Page 345 - Fundamentals of Gas Shale Reservoirs

15

MOLECULAR SIMULATION OF GAS ADSORPTION
IN MINERALS AND COAL: IMPLICATIONS FOR GAS
OCCURRENCE IN SHALE GAS RESERVOIRS

Keyu Liu , Shuichang Zhang , Shaobo Liu and Hua Tian 1
1,2
1
1
1 PetroChina Research Institute of Petroleum Exploration and Development, Beijing, China
2 CSIRO Division of Earth Science and Resource Engineering, Perth, WA, Australia

15.1 INTRODUCTION
The natural systems usually consist of a vast number of
particles. It is impossible to find properties of such complex
15.1.1 Molecular Dynamics Simulation
systems analytically. MD circumvents this problem by using
Molecular simulation (MS) includes molecular dynamics (MD) numerical methods. It has been widely used as a useful
and Monte Carlo (MC) simulations. MD simulation (Allen predictive tool for new materials without synthesizing them
and Tildesley, 1989; Frenkel and Smit, 1996; Sadus, 1999) is and for virtual experiments to provide an insight into rela­
based on the principles of Newtonian mechanics. It simulates tionship between microstructures and macrothermodynamic,
physical movements of interacting atoms and molecules in a elastic, dielectric and transport properties under some extreme
complicated and complex system. The trajectories of atoms conditions, where experiments are impractical or impossible.
and molecules are determined by numerically solving the MD serves as a bridge between the microscopic and the
Newton’s equations of motion, where forces between the macroscopic realms and a bridge between theoretical and
atoms and molecules are defined by molecular mechanics experimental work. Testing agreement between theory
force fields. Thermodynamic, structural, and transport prop­ and simulations allows theory to be refined and comparing
erties, such as potential energy, radial distribution function simulation experiment results to be extrapolated.
(RDF), and diffusion coefficient are derived from the trajec­ Advantages of MD over experiments lie in the possibility
tories of a system according to the principle of statistical of obtaining detailed time evolution of a system and allow­
mechanics. MS can also be performed in the grand canonical ing the information on the dynamics of the system and the
statistical ensemble (GCMC) (Zhang et al., 2014a). atomic processes to be fully presented, which otherwise
The basic procedure of MD simulation includes (i) cre­ might be difficult or impossible to realize in the laboratory.
ating a system including hundreds and thousands of atoms It can solve problems with many degrees of freedom and
and molecules, (ii) applying the law of classical Newtonian under a variety of initial and external conditions.
mechanics to all the particles in the system, (iii) solving the The disadvantage of MD is that the reliability of the cal­
coupled Newtonian’s equations of motion numerically and culations depends on the quality of the employed interaction
obtaining coordinates and velocities and each particle by model, as the construction of the model always involves
integration, (iv) updating the positions and velocities of compromises between accuracy and speed of calculation. It
the particles and recording the trajectories and velocities has its intrinsic time and length scale problem, where simu­
of the particles for every instant of time, and (v) computing lation size is at nanoscale level with simulation durations
properties of interest. The MD simulation workflow is shown being typically no more than hundreds of nanoseconds.
in Figure 15.1. Nowadays, MD has been developed into an extremely useful,

Fundamentals of Gas Shale Reservoirs, First Edition. Edited by Reza Rezaee.
© 2015 John Wiley & Sons, Inc. Published 2015 by John Wiley & Sons, Inc.