Approaching Water Sensitive Cities with Adaptive Rainwater Diversion

Total water cycle management (TWCM) discussion mostly limits the role of residential rainwater harvesting to providing an alternate water supply to a fraction of fit-for-purpose end uses. However, with operational improvements, greater outcomes can be achieved. By increasing the portion of roof area connected and developing adaptive rainwater diversion (ARD), reliable stormwater management outcomes can also be achieved. ARD controls tank drawdown by adapting to changes in dwelling consumption and rainfall, thus allowing the available storage to mimic the pre-urbanised catchment storage recovery. The ARD approach has the basis that mains water savings can be achieved in two ways: 1) Rainwater supply where the rainwater harvest is used directly to reduce mains consumption of that dwelling; and 2) Rainwater diversion where rainwater is diverted from the dwelling. This does not directly reduce mains consumption of the dwelling but produces a water resource that is used by others to reduce mains consumption. In this way, total rainwater yield and mains water savings is the sum of rainwater supply and diversion. This research investigates rainwater supply, rainwater diversion, runoff volume and runoff flow frequency for South East Queensland. Results show, the average sized detached dwelling when fitted with a 5 kL tank and ARD system is compliant with the mandated water saving targets and the Queensland Best Practices Environmental Management Guidelines for stormwater flow frequency management. It is recommended that rainwater is diverted into the existing stormwater system where reuse facilities exist. Otherwise, discharging into the sewer, has the potential to reduce sewer fouling and increase the substitution of mains supply with treated effluent. This improves sewerage reticulation by adding a secondary purpose and, by using existing infrastructure, removes many barriers for retrofitting TWCM and water sensitive urban design (WSUD). Also, as ARD brings adaptive and multifunctional infrastructure into our urban design, we begin to develop water sensitive cities. The outcomes of this research are most promising to established and future planned high density residential suburbia, where TWCM policy and WSUD is chiefly needed. INTRODUCTION Water sensitive cities must have adaptive and multifunctional infrastructure, among many other capacities. Achieving this, is the ultimate goal of water sensitive urban design (WSUD). WSUD focuses on overturning our culture of misuse and waste of water resources by recognising the lifesustaining qualities of water in the design of our urban environment (Water by Design 2009). WSUD applies to all elements of our built environment from the urban core to rural living and extends into rural production. Incorporated are aspects of water conservation in addition to the management of stormwater flow frequency, waterway stability and stormwater quality. Rainwater harvesting at either the allotment or cluster scale contributes to WSUD in all elements of the built environment in the following ways: • Contributes to water conservation With first flush devices fitted, rainwater is fit-forpurpose to substitute mains water for all non-potable end uses (Qld DIP 2007), which can constitute more than 80% of residential consumption (Willis et al. 2011); • Contributes to stormwater quality management Storing water or running water through a rainwater tank can reduce pollutant loading; and • Contributes to waterway stability management The airspace above the obvert of the overflow outlet provides storage that can fill during excessive rain events. This occurs when the inflow rate exceeds the outflow rate. The result is, attenuation of the peak discharge. Rainwater harvesting is however, unreliable at managing stormwater flow frequency. Flow frequency management is important for urban catchments. The introduction of highly efficient stormwater delivery systems and smooth impervious areas, in the way of roads, rooves and driveways, has resulted in runoff occurring from smaller and more frequent rain events. These repeated impacts are degrading our urban creeks and streams in a process referred to as the urban stream syndrome (Meyer et al. 2005; Walsh et al. 2005a; Walsh et al. 2005b). Reliable flow frequency management is achieved when adequate rainwater storage capacity is available prior to a rain event. Usually, the daily consumption of rainwater is too low and rainwater storage is insufficient. Also, there are many reasons why rainwater consumption from any dwelling may decrease, such as reduced irrigation in winter or a period of vacancy due to work, holiday or a change in tenants. This is evident in the variability of rainwater yield reported by recent studies (Beal et al. 2010). With diminishing consumption and performance variability, the rainwater harvesting contribution to flow frequency management is