REPORT ON DYCORD+'95
The fourth IFAC (International Federation of Automatic Control)
Symposium on Dynamics and Control of Chemical Reactors, Distillation
Columns and Batch Processes was held in Helsingoer, Danmark on June 7
- 9, 1995. This series of conferences is held every 3rd year since
1986 (the last meeting was in Maryland, near Washington, 1992).
Chairman and editor of the international program committee was James
B. Rawlings, University of Texas at Austin. Co-chairman and
co-editors for the main sessions was: Wolfgang Marquardt (Reactors),
Sigurd Skogestad (Distillation), and Dominique Bonvin (Batch). Sten
Bay Joergensen (DTU, Lyngby) was chairman of the organizing committee.
There were about 130 delegates from 21 countries present. 15 from
Germany, 15 from USA, 7 from Finland and only 2 from Sweden. Norway
was represented by:
Bjarne Foss (Professor, Engineering Cybernetics - NTH)
Stein Wasboe (Ph.D. student, Enginnering Cybernetics - NTH)
Helge Hole (M.Sc., Hyprotech Scandinavia)
Sigurd Skogestad (Professor, Chemical Engineering - NTH)
Bernd Wittgens (Ph.D. student, Chemical Engineering - NTH)
Katrine Hilmen (Ph.D. student, Chemical Engineering - NTH)
plus two "half"-Norwegians
Elling W. Jacobsen (KTH)
Eva Sorensen (Imperial College)
Unfortunately, Prof. Jens Balchen could not participate because he has
had a heart attack and is now recoverying at home in Norway. Still,
some of the main ideas of his scheduled presentation "How Have We
Arrived at the Present State of Knowledge in Porcess Control ? Is
There a Lesson to Be Learned ?" was given by Sigurd Skogestad. All
the participants at the conference signed a huge "get well" card that
was sent to Jens Balchen on Thursday 8.
The DYCORD+'95 Symposium was focused on three important areas of
process dynamics and control:
(1) Chemical Reactors
(2) Distillation Columns
(3) Batch Processes
There were plenary sessions from 8.30 to 13.00. Then after lunch,
there were three parallell sessions on the subjects named over from
14.30 to 16.30 and poster presentations from 16.30 to 18.30 the first
day. I will give a short summary of the three days of the symposium.
Wednesday, June 7
-----------------
*** Dr. Babatunde A. Ogunnaike (DuPont Central Science & Eng., USA) held a
very interesting talk on "Industrial Perspectives on Process Control".
He adressed the problem of collaboration between industry and academia
and the R&D and manufacturing groups in industry. Often the R&D group
in a process industial company are called "recommend and dissapear" by
the manufacturing people ("They recomend a Kalman filter, but don't
stay the night-shift to se if it works...").
According to Babatunde Ogunnaike, the "people in industrial R&D
understand the language of the most abstruse academic theoretician and
can also communicate with the most unrepentant practitioner (the one
who is naturally suspicious of anything more complex than a PI
controller)". He pointet out that the main responsibility for making
better communication between academia and industry lies with the
people working in R&D. They play a two-sided role and have to
"interlink with both academia and the mechanics", he said.
It is clear that the "process control people" both in industry and in
academia are aware of the need for better transfer of knowledge and
stronger collaboration to be able to implement more of the new results
in the field of process control research. Babatunde Ogunnaike pointed
out three specific recommendations to improve this:
1. Industry must be more open (give academic researchers access to
problems of real significance, and must be more willing to implement
new technology)
2. Academic researchers must refocus their efforts (familiarity with
the peculiarities of the industrial process is the Process Control
researcher's competitive advantage over other control theorist)
3. The academic/industrial partnership should be strengthened
(Govermental founding of joint industrial/academic collaborative
research will strengthen this partnership further)
Sigurd Skogestad will be spending two weeks with DuPont (D. Smith,
B. A. Ogunnaike) at the end of July 1995.
*** Wednesdays subject for the plenary sessions was "Chemical
Reactors". Hans-Joerg Preisigke (Bayer AG, Germany) held a
interesting presentation on "Computational Fluid Dynamics as a Tool
for Chemical Reactor Optimization".
*** Elling W. Jacobsens held his talk on input multiplicity in ideal
two-product distillation. *** Eva Soerensen presented the results on
"Comparison of Inverted and Regular Batch Distillation". Both
presentations were very good.
