(347e) Energy Integration for Waste Heat in Industrial Plants through a Metaheuristic-Deterministic Approach

Several industrial processes waste great amounts of energy that has no practical use and it is discharged into the environment (causing a double problem). Waste heat is classified according temperature ranges as high temperature (waste heat > 400 °C), medium temperature (100 °C < waste heat < 400 °C) and low temperature (waste heat <100 °C). It has been reported that annually in U.S. a total of 18.9 exajoules of waste heat are discharged by thermal power plants, where 4% of this energy corresponds to waste heat at temperatures above 90 °C. In this context, there is a set of systems available for waste heat recovery (WHR), with the purpose of reusing the residual energy and as consequence to improve the overall energy efficiency. Thus, organic Rankine cycle (ORC) is a thermodynamic cycle for the generation of electricity, which uses an organic liquid with a high molecular weight and a low liquid-vapor phase change as a working fluid; the main application of an ORC is the WHR. Other auxiliary plant that can be employed in order to take advantage for the residual energy is the absorption refrigeration (AR) cycle, which represents an efficient alternative to satisfy the refrigeration requirements in multiple processes. Finally, heat exchangers networks (HENs) deal with the heat recovery between process streams; with this approach, the consumption of utilities (i.e., steam, fuels, cooling water, refrigeration) is reduced, and consequently, economic and environmental benefits are obtained.

However the main disadvantage when HENs are integrated and modeled with thermal engines for waste heat recovery (i.e., SRC, ORC or AR) is the significant number of assumptions related to the operation of the thermal engines because in deterministic optimization approaches is extremely complicated to involve rigorous thermodynamic models in order to simulate them; this is associated to the highly non-convex mathematical terms, which lead to difficult convergence problems. Therefore, in order to overcome previous limitations, in this work we propose a hybrid optimization approach, which combines metaheuristic and deterministic techniques together with process simulators; this way, thermal engines are properly simulated and the optimization of an entire integrated system between HENs and thermal engines can obtain more accurate results.

Moreover, metaheuristic optimization is established through the MS Excel-VBA-Aspen Plus link to obtain accurate modelling results. Deterministic optimization is implemented in the GAMS platform, where the mathematical formulation is based on a superstructure that considers all the energy interconnections between the heat exchanger network, utilities, and thermal engines. Furthermore, economic, environmental and social targets are evaluated. A case study is presented to show the applicability of the proposed methodology. The operating conditions obtained are presented (working fluid flowrate, temperatures, pressures and efficiencies) for each thermal engine. Finally, the results show an increase in the total annual profit of 148%.