Supply chain network design (SCND) is a strategic decision aimed to establish the network structure. Decisions such as facilities location, routes selection, materials flows, transportation modes selection, and inventory levels, are part of SCND problem. However, according to the literature review, optimization models considering capacity level as a decision variable (dynamic capacity) are almost non-existent. In real cases, capacity level is a variable that significantly affects the network structure and its economic, environmental and social performance. Consequently, allocating a fixed level of capacity to facilities prevents to analyze its effects on supply chain performance. Therefore, the aim of this contribution is to present an optimization model to support a biofuel SCND using coffee residues in the Colombian context. The model addresses a dynamic capacity strategy (expansion or closing of facilities), establishing long-term changes in facilities, transportation and inventory decisions and its effect on sustainable performance (economic, environmental and social). When compared with a fixed capacity strategy, a dynamic capacity strategy presents better performance in economic (Net Present Value) and environmental dimensions. However, negative effects in the social dimension (job creation) were observed. The proposed model is a useful and realistic tool to establish long-term investment portfolios to support the development of new supply chains.