Urban drainage system modelling

From early computer programs written in the 1970s and 1980s, desktop software packages for urban runoff modelling have evolved over the past decades. An important line of development was the WASSP family of products, which from the early 1980s provided a computer implementation of the Wallingford Procedure. This led to the WALLRUS package and subsequently HydroWorks and InfoWorks. In the UK there is now a range of urban drainage modelling software in use, including software packages such as MIKE URBAN and SWMM. Two of the most widely used software packages are InfoWorks and WinDES, both of which have been developed in the UK and include support for UK runoff models and design standards.

Modern urban drainage software applications allow users to design, analyse and run scenario tests on large, complex systems of linked surface and subsurface drainage. They are capable of modelling the control features commonly found in urban drainage networks, including gullys, conduits, open channels, weirs of various types, orifices and penstocks. Some packages include optimisation methods to assist in sizing of pipes and setting channel capacities. There are also now capabilities to link the drainage network with two-dimensional (2D) depth-averaged flow models to simulate overland flood flow pathways.

An example of the type of complex drainage network that can be modelled is shown in Fig. 18.8, which is a pipe and gulley network for an area in the south-west of England.

It is typical of urban drainage data that despite the best efforts of the responsible operating authorities, there can still be uncertainty about the where some drains are,

Fig. 18.8 Schematic connectivity diagram for a modelled urban drainage system (data used with permission of Somerset Drainage Boards Consortium).

their levels, material condition and even which drains are connected. In this example, investigation centred on a major culverted drain denoted 'WR' of approximately 700 m length that has been linked with reported flooding incidents. In order to assess the capacity of the whole urban drainage system, all known highway gullys were included in the model. This enabled the model to be used to distinguish surface water flooding, where water ponds and flows over the surface as it is unable to enter the underground systems, from sewer and culvert flooding where water exits from overloaded sewers and culverts.

The hydrological inputs to the system were discretised into a series of sub-catchments. For urban sub-catchments, the 'new UK' runoff model was applied because it is suited to the mixed permeable and impermeable surfaces present in the study area, and models the change in runoff production as permeable areas become saturated during the storm event. In this case, areas drained by highway drainage and permeable surfaces were assessed as one layer of sub-catchments, while impermeable roof areas were assessed separately. In line with the methods adopted by some UK water companies, antecedent wetness was represented by the net antecedent precipitation index for the 30 days preceding an event (NAPI30), which was estimated using the 95 percentile NAPI30 value for summer and winter seasons. This was calculated using gauged rainfall from a local site for 1998 to 2007. Average evaporation values of 3 mm day-1 in summer and 1 mm day-1 in winter were used.

For rural areas, ReFH unit hydrograph models (Chapter 13) were used to generate inflows to the drainage system based on FEH design rainfall profiles. One of the major sources of uncertainty in small catchment runoff estimation can be the difficulty of establishing the true catchment area. In this study, a combination of flat topography and artificial land drainage made it difficult to determine the exact drained area at the upstream inflow to the 'WR' culvert system. Hence two possible inflow areas were modelled at this location, one assuming a 40 ha sub-catchment and one assuming a 190 ha sub-catchment.

Water surface profiles for the culvert at 'WR' are shown in Fig. 18.9 for the two possible upstream drained areas. Of particular note is a constriction from 600 mm to twin 400 mm diameter pipes downstream of a manhole at 229 m, where the culvert passes under a gas main. The twin pipes form a constriction in pipe full capacity, from 0.46 m3 s-1 to 0.31 m3 s-1. It can be seen from the modelling that this constriction would lead to surcharging and surface flooding for the 100-year return period (1 per cent AEP) storm event should the larger of the assumed upstream catchment areas be the true drained area.

In addition to allowing hydrological input scenarios to be tested, the drainage model in this case was used to assess the implications of differing levels of sediment build up within the WR culvert. Table 18.5 shows the predicted volumes of flood water produced during design storms of 2-100-year return periods with the culvert modelled either in its 'clean' state (with data taken from a survey following cleaning operations) or in one of two sedimentation scenarios. The impact of allowing sediment depths to build up would not necessarily be noticed for a 2-year event. However, at a 10-year return period and above, a substantial increase in flood risk was found.

