Sustainable stormwater management in Yinchuan New Town

The configuration and design objectives for water and solid waste management in the Yinchuan New Town project were to:

• Determine a general water balance for the project area

• Develop guidelines and a design for the stormwater drainage in the area

• Design a wastewater system integrated with the solid waste system

• Design a wastewater treatment plant including nutrient reduction

• Prepare a proposal for potable water use reduction by separating black and grey waters

The water balance included an estimate of water need for the area, including evaporation from Mingcui Lake, and predicted water use in Yinchuan New Town, as well as estimates of water availability from different sources like water reuse and abstraction from existing irrigation channels.

The objective of this paper is to present the methodology applied in order to achieve SuDS in the project area and the results obtained.

Yinchuan is the capital of Ningxia Hui Autonomous Region, China (People's Republic of China (PRC)). It has a typical, temperate continental monsoon climate, with little rain and high evaporation rates. Given its proximity to the Yellow River and a developed channel system, the area has relatively good surface- and ground-water conditions. The surrounding region, however, is very short of water.

To achieve sustainable urban development, an approach integrating urban planning, transport, energy, and solid waste and water resource management is necessary, with the latter playing a key role. The design of these components was an integral part of the urban planning and design process.

The approach to achieving sustainable urban development was based on the SimbioCity methodology (Ranhagen et al. 2007). The principles of SimbioCity incorporate a multi-disciplinary mode of operation that enables combined problems to be solved successfully, taking into account social, institutional, environmental, technical, economic and spatial aspects (Ranhagen et al. 2012). A basic tool in the methodology is the eco-cycle model, where reduced resource use and pollutant loads are sought by creating cycles where materials and water are recycled to the largest possible extent, or used in energy generation. Resource use is reduced by introducing efficient techniques, and taking advantage of the synergies created by integrating the water, waste and energy cycles. An example of this is the generation of biogas (energy cycle) from organic waste (waste cycle) and wastewater sludge (water cycle). Similar approaches have been proposed by others (Novotny 2011).

The relevant standards for urban drainage in China are GB 50318-2012 Code of Urban Drainage Engineering Planning, GB 50318-2000 Code of Urban Wastewater Engineering Planning, and GB 50014-2006 Code for the Design of Outdoor Wastewater Engineering.

Average precipitation around Yinchuan is 189 mm/year, with average water surface (potential) evaporation of 1,197 mm/year. Precipitation occurs from June to August, inclusive, and the twenty first century has seen some flooding in the existing urban areas, such as that in the central areas of the city in July 2012.

The relevant Chinese water quality standard is GB 3838-2002. Water quality in Mingcui Lake is below grade III, where grade I is the best and grade V the worst. Water quality in the lake is moderate, due to agriculture drainage and the discharge of untreated wastewater from rural villages in the area.

The Chinese standard states that stormwater pipes must be designed for a storm with a return period up to 3 years, based on the characteristics of the catchment, and its topography and climate conditions, and from 3 to 5 years for important roads (GB 50014-2006).

With respect to stormwater management and flood prevention, alternatives to a conventional stormwater sewer system were needed because flood events have been relatively frequent, even though the project area has low average annual precipitation, because of the very concentrated rainy period. Moreover, the stormwater management system had to avoid polluting existing water bodies during rainfall events. In order to comply with these demands, a new approach was applied based on stormwater detention and treatment.

The method and solution applied had two objectives:

• To avoid flooding important roads, as well as residential and commercial areas, for a 100-year return period flood, although the costs for achieving this must be reasonable.

• To introduce stormwater treatment to maintain or improve water quality in Mingcui Lake.

To achieve these goals, a SuDS approach was applied during configuration and design of Yinchuan New Town's infrastructure.

The approach proposed in the CIRIA SuDS Manual (Woods-Ballard et al. 2007) was used as the design basis. According to CIRIA, stormwater should be managed in small, cost-effective landscape features in small sub-catchments, rather than being conveyed to and managed in large systems at the bottom of drainage areas (end of pipe solutions). In addition, solutions used in Malmö, Sweden (Stahre 2006, 2008) as well as Portland, USA (Liptan 2011) were applied, with some adaptation to local conditions in Yinchuan.

where:

• V_detention: detention volume (m³)

• t_r: rain duration (min)

• Q_in: runoff flow, calculated as

  

  2

• Δ_k: Area k within the plot or channel watershed (ha)

• i_k: runoff coefficient for Area k (dimensionless)

• Q_out: given outflow from the channel or detention structure at plot level (l/s)

• L rainfall intensity (l/s/ha)

where T is the return period (years). The return period for the plot level detention structures was 10 years, that for the channels 100.

Detention volumes were calculated using formulas (1) to (3) above, for different values of rain duration (tr). The maximum of these detention volumes is thus that required for the catchment.

Stormwater discharge was identified as one of the major polluting emissions in urban areas in Sweden (Alm et al. 2010). StormTac is used to estimate the mean pollutant concentration based on the specific land use(s) in the catchments. It uses ‘standard concentrations' per land use type, to calculate pollution loads at the catchment discharge point.

