This work aims to develop and validate an algorithm to identify potential photovoltaic panel users among a database of customers. This need is influenced by the trend of increasing consumption in the large urban and industrial centers of Rio de Janeiro, especially at peak hours, driven by the growing middle class and the economy. This work’s motivation, serving the interests of the project, is to offer a tool to aid decision making for energy utilities and infrastructure planning. As a result, the proposed algorithm was able to classify medium, and high, voltage consumers, using data provided by the IT team at Light, into two categories: ‘potential’ or ‘not potential’ (to the deployment of photovoltaic panels on their grid). Among the conclusions, the algorithm was effective in reducing levels of consumption at customer’s peak hours; however, the economic feasibility study of the project deemed it to be impractical, since the return on investment is too small.