Molecular simulation of the thermophysical properties and phase behaviour of impure CO₂ relevant to CCS

Abstract

Impurities from the CCS chain can greatly influence the physical properties of CO 2 . This has important design, safety and cost implications for the compression, transport and storage of CO 2 . There is an urgent need to understand and predict the properties of impure CO 2 to assist with CCS implementation. However, CCS presents demanding modelling requirements. A suitable model must both accurately and robustly predict CO 2 phase behaviour over a wide range of temperature and pressure, and maintain that predictive power for CO 2 mixtures with numerous, mutually interacting chemical species. A promising technique to address this task is molecular simulation. It offers a molecular approach, with foundations in firmly established physical principles, along with the potential to predict the wide range of physical properties required for CCS. The quality of predictions from molecular simulation depends on accurate force-fields to describe the interactions between CO 2 and other molecules. Unfortunately, there is currently no universally applicable method to obtain force-fields suitable for molecular simulation.In this paper we present two methods of obtaining force-fields: the first being semi-empirical and the second using ab initio quantum-chemical calculations. In the first approach we optimise the impurity force-field against measurements of the phase and pressure-volume behaviour of CO 2 binary mixtures with N 2 , O 2 , Ar and H 2 . A gradient-free optimiser allows us to use the simulation itself as the underlying model. This leads to accurate and robust predictions under conditions relevant to CCS. In the second approach we use quantum-chemical calculations to produce ab initio evaluations of the interactions between CO 2 and relevant impurities, taking N 2 as an exemplar. We use a modest number of these calculations to train a machine-learning algorithm, known as a Gaussian process, to describe these data. The resulting model is then able to accurately predict a much broader set of ab initio force-field calculations at comparatively low numerical cost. Although our method is not yet ready to be implemented in a molecular simulation, † Electronic Supplementary Information (ESI) available: [details of any supplementary information available should be included here]. See

Molecular Models for the Hydrogen Age: Hydrogen, Nitrogen, Oxygen, Argon, and Water

“…The force fields for Ar, N 2 , and O 2 were taken from the work of Vrabec et al These force fields were adjusted to experimental VLE data and it was shown that they are an adequate choice for numerous fluid mixtures. ,− For H 2 O, the widely appreciated TIP4P/2005 force field, which performs well for a large variety of thermodynamic properties, − was employed.…”

“…Several literature studies already dealt with some of these subsystems. Stoll et al and Vrabec et al performed molecular simulation studies on binary mixtures related to dry air (N 2 + O 2 , N 2 + Ar, and Ar + O 2 ), which were then used to predict cryogenic vapor–liquid equilibria (VLE) of dry air (N 2 + O 2 + Ar). , Furthermore, humid air (N 2 + O 2 + Ar + H 2 O) was examined by Eckl et al Related studies with a different objective, that is, CO 2 mixtures for carbon capture and sequestration (CCS) and the like, were conducted by means of molecular modeling and simulation by Vrabec et al (N 2 + O 2 + CO 2 ), Tenorio et al (H 2 + CO 2 ), and Cresswell et al (binary mixtures of CO 2 with H 2 , N 2 , Ar, and O 2 ). Moreover, several studies employed the statistical associating fluid theory (SAFT) in its various forms or cubic EOS to study mixtures containing CO 2 . , Solubilities of binary aqueous systems (N 2 + H 2 O and H 2 + H 2 O) were examined by Sun et al with the SAFT–Lennard-Jones EOS.…”

Gaussian process models of potential energy surfaces with boundary optimization

“…Previously, Latin hypercube (LHC) sampling, which is a space-filling design, has been employed to build data sets when modelling PESs with GPs [17][18][19]61 . The approach of Uteva et al 17 entailed designing LHCs in inverse separations over a range of angles that specified the symmetry-distinct region 17 .…”

“…This approach proceeds by training a statistical technique on a relatively small set of data from ab initio calculations on the PES of interest, known as the training set. Many such techniques have been employed to predict the energy in these algorithms, including neural networks [9][10][11][12][13] , moment tensors [14][15][16] and Gaussian processes [17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34] (GPs).…”

“…GP models have also produced promising results in applications to intermolecular interactions [17][18][19] . Initially 17 , the traina) Electronic mail: jack.broad@nottingham.ac.uk (author to whom correspondence should be addressed) b) Electronic mail: richard.wheatley@nottingham.ac.uk ing sets for these models were constructed with Latin hypercube sampling [46][47][48] though sequential design strategies 49,50 , which achieve a prescribed accuracy with a smaller training set, have been shown to outperform such methods 19 .…”