Panel_14427
Panel_14427
8:30 AM
5:00 PM
8:30 a.m.
Shale Matrix Permeability Evolution During Reservoir Depletion: Fluid Dynamic and Poroelastic Aspects
Exhibition Hall
By Y. Gensterblum, R. Heller, M. D. Zoback
The evolution of reservoir permeability is of fundamental importance in reservoir engineering. However, many of the commonly used concepts in reservoir may not be directly applicable to unconventional reservoirs due to their extremely low (and highly anisotropic) permeability, extremely small pore throat sizes. Therefore, the transport processes of compressible micro- to meso-porous rocks such as shales require a critical revision of the classical concepts. Due to the higher compressibility of shale reservoirs as compared to conventional reservoirs, some processes have to be considered as coupled such as the transition from Darcy-flow to slip-flow and the stress sensitivity of the permeability to pore throat compressibility, which is a poroelastic effect. We also develop a detailed description of the coupling between slip-flow and the stress sensitivity in unconventional reservoirs, and interpret experimental observations in light of this description. We characterize the transport properties of shales in a manner that includes a zero-effective-stress permeability coefficient, a stress sensitivity coefficient, an effective stress coefficient and slippage as a function of effective stress. We model single-phase matrix gas permeability during reservoir depletion for the Eagle Ford and Marcellus shale. This case study shows the significant influence of slip-flow, which can lead up to a two times higher permeability when the pore pressure decline below 10 MPa. Furthermore, considering the coupling between slip flow and poroelasticity in the permeability model, the permeability can be 50% higher for selected samples.
The evolution of reservoir permeability is of fundamental importance in reservoir engineering. However, many of the commonly used concepts in reservoir may not be directly applicable to unconventional reservoirs due to their extremely low (and highly anisotropic) permeability, extremely small pore throat sizes. Therefore, the transport processes of compressible micro- to meso-porous rocks such as shales require a critical revision of the classical concepts. Due to the higher compressibility of shale reservoirs as compared to conventional reservoirs, some processes have to be considered as coupled such as the transition from Darcy-flow to slip-flow and the stress sensitivity of the permeability to pore throat compressibility, which is a poroelastic effect. We also develop a detailed description of the coupling between slip-flow and the stress sensitivity in unconventional reservoirs, and interpret experimental observations in light of this description. We characterize the transport properties of shales in a manner that includes a zero-effective-stress permeability coefficient, a stress sensitivity coefficient, an effective stress coefficient and slippage as a function of effective stress. We model single-phase matrix gas permeability during reservoir depletion for the Eagle Ford and Marcellus shale. This case study shows the significant influence of slip-flow, which can lead up to a two times higher permeability when the pore pressure decline below 10 MPa. Furthermore, considering the coupling between slip flow and poroelasticity in the permeability model, the permeability can be 50% higher for selected samples.
Panel_14923
Panel_14923
8:30 AM
5:00 PM
8:30 a.m.
Gas Transport Properties and Its Anisotropy of the Shale Matrix — A Review
Exhibition Hall
By Y. Gensterblum, M. D. Zoback
For a number of decades, scientists and engineers have been investigating and describing storage and transport mechanisms in porous media such as reservoir rocks. This work has resulted in the development of concepts such as single phase and multi-phase flow, which describe the movement of fluids in conventional reservoir rock types such as sandstones and carbonates. However, many of these concepts may not be directly applicable to unconventional reservoirs. For example, shale gas reservoirs consist of organic-rich shale matrix, which have high compressibility, very small pore throats, extremely low and anisotropic porosity and permeability. The models developed to describe conventional reservoirs may not accurately describe the processes at work in these rocks. We aim to characterize the transport processes occurring in unconventional reservoirs. We will examine processes occurring at various spatial scales, ranging from fracture flow on the centimeter scale down to slip-flow on the nanometer scale. Due to the softer nature of tight shales, many processes, such as slip-flow and the pore-throat compressibility, will have to be considered as coupled. We collect more than 300 publications and interpret experimental observations in light of this description.
