Process Systems Enterprise Limited
email this page print this page
pdf overview

PSE process design optimisation services

Truly optimal process design

   multitubular reactor design optimization variables

Multitubular reactor design variables. Catalyst design decisions can also
be included

PSE's design optimisation service applies model-based engineering approaches to determine optimal designs for process plants or individual items of equipment.

Typical examples

Model-based design can be applied at a unit or process level, and at the front-end-engineering design (FEED) or detailed design stage. Typical examples are:

  • optimal reactor design. High-fidelity predictive models of physics and chemistry are used – coupled with CFD models where required – for detailed design of complex reactors, including internal geometry.
  • optimal design of distillation column configuration. Mixed integer optimisation techniques are used to determine simultaneously optimal feed tray location, number of trays and heat integration as well as optimal values of 'continuous' design variables such as column diameter or operating conditions.
  • optimal design of separation sections. Similar techniques are used to determine the configuration of entire separation sections for optimal CAPEX and OPEX. This can result in consolidation or elimination of columns.
  • simultaneous (whole plant) optimisation of reactor and separation sections. Reactor and separation section designs are highly interdependant. Considering both sections within the same optimisation results in truly optimal design.

   Repsol whole plant optimization

Hydrocarbon Processing
December 2010
Improve engineering via whole-plant design optimization [Repsol, Spain]

What are the benefits

The benefits are, quite simply, better design and better operations, leading to:

  • lower capital and operating cost (CAPEX and OPEX)
  • higher profitability through increased throughput and better and more consistent product quality
  • reduced technology risk
  • better compliance with health, safety and environmental requirements.

In one recent reactor-and-separation design case two columns were eliminated completely, resulting in significant capital savings.

Hydrocarbon Processing reactor optimization

Hydrocarbon Processing
June 2007
Optimize terephthaldehyde reactor operations [LG Chem, Ltd, Korea]

How does it work?

Model-based engineering uses detailed predictive models of the process to accurately quantify the effect of design decisions. Models are validated against experimental, pilot or operating data where necessary.

Models are frequently used in conjunction with rigorous mathematical optimisation techniques to achieve design objectives directly – while taking into account design and operating constraints – without the need for trial-and-error iteration.

Many different aspects of design

Even without optimisation, it is often the case that significant process and equipment improvements are identified simply by the application of accurate predictive models.

gPROMS helps us to minimise development risks during process design and operation

Arkema

For example, a detailed model of a reactor can be used to rank catalyst alternatives, in support of scale-up, to support detailed mechanical design, to design optimal operating policy, to design and test control schemes, and so on.

What does PSE provide?

PSE provides models and modelling technology, expertise and a set of well-developed methodologies built on the foundations of Model-Based Engineering.

These are summarised as:

Optimal / optimized distillation column configuration and design
Distillation: optimal configuration and conditions – click to enlarge. Annualised cost was reduced by 16%
  • The gPROMS advanced process modelling platform.
    gPROMS underpins the complex calculation required to generate the high-accuracy predictive information on which key design and operating decisions are based. It also integrates theoretical models and real-world data via sophisticated parameter estimation capabilities.
  • The state-of-the-art gPROMS optimisation technologies, used to determine optimal design parameters or operating trajectories without the need for trial-and-error simulation.
  • The state-of-the-art gPROMS advanced process models for catalytic reaction, complex separation, solution crystallisation, polymerisation and many other complex processes.
  • PSE's hybrid modelling technologies which combine modelling in gPROMS with computational fluid dynamics (CFD) tools for ultimate scale-up accuracy.
  • Consulting expertise gained through years of providing expert services across many process industry sectors.