided near the base to convey the collected water to the downstream drainage component. They occupy less space, have a lower storage capacity than swales and can clog more easily. The depth of filter drains should be between 1 and 2 m, and the filler medium should be at least 0.5 m beneath any inflow and outflow pipe network. When located next to filter strips or flow spreaders, filter drains can replace kerbs and gullies. The main cause of damage to filter drains involves vehicles scattering the filter material. Adequate protection from this as well as designing for any possible vehicular traffic are recommended [45,49].

4.4. Detention Basins

Detention basins are used in locations where space is ample, for example, motorways and trunk roads. They are also suitable for junctions and roundabouts [45]. An example of a detention basin [47] is shown in Figure 11.

4.5. Bioretention Systems

These are shallow landscape depressions that collect runoff from frequent rainfall events which filter down through the vegetation and specialised engineered soils. The filtered soil can be collected by an underdrain or can be allowed to infiltrate into the surrounding soil. Their main purpose is interception, and the design consists of an inlet, temporary storage of water on the surface (150–300 mm), vegetation, filter layer (750–1000 mm deep), transition layer or geotextile, drainage layer, perforated pipes and overflow. Raingardens are smaller than bioretention systems and have an inflow of 150 mm maximum depth that is underlain by thin (200–500 mm) layer of compost or sand-amended native soils or engineered soils [45,51].

4.6. Ponds and Wetlands

Ponds are depressions with a permanent volume of water that provide attenuation as well as water treatment, while wetlands contain shallow layers of water that promote the growth of bottom-rooted plants. They require upstream pre-treatment systems, and a flow control system at the outfall controls the rate of discharge into the pond. The aesthetics and integration into the landscape are important parts of the design of ponds and wetlands [45,51].

4.7. Attenuation Storage Tanks

These are spaces provided below the ground level for the temporary storage of water prior to infiltration, controlled release or use. The storage tank can be made of geocellular materials, plastic corrugated arches, oversize concrete or plastic pipes, corrugated steel pipes, hybrid structures using reinforced earth walls, concrete box culvert sections and tanks, glass-reinforced plastic tanks and concrete roof panels. The main benefits of these tanks over other systems of storage are their high storage volume and the option to be installed beneath roads and car parks. High values of the peak runoff rate are common in built-up areas, as there is no natural surface to stop or slow the water flow. Attenuation storages are introduced to store the water and discharge it at a lower rate to the receiving water body [45,52].

4.8. Comparison of SuDS Measures

The various SuDS options can be combined in different ways to deliver the required performance on any site. For long stretches of roads, filter strips can act as a first line of treatment and can be discharged into swales. The purpose of filter strips is the pre-treatment and removal of sediments prior to their discharge into swales and bioretention units. Filter strips also have a low level of infiltration and are not suitable for areas with shade that might limit grass growth. Their effectiveness depends on their length in the direction of the flow, so they are not suitable for limited spaces. If placed too close to impervious surfaces they face a risk of clogging. They can be installed close to industrial sites but must be lined if used in contaminated or brownfield sites.

Pre-treated water can move into swales or filter drains depending on the site type and requirements. Swales require more space and are suitable for sites with steep slopes while filter drains are suitable for flat slopes (less than 2% longitudinal slope). Kerb drains can also discharge into swales which can convey and attenuate, overlay underdrain systems or deliver to wet and marshy land, depending on the type of the swale. The major advantage of using swales is that if used with filter strips and flow spreaders, they can replace conventional pipe work. However, the requirement for space makes them unsuitable for dense, urban developments or embankments. Swales are not suitable for embankments where infiltration can cause instability. They are also not suitable for areas where excessive fertiliser or weed killer has been applied and require lined underdrains with levels above the seasonal highest groundwater level if built on brownfield or contaminated sites.