For a number of decades, scientists and engineers have been investigating and describing storage and transport mechanisms in porous media such as reservoir rocks. This work has resulted in the development of concepts such as single phase and multi-phase flow, which describe the movement of fluids in conventional reservoir rock types such as sandstones and carbonates. However, many of these concepts may not be directly applicable to unconventional reservoirs. For example, shale gas reservoirs consist of organic-rich shale matrix, which have high compressibility, very small pore throats, extremely low and anisotropic porosity and permeability. The models developed to describe conventional reservoirs may not accurately describe the processes at work in these rocks. We aim to characterize the transport processes occurring in unconventional reservoirs. We will examine processes occurring at various spatial scales, ranging from fracture flow on the centimeter scale down to slip-flow on the nanometer scale. Due to the softer nature of tight shales, many processes, such as slip-flow and the pore-throat compressibility, will have to be considered as coupled. We collect more than 300 publications and interpret experimental observations in light of this description.
Panel_14919
Panel_14919
8:30 AM
5:00 PM
8:30 a.m.
Eagle Ford Pores, Pore Throats, and Fluid Content Identified Through NMR Analysis
Exhibition Hall
By J. M. Shultz, H. Daigle, U. Hammes, K. Smye, S. C. Ruppel
Mudrock porosity is associated with both inorganic and organic matter, and hydrocarbons are found in both. The upper Eagle Ford is considered to be dominated by inorganic porosity, while the lower Eagle Ford is considered to have more organic-hosted porosity related to high organic content. The differences in inorganic versus organic pore types play a large role with regard to pore networks. This study investigates Eagle Ford mudrock pores through the use of nuclear magnetic resonance (NMR) in order to better quantify porosity values for these unconventional reservoirs. Recently, laboratory-based NMR has been used to measure fluid content and pore volume of mudrocks affordably and nondestructively. While NMR is quite versatile, important limitations exist, including the inability to directly measure pore-throat sizes or pores that are not fluid filled. However, calibration to mercury injection capillary pressure (MICP) measurements from finely grinding the same sample into powder yields interpretable results. Using a total of 28 samples from three wells in Karnes and Maverick Counties, South Texas, this study examines how pore-throat size distribution and fluid content differ vertically in the succession. The lithology and facies vary, both vertically within each well and laterally between wells. Six facies were identified among the three wells. With the aid of geochemical analysis such as XRF these lithologies can be broken down into subfacies based on calcite content and minor elements, such as molybdenum and vanadium, related to anoxia. These subfacies were used to pick the sample points. While a mixture of pore types is anticipated, samples with higher organic-hosted porosity are expected to have smaller pore throats and contain more oil. Early results indicate measurable pore throat sizes between 0.001 and 0.01 microns filled with low to medium-viscosity fluids, likely a mixture of water and oil, and intrusion corrected porosities ranging between 5.2% to 9.4% in the lower Eagle Ford. Greater heterogeneity of the pore system in the lower Eagle Ford has been confirmed through an SEM study. Using NMR to evaluate the linkage between facies changes and pore-throat-size distribution and content will aid future play-wide analysis.
Mudrock porosity is associated with both inorganic and organic matter, and hydrocarbons are found in both. The upper Eagle Ford is considered to be dominated by inorganic porosity, while the lower Eagle Ford is considered to have more organic-hosted porosity related to high organic content. The differences in inorganic versus organic pore types play a large role with regard to pore networks. This study investigates Eagle Ford mudrock pores through the use of nuclear magnetic resonance (NMR) in order to better quantify porosity values for these unconventional reservoirs. Recently, laboratory-based NMR has been used to measure fluid content and pore volume of mudrocks affordably and nondestructively. While NMR is quite versatile, important limitations exist, including the inability to directly measure pore-throat sizes or pores that are not fluid filled. However, calibration to mercury injection capillary pressure (MICP) measurements from finely grinding the same sample into powder yields interpretable results. Using a total of 28 samples from three wells in Karnes and Maverick Counties, South Texas, this study examines how pore-throat size distribution and fluid content differ vertically in the succession. The lithology and facies vary, both vertically within each well and laterally between wells. Six facies were identified among the three wells. With the aid of geochemical analysis such as XRF these lithologies can be broken down into subfacies based on calcite content and minor elements, such as molybdenum and vanadium, related to anoxia. These subfacies were used to pick the sample points. While a mixture of pore types is anticipated, samples with higher organic-hosted porosity are expected to have smaller pore throats and contain more oil. Early results indicate measurable pore throat sizes between 0.001 and 0.01 microns filled with low to medium-viscosity fluids, likely a mixture of water and oil, and intrusion corrected porosities ranging between 5.2% to 9.4% in the lower Eagle Ford. Greater heterogeneity of the pore system in the lower Eagle Ford has been confirmed through an SEM study. Using NMR to evaluate the linkage between facies changes and pore-throat-size distribution and content will aid future play-wide analysis.
