INGAR   05399
INSTITUTO DE DESARROLLO Y DISEÑO
Unidad Ejecutora - UE
congresos y reuniones científicas
Título:
Simultaneous design and scheduling of a plant for producing ethanol and derivatives
Autor/es:
FUMERO, YANINA; CORSANO, GABRIELA; MONTAGNA, JORGE MARCELO
Lugar:
Chalkidiki
Reunión:
Congreso; ESCAPE 21 - European Symposium on Computer-Aided Process Engng; 2011
Institución organizadora:
CAPE-EFCE
Resumen:
The inherent operational flexibility of multiproduct plants gives rise to considerable complexity in the design of such plants. In many cases, scheduling strategies are not incorporated or well integrated. Usually the simplest scheduling sequence, single product campaign (SPC), is considered, which may lead to over-design or under-design. Besides, in some cases, SPC are not acceptable, for example when perishable products must be supplied during the entire time horizon. In order to solve this problem with a different approach, Corsano et al. (2007) presented a heuristic strategy for the synthesis and design of multiproduct batch plants where the production scheduling is considered through mixed product campaigns (MPC). They formulated a two stage methodology, where non linear programming (NLP) models were solved at each stage. Taking into account the ratios among the number of product batches obtained in the first stage, different configurations for the MPC were proposed by the designer in the second stage. In this way, the number of batches of each product and the batches sequencing in each campaign were predetermined. Then, a model was solved for each recommended campaign, and the best solution was selected among the proposed MPCs. In this work, the design and scheduling with MPC of a semicontinuous/batch plant are simultaneously optimized through a mixed integer linear programming (MILP) model. The number of parallel units out of phase for batch stages, the unit sizes, the product batch sizes, the number of batches for each product in the campaign and production sequence on each unit are determined in order to fulfill the product demands in the time horizon. The objective function minimizes the investment cost. In order to avoid non linear formulations, the assumption of discrete sizes for semicontinuous and batch units (which is the usual commercial procurement policy), and a maximum number of batches of each product in the campaign are adopted. The proposed approach is applied and solved for a plant with batch and semicontinuous units to produce torula yeast, ethanol, and bakery yeast. The bakery yeast is a byproduct of ethanol production that is obtained by the evaporation and drying of the centrifugation residue of this process. Due to yeast degradation is fast and a continuous supply must be assured, the MPC policy is the most appropriate for these productions. Also, this policy allows reducing the idle times, caused by the difference among processing time and different processing routes for products, and avoids unit sizes overestimation. The proposed MILP model is a novel approach where the simultaneous optimization of design and scheduling considering MPCs, for a multiproduct semicontinuous/batch plant, can be solved to global optimality with reasonable computational effort.