Due to an important spatial dispersion of raw materials in many farms and a competition between agricultural activities for the use of production factors (land in particular), strongly dependent on the CAP, the cost estimates of these raw materials pose specific problems. Thanks to supply models, based on linear programming, it is possible to correctly estimate these costs, their diversity and finally to aggregate them in order to obtain raw material supply for industry. At the same time, one estimates the agricultural producers' surplus, one of the macroeconomic advantages of biofuels.