PROST - Annual Report 1996

The strong point center on process systems engineering (PROST) was established in June of 1994. It was awarded by NTNU and SINTEF, in recognition of the strong international standing of the group. The center receives one million NOK annually for 5 years from NTNU and SINTEF. PROST originally consisted of four professors at the Department of Chemical Engineering, NTNU, and Department of Chemical Engineering at SINTEF Applied Chemistry. In August 1995, PROST was expanded by including the process control activities at Electrical Engineering at NTNU and at SINTEF Automatic Control. The main research areas are:

w Process modelling, simulation and optimization, both stationary and dynamic

w Process control

w Process synthesis

w Process operations

About 50 persons are associated with PROST. The member groups of PROST are:

w Process Control Group, Dep. of Chemical Engineering, NTNU,

Professor Sigurd Skogestad, (head of PROST)

w Process Modelling and Simulation Group, Department of Chemical Engineering, NTNU,

Professor Terje Hertzberg.

w Process Synthesis and Optimization Group, Department of Chemical Engineering, NTNU,

Professor Kristian Lien.

w Process Systems engineering group, SINTEF Chemical Engineering.

Research Manager Ole Wærnes

w Systems Engineering Group, Department of Thermal Energy, NTNU

Professor Odd Andreas Asbjørnsen

w Process Integration Group, Department of Thermal Energy, NTNU

Professor Truls Gundersen

w Process Control Group, Department of Engineering Cybernetics, NTNU,

Professor Bjarne A. Foss

w Process Control Group, SINTEF Automatic Control,

Research Manager Peter Singstad

The group of Professor Asbjørnsen joined PROST formally in June 1996 .

In addition we are happy to have Professor Jens G. Balchen working closely with us.

A short summary of the activities are given below. In addition, there are separate reports from the groups and the Ph.D. students.

Bjarne Foss was on sabbatical leave 1995-96 at the University of Texas at Austin and Terje Hertzberg is on sabbatical leave 1996-97 at University of California at Berkeley.

An important event in 1996 was the publication by Wiley of the book "Multivariable feedback control-analysis and design" by Sigurd Skogestad and Ian Postlethwaite.

24 international journal papers and 18 conference papers were published by researchers, professors and students at the strong point center in 1996.

PROST employed a total of 26 Ph.D. students at the end of 1996, and 5 finished their Ph.D during the year. In total, 14 diploma students were associated with the center in 1996.

The SINTEF groups have worked on many projects in 1996. These are described in the reports from the two SINTEF groups.

One of our main goals has been to pass on information about our activities, and to establish a better communications with the industry and other users of our research.

PROSTs annual scientific meeting was held on the 11 April 1996, where the latest research results were presented. Further co-operation between the industry and PROST was discussed.

The PROST homepage was established in July 1996. Its present address is: www.kkt.chembio.ntnu.no/research/PROST

The PROST email list includes 19 industrial people who get updated information about the activities. Previous email messages are available on the PROST homepage. A four-page color brochure about the center was published in October 1996.

The Prost Seminar series was revitalized and we had 11 seminars with speakers from industry and academia during the autumn of 1996. A PROST workshop on on-line optimization with 20 participants was held in Trondheim on 28 Oct.1996.

27 - 29 May 1997 PROST organises the combined international conference "Process Systems Engineering" and "European Symposium on Computer Aided Process Engineering"

(PSESCAPE - 97), with Kristian Lien as general chairman and Sigurd Skogestad as the program committee chairman. 200 papers have been accepted for presentation and we expect about 300 participants. (the actual number turned out to be about 400)

At the end of 1996 the following persons were associated with the group:

w Sigurd Skogestad (Professor)

w Atle Christer Christiansen (dr.ing. student)

w Ivar Halvorsen (dr.ing. student)

w Kjetil Havre (dr.ing. student)

w Eva-Katrine Hilmen (dr.ing. student)

w Truls Larsson (dr.ing. student)

w Bernd Wittgens (dr.ing. student)

w Valeri Kiva (visiting professor)

w Tove Krokstad (secretary)

In addition, two part-time dr.ing. students are associated with the group:

Pål Flatby (Statoil) still has to finish up his dr.ing. thesis, whereas Audun Faanes (Norsk Hydro) still has not started his research work. Ingvild Løvik was associated with the group during the autumn of 1996 since Terje Hertzberg was on sabattical leave. John Morud (SINTEF/Norske Skog) and Morten Hovd (Fantoft) participated actively in the group meetings. Ying Zhao graduated and left Trondheim to go to Canada in September 1996.

The objectives of our research are:

w To educate engineers with a strong background in both chemical engineering and control. We attempt to bridge the gap in understanding, that has traditionally existed between the two disciplines.

w To perform high-quality and relevant research. In addition to being valuable in itself, research is vital in order to give out engineers the right training. It also keeps us up dated with international research, and thus enables us to bring international research to industry.

w John C. Morud, "Studies on the dynamics and operation of integrated plants". The sis submitted in December 1995 and defended 15 April 1996 with examiners Manfed Morari (ETH), Magne Hillestad (Statoil) and Odd Andreas Asbjørnsen (NTNU). John works for Norske Skog in Skogn, and we have been happy to have him located in our Department during the last year. He moved to Skogn in April 1997.

w Ying Zhao, "Studies on modeling and control of continuous biotechnical processes". Thesis submitted in August 1996 and defended 13 September 1996 with examiners Sten Bay Jørgensen (DTU) and Steinar Sælid (Prediktor AS). Ying Zhao presently lives in Toronto, Canada.

Five Diploma students graduated from the group in 1996:

1. Magnus Egge Andresen, Analysemodell for evaluering og optimal drift ved

Slagen-raffineriet. Utført ved Esso, Slagentangen

2. Berhanu Assefa Demessie, Experimental verification of multivessel batch distil-

lation.

3. Marius Støre Govatsmark , Modellbasert estimering for regulering av destilla-

sjonskolonne i metanolanlegg. Utført ved Statoil Trondheim.

4. Vegard Røbech Heggem, Kapasitetsøkning av etylen kjølekrets. Utført ved

Hydro Rafnes

5. Jørn Schumann, Optimal drift av destillasjonsanlagg for gjenvinning av løsnings- midler. Utført ved NTNU i samarbeide med Borregaard Fine Chemicals.

In addition, 7 students have done project work in cooperation with the group.

w The most important milestone in 1996 was the publication of the book "Multivariable feedback control - Analysis and design" by S. Skogestad and I. Postlethwaite (Wiley, May 1996, 572 pages). The book was very well received and an extensive home page with comments, corrections, solutions to exercises, computer (MATLAB) files, etc. has been set up on the internet.

w Sigurd Skogestad was in September 1996 appointed to Editor in the journal Auto matica, which is one of the most prestigious journals in the area of control theory.

w The group has during 1996 published 9 papers in international journals / book chap ters, and 6 papers at international conferences. A complete list of publications is en closed.

w We have established an internet web site where most of our latest publications are available, and also the six Ph.D. thesises from the last four years, see http://www.kkt.chembio.ntnuno/research/Process_Control, has been set up on the internet.

w A new project on complex (Petlyuk) distillation supported by the European Union' Joule project started in April 1996. The project is coordinated by Robin Smith at UMIST in Manchester, and Ivar Halvorsen has been most actively involved from our side. Our role is to investigate the operation, optimization and control of such units. The first progress meeting was held in Trondheim in December 1996.

Atle C. Christiansen: Studies on dynamics and control of integrated plants.

Atle started his work in 1994. He is now focusing his work on integrated (Petlyuk) distillation columns, and plans to finish at the end of 1997.

Pål Flatby: Modelling and control of distillation columns. Pål started his work in 1992. He is now back with Statoil Research Center in Trondheim, but has yet to write up his thesis.

Audun Faanes: Plantwide control. Audun is with Norsk Hydro Research Center in Porsgrunn. He has finished most of his course work, but he has not yet started his research

work.

Ivar Halvorsen (from 16 April 1996): Control and optimization of complex industrial processes. Ivar works with SINTEF Automatic Control, and was initially involved in the European project on complex distillation columns - DISC. From January 1997 he is financed by the Research councils reactor program (REPP).

Kjetil Havre: Control structure selection for integrated plants. Kjetil started his work in 1993, but he has had one year on leave. He spends part of his time at IFE at Kjeller near Oslo, and plans to finish up his work during 1997.

Eva-Katrine Hilmen: Batch distillation. Katrine started her work in 1994 and she is focusing her work on azeotropic batch distillation close cooperation with Dr. Valeri Kiva. She plans to finish her work in 1998.

Truls Larsson (from 1 May 1996): Optimization and control of methanol processes. Truls is working in cooperation with Statoil.

Bernd Wittgens: Experimental verification of distillation control theory. Bernds started in 1990 and his works involves experimental dynamic and control studies of continous and batch distillation. He plans to finish his thesis during 1997.

Dr. Valeri Kiva, Karpov Institute of Physical Chemistry, Head of Laboratory of Mixture Separation. Visit Oct. 1996 - June 1997 financed by PROST.

Conferences attended: see publications

The group is closely associated with the strong point center in process systems engineering in Trondheim (PROST), where Sigurd Skogestad is the head.

Sigurd Skogestad is the IPC chairman for the PSE-ESCAPE'97 symposium in Trondheim during 25-29 May 1997, and one of his main activities during the last part of 1996

was in supervising the review procedure for selecting abstracts and papers and for putting together the technical session program.

The organization of this symposium will be one of the main activities for the group during the first part of 1997.

Sigurd Skogestad chairs the university dr.ing. committee, and is the chairman of NTH's fund (from 1997). He is also member of Faculty Board (from 1996).

w We are a member of EU network on chemical process control.

w We are a member of the informal European network on computer aided

process engineering (CAPE network)

w DISC: We are involved in a new EU project on integrated "Petlyuk" distillation

which started in April 1996

w INCOOP: A new cooperation with AspenTech, Shell, Dow, Statoil, RWTH Aachen

and TU Delft is due to start in 1997.

w Valeri Kiva from Karpov institute in Moscow will spend the last part of 1996 with us and will stay with us until June 1997.

w Sigurd Skogstad was the examiner for two Ph.D. thesises:

1. Johan Pensar, "Parametric methods for optimaland robust control", Åbo

Akademi University, Feb. 1996.

2. Yi Cao, "Control structure selection for chemical processes using input-output controllability analysis", University of Execeter, March 1996.

w S. Skogestad, "A procedure for SISO controllability analysis - with application to design of pH neutralization processes", Comp.Chem.Engng., 20, 373-386, 1996.

w E.A. Wolff and S. Skogestad, "Temperature cascade control of distillation columns",

Ind.Eng.Chem.Res., 35, 475-484, 1996.

w J. Morud and S. Skogestad, "Dynamic behaviour of integrated plants",J. Process Control, 6, 145-146, 1996.

w S. Skogestad and K. Havre, "The use of the RGA and condition number as robust ness measures",

Computers and Chem. Engng., 20, Suppl., S1005-1010 (1996). (Supplement from symposium ESCAPE-6, Rhodes, Greece, May 26-29 , 1996).

w B. Wittgens, R. Litto, E. Sørensen and S. Skogestad, "Total reflux operation of mul- tivessel batch distillation", Computers and Chem. Engng., 20, Suppl., S1041-1046

(1996). (Supplement from symposium ESCAPE-6, Rhodes, Greece, May 26-29 , 1996).

w E. Sørensen and S. Skogestad, "Optimal startup procedures for batch distillation", Computers and Chem. Engng., 20, Suppl., S1257-1262 (1996). (Supplement from

symposium ESCAPE-6, Rhodes, Greece, May 26-29 , 1996).

w S. Skogestad and E.A. Wolff, "Controllability measures for disturbance rejection" Modeling, Identification and Control, 17, 167-182 (1996). (Reprint of paper pre- sented at the IFAC Workshop on interactions between process design and control, London, Sept. 1992;

w E. Sørensen and S. Skogestad, "Comparison of regular and inverted batch distilla- tion", Chem. Eng. Sci., 51, 4949-4962, 1996.

w R. D. Braatz, M. Morari and S. Skogestad, "Loopshaping for robust performance", Int. J. of Robust and Nonlinear Control, 6, 805-823, 1996.

w A.C. Christiansen, J.C. Morud and S. Skogestad, "A comparative analysis

of methods for solving systems of nonlinear algebraic equations," Proc. of the 38th SIMS Simulation Conference, Trondheim, June 1996, 217-230.

w K. Havre and S. Skogestad, "Input/output selection and partial control," IFAC World Congress, San Francisco, July 1996, Volume M, 181-186.

w K. Havre, J. Morud and S. Skogestad, "Selection of feedback variables for implement ing optimizing control schemes," UKACC Conference Control'96, Exeter, UK, Sept.