In comparison, lined filter drains can manage runoff areas with high groundwater contamination risk, they are suitable for sites where vegetated solutions are not feasible and can be easily incorporated into the landscape and public places. They can accept pre-treated flow by a filter strip or equivalent which they will attenuate to remove fines, metals and hydrocarbons. They need to be at some distance from the carriageway to prevent stone scatter by vehicles and water ingress in pavements. When placed next to pavements, filter drains need to be lined with impermeable membranes, and they are suitable for catchments with small impermeable areas. While suitable for sites with a lower slope (unlike swales), they have a lower storage capacity as compared to swales, clog easily, have a risk of slope and pavement failures, do not trap sediments and are not suitable for unstable ground or areas with a continuous groundwater flow. They are also not suitable for retrofitting.

Detention basins, however, can be used for retrofitting in both residential and non-residential multi-functional spaces. Detention basins provide interception for small rainfall events and remove sediments and buoyant materials. This contrasts with bioretention systems, which provide interception for frequent rainfall events, followed by infiltration into an underdrain or surrounding soil. Detention basins have the disadvantage of clogging at the outlets, are not suitable for brownfield sites and may require a pool structure at the discharge site for the settlement of sediments. It is essential to check the groundwater historic record of the area prior to their development. While detention basins, as well as ponds and wetlands, are suitable for use in sites with large spaces like junctions, roundabouts, motorways and trunk road network, bioretention systems are appropriate at traffic calming zones, road buildouts, dead spaces in car parks and turning areas. The major benefits of bioretention systems are the reduction in the runoff rate and volume, treatment of pollution, improvement of landscape and biodiversity, cooling effect and ability to integrate in both low- and high-density developments. However, they require an overflow or bypass to downstream drainage components during extreme events, and they should not have impermeable liners unless there is a need to prevent infiltration, for example, close to structural foundations or for groundwater protection.

Attenuation storage tanks are suited to a location with space constraints. Attenuation storage tanks provide below-ground storage prior to infiltration, release or use, while ponds and wetlands provide flood attenuation for all events, including 10-, 30-, 100- and 200-year events. The disadvantage with attenuation tanks is the lack of accessibility for maintenance and a high cost compared to surface solutions. They lack treatment performance if used in isolation and so require integration with above-ground treatment systems. Geocellular and plastic arches require sediment treatment upstream due to the lack of accessibility, and geocellular systems require a more intense and precise design to protect them from failure. The main benefits of attenuation tanks are that they are flexible in size and shape, have a high storage volume and save space as they can be installed beneath roads and car parks.

In comparison, ponds and wetlands require large surface spaces, but the vegetation supported by them ensures the adhesion of contaminants, aerobic decomposition of pollutants, stabilisation of settled sediments and treatment through settling and biological uptake. Ponds and wetlands require upstream pre-treatment and are not suitable for unstable grounds, steep slopes, or sites with water-fill, uncontrolled-fill or non-engineered-fill. They do not provide interception and will not reduce the contaminated load from volumetric runoff. They should not be designed without a landscape architect, should be oriented in the direction of prevailing summer winds, should be overlooked by housing and require impermeable liners and separation from groundwater if designed on contaminated land.

In contrast to the other SuDS measures, permeable pavements serve the purpose of filtration, absorption, biodegradation and sedimentation. They are suitable for use in highways with a low-volume or low-speed traffic, car parks and non-trafficked areas. However, research on their use at heavily trafficked roads has been conducted. The major advantage is that they absorb the flow close to the source, which allows other attenuation features to be smaller and shallower. They can work in relatively flat terrains, and as they replace impervious pavements, there is no excess land use. The limitation is that they are not suitable for areas with a high silt load and must not be unlined if they receive runoff from brownfield or contaminated sites [45]. As the focus of this paper is adaptations of pavements, permeable pavements are explored in further detail in Section 4.9.

4.9. Permeable Pavements

Permeable pavements are a structural Stormwater Control Measure (SCM). The pavements include a surface layer through which the water infiltrates into a permeable base or subbase layer that acts as a water reservoir and supports the traffic load. The subbase thickness can be designed based on the structural and hydrological requirements. Perforated underdrain pipes are present at the base. Weirs and upturned elbows can be added to promote infiltration, increase temporary storage and avoid rapid dewatering by storing the water in the reservoir. The underdrain can be eliminated from the design due to various site conditions, e.g., the low permeability of natural soil, existing high groundwater elevation, contamination of soil or presence of shallow bedrock. The designs are illustrated in Figure 12 [53].

