Page 339 - Sustainability in the Process Industry Integration and Optimization

316 Cha p te r F o u r tee n

14.2 Development Trends
At the 2009 PRES conference, development trends in PI, optimization,
and Mathematical Programming as tools for sustainability in
process industries were reviewed by Friedler (2009, 2010). Some of
these trends are discussed in this section.

14.2.1 Top-Level Analysis
A typical industrial site comprises different process production units
linked to a common utility system. The centralized utility system
meets the demands for heat and power, creating indirect links
between the processes. Considerable amounts of steam can often be
saved by optimizing and retrofitting site processes and their heat
exchanger networks (Dhole and Linnhoff, 1993a; Asante and Zhu,
1997). Although such projects might have the potential to save steam
at different pressure levels, they may not always save costs or even
fuel. Moreover, steam at a given level of pressure can have different
prices depending on how much is saved, despite the common practice
of attributing a single value to steam at a given level. This issue was
first addressed systematically by Makwana and colleagues
(Makwana, 1998; Makwana, Smith, and Zhu, 1998), who provided an
approach to screening energy projects without the need to collect
data from every site process. Instead, their methodology of top-level
analysis required the collection and analysis only of basic data
relating to the central utility system. This approach was based on
comparing current and optional heat flow paths through the utility
system. The procedure provided a good conceptual view of the
problem, but the required number of manual steps and calculations
was not amenable to automation and too complex for application to
real sites. However, an important concept derived from this initial
version of top-level analysis is that, given marginal steam prices, a
small part of the total site can be analyzed to rank the most promising
directions for steam-saving projects. Those marginal steam prices
are key because they can be computed in a different way—namely,
by utility system optimization. This approach led to the development
by Varbanov et al. (2004) of a stepwise optimization procedure that
can be used to determine the true marginal price of steam at different
pressure levels as well as the constraints on steam savings imposed
by the utility system. This approach builds on the optimization
framework for industrial utility systems presented by Varbanov,
Doyle, and Smith (2004), and it enables the analysis of complex utility
systems using automated tools for utility system optimization.
14.2.2 Maintenance Scheduling, Maintainability,
and Reliability
A chief concern of companies and operators is that industrial plants
should operate safely and reliably, with minimal costs for maintenance.