Research Interests
Dr. James Riggs
| Process Control | Process Optimization |
Process Control
Distillation Control:
Distillation is one of the
most underestimated fields of chemical
engineering. Just because distillation
has been around for well over a
hundred years does not mean that
it is well understood. This is particularly
true for distillation control. We
have and continue to study a range
of single distillation columns using
detailed dynamic simulators: a propylene/propane
splitter, a xylene/toluene vacuum column,
a refinery depropanizer, and a main fractionator
from a FCC unit. In addition, we are studying
extractive distillation, azeotropic distillation,
and reactive distillation. For each of
these columns we are identifying
the best control configuration as
well as the relative performance
of conventional PI controls, advanced PI
controls, model predictive control (e.g.,
DMCPlus), and a variety of nonlinear controllers.
Finally, we are studying a series of columns
(a demethanizer, a de-ethanizer, and a
depropanizer) in order to study
the effect of constraint control
on the overall control performance
using advanced PID and DMCPlus. We benchmark
our simulators against industrial data
to ensure that our simulators represent
industrial distillation columns.
pH Control:
pH control is one of the
most challenging process control problems
due to the extreme nonlinearity that
can result and the enormous disturbances
that occur. We are using detailed
simulations and a laboratory system
to study waste acid neutralization
using an in-line process. The in-line
process is desirable because it requires
much less capital than the conventional
design with well mixed reactors, but
is a more challenging control problem.
Reactor Control:
Temperature control of reactors is
vitally important to the production
rate and quality of many final products
in the chemical industry. We are
studying temperature control of a variety
of industrial fixed bed reactors. In
each case we are working with industry
to benchmark our simulators and ensure
that they properly represent the
industrial problem. We are also evaluating
the choice of manipulated variable
for temperature control of industrial
CSTR’s.
Facilities:
We have a Linux network of 10 Pentium
PC’s and a Dec Alfa and a Windows
NT network of 5 Pentium PC’s that
we use for running our simulation
programs. In addition, we are one
of only two US schools that has a
DMCPlus (Aspentech) site license.
[DMCPlus is the leading advanced
control software in the world.
Process Optimization
Facilities:
We have a Linux network of
10 Pentium PC’s and a Dec Alfa
and a Windows NT network of
5 Pentium PC’s that we use
for performing our optimization
studies.
General:
Process optimization involves
maintaining a process at
the economic optimum operating
conditions in the face of
changes in feedstock and
changes in the product pricing
structure. Due to the scale of
most chemical processes, the economic
benefit associated with typical
optimization projects, which usually
yield in the area of 5% improvement,
can be enormous even though
the capital requirements
are relatively small. Therefore,
process optimization is expected
to become even more important
in the near future. Areas
of concern for industry
include what degree of model fidelity
is required to reap the benefits
of process optimization and
can simplified approaches
to process optimization
yield very nearly the same benefits
as an optimization approach
based on rigorous first principles
modeling. These questions
and others we are studying
in our research effort.
Refinery Optimization:
We are currently studying the
unit optimization problem for
crude units, FCC units, reformer
units, hydrocracking units,
alkylation units, and gasoline
blending operations. In these
studies we are developing detailed
models for each unit and studying
the optimization of each unit
by itself. In these optimization
studies we consider the incremental
benefit associated with "operator" optimization,
constraint control, off-line
optimization, and on-line optimization.
In addition, we are studying
the problem of refinery-wide
optimization in which the total
refinery is considered in the
optimization analysis. Refinery-wide
optimization is just now being
attempted by industry and represents
a whole new class of optimization
problems. We are currently
working with refining companies
to ensure that our studies
are industrially relevant.
Ethylene
Plant Optimization:
We are currently
studying
ethylene plant optimization.
We are applying our optimization
models to a Dow Chemical
ethylene plant. From this
study, we will
determine how our process
models need to be improved
in order to
effective study this class
of process optimization
problems.