These pavements have a higher resistance to freeze and thaw damage as there is a reduction or complete elimination of thin ice accumulation on the pavements in the winter season. Other advantages include reduced runoff volumes and peak discharges, increased infiltration and groundwater recharge, improvement in water quality, reduction in stormwater temperature and a heat island effect, reduction in drainage system infrastructure requirement and cost and pavement surface benefits like wet weather glare, reduced heat island effect, reduced hydroplaning risk and standing water and increased road safety due to these pavement surface benefits. Permeable pavements can be used for the construction of walkways, driveways, sidewalks, low-volume roadways, patios, courtyards, parking lots, alleys, playground spray pools, etc. They are suitable for geographically constrained urban areas where they provide underground drainage, thus saving overground space for other uses. The material and design depend on the requirements of the site. However, they are not recommended for roads with high-speed, high-volume traffic, consistent heavy loads, or roads with a high pollution potential, sediment deposition or organic material accumulation. Permeable pavements are also not suitable for locations where high turning motions are expected to occur in the same spot [54]. Permeable pavements can be classified into two types and four main categories as follows: monolithic (Permeable Asphalt (PA), Permeable Concrete (PC)), and modular (Permeable Interlocking Concrete Pavement (PICP) and grid pavement systems) [55], as shown in Figure 13.

4.9.1. Porous Asphalt

Porous Asphalt (PA) pavements consist of a permeable asphalt surface layer without fines to create greater void space, laid over an open-graded aggregate choker course and a reservoir bed. Figure 14 shows a typical PA pavement. The use of additives and binders provides higher durability and makes it possible to consider these pavements for use in roads with high loads and traffic volume [53]. The choice of asphalt grade and polymer binder can balance or minimise the performance degradation of PA pavements. Gap-graded mix designs have been found to be better at meeting the air void requirements of PA [56].

PA pavements can be used in climates that are appropriate for conventional asphalt and show good performance in cold climates because the permeable, open graded components allow for water expansion during freezing, which makes them less susceptible to damage due to freeze and thaw as compared to conventional pavements. However, based on the depth and intensity of frost penetration, the base depth can be increased. For example, in somewhat frost-susceptible soils, base thickness can be 35–55 cm, for a substantial frost penetration base or a subbase it can be 65% of the local frost design depth and in extremely frost-susceptible soils, the improved base or subbase layers can be provided up to the full frost depth. PA pavements are less susceptible to the formation of black ice, require less ploughing and de-icing, and they require 50 to 75% less winter maintenance salt than conventional pavements. In warmer climates, the draindown of the asphalt binder into the pavement layers and the consequent clogging may occur during construction when the hot mix asphalt/wet mix macadam is at a high temperature and can be prevented by using the appropriate asphalt binder grade, mix temperatures and additives [53].

Ma et al. (2016) [57] examined the use of high-viscosity binder in PA pavements and found that they increased the compressive strength, resilient modulus, durability and the low-temperature cracking resistance and reduced its draindown ratio. In comparison, adding DBS (Dibenzylidene Sorbitol) additives, fibre (polyester, mineral and cellulose) and hydrated lime improved performance against some criteria but reduced performance against others [57]. Cellulose fibres are the fibres commonly used in PA pavements to stop binder drainage, thus preventing loss during storage and transport, and their addition has been found to increase vertical and horizontal permeability and resistance to permanent deformation [58]. Thus, the ratios of the additives in PA pavements should be carefully selected to obtain the best results. The structural design of PA pavements should vary with the functional and service requirements of the site [56].

Though, PA pavements have been found to favourably affect some risk factors associated with accident occurrence, the net impact of this on accident occurrence has been found to be inconclusive [59]. While the ravelling of PA pavements due to traffic leads to a shorter life, PA pavements have also been found to self-heal and recover up to 90% of their original resistance when kept under pressure at a temperature of 25 °C [60].

In the Netherlands, the double-layer PA pavements have been extensively constructed. The pavement structure consists of an upper layer consisting of fine aggregates and small air voids overlaying a lower layer of course aggregates with larger air voids. A typical design [61] is shown in Figure 15.

