Page 69 - Design of Simple and Robust Process Plants

3.3 Design Philosophies 53
The process optimization models (with the exception of planning models) are in
general based on non linear programming (NLP) techniques and are equation-based.
Most process plants have a strong nonlinear behavior, while at the same time the
profit margins are only a few percent of the operational costs. These factors require
the models to be highly accurate, and to be validated over their operational range.
A ªprofit meterº designed as an envelope over the total process plant is an excel-
lent tool for use in overall model validation (Krist et al., 1994). The profit meter can
also be used as an instrument to obtain and evaluate real-time plant performance
data for production, as well as for business people to maximize profit by varying
process conditions.
The technology for dynamic and continuous optimization is relative new, but is
maturing. Savings realized through optimization are considerable, but are differen-
tiated by type and by application. Savings for each type of optimization are of the
order of magnitude 2±4% of operational costs, and may be 10% if savings on con-
strained controllers and capacity utilization are included. The impact on actual profit
might be of the order 10±20%, or even higher, depending on the profit margin.

Summary
. Operation optimization can be split into business and plant optimization ap-
plications.
. Business optimizations are applied to planning, supply chain, feedstock eva-
luations, and overall business operation.
. Process optimizations are applied to scheduling, operational conditions selec-
tion, and transient operations.
. Process optimization models need to be highly accurate as the effect is in the
order of a few percent of operational costs.
. Profit meters are designed to measure plant performance in real time; they
are used for overall model validation and evaluation of plant performance by
operation and business.
. Operation optimization may make a significant contribution to profit mar-
gins.

World-class Manufacturing Perspective

3.3.9
Just-in-Time Production (JIP)

ªJust-in-time productionº is a term that was introduced in the 1970s and 1980s, and
was part of the quality wave which started in Japan. In particular, JIP came from the
car industry, where the logistic costs are extremely high, but since then it has found
its way through all industries. The principle is that a product is produced at quality
(total quality control) when isit is needed, the objective being to reduce the logistic
costs. The same applies for feed streams which are to be received, which one might
call ªjust-in-time receivingº.