A cooler network is a significant part in a heat exchanger network, which is also vital to energy consumption and economic performance of the entire chemical manufacturing process. Currently, the work on the optimization of cooler networks chiefly has attached to optimization of circulating cooling water systems. However, few researchers have taken different cooling types with series and parallel configuration into consideration in the process of optimization on cooler networks. On the basis of lectotype among the dry type air cooler, spray type air cooler on behalf of wet type air coolers, and circulating cooling water system, a superstructure-based optimization model with minimizing total annual cost as the objective function is established in this paper, formulated through mixed-integer nonlinear programming. Coolers with series and parallel configuration are allowed to reduce energy and water consumption and improve economic performance. Corresponding case studies show that the model can give a more economic efficient solution of the cooler network.