It is necessary to evaluate the reliability of the water supply network, when the water supply network is damaged by an earthquake. Therefore, this paper researched the feasibility and characteristics of the surplus power entropy as the reliability index of the water supply network, and established a scheme framework for optimizing and improving the reliability of the water supply network. This paper developed a reliability evaluation model for the water supply network after an earthquake. Combined with the Monte Carlo stochastic simulation hydraulic analysis, this model is also based on the pressure-driven nodes water demand model. In the case study, the surplus power entropy method was applied to test the reliability of the model. The statistical curves of the surplus power entropy of nodes and pipe networks, the distribution of the surplus power entropy with different intensities in pipe networks, and the comparison results of three reliability improvement schemes, before and after, were obtained. The influence factors of the surplus power entropy were obtained from the data analysis. The high consistency between the surplus power entropy and flow entropy verifies the feasibility of the surplus power entropy as a reliability index. The three schemes show that the surplus power entropy index can be used as a beneficial supplement to the reliability evaluation index of the pipe network.Sustainability 2020, 12, 1591 2 of 17 variety of specified working conditions, investigate the pipe network's ability to meet the demand, and test the reliability of the water supply network. The commonly used simulation methods include the Monte Carlo method [10], the quasi-Monte Carlo method [11], and so on. The agency methods do not directly test the reliability of the pipe network, but use the appropriate agency model to test the reliability [12]. The agency methods include information entropy [13,14], flow entropy [15], the path entropy method [16], and the recoverable index method [17]. Many academics and researchers used a combination of reliability methods in their studies. For example, Deng et al.[18] proposed a system reliability solution method for a water supply pipe network by combining analytics methods and simulation methods, while calculating path importance and connectivity ratio. The simulation methods can simulate and perform hydraulic analysis under various working conditions. However, this method requires many rounds of hydraulic calculation to reach the results. In addition, the agency method models are simple and easy to program and calculate, but they lack the source of certainty.This study combined the advantages of simulation methods and agency methods, and used Monte Carlo stochastic simulation to obtain the hydraulic analysis of the water supply network after earthquakes, based on the pressure drive node water demand model. Surplus power entropy was taken as the reliability evaluation index of the water supply network. On the basis of solving the entropy value, the paper developed the structural optimiz...