1996, 491-496.

w K. Havre and S. Skogestad, "Effect of RHP zeros and poles on performance in multi variable systems,"UKACC Conference Control'96, Exeter, UK, Sept.1996, 930- 935.

w K. Havre and S. Skogestad, "Selection of variables for regulatry control using pole directions, AIChE Annual Meeting, Chicago, Nov. 1996, Paper 45f.

w S. Skogestad, "Control structure design" AIChE Annual Meeting, Chicago, Nov. 1996, Paper 149a.

w S. Skogestad, "Control structure design", Åbo Akademi, Finland, 15 Feb. 1996.

w S. Skogestad, "Control structure design", Workshop on controllability analysis and plantwide control, DTU, Denmark, 21 May 1996.

w S. Skogestad, "Selection of feedback variables for optimizing control of Petlyuk col- umns", Workshop on Petlyuk distillation, NTNU, Trondheim, 9 Sep. 1996.

ANNUAL REPORT

By the end of 1996 the group consisted of five persons ( one professor and four Ph.D students).

w Terje Hertzberg, (Professor)

w Ingvild Løvik, Ph.D. student

w Hans Kristian Riksheim , Ph.D. student

w Sverre Støren, Ph.D. student

w Helge Tørvi, Ph.D. student

The objectives of the activities can be summarized as:

w Development of efficient methods and tools for process modeling, simulation and optimization through the synthesis of knowledge in mathematics, statistics, numerical methods, computer science and process technology. A thorough understanding of the different technologies and their interaction form the basis for our approach to the different problem areas.

w To educate candidates with a strong background in modeling of chemical engineering systems, from macro to full plant scale.

Cooperate with industry and the international research community in our area to ensure the quality and relevance of our work.

w Håvard I. Moe. The Thesis :Dynamic Process Simulation. Studies on Modeling and Index Reduction was defended in February with professor Wolfgang Marquardt, Aachen Univ. of Tech., Germany as the foreign external examiner. Håvard has joined Statoil in Stavanger.

w Terje Hertzberg was on sabbatical leave at the University of California at Berkeley from September 1^st. He is working with Prof. Alan Foss in the Chemical Engineering Department on a system for computer aided process modeling to be used in teaching. This is a joint project with Prof. Georg Stephanopoulos of MIT.

w Sverre Støren finished his thesis late 96 and has joined Statoils Research Center

in Trondheim. He will defend his thesis on March 18^th 97 with Professor Rafique Gani, DTU as the main opponent.

w Hans Christian Riksheim and Helge Tørvi are visiting Carnegie Mellon University for six month from October 96. Hans Christian has been working with Prof. Art Westerberg on partitioned integration of differential algebraic systems of equations. Helge has been working with Dr. Urmila Diwekar on uncertainty estimation in simulation of batch processes.

w In addition to H. Moes thesis two international journal papers have been published

by the group last year and four has been accepted for presentation/publication. A full list of the publications is included

w No diploma projects have been supervised by the group this autumn since Hertzberg, Riksheim and Tørvi have spend the period abroad.

Works with efficient dynamic process simulation and optimization. He finished his work this autumn and joined Statoils Research Center in Trondeim in September. His thesis is titled: Dynamic Process Optimization, Studies on Local Thermodynamic Models and Sensitivity Estimation. The defense is scheduled for March 97. In addition to finishing the thesis he has completed two publications one on sensitivity estimation that has been accepted for publication by Compters and Chemical Engineering and a presentation on localtherodynamic models in sensitivity estimation that has been accepted for PSE97/ESCAPE7.

Helge Tørvi

His field of interest is uncertainty in dynamic simulation. These calculations have mainly been done by Monte-Carlo methods, which can be slow and computationally expensive. Helge tries to develop methods with high accuracy, and still low computational demand. One new approach for calculation of uncertainties in dynamic simulation has been developed and tested. The method is based on Gaussian quadrature and uses the uncertain parameters pdf’s as weight functions.

w I lectured one undergraduate cource in process simulation in the spring semester.

w I have reviewed papers for Comp.&Chem. Eng., Biotechnology and Chem. Eng. and Processing.

w Assistant Dean of the Faculty

w Chairman "Education and Curriculum Committee" at the Faculty

w Chairman "Continuing and Further Education Committee", NTH

w Chairman "Conference Committee", NTH

w Coordinator NFR program "Process System Engineering"

w Program board INPRO

w Editorial board Computers & Chemical Engineering

w Editorial board Chemical Engineering and Processing

w ESCAPE 6, Sixth European Symposium on Computer Aided Process Engineering, Rhodes, Hellas, June 1996. (T. Hertzberg, H.C. Riksheim, I. Løvik)

w Inverse-Problems in reservoir evaluation, workshop, NTNU, Trondheim, May 1996. (H. Tørvi)

w The work in 1997 will follow the trend established the last years.

w Sverre Støren will defend his thesis March 18^th and Helge Tørvi and Hans Christian Riksheim plan to finish their work by the end of the year.

w The participation in the PSESCAPE-97 in May will be a major event this year with the group giving two presentations.

w Terje Hertzberg will return from sabbatical leave at The University of California, Berkely during the summer and will continue the cooperation with the "modeling project at MIT & Berkely

ANNUAL REPORT 1996

In 1996, the group consisted of the following members:

w Kristian M. Lien, Professor

w Steinar Hauan, dr. ing. student

w Harald Nordgaard Hansen, dr. ing. student

w Eirik Bjerve Sund, dr. ing. student

w Jahn Olaf Olsen, dr. ing. student

w Catharina Lindheim, dr. ing. student

w Erik Sauar, dr. ing. student

In 1996, the group consisted of the following members:

Kristian M. Lien (Professor), whose main research interests are in the areas of process synthesis, and in the interface between technology and work organization in process operations.

Steinar Hauan (dr.ing.student) develops methods for design of multifunctional units, with emphasis on reactive distillation. A close cooperation has been established with Prof. A.W.Westerberg's group at Carnegie Mellon University (CMU) in Pittsburgh in this area. Hauan spent approx. 6 months at CMU in -96.

Harald Nordgaard Hansen (dr.ing.student) develops models for optimization of synthesis gas processes. He spent the spring of -96 at Imperial College (IC) in London, visiting dr. Costas Pantelides, working with implementing his models in IC's new modeling system gPROMS.

Eirik Bjerve Sund (dr.ing. student) develops new methods for design and optimization of reactor systems, based on the Attainable Region approach. In -96 he has been partly funded by the Nordic Process Integration program, and he has spent some time in Aabo, Finland,

visiting prof. Tapio Westerlund's group.

Jahn Olaf Olsen (dr.ing. student) studies the interactions between organization and technology, partly funded by his host company Borregaard Chemcell Ltd. He is a member of

the crossdisciplinary INPRO group. He spent parts of 1996 in Pittsburgh visiting prof. A.W.Westerberg and his n-dim group at the Engineering Design Research Center at CMU.

Catharina Lindheim (dr.ing. student) studies the interactions between organization and technology, partly funded by her host company Statoil. She is a member of the crossdisciplinary INPRO group. She spent parts of 1996 in a DuPont plant in Canada (Gibbons, close to Edmonton), which has functioned as "model plant" for Statoil's new methanol plant at Tjeldbergodden.

Erik Sauar (associated dr.ing.student) develops new principles for energy efficient process designs. His main affiliation is with Physical Chemistry, and he applied principles from Irreversible thermodynamics to process design.

Hans Skrifvars (Åbo) and Jens Mikkelsen (DTU) have been visiting researchers in the group in parts of -96. Both are funded by the Nordic Process Integration program.

Torbjørn Pettersen joined Sintef Chemistry in -96, after being a postdoc in the group. He worked initially with modeling of membrane systems, and later also more in reactor modeling.

Students Graduated:

No dr.ing. students graduated from the group in 1996. Altogether six diploma students graduated from the group in -96, one of them a visiting student from Aachen.

Process Synthesis, Fall -96. (4. year chem.eng. course)

See below.

Prof. Lien has in 1996 been involved in a number of other activities, related to the area of Process Systems Engineering:

w General chairman of the 1997 joint PSE/ESCAPE-97 conference in Trondheim.

w Member (and co-founder) of the European Cape Forum. This informal network meets once a year at alternating places in Europe. It has recently had a formalized expansion in the CAPE.NET EU Brite/EuRam network program, where we are partners together with approx. 15 other academic and industrial groups.

w Member of the Board for NTNU's PAKT program (Coordination Science).

This research program works in the interface between Information Technlogy, Engineering and Social Sciences.

w Member of the Board for the Nordic Process Integration research program.

This program consists of nine doctoral students from chemical and mechanical engineering faculties in the Nordic countries.

w Organizer of The Nordic Membrane Seminar, Oslo, April -96.

(with T. Pettersen)

w Member of the EOLSS (Encyclopedia of Life Support Systems) Editorial

Board on Process Systems Engineering.

w Co-organizer of NHO/NITO study tour on new information technology,

USA, Febr. -96.

w Organizer of the "Technology and Social Sciences" seminar, Jektvolden, June -96. This is part of the INPRO / PAKT cooperation to establish a joint educational program on the interface between engineering and the social sciences.

In addition, he has also served on a number of different committees in 1996: The Faculty Research Committee and its Information Technology Committee, the Rector's Cross-disciplinary activities Advisory panel, NTNU's Kollegieraad and the Board of NTNU's Professorforum.

w The group has produced 5 journal and 6 conference papers in 1996, in addition to a number of presentations without formal papers:

w S.Hauan and K.M.Lien," Geometric visualisation of reactive fixed points", Comp.chem.Engng, Vol 20 Suppl., pp. S133-S138,

w S.Hauan, T.Omtveit and K.M.Lien, 'Analysis of reactive separation systems', Presented at the 1996 AIChE Annual Meeting, Chicago, paper 5f.

w Sauar, E., Ratkje, S.K., Lien, K.M., Equipartition of Forces - a new Principle for Process Design and Optimization, I&EC Research , 35, 4147-4153, 1996.

w Sauar, E., Rivero, R., Ratkje, S.K., Lien, K.M., Diabatic Column Optimization compared to Isoforce Columns. In Efficiency, Costs, Optimization, Simulation and Environmental Impact of Energy Systems 1996, Alvfors P., Eidensten, L., Svedberg,G., Jinyue, Y. (Eds.), ISBN 91-7170-664-x, Stockholm, pp. 635-640

w Sauar, E., Ratkje, S.K., Lien, K.M., Equipartition of Forces - a new Principle for Process Design, AIChE Annual Meeting , Chicago,November, 1996.

w H.N. Hansen, C.C. Pantelides and K.M. Lien. Modelling and Optimisation of Integrated Synthesis Gas Production Plants. AIChE Annual Meeting, Chicago, November 1996.

w Sund, E. B., Lien, K. M., "An optimal control formulation of the reaction-mixing problem" , AIChE-96, Chicago.

w Lindheim,C., Totland,T. and Lien,K.M., "Enterprise Modeling, - A new task for process systems engineers?", presented at the ESCAPE-6 conference, Rhodes, May 26th-29th 1996.

w J.O.Olsen and K.M.Lien, "Modeling for Improved Process Operation - A Socio- Technical Approach", AIChE, Chicago Nowember 1996.

w Lien, K.M. and T. Perris: "Future directions for CAPE research. Perceptions of industrial needs and opportunities." Comp.Chem.Engng. Vol 20, Suppl. S1551- 1557, Pergamon, 1996.

w Lien, K.M. and T. Perris: "Industry and Academia. Rebuilding the partnership", Comp.Chem.Engng. Vol 20, Suppl. S1545-1550, Pergamon, 1996.

w Lien K.M. "Towards Phenomena Based Design", Invited lecture, Carnegie

Mellon Univ., Chem. Eng. dept. Seminar Series, Febr. -96.

w Lien, K.M. "Phenomena based design", Invited lecture, Imperial College, Center for Process Systems Engineering Seminar series, Oct. -96.

w Lien, K.M., presentation at INPRO forum, Trondheim, febr. -96.

w Lien, K.M., presentation on Reactive Distillation seminar, Helsinki, June -96.

w Lien, K.M., presentation on BISP seminar, Porsgrunn, June -96.

