This study deals with a real-life batch sizing and sequencing problem where sequence-dependent setup times are necessary when changing the production from one type of product to another. Given an internal demand to be fulfilled at each period t from buffers, one has to organize the production to replenish the buffers so as to keep the inventory levels between predefined bounds at any time, while respecting a minimum batch size. To simultaneously define which type of product and for how long time to produce, we propose an Integer Linear Programming (ILP) formulation which determines the optimal sequence of machine states over the planning horizon. We distinguish three possible states: production, setup or changeovers and idle states. The first experiments are carried out on small-size instances, using Cplex solver. As future work, we will extend the model to better represent the real case and then we aim to develope approximate resolution methods to solve larger instances based on real data.