However, the higher volume of voids results in lower strength and moisture stability making the pavement susceptible to permanent deformation and ravelling. To address these issues, high-viscosity asphalt is used with the addition of thermoplastic elastomers, fibres, nanometre and cement as additives. Among these, Tafpack-Super (TPS) [62], polymer modified bitumen (PMB) containing more than 9% SBS (styrene-butadienestyrene) [63], and Crumb Rubber (CR) have been extensively tested. Other benefits of double-layered PA pavements include a reduction in urban heat island effect, skid resistance, pollution and tyre–pavement noise.

4.9.2. Pervious Concrete

Pervious Concrete (PC) pavements are created by adding open-graded aggregates into a hydraulic cementitious mix which creates an approximately 15% to 25% interconnected void space [53] and interconnected pores between 2 and 8 mm size [64,65,66] that allow the rapid infiltration of water. The addition of aggregates makes the appearance coarse, and additives can be added to improve strength and binding [53]. The water-to-cement ratio is lower than that of conventional concrete and typically varies in the range of 0.26–0.45 [66]. The aggregate size, shape and distribution as well as aggregate type influence PC pavement properties, e.g., dolomite aggregates provide a higher compressive strength compared to limestone or slag aggregates and granite aggregates. PC has a superior freeze–thaw performance as compared to other aggregate materials [67]. An increase in fine aggregates has been found to cause a decrease in the volume of voids, which in turn increases the compressive strength, flexural strength and split tensile strength [67,68,69,70,71,72]. Alternatively, large size aggregates have a higher surface area leading to a higher void ratio. Thus, the aggregate size is an influential control factor for the tensile and flexural strength, while the paste content is more influential for compressive strength [70]. A greater thickness of cement paste has been found to increase the strength properties but reduce the permeability and porosity of PC pavements. Water-reducing admixtures are added to increase the workability and retardants are used to increase the setting time of PC. The structural integrity of PC has been found to be dependent on the base and subbase conditions [67]. An illustration of a typical PC pavement is shown in Figure 16. The major advantages of this pavement type are the reduction in stormwater volume, impervious cover, peak flow rate, heat island effect and noise and the increase in groundwater recharge, pollutant removal and skid resistance [53].

The design can be optimised to handle heavy loading, but PC tends to deteriorate under high turning loads. A major indicator of the PC pavements’ performance is their permeability rate, which decreases over time due to clogging [64]. For this reason, PC pavements need commercial vacuuming, approximately twice a year. While snow ploughing is permissible, the use of de-icing materials and sands can cause spalling of the concrete and should not be used, especially in new applications. While PC is generally more expensive than traditional concrete, cost savings can be achieved if storm drainage costs are included in the costs. Maintenance costs for PC pavements are lower than those for traditional pavements due to the absence of catch basins, pipes, ditches, etc. While commonly used in parking areas, it has also been used in sidewalks, trails, pathways, curbs, gutters, overlays, slope stabilisation, pool deck areas, erosion protection for bridge abutments, residential driveways, residential or low-volume streets, ditch linings, highway shoulders, drainage layers, etc. [53]. The compressive strength, water penetration, abrasion resistance, and freezing and thawing durability of PC has been found to increase with the use of smaller-sized aggregates and by adding silica fume (SF) and superplasticizer (SP) [69]. Typical PC pavements with a porosity between 15 and 30% have a compressive strength of 7 to 25 MPa [71]. Reducing the aggregate-to-binder ratio can increase the compressive strength to over 20 MPa, while supplementary cementitious materials (SCMs), such as SFtab and fly ash (FA), polymer modification of the matrix, etc., have been added to PC to achieve compressive strengths exceeding 40 MPa [71]. Li et al. explored the use of reactive power concrete (RPC) to develop a high strength pervious concrete (HSPC) pavement which showed a peak 7-day compressive strength of 61.37 MPa [72]. Other additives that have been used to improve the mechanical properties of PC include rice husk ash (RHA) and glass, steel and polyphenylene sulphide (PPS) fibres [68]. Recycled concrete aggregate (RCA) has been used to replace Natural Aggregates (NA) and was found to improve the hydrological properties but reduce the mechanical behaviour of PC pavements [73,74].

