Page 46 - Sustainability in the Process Industry Integration and Optimization

CHAPTER 3

Process Optimization

his chapter deals with process optimization: its definition,
goals, and application areas within sustainable industrial
Tprocess design and integration. The aim is to provide
information on how to formulate sustainability tasks as optimization
problems and on what tools to employ for solving them. The chapter
begins with a brief description of the general framework for model
building and optimization; this is followed by basics of optimization
problems and their classes as well as descriptions of the most common
algorithms for solving optimization problems. Finally, the chapter
discusses how to build models efficiently, how to handle complexity,
and how to ensure model adequacy and sufficient precision. The
details of computational implementations of optimization solvers
and other software tools are given in Chapter 9.

3.1 Introduction
Building and operating industrial processes entail costs and
environmental impacts. Emissions and effluents include: gaseous
waste streams, which may harbor CO , SO and NO ; wastewater and
2 x x
various aqueous streams; and flue gases. When attempting to
improve the environmental and economic performance of process
systems, it is important to keep in mind that the processing paths,
which connect the various system inputs and outputs, usually
interact with each other. Therefore, minimizing resource demands
and environmental impacts is greatly facilitated by properly modeling
the process systems and then deciding which designs and operating
policies to pursue—in what priority and to what extent.
Maintaining a balance between model accuracy and simplicity is
necessary in order to derive meaningful results with minimal
computational expense. A system model can be created for different
purposes. A lumped steady-state model (i.e., one that neglects
variations in time and space) will contain only algebraic equations.
To simplify the problem, steady-state models assume that the
operating units are “black boxes” or “gray boxes” (a black box is a model
that represents an empirical process in terms of its input, output, and
transfer parameters but does not describe any internal physics; a gray

23