CHAPTER 14

Optimization Process
Set up an optimization problem:

Define decision variables as independent variables, either continuous (e.g., temperature) or discrete
(e.g., number of stages).

Quantify the value of potential improvement:

Utilize an objective function that reaches a minimum or maximum for optimal values of decision
variables.

Objective function types depend on user preference, such as profit or cost functions.

Identify potential for improvement:

Explore techniques to quickly identify areas for improvement.

Identify constraints, barriers, and bottlenecks:

Constraints limit decision variable values and can be linear or nonlinear, equality or inequality.

Choose an appropriate procedure:

Distinguish between linear programming (for linear objectives and constraints) and nonlinear
programming (for other cases, including mixed-integer nonlinear programming).

Evaluate the result:

 Differentiate between global optimum (best for all allowable values of decision variables)
and local optimum (no small, allowable change improves the objective function).
Decision Variables:

 Independent variables, continuous (e.g., temperature) or discrete (e.g., number of stages).

Objective Function:

 Mathematical function reaching a minimum or maximum for optimal decision variable

values.
 Represents the measure of value or goodness for the optimization problem.

Constraints:

Limitations on decision variable values, linear or nonlinear, equality or inequality.

Optimization Types:

 Linear programming for linear objectives and constraints.

 Nonlinear programming for other cases.
 Quadratic programming for second-order objective with linear constraints.
 Mixed-integer for problems with both discrete and continuous variables.

Common Misconceptions:

 Optimization is not necessarily a complex, esoteric, mathematical exercise.

 It involves dynamic, creative activities such as brainstorming and exploring alternatives.
 Real problems are often more interesting and require careful consideration of constraints.

Example:

 Example of a chemical engineering problem: evaluating the optimal heat exchanger for
heating a stream from 30°C to 160°C.
 Includes continuous variables (area of the heat exchanger, temperature) and integer
variables (choice of low-, medium-, or high-pressure steam).
 Constraints include materials of construction dependent nonlinearly on factors like pressure,
temperature, and composition.

Additional Note:

 Optimization is a Dynamic, Creative Activity:

 Involves brainstorming, exploring alternatives, and asking "What if...?"
 Analytical solutions are possible for some problems, but real problems are often more
dynamic and require creative thinking.
Visualization Example:

 Figure 14.1(a) shows a classical curve of annualized pumping cost versus pipe diameter.
 The analysis involves annualized equipment and power costs, leading to a total annualized
cost curve.
 Analytical determination of optimal pipe diameter involves considering the slope and second
derivative of the combined cost curve.

Optimization in Process Engineering:

Curve Intersection Misconception:

 Optimization often involves balancing annual operating cost and annualized capital cost.
However, assuming the intersection of these cost curves is the optimum is a common
misconception.
 Real-world cost functions are influenced by standardization of equipment sizes. Some pipe
diameters and pump sizes are standard, and deviating from these incurs higher costs.
 Cost evaluations are typically limited to standard pipe sizes, emphasizing the need for careful
consideration of available equipment options.

First Derivative Misconception:

 The assumption that the optimum occurs where the first derivatives of the cost function are
zero is another misconception.
 Continuous and smooth cost functions rarely have optima at points where derivatives are
zero. Optimal solutions often lie along constraints, requiring a nuanced understanding of the
problem.
 The misconception highlights the need for a more sophisticated approach, considering
constraints and exploring the problem space thoroughly.

Locating the Optimum:

 Bracketing a solution, while a common strategy, doesn't guarantee finding the best solution.
 Proper optimization requires more than bracketing; it involves iterative evaluations between
points to ensure accuracy.
 Figure 14.3 illustrates the importance of evaluating more points within the bracket to refine
and confirm the optimum location accurately.

Estimating Problem Difficulty:

 Estimating the difficulty of a problem is crucial for effective problem-solving.

 Characteristics such as the number of decision variables, their correlation, and the
involvement of discrete or continuous variables help classify problems as easy or difficult.
  Recognizing problem complexity guides the allocation of resources and the development of
appropriate optimization strategies.

Top-Down and Bottom-Up Strategies:

 Top-down and bottom-up strategies offer complementary approaches to optimization.

 Top-down involves looking at the big picture first, encouraging bold changes in process
configuration or variable values.
 Bottom-up focuses on detailed study, confirming whether changes are genuine
improvements. Flexibility is key, allowing for adjustments based on both overview and
details.

Communication of Optimization Results:

 Effectively communicating optimization results is essential for decision-makers.

 Presenting results involves explaining the range of decision variables explored, emphasizing
that the solution is not merely a local minimum.
 Visual aids, like families of curves, help convey the validity of the optimum and the sensitivity
of the solution to changes in decision variables.

Base Case in Optimization:

 The base case serves as the starting point for optimization, representing an existing process
or a conceptual design.
 It must be detailed enough to calculate the objective function, which often includes capital
and operating costs.
 While not necessarily a completed design, the base case provides a foundation for
comparison and improvement.

Objective Functions:

 Choosing the right objective function is critical for optimization success.

 Commonly, minimizing equivalent annual operating cost (EAOC) or maximizing net profit are
chosen as objective functions.
 The objective function should capture the most desired condition and align with the overall
goals of the optimization effort.

