As the demand for national electricity grows, the selection of distribution transformer capacity becomes more and more important. The economics and safety of distribution transformers directly affect the economic benefits of the entire power sector. Reasonable selection of transformer capacity can effectively guide the reasonable economic investment of relevant departments, and also plays a key role in the energy-saving operation of distribution transformers. This paper first clusters the distribution transformers working in different scenarios, and then performs load forecasting based on gray theory for each type of scenario to obtain the load forecast value after thirty years. A full life cycle cost model is established, taking into account all costs in the actual selection of transformers. After the actual data test, this method has good guiding significance for the selection of actual transformers.