In the spring, much of the available time will be allocated to the planning of the PSE/ESCAPE-97 conference. Most of the doctoral students will concentrate on the write-up of their theses in 1997, planning to finish in -97/-98.

Prof. Lien will in the fall go on Sabattical to Carnegie Mellon Univ. in Pittsburgh and work in prof. A.W. Westerberg's group there.

Ph.D. Research

ANNUAL REPORT 1996

The Process Cybernetics group is part of the Department of Engineering Cybernetics. The Department has a faculty of 7 professors, and approximately 40 graduate students working towards their Dr. Ing. degree.

The activity has the following objectives:

w Perform research focusing on modeling methods, and the application of these for plant operations. This includes control, surveillance, diagnosis, and optimization.

w Educate highly-qualified persons suitable for research and industrial challenges.

w Perform research with both high academic quality and industrial relevance.

In 1996, two candidates have graduated with a Dr. Ing. degree. These are:

w Grethe Tangen. The thesis 'Decision Making Support Applied to Hydropower Plant Upgrading' was defended in October with Prof. Morten Lind from The Danish Technical University as the main opponent. Grethe now works at EFI in Trondheim, Norway.

w Stein Ommund Wasbø. The thesis 'Ferromanganese Furnace Modelling Using Object-Oriented Principles' was defended in November with Prof. Bo Egardt from Chalmers Institute of Technology, Sweden as the main opponent. Stein now works with Elkem AS in Trondheim, Norway.

The group has 4 publications in international journals and as book chapters, and 5 papers published at international conferences in 1996. A complete list of journal and conference publications is enclosed.

w The development of a software package (ORBIT) for operating regime based model identification and gain scheduled control has continued.

The development is performed by Dr. Tor A. Johansen at SINTEF Automatic Control. The package is currently used in a joint project with Daimler-Benz and by other universities and research centers.

w Elkem AS is using a simplified version of the comprehensive ferro-manganese model developed by Stein at their Porsgrunn plant.

w The Kamyr Digester model and simulator form the basis for Dr. Finn Are Michelsen's private consultancy work. This simulator has in 1996 been distributed to two universities in the U.S.

w Grethe is pursuing her work on decision support at EFI in cooperation with Statkraft AS

Tormod Drengstig started his research in Feb. 94 and is financed by The Research Council of Norway through the Prosmat program and Hydro Aluminium AS. Tormod is working on a graphical process representation scheme with the aim to use this for model development. He has during 1996 also worked with modeling and estimation on an aluminium electrolysis process. The main focus has been on AlF$_3$ and temperature dynamics. Tormod will complete his work in 1997.

Kjell Støle-Hansen started his research in Sept. 94 and is financed by The Research Council of Norway through the INPRO program. He will emphasize improved plant operations by the use of dynamic models. Kjell is using precipitation and cementation processes at the Falconbridge company as cases, and has developed dynamic models and control structures for these processes. In 1996 he spent 6 months in Canada working as a control engineer at one of

Falconbridge's process plants. Kjell's plans for 1997 is to complete a controllability analysis and propose an improved control scheme for a precipitation process, and look into non-technological aspects concerning implementation of new control strategies in a process plant. Kjell plans to complete his work by March 1998.

Jo Simensen started his research in Sept. 94 and is financed by The Research Council of Norway through the INPRO program. Jo focuses on information systems supporting the operation of batch plants. In addition, Jo is investigating the role of modelling in the

interplay between a process plant and the organization running it. He is using a pharmaceutical plant at Nycomed Imaging AS as case. In 1996, Jo spent three months at the Department of Automatic Control at Lund Institute of Technology in Sweden. In cooperation with Karl-Erik

Årzén and Charlotta Johnsson, he implemented a prototype for an integrated batch plant operation system. Jo plans to complete his research in 1997.

Olav Slupphaug started his research in Feb. 95. He is working with control of hybrid systems, that is, systems which consist of both discrete and continuous parts. In particular he is looking into the use of model predictive control (MPC) for such systems. Olav had a three months leave (June-August) to develop an MPC tool for hybrid systems together with

Jostein Vada. From Sept. to Dec. Olav stayed with Prof. James B.Rawling's research group at

the Department of Chemical Engineering at the University of Wisconsin at Madison. Olav will complete his research in the spring of 1998.

Jostein Vada has been a teaching assistant and a doctorate student during 1996. He has followed several courses and taken exams. He starts his research period in 1997 in the area of analysis and control synthesis of complex process plants. This topic has many similarities with

the work being performed by Olav Slupphaug. Jostein expects to complete his research in the autumn of 1999

The group has participated at several conferences in 1996. These conferences are listed below, with the names of the people representing our group given in brackets.

w The sixth European Symposium on Computer Aided Process Engineering (ESCAPE-6), Rhodes, Greece, May. (Støle-Hansen, Simensen)

w IFAC World Congress, San Francisco, USA, July. (Foss)

w IEEE Int. Conference on Fuzzy Systems, New Orleans, USA, Sept. (Johansen)

w IEEE Int.Conference on Decision and Control, Kobe, Japan, Dec. (Johansen)

w The American Institute of Chemical Engineers (AIChE) annual meeting, Chicago, Illinois, Nov. (Foss, Slupphaug)

w IEEE Conf. on Systems, Man and Cybernetics, Vancouver, Canada, Oct. (Tangen)

Bjarne gave guest lectures at Georgia Institute of Technology, Atlanta, Auburn University, Alabama, Texas AM University, Texas, and The University of Texas, Austin.

A 3-day course in Optimization and Control was held with participants from Elkem AS in November.

We are engaged in several research programs and networks.

w We are an active member in the Strong Point Center in Process Systems Engineering (PROST) at the Norwegian University of Science and Technology.

w We are actively engaged in ORD, a departmental strategic research program.

w We are tightly linked to INPRO. The INPRO-program focuses on process plant operation, in particular the cross-disciplinary nature of this problem. There is a heavy engagement from three departments; Department of Organization and Work Science, Department of Chemical Engineering and our department. The INPRO-program is funded by the Federation of Norwegian Process and Manufacturing Industries (PIL),

The Norwegian Oil Industry Association (OLF), the Research Council of Norway and the nine companies that are associated to INPRO.

w We participate in COSY, a European research network program.

w Further, the group has close ties with other research groups at our department, and SINTEF Automatic Control.

Our activities are financed by a number of projects and fellowships. Most of the resources are supplied by The Research Council of Norway. Additional funding from collaborating industry is, however, appreciated as it gives us flexibility to pursue high-risk avenues, attend seminars and conferences, visit other research groups, buy equipment, and perform experiments.

Work in 1997 will develop along the same lines as in 1996. Two Dr. Ing.students (Tormod Drengstig and Jo Simensen) plan to defend their theses in 1997. Further, several publications, and conference presentations, invited and regular, are due next year. Further, we will continue the effort to transfer our research results to the industrial domain.

The main event in 1997 is a 2-day international workshop on 'Multiple model approaches to modeling and control' to be held in Trondheim in September.

The collaboration with the other members of PROST will continue, and cooperation activities will continue.

The INPRO-program will be used as a basis for expanding the activities within plant operations and to develop an inter-departmental curriculum on plant operations at the Norwegian University of Science and Technology.

Theses

w G. Tangen "Decision Making Support Applied to Hydropower Plant Upgrading",

Dr.Ing. thesis, Report. no NTNU 1996:73, ITK 1996:51-W

w S. O. Wasbø "Ferromanganese Furnace Modelling Using Object-Oriented Principle", Dr. Ing. thesis, Report. no NTNU 1996:80, ITK 1996:45-W

w T. A. Johansen "Identification of Non-linear Systems using Empirical Data and Prior Knowledge - An Optimization Approach", Automatica, Vol. 32, pp. 337-356.

w T. A. Johansen and P. A. Ioannou "Robust Adaptive Control of Minimum phase Nonlinear Systems", Int. J. Adaptive Control and Signal Processing,Vol. 10, pp. 61- 78.

w F. A. Michelsen and B. A. Foss "A Comprehensive Mechanistic Model of the Continuous Kamyr Digester", Applied Mathematical Modelling, Vol. 20, pp. 523- 533.

w F. A. Michelsen and B. A. Foss "A Dynamic Model of the Interaction between the Chemical Reactions and the Residence time in the Continuous Kamyr Digester", Tappi Journal, Vol. 79, pp. 170-176.

w T. A. Johansen, "Stability, Performance and Robustness of Fuzzy Model-Based Control", Proc. IEEE Conf. Decision and Control, pp. 604-609, Kobe, Japan.

w T. A. Johansen, "Robust Identification of Takagi-Sugeno-Kang Fuzzy Models using Regularization", Proc. IEEE Conf. Fuzzy Systems, pp. 180-186, New Orleans, Louisiana.

w K. Støle-Hansen and O. M. Dotterud, "Dynamic modelling of pH controlled precipitation and dissolution in an aqueous solution", Comput. chem. Engng., Vol. 20, Suppl. pp. S629-S634, Presented at ESCAPE-6, Rhodes, Greece.

w J. Simensen and B. A. Foss, "A Concept for Modelling Batch Plant Operation", Comput. chem. Engng., Vol. 20, Suppl. pp. S1245-S1250, Presented at ESCAPE-6,

Rhodes, Greece.

w G. Tangen and P. Støa, "Decision Support for Hydropower Plant Upgrading - Integrating Multi Attribute Decision Making and Knowledge Based Systems'', Intelligence Systems Applications to Power Systems (ISAP'96), pp.\ 145-150}, Orlando, Florida.

w B. A. Foss, "Optimization based control", The Texas-Wisconsin Modeling and Control Consortium workshop, Austin, Texas.

w B. A. Foss, "Modeling methods for first principles and semi-empirical models, COSY workshop on Mathematical Modeling of Complex Systems, Lund, Sweden.

w S. O.Wasbø, "Operator Support System for HC-FeMn furnaces", Manganese Symposium, Trondheim, Norway.

w O. Slupphaug, J. Vada "Predsim Users Guide. A Platform for Simulating MPC.

Release 1.0.", Report no. ITK 96-52-T, Department of Engineering Cybernetics,

Norwegian University of Science and Technology, Trondheim, Norway.

w T. Drengstig, S. O. Wasbø and B. A. Foss "A Formal Graphical Based Process Modeling Methodology", Report no. ITK 1996-41-W, Department of Engineering Cybernetics, Norwegian University of Science and Technology, Trondheim, Norway.

w J. Simensen, C. Johnsson and K.-E.Årzén, "A Framework for Batch Plant Information Models", Report ISRN LUTFD 2 TFRT-7553-SE, Department of Automatic Control, Lund Institute of Technology, Lund, Sweden.

w O. Slupphaug, "Model Predictive Control for a class of Hybrid Systems", Report no. ITK 1996-46-W, Department of Engineering Cybernetics, Norwegian University of Science and Technology, Trondheim, Norway.

Annual Report 1996

In the department of Chemical Engineering at SINTEF Applied Chemistry there is a group consisting of five research scientists which is involved in the strong point center PROST. The group which is lead by Dr. Thor Mejdell, consists of the following persons:

Telephone e-mail

Dr. ing. Thor Mejdell +47 735 94128 Thor.Mejdell@chem.sintef.no

Dr. ing. Magne Lysberg +47 735 94123 Magne.Lysberg@chem.sintef.no

Siv ing. Trude Sundset +47 735 92811 Trude.Sundset@chem.sintef.no

Siv ing. Terje Strøm +47 735 92812 Terje.Strom@chem.sintef.no

Dr. ing. Torbjørn Pettersen +47 735 94492 Torbjorn.Pettersen@chem.sintef.no

Jostein Sogge left the group for a position at Reinertsen in March 1996. Trude Sundset had a maternity leave from oktober 1995 to july 1996. She will leave the group in February 1997, and Arne Grislingås in March 1997, both for positions in STATOIL. The research manager will be Dr. ing. Ole Wærnes.

Sundset, Strøm and Pettersen have been integrated with the group of Prof. Kristian Lien, Lysberg and Mejdell with the group of prof. Terje Hertzberg, and Mejdell also with the group of prof. Sigurd Skogestad.

