(521c) Simultaneous Optimization of Flowsheet, Heat Recovery, and Water Network

Significant research work has been done in the area of heat integration in the past thirty years due to the incentive of improving energy recovery.  Also, stricter environmental regulations have caused an increasing demand for minimization of water consumption and effluent generation for industrial processes.  Various techniques have been developed in response to this need for process integration and they are presented in these reviews^[1][3][4].  However, most of this research has been focused on minimization of energy and water separately.  Furthermore, traditionally, process synthesis design is performed sequentially - the flowsheet is optimized first, then the resulting operating parameters are used in heat exchanger network synthesis (HENS) and water network (WN). 

This presentation describes a novel and effective approach for the simultaneous optimization of process flowsheet subject to heat and water integration.  Through simultaneous optimization approach, the idea is to represent the HENS and the WN through design targets such as energy consumption and water consumption to avoid solving a very large MINLP problem. The advantage of performing simultaneous optimization is that proper trade-offs of operation cost and investment cost can be taken into account.

While effective NLP targeting formulation for simultaneous process optimization with heat integration has already been presented by Duran & Grossmann^[2], this is not the case for targeting minimum water use.  Typical WN problems are formulated using NLP or MINLP models using a detailed superstructure as reported in many references^[5].  Freshwater is first used for process units, and then wastewater generated from these processes is treated in various treatment units.  Process units are defined by their allowable contaminant concentration level, whereas wastewater treatment units are defined based on their contaminant removal ratio.  The superstructures consider systematic alternatives for water recycle and water reuse with the objective of minimizing the freshwater consumption, or more generally the total cost. In general, single-contaminant systems can be modeled using LP formulations, while multi-contaminant systems take the form of NLP models. 

In order to incorporate water integration in the simultaneous optimization framework, a new LP formulation is presented to determine the target for a multi-contaminant water network with process units only.  In particular, the bilinear equations of the mixer are replaced by a set of inequalities that are evaluated at the limits of inlets and outlet concentrations of the units. This relaxation, which follows the optimality conditions proved by Bagajewicz & Savelski^[6], yields an LP that can be shown to predict the exact minimum amount of freshwater that is required by the process units. The LP targeting model is incorporated in the simultaneous optimization of a methanol synthesis process flowsheet with heat and water integration and resulted in an improvement of the economical objective function.  The important observation here is that in the simultaneous approach, heating and cooling are integrated as part of the optimization so that improved operating conditions can be achieved. It is shown for a specific instance that the resulting flowsheet from the simultaneous method improves the profit by 17%.

For the case of multi-contaminant water network problems including wastewater treatment units with process units, the proposed LP targeting formulation is shown to predict a rigorous minimum freshwater requirement for special cases of the WN problem. For the general case, the lack of exact upper bound of contaminant concentration in the treatment units complicates the interactions in the mixer units.  To circumvent this problem, a number of linear relaxations are used to obtain approximate but tight water targets for general WN problems.  The proposed targeting model yields an MILP problem. The application of this model is also applied in the simultaneous optimization, heat and water integration of the methanol process yielding further improvements in the profit.

References:

[1] Bagajewicz, M., (2000). A review of recent design procedures for water networks in re0neries and process plants. Computers & Chemical Engineering, 24 (9), 2093-2115.   

[2] Duran, M.A. & Grossmann, I.E. (1986). Simultaneous optimization and heat integration of chemical processes.  AIChE Journal, 32, 123-138.

[3] Foo, D.C.Y. (2009). State-of-the-art review of pinch analysis techniques for water network synthesis. Ind. Eng. Chem. Res. 48, 5125–5159.

[4] Furman, K.C. & Sahinidis, N.V. (2002). A critical review and annotated bibliography for heat exchanger network synthesis in the 20th Century. Ind. Eng. Chem. Res., 41, 2335-2370.

[5] Karuppiah, R. & Grossmann, I. E. (2008). Global optimization of multi-scenario mixed integer nonlinear programming models arising in the synthesis of integrated water networks under uncertainty. Computers & Chemical Engineering, 32(1-2), 145-160.

[6] Savelski, M. & Bagajewicz, M. (2003). On the necessary conditions of optimality of water utilization systems in process plants with multiple contaminants. Chemical Engineering Science, 58(23-24), 5349-5362.