Page 444 - Design of Simple and Robust Process Plants

10.3 The Design of High-quality Plants 431
10.3.2.6 Design to capacity
The objective is to avoid unnecessary accumulation of over-design, safety factors and
design allowances to prevent high cost. We can differentiate different types of over-
design, which as such might be defendable, but the combination of these factors
will lead to high cost.
Let us give an example of accumulation of over- design. For the equipment design
of a process plant, the hourly mass balance is taken as the basis. The mass balance
is often considered as to be fulfilled at the most extreme feed and product conditions
on a hot summer's day afternoon, and also for a very cold night in the winter. The
end-of-run conditions of the reactor and the maximum fouling will also be additive
factors used for equipment sizing. The designer must also have some concern about
the physical properties, and he/she will add some ªfatº for compensation. However,
he/she must also be concerned about the calculation method, with its reference
data. So, for a distillation column, they may design at 50% of the required product
specification, and at 60±70% of flooding, under extreme conditions. They will most
likely size the heat exchangers some 10% above the required area at extreme envir-
onmental conditions. This example is a reality in many cases, and that is why a
plant can easily have 20% excess capacity in its equipment. This capacity shows up
after start-up, when the size of some control valves is increased (this often being the
first bottleneck). No consideration is made as to the excess capacity required by the
business for future expansion as over-design, although it should be documented
within the scope of the project.
The different types of over-design to be recognized are:
. Mass balance and hourly capacity: some projects size all equipment on a
higher than capacity mass balance (engineering companies may do so to
ensure that the capacity guarantee is met ± the customer pays the bill).
. Raw material and product mixes and specification.
. Design variables, such as environmental conditions maximum and mini-
mum air temperatures, humidity, cooling water temperatures, catalyst aging,
fouling, utility design pressures and temperatures.
. Unit operation and equipment design parameters.
. Uncertainty in physical properties and correlations.
. Safety factors.
As a standard approach we could say that:

A process engineer should cope with uncertainties where they really exist and avoid over-
design on over-design.

An example could be the design of a heat exchanger. If there is uncertainty about
the fouling factor, we either make a comparison in a similar application or take our
best estimate, but we do not add an additional area for uncertainties in the calcula-
tion method. In general, these calculation correlations already have (as a standard) a
probability of over 95% on the calculated result.