*** At the poster session, the poster on "Neural Network
Identification of an In-Line pH Process" S.K. Dohery, J.B. Gomm and
D. Williams (LiverpoolJohn Moores Univ.) got a lot of interest.
Thursday, June 8
----------------
Thursdays subject for the plenary sessions was "Distillation Columns".
*** The first talk was given by Sigurd Skogestad under the title "A
System's Approach to Distillation - the Rebirth of a Research Area" -
or - "Bits and pieces about distillation". Here he showed some of the
complex behaviour that may occure in a 'simple' distillation column
and overruled some of the myths in control of distillation columns. He
also gave some comments on the understanding of Petluyk columns
('parallell paths') and showed some recent simulation results on the
multivessel batch column design. The overall conclusion of his talk
was that distillation is an old but still active research area ('not
dead yet'). Sigurd got very good responses on his talk.
*** The second plenary talk was held by David Q. Mayne (Univ. of California
at Davis) gave a talk on "Optimization in Model Based Control" where
he stated that with nonlinear control for nonlinear constrained
problems one could achieve a much better performance of the controller.
Plenary session 3 was on distillation column. Five papers were
presented covering a rather broad area in distillation, with Sigurd as
chairman.
*** Manfred Morari (California Institute of Technology - at
present: ETH in Zurich) presented a paper on "Multiple Steady States
in Azeotropic Distillation and their Effect on Column Operation and
Design". Here it is given a method for prediction of the existence of
multiple steady states in azeotropic distillation in the case of
infinite reflux and an infinite number of trays. This has relevant
implications for columns operation at finite reflux and finite number
of trays.
*** Shinji Hasebe (Kyoto Univ. - Japan) held a very interesting talk
on a new type of separation system: multi-effect batch distillation
(MEBAD). The separation performance is compared with a continous
distillation system. A simulation study showed energy savings and
better separation for the multi-effect batch distillation column for
mixtures with more than five components. He also showed that the MEBAD
is very easy to control and he proposed a single loop cascade control
of the holdup of the vessels by temperature-control (inner-loop) and
level-control (outer-loop). The conclusion of his talk was that a
multi-effect batch distillation system has wide applicability as a
flexible and energy-efficient separation system. Here in Trondheim, we
have proposed a new simple way of controlling such columns, and Rajab
Litto will implement this experementally as part of his Ph.D. thesis.
*** Bernd Wittgens' presentation on evaluation of dynamic models of
continous distillation columns went very well and he got some
questions and comments from the audience. *** Thomas J. McAvory
(Univ. of Maryland) gave a benchmark problem on plantwide control.
In the afternoon parallel sessions were held. The presentation of the
benchmark problems for the distillation session gave some fruitful
discussion, but the time avaible for this was maybe insufficient?
At the dinner at Kronborg Castle, Jim Rawlings gave some special
thanks to the international program committee. Sigurd Skogestad got a
Viking-helmet with the inscription "The paty begins here". (He lived up
to the slogan later on the evening.)
Friday, June 9
--------------
*** Fridays plenary sessions were on "Batch Processes". E. D. Gilles
(Stuttgart) held a talk on "Some Aspects of Integrated Process
Operation", which was not on batch....
The parallel sessions were reduced to "Chemical Reactors" and
"Distillation Columns", and (as usually for the last day) some of the
participants had left already. The SAS strike made severe problems for
some of the participants flying with SAS (like me!). Others rebooked
their flight to another company or stayed the night in Cobenhagen to
get a plain the next day.
In summary, the DYCORD+ series provides a good opportunity for the
modelling and process control communities to meet each other. The
DYCORD+'95 Symphosium was well organized with high quality
presentations. Attendance of participants both from industry and
academia could make it easy to get new impulses for research
work. However, the participants from industry were mainly consuming
the presented material, so that it was still difficult for the
academic community to figure out were the "real" problems are.
From the meeting it seems clear that distillation column research is
"not dead yet', wheras it seemed like the activity on reactors is
somwehat less than some years ago.
The next IFAC Symposium on Dynamics and Control of Chemical Reactors,
Distillation Columns and Batch Processes, DYCORD+'98, is to be held in
May 1998 at Corfu, Greece. :)
Comments:
---------
We will put all the papers from Proceedings in a new
UNIX litterature database file made for the Process Control Group.
Proceedings can be found at the Process Control Library at the
Department of Chemical Engineering at NTH.
Trondheim, 15. juni 1995,
-Katrine Hilmen