With a system-wide network model in place, it is also possible to study the implications of changes to the connectivity of the system (Fig. 18.10). In this case, a number

m 0 47 69 96 140 196 229 267297 354 386 438 494 538 595 628 662 736

Fig. 18.9 Long section water surface profiles forthe culvert'WR'showing sensitivity under 1/100-year rain storm to assumed drained areas of 40 ha (top panel) and 190 ha (lower panel). Data used with permission of Somerset Drainage Boards Consortium.

m 0 47 69 96 140 196 229 267297 354 386 438 494 538 595 628 662 736

Fig. 18.9 Long section water surface profiles forthe culvert'WR'showing sensitivity under 1/100-year rain storm to assumed drained areas of 40 ha (top panel) and 190 ha (lower panel). Data used with permission of Somerset Drainage Boards Consortium.

of drainage areas were identified that had no known outfall to the main 'WR' culvert system and hence could be unable to discharge storm water effectively, leading to flood risk. These 'orphaned' drains were connected to the 'WR' culvert and the design events run for storm return periods as above (see Table 18.6). The culvert appears to have some available capacity to accept these additional flows without raising flood risk to properties during the 100-year event. The solutions tested would require major capital works and hence detailed feasibility studies, including flow survey. The drainage system model enables a set of hydrological 'design' inputs to be tested against various scenarios for management or development of the sewer network thus providing a cost-effective basis for planning and future investment.

Table 18.5 Total flood water volume over the whole modelled area for different scenarios about maintenance of the 'WR' culvert system

Return period in years Total flood volume (m3)

Clean (sediment Sediment 20% of Sediment 40% of depths from pipe/channel height pipe/channel height post-cleaning survey)

Clean (sediment Sediment 20% of Sediment 40% of depths from pipe/channel height pipe/channel height post-cleaning survey)

2

463

483

5I2

I0

I226

I237

2I74

30

I765

229I

4660

I00

399I

5472

9489

Fig. 18.10 Possible new surface water sewer (grey dotted line) to improve connectivity of 'orphaned' drains. Data used with permission of Somerset Drainage Boards Consortium.

18.7 Sustainable urban drainage systems (SUDS)

Historically, urban drainage has been based on surface drains and underground pipes, which are designed to allow water to flow away from the built-up area as quickly as possible to prevent flooding. There is then a risk that high-intensity rain storms may overwhelm the system, leading to flooding, either because the drains do not have a designed capacity to convey water at a high enough rate or because of degradation in the condition of the pipe network, possibly including sewer collapses or siltation. Of course the fact that the piped network is underground makes it all the more difficult

Table 18.6 Changes in total flood volumes over the whole modelled area for design events with two different drainage network connectivity scenarios

Return period (1 in x years)

Total flood volume (m3)

Network connectivity as-is

'Orphaned' areas connected

2

463

201

10

1226

464

30

1765

843

100

3991

2081

to inspect and maintain the infrastructure. Drainage systems often take foul water to waste water treatment works in a sewerage system separate from the surface runoff drains. In some cases, surface runoff is routed together with foul water through a combined main sewer. In both cases, heavy rainfall can exceed the capacity of the system causing discharge of the foul water and surface water runoff through combined sewer overflows (CSOs), leading to contaminated water entering receiving water-courses such as lakes and rivers. High sediment and solute loads can by-pass urban watercourses and be discharged directly into a downstream river, causing strong spatial and temporal variations that can affect ecological status and channel morphology (Old et al., 2006) In contrast to these conventional systems, 'SUDS' are urban drainage systems that follow sustainable development principles. Sustainable drainage offers benefits brought by mitigating the adverse impacts of urban development on storm runoff, particularly through:

• reduced runoff rates, hence reduced downstream flood risk;

• encouraging groundwater recharge in areas where urban drainage might otherwise cause rainfall and runoff to be exported from the catchment too quickly;

• improving water quality by holding back overflows from CSO spills and preventing direct discharge of high concentrations of contaminated water;

• providing amenity and habitat within in the urban environment.

