Computational Applied and Geology Lab.

1. What is computational Rock Physics?

Fundamentally, as rock physicists/Earth scientists, we would like to understand and predict the properties of the Earth’s material and their evolutions in various situations. We usually perform lab measurements, find relations, and make theoretical but simplified models. This approach contributes to solve many problems and predicts important rock parameters that are hard to measure from other parameters, which are easier and/or cheaper to get, for example, Kozeny-Carman relation for hydraulic permeability, Gassmann’s equation, Biot effect, and squirt flow for velocity changes by fluid changes in rocks, grain contact models for elastic properties of granular media, many effective medium theories for seismic velocities for composite materials, and of course, many empirical relations between rock properties. Although these models give tremendous help to solve problems in situ and also give great insight on our understanding on physical properties of rocks, they have several drawbacks – (1) the model is sometimes over-simplified, and (2) those simplifications are model-dependent, i.e. a model is good for elastic property estimation does not work for the situations where we need to understand fluid flow.

Our attempt is to start with realistic pore geometry of Earth’s material. Using the real pore geometry, which consists of grains and pore space and which contains all the complexities, we try to calculate physical properties of the materials directly. Due to recent development of computational technology and new algorithms, now this is possible. From this approach, we can predict and understand the physical properties of Earth material more accurately, and can relate one property to another more easily since we use exactly the same models for all properties. Furthermore, this framework enables us to explore complex and coupled problems, such as two-phase flow in porous media and diagenesis, which involves multiple physical properties.