not assured and should be considered in WSUD cautiously. Periodic diversion of rainwater can overcome this limitation. This diversion aims to control the tank drawdown rate, or critical period, and provide sufficient storage for reliable flow frequency management. Ideally, rainwater should be diverted to a system that can treat and return the water to municipal supplies. In this case, the diversion should be slow enough not to overwhelm the receiving system. This is a desirable outcome, as the diversion could occur under gravity flow, thereby allowing rainwater to be used without the high energy costs associated with small decentralised pressure pumps. (Brodie 2009) introduces the concept of trickle diversion from communal rainwater tanks to achieve flood discharge reduction, stormwater flow moderation and to overcome the management and ownership constraints of individual residential tanks. In this concept, the diverted rainwater is returned to the municipal water supply using a stormwater harvesting scheme. In many situations, the barriers for retrofitting communal rainwater tanks and stormwater harvesting schemes are too great. Thus, an alternative is needed that promotes multifunction of existing infrastructure at the household level. With treatment at the household level, rainwater diversion would need to adapt to changes in rainfall and consumption to ensure optimum performance. Otherwise, when rainwater consumption returns to normal or when rainfall reduces the system will drain too quickly, and the reliability of rainwater supply will be compromised. To be adaptive, the diversion would be controlled by an electronic device which monitors the rainwater tank. This device would divert rainwater, via a trickle outlet, under the right conditions. It is therefore hypothesised that introducing a system of adaptive rainwater diversion (ARD) will extend the role of rainwater harvesting to runoff flow frequency management, while simultaneously achieving water conservation and providing adaptive and multifunctional infrastructure, as needed to develop water sensitive cities. The objectives of this study are: • Assess the dual performance (water conservation and runoff frequency management) of rainwater harvesting, with and without ARD, and while configured in accordance with the Queensland rainwater harvesting operating policy (Qld DIP 2007) for detached dwelling in South East Queensland (SEQ); • Assess the dual performance of rainwater harvesting, with and without ARD, and while optimised for dual operation; • Assess the difference in rainwater supplied to the dwelling resulting from introducing ARD; • Assess the response of the ARD system to changes in consumption and rainfall. METHODOLOGY To achieve the study objectives many scenarios were simulated using MUSIC and Excel. Table 1 defines the parameter values and purpose of each scenarios. Table 1 Modelling scenarios Parameter Adopted values Comments Connected roof area ¤ 100 m 2 Mandate minimum for assessing mandate performance 150 m 2 Intermediate value for assessing intermediate performance 200 m 2 Practical maximum value for assessing ultimate performance Rainfall € Below long term average Reduced rainfall scenario for assessing Brisbane performance Equivalent to long term average Average rainfall scenario for assessing Caloundra performance Above long term average Elevated rainfall scenario for assessing Gold Coast performance Consumption ¥ 184 L/pc/d supplying all non-potable fittings μ 100% occupancy or 2.5 persons for assessing ultimate performance annual consumption from rainwater connected fittings 134 kL/dwelling/yr 184 L/pc/d supplying the minimum fittings § 100% occupancy or 2.5 persons for assessing mandate performance annual consumption from rainwater connected fittings 84 kL/dwelling/yr 92 L/pc/d supplying the minimum fittings § 50% occupancy or one full time occupant for assessing diminishing consumption annual consumption from rainwater connected fittings 42 kL/dwelling/yr 37 L/pc/d supplying the minimum fittings § 20 % occupancy or one intermittent occupant for assessing diminishing consumption annual consumption from rainwater connected fittings 16 kL/dwelling/yr 4 L/pc/d Vacant dwelling for assessing diminishing consumption annual consumption from rainwater connected fittings 2 kL/dwelling/yr Rainwater system 5 kL tank only Mandate minimum for assessing conventional rainwater harvesting performance 5 kL tank and ARD system For assessing the proposed adaptive rainwater diversion system ¤ Based on the total roof area of 215 m 2 from the conceptual design of a residential estate with 15 allotments/ha (Water by Design 2010). € Refer to Table 2 for rainfall data statistics. ¥ Daily per capita consumption derived from summer 2009/2010 survey of Gold Coast dual reticulated dwellings (Willis et al. 2011). This is equivalent to a total annual consumption of 168 kL/dwelling, for the average sized dwelling of 2.5 occupants (ABS 3236.