The basic philosophy of SUDS is to reduce the impact of development and maintain as far as possible a natural runoff regime through a 'little and often' approach. The main components of SUDS are designed to manage localised runoff throughout the developed area, rather than conveying storm water artificially and then requiring large control structures or flood storage, as might be the case in more conventional drainage systems. A comprehensive guide to SUDS, including design principles, construction, maintenance and examples is the 'SUDS Manual' published by CIRIA, formerly known as the Construction Industry Research Association (Woods-Ballard et al., 2007). Typical SUDS components are:

• filter strips - wide, gently sloping vegetated areas adjacent to impermeable surfaces;

• swales - broad, shallow channels planted with grass or other vegetation to convey or store runoff and allow infiltration;

• infiltration basins - depressions similar to swales but designed for storage and infiltration rather than conveyance;

• detention basins - larger basins to provide runoff storage, usually dry though they may contain permanent ponds;

• wet ponds - basins designed to have permanent ponds of water to treat water quality, often providing wildlife or amenity value and additional runoff storage;

• constructed wetlands - ponds with shallow water and vegetation planted to enhance water quality and habitats;

• filter drains and perforation pipes - trenches filled with gravel or other permeable material typically placed along paved areas or across slopes;

• infiltration devices - temporary, localised storage of runoff to allow infiltration into the ground;

• pervious surfaces - paved surfaces that allow infiltration of rainwater into a permeable storage layer;

• green roofs - vegetated roofs planted over a drainage layer.

Fig. 18.11 shows a typical SUDS feature in an area of new housing development.

Achieving a sustainable urban drainage system requires planning, careful development and on-going management. There is a series of elements that make up the whole system and that should be considered during development planning. Source control refers to control of runoff at a very small spatial scale, through soakaways and filter strips or similar infiltration mechanisms, green roofs and permeable paving. At a slightly larger scale, site control refers to routing and attenuation of runoff from an

Fig. 18.11 Drainage ponds associated with new housing development.

area to larger soakaways or detention ponds. Larger still, runoff from a whole site or a group of developments may be routed into a wetland or balancing pond. In addition to these runoff control and attenuation strategies, reduction in runoff production can be achieved by good practice in maintenance, for example, clearing trash from car parks to prevent blockage of flow routes and rainwater harvesting.

By combining many small features in a large SUDs system, the aim is to prevent the system failing completely in the event of any one component not performing. The links in the system should include surface routing through natural swales and trenches, if possible, rather than artificial pipes. This promotes a highly distributed, attenuating drainage system that should be resilient if well maintained. A SUDS system therefore requires maintenance, such as vegetation cutting and removal of debris and sediment, to ensure long-term sustainability. The maintenance regime should be an integral part of the management plan for a development. In general, where sustainable development principles have been embraced, there is a shift from 'design' and 'periodic maintenance' of individual structures towards concepts of 'whole system' management, where the operation of the system is planned and managed over the entire life cycle of the development. This shift in thinking is enshrined in the UK in policies on planning and development appraisal. One practical example of this shift that is of relevance to urban hydrology is found in the revision of guidance on culvert design; the Culvert Design Guide, published in 1997 by CIRIA, has been updated and extended in the newer Culvert Design and Operation Guide (emphasis added) taking a somewhat wider environmental and whole life-cycle view.

In addition to their benefits for flood management, SUDS can also improve water quality through a number of processes. Many pollutants are bound to sediments, which can be trapped by SUDS features or removed by filtration through soil and using geo-textile layers. Heavy metals, solid wastes, organic material, hydrocarbon and pesticides can all be removed from runoff by a combination of filtration, sediment deposition, biodegradation and plant uptake. Again, maintenance is required to ensure sustainable effectiveness, e.g. by removal of plants and replanting to stop metals being returned to the water-course and removal of sediment to prevent flushing of contaminated solids in a subsequent flood event.

0 -1

Post a comment