**** EUROPEAN FORUM FOR COMPUTER-AIDED PROCESS ENGINEERING RESEARCH ****
(CAPE FORUM- 4 TH EVENT)
ENSIGC-INPT, 18 chemin de la Loge, TOULOUSE, France
on
Friday, Saturday and Sunday
22, 23 and 24 March 1996
KRISTIAN M. LIEN, APRIL 25. -96
This year's CAPE Forum followed the previous meetings in Trondheim, -93,
Aachen, -94 and Edinburgh, -95. The purpose of the forum is to meet informally
once a year and discuss matters of common interest to the participants, and
help contribute to more collaboration in the CAPE area in Europe. Here are
my perceptions of what went on:
In addition to approx. 10 participants from Toulouse, the foolowing people were
present:
David Bogle (University College London, UK)
Bertrand Braunschweig (Institut Francais du Petrole, France)
David Cameron (Norsk Hydro a.s., Norway)
Stephane Dechelotte (ProSIM S.A., France)
Waquar Elahi (BP Chemicals, UK)
Rafiqul Gani (Technical University of Denmark, Denmark)
George Heyen (Universite de liege, Belgique)
Antonis Kokossis (University of Manchester, UK)
Andrej Kraslawski (Lapeenranta University of Technology, Finland)
Zdravko Kravanja (University of Maribor, Slovenia)
Kristian Lien (Norwegian Institute of Technology, Norway)
Sandro Macchietto (Imperial College, UK)
Tom Malik (ICI Engineering Technology, UK)
Wolfgang Marquardt (RWTH Aachen, Germany)
Thor Mejdell (Norwegian Institute of Technology, Norway)
Costas Pantelides (Imperial College of Science, UK)
Tony Perris (Divocina, UK)
L. Phan (CGE, France)
Christian Schulz (Universitdt Dortmund, Germany)
Pierre Sere-Peyrigain (ProSIM S.A. France)
Sigurd Skogestad (Norwegian Institute of Technology, Norway)
Hans Skrifvars (Abo akademi University, Finland)
Eva Soerensen (UCL, London))
Esben Sorensen (Haldor Tops_e A/S, Denmark)
Tapio Westerlund (Abo Akademi University, Finland)
As you see, this means that we are getting closer and closer to our goal to
have an audience that is equally distributed between academics and
non-academics.
THE PROGRAMME, WITH SOME COMMENTS:
Saturday, March 23rd , 1996 :
9.00-10.45 Welcome
News from Bruxelles (European funding, CAPE actions and Networks)
(30mn) more especially CAPE.NET Thematic Network
Tony Perris- Divonica (UK)
David Bogle- University College of London (UCL, UK)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Here, David presented the CAPE.NET initiative and informed about the goals
and general setup of a network to be funded by Brite/EuRam (hopefully).
The first phase has already been funded, and interviews with prospective
industrial participants have been carried through. David handed out the
questionnaire that has been used in the companies, and asked all participants
to fill it in. Tony would then compile the results and compare with the
responses received from industry at the end of the meeting (see below).
-----------------------------------------------------------------------
Presentations concerning industrial activities and needs
(30mn) "CAPE OPEN" : B. Braunschweig (IFP-Rueil-Malmaison)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Bertrand presented an initiative involving major players in European
Chemical industry together with the major (alas, USA-owned) software houses.
They plan to spend a lot of money on coming up with standards for how to
create more open simulation environments.
Personally, I'm a little sceptical to this. In particular the fact that they
plan to use EU money to consolidate the position of the major US based vendors.
Secondly, I'm not sure to what extents their standards thinking is sufficienly
directed towards the future (Is this another HD TV kind of project ?)
Anyway, it is probably worth watching what happens with this initiative.
------------------------------------------------------------------------
10.45-11.00 Tea/coffee break
11.00-12.30 Short presentations about CAPE field of applications :
(about 20 mn including one pertinent question)
Quality product/quality design :
Rafiqul Gani (Technical Un. of Denmark, Denmark)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Rafique talked about interactions between product and process design.
(How to design products with desired properties that are also easy
to produce.)
---------------------------------------------------------------
Complex distillation systems :
Sigurd Skogestad (Norv. Instit. of Techno., Norway)
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Sigurd gave an interesting talk on new kinds of batch distillation columns,
with intermediate vessels. Good response from the audience.
-------------------------------------------------------------------
Fine chemical /Biochemical processes :
David Bogle (UCL, UK)
Environmental Aspects and applications :
L. Phan (Compagnie Generale des Eaux-Silic, France)
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Overview presentations. Phan talked about challenges in design and
control of municipal waste water treatment.
-------------------------------------------------------------------
12.30-14.00 Lunch at the school residence restaurant
14.00-16.00. Group round table session on CAPE advanced methodologies ,
tools and emerging new areas in CAPE applications
(one plenary session (1 hour) and two Parallel sessions,
each session 45 min, following by reports from chairmen)
in plenary Process Synthesis/Optimisation :
Kristian Lien (Norv. Instit. of Techno., Norway),
Zdravko Kravanja (Un . of Maribor, Slovenia)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Zdravko gave an overview of methods and requirements.