This project, which was funded by the Norwegian Research Council, was a technical evaluation of production of a hydrogen/nitrogen mixture for ammonia production. Comparison between membrane reactor technology and conventional Gas Heated Reforming (GHR) was made. The main part of the study was HYSIM simulations of various GHR concepts with either air or oxygen as feed to the secondary autothermal reformer. Pure membrane reactor steam reforming was also simulated. One important conclusion was that GHR with use of air as oxidant was not technically feasible if stoichiometric hydrogen/nitrogen is necessary. Use of enriched air with 30% oxygen is possible, however.

This is a part of an EU project hydrogen selective membrane reactors applied to synthesis gas production by steam reforming. The project, which is co-ordinated by Rune Bredesen, SINTEF Materials Technology, is financed within the Framework Program IV. Our task will be to perform design and technical/economic evaluations on the membrane reactor processes applied to hydrogen production as a feedstock to ammonia plants. This work is performed

in co-operation with Norsk Hydro in Norway, and with KEMA in The Netherlands. Most of the work in this project will be performed during 1997.

This project starded in 1996 (4.2 MNOK a year), and will run for six years. It is supported by PROSMAT program in NRC. The project is a cooperation with other departments at SINTEF and NTNU, and with Hydro PVC, Borealis, Jotun, Dyno and Statoil. Reactor modelling, thermodynamics and model based control is among the activities in the prosject.

In this activity in REPP we are modelling polymerisation reactors for the cooperating industrial companies. The models are implemented in MATLAB, and will be used for simulation, optimization and control.

This project is part of a long-term project on polymer technology. Last year's efforts were concentrated on literature studies which led to recommendations on how to proceed with developing tools for estimating phase equilibria and transport properties. Plans for doing selected experiments were also considered.

The aim of the project was to identify energy sources and requirements, and to develop suitable design solutions for utilisation of waste heat in an anode baking plant. This work will continue in 1997 with emphasise on implementation of a design solution at an anode baking plant.

One activity in the project is on developing a kinetic model which describes the solution of a mineral salt into a mineral salt melt as a function of parameters like temperature and composition of the mineral salt melt. The other part of the project is devoted to building and verifying a design model for a mixing process. Both theoretical and experimental work is part

of the project.

The goal has been to develop a pre-project targeted on small to medium size businesses (SMB). The pre-project is supposed to consist of an evaluation of the hole or parts of the production process in an SMB, a so-called "health-check". The goal of the project is to identify bottlenecks and the potential for improvement of the process. The work on preparing this pre-project has been financed through PROST, and has consisted of three main

activities.

* Identification of sub tasks to be part of a "health-check".

* Development of tools and knowledge needed for work with this type of industry.

* Preparing the marketing of the per-projects.

This work will be continued in 1997, with special emphasis on marketing of

the "health-check".

An evaluation of a production line for a dairy product as been performed at

two dairies. The aim of the work was to identify bottlenecks and problem

areas, and the potential for improvement of the production processes. This

project had the same aim as the "health-check" projects mentioned previously.

This is a joint project with Hydro, Statoil and Tel-Tek, and it is supported by NRC. The

activities consist of a) development of measurement equipment and numerical procedures

for estimating bubble size and rise velocity, b) verification of parameters in different types

of reactor models.

The work was performed for Hydro. The model consist of a one dimensional dispersion

model with heat exchange and was implemented in Matlab.

A technical and economic evaluation of some processes were made for STATOIL.

The objectiv in this project is to control and improve the quality of ruminant feed in an expander process. The expander treatement, increases the bypass of protein in the rumen. This year a factorial experimental design was performed, and the main effects of the expander on protein degradation in the rumen was revealed.

H.I. Farag, T. Mejdell, K. Hjarbo, P. Ege, M. Lysberg, A. Grislingås and H. DeLasa

"Fibre Optic and Capacitance Probes In Turbulent Fluidized Beds." Accepted for publication in Chem Eng Comm.

T. Pettersen, A. Argo, R.D. Noble, C.A. Koval, "Design of combined membrane and distillation processes", Separation Technology 6:175-187, (1996).

R. Bredesen, J. Sogge, "A technical and economic assessment of membrane reactors for hydrogen and syngas production", Seminar on the Ecological appl. of innovative membrane techn. in the chemical industry, May 1.-4., Cetraro, Italy (1996).

ANNUAL REPORT 1996

The PROST group at the Automatic Control department consists of the following persons:

w Department Manager Peter Singstad (Head of PROST group)

w Research Scientist Tor Steinar Schei

w Research Scientist Tor Ivar Eikaas

w Research Scientist Ivar Halvorsen

w Research Scientist Tor Arne Johansen

w Research Scientist Per Eilif Wahl

w Research Scientist Hans Berntsen

w Research Scientist Kjell Støle Hansen

REPP was established in 1996 as a co-operation between Jotun, Dyno, Norsk Hydro, Borealis, Statoil and SINTEF. The Automatic control department is responsible for activities within model based process control. General software tools for on-line state and parameter

estimation and non-linear model predictive control, previously developed at SINTEF, are refined and further developed for batch and continuous reactor control applications, with a focus on exothermic polymerization reactors.

DISC is a research project financed in part by the European Commission through the JOULE III programme. The objective of this project is to develop methods for design and control of the integrated "Petlyuk", or "dividing wall" distillation column. This column separates feed with three components into three product streams, within a single columns shell. Possible savings in both capital cost and energy consumption at about 30% have been reported. But in order to achieve this in practice, some kind of optimizing control will probably be required. Our focus is on control problems and control design.

The process control system for the Norne platform is tuned in advance of oil production start up. The tuning is performed on a combined engineering and training simulator previously developed at SINTEF. The tuning tool "MultiTune", which is also developed at SINTEF, is used for computation of controller parameters. A similar project has previously been done for the Heidrun platform. MultiTune will be integrated with the Norne process control system from Kongsberg Simrad in 1997. Further fine tuning will then be performed directly on the

Norne process plant.

MultiTune is a software tool for automatic tuning of conventional industrial process control systems consisting of PID controllers in single loops and cascades, and simple feed-forward and decoupling transfer functions.

ORBIT is a MATLAB based software tool for computer aided grey box modelling and parameter estimation developed by SINTEF and NTNU. The model representation is based on multiple local models. Furthermore, ORBIT contains methods for model based gain scheduled control design. ORBIT is currently being applied in a research project for Daimler-Benz.

The activities at SINTEF are concerned with the development of an ontology for modelling of operational procedures and functional requirements for process control of oil and gas processing plants. SINTEF has demonstrated an application where a file describing

the process topology is read, and functional requirements for control of this process is generated automatically. We have also demonstrated automatic generation of operational procedures based on a model of the process topology and functional structure.

w T. A. Johansen and B. A. Foss, "Operating Regime Based Process Modeling and Identification", Computers and Chemical Engineering, Vol. 21, pp. 159-176, 1997

w T. A. Johansen, "On Tikhonov Regularization, Bias and Variance in Nonlinear System Identification", to appear in Automatica, March 1997

w K. J. Hunt and T. A. Johansen, "Design and Analysis of Gain-Scheduled Control using Local Controller Networks", to appear in Int. J. Control, 1997

w T. A. Johansen, "Identification of Non-linear Systems using Empirical Data and Prior Knowledge - An Optimization Approach", Automatica, Vol. 32, pp. 337-356, 1996.

w T. A. Johansen and P. A. Ioannou, "Robust Adaptive Control of Minimum phase Nonlinear Systems", Int. J. Adaptive Control and Signal Processing, Vol. 10, pp. 61- 78, 1996.

w T. A. Johansen and M. M. Polycarpou, "Stable Adaptive Control of a General Class of Non-linear Systems", in "Neural Adaptive Control Technology I", R. Zbikowski and K. Hunt (Eds), World Scientific, pp. 177-206, 1996.

w T. A. Johansen, "Robust Identification of Takagi-Sugeno-Kang Fuzzy Models using Regularization", Proc. IEEE Conf. Fuzzy Systems, New Orleans, 1996, pp. 180-186.

w T. A. Johansen, "Stability, Performance and Robustness of Fuzzy Model-Based Control" Proc. IEEE Conf. Decision and Control, Kobe, 1996, pp. 604-609.

w R. Murray-Smith and T. A. Johansen (Editors), "Multiple Model Approaches to Modelling and Control", Taylor & Francis, London, 1997

I started my doctoral studies in March 1994 under the supervision of Professor Sigurd Skogestad and plan to finish within the end 1997. My doctoral work is financed by the NFR-supported project on computer aided process engineering.

The working title of my Ph.D. is "Studies on Dynamics and Control of Integrated Plants", in which the main objectives and aims are to contribute towards a greater understanding of the behavior of integrated plants. By integrated plants we understand here processes that are interconnected either in series or in parallel, or simpler processes that involve material recycle or heat integration.

In terms of research activities I have extended a previous work on dynamics of batch distillation columns, with emphasis on parametric sensitivity. We previously showed that columns operated under constant reflux and with mass inputs may become very sensitive to small changes in some operating parameters. The effects may be dramatic if the potential sensitivity is not considered during operation and without proper feedback mechanisms. We have now also included examples for columns operated with a pre-computed input trajectory to maintain constant distillate composition. We have showed that sensitivity is displayed in this case.

I also continued a work on numerical methods for solving systems of nonlinear algebraic equations (NAE's), in which I compared ahomotopy-continuation algorithms with a novel tear and grid method. A paper was written and published in the proceedings of the SIMS'96

conference in Trondheim. I have also done some work together with John Morud on making efficient continuation methods to explore solutions in the parameter space of systems of NAE's.

Finally, I have been working on modeling, optimization and design of a certain class of complex distillation arrangements, known as Petlyuk arrangements. The promising aspects of such columns lie in the potential for large savings in both capital and operational costs. We

have shown through simulations that it is possible to extend the Petlyuk ideas for ternary mixtures, to include separations of mixtures with four or more components. Two conference papers have been submitted, to be presented during 1997.

w Co-advised a project work for a 4th year project on reactive distillation columns.

w Organized the exercizes and a small project work for the process control course at the Chemical Engineering department.

w Held the lectures in the process control course, during Professor Skogestads attendance at a conference.

w Attend two workshops in an EU Joule project (Disc) project in Manchester (England) and Trondheim.

w Attended the ESCAPE'6 conference in Rhodes, Greece.

w Christiansen A. C., J. Morud and S. Skogestad. "A Comparative Analysis of Numerical Methods for Steady State Analysis : Tear and Grid Approach or Homotopy Continuation".

In the proceedings for the 38th SIMS Simulation Conference, pp 217-230, Trondheim, 1996

w Christiansen A.C., S. Skogestad and K. Lien. "Integrated Distillation Arrangements - Extending the Petlyuk Ideas". Accepted for presentation at the PSE'97-ESCAPE'7 in Trondheim 1997.

w Christiansen A.C., S. Skogestad and K. Lien. "Partitioned Petlyuk Arrangements for 4-component Separations". Submitted to the Distillation & Absorption'97 conference in Maastricht.

During 1997 I will finally publish the work on batch distillation. I will continue my work on Petlyuk arrangements, particularly focusing on issues related to design and optimization. I also aim to do a control study of one of the proposed arrangements. I will present a paper on the PSE97-ESCAPE'7, and one paper at the Distillation & Absorption'97. I also hope to defend my thesis within 1997.

I started my study in February 1994 and will finish during 1997.

My doctoral study is financially supported by The Research Council of Norway and Hydro Aluminium AS through the PROSMAT program. My supervisor is professor Bjarne A. Foss.

I am working on a graphical process representation scheme with the aim to use this scheme for model development of unit processes. An aluminum electrolysis cell has been used as a case study, and the main focus has been on AlF$_3$ and temperature dynamics.

My industrial contact is Hydro Aluminium AS.

w I wrote a report and a paper together with Stein O. Wasbø and Bjarne Foss on the graphical representation scheme. The paper is accepted at PSE-ESCAPE-97.

w During the autumn, I have developed and validated several dynamic models of the aluminum electrolysis cell. A conclusion of this work is that cell behavior is

dominated by energy disturbances and that AlF$_3$ disturbances have less influence. This implies that improved cell control can be achieved by performing tighter control and monitoring of the crust thickness. Based on these results, I wrote a paper to be

presented at Nordic Process Control Workshop-7 in 1997.

w Drengstig, T. S. O. Wasbø, B. A. Foss (1996), "A Formal Graphical based Process Modeling Methodology", Report 96-41-W, Department of Engineering Cybernetics, Norwegian University of Science and Technology, Trondheim, Norway.

