Automated Prediction of Side Reactions to Support Chemical Process Plant Design
Hafnium Labs and Software for Chemistry & Materials (SCM) proudly announce their collaboration on Automated Prediction of Side Reactions to Support Chemical Process Plant Design. The AutoReactPro Eurostars project started in October 2021 to develop the first automated tool to predict side reactions and their effects on chemical process plant design.
Faster and more optimized process design is necessary to make chemical plants greener. This includes dealing with unexpected side reactions, which often cause issues such as reduced efficiency or damage. For example, trace chemicals in waste feedstocks can cause unexpected corrosion in pipes and vessels, leading to expensive repairs, lost revenue, and slowing the transition from fossil to carbon-neutral feedstocks.
The chemical industry can save costs and speed up the green transition by identifying side reactions that may cause problems during early design: an estimated €750m/yr in value and 100 Mt/yr CO2 savings can be delivered through faster process innovation and avoidance of plant failures.
By combining Hafnium Labs’ expertise in physical property modeling and SCM’s expertise in atomistic modeling and reaction prediction methods, the goal of AutoReactPro is to develop software to automatically predict unforeseen side reactions and their impact on process streams. This will guide experts on possible side reactions to investigate further during early chemical process design.
The SCM and Hafnium Labs expert software developers and scientists look forward to hearing about your next green plant design – and discuss how AutoReactPro can help you save time, money, and resources!
SCM’s core strengths are related to atomistic scale modeling of molecules and periodic structures in chemistry and materials science. For this SCM offers the integrated Amsterdam Modeling Suite (AMS). AMS includes the well-known Density Functional Theory (DFT) code ADF, the DFT code BAND for periodic structures, fast semi-empirical methods (DFTB & MOPAC), reactive molecular dynamics (ReaxFF), classical and Machine Learning potentials, as well as the fluid thermodynamics code COSMO-RS, and convenient interfaces to some external codes. All modules are accessible through an intuitive common Graphical User Interface and python scripting environment, making it easy to switch between different levels of theory or to combine sequences of computational chemistry calculations in more complex workflows. Both independently and with partners, SCM is targeting more complex problems to enable high-throughput and multiscale simulations and improve speed and accuracy through machine learning methods. The collaboration with Hafnium Labs fits nicely with our goal to further develop automated reaction detection and analysis tools.
Hafnium Labs supports chemicals, energy, and pharma companies by solving one of the biggest challenges in chemistry: Obtaining reliable physical properties fast.
Taking an “all data, all models” approach, our Q-props software drastically improves the accuracy and reliability of physical property modeling – from pure compounds to complex mixtures.
Combining advanced models from quantum chemistry, thermodynamics, and machine learning with experimental data, Q-props always provides the best possible property prediction and a prediction-specific uncertainty, so you know how reliable each prediction is and how to improve it. Q-props comes with the world’s largest database of combined experimental and calculated physical property data, allows integration of customer in-house data and models, and integrates in 3rd party applications such as process simulators. Q-props supports decision-making from molecular discovery to process optimization, helping researchers and engineers obtain reliable properties for new and complex chemistries and integrating them in their digital designs. AutoReactPro, developed with SCM, is a natural extension, helping engineers get more value out of their process simulations by highlighting the potential effect of unforeseen side reactions during early design.
Contact information
For Hafnium Labs, please contact Dr. Bjørn Maribo-Mogensen bmm@hafniulabs.com
For Software for Chemistry & Materials, please contact Dr. Fedor Goumans, goumans@scm.com
