Advanced process modelling

High-quality information for decision support

First-principles modelling validated against experimental data

Advanced process models combine first-principles theory with observed data. This is what gives them such unprecedented predictive accuracy.

Advanced process modelling involves applying detailed, high-fidelity mathematical models of process equipment and phenomena to provide high-quality information for decision support in process innovation, design and operation.

Much of the predictive power of such models results from the combination of first-principles chemical engineering, physics and chemistry with observed ("real-life") data such as laboratory, pilot plant or operational measurements.

Principles

Advanced process models typically combine first-principles theoretical models with observed – experimental, pilot plant or operating – data.

Theoretical models can come from a variety of sources, but typically come from:

research literature, where many papers are published containing models in equation form. gPROMS^® is specifically designed to allow easy transcription of mathematics from paper to model.
PSE's extensive set of advanced process models, which can be supplied or tailor-made on request.

First-principles theoretical process model

Once a suitable theoretical model is available, it is validated against data. In practice this typically involves fitting the empirical elements – for example, the reaction kinetic parameters, heat transfer coefficients and so on – to real-life data.

It is the combination of first-principles theory and real-life data than provides the powerful predictive capability of such models.

Why use advanced process models?

High-fidelity predictive models provide companies with significant competitive advantage by providing a means to :

capture, deploy and transfer corporate knowledge effectively
make decisions based on accurate quantification
innovate rapidly, by enabling exploration of a wide design space rapidly and at relatively low cost (especially when compared to building pilot plants)
manage technology risk effectively, in particular by identifying areas of low data reliability and helping to address these.

PSE is the leading supplier of advanced process modelling technology and services to the process industries. Our gPROMS modelling and optimisation platform provides a sophisticated, modern software environment created specifically for construction, validation and execution of high-accuracy models, and the company is a pioneer in the growing application of model-based engineering.

What are typical applications?

High-fidelity predictive models can be used to optimise many aspects of multitubular reactor design

Advanced process models can be – and for maximum effectiveness and return-on-investment, should be – applied across the process lifecycle, from laboratory conceptual work, through process and detailed design to online operation.

Typical application areas are those that involve complex physical and chemical phenomena, such as reaction engineering, crystallisation, complex separation processes and fuel cell component and system design.

You can see many typical examples of application here.

What are the benefits of advanced process modelling?

A key benefit of using advanced process models is that they provide high-quality information to underpin decision support in process and product innovation, and process design and operation.

In particular when used as part of a model-based engineering approach, typical benefits are:

better process designs, with lower design margins, leading to reduced capital expenditure, higher reliability and lower operating cost
better equipment designs, with lower design margins, again leading to reduced capital expenditure, better performance and controllability and lower operating cost
confident scale-up
accelerated innovation, with reduced innovation risk, cost and time-to-market. Model-based innovation techniques allow rapid screening and verification of designs, with reduced reliance on physical testing
better operations, with greater flexibility of operation through better understanding of operating limits and the ability to operate closer to constraints with confidence
better product designs, through the ability to test product designs and tailor manufacturing processes to deliver required product attributes
better compliance with health, safety and environmental requirements
effective risk management, by using the reliable quantitative data on which to base R&D and design decisions, allowing you to manage risk with confidence.

gPROMS models are documents, not software.

A gPROMS model contains just the description of the physical and chemical relationships governing the process. Solution of the resulting equations is taken care of automatically.

This means that a well-written model is effectively a document.

It can be easily read and understood by someone with the right level of process knowledge, making it easy to maintain and extend.

For this reason, models are ideal vehicles for capturing and building vital corporate knowledge, as well as transferring that knowledge between different stakeholders across the corporation.

How does gPROMS advanced process modelling relate to other modelling and simulation technologies?

The type of advanced process modelling made possible by gPROMS differs from the traditional technologies available to the process industries, such as steady-state process (flowsheeting) simulation and Computational Fluid Dynamics (CFD). Here are some key differences:

Differences from steady-state flowsheeting:

easy custom modelling capability
steady state and dynamics
parameter estimation from data
robust optimisation (including mixed-integer and dynamics) capable of handling a large number of decisions simultaneously
high degree of predictive accuracy for complex chemical phenomena - for example, catalytic reaction
multiscale modelling capabilities

Differences from CFD:

easy custom modelling capability capable of modelling complex, implicitly-related chemical phenomena
flowsheeting capabilities
parameter estimation from data
ability to model dynamic aspects such as control, operating procedures
robust optimisation (including mixed-integer and dynamics) capable of handling a large number of decisions simultaneously

gPROMS modelling is entirely complementary to many existing technologies, and can augment their application significantly.

For example:

a high-accuracy reactor or distillation model can be deployed within a traditional process flowsheet simulator using the gO:CAPE-OPEN object, in order to investigate process design alternatives once the detailed unit model has been designed.
gPROMS can link to CFD programs in many different ways