I will continue to work on the modeling methodology, and I will start writing my thesis. I plan to be finished by the end of the summer.

My Ph.D. study started in June 1994 and is scheduled to finish in the summer of 1997. I am situated at Tel-Tek in Porsgrunn and my supervisor is Prof. Truls Gundersen. My co-supervisor is Prof. Sigurd Skogestad, and the work is financed through the Norwegian Research Council program in Computer Aided Chemical Engineering (DAAK).

The research focuses on further development of methods for design and analysis of integrated processes. Emphasis is on controllability, operability and flexibility of the prosesses, and the work is carried out in co-operation with department of Chemical Engineering at NTNU. Methods for optimal operation of heat exchanger networks are investigated, and the goal is to develop control strategies that maximize heat recovery together with reasonable dynamic behaviour of the process.

• Participated at ESCAPE 6 on Rhodos in May (with poster presentation).
• Participated (talks without written papers) at Plantwide Control an Controllability workshop at DTU, Denmark (May 20), Petlyuk workshop at NTNU (September 9) and On-line Optimization workshop at NTNU (October 28).
• In the spring I was co-supervisor for a diploma-student on dynamic modelling and simulation of heat exchanger networks.
• A method for on-line optimization of heat exchanger networks based on periodic steady state optimization and optimal choice of measurements has been developed.

• Glemmestad, B., K.W. Mathisen and T. Gundersen (1996). "Optimal Operation of Heat Exchanger Networks based on Structural Information". Presented at ESCAPE 6 on Rhodes, May 1996.
• Glemmestad, B., S. Skogestad and T. Gundersen. "On-line Optimization and Choice of Optimization Variables for Control of Heat Exchanger Networks". Accepted for presentation at PSE-ESCAPE'97 in Trondheim, May 1997.

Finish thesis and get a job.

I started my study in 1996 and will be finishing my work in 1999.

My supervisor is professor Sigurd Skogestad.

I am working on optimization and control of complex industrial processes.

Uncertainties and unknown disturbances makes optimization difficult to realize on a real plant. How should criteria be formulated? How should measurement and control structures be chosen in order to make optimizing control possible for complex processes?

How do we know if an action takes the process closer to optimum if we do not know for certain what the current state is, or where the real optimum is. Such questions are central for the problem area.

An interesting case is the optimal use of control inputs for the integrated "Petlyuk" distillation column. Possible savings in both capital cost and energy consumption at about 30% have been reported. But in order to achieve this, some kind of optimizing control will probably be required.

Chemical reactors is chosen as the second industrial case, in order to check generality of methods. Chemical reactors are very central process units in industrial plants, and such units have varying dynamic properties which makes control difficult.

The potential for application of results are good. Complex integrated units have often significant advantages regarding both capital- and operational costs. But in order to realize the economic potential, improved control and optimization are required. Thus better methods for control and optimization will increase the realizable benefit potential, and advanced control is very often a requirement for operation of a complex unit.

w Four courses have been completed.

w Controllability analysis of Petlyuk distillation columns has started. In this work we also cooperate with 5 European partners in the the European Community project: "Complex distillation columns, (DISC)", financed through the JOULE programme.

w Ivar J. Halvorsen and Sigurd Skogestad (1997), 'Optimizing Control of Petlyuk Distillation: Understanding the Steady-State Behavior.', Accepted for publication at PSE/ESCAPE 97 Trondheim May 1997.

I will complete the rest of my courses. The work on analysis and control og Petlyuk Distillation will be continued. I will start to look into chemical reactor control as the second major case.

I started my Ph.D. studies in August 1994 and plan to complete my thesis by September 1998. This doctoral work is financed by a scholarship from Telemark Institute of Technology. My advisor is Professor Truls Gundersen and Professor Kristian Lien is my co-advisor.

"Use of Mathematical Programming (MP) in Proces Integration" is the title of my Ph.D. work. Currently the focus is on further extensions of existing models for design and optimization of Heat Exchanger Networks (HENs). The goal is to provide a sequential framework (LP, MILP, NLP) that does not suffer from the numerical problems caused by nonlinearities, discontinuities and combinatorial explosion in simultaneous MILP models. Developing a new MILP Transshipment model to provide sets of matches that gives designs with close to minimum area and cost when supplied to an NLP model for network generation and optimization is one of the major tasks. Also, the idea is to develop a convex NLP model which generates the global optimum network rather than the local on.

w Teaching assistant in the courses Process Synthesis, Thermodynamics and Process Modeling II for Msc students at Telemark Institute of Technology.

w Last spring I was co-advisor for two diploma students on synthesis of heat exchanger networks and simultaneous optimization and heat integration.

From January 1, through June 30,1997 I will be visiting Department of Chemical Engineering, Carnegie Mellon University, Pittsburgh, USA. During this period I will be working in Professor Ignacio Grossman’s group in the area of global optimization of heat exchanger networks with fixed topology.

I started my Ph.D. study in 1992, and will finish in 1997. My work is financed by the Royal Norwegian Research Council, and my advisor is Prof. Kristian M. Lien.

I am currently in the final stages of my studies, and am writing on my thesis. The working title for the thesis is 'Some Issues in Large Scale Process Synthesis,' and the topic is how to go about building large models, involving several people in a modelling project, making models that can be used by others (or yourself at a later stage) and so on. The thesis is the result of our work on modelling synthesis gas processes, as well as earlier work on parallel algorithms in process synthesis.

w The spring of '96 was spent visiting at the Centre for Process Systems Engineering, Imperial College of Science, Technology and Medicine, in the group of prof. C.C. Pantelides. During this stay the synthesis gas reactor models were rewritten using the gPROMS process simulator developed at the centre. This allowed an implementation of the system of partial differential and algebraic equations (PDAEs) describing the reactor models directly in the simulator, and during this stay much progress was made on the reactor models.

In the fall, the models were further developed, and a presentation for the AIChE Annual Meeting in Chicago in November was prepared. This fall, we also had a diploma student working on a model for a convective steam reformer, which was an extension of the models developed in this work. In addition, some time was spent assisting in a course in process synthesis.

After the AIChE presentation, focus was shifted completely on to the thesis itself. Looking at what had actually been achieved both in this project and prior work, the expected contents of the thesis was formulated, and the contents list is included below. The focus of the thesis itself has also shifted slightly, as the contents will show.

w H.N. Hansen, C.C. Pantelides and K.M. Lien. Modelling and Optimisation of Integrated Synthesis Gas Production Plants. AIChE Annual Meeting, Chicago, November 1996.

The thesis is expected to be finished in mid '97, the schedule for the spring is already set and mainly involves writing. There is also planned a small trip to some key academic groups some time this spring.

I started my Ph.D studies in 1994 and expect to finish in 1998. Until june -96, I was employed as a scientific assistant with responsibility for our UNIX computer system and network.

Since then, I've been visiting prof Art Westerberg at Carnegie Mellon University, a visit that will continue into 1997. My advisor is prof. Kristian M. Lien.

The working title of my thesis is 'Integrated systems of reaction and separation'. Design strategy/algorithms for reactive distillation columns is the current focus, but multifunctional units on a broader basis is of interest. Related research areas are design of azeotropic

distillation systems and optimization of reactor networks.

w Completed 3-weeks course in 'Advanced Optimization' Course at DTU-Denmark (january)

w Joined ESCAPE-6 conference at Rhodes (May)

w Presented talk on 'reactive fixed points' at HUT-Finland (june)

w Research visit at Carnegie Mellon University (june-december)

w Implemented interfaces for electronic review process to PSE'97/ESCAPE-7

w Prepared research note on mechanism in MTBE dynamic reactive distillation in cooperation with Shell Research, Amsterdam

w Presented talk on 'reactive separation systems' at AICHE-Chicago (november)

w S.Hauan and K.M.Lien, "Geometric visualisation of reactive fixed points", Comp.chem.Engng, Vol 20 Suppl., pp. S133-S138, 1996

w S.Hauan, T.Omtveit and K.M.Lien, 'Analysis of reactive separation systems', Presented at the 1996 AIChE Annual Meeting, Chicago, paper 5f

My research visit to CMU will come to an end in late april at which time I'll return to Trondheim to assist in arranging the process systems engineering conference, PSE'97/ESCAPE-7.

Personal research goals are:

w Expand the analysis of reactive separation systems to the (pre-)design phase of multifunctional units

w Investigate geometric details in azetropic distillation colums (with A.Westerberg)

The Ph.D. study started in May 1993 and I had a leave from the study for one year (June 1994 to June 1995). I plan to finish the study in the summer of 1997. The work is financed

through the Norwegian Research Council strategic technology program in Computer Aided Chemical Engineering.

The title of the thesis is "Control Structure Selection for Integrated Plants". It mainly deals with finding a control structure which is both robust with respect to model uncertainty and operational changes and which yields acceptable performance. The main focus is on selection of inputs and outputs to be controlled on the regulatory control layer, the interconnection of inputs and outputs (control structure selection) and limitations on achievable performance in multivariable systems.

w Attended European network, HCM-meeting in Lyngby, 19-20 May, 1996.

w Attended European Symposium on Computer Aided Process Engineering - 6, ESCAPE-6, 26-29 May 1996, Rhodes Greece.

w Attended the 13'th IFAC world congress, 30 June - 5 July 1996,San Francisco, USA.

w Attended the UKACC International Conference, CONTROL'96, 2-5 Sept. 1996,University of Exeter, UK.

w The use of RGA and condition number as robustness measures, S. Skogestad and K. Havre, ESCAPE-6, Rhodes, May 1996.

w Input/output selection and partial control, K. Havre and S. Skogestad, IFAC World Congress, San Francisco, July 1996.

w Directions and factorizations of zeros and poles in multivariable systems,

(TECHNICAL REPORT) K. Havre and S. Skogestad, Norwegian University of

Science and Technology, Dept. of Chemical Engineering, June 1996.

w Effect of RHP zeros and poles on performance in multivariable systems,

K. Havre and S. Skogestad, Conference CONTROL'96, Exeter, UK, September 1996.

w Selection of feedback variables for implementing optimizing control schemes,

K. Havre, J.C. Morud and S. Skogestad, Conference CONTROL'96, Exeter, UK, September 1996.

w SAPHYS: A Joule II E.C. Project for solar hydrogen, S. Galli, M. Stefanoni, K. Havre, P. Borg, W.A. Brocke and J. Mergel, presented at Hydrogen`96 - 11th World Hydrogen Energy Conference, Stuttgart, Germany, June 23-28, 1996.

w Selection of variables for regulatory control using pole directions, K. Havre and

S. Skogestad, AIChE Annual Meeting, Nov. 1996.

At the moment I am writing on a couple of Journal articles. I will finish writing my thesis during the summer of 1997 and defend my thesis in the autumn 1997.

Status

I started my Ph.D. studies in February 1994 and will finish in the Spring of 1998. My doctoral work is financed by the Research Council of Norway and partly Nycomed Imaging, Borregaard Fine Chemicals, Norsk Hydro and Dyno. My advisor is Prof. Sigurd Skogestad.

The working title of my thesis is ``Batch distillation - control and optimization'' with a focus on separation of multicomponent azeotropic mixtures. The motivation for the work is the need for recovery and recycle of waste solvents in the pharmaceutical- and specialty chemical industry. This typically involves purification of homogeneous azeotropic mixtures which is inherently difficult to separate into pure components. The most common method to separate

such mixtures is by means of extractive distillation. The aim is to extend continuous azeotropic distillation insight to batch azeotropic distillation, which is more suited in small-scale production of high added value products with great demands for flexibility and productivity.

My current research activity is on operation and control of batch extractive distillation using a middle vessel column. Recent simulation studies show that the middle vessel column perform better than the batch rectifier in almost all cases for separation of ternary, ideal mixtures. This column configuration may be advantageously used in batch extractive distillation and this is a new area of research.

w Presented my work on holdup effects (dynamics) in batch distillation of azeotropic mixtures and the initial results on batch extractive distillation at the 1996 Annual Process Design and Control Center Meeting in Amherst - Massachusetts, May 1996.

w Returned to NTNU - Trondheim in August 1996, after spending 10 months in the group of Prof. M. F. Doherty and Prof. M. F. Malone at the Department of Chemical Engineering, University of Massachusetts - Amherst, USA.

w Industrial project on high-purity distillation for Hydrogas AS.

w Teaching assistant in an undergraduate course in process control.

