Process Systems Engineering
Process systems engineering deals with the overall system behaviour and how the individual units should be combined to achieve optimal overall performance. Important topics are multi-scale process modelling, operation and control, design and synthesis, simulation, statistics and optimization. Applications include distillation, reactors, system biology and subsea processing (SUBPRO)
The group presently consists of more than 20 people, in addition to about 6 Diploma and project students. The group closely cooperates with other systems-oriented departments at the university, including Engineering Cybernetics and SINTEF. The process systems engineering activity at NTNU (PROST) holds high international standards and was already in 1994 recognized as a strong-point center, both by NTNU and SINTEF.
At present, the main activities in the group are within process control and process modelling including efficient thermodynamic calculations. A new area from 2008 is systems biology, where Nadi Bar started as Associate professor in January 2008.
- Process and plantwide control
- Process modelling
- Systems biology and bioinformatics
- Subsea processing (SUBPRO)
Small-scale experimental rigs have been built to study anti-slug control and novel distillation arrangements. In most cases, control is an "add-on" to enable and improve operation, but the anti-slug rig demonstrates how control in some cases can be used to operate the system in a completely different manner. The Kaibel distillation column is a 6 meter high and 5 cm in diameter and can be used to study "thermally coupled" columns, including the threeproduct Petlyuk column and the four-product Kaibel column. The group also has an automatic drink mixer, which is used for demonstration purposes and to study sequence control based on automata theory (Preisig).
The model generally needs to be fitted to experimental data, and the group has always has a strong focus on statistical methods and experimental design (Hertzberg). The group applies advance statistical methods such as Principle Component analysis, Singular value decomposition, network component analysis and iterative methods and Fourier transform to analyze and understand large data quantities. We also apply parameter estimation search using parallel computing (Skjøndal-Bar) in order to fit experimental data to dynamic models.
Comes from the Norwegian Research Council, the Gas Technology Center at NTNU and SINTEF, from industry (Statoil, Gassco, Hydro) and from the EU.