In Yinchuan New Town, the treatment facilities considered will comprise the channels or ditches, and the treatment ponds upstream of the discharge point. There is also some treatment at plot level; however, since there will be some stormwater detention facilities at this level, so it is not possible to establish which load reduction is applicable. Because of this, the treatment at plot level was excluded from the load reduction estimates – i.e., no load reduction is considered in the estimates at plot level – yielding a conservative approach.

where:

• A_p: Pond or constructed wetland area (m²)

• A_red: Reduced catchment area, calculated as ⁠, that is the area multiplied by its runoff coefficient

• A_k: Area k within the catchment (ha)

• C_k: runoff coefficient for Area k

• k₁, k₂: regression coefficients for the individual pollutant

• f_Cin : inlet concentration factor

• f_veg : vegetation factor

• f_bypass : bypass factor

• f_Vd : detention volume factor

• f_Cirr : irreducible concentration factor

• f_temp: temperature factor.

• f_shape : shape factor.

The runoff and regression coefficients, and all factors above, are dimensionless.

The minimum outlet concentration, the ‘irreducible concentration’ (C_irr), arises from influent content and internal processes in ponds and wetlands (decomposition of plants, seepage of anaerobic liquors from the bottom, exchange with sediment, sediment disturbance by benthic animals, etc.), which limit the extent to which pollutants can be removed. In StormTac, the reduction efficiency is adjusted so that nothing less than minimum concentrations are obtained at the outlet. The irreducible concentrations were estimated from effluent concentration data from facilities in Europe. Some examples of these are C_irr (P_tot) 20–30 μg/l, C_irr (Cu) 6–7 μg/l, C_irr (Zn) 14–25 μg/l and C_irr (SS) 5–10 mg/l (Larm 2014).

where:

• A_ditch: channel or ditch area (m²)

• A_red: reduced catchment area (ha)

• k₁, k₂: Regression coefficients for the individual pollutants, dimensionless.

The regression coefficients and factors used in the equations above have been obtained from existing ponds, swales and other stormwater detention and treatment facilities, mainly in Europe and North America (Larm 2000; Larm & Hallberg 2008). The annual precipitation at these overseas sites is higher than that at Yinchuan but, as precipitation in Yinchuan occurs mainly between June and August, rainfall intensity is similar to that at the sites used as a base for StormTac. Other factors which may influence removal efficiency, apart from pond size and form, are the vegetation and soil types used in the treatment ponds, which are basically the same or very similar in Yinchuan and the sites where the factors and coefficients were deduced.

The recommended surface area for a channel or ditch to achieve a satisfactory pollutant removal rate is at least 2% of the reduced catchment area (Larm & Banach 2011).

The project area was divided into sub-catchments on the basis of topographic levels, and road and green areas, as well as stormwater infrastructure planned for Yinchuan New Town. The topographic levels, and the proposed infrastructure and landscape, were all based on the site's existing natural characteristics. The main roads usually established the watershed limits, while the green corridors and channels constituted the flooding areas. This required an iterative process involving different professionals – e.g., hydrologists, road engineers, landscape architects and city planners.

The required detention volumes at plot level for a 10-year return period and channel level for a 100-year period for the sub-catchments in Yinchuan New Town are presented in Table 2.

The length to width ratio of a detention pond should be 3:1 or greater, and vegetation in the shallow zones should cover 25 to 45% of the pond's area (Larm & Banach 2011). The anticipated pollutant concentrations in stormwater runoff were calculated on the basis of the proposed land use(s), see Table 4:

As shown in Table 4, the anticipated – i.e., calculated – concentrations of Tot P, Tot N, oil index, Chemical Oxygen Demand (COD) and Biological Oxygen Demand (BOD), in the untreated run-off waters all exceed the permissible Grade III water quality levels. The values given in Table 4 were used to determine the pollutant load reduction required from the stormwater detention and treatment facilities. The proportional pollutant load reduction given by channel areas of between 2 and 6% of the reduced sub-catchment and pond areas – Table 3 – are shown in Table 5:

These pollutant load reductions, from those anticipated in the storm run-off waters, would lead to the post-treatment pollutant concentrations shown in Table 6. All of the latter now meet the Grade III water quality requirements.

Stormwater detention – i.e., flood control – was achieved for residential and commercial areas, and roads using a combination of detention at plot and building level, and planned green corridors and spaces. Between them, these provide detention volumes sufficient to maintain the water level below building thresholds for a 100-year return period flood event.

Stormwater treatment is achieved initially at plot level, then at channel level and finally at pond level, before discharge to the natural wetland.

An important factor in the introduction of SuDS in Yinchuan New Town has been its cost-effectiveness, due to the use of multifunctional areas. The green structure was used in an effective way to enable both flood management and stormwater run-off treatment.

High levels of cooperation between different professions – e.g., hydrologists, road engineers, landscape architects and city planners – were fundamental to achieving the necessary agreement on subjects such as topographic planning settings and land use. This, among other factors, helped make a successful stormwater detention and treatment strategy and application possible.

It is recommended that monitoring points are set up to obtain real water quality values. Since the pollutant load calculations are based on development and stormwater detention facilities in Europe and North America, measurements of the actual loads and load reductions should be made. They will form the basis of a follow-up strategy, allowing adaptive measures to be taken to ensure a successful outcome, when the new development and proposed stormwater facilities are complete and functioning.

The authors would like to thank Sweco personnel in Malmö and Beijing, in particular Peter Krigström, Project Manager and Main Planning Designer, as well as personnel from Minsheng Real Estate in Yinchuan.