Finish my work on alternative operation strategies and control configurations in extractive batch distillation. Take a course on azeotropic distillation with Guest Professor Dr. V. Kiva from Karpov Institute of Physical Chemistry, Moscow as advisor.

Extend the work on batch distillation of azeotropic mixtures to heterogeneous systems using the multivessel batch distillation column.

Co-advisor on a Diploma project on improved performance of a heterogeneous azeotropic distillation process at Dyno Sæthre.

I started my study in january 1996 and will be finishing my work in the beginning of year 2000. The financial support is from The Research Council of Norway. My main supervisor is prof. Rolf Henriksen and my support supervisor is prof.emeritus Jens Glad Balchen.

For the time being I am still taking courses in connection to my planned study.

I will be finished with my courses in desember 1997. So far I have done a litteratur search and study on gasnetoptimization.

I have been partisipating in meetings at Statoils research center to gain insight in the challenges and problems that is a fact in gastransportationsystems today.

I have visited the Gas Transportation Centre at Bygnes which is responsible for the control of Norways natural gasnet for export. I have also visited Statoils Gas Sales Centre in Stavanger to gain insight into their main workingareas and cooperation with the Transportation Centre and the gas fields.

I will be working on optimization of natural gas nets. Today most of the gasnets are controlled non automaticaly by experts which have advanced operating support systems for help to take the right desitions. There is a wish to minimize the total fuel cost in the gasnet and the enviromental costs in connection with these chemical reactions. To control the net optimal it is

necessary to know all the offtakes and supply capasities to a high degree of accuracy. To store unnecessary gas amounts in the physical net, and to have big artificial and physical drygasstorages cost a great deal of money and must be minimized without risking the delievery regularity to the customers.

From the summer 1997 I will try to spesify a spesific topic in this area that will probably be of important significanse in gasnet optimization.I feel that it has been important to get insight into the many areas of gasnet before I start to consentrate on a main area.

I will be finishing my courses and make a status report on gasnetoptimization for the time up to 1997. Then I will spesify my working area which will be the main work for my thesis and do some research in autumn 1997. In spring 1997 I have to focus on my four courses and the fifty percent job as scientifical assistant.

I started my study in may 1996. My advisor is Prof. Sigurd Skogestad.

The title for my project is optimization and control of methanol plants. Since I started last year I have not had much time for research. During the summer I stayed at Statoil FoU, where I did some simulations on the training simulator for the methanol plant and looked at the control structure for the plant.

w In the last semester I worked as a supervisor for a laboratory exercises.

w I have also attended the following courses: Matrix methods, Optimization theory, Stocastical and adaptive systems, Heat and mass transport and Optimal control.

w As a part of Optimal Control I made a presentation for the Nordic Process Control Workshop.

Take the courses: Linear analysis, Thermodynamics for hydrocarbons, Advanced process synthesis and Multi variable frequency analysis.

Supervision of laboratory excersises.

Work on my thesis.

I started my Ph.D studies September 1994 and expect to finish by the end of 1998. I am participating in the INPRO-program which is a cross-disciplinary Ph.D-program. My advisor is Prof. Kristian M. Lien. The work is financed by the INPRO-program, except for one year which is financed by the Nordic Energy Research Program.

My research activities are related to "operation of process plants". During my doctoral work I am cooperating with Statoil Research Center in Trondheim and Statoil Tjeldbergodden.

My focus is on "handling of abnormal situations in operation of process plants". Research shows there exist a huge potential of economical savings in improving the handling of such situations. More specific, my focus is on improving the handling process, the utilization of the information collected, and the process of learning from previous situations. I also pay some attention to the structure of the systems and tools used in the handling of abnormal situations, and how the systems and the organization structures are adapted to each other.

w Invited lecture at Hydro Porsgrunn, June 3rd, "Operator support system in `new' organization structures"

w BISP Project-seminar, June 4th-6th

w I attended a course in Boston, USA, on G2 (Gensym Computas) 8th-12th of July 1996.

w Two visits to DuPont Gibbons, Canada

January 1996: I attended a three weeks course for new operators at DuPont Gibbons

August 1996 : The main purpose of this visit was to observe and study their organization structure, and compare with my impression from the first visit. I spent time with the operators during shifts, observed and made interviews.

w Presentation at Statoil Research center and Statoil Tjeldbergodden: "Experiences and impressions from DuPont Gibbons"

w Participating in a project at Statoil Tjeldbergodden on design of a variance system.

w The Nordic Energy Research Program, Annual meeting in Copenhagen, 21th-22nd November.My presentation: "Handling of abnormal situations in operation of process plants"

w INPRO-activities:

Seminars at associated plants: Seminar Phillips Stavanger, June 18th, "Operation and design"

w Conference attendance: ESCAPE-6, Rhodes, May

w Lindheim,C., Totland,T. and Lien,K.M., "Enterprise Modeling, - A new task for process systems engineers?", presented at the ESCAPE-6 conference, Rhodes, May 26th-29th 1996.

w Lindheim,C. and Lien, K.M., "Operator Support Systems for New Kind of Process Operation Work" submitted to the ESCAPE7/PSE97 conference in Trondheim, Norway, May 1997.

Collect data for case study: Throughout the next year I will spend time at Statoil, Tjeldbergodden, which is my case during the work. I am going to study the production process, the organization structure, their information systems, operator support systems and process control system.

Attend the ECSAPE7/PSE97 conference in Trondheim in May, and a 6 weeks stay at DTU in Copenhagen to visit prof. Rafiqul Gani's group.

I started my Ph.D. studies in april 1996 under the supervision of Professor Terje Hertzberg and plan to finish in the spring 2000. My work is financed by the NFR - project on chemical conversion of naturalgas.

My research area is dynamic behaviour of a methanolreactor in Statoil's methanolplant at Tjeldbergodden. The methanolsythesis is an integrated process with heat and material recycling, and the catalyst is decactivating over time. My focus is on stability, optimal operation and state surveillance.

w I stayed three months at Statoil Research Center in Trondheim whereI participated in a project on methanol kinetics.

w Conference attendance: ESCAPE-6, Rhodes, May

w Supervision of the process control laboratory for 4. grade students.

w I attended the following courses: Reactor modelling

Mathematical modelling and model fitting

Industrial ecology

Matrix methods

During 1997 I will finish my course work with the completion of the following courses:

Matrix methods

Industrial ecology

Numerical analysis

Advanced process synthesis

Advanced process simulation

Organize the exercises for the process calculations course

Attend the ECSAPE7/PSE97 conference in Trondheim in May

In the summer and autumn I will work mainly on my research

I'm applying for the Ph.D. studies in january 1997 which is scheduled to be finished in the year 2000. I have not yet received any financial support for my studies. My supervisor is prof. Bjarne A. Foss.

The research area is planned as 'Nonlinear and robust model predictive control'. I.e. some nonlinear problems can be represented as a convex combination of locally linear models. We shall develop robust and feasible methods for model predictive control (MPC) of such nonlinear systems. To achieve this, linear matrix inequalities (LMI) and new stability theory for MPC and optimization methods will be applied.

w I was engaged to revise a compendium in instrumentation systems at the departement of engineering cybernetics last autumn.

w I have now completed these Ph.D. courses: Identification and estimation theory

Models fault detection and diagnosis

None.

I will be working as a teaching assistant at the departement of engineering cybernetics this spring. Provided financial fundation, I will become a Ph.D. student next autumn. Planned courses 1997:

Linear analysis

Dynamical systems

Optimization theory

Technical writing

LMI-methods in nonlinear MPC (self study)

I have been student on the PhD project INPRO (Integrated production systems for the process industry) since 1. June 1994 and will finish my doctoral degree in 1997. My project is partly financed by Borregaard Industries Ltd in Sarpsborg and they are my co-operating process company during the PhD study. My advisor is prof. Kristian M. Lien.

The working title of my thesis is "Modeling and Optimization of Existing Process Plants". I am developing a method for modeling of existing plant. The main object is to facilitate the change process in the organization which is necessary to support the optimization.

I was visiting prof. A. W. Westerberg at Carnegie Mellon University in Pittsburg from february to September 1996.

J.O.Olsen and K.M.Lien, "Modeling for Improved Process Operation - A Socio-Technical Approach", AIChE, Chicago Nowember 1996.

Write the thesis.

Started with my studies in November 1994. Soon to be finished with courses. Numerical Analysis is left.

Simulation of chemical processes. Numerical integration of Differential-Algebraic equation systems.

w Finished courses in Matrix methods, Thermodynamics for hydrocarbons and Reactor modeling. Presented a paper at ESCAPE-6 in Greece.

w From October I have been visiting Carnegie Mellon University in Pittsburgh to continue my work on Differential-Algebraic equation systems.

w Stiffness reduction of gas system models H.I. MOE, H.C. RIKSHEIM, and T. HERTZBERG (Paper presented at ESCAPE-6, May 1996, Greece)

Do the research necessary to write the thesis.

I started my Ph.D. studies in 1995, and will finish my doctoral degree in 1998. My doctoral work is financed by the Norwegian Research Council, and my advisors are Prof. Signe Kjelstrup and Prof. Kristian M. Lien. Part of the work is done on processes

belonging to Norsk Hydro, where I also have an office (Hydro Vækerø, Avd. for Prosessteknikk until June 1996 and Hydro Research Center from August 1996).

The general topic of my research is energy efficient process designs and optimization. So far I have, together with my supervisors, developed a new principle for process design and optimization, called equipartition of forces. The principle states that the best trade-offs between energy dissipation (related to operating costs) and transfer area (related to investment costs) are achieved when the thermodynamic driving forces are uniformly distributed over the transfer area. This has been applied so far to distillation columns, batch and plug flow reactors. One of the columns optimized belongs to Norsk Hydro. Another optimization has been studied together with Prof. Dr. Ing. Ricardo Rivero at the Mexican Institute of Petroleum. The work uncovered an important failure in our application of the principle to distillation, which we can now correct for. The results were published at ECOS'96 in

Stockholm.

In 1996 my major work has been on applying the Equipartition of Forces principle to chemical reactors. We have been able to explain and generalize more tentative, numerical results in

the literature. The results will be published at PSE ESCAPE-7 in Trondheim 1997.

w I finished a M.Sc. in Social Anthropology in 1996 on "Turkeys Struggle with democracy and Kurds"

w I presented our papers at ECOS'96 and AICHE 1996.

w I took courses in Numerical Calculations, Reactor modelling and Chemometrics.

w Sauar, E., Ratkje, S.K., Lien, K.M., Equipartition of Forces - a new Principle for Process Design and Optimization, I&EC Research , 35, 4147-4153, 1996.

w Sauar, E., Rivero, R., Ratkje, S.K., Lien, K.M., Diabatic Column Optimization compared to Isoforce Columns. In Efficiency, Costs, Optimization, Simulation and Environmental Impact of Energy Systems 1996, Alvfors P., Eidensten, L., Svedberg,G., Jinyue, Y. (Eds.), ISBN 91-7170-664-x, Stockholm, pp. 635-640.

w Sauar, E., Ratkje, S.K., Lien, K.M., Equipartition of Forces - a new Principle for Process Design, AIChE Annual Meeting , Chicago,November, 1996.

w Sauar, E., Turkeys Struggle with democracy and Kurds - an Authority, Information and Terrorism perspective, M.Sc. thesis, Dep. of Social Anthropology, NTNU, ISBN 82-91782-00-8, Trondheim 1996.

In January 1997 I will go to Carnegie-Mellon University in Pittsburg, and stay with Prof. Erik B.Ydstie for 3-4 months. I am also planning to make a joint study on reactive distillation with Steinar Hauan and intend to write a paper on energy pricing and energy taxing.

I started my Ph.D study in September 1994, and plan to finish my doctoral thesis in 1997. My research is financed by the Research Council of Norway, the Norwegian Federation of Process and Manufacturing Industries (PIL) and the Nycomed Imaging plants in Norway through the

INPRO program. My advisor are Prof. Bjarne A. Foss.

