Page 316 - Sustainability in the Process Industry Integration and Optimization

T y p i c a l P i t f a l l s a n d How t o Avo i d T h e m 293

fashion, and a common control task is to maintain such processes in a
steady state. However, there are also many situations in which the
working regime must be changed. This occurs not only at start-up
and shutdown but also in response to changes in operating conditions
(e.g., outside temperatures) and production conditions (e.g., capacity,
volume). Problems involving such variation are generally addressed
in one of two ways: (1) accommodating various scenarios via a design
objective of minimizing sensitivity to the possible changes or
(2) designing to optimize the total system’s overall dynamic
performance. The second option is much more complicated and
requires more time and resources; hence it is used only when
substantial deviations from steady-state operations are anticipated.

12.4 Interpreting Results

After data extraction, the most important aspect of PI analysis and
optimization is the correct interpretation of results. These results are
typically presented in the form of a printout generated by some
software tool; in most cases, the software output is in the form of
a grid diagram (e.g., STAR, 2009) or PFD supported by printout
tables—as with ASPEN (AspenTech, 2009a, 2009b, 2009c) and
gPROMS (PSE, 2009). In order to minimize typos and misinterpretation,
many software tools feature an interface that facilitates the transfer
of data files. This technology is now fairly well developed.
The most challenging aspect of data interpretation is assessing
the results in terms of possible further development or correction of
the important process features. The interpretation will depend on the
extent to which the factors mentioned previously (i.e., data uncertainty,
data softness, flexibility, operability, controllability, safety, availability,
and maintenance) have already been incorporated into the data
extraction. If the data extraction did not reflect these considerations,
then they must be addressed during the data interpretation. At this
stage, the close collaboration of a team that includes the main involved
professionals is highly recommended.
Designers are strongly advised not to stick with just one solution
but rather to explore different potential scenarios associated with
various operating conditions and then test their design’s sensitivity
to the possible variations. For screening and scoping it is helpful to
use all types of Composite Curves as well as information (if provided
by software) on which streams are contributing to specific parts of
the curves.

12.5 Making It Happen

Even when a sustainable and near-optimum design is developed,
it must still be put into practice. This involves selling the