4.9.3. Permeable Interlocking Concrete Pavement (PICP)

PICPs are made of interlocked concrete units with stone-filled, permeable joints between them to facilitate drainage. The gaps are present over 5% to 15% of the surface area and contain small-sized aggregates that maintain high permeability, store water, reduce runoff volume and improve water quality. The advantages of using this system are that it is visually appealing, requires less maintenance, is easily repaired and can withstand high vehicle loads. PICP has been seen to remain stable in freezing and thawing climates, with no heaving occurring because the base, subbase and subgrade usually drain before the freeze can occur, the air in the aggregate voids provides an insulating effect, the ground provides heat that delays freezing and the presence of spaces in aggregate voids allows water to expand while within the base or subbase. A typical PICP design is shown in Figure 17, while different types of paving units are illustrated in Figure 18 [53].

Other advantages of the PICP are durability, high strength, traffic calming and high solar reflectance and resistance to degradation by de-icing materials. Thus, they can be used at primary parking areas, overflow parking areas, sidewalks, pathways, drives, and aisles. Their use for road shoulders, rest areas, access drives, ring roads, areas for loading and frequent truck traffic is limited to the allowable design load repetition [53].

4.9.4. Grid Pavement Systems

These are concrete or plastic open-celled modular grids filled with turf and small uniformly graded gravel. The grid is laid over a dense-graded aggregate base and is used for emergency drives, parking areas and low-use drive lanes. They are suitable for light and infrequent vehicular loading applications. Open-graded aggregates and bases can also be used for higher water storage and seepage. Concrete grids [53] can be latticed or castellated as shown in Figure 19.

Plastic grid pavement units are placed on an aggregate base. Grass-filled grid pavements are used for fire lanes, temporary events, maintenance access roads, and overflow parking, while aggregate-filled grid pavements are suitable for areas where grass may not survive or where quick installation is required. Both types are illustrated in Figure 20.

This pavement type is not suitable for roads with consistent heavy loads, high-speed and high-volume traffic, high pollution potential, sediment deposition or organic material accumulation. Locations with expected high turning motions in the same spot are also not recommended for this design. Clogging with sediments can be a problem with both types of grid pavements and the cracking of concrete can be avoided by using reinforced grids or smaller sized units [53].

4.9.5. Comparison of PA, PC, PICP and Grid Pavements

PA, PC, PICP and grid pavements are all recommended for overflow parking areas, sidewalks/pathways, and drives/aisles. For use in primary parking areas with heavy traffic, PC and PICP can be used without any changes, PA pavements require high-durability mixes with base and subbase designs, grass-filled grid pavements have limited use and aggregate-filled grid pavements with well-designed base thicknesses are recommended for areas with heavy use. Use in roads and highways is limited for all the four mentioned permeable pavement types, with PICPs being used for shoulders, rest areas, etc., up to the design load repetition limitations and grid pavements being used for median crossovers, shoulder, maintenance access roads, parking, rest areas, etc., with base depth designed according to the load. For access drives and ring roads, PC pavement is allowed to be used, while the use of PICPs is subject to load-repetition limitations; grid pavements require designed base thickness and PA pavements are not recommended for heavy traffic. PA and grid pavements are not suitable for loading areas and PICPs and PC pavements can have limited use only. Truck traffic is also limited for all four pavement types.

PC pavements have been found to have a higher infiltration rate compared to PICPs and concrete grid pavements. The volume reduction in stormwater, ground water recharge of soil and peak discharge rate control is high for PA pavements, PC pavements and PICPs, if the soil permeability is good, but all the performance metrics are low or medium for grid pavements. Alam et al. [55] found that PC pavements showed a reduction in runoff volume and peak flow as compared to PICPs. However, the detention capacity of PICPs was found to be higher than PA pavements by Madrazo-Uribeetxebarria et al. [75]. The stormwater quality functions like the removal of total suspended solids, total nitrogen, total phosphorus, metals and temperature control are comparable for all four pavement types. However, PA’s require vacuuming to minimise sediments, soils and other materials from flowing into the pavement, while PC pavements, PICPs and grid pavements require pre-treatment of runoff.

