This paper deals with production activities planning on a tactical horizon for the phosphate industry. The studied supply chain is organized depending on the transportation modes used. There exist trains and pipeline that transport phosphate rock and pulp (respectively) to export and local customers. Due to the predetermined allocation of raw materials used currently in producing the final product, planning production activities in this supply chain should be then reorganized in such a way as to consider all the production configurations related to the use of each of the two transportation modes using such a flexible allocation instead. Moreover, customer requirements in terms of product quality represented by quality charters may be changed by the customer over tirne and from one order book to another for the same final product, which makes satisfying those requirements difficult. Thus, an integer linear programming mode! is developed and allows a global optimization of allocating raw materials and routings dynamically to satisfy the two types of customers at a lower cost. This work concerns a real industrial context and the numerical results provided prove the efficiency and flexibility ensured by our approach in comparison to the practices adopted currently