Panel_14918
Panel_14918
8:30 AM
5:00 PM
8:30 a.m.
Occurrence of Organic-Matter Pores in Sub-1.0% Vitrinite Reflectance Mudrocks: Examples From the Devonian New Albany Shale, the Mississippian Barnett Shale and the Cretaceous Eagle Ford Formation
Exhibition Hall
By R. M. Reed, R. G. Loucks, S. C. Ruppel
There is controversy about the development of organic-matter (OM) pores in lower maturity organic-rich mudrocks, with some studies concluding OM pores are not formed until the gas window is reached (vitrinite reflectances > 1.3%). However, we have observed many samples in which OM pores are present in sub-1.0% vitrinite reflectance mudrocks from the Devonian New Albany Shale of the Illinois Basin (siliceous-argillaceous lithologies), the Mississippian Barnett Shale of the Fort Worth Basin (dominantly siliceous lithologies) and the Cretaceous Eagle Ford Formation from South Texas (dominantly calcareous lithologies). In these same rocks, OM pores are generally absent or rare at vitrinite reflectances less than 0.75%. Further controversy exists over being able to differentiate maturation-related OM pores from pores that remain when mobile organic matter incompletely fills pre-existing interparticle or intraparticle pores due to the presence of other fluid phases. Such pores tend to be smooth walled and larger than maturation-related OM pores and this mechanism typically forms only one or two pores in the center of an area of OM. Three cores were sampled in the Barnett, one core was sampled in the New Albany and four cores were sampled in the Eagle Ford. All samples were in the 0.75 to 1.0% vitrinite reflectance thermal maturity range. Pores were examined using a field-emission scanning electron microscope on surfaces prepared with broad-ion-beam milling using Ar-ions. OM pores observed are typically less than a micrometer in diameter. OM pore shapes are highly variable from nearly circular to highly elongate in outline. OM pore abundances are highly variable ranging from a scattered few to densely concentrated spongy patches. Differences in pore size, shape, and abundance may be related to underlying differences in the composition and structure of the original OM. OM pores are overall lower in abundance in sub-1.0% vitrinite reflectance mudrocks than in more mature mudrocks. However, heterogeneity of pore development is such that nonporous to highly porous organic matter is found across a range of thermal maturities. Variation occurs both within and between samples.
There is controversy about the development of organic-matter (OM) pores in lower maturity organic-rich mudrocks, with some studies concluding OM pores are not formed until the gas window is reached (vitrinite reflectances > 1.3%). However, we have observed many samples in which OM pores are present in sub-1.0% vitrinite reflectance mudrocks from the Devonian New Albany Shale of the Illinois Basin (siliceous-argillaceous lithologies), the Mississippian Barnett Shale of the Fort Worth Basin (dominantly siliceous lithologies) and the Cretaceous Eagle Ford Formation from South Texas (dominantly calcareous lithologies). In these same rocks, OM pores are generally absent or rare at vitrinite reflectances less than 0.75%. Further controversy exists over being able to differentiate maturation-related OM pores from pores that remain when mobile organic matter incompletely fills pre-existing interparticle or intraparticle pores due to the presence of other fluid phases. Such pores tend to be smooth walled and larger than maturation-related OM pores and this mechanism typically forms only one or two pores in the center of an area of OM. Three cores were sampled in the Barnett, one core was sampled in the New Albany and four cores were sampled in the Eagle Ford. All samples were in the 0.75 to 1.0% vitrinite reflectance thermal maturity range. Pores were examined using a field-emission scanning electron microscope on surfaces prepared with broad-ion-beam milling using Ar-ions. OM pores observed are typically less than a micrometer in diameter. OM pore shapes are highly variable from nearly circular to highly elongate in outline. OM pore abundances are highly variable ranging from a scattered few to densely concentrated spongy patches. Differences in pore size, shape, and abundance may be related to underlying differences in the composition and structure of the original OM. OM pores are overall lower in abundance in sub-1.0% vitrinite reflectance mudrocks than in more mature mudrocks. However, heterogeneity of pore development is such that nonporous to highly porous organic matter is found across a range of thermal maturities. Variation occurs both within and between samples.
