Process integration applied to design and operation of distillation columns

 

Project description

 

Date: 25 Aug. 1999 (sendt to Nordic Energy Research Program for Process Integratyion, NEFP)

 

Sigurd Skogestad and Hilde Engelien, NTNU

 

 

Introduction

 

The energy costs involved in separation, and in particular distillation, consitute a large fraction in many chemical plants (often above 50%). Traditionally, the distillation column design for a given plant has been based on a sequence of regular columns. The process integration is traditionally done afterwards and involves designing a heat exchanger network, based on method such as pinch design and mathematical programming. In some cases the column pressures are adjusted to move energy requirements from above to below the pinch, and heat pumps may be introduced for the same reason.

 

More recently, there has been increased interests in integrated column designs, where the distillation columns are more directly integrated, e.g. prefractionator arrangements and the even more directly integrated Petlyuk columns. Such columns may also result in significant savings in capital costs (indirect energy costs. An important rule of thumb for reducing the capital costs and also the energy costs is to reduce the number of process units. By use of the Petlyuk design one replaces two columns with two reboilers and two condensers with a single shell (with an intermediate wall) with one reboiler and one condenser, and it has been shown that offers potentiall savings of both capital of energy of about 30%.

 

However, we have found that it may be difficult to obtain the theoretical energy savings for the petlyuk column in practice, unless the column is operated correctly, and this may explain why industry has been somewhat reluctant to use this design. (We think we know by now how to operate the column in practice to achieve the theoretical savings).

 

The project

 

The goal of the project is to study how distillation columns may be integrated into the process in order to minimize the overall energy consumption. Both design and operation aspects will be studied. We plan to consider conventional heat integration, integrated columns (e.g. prefractionator and Petlyuk designs) and use of pressure as a parameter.

 

 

Experience and of the applicants (research group)

 

The supervisor (Sigurd Skogestad) has experience in process integration dating back to the period 1981-83 when he worked with exergy analysis and also a little on heat exchanger network design at Norsk Hydro (arranging and attending one of the early pinch courses of Linnhoff in 1983). More recently (1996-98), he was involved in an EU Joule project on integrated (Petlyuk) distillation (DISC), working together with Robin Smith at UMIST. In general, his expertice is on distillation and on optimal operation and control of process plants. He has written a textbook on multivariable control (Wiley, 1996). He heads a group of about 7 PH.D. students, and 11 Ph.D.'s have graduaded from the group since 1990.

 

Hilde Engelien has a Master degree in chemical engineering from University of Surrey, UK (1997). Since then she has two years of experience working with consulting companies for the process industries in the UK. She is very motivated for further studies with the aim of obtaining a Ph.D. degree with emphasis on process integration.

 

The following graduated Ph.D.'s from the group have been directly related to process integration:

 

 

 

 

 

Sigurd Skogestad was also the co-supervisor of

 

 

At present, there are several students in the group working on process integration, of which the most relevant are

 

 

 

 

Ivar Halvorsen, who was involved on the EU DISC-project together with UMIST, will graduate soon. It is planned that the work of Hilde Engelien will extend the results of Halvorsen on Petlyuk distillation and compare with more conventional solution with heat integration and use of pressure as a design variable.