Chemical Engineering August 2014 - 37

Adoption of AdvAnced control technologies
A
dvanced control techniques offer opportunities for effectively
handling combustion systems that have strong nonlinearities
or internal couplings, and that therefore cannot be satisfactorily
handled by the commonly applied traditional PID (proportional,
integral derivative) controller, which is characterized by
linear and scalar behaviors. Published data are available on the
shares of advanced control methods for several years up to 2008,
which clearly show that the application of advanced control methods
for power plants and industry boilers are far below those in
the chemical, oil-and-gas and petroleum refining industries, for
example. In 2008, the power industry spent the most on DCS (distributed
control system) upgrades and replacements, while it spent
the least on advanced process control (see Table 1 in Ref. 18).
What reasons could explain these data? Instead of exact answers,
some differences will be identified and discussed here. It
may be that combustion processes do not dictate so definitely the
applications of advanced controllers, because some other chemical
processes are characterized by much stronger nonlinearities.
But in reality, combustion processes are also nonlinear, especially
when considering their entire load ranges. It is true that these nonlinearities
are less significant, and they predominantly can still be
kept in operation by means of traditional linear tools. However,
industrial practice shows clearly that this approach leads to efficiency
reductions already in the simplest case of a power plant
boiler running below its base load.
Table 1. RaTios of woRldwide sales of aPC
RelaTive To woRldwide sales of dCs*
industry
Ratio of APC
sales over
DCS sales
Refining oil and
gas
0.092
0.061
Chemical Power
0.050
0.010
* Ratios of worldwide sales of APC (advanced process control) relative
to worldwide sales of DCS (distributed control systems) in several industries
in 2008. Adapted from Ref. 18.
In some cases of combustion technologies, the suboptimal efficiency
associated with traditional control methods is no longer acceptable,
or the traditional control methods simply do not work. In these cases,
the application of advanced controllers is dictated already. Many
such examples were reported together with success stories of solving
them [21-24]. They include a wide range of combustion processes,
from compensating for unacceptable fluctuations of biomass boilers
with soft sensors, to lambda control of internal combustion engines
by means of a robust predictive controller.
A further question is, of course, what factors block the dissemination
of advanced control methods in those cases in which an obvimore,
MPC is also able to consider
other constraints without direct relationships
to the process in question.
If the control system is well designed,
this ability can be perfectly
utilized for limiting temperature
stresses in key structural elements
of a combustion system during load
shifts, for example.
Another advantage of MPC
against the traditional PID is that
is it multidimensional. As a result,
internal cross-couplings of the combustion
process can be handled
inherently. There is no need for
figuring out the significance of the
cross-couplings. An example is the
cross-couplings between load conous
limitation of traditional control may not necessarily be present,
but where advanced control methods would assure significant increases
in efficiency and flexibility, or decreases in lifetime consumptions
of critical parts.
For collecting the answers to this question, and for formulating
some advice on overcoming most blocking factors, a workshop
took place in May 2012 at the Budapest University of Technology
and Economics. A group of experts in three categories - plant
operators, equipment suppliers and researchers - came together
from four countries to discuss the issue in an unstructured, interactive
format. From the workshop, a number of basic statements
emerged. They are summarized here.
An important reason for the low level of advanced control methods
in combustion processes in the energy industry is the presence
of gaps between several actors. A lack of communication exists
among people and institutions working on the same topic. Key
people involved working on industrial combustion projects in the
areas of control, energy, management and financials are often
unable to balance benefits and drawbacks from the perspective of
each of their separate fields. Several platforms for bringing people
together could help the situation, as could the use of a common
language. This common language could be money. However, expressing
certain benefits (such as safety, usability and lifetime reduction)
in the context of money is not straightforward. Education
could also help in connecting control theory with system theory,
university with industry, and practice with theory.
Safety is another issue that slows down the utilization of advanced
control in combustion processes. Safety is the highest priority,
however, it can be warranted not only by traditional controllers
but also by means of a proper structure in the control system. Traditional
controllers should be used on a higher, supervising level,
and very clear interfaces should be built with the advanced control
elements [21]. Consider that this approach will also assure further
benefits. Additional certification costs for safety-related control
elements can be avoided, as can some of the risk arising from
frequent version changes of the hardware and software platforms
running the advanced control computation methods.
Software elements of advanced decision makers must follow the
same safety-related requirements as their hardware components,
as discussed above. That's why, also the above discussed considerations
in applying them should be followed in this case - especially
regarding the structure and clear interfaces joining traditional
and advanced elements. The same idea also suggests that
the range of control tasks to be realized on the basis of advanced
methods must be selected carefully. Positioners of actuators are, for
example, advised to be kept unchanged for simplicity and reliability.
Also, simple tasks of low significance should not be involved in
the area of tasks to be realized by advanced control methods. ❏
trol and lambda control in a boiler.
The basic MPC algorithm itself
is a discrete time on-line optimization,
which requires a dynamic
process model. The more accurate
the model, the better the control
performance that can be expected.
However, model inaccuracies will be
effectively compensated by the controller.
Prior to the controller design,
the constraints must be given, along
with two further parameters, the socalled
weight matrices. They form a
very clear representation of the design
criteria for balancing between
two contradictory requirements of
all control tasks - accuracy and
quiet actuator movements.
Intelligent control methods.
With respect to control methods,
the term " intelligent " suggests
those with origins in artificial intelligence
research. Two basic control
methods fall into this category: artificial
neural networks and fuzzy
control. Both are excellent tools for
controller design in cases where the
formulation of control-system rules
on the basis of traditional mathematical
formalism is very difficult.
In many situations, the control
rule is available by verbal statements
only, which arise from human
thinking. This set of control rules can
be called " expert knowledge, " and a
very effective way of representing
ChemiCal engineering www.Che.Com august 2014 37
http://www.Che.Com
Chemical Engineering August 2014

Table of Contents for the Digital Edition of Chemical Engineering August 2014
