Distributed energy storage system (DESS) is very important for peak shaving of the power system. Its location and capacity arrangement has traditionally made it a focus for field study. However, poor economic and technical analyses, as well as DESS's high processing cost, remain issues. In light of this, this research provides a greedy algorithm-based optimum capacity allocation strategy for a DESS. First, a comprehensive DESS economic model and operation constraint model are established. Compared with the traditional defect of only considering investment and operation costs, it increases the economic benefits brought by energy storage operation scheduling. Then, using power loss sensitivity, site selection can lower the dimension of the addressing problem and enhance optimization efficiency. Then, the greedy algorithm is used to divide the DESS into many units and optimize them, and the decision-making process of each unit is simplified into a simple model, which can significantly improve the solution efficiency. The simulation study is conducted in MATLAB R2015b to validate the efficiency of the suggested technique, using load data from a neighborhood in Jiaxing City, Zhejiang Province, as an example. The results show that (1) when compared with evolutionary algorithms, this approach can only find the local optimal solution, and the economic gain is slightly smaller (albeit the difference is modest), but it can greatly improve computing efficiency and (2) the optimization results are identical to the entire optimization, but it excludes the computation of network loss cost, resulting in increased calculation efficiency.