In my research, I focus on batch plants and information systems to support the operation of them. This include an investigation of the role of models and modelling in plant information systems as an interplay between a process plant and the organization running it.

w In 1996, I spent three months at the Department of Automatic Control at Lund Institute of Technology in Sweden. In cooperation with Karl Erik ]rzen and Charlotta Johnsson, I implemented a prototype for an integrated batch plant operation system. This work is reported in "A Framework for Batch Plant Information Models" (Report ISRN LUTFD2/TFRT--7553--SE).

w In May, I attended the ESCAPE-6 conference in Rhodes were I presented the paper "A Concept for Modelling Batch Plant Operation" (cf. Comput Chem Eng 1996, 20, S1245-S1250). I was also sponsored by PROST to attend a G-2 course arranged by the Gensum Corp. in Boston, Mass.

I plan to complete my thesis in 1997.

I started my Ph.D. study in february 1995, and will finish my doctoral degree in 1998. My work is financed by the Norwegian University of Science and Technology. My advisor is Prof. Bjarne A. Foss.

I work within the area of hybrid control, which in this context means control of processes which exhibits both discrete and continuous phenomena such as a mixture of discrete- (on/off elements) and continuous control inputs. In particular I investigate usage of the principle of model predicitive control (MPC) for controlling such systems.

w During the summer, Jostein Vada and I developed a Matlab program called PredSim, which is a plattform for simulating model predictive control. The problem formulation in PredSim is general and is suitable for simulating MPC for hybrid systems. I was on leave from my stipend during this period.

w I spent the fall semester at the University of Wisconsin at Madison in Professor Jim Rawlings' research group. Professor Rawlings is highly recognized for his work in the MPC area.

w Attended the AIChE annual meeting in Chicago.

Two accepted. One for ECC'97 and the other one for ACC'97.

During '97 I plan to continue the work on MPC of hybrid systems. I will also work with robust MPC (not for hybrid systems). I will present my work at least at two conferences.

I started my Ph.D. work in September 1994 and will finish my doctoral degree in march 1998. My work is financed by the INPRO program, and my advisor is Prof. Bjarne A. Foss.

The preliminary title of my thesis is "Studies of some methods and phenomena in control engineering projects - with application to precipitation and cementation processes". My work is within process modeling, analysis and control design, and aspects within implementation of new control strategies in a process plant. My industrial partner is Falconbridge Nikkelverk in Norway and Falconbridge Kidd Creek in Canada.

w Attended and presented a paper at ESCAPE-6, Rhodes, Greece.

w Spent six months at Falconbridge Kidd Creek in Canada working within process

control.

w Co-advisor for one diploma work, and one student project work.

w Støle-Hansen, K. and Dotterud, O.M. (1996). Dynamic modelling of pH controlled

precipitation and dissolution in an aqueous solution. Computers chem. Engng,

Vol 20, Suppl pp. S629-S634.

Present papers at AdChem'97 and PSESCAPE-97.

Work with control design and issues related to the process of implementing new

control strategies.

I have just finished the writing of my thesis, the defence is on march 18. 1997. The study started in 1991. The major part of my study has been funded by the PROMIA program, but I have also received support from the Nordic Energy Research Program. My supervisor is Prof. Terje Hertzberg.

The title of my thesis is: 'Dynamic optimization, studies on local thermodynamic models and sensitivity estimation'. Local thermodynamic models have been tried out in both dynamic simulation and dynamic optimization context. The negative impact local thermodynamic models have on higher order integrators has been studied, and two possible solutions have been proposed. The most computer intensive task in dynamic optimization is the evaluation of

the sensitivities of the objective function and constraints w.r.t. the parameters. A alternative approach based on perturbations is tried out. All N+1 trajectories needed are solved simultaneously by the DAE integrator. This means that the same Jacobian, steplength etc. is used in all N+1 trajectories, the result is a more efficient calculation, and more accurate gradient information.

w Co-advisor of a student project on the energy-optimization of the nitric acid process.

w Finished my Ph.D.

w Got a job

w Støren, S., T. Hertzberg (1997) 'Obtaining sensitivity information in dynamic

optimization problems solved by the sequential approach', Accepted for Comp.Chem.Eng.

Defend my Ph.D.

My Ph.D. studies started in november 1993. Until the end of 1996 I've been funded by the Norwegian Research Council's PROGIT program. My fundings for 1997 are provided by the Nordic Energy Research Program. My advisor is prof. Kristian Lien. I plan to finish my doctoral work within the end of 1997.

The scope of my thesis is design and synthesis of chemical reactor systems. During the first part of my studies I've in particular been looking at the mixing patterns in the reactor system. During 1996 we have developed a new, general mathematical model for continuous reactors, which is suited for manipulating mixing patterns so as to give the best possible selectivity,

yield, conversion and reactor investment cost. The model is very flexible, and may represent different mixing extremes such as plug flow, continuously stirred tank reactor and differential sidestream reactor. Our model also overcomes many of the problems associated with existing methods, and is thus a promising tool for reactor system design and synthesis. We have applied the model on simple example systems taken from litterature, as well as on industrial case studies. The oxidation of SO2 and the production of polyethylene are two interesting main case studies, which illustrates how differential sidestream addition may be optimal.

We have also turned our attention towards the energy integration of the reactor system with the rest of the process. It is clear that there is a strong interaction between the reactor system and the surrounding process, but it is difficult to quantify the trade-offs that has to be made. The knowledge in this field is limited, so there is a great incentive to gain more insight about these interactions.

The following course has been completed during the year:

w Philosophy and the theory of science.

w July/August: visit to prof. L. Bieglers group at Carnegie Mellon University, USA,

w November: Presentation at the AIChE Annual Meeting 1996, Chicago, USA.

w November: Started my research visit with prof.\ Tapio Westerlunds group at Åbo Akademi, Åbo, Finland,

w Sund, E. B., Lien, K. M., "An optimal control formulation of the reaction-mixing problem" , AIChE-96, Chicago.

This year I will concentrate the work on the energy integration of reactor systems. We have planned two publications/presentations; at the PSE-ESCAPE97 conference in Trondheim, May 25-29., and at the AIChE Annual Meeting 1997, Los Angeles, November 16-21. I will spend approx. two months of 1997 at Åbo Akademi. Finally, I plan to finish my PhD thesis at the end of the year.

I started my Ph.D. study in august 1994, and expect to complete my thesis in 1997.

My doctoral work is financed by the DAK II project, and my supervisor is Prof. Terje Hertzberg.

In my research I focus on the propagation of uncertainties in dynamic process modells. These calculations have mainly been done by Monte-Carlo methods, which can be slow and computationally expensive. In my work I try to develop methods with high accuracy, and still low computational demand.

w In january I developed a steady-state simulator of a oxy-chlorination reactor. This work was initiated and funded by SINTEF Chemical Engineering. During the spring I arranged and participated in the colloquium on process systems technology.

w In may I participated in a workshop on Inverse-Problems in reservoir evaluation.

During spring and summer I worked on a new approach for calculation of

uncertainties in dynamic simulation results. The method is based on Gaussian- quadrature and uses the uncertain parameters pdf's as weight functions.

w Since October I have been visiting Dr. Urmila Diwekar at CMU. I have studied her recent work on efficient sampling techniques using quasi-random sequences, and used her batch-distillation software for stochastic simulation.

w Tørvi, H. Hertzberg T. (1996), 'Estimation of uncertainty in dynamic simulation results'. This will be presented on the ESCAPE-7 conference in Trondheim 1997, and appear in Comp. chem. Eng.

I will continue my work on methods for calculating propagation of uncertainties. Hopefully I will finish my thesis by the end of the year.

I started my Ph.D. study in September 1995 and expects to finish my doctoral degree in 1999. My doctoral work is financed by the Norwegian Research Council, and my advisor is Professor Bjarne A. Foss.

During my studies, I will work within the area of analysis and control of complex process plants.

w In September 1996 I terminated my one year duty as teaching assistant, and will now continue my studies on full time.

w During 1996, I fullfilled the following courses: Linear analysis

Optimal control of dynamic systems

Models in fault detection and diagnosis

Termodynamics

Representation and analysis of hybrid systems (self study)

w During the summer, Olav Slupphaug and I developed a Matlab program called

PredSim, which is a plattform for simulating model predictive control. The problem formulation in PredSim is general and is suitable for simulating MPC for systems with both discrete and continuous inputs.

None

I am planning to take the courses 75045 Dynamic Systems and 990 Information Retrieval during the spring. Besides that, I will attend ACC97 in Albuquerque in June. At ACC97, Olav Slupphaug will present a paper where I am co-author. Further, I will decide more specifically my Ph.D.research topic and study the contributions whithin that area.

I started my Ph.D. work in spring 1990 and plan to finish by spring 1997. From 1990 to autumn 1994 I was financed by DEMINEX. Since autumn 1994 I am employed at the department of Chemical Engineering as researcher assistant, this work include supervision of laboratory exercises of third and fourth grade students.

The working title of my PhD. thesis is ' Studies on the dynamics of continuous and batch distillation processes'. At the department a pilot plant distillation column are installed and offers the possibility of practical experiments on distillation column operation. During the experimental work the following issues will be addressed:

1) continuous distillation,

Validation of model assumptions regarding complexity of the model, simplification and assumptions made during the modeling process. Measurement types to interfere the product compositions

2) batch distillation

Operation and control of the newly developed multi-vessel batch distillation column

w Simulation: Simulations of a newly developed multivessel batch distillation

column for the separation of multicomponent mixtures. During the work

the feasibility of the process was show and possible energy savings are investigated. A new method of controlling the column was developed which is capable to control the process properly.

w Experimental: A multivessel batch distillation pilot plant was designed and

build. Experiments have been performed which show the operability of

the process and verify the simulation results. The control structure

developed in the simulation work was implemented on the pilot plant.

w Work at the department:

Supervision of Process-laboratory for third grade students. Supervision a 4'th grade student project on 'Experimental verification of Multivessel Batch Distillation': Investigation concerning the influence of initial feed-composiiton on the product

compositions.

Supervision a diploma-thesis on 'Experimental verification of Multivessel Batch Distillation': Investigations on the start-up procedures, set-point choice and its influence on the product quality. Further initial research on the energy consumption of the process.

Supervision and administration of the process control laboratory for 4'th grade students.

Supervision and administration process laboratory for Trondhjems Ingeniør Høgskole.

Through the "Batch Project" for the production of Fine Chemicals, industrial contact to Boregaard Fine Chemicals, Dyno, Norsk Hydro and Nycomed is established, in this program I am responsible for the experimental activities, including maintenance of analysis and distillation equipment.

w ESCAPE'6, European Symposium on Computer Aided Process Engineering,

26-29. may 1996, Rhodes, Greece

w ESCAPE'6; 'Total reflux operation of Multi-vessel Batch Distillation', B. Wittgens, R. Litto, E. Sørensen & S. Skogestad, Rhodos, Grecce in: supplement to Computer & Chemical Engineering, 1996, Elsevier Science Ltd.

w "Multivessel Batch Distillation - Experimental verification",B. Wittgens & S. Skogestad, accepted for presentation on: Destillation and Absorption'97, Maastricht, The Netherland

w "Multivessel Batch Distillation'', S. Skogestad, B. Wittgens, E. Sorensen & R. Litto, reviewed oct. 1996 AIChE-Journal

The work on the multivessel batch distillation column will be continued to investigate further features of the process including start-up and operation procedures. The work at the department will include the supervision and adminiatration of the process laboratory of third grade students, the process control laboratory for the forth grade course. Further I will supervise a project of a forth grade desgin course on energy consumption of different distillation column configurations.

My Ph.D. started in September 1995, and the scheduled finish is in September 1998.

I am situated at Tel-Tek in Porsgrunn, where I attend the Process Integration group at Telemark Institute of Technology. My supervisor is Prof. Truls Gundersen, and the doctoral work is financed by the Nordic Energy Research Programme on Process Integration.

My research will be focused on integrating and optimizing Water consuming processes in Nordic industry. This will also include energy, raw material and and environmental considerations. I will will start with the young "Waterpinch" methodology, and

try to extend its application to handle typical large Nordic Water consuming processes.

Spring 1996 was used mainly to pass compulsory courses at NTNU. In November

I visited the company Danish Energy Analysis in Copenhagen, and stayed for 3 months.