Panel_14928
Panel_14928
8:30 AM
5:00 PM
8:30 a.m.
Variation of Wettability of Organic-Rich Shales With Thermal Maturity and the Implications for Oil and Gas Distributions
Exhibition Hall
By S. Peng, T. Zhang, G. S. Ellis, M. D. Lewan
Wettability of pore surfaces in organic-rich shales can significantly affect the relative distribution, storage, and productivity of gas and oil in these lithologies. Changes in organic matter structure during thermal maturation will affect its surface chemistry, therefore changing the wettability, as indicated by our recent experiments on methane and water vapor sorption. Consequently, the relative distribution of gas, oil, and water in shale systems is closely related to the thermal maturation of organic matter. An immature sample of the Woodford Shale was prepared by hydrous pyrolysis at five different time-temperature conditions to generate a suite of samples having different thermal maturities. Methane (CH4) adsorption isotherms were measured on the moisture-equilibrated and dry samples at 35°C and pressures up to 14MPa. Nitrogen and water vapor adsorption experiments were also conducted on the dry samples to characterize the pore-size distribution, surface area, and surface wettability. Water distribution and morphology were observed for these samples under various vapor pressure conditions with environmental SEM (ESEM). Comparison of water-wet and dry samples shows that moisture added to the system does not substantially reduce CH4 adsorption. This is consistent with observations from water vapor sorption isotherms that indicate the dominant hydrophobic nature of all the samples. However, the reduction of CH4 adsorption capacity is larger for the samples in the late oil-generation (367°C/72 hr, 40% reduction) and early oil-cracking (400 °C/72 hr, 53% reduction) stages compared to the early bitumen (300 °C/72 hr, 35% reduction) and maximum bitumen (333 °C/72 hr, 34% reduction) generation stages. These findings suggest that although organic matter is generally hydrophobic, there is a wettability trend for the organic matter of the studied samples that correlates with higher thermal maturity stages. These stages of petroleum formation are notably associated with the formation of insoluble hydrocarbon residue or pyrobitumen. In addition, the distribution and morphology of water observed by ESEM illustrates the increase in wettability of the high-maturity samples. This provides further evidence that pyrobitumen may play an important role in controlling the wettability of organic matter. These results also suggest that whole-rock wettability may be a useful proxy for determining the proportion of kerogen, bitumen, and pyrobitumen in shales.
Wettability of pore surfaces in organic-rich shales can significantly affect the relative distribution, storage, and productivity of gas and oil in these lithologies. Changes in organic matter structure during thermal maturation will affect its surface chemistry, therefore changing the wettability, as indicated by our recent experiments on methane and water vapor sorption. Consequently, the relative distribution of gas, oil, and water in shale systems is closely related to the thermal maturation of organic matter. An immature sample of the Woodford Shale was prepared by hydrous pyrolysis at five different time-temperature conditions to generate a suite of samples having different thermal maturities. Methane (CH4) adsorption isotherms were measured on the moisture-equilibrated and dry samples at 35°C and pressures up to 14MPa. Nitrogen and water vapor adsorption experiments were also conducted on the dry samples to characterize the pore-size distribution, surface area, and surface wettability. Water distribution and morphology were observed for these samples under various vapor pressure conditions with environmental SEM (ESEM). Comparison of water-wet and dry samples shows that moisture added to the system does not substantially reduce CH4 adsorption. This is consistent with observations from water vapor sorption isotherms that indicate the dominant hydrophobic nature of all the samples. However, the reduction of CH4 adsorption capacity is larger for the samples in the late oil-generation (367°C/72 hr, 40% reduction) and early oil-cracking (400 °C/72 hr, 53% reduction) stages compared to the early bitumen (300 °C/72 hr, 35% reduction) and maximum bitumen (333 °C/72 hr, 34% reduction) generation stages. These findings suggest that although organic matter is generally hydrophobic, there is a wettability trend for the organic matter of the studied samples that correlates with higher thermal maturity stages. These stages of petroleum formation are notably associated with the formation of insoluble hydrocarbon residue or pyrobitumen. In addition, the distribution and morphology of water observed by ESEM illustrates the increase in wettability of the high-maturity samples. This provides further evidence that pyrobitumen may play an important role in controlling the wettability of organic matter. These results also suggest that whole-rock wettability may be a useful proxy for determining the proportion of kerogen, bitumen, and pyrobitumen in shales.
