Data driven multi-stage sizing of microgrids with grid capacity connection cost

Properly sizing a microgrid is of paramount importance to reach an optimal operation during its entire lifetime. This paper presents a mixed-integer linear optimization formulation of this sizing problem, with a particular emphasis on generation and battery storage system degradation, the corresponding necessary reinvestments along the life of the project, and connection costs to the public grid. This methodology is applied to a wide range of case studies in different locations around the globe. The generated results are then used to train a supervised learning method that can approximate the outcome of the sizing optimization problem in almost no time, allowing for quasi-instantaneous pre-designs.