Javier A. Francesconi

Simultaneous Flowsheet Optimization and Heat Integration of a Bioethanol Processor for PEM Fuel Cell System

Abstract The conceptual design of a bioethanol processor for fuel cell applications with heat integration was analyzed via model-based optimization formulation. Basically, the system consists of steam reforming of bioethanol and hydrogen purification process that are coupled to a proton electrolyte membrane fuel cell (PEMFC). The system was implemented within General Algebraic Modeling System (GAMS). The model simultaneously handles the problem of optimal heat integration while performing the optimization of process flowsheet. A 50 kW power generation system is presented as case study; the objective was to obtain the operative conditions for the process units that maximize the system efficiency. The operating variables of the system considered as decision variables were: system pressure, water-ethanol molar ratio and reforming temperature, input temperatures to hydrogen purification reactors, fuel cell temperature, and fuel cell hydrogen and oxygen utilizations. The results obtained predict a global efficiency of 43%, about 5% higher than the reference case. These results demonstrate the importance of simultaneous optimal heat integration for fuel cell-based processes.

Optimal Synthesis of Heat Exchanger Networks Using Enthalpy-Temperature Functions to Describe Streams

Abstract This work presents a mixed integer non linear programming (MINLP) model for heat exchanger network synthesis including a detailed description of the process streams resorting to a mathematical function that matches temperature with enthalpy. This allows the model to account for non linear behavior of streams by taking into consideration their composition. To achieve the objective, the energy balance equation proposed in the “synheat” model by Yee and Grossmann (1990) is modified but equations involving temperature are kept the same. A new equation is added to that model to describe the relationship between temperature and enthalpy by correlation. As the original model, this modified MINLP model provides the network structure that minimizes the total annual cost. The mathematical program has non-convex equations, and only locally optimal solutions can be guaranteed. The approach between cold and hot streams has more detailed information because describes the real behavior of streams, and avoids matches that a linear functionality or a classical model may allow. Correlations between temperature and enthalpy are a function of the components present in a particular stream. The required information can be collected from bibliography or public databases as those provided by NIST (National Institute of Standards and Technology). The proposed methodology is a useful tool when the minimum temperature approach between streams is small. Examples are presented and discussed to compare results when streams are described by linear and non linear functionality between temperature and enthalpy.