This paper proposes a mixed integer linear program (MILP) for the design and planning of a production-distribution system. This study aims to help strategic and tactical decisions: opening, closing or enlargement of facilities, supplier selection, flows along the supply chain. These decisions are dynamic, i.e. the value of the decision variables may change within the planning horizon. The model considers a multi-echelon, multi-commodity production-distribution network with deterministic demands. We present one application: how to plan the expansion of a company that has to face increasing demands. We report numerical experiments with an MILP solver.