Analysis of Base Costs:

 Analyzing base costs involves establishing targets for an idealized process to serve as a
benchmark.
 Pareto analysis helps identify major contributors to the objective function, focusing efforts
on areas with the most significant impact.
  The base cost analysis provides a roadmap for optimization, highlighting key areas for
improvement.

Identifying Decision Variables:

 Decision variables significantly impact the outcome of the optimization process.

 Identifying key decision variables involves considering their contribution to the objective
function and their sensitivity.
 Techniques such as sensitivity analysis and experimental design assist in prioritizing decision
variables.

Strategies for Decision Variables:

 Decision variables can be categorized as controlled variables or equipment characteristics.

 Controlled variables, influenced by control systems, are secondary decision variables, while
primary decision variables are manipulated by the final control element.
 Equipment characteristics, such as reactor volume and number of stages, are critical decision
variables that influence the design of individual units.

Topological Optimization
Topological optimization involves the arrangement of process equipment, while parametric
optimization deals with operating variables. This section focuses on topological optimization.

Why Topological Optimization First:

1. Impact on Profitability: Topological changes often have a significant impact on plant profitability.

2. Parametric Optimization Interpretation: Parametric optimization is easier to interpret when the

flowsheet's topology is fixed.

3. Simultaneous Use: Both types of optimization may be needed, but major topological changes are
best addressed early in the process.

Questions in Topological Optimization:

When considering the topology of a process, engineers must answer key questions:

1. Can unwanted by-products be eliminated?

2. Can equipment be eliminated or rearranged?

3. Can alternative separation methods or reactor configurations be employed?

4. To what extent can heat integration be improved?

Elimination of Unwanted By-Products:

 Addressing unwanted nonhazardous by-products or hazardous waste streams is crucial early

in the design process.
 Example 14.2 explores process changes to minimize or eliminate the production of a costly
waste stream (DIPB) in a cumene production process.

Elimination and Rearrangement of Equipment:

 Both can lead to significant process improvements.

 Elimination involves removing equipment based on changes in operating conditions. Example
14.3 evaluates changes in a cumene process by reducing per-pass conversion of propylene.
 Rearrangement should follow certain guidelines. For instance, easier separations should be
performed first, and special consideration is needed for mixtures forming azeotropes.

Alternative Separation Schemes and Reactor Configurations:

 Early in the design process, engineers should explore alternative separation technologies and
reactor configurations.
 While distillation is common, recognizing when nontraditional technologies might be
beneficial is crucial.
 Example 14.6 explores the benefits of separating hydrogen from methane in a toluene
hydrodealkylation process and discusses membrane separation and pressure-swing
adsorption as potential technologies.

Parametric Optimization:
 Definition: Parametric optimization involves adjusting key decision variables within a
specified range to find the optimal conditions for a process.
 Application in the Text: The text describes the optimization of the distillation column T-201
by varying parameters like reflux ratio, operating pressure, and DME recovery to maximize
NPV.

Objective Function:

Definition: An objective function is a mathematical expression representing the goal to be optimized

in an optimization problem. It quantifies the performance of the system under different conditions.

Examples in the Text:

  The text uses the net present value (NPV) as the objective function for the optimization of
the distillation column. The NPV is a measure of the profitability of the process over time.
 In the context of the DME separation process, the NPV is influenced by factors such as reflux
ratio, operating pressure, and percent recovery of key components.

SHORT ANSWER QUESTIONS:

1. Describe a Pareto analysis. When is it used?

Pareto Analysis:

 Definition: Pareto analysis is a technique used for identifying the most significant factors
contributing to a specific problem or outcome. It is based on the Pareto Principle, which
states that a small number of causes (20%) are responsible for a large percentage (80%) of
the effects.
 Application: Pareto analysis is used when there are multiple factors influencing a problem,
and there is a need to prioritize efforts to address the most critical issues. It helps in focusing
resources on the factors that have the most significant impact.

2. What is the difference between parametric optimization and topological optimization? List one
example of each.

PARAMETRIC OPTIMIZATION TOPOLOGICAL OPTIMIZATION

Definition: Parametric optimization involves Definition: Topological optimization involves
adjusting specific numerical parameters or modifying the structure or configuration of a
variables within a system or process to optimize system or process to achieve the best
its performance. performance.

Example: In chemical engineering, adjusting the Example: In structural engineering, changing

operating temperature and pressure of a the design of a bridge to minimize material
reactor to maximize the yield of a desired usage while maintaining strength is an example
product is an example of parametric of topological optimization.
optimization.

3. What is an objective function? Give two examples of one.

Objective Function:

Definition: An objective function is a mathematical expression that quantifies the goal or outcome to
be optimized in an optimization problem. It represents the measure of success or performance.

Examples:

Net Present Value (NPV) in Financial Optimization: In financial decision-making, the objective
function could be the NPV, which considers the present value of cash flows to determine the
profitability of an investment over time.
Total Cost in Supply Chain Optimization: In supply chain management, the objective function might
be the total cost, including production, transportation, and inventory costs. The goal is to minimize
the overall cost for efficient supply chain operations.