07122-22

Goal
The goal of this project is to determine the physical mechanisms that limit gas recovery from tight rock formations, using high-resolution 3D imaging of low-permeability reservoir rocks combined with two-phase, pore-scale fluid flow modeling. The results will be used to investigate methods of changing formation properties volumetrically to optimize gas production from tight reservoirs.

Performers
Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA 94720

Background
Gas production from gas shales and tight sands has grown significantly over the past 20 years. However, gas recovery from unconventional reservoirs remains low, primarily because of the low permeability of the reservoir rock. Rapid production rate decline is typical for tight gas wells, and much of the in-place gas remains unrecoverable with current technologies. Although reservoir stimulation through hydraulic fracturing increases the permeability close to the well-bore, it does not provide access to more distal gas that is essentially trapped in the formation.

This project attempts to address these problems by looking in detail at the exact physical mechanisms that act to block gas flow through tight formations. Part of the problem is that the pore-scale geometry of tight reservoir rocks is not well-known. Thin sections do not provide a 3D picture of the rock’s structure, and traditional scanning technologies do not have sufficiently high resolution to image pore-scale features.

This project will carry out high-resolution, 3D digital imaging of low-permeability sandstones and shales using Advanced Light Source and Focused Ion Beam technologies. Imaging results will be used to develop a 2-phase fluid-flow simulation code to model gas flow at the pore scale. Pore-scale modeling will be verified using petrophysical data acquired in the laboratory and in the field. In addition, depositional models for gas-bearing, low-permeability formations will be developed. The project will be carried out by a team of LBNL scientists and engineers using tight gas reservoir data and core samples provided by industry partners Schlumberger, BP, and Chevron.

Deliverables for the project will include: 1) processed 3D images of low-permeability reservoir samples; 2) algorithms and codes for simulating pore-scale gas flow; 3) a model of retrograde gas condensation; 4) depositional models for the formation of tight gas rocks; and 5) a final report summarizing all technical aspects of the project.

Potential Impacts
This project should result in a better understanding of the physical mechanisms that restrict gas flow in tight formations. In the short term, this knowledge should lead to a better understanding of the basic processes governing gas recovery in reservoir rocks with low matrix permeability. Over the longer term, this improved understanding may lead to the development of new strategies and tools to enhance gas production from tight formations.

Such new strategies, if adopted, could lead to an increase in domestic gas production in the U.S., improved ultimate recovery for tight gas wells, and extended field life for unconventional gas plays. Increased domestic gas production would result in increased tax revenues, royalties, and regional economic benefits. Improved recovery for individual, existing wells has the added benefit of decreasing the environmental footprint of a field development program.

Accomplishments
The effort over the past year was focused in two principal areas:

1. High-resolution imaging of tight sand and shale samples at three user facilities at the Lawrence Berkeley National Laboratory: the Advanced Light Source (ALS) facility, Molecular Foundry (MF), and the National Center for Electron Microscopy (NCEM).
2. Development of efficient analytical and numerical methods and codes for numerical evaluation of the capillary pressure, capillary-equilibrium two-phase fluid distribution, and relative permeability functions from the pore space geometry.

The industrial partners, BP, Chevron, and Schlumberger, shared a number of cores and samples of reservoir rock, as well as some petrophysical data from laboratory tests. A New Albany shale sample has been provided by the Gas Technology Institute (GTI). The high-resolution images suggest the following observations:

1. The micron-scale resolution x-ray CT imaging at the ALS provides data suitable for 3D reconstruction of the pore space geometry for the tight sands. Submicron-scale features characterizing the microporosity of the rock may require a higher resolution.
2. Scanning Electron Microscope (SEM) images show the geometry of clay crystal clusters attached to individual grains. A thorough quantitative analysis is needed to evaluate the relative volume of the pore space filled with clay clusters.
3. The x-ray micro-CT imaging of shale showed the geometry of the very few, if any, cracks and voids, and the geometric distribution of grains. However, micron-scale resolution is insufficient for studying the geometry and the mineral composition of shale in detail.
4. The SEM/Focused Ion Beam imaging technique (SEM/FIB) was applied for 3D reconstruction of shale structure. This imaging technique resolved the nanometer-scale features with amazing quality. The SEM/FIB imaging is based in sequential FIB milling and SEM imaging and, therefore, is destructive.
5. A successful test of mapping the elemental distribution in shale using the SEM/FIB/Energy Dispersive x-ray Spectroscopy (EDS) technique was performed at the NCEM. This work will be continued.

The modeling effort was focused on developing tools and methods for analysis of the rock properties from the 3D images described above. The software tools and codes expand and customize the family of algorithms based on the method of Maximal Inscribed Spheres (MIS). The segmentation and visualization procedures involve custom-made original object-oriented cluster-search and connectivity codes along with open-source and commercial tools (ImageJ, ImageMagick, GraphicsMagick, Scilab, ParaView, GNUPlot, and Avizo). Flow simulation codes use the high-performance computational libraries: SparceLib++, VT++, and IML++ by NIST.

