Computational science augments traditional research methods by accelerating and guiding experimental work and providing insight into processes and results. It is used across Shell’s businesses to predict everything from the chemistry of catalysts and batteries to capturing flow through reactors, pipelines and rocks. These are complex simulations which require high performance computing and algorithm optimisation.
A key aim of computational science is to use computer models to predict the performance of materials and systems in specific situations. One of the most striking aspects of computational science projects is their breadth of scale. This multiscale modelling covers interactions at the atomic and molecular levels to the design of reactors in industrial plants.
Our grasp of computational technology helped us to lead the way in technological developments in exploration in the 1960s, 70s and 80s. Demand for computational design and analysis has increased dramatically since mid-2000’s across increasingly varied domains. The growth in computer power from Moore’s law has made realistic catalyst modelling and complex fluid flow studies possible that were unthinkable only 15 years ago. Shell has a diverse team of chemical engineers, mechanical engineers, aerospace engineers, chemists, material scientists, mathematicians, physicists and computational scientists. This expertise in mathematics and computing is what gives us such a strong advantage today in developing and adopting digital technology.
By combining data-based models with physics/chemistry based computational models, we augment the power of both by integrating the speed and agility of AI with the interpretability and explainability of Computational Science, we move towards an era of Augmented Intelligence, where we augment our decision making manifold. Find out more in our recent publication on developing machine learning models for materials datasets.
