Page 63 - Design of Simple and Robust Process Plants

3.3 Design Philosophies 47
will be selected for the simultaneous optimization of process conditions and equip-
ment dimensions, see Chapter 4.
This methodology has as advantage that all decision steps can be followed and
intermediately validated while active constraints might be reconsidered.
Process simulation with a sufficiently detailed reactor model, with options for
mixed integer problems including cost estimation and economic sections, are a
requirement for design optimizations.

Summary
. Design optimization will increasingly be applied; the driving force for this
increased emphasis will be ± next to economic consideration ± the need to
comply with a sustainable world.
. A layered optimization approach will be followed by the engineer to maintain
an insight in all decision steps during the design process. Ultimately, an over-
all optimization model will be used to optimize design of the selected overall
flowsheet.
. The simulation environment needs to be mixed integer, and equipped with
estimation and economic sections to handle the design as a single optimiza-
tion problem.

3.3.6
Clever Process Integration

Process integration is a development that is inherent to the trend of further design
optimization. All types of integration within between± as well as between ± pro-
cesses will be studied in order to optimize the design. Next to heat integration,
power, cooling, refrigeration, water and hydrogen integration are also studied to
achieve better overall performance.
In general, the integration studies are performed under steady-state conditions,
without any consideration for either the dynamic effect or the availability of the
streams. As such, the integration might be seen as a contradiction to process simplifi-
cation because it makes the design more complicated. Referring to the formulation of
complexity in Chapter 2, it is clear that this will add to the complexity. Integration will
have its effect on complexity, as the following terms will be increased: the number of
equipment pieces, interconnections, input and output DI&DOs and AI&AOs, but par-
ticularly the disturbances will increase ± and this as such may introduce interaction.
There are two ways to reduce complexity:
1. By application of a higher automation level.
2. By de-coupling the system to achieve robust control.

The previously mentioned examples of a complicated piece of equipment such as a TV
set or a household refrigerator are simple to use due to their high level of automation.
However, de-coupling must be carried out in terms of both hardware and software
wise.