1. The verification of the flow simulations against exact known solutions showed high accuracy of the numerical results.
2. A variety of boundary conditions, including slip and no slip conditions were implemented. The impact of slip condition was correlated to the computed permeability of the sample.
3. The MIS calculations were used to evaluate the capillary-equilibrium fluid distributions at different capillary pressures, saturations, see Figure 1, left. Simulations showed that even a relatively small volume of the wetting fluid can make the gas phase disconnected.

   
   Figure 1. Computed two-phase fluid distribution in a 1.8-micron resolution 3D microtomography image of a 1 mm3 sample of tight-sand, left. The red color denotes gas-occupied voxels, whereas water-occupied voxels are blue. Relative permeability curves are estimated from the computed pore-scale fluid distribution, right. To compute the capillary equilibrium fluid distribution, we employ the Method of Maximal Inscribed Spheres.

4. The computed fluid distributions were used as input data for evaluation of relative permeability curves see Figure 1, right. Additional connectivity analysis facilitates permeability evaluation.
5. Flow simulations is the most computationally demanding part. The parallel nature of the relative permeability evaluation allows for splitting the problem into independent computations, which can run on separate computers. All simulations were performed on a number of desktop computers, using the time when the computers were not loaded by other tasks.

Current Status
The key tasks to be undertaken are outlined below.

High Resolution Digital Imaging Using x-ray micro CT and FIB/SEM Technologies The LBNL research team will continue to select and prepare representative core samples of tight sand and shale reservoir rocks obtained from their industry partners. Using combined Focused Ion Beam and Scanning Electron Microscopy technologies, the researchers will continue developing high-resolution, digital, 3D images of the pore and grain structure in tight sand and shale samples. Micron-scale imaging of 3D pore structures will be accomplished using LBNL’s Advanced Light Source technology. EDS elemental mapping will complement the nanometer-scale FIB/SEM imaging.

Model Simulation of Pore-Scale, Two-Phase Fluid Flow The LBNL researchers will continue developing codes for simulating pore-scale fluid distribution and flow of gas and gas condensate through low-permeability rocks. In addition, the models will incorporate the impact of retrograde condensation at fractures and narrow pore throats on the transport of gas. The objective will be to study the pore-scale mechanisms of gas flow blockage by the gas condensate. Pore-scale models will be verified using available petrophysical data measured in the laboratory and in the field.

Depositional Model Development The research team will develop depositional models suitable for simulation of tight gas sand and shale formations. The models will be verified using digital imaging results as well as petrophysical information from the field.

Technology Transfer LBNL will work closely with RPSEA to develop and implement an effective program for technology transfer. Technology transfer will include journal publications, conference presentations, and technical reports. There are plans of setting up a project-dedicated website for posting new results for public access.

Project Start: November 1, 2008
Project End: November 1, 2011

DOE Contribution: $1,071,105
Performer Contribution: $420,000

Contact Information:
RPSEA – Charlotte Schroeder (cschroeder@rpsea.org or 281-690-5506)
NETL – Virginia Weyland (Virginia.Weyland@netl.doe.gov or 281-494-2517)
LBNL – Dmitriy Silin (dsilin@lbl.gov or 510-495-2215)

Presentations The project results have been presented at a number of conferences.

Silin, D.,. Ajo Franklin, J. B., Cabrini, S., Kneafsey, T. J., MacDowell, A. Nico, P. S., and Tomutsa, L. Pore-scale studies of unconventional reservoir rocks. Eos Trans. AGU, 90(52), Fall Meet. Suppl., Abstract H23F-1018, 2009

Analyzing microtomography images of natural rocks with the method of maximal inscribed spheres. Numerical Modeling with Large 3D Data Sets ALS Workshop October 17, 2009

Silin, D. and Patzek, T. Predicting Relative-Permeability Curves Directly From Rock Images. SPE paper 124974 presented at the SPE Annual Technical Conference and Exhibition, 4-7 October 2009, New Orleans, Louisiana, 2009. Society of Petroleum Engineers.

Nico, Peter S. Jonathan B. Ajo-Franklin, Alastair McDowell, Dmitriy B. Silin, Liviu Tomutsa, Sally M. Benson, Yuxin Wu Synchrotron X-ray Micro- Tomography and Geological CO2 Sequestration. In Advances in Computed Tomography for Geomaterials, GeoX 2010. Ed. Khalid .A. Alshibi and Allen H. Reed. Wiley, & Sons, Hoboken, NJ, p. 374-380, 2010

The researchers were involved in organizing the “Numerical Modeling with Large 3D Data Sets ALS Workshop” held on October 15-17, 2009 at the Lawrence Berkeley National Laboratory.