Panel_14920
Panel_14920
8:30 AM
5:00 PM
8:30 a.m.
The Nature of Pore Structure in Middle Devonian Organic-Rich Black Shale From West Virginia and Pennsylvania
Exhibition Hall
By L. Song, T. Carr
Analyses of organic-rich mudstones from three wells that penetrated the Marcellus Shale in West Virginia and Pennsylvania were performed to evaluate the nature of pore structure (porosity and permeability) in both organic matter (OM) and inorganic matrix (IM). Samples include 32 core plugs with different levels of thermal maturity and TOC from Mahantango Formation and Marcellus Shale. For each sample, 10 to 15 SEM images were digitalized to quantify shape, size, and SEM-visible porosity. Also, X-ray Fluorescence (XRF) was conducted on every sample and converted to weight percent of quartz and feldspar, clay, and carbonate to study the heterogeneity and the influence of the inorganic matrix (IM) on pore structure and SEM-visible porosity. Pore types are categorized according to the shape and location. Inter-clay-particle (also referred to as phyllosilicate framework) pores and spongy pores are the most ubiquitous pore type in IM and OM respectively. Inter-clay-particle pores usually show triangular or elongate shapes. Their sizes are from tens to hundreds of nanometers. Spongy pores in OM are usually roundish, and show a relatively uniform size in a single OM particle. Their sizes range from ten (resolution-limitation) to several hundreds of nanometers. We defined OM degradation index (OMDI) to describe the extent of development of OM pores, which is porosity in organic matter divided by whole area OM covered and the voids in it. There is no systematic change in SEM-visible porosity as a function of the abundance of OM. However, there is a negative correlation between abundance of OM and OMDI. Samples with higher content of OM usually have large OM particles without pores, which can be explained by differences of maceral types. The ten nanometer-per-pixel resolution of the SEM could hinder recognition of smaller pores. The stratigraphic distribution of pore structure in the Mahantango, upper Marcellus, and lower Marcellus is strongly affected by heterogeneity of mineral composition. OMDI indicates a positive correlation between thermal maturity and development of OM pores in lower Marcellus, which hasn’t been found in other formations. OM pores have been considered as a secondary pore type that formed during post-depositional process primarily affected by thermal maturity. However, the derivation and supply of different maceral types exert a significant control on development of porosity in organic-rich Marcellus Shale.
