Major challenges of water use in the drilling and fracturing process in shale gas production are large volumes required in a short-period of time and the nonsteady nature of wastewater treatment. A new mixed-integer linear programming (MILP) model for optimizing capital investment decisions for water use for shale gas production through a discrete-time representation of the State-Task Network is presented. The objective is to minimize the capital cost of impoundment, piping, and treatment facility, and operating cost including freshwater, pumping, and treatment. The goal is to determine the location and capacity of impoundment, the type of piping, treatment facility locations and removal capability, freshwater sources, as well as the frac schedule. In addition, the impact of several factors such as limiting truck hauling and increasing flowback volume on the solution is examined. A case study is optimized to illustrate the application of the proposed formulation. (c) 2015 American Institute of Chemical Engineers AIChE J, 61: 1770-1782, 2015