Electric energy substitution can improve residents’ electrification level and reduce environmental pollution. This is in accord with requirements of sustainable development for energy consumption. In this paper, a proportional model-based Non-Intrusive Load Monitoring (NILM) technique is proposed to estimate the total electric energy increment of a community after electric energy substitution. First, this paper analyzes a real scene when residents use energy and get the electric appliance that can be used in electric energy substitution. These appliances are modeled by NILM based on a public dataset, and then current collectors are installed for a few users in the community to obtain low-frequency electric energy data. Based on the NILM model and the collected data, a proportional correlation between real power consumption and replaceable power consumption increment is built. Then the replaceable power consumption increment is estimated based on the established correlation and the actual power consumption of the known majority of users in the community. Finally, the estimation results are evaluated using real gas consumption data. Results show that the cost of sampling a few of users’ electricity data is low, and this method can estimate total electric energy increment in a certain area caused by electric energy substitution precisely. This is very helpful when carrying out electric energy substitution. This work is supported by a Project of State Grid Corporation of China (No. SGZJJXI0JCWT1800397) and the National Key Research and Development Program of China (No. 2018YFB0904904).