The invention discloses a load-side energy storage optimization configuration method based on a peak-shaving effect, and the method comprises the steps: firstly configuring the typical daily load dataof an energy storage load point, and generating a candidate configuration set according to the standard deviation of load data and the mean unidirectional duration; secondly taking each configurationelement in the candidate configuration set as a boundary condition, modifying the inequality constraints for calculating the peak-shaving effect, and solving the optimal value of variance of a new load curve through mixed integer programming, wherein the decrease rate of the corresponding standard deviation relative to an original curve standard deviation is taken as the benefit score of configuration; finally calculating the cost of each candidate configuration point according to power and capacity, integrating the benefit scores and the cost, such as the ratio of the decrease ratio of the standard deviation to the cost, and selecting the optimal selection from the candidate set. The method can reduce the equipment cost of power grid energy storage configuration, can quickly and accurately obtain the optimal configuration of the load-side energy storage configuration, is high in accuracy of load-side energy storage configuration, and is high in reliability.