Hu et al. [76] found PC pavements to be superior in performance for flood mitigation compared to PA pavements and PICPs. The regular maintenance of permeable pavements is important to prevent clogging, which can reduce the performance of permeable pavements from 60 to 92%. In a study on flood risk to urban road users, Qiao et al. [77] found the cost of retrofitting a road by creating permeable pavements is higher than the total risk to the road, rendering it uneconomic, but they still found it beneficial for places experiencing repeated flooding.

The Counter Creek Flood Alleviation Scheme for the London Borough of Hammersmith and Fulham designed by Thames Water is an excellent example of the use of permeable pavements. In this design, bio-retention areas or green SuDS were not added due to lack of space in the presence of parking bays and to preserve parking space. Control of runoff volume through infiltration was also not considered suitable due to the proximity to property basements. Block permeable pavements were selected as the suitable SuDS option, as these could be constructed in modules and retrofitted on the streets. A geocellular subbase replaced the permeable unbound subbase and reduced the volume required which, in turn, reduced the depth of excavation and risk of clashing with underground utilities [78].

The water from the carriageway and the footway drained into the permeable parking bays present on both sides of the road. As shown in Figure 21, a capping layer of clause 803 type 1 material is installed first. On top of the capping layer, a 300 mm geocellular subbase material is installed, which is then covered with a 220 mm Pervious Concrete (PC) base, 50 mm bedding and block pavers on the top of the pavement. An impermeable membrane restricts infiltration into the ground and a geotextile laid above the geocellular units captures silts and oils. The terminal manholes were connected to the sewer system and raingardens, and tree pits were integrated into the design at strategic spots without compromising any space for the parking bays. As shown in Figure 22, the hydraulic system of the pavements included the geocellular units being connected to the subbase storage on both upstream and downstream sides, which include gravel and rockwool material to provide irrigation to the vegetation by capillary action [78]. Other examples of the successful use of permeable pavements in public spaces is shown in Figure 23.

5. Analysis and Key Points

The primary findings and insights from this literature review study can be organised into the following main points:

• The pre-flooding mitigation techniques include preventing saturation and increasing the saturation time by changing the pavement design, effective maintenance, permeability (decrease for flexible pavement surface layer and increase for permeable pavements or for a flexible pavement drainage layer) and cracking potential by altering the use of pavement materials, for example, changing the viscosity, void ratio, gradation, particle size, shape and density, or using asphalt mixes with low deterioration of tensile strength with time. The other method used to achieve mitigation is to use innovative materials that are resilient to damage at high levels of saturation. This is identified as a topic for further research.
  These finding also correspond to Nivedya et al.’s [79] findings that a higher value of the diameter of base material corresponding to 60% passing in the sieve analysis corresponds to a higher hydraulic conductivity of the base and, hence, better drainage [25] have also found the base and subgrade materials that influence the change in vertical strain and are thus relevant for assessing flooded pavement performance with regard to rutting. Tarefdar and Ahmad [80] also found a relationship between the permeability and tensile strength ratio. The relationship between the void ratio and tensile strength of asphalt mixtures has also been established by Hong et al. [81]. Thus, the findings of the study are compatible with other research in this field and a further study establishing models of the relationships between the various parameters mentioned is another topic identified for further research.
• The results of an experimental study on three innovative products, Erosion Control Blanket, Turf Reinforcement Mat paired with Fibre Reinforced Matrix and flexible concrete geogrid mat, were studied, and it was concluded that while all three provided partial protection to embankments against an overtopping flow, the flexible geogrid, though the most expensive, showed better performance as compared to the other two options. However, even though the flexible geogrid was cheaper compared to the fully grouted riprap, its performance is poorer and only provides partial protection in comparison. This is a product identified for further development. The findings are corroborated by studies conducted by Summer et al. [82].
• Pre-construction design changes or post-construction mitigation measures can save roads from damage when exposed to flooding hazards. As with most interventions, the stage when application is easiest is the design stage, because post-construction mitigation is restricted by what has already been constructed. This is in alignment with Alam et al. [83]’s conclusion that long-term inundation should be included as a design parameter in the road design of flood-prone areas. Chan and Wang [84] conducted case studies to show pavement structural design optimisation with the inclusion of flooding considerations. For example, they found that increasing roadbed height in the design stage reduced the need to increase the structural capacity of the road for flooding resilience.
• Post-flooding studies demonstrated that even during post-flooding rehabilitation, strengthening the existing subgrade or materials is the primary measure of rehabilitation, and even the complete reconstruction would involve a flooding-resilient design. This highlights the importance of proper design. The superior performance of cement-treated bases is highlighted. Funding and road priority were found to be important determinants of pavement rehabilitation decisions in post-flooding scenarios.