The purpose of this stay, was to get industrial partners, and thereby get real cases to base

my work on. We did succeed in this matter. Other activities I have attended;

w 2 day course in "Wastewater Minimization" at UMIST in Nov. 95

w 2 week course in "Advanced Pinch Technology" at Chalmers Aug. 96

w 2 day course in "Energy and Environmental Management" at Copenhagen Nov. 96

w Participating in seminars held by The Nordic Programme of PI ,at Chalmers Dec. 96

None

During spring 1997 I will complete my compulsory courses, staying part time in Trondheim at NTNU. During this year, I will start to work on my main project based on real cases from the Pulp and Paper industry, and hopefully also from a Petroleum Refinery. As to fulfil an obligation within my scholarship, I plan to spend a period of 3 months during fall 97/spring 98 at a Technical University in Finland.

A.C. Christiansen, J.C. Morud and S. Skogestad, "A comparative analysis

of methods for solving systems of nonlinear algebraic equations," Proc. of the 38th SIMS Simulation Conference, Trondheim, June 1996, 217-230.

Glemmestad B., Mathisen K.W. and Gundersen T. "Optimal Operation of Heat Exchanger Networks based on Structural Information", Comput. and Chem. Engng., vol. 20, Suppl. (Proc. from the European Symposium of Computer-Aided Process Engineering, ESCAPE-6, Rhodes, May 1996), pp S823-S828.

Gundersen T., Duvold S. and Hashemi-Ahmady A. "An Extended Vertical MILP Model for Heat Exchanger Network Synthesis", Comput. and Chem. Engng., vol. 20, Suppl. (Proc. from the European Symposium of Computer-Aided Process Engineering, ESCAPE-6, Rhodes, May 1996), pp S97-S102.

Gundersen T. "IEA Implementing Agreement on Process Integration - Survey of Process Integration Methods, Software, Applications and End-User Needs", UN Workshop on Process Integration, Manchester, 11 June 1996:

Gundersen T. "In-house or commercial Software for Process Integration in Norway - Experiences from Industry and Academia", Nordic seminar on Process Integration, Gothenburg, 6 September 1996.

H. N. Hansen, C.C. Pantelides and K.M. Lien. Modelling and Optimisation of Integrated Synthesis Gas Production Plants. AIChE Annual Meeting, Chicago, November 1996.

S.Hauan and K.M.Lien," Geometric visualisation of reactive fixed points", Comp.chem.Engng, Vol 20 Suppl., pp. S133-S138,

S.Hauan, T.Omtveit and K.M.Lien, 'Analysis of reactive separation systems', Presented at the 1996 AIChE Annual Meeting, Chicago, paper 5f.

K. Havre and S. Skogestad, "Input/output selection and partial control," IFAC World Congress, San Francisco, July 1996, Volume M, 181-186.

K. Havre, J. Morud and S. Skogestad, "Selection of feedback variables for implement ing optimizing control schemes," UKACC Conference Control'96, Exeter, UK, Sept. 1996, 491-496.

K. Havre and S. Skogestad, "Effect of RHP zeros and poles on performance in multi variable systems," UKACC Conference Control'96, Exeter, UK, Sept. 1996, 930- 935.

K. Havre and S. Skogestad, "Selection of variables for regulatry control using pole directions, AIChE Annual Meeting, Chicago, Nov. 1996, Paper 45f.

T. A. Johansen "Identification of Non-linear Systems using Empirical Data and Prior Knowledge - An Optimization Approach", Automatica, Vol. 32, pp. 337-356. 1996

T. A. Johansen and P. A. Ioannou "Robust Adaptive Control of Minimum phase Nonlinear Systems",Int. J. Adaptive Control and Signal Processing,Vol.10, pp.61-78. 1996

T. A. Johansen, "Stability, Performance and Robustness of Fuzzy Model-Based Control", Proc. IEEE Conf. Decision and Control, pp. 604-609, Kobe, Japan.

T. A. Johansen, "Robust Identification of Takagi-Sugeno-Kang Fuzzy Models using Regularization", Proc. IEEE Conf. Fuzzy Systems, pp. 180-186, New Orleans, Louisiana

T. A. Johansen and B. A. Foss, "Operating Regime Based Process Modeling and Identification", Computers and Chemical Engineering, Vol. 21, pp. 159-176, 1997

T. A. Johansen and M. M. Polycarpou, "Stable Adaptive Control of a General Class of Non- linear Systems", in "Neural Adaptive Control Technology I", R. Zbikowski and K. Hunt (Eds), World Scientific, pp. 177-206, 1996.

Kvamsdal, H.M. and T. Hertzberg, " A Preliminary Design Study of a Multicomponent PSA Gas Separation System, Chemical Engineering and Processing, 35, pp. 213- 224,1996.

Lien, K.M. and T. Perris: "Future directions for CAPE research. Perceptions of industrial needs and opportunities." Comp.Chem.Engng. Vol 20, Suppl. S1551- | 1557, Pergamon, 1996.

Lien, K.M. and T. Perris: "Industry and Academia. Rebuilding the partnership", Comp.Chem.Engng. Vol 20, Suppl. S1545-1550, Pergamon, 1996.

Lindheim,C., Totland,T. and Lien,K.M., "Enterprise Modeling, - A new task for process systems engineers?", presented at the ESCAPE-6 conference, Rhodes, May 26th- 29th 1996.

F. A. Michelsen and B. A. Foss "A Comprehensive Mechanistic Model of the Continuous Kamyr Digester",Applied Mathematical Modelling,Vol.20, pp.523-533.

F. A. Michelsen and B. A. Foss "A Dynamic Model of the Interaction between the Chemical Reactions and the Residence time in the Continuous Kamyr Digester", Tappi Journal, Vol. 79, pp. 170-176.

Moe, H.I., H.C. Riksheim and T. Hertzberg, "Stiffness Reduction of Gas System Models", Computers and Chemical Engineering, 20, Suppl., pp. 901-906, 1996

J. Morud and S. Skogestad, "Dynamic behaviour of integrated plants", J. Process Control, 6, 145-146, 1996.

J.O.Olsen and K.M.Lien, "Modeling for Improved Process Operation - A Socio- Technical Approach", AIChE, Chicago Nowember 1996.

T. Pettersen, A. Argo, R.D. Noble, C.A. Koval, "Design of combined membrane and distillation processes", Separation Technology 6:175-187, (1996).

Sauar, E., Ratkje, S.K., Lien, K.M., Equipartition of Forces - a new Principle for Process Design and Optimization, I&EC Research , 35, 4147-4153, 1996.

Sauar, E., Rivero, R., Ratkje, S.K., Lien, K.M., Diabatic Column Optimization compared to Isoforce Columns. In Efficiency, Costs, Optimization, Simulation and Environmental Impact of Energy Systems 1996, Alvfors P., Eidensten, L., Svedberg,G., Jinyue, Y. (Eds.), ISBN 91-7170-664-x, Stockholm, pp. 635-640

Sauar, E., Ratkje, S.K., Lien, K.M., Equipartition of Forces - a new Principle for Process Design, AIChE Annual Meeting , Chicago,November, 1996.

J. Simensen and B. A. Foss, "A Concept for Modelling Batch Plant Operation", Comput. chem. Engng., Vol. 20, Suppl. pp. S1245-S1250, Presented at ESCAPE-6,

Rhodes, Greece.

S. Skogestad, "A procedure for SISO controllability analysis - with application to design of pH neutralization processes", Comp.Chem.Engng., 20, 373-386, 1996.

S. Skogestad and K. Havre, "The use of the RGA and condition number as robustness measures", Computers and Chem. Engng., 20, Suppl., S1005-1010 (1996). (Supplement from symposium ESCAPE-6, Rhodes, Greece, May 26-29 , 1996).

S. Skogestad and E.A. Wolff, "Controllability measures for disturbance rejection" Modeling, Identification and Control, 17, 167-182 (1996). (Reprint of paper

presented at the IFAC Workshop on interactions between process design and control, London, Sept. 1992.

S. Skogestad, "Control structure design" AIChE Annual Meeting, Chicago, Nov. 1996, Paper 149a.

R. D. Braatz, M. Morari and S. Skogestad, "Loopshaping for robust performance", Int. J. of Robust and Nonlinear Control, 6, 805-823, 1996.

E.A. Wolff and S. Skogestad, "Temperature cascade control of distillation columns",

Ind.Eng.Chem.Res., 35, 475-484, 1996.

R. Bredesen, J. Sogge, "A technical and economic assessment of membrane reactors for hydrogen and syngas production", Seminar on the Ecological appl. of innovative membrane techn. in the chemical industry, May 1.-4., Cetraro, Italy (1996).

K. Støle-Hansen and O. M. Dotterud, "Dynamic modelling of pH controlled precipitation and dissolution in an aqueous solution", Comput. chem. Engng., Vol. 20, Suppl. pp. S629-S634, Presented at ESCAPE-6, Rhodes, Greece.

E. Sørensen and S. Skogestad, "Optimal startup procedures for batch distillation", Computers and Chem. Engng., 20, Suppl., S1257-1262 (1996). (Supplement from

symposium ESCAPE-6, Rhodes, Greece, May 26-29 , 1996).

E. Sørensen and S. Skogestad, "Comparison of regular and inverted batch distillation", Chem. Eng. Sci., 51, 4949-4962, 1996.

Sund, E. B., Lien, K. M., "An optimal control formulation of the reaction-mixing problem" , AIChE-96, Chicago.

G. Tangen and P. Støa, "Decision Support for Hydropower Plant Upgrading - Integrating Multi Attribute Decision Making and Knowledge Based Systems'', Intelligence Systems Applications to Power Systems (ISAP'96), pp.\ 145-150}, Orlando, Florida.

B. Wittgens, R. Litto, E. Sørensen and S. Skogestad, "Total reflux operation of mul- tivessel batch distillation", Computers and Chem. Engng., 20, Suppl., S1041-1046

(1996). (Supplement from symposium ESCAPE-6, Rhodes, Greece, May 26-29 , 1996).

H. I. Moe "Dynamic Process Simulation. Studies on Modeling and Index Reduction", Dr.ing. thesis, report no. NTH 1995: 122, NTH, Trondheim, 1995.

J. Chr.Morud "Studies on the Dynamics and Operation of Integrated Plants"

Dr.ing thesis, NTNU,Trondheim 1995

G. Tangen "Decision Making Support Applied to Hydropower Plant Upgrading",

Dr.Ing. thesis, Report. no NTNU 1996:73, ITK 1996:51-W

S. O. Wasbø "Ferromanganese Furnace Modelling Using Object-Oriented Principle", Dr. Ing. thesis, Report. no NTNU 1996:80, ITK 1996:45-W

Dr. ing thesis, NTNUTrondheim 1996

w 13 sep. No seminar. Instead: Disputas (Ph.D. defence) Ying Zhao. Supervisor: Sigurd Skogestad. Opponents: Sten Bay Jørgensen, Steinar Sælid and David Levine.

w 20 sep. Dag Ljungquist , Inst. for kybernetikk, NTNU:

Prosessregulering i norsk aluminiumsindustri - Stort potensiale/ store utfordringer

w 27 sep. Hasse Skrifvars , Abo Akademi, Finland: Alternative stategies for solving MINLP problems.

w 04 okt. Sigurd Skogestad , NTNU: Høydepunkter fra sin nye bok.

w 11 okt. Odd Andreas Asbjørnsen, NTNU: Life Cycle Assessment (LCA) in a control engineering perspective.

w 18 okt. Kristian Lien, Institutt for kjemiteknikk, NTNU: Towards phenomena based process design.

w 25 okt. No seminra -instead: Online optimaliserings-workshop på mandag 28. oktober.

w 01 nov. Audun Faanes, Hydro F-senter: Ny PVC-reaktor: Erfaringer fra implementering av nytt reguleringssystem.

w 08 nov. Dan Rivera, Arizona State University: A Control-Relevant Methodology for Multivariable System Identification with Application to High-Purity Distillation.

w 15 nov. (AIChE Annual meeting -no semianar)

w 22 nov. Steinar Sælid, Prediktor AS: "Modellbasert reguleringsteknologi. Eksempel fra potensiell anvendelse: Prilleprosess."

w 29 nov. Costas Pantelides, Imperial College: Process Design Revisited.

w 06 des. Roger Hansen,. Statoil: Metanolanlegget på Tjeldbergodden.

w 13 des. Jens Balchen, NTNU: Eksperimenterende optimalisering.