Reducing customer complaints of discoloured water is a major challenge for the water industry. The ability to predict the spatial probability and severity of discolouration events in distribution systems could lead to the implementation and optimisation of proactive operational and maintenance strategies to minimise discolouration. This paper explores the transfer of a predictive, semi-empirical model developed to describe iron dominated discolouration problems in the UK to an Australian system, where discolouration is primarily associated with clay type material. The paper presents the application of the model to a large diameter water main that forms part of the Melbourne system within a single source water quality zone. The model is based on cohesive transport theory and for this application includes the concept of a ‘self-cleaning threshold’, defined as a shear stress that the pipe experiences regularly, due to normal daily demand, that prohibits the accumulation of sufficient material within the pipe and hence poses no discolouration risk.

Results presented here show that the model can be calibrated to simulate the turbidity response measured in real systems due to changes in hydraulic conditions, for clay driven discolouration problems. These changes in hydraulic condition resulted from various ‘natural’ events which occurred during the available data period. Through such simulation the semi-empirical model parameters were evaluated. These simulations demonstrate the capabilities of the model, which could be applied to operational and maintenance practice. For example, to identify and prioritise network cleaning operations to minimise discolouration risk, at pipe level, through simulation of discolouration responses to various possible events (burst, fire fighting etc) and ranking the resulting predicted discolouration, or through prior simulation to identify the cleaning necessary to mitigate discolouration risk of planned operations such as valve movements.

You do not currently have access to this content.