Analyses of organic-rich mudstones from three wells that penetrated the Marcellus Shale in West Virginia and Pennsylvania were performed to evaluate the nature of pore structure (porosity and permeability) in both organic matter (OM) and inorganic matrix (IM). Samples include 32 core plugs with different levels of thermal maturity and TOC from Mahantango Formation and Marcellus Shale. For each sample, 10 to 15 SEM images were digitalized to quantify shape, size, and SEM-visible porosity. Also, X-ray Fluorescence (XRF) was conducted on every sample and converted to weight percent of quartz and feldspar, clay, and carbonate to study the heterogeneity and the influence of the inorganic matrix (IM) on pore structure and SEM-visible porosity. Pore types are categorized according to the shape and location. Inter-clay-particle (also referred to as phyllosilicate framework) pores and spongy pores are the most ubiquitous pore type in IM and OM respectively. Inter-clay-particle pores usually show triangular or elongate shapes. Their sizes are from tens to hundreds of nanometers. Spongy pores in OM are usually roundish, and show a relatively uniform size in a single OM particle. Their sizes range from ten (resolution-limitation) to several hundreds of nanometers. We defined OM degradation index (OMDI) to describe the extent of development of OM pores, which is porosity in organic matter divided by whole area OM covered and the voids in it. There is no systematic change in SEM-visible porosity as a function of the abundance of OM. However, there is a negative correlation between abundance of OM and OMDI. Samples with higher content of OM usually have large OM particles without pores, which can be explained by differences of maceral types. The ten nanometer-per-pixel resolution of the SEM could hinder recognition of smaller pores. The stratigraphic distribution of pore structure in the Mahantango, upper Marcellus, and lower Marcellus is strongly affected by heterogeneity of mineral composition. OMDI indicates a positive correlation between thermal maturity and development of OM pores in lower Marcellus, which hasn’t been found in other formations. OM pores have been considered as a secondary pore type that formed during post-depositional process primarily affected by thermal maturity. However, the derivation and supply of different maceral types exert a significant control on development of porosity in organic-rich Marcellus Shale.
Panel_14925
Panel_14925
8:30 AM
5:00 PM
8:30 a.m.
Porosity, Permeability, Pore Characterization and Rock Mechanics of the Triassic Cow Branch and Walnut Cove Formations: A Continuous Gas Assessment Unit, Dan River Basin, Stokes and Rockingham Counties, North Carolina, USA
Exhibition Hall
By J. C. Reid, K. B. Taylor, M. M. McGlue
The Late Triassic (Norian) Dan River basin is a continuous gas assessment unit (AU) and a total petroleum system. The source rocks (Walnut Cove and Cow Branch Formations) are thick grey and black freshwater shales; the stratigraphically lower Walnut Cove Fm. has a thin basal coal. These lacustrine strata were deposited in a rift basin near the paleo-equator after Pangea’s breakup. The AU has an estimated mean gas content of 49 BCFG, and a natural gas liquids content of 0 MMBNGL (USGS Fact Sheet 2012 - 3075) based on limited 2011 data. The Walnut Cove Fm. is up to 600 feet thick with an outcrop strike of ~22 miles and a width of several miles. The Cow Branch Fm., up to 1,500 feet thick, is exposed in broad patches in Stokes and Rockingham counties, NC, and isolated localities in southernmost Virginia. The hydrocarbon potential of these two formations as shale gas reservoirs was characterized from diamond drill core hole SO-C-2-81 (Walnut Cove Fm.), and the Cow Branch Fm. exposed continuously in the Ararat (aka Solite or Cemex) quarry. Characterization results reported herein include substantial new data not available for the 2011 USGS assessment. They are: 1) organic geochemistry and thermal maturation data (doubling the 2011 data set); 2) down hole XRD whole rock mineralogy for Walnut Cove Fm., n= 34, outcrop whole rock XRD for the Cow Branch Fm., n = 13; 3) rock mechanics including triaxial compressive strength tests with acoustic velocities, pre- and post CT scans, and Young’s Modulus and Poisson’s Ratio (Walnut Cove Fm., n = 7, Cow Branch Fm., n = 13); 4) mercury injection capillary pressure data obtained to characterize porosity and permeability in the both formations (Walnut Cove Fm., n = 14, Cow Branch Fm., n = 27) with maximum pressure of 60,000 psia providing a pore aperture frequency distribution down to nanometer-scale diameter; and 5) pore characterization using SEM and ion beam milled samples. This is the third report characterizing the continental Triassic rift / lacustrine deposits in NC. Previous reports on the time equivalent Cumnock Fm., Deep River basin, NC, are available on ‘Search and Discovery’. The Cumnock Fm. consists of black organic-rich mudrocks with significant porosity. Initial results from characterizing these formations show them as petroleum source rocks (wt% TOC > 2.0). Mineralogical composition varies among the three different mudrocks, which likely has implications for inter- and intra-particle porosity trends.
