As a high water consumption industry, water conservation is one of the challenges needed to be conquered in order to achieve the sustainable development for Chinese iron and steel industry. Based on the idea of Whole Process Pollution Control (WPPC), global optimization of water network in the steel industrial park was carried out in this work. Firstly, the configurational and operational characteristics of the water network in the typical steel industrial park were investigated, and a multi-scale modeling method was proposed to describe the water network. Wherein different scale water systems, including water-use/treatment unit, process-scale water network, park-scale water network, and their interactions were discussed. Secondly, a superstructure was built to represent all possible water network configurations. Thirdly, an optimization model tailored for the steel industrial park was set up with the aim of minimizing the comprehensive water-use cost. The water conservation and pollutants balance of different scale water systems, together with some limits on structure and operation of the water network, were described as constraints. Lastly, to validate the proposed model, global optimization of water network in a steel industrial park with 5 million tons of crude steel production capacity per year was studied by constructing and solving the mathematical model. The results of study cases showed that the case driven by WPCC strategy achieved the best performance compared to other cases, i.e., lowest comprehensive water-use cost and lowest fresh water consumption. Compared to the current water-use index of the studied park, the index of the case driven by WPCC decreased by more than 20% at least. In terms of the results, the case studies also indicated the applicability of the proposed multi-scale optimization model, and all these data can help the steel industrial park to make decisions for implementing global optimization of water network.