Molecular modeling of organic materials for flow simulation and digital source-rock physics
Conference Paper
Overview
Overview
abstract
Copyright 2017, Society of Petroleum Engineers. Model development for organic materials such as kerogen and bitumen using molecular building blocks is an important and fast-evolving science for source rock characterization. However, the size of the current models is much smaller than the representative elementary volume of organic in order to describe the macroscopic quantities such as diffusion coefficents and permeability. In addition, pore size distribution of the current models is skewed towards the lower end such that the predicted quantities are inaccurate. A new methodology is presented to build larger organic models to overcome the scale-dependence issue. A solid organic skeleton can be built using 3D tomographs which can be obtained from high-resolution microscopy such as TEM. The skeleton is populated with atoms distributed based on the organic matters maturity and elemental composition. As part of the new methodology to build larger organic model, we replace the atoms that make up the skeleton with an average representative atom whose bond length with the surrounding representative atoms is tuned to maintain the solid density and the structure of the skeleton unchanged. The average force field parameters are calculated based on kerogen's elemental composition. Permeability of this simplified organic model is measured using molecular dynamics simulation of steadystate fluid flow through the model pore-network. When the transport simulation results of the simplified organic model are compared to its counterpart carrying exact molecular description, the simplified model is accurate for the calculations of permeability, tortuosity, and saturations and reduced the computational cost significantly. The simplified model can be applied to large samples and plugged into the existing digital rock workflows, to utilize meaningful pore connection information provided from tomograhy.