AWWA JAW20485

Objectives of this study were to evaluate the success of the Seattle, Washington, program for reducing the corrosion rate of residential copper plumbing, to determine how instantaneous copper corrosion rates are affected by short-term variations in selected water quality parameters, and to develop a statistical model of the copper corrosion process as a function of selected water quality parameters. Results indicate that Seattle's program of corrosion control through pH adjustment and alkalinity addition has been successful in reducing the overall corrosion rate on aged copper surfaces. Variation in corrosion rates can be successfully evaluated as a function of short-term changes in water quality conditions. Regression analysis of these variations can produce linear models to predict corrosion rates. The oxide film layer on aged copper surfaces, as compared with the surfaces of new copper plumbing, reduces the corrosion rate by about fifty percent. Free chlorine residual was the most important parameter tested in predicting corrosion rates. pH as a predictor was of little importance for corrosion on relatively clean copper surfaces, but was of value on surfaces with a heavy oxide layer. Conductivity was of little importance as a predictor in low alkalinity water. Includes 9 references, tables, figures.