Different SuDS measures were compared, and it can be concluded that the application of SuDS is site-specific and complex. Several tools are available to support design decisions regarding choice of SuDS for any location [85,86].

An important point to consider is that many SuDS measures are designed for small and frequent events, while ponds and wetlands have the potential to provide flood attenuation for all events, including 10-, 30-, 100- and 200-year events. Permeable pavements are a promising area for research. While previously used on low-traffic roads and found unsuitable for heavy turning loads, research and changes in their design are being conducted to make them suitable for high-traffic roads.

Research shows that changing the mix, percentage and application methods of commonly used pavement materials can obtain better results in terms of flood resilience. This calls for research by construction agencies and the public highway authority to determine changes in conventional design and to plan for the frequent flooding scenarios which are becoming increasingly common across the world. Similarities in post-flooding rehabilitation design and pre-flooding design measures, for example, providing suitable drainage, shows the importance of including flooding as a criterion in the earliest stages of road design. The increasing use of SuDS provides an effective drainage strategy in urban spaces. However, these needs to be coordinated with higher capacity measures to plan for higher flows, which should in no scenario overwhelm the system.

The aim of this study was to study the various flooding mitigation and adaptation measures available to planners and designers. The study can be used as a comprehensive guide to the different avenues available to deliver a robust design. The opportunities for further research identified in this study include the following: 1. The use of innovative materials that are resilient to damage at high levels of saturation. These must also be sustainable and their capacity to resist water damage should be determined and compared to conventional materials to determine the extent of resilience added by conducting tests in turbulent flows and high saturation times. 2. Innovative products for embankment protection are available and can be improved to provide options with more sustainable materials, which should, however, deliver on performance when compared to traditional concrete-based or mechanical options. 3. Innovations in SuDS applications and integrating these with flood management for different flooding scenarios.

The limitations of this research include the measurement of flooding damage to road pavements through quantitative methods. Several methodologies for measuring the structural damage caused to flexible pavements exist (machine learning, mechanistic-empirical, finite-element analysis, fragility analysis, multiple hazard analysis, AASHTO’s structural number approach, etc.) and these can be compared to determine their effectiveness and suitability. This topic was outside the scope of this study but is recommended as a promising and essential topic for further research.

6. Conclusions

The excessive economic loss due to frequent and catastrophic flooding make a strong case for the research, planning and implementation of the climate adaptation and mitigation of road and transport networks.

The objective of the research was to identify structural mitigation and adaptation strategies and determine when these strategies can be implemented to reduce the effect of flooding events. This objective was met, and the study provides a comprehensive review of various mitigation and adaptation strategies available to engineers and road designers to plan and design for the flooding resilience of roads. The concepts of the study were changes in the design of roads and the associated transport infrastructure, structural changes in existing flood-affected or flood-prone roads and the context of the study was changes in construction materials and methods to achieve the concepts. The main strategies identified for this include altering the use of existing materials, the use of innovative materials, designing roads and ancillary structures considering flooding impact, the importance of pre-flooding mitigation to reduce damage and the need for post-flood rehabilitation and the various opportunities provided by SuDS solutions to improve the drainage capabilities of urban spaces, increase resilience and build capacity and redundancy into the system with respect to flooding events. The primary questions of structural mitigation, adaptation strategies and the flooding resilience of roads as well as the secondary question of increasing the flooding resilience were addressed and discussed. These strategies were compared to gain insight into the topic and identify viable research problems for further work.