CHARACTERIZING STRUCTURE AND GEOCHEMISTRY OF SHALE PORES BY NEUTRON SCATTERING
Open Access
- Author:
- Gu, Xin
- Graduate Program:
- Geosciences
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 24, 2017
- Committee Members:
- Susan L. Brantley, Dissertation Advisor/Co-Advisor
Susan L. Brantley, Committee Chair/Co-Chair
Peter J. Heaney, Committee Member
Terry Engelder, Committee Member
Jason Kaye, Outside Member - Keywords:
- neutron scattering
shale
weathering
microfracture
porosity
Marcellus shale - Abstract:
- The commercial success of shale gas and oil plays have stimulated interest in the pore structure of shale, which controls the storage and transmission of oil and gas in shale reservoirs. The pore space also provides much of the surface area accessible to solutes, colloids, or microorganisms in shale and consequently determines the fluid-rock interactions that occur in this important rock type. However, the pore structure of shale is still hard to evaluate using routine core analysis or petrophysical techniques because of the fine–grained texture, presence of extremely small pore sizes with dimensions ranging over several orders of magnitudes (sub-nanometer to hundreds of micrometers), extremely low permeability, and complex organic-mineral associations. One technique that has proven useful in evaluating the porosity of shales is small-angle and ultra small-angle neutron scattering. These neutron scattering techniques can evaluate a wide range of pores with length scales from nanometers to micrometers in size. Additionally, neutron scattering data provides information on the fractal properties, pore orientation, pore volume, pore size distribution, specific surface area and fluid accessibility. In this dissertation, neutron scattering is used as the primary characterizing tool to probe the pore structures of shale. To use neutron scattering to investigate growth and development of pores in shale, both organic-rich and organic-poor shales are investigated under both weathered and unweathered conditions. To use neutron scattering measurements to interpret shale porosity, a new data processing procedure and a new method for porosity calculation was developed to account for shale anisotropy. The degree of anisotropy of shale that affects neutron scattering depends on sample orientation: for most thin sections cut in the plane of bedding, the scattering pattern is isotropic, while for thin sections cut perpendicular to the bedding, the scattering pattern is anisotropic. Observations from focused ion beam scanning electron microscopy (FIB-SEM) show that the anisotropic scattering patterns can be attributed to elongated pores predominantly associated with phyllosilicates. A combination of neutron scattering, FIB-SEM, and nitrogen adsorption measurements were used to characterize pore structures in samples from two boreholes of Marcellus shale, the spatially most extensive shale gas play in the United States. The distribution of pores in Marcellus shale comprises a random fractal over several orders of magnitude in spatial dimension (nanometer to tens of micron). Both porosity and specific surface area estimated through neutron scattering are in good agreement with the measurements by nitrogen adsorption. A novel approach was developed to characterize the structure of pores in organic matter (OM) in the shale. Additionally, water-accessible porosity determined by neutron scattering was distinguished from total porosity by exploiting contrast matching techniques. In contrast matching, the shale matrix is saturated with a suitable mixture of deuterated and protonated water, made to match the scattering length density of the shale. In addition, the porosity within the mineral matrix was distinguished from porosity in the OM by comparing neutron measurements on samples before and after combustion and removal of OM. In the Marcellus, OM generally hosts 20−50% of the total porosity measured by neutron scattering. This porosity occupies as much as one third of the OM volume in some samples. OM pores in some samples are <2 nm in width and water-inaccessible. In contrast, sponge-like OM with pores >20 nm in width are dominant in some samples and these pores exhibit characteristics of water accessibility. This latter observation contradicts the current paradigm in the literature that OM porosity is organophilic and therefore not likely to host water. A combination of neutron scattering and SEM imaging was also used to investigate porosity formation during weathering of both organic-rich and organic-poor shales. To understand the controls on the evolution of weathering profiles that underlie hilly and mountainous regions, shale samples from boreholes and soil cores were collected from catchments located in Pennsylvania, USA (Shale Hills Critical Zone Observatory, SSHCZO), California, USA (Eel River Critical Zone Observatory, ERCZO), and Yilan, Taiwan (Fushan Experimental Forest, Fushan). The shale recovered from these three sites have similar mineralogical compositions, but are located in vastly different climate and tectonic settings. In particular, the erosion rate at Fushan (3-6 mm/yr) is much faster than at ERCZO (0.2-0.4 mm/yr) and SSHCZO (0.015 mm/yr); the average annual precipitation at Fushan is higher (3.4-4.3 m/yr vs. 1.7 m/yr at ERCZO and 1 m/yr at SSHCZO); and the mean annual air temperature at Fushan is higher (18 °C vs. 10 °C at ERCZO and SSHCZO). Regardless of the differences in erosion rate, the chemical weathering profiles at these three site are similar: pyrite and carbonate are depleted at depth; illite is not depleted in the consolidated rock material but is lost from the soil; the depletion profiles for chlorite and plagioclase initiate between the pyrite and illite fronts. At SSHCZO and ERCZO, the interface between weathered and unweathered rock roughly coincides with the zone of fluctuations in the water table and the samples above the water table have higher porosity and water-accessibility than the samples below. However, the mechanisms of porosity formation in shale chips are different at these three sites: most porosity was generated through chemical weathering (primarily chlorite dissolution) at SSHCZO; whereas, microfracturing dominated porosity generation at ERCZO and Fushan. It is possible that the eroding landscapes at each of the three sites are moving toward balance between rates of erosion and weathering advance. In the rapidly eroding Fushan and ERCZO sites, microfracturing may promote faster water infiltration and weathering advance rates relative to the more slowly eroding SSHCZO. Weathered Marcellus shale samples were collected from an outcrop at Frankstown, Pennsylvania, USA. The Marcellus shale formation in outcrop overlies a layer of carbonate (Onondaga Limestone) at ~ 10 m below land surface which is characterized by low porosity as measured by neutron scattering (<3%). All the shale samples above the carbonate layer are almost completely depleted in carbonate, plagioclase, chlorite and pyrite. The porosities in the weathered Marcellus shale are twice as high as in the unweathered samples from deep boreholes elsewhere in Pennsylvania. The pore size distribution exhibits a broad peak for pores of sizes in the range of tens of microns, likely due to the loss of OM and/or dissolution of carbonate during weathering. The weathering extent of silicates in the Marcellus shale has proceeded to a greater extent than in the organic-poor, Rose Hill shale collected in nearby SSHCZO. The greater weathering extent in the Marcellus shale despite the similarity in climate and erosion rate in these two neighboring locations is attributed to 1) the formation of micron-size pores as OM is oxidized that increase the infiltration rate and the extent of silicate weathering in weathered Marcellus shale; 2) the higher pyrite/carbonate ratio in the Marcellus shale as compared to the Rose Hill shale. The latter characteristic generates excess acidity as the pyrite oxidizes, enhancing the dissolution of the silicates. The characterization approaches developed in this study provide quantitative descriptions of the size, roughness, distribution, and fluid accessibility of the complex pore structure of shale. The results from this study on pore size distribution and water-accessibility in both organic and mineral phases in Marcellus shale may facilitate more accurate models of gas transport and storage capacity and of water−mineral interactions during hydrofracturing. The results from this study on shale weathering reveal the pathways of porosity formation as a function of tectonics, climate, and the mineral and OM content of the bedrock. These observations may ultimately yield better estimates of regolith depth and weathering rates on shale.