Multicriteria analysis and optimization of a of heat transfer and drying process for application in West Africa

Abstract

The reinforcement of the food processing sector is recognized as a driving factor for the development of sub-Saharan African countries, faced with considerable major demographic growth accompanied by a high rate of urbanization. While the agribusiness companies generate added value locally and boost agricultural production, they find difficulties in obtaining efficient equipment and securing their energy supply.Agribusiness process design and optimization methods are still underdeveloped, due to the complexity of these systems (food composition and properties, variable and changing quality, etc.), modelling of which is not easy since it requires multidisciplinary knowledge.This work proposes to implement an integrated method, already proven in other industrial fields, the “Observation-Interpretation-Aggregation” method (OIA), and apply it to a process coupling a biomass energy conversion unit to a cereal products dryer. The bioenergy supply for drying, a very common practice in West Africa despite being energy-intensive, represents a challenge for the companies. The design of this process takes into account the various objectives such as quality of the dried product, local manufacture and the energy efficiency of the equipment, in order to guarantee better sustainability.First of all, the models for heat transfer and pressure loss associated with an innovative elliptic turbulator are created. This component is inserted into the tubes of a heat exchanger, and significantly improves heat transfer. Secondly, the process design and observation variables are defined and justified. The representation models of the various unit operations are developed and brought together in a simulator, in order to predict the process performances. Finally, the simulator is integrated into a multicriteria optimization environment able to formalize, interpret and then aggregate end user preferences. This procedure is based on a genetic algorithm. The relevance of the hig