Permeability Simulation with OPENFOAM and OPENPNM

Permeability and Filtration coefficients are some must-know factors for oil/gas reservoir management and optimization. OpenFOAM and Network-based software (OpenPNM for example) are our chosen open-source software toolkits for permeability simulation.

It’s hard to define an audience for this presentation, really. Anyone who wants to:

• Kill the time by learning about permeability of soils, or
• Prepare for a numerical simulation of permeability of soils

And understands the basics of:

• Flow through porous medium,
• OpenFOAM, or Python programming language,
• Image processing (we are using FIJI – Fiji Is Just ImageJ – ) Should be able to benefit from this presentation and related files (The way I presented it: Split it up to parts and give each part to the appropriate audience).

Abstract

One of the most important properties of hydrocarbon reservoirs is their capacity of passing fluids; ie, their permeability to fluids in question. The engineering knowledge have accumulated over the years to explain, optimize and use the behavior of reservoirs in this matter.

With a great focus on granular reservoirs, this presentation introduces key concepts regarding this interesting property of rocks, including the most effective way of comparing two reservoirs in respect to their permeability: The use of permeability coefficient; In addition, we’ve tried to present the most important factors affecting rock permeability, and how they affect it.

Based on these factors, several mathematical models -adhering to certain conditions- were formed; we’ve picked two well-known well-accepted models to feature: Kozeny-Carman (relatively simple porosity-permeability relationship); Happel-Brenner (more complex porosity-permeability relationship).

With their “very limited” porosity range, these models fail in predicting the permeability of many rock sample; especially if they require the determination of grain shape descriptors (sphericity, roundness … etc).

Thus, results from mathematical predictions are always treated with caution; The most accurate method to estimate permeability of a sample rock to a certain test fluid is to actually measure it (More precisely, calculate it from measured data). This paper features the most-common permeability tests methods (Principle: apply pressure gradients, measure the outgoing flow rate, and use Darcy’s law to determine permeability). We’ve also tried to show how field-based methods (Well tests and logs) are used to estimate permeability.

To formulate a better understanding of this concept, we’ve tried to use common simulation approaches to determine the permeability of a piece of porous material (a 2-D 1.024mm×0.728mm image generated with PoreSpy - a Python Library -):

• Directly simulate a viscous flow through medium’s pores using OpenFOAM (The most trusted method).
• Instead of real pore (complex) geometry, assemble the porous material to a network of spheres (pores) connected by cylinders (throats), and solve the flow there, using PpenPNM - also a Python Library -.

Downloads