Abstract:Significant iron release from cast iron pipes in water distribution systems(WDSs),which usually occurs during the source water switch period,is a great concern of water utilities because of the potential occurrence of "red water" and customer complaints.This study developed a new method which combined in-situ water stagnation experiments with mathematical models and numerical simulations to predict the iron release caused by source water switch.In-situ water stagnation experiments were conducted to determine the total iron accumulation in nine cast iron pipes in-service in Beijing when switching the local water to tre ated Danjiangkou Reservior water.Results showe d that the difference in the concentration increment of total iron in 24 hr(ΔCI TI,24 ),i.e.short-term iron release,caused by source water switch was mainly dependent on the difference in the key quality parameters(pH,hardness,nitrate,Larson Ratio and dissolved oxygen(DO)) between the two source waters.The iron release rate(RFe) after switch,i.e.long-term iron release,was closely related to the pipe properties as well as the DO and total residual chlorine(TRC) concentrations.Mathematical models of ΔCI TI,24 and RFe were developed to quantitatively reveal the relationship between iron release and the key quality parameters.The R Fe model could successfully combine with EPANET-MSX,a numerical simulator of water quality for WDSs to extend the iron release modeling from pipe level to network level.The new method is applicable to predicting iron release during source water switch,thus facilitating water utilities to take preventive actions to avoid "red water".

Modeling Iron Release from Cast Iron Pipes in an Urban Water Distribution System Caused by Source Water Switch