Greywater re-use has much potential as a water conservation measure although its uptake has been tempered somewhat by concerns over the potential risk to health, financial viability and the absence of formal legislation. The focus of these concerns varies according to the planned scale of re-use. At the domestic scale, research has shown that although greywater re-use is technically feasible and conceptually attractive to a proportion of homeowners, presently, the financial returns from water savings are too small to encourage wider uptake. It has been suggested that the financial viability of greywater re-use improves with an increase in scale.
This research reports on the performance of water re-use systems using a computer simulation model. Re-use system design is evaluated and re-use system performance predicted. The importance of estimating the water saving potential and water quality implications of a planned re-use project becomes more acute as the scale increases. In order to estimate water saving efficiency and evaluate system design, it is necessary to understand the patterns of water flow in a building. Yet, there is a limited availability of detailed appliance usage data for multi-occupancy building. The appliance usage of multi-occupancy residential buildings can be estimated from single-family water usage data.
Results show that increases in storage size give improved re-use system performance up to a point, after which increases in capacity result in reduced gains in performance. For a single family household of 4 occupants this point is around 250 litres and for multi-occupancy buildings it is around 1000 litres for all occupancies. Disinfection performance in lower occupancy buildings is compromised if the storage capacity is increased beyond these values. Simulating different levels of component reliability and frequency of maintenance revealed that performance was significantly reduced if systems were maintained at 3 month or even 1 month intervals, although, weekly maintenance checks maintained performance levels even for frequently failing components.