How can Model-Based Innovation help me?

First of all, what is Model-Based Innovation?

Mode-Based

Combining chemistry and process models

"Chemical modellers develop detailed chemistry models that are delivered back into the chemical engineering ‘unit-and-stream’ environment to provide process modellers with increasingly-accurate predictive models."

– Scott Owens & Richard Jarvis, Nexia Solutions, working on decommissioning models for complex nuclear systems.

Model-Based Innovation uses high-accuracy Advanced Process Models within modern R&D methodologies to:

accelerate innovation cycles by application of a variety of model-based and model-centric techniques
manage risk through the use of high-accuracy quantitative information and formal mathematical risk analysis methods
integrate the R&D and engineering functions of an organisation

Integrating R&D and Engineering

Process organisations have long sought a way to integrate their R&D and Engineering functions more closely.

Frequently, R&D and Engineering are performed in a sequential fashion. A programme of experimentation results in large amounts of data, which is then used in the Engineering phase some months later.

With luck, the data will be appropriate. However, more often than not the data is less-than-ideal: it might have been collected based on incorrect assumptions about the design; critical areas might have been overlooked; or fundamental aspects of the design might have changed since the data was collected.

There are only two alternatives in such a case: embark on another – potentially expensive and lengthy – experimentation programme, or design with poor data.

How does integrating R&D and Engineering help?

By more closely integrating the two functions, and performing R&D in parallel with Engineering as far as possible:

it is possible to ensure that R&D is targeted at specific elements of the design
any new data requirements identified by Engineering can quickly be incorporated into the current experimental programme
Engineering can provide R&D with very specific requirements – for example, to "measure x and y in the range t1 to t2"
R&D can provide Engineering with validated models that can simply slot into the Engineering modelling effort.

The approach sounds obvious. However, in practice the challenge has been in finding a common framework between these two very different parts of the organisation.

Advanced Process Models provide the medium for capturing information and transferring it between sections of the organisation in a form that can generate value in many different ways.

Over and above this, modelling can be used for data capture and analysis – for example, to build a definitive reaction set and accurately estimate reaction kinetic parameters – which can then be used to refine the R&D process itself as well as assist Engineering.

For example, a detailed model of a reaction set can not only be used to design a reactor, it can also be used in model-based experiment design to design the optimal set of experiments to enhance the accuracy of its own parameters.

In Nexia Solutions, for example, chemists create definitive models of the chemistry of nuclear processes, which are then incorporated as part of Engineering's process models. However, full Model-Based Innovation techniques can take this a step further.

So how does Model-Based Innovation work?

Model-Based Innovation is more than just simulating a reactor. By applying high-accuracy Advanced Process Models withing modern R&D methodologies, MBI helps to simultaneously refine research information and provide data and models for engineering design. This helps to accelerate the innovation process significantly.

One of the key aspects of MBI is model-centred or model-centric experimentation. This involves:

Model-based data analysis. All experimental data ultimately needs to be analysed and used to adjust key design parameters such as reaction kinetic coefficients or heat transfer coefficients. gPROMS provides a way to do this using rigorous first-principles models rather than statistical methods, thereby introducing a new level of accuracy.
Model-targeted experimentation: MBI's approach is, rather than using experimentation to improve the design, using experimentation to improve the models that are used to optimise the design.

This significantly leverages and enhances the value of the information in the experiments, and provides a powerful and flexible tool to designers that allows them to investigate many different design possibilities and quickly and confidently screen alternatives.

Model-based experiment design. If the data are inadequate, it is possible to use models of the experimentation process to design experiments that provide the maximum amount of information from the smallest number of experiments.

Proceeding in parallel with the model-centred experimentation programme, as soon as viable data are available, is model-based design optimisation. This uses the validated chemistry and other fundamental models within a detailed engineering framework. The Model-Based Innovation cycle - from model-centric experimentation to model-based design optimisation

So how does MBI help in risk management ?

Modelling analysis tools provide the ability to identify where areas of uncertainty lie, and which parameters make a critical difference to the design. It is on refining the knowledge and quantification of these parameters where R&D effort needs to be focused.

Model-Based Innovation helps manage risk in two ways:

Implicit risk management

By using models with an increasingly high degree of predictive accuracy, design risk is continually being reduced. High-accuracy models can eliminate the need for a lot of physical testing, meaning that more time and money can be spent on the physical testing that is performed.

Explicit risk management

Risk can be quantified formally through the application of mathematical risk analysis tools based on the accurate quantitative predictions of the model.

The plots on the right show the effect of uncertainty in a parameter theta on the profitability of the company.

This tool can be applied to determine where R&D funds are best spent.

What does this all mean to me and my organisation?

What this means is that you can accelerate innovation and reduce or confidently manage risk.

We are happy to discuss this in much more detail if you would like further information; please call one of our offices or email info@psenterprise.com.