Page 167 - Sustainability in the Process Industry Integration and Optimization

144 Cha p te r S i x

facilitate a closer match between hot process and cold refrigeration
streams. Del Nogal et al. (2008) described a methodology for mixed
refrigerant system design based on a superstructure arrangement.
The problem is highly nonlinear and features many local optima, which
behave as unwanted “traps” for traditional deterministic optimization
methods. This paper therefore suggests that a genetic algorithm (GA)
be used to solve the optimization problem; see Figure 6.19.
The interactions between the GA and the simulator result in a set
of the best solutions found over a discretized solution space. The
preliminary solutions so obtained then serve as starting points for
standard nonlinear programming (NLP) optimization techniques to
fine-tune the results and, finally, report the optimal solution. One of
the important aspects of this model is that it ensures the feasibility of
heat recovery in every exchanger. The design produced during
optimization is then simulated, and cold and hot Composite Curves
are produced. Finally, the CCs are rigorously checked against the
stipulated ΔT .
min
6.6 Integrating Reliability, Availability, and
Maintainability into Process Design

6.6.1 Integration

Current practice often views reliability as a mere afterthought to the
design process. As a result, systems go through repeated design and
redesign in search of greater process reliability, availability, and
maintainability (RAM). An alternative approach, as described by
Yin and colleagues (2009), is illustrated in Figure 6.20. This new

FIGURE 6.19 Integrated Genetic
design for low-temperature Algorithm
energy systems (after Del
Nogal et al., 2008). Set of operating
condition with structural
options
Objective
Refrigeration
simulator function
Updated power
demands
Driver selection

Integrated design