I (Kristian) presented our new method for design of multifunctional units.
Good response, several questions afterwards, and heated discussions in the
afternoon.
Costas Pantelides (Imp. College) presented a new optimization method
that has much fewer
pre-assumptions about structure than I have ever seen before. Interesting.
Seems to require quite some computational power.
Antonis Kokossis (UMIST) presented results from a new method for synthesis of
multiphase reactirs. (Unfortunately, he was very secretive about the method,
he only wanted to show the results.)
Tapio Westerlund (Aabo) showed results from an application they have been doing
in production planning in a paper mill. Very impressive results. Has already
saved a lot of money.
Bill Daly (Uniliver) presented a new design environment under development
internally in Uniliver.
To me this seemed to be similar to a.k.a. "browser" (like Netscape ?) that
acts as a common front end towards various design tools in Uniliver.
An ambitious project. Worth to follow up on what happens later here.
-------------------------------------------------------------------------
in parallel Numerical techniques & Strategies
(DAE/PDAE/IPDAE ; EO , Modular) :
Costas Pantelides (Imperial College, UK )
Wolfgang Marquardt (RWTH Aachen, Germany)
Fine Chemical processes/clean processes/
Environmental & Safety applications
MV Le Lann (ENSIGC-INPT, France)
Tom Malik (ICI Engineering, UK)
15 min chairman report:summary of the round table discussions
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Here I followed the first parallel. A lengthy discussion about use of
more detailed models. In my opinion, not too much new here.
------------------------------------------------------------------
16.00-16.30 Tea/coffee break
16.30 -17.30 Round table discussion (parallel session)
in parallel Batch processes/Manufacturing systems/Hybrid systems :
Sandro Macchietto (Imperial College, UK)
Herve Pingaud (ENSIGC, France)
Data reconciliation, Parameter estimation,
On-line applications
George Heyen (Un. de Liege, Belgique),
David Cameron (Norsk Hydro, Norway)
Chairman report : summary of the round table discussions
++++++++++++++++++++++++++++++++++++++++++++++++++++++++
I followed the first parallel here. At this point, the audience
was quite "tame". I did not discover any really significant news.
The other parallel reported that they had mainly been discussing data
reconciliation, and concluded that this is an area which is still
not very well understood. The recommendation was to use models for
reconciliation with no adjustable parameters, in order to "tune the
measurements to the model, and not the other way around."
-------------------------------------------------------------------
Sunday, March 24th , 1996
9.30-11.00 New tools and Emerging technologies for Modelling and Simulation
Environments :
(about 30 mn including more than one pertinent question)
Modelling Environment :
W. Marquardt, (RWTH Aachen, Germany)
Object-oriented Modelling & Simulation :
L. Jourda (ENSIGCT-INPT, France)
Open discussion & flash presentations
by volunteers from the audience
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Wolfgang made an interesting presentation of his research in the area.
The "conceptual model" that he proposes looks strikingly similar to
the Blackboard models from AI research in the -80ies, in my opinion.
(My own dr.ing. work, for example)
Wolfgang and I do not agree at all points about the role of traditional
object oriented methods in modeling. Wolfgang uses it. I claim that one
should instead be using a more flexible approach, where evolution of
class definitions is possible (e.g. protype based approaches). Wolfgang
disagrees because he feels that the lack of structure that these systems
have, will compromise consistency. My response is that while you should
strive for consistency on the final model, the struggle for consistency
WHILE DEVELOPING the model is not only a dream, but a nightmare:
In my opinion, modeling is learning about the problem. Learning means being
open to what you could not predict. Fixing the object classes means that you
have to anticipate what you will learn. Therefore, in my opinion there is a
basic choice to be made: EITHER you have guaranteed consistency (which would
be nice), OR you have the possiblity to learn (which is crucial). Asking for
both seems to me like trying to be just a little bit pregnant. I'm pretty
sure this discussion will be brought up again in the meeting next year.
(I have to admit, though, that the audience seemed to be quite provoked by my
suggestions. Which is what I hoped for.)
--------------------------------------------------------------
11.00-11.15 Tea/Coffee break
11.15-11.45 Future directions for CAPE research perception of
Industrial needs and opportunities :
Kristian Lien (Norv. Instit. of Techno., Norway),
Tony Perris (Divonica, UK)
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Tony went through the responses to the questionnaires (see above).
The industry responses and the responses from the Forum participants were
on most cases quite similar. This means that the forum should have a pretty
good chance to respond to the needs that industry have stated that they have.
-----------------------------------------------------------------
11.45-12.30 Other businesses, time to do the point, feedback, next meeting
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Everybody agreed that the French had organized a superb meeting.
The next meeting will be at DTU in Copenhagen approx. a year from now.
--------------------------------------------------------------
==================================================================
Sigurd og Thor: Send gjerne ut deres inntrykk, og. (Spesielt hvis der er noe
i min rapport dere vil korrigere.)
-Kristian