The Late Triassic (Norian) Dan River basin is a continuous gas assessment unit (AU) and a total petroleum system. The source rocks (Walnut Cove and Cow Branch Formations) are thick grey and black freshwater shales; the stratigraphically lower Walnut Cove Fm. has a thin basal coal. These lacustrine strata were deposited in a rift basin near the paleo-equator after Pangea’s breakup. The AU has an estimated mean gas content of 49 BCFG, and a natural gas liquids content of 0 MMBNGL (USGS Fact Sheet 2012 - 3075) based on limited 2011 data. The Walnut Cove Fm. is up to 600 feet thick with an outcrop strike of ~22 miles and a width of several miles. The Cow Branch Fm., up to 1,500 feet thick, is exposed in broad patches in Stokes and Rockingham counties, NC, and isolated localities in southernmost Virginia. The hydrocarbon potential of these two formations as shale gas reservoirs was characterized from diamond drill core hole SO-C-2-81 (Walnut Cove Fm.), and the Cow Branch Fm. exposed continuously in the Ararat (aka Solite or Cemex) quarry. Characterization results reported herein include substantial new data not available for the 2011 USGS assessment. They are: 1) organic geochemistry and thermal maturation data (doubling the 2011 data set); 2) down hole XRD whole rock mineralogy for Walnut Cove Fm., n= 34, outcrop whole rock XRD for the Cow Branch Fm., n = 13; 3) rock mechanics including triaxial compressive strength tests with acoustic velocities, pre- and post CT scans, and Young’s Modulus and Poisson’s Ratio (Walnut Cove Fm., n = 7, Cow Branch Fm., n = 13); 4) mercury injection capillary pressure data obtained to characterize porosity and permeability in the both formations (Walnut Cove Fm., n = 14, Cow Branch Fm., n = 27) with maximum pressure of 60,000 psia providing a pore aperture frequency distribution down to nanometer-scale diameter; and 5) pore characterization using SEM and ion beam milled samples. This is the third report characterizing the continental Triassic rift / lacustrine deposits in NC. Previous reports on the time equivalent Cumnock Fm., Deep River basin, NC, are available on ‘Search and Discovery’. The Cumnock Fm. consists of black organic-rich mudrocks with significant porosity. Initial results from characterizing these formations show them as petroleum source rocks (wt% TOC > 2.0). Mineralogical composition varies among the three different mudrocks, which likely has implications for inter- and intra-particle porosity trends.
Panel_14930
Panel_14930
8:30 AM
5:00 PM
8:30 a.m.
Gases Released From Organic-Rich Shales by Crush Analysis: Implications for Gas Generation, Desorption, Storage and Deliverability
Exhibition Hall
By T. Zhang, K. Milliken, S. C. Ruppel, X. Sun
The gases that fill nanoscale pores in organic-rich shales can be released during crushing. Our results show that both thermal maturity and gas desorption contribute to changes in the CH4/CO2 ratio of gases released during rock crushing. CH4/CO2 ratios of these gases are lower at low thermal maturities and higher at high thermal maturities because more CH4-rich gas is generated at higher maturity levels. However, no obvious compositional fractionation occurs among C1, C2, and C3 of crushed-rock gas, and C1/C2 and C2/C3 ratios remain nearly constant during crushing although these ratios are greatly increased overall when the level of thermal maturity is high. Gas geochemical parameters (C1/C2, C2/C3, and i-C4/n-C4) of released gas are good indicators of thermal maturation of organic-rich shales. Gas yields from a set of samples of varied TOC content but uniform thermal maturity, vary directly with TOC and porosity, suggesting greater pore connectivity for high TOC samples because of the development of organic matter-hosted pores critical to gas storage and deliverability. Trends in released gas yield and gas chemistry during rock crushing relate to gas storage states and pore connectivity. The d13C1, d13C2 and d13C3 values of gas released from particles of coarser size (> 250 µm) are similar to values of gas produced from Barnett shales after hydraulic fracturing. CH4-dominated gas appears to be stored in larger connected pores and is therefore released during the initial stages of crushing. The carbon-isotope values of C1, C2, and C3 are heavier in the more thermally mature samples, suggesting that this released gas is representative of the gas chemistry of the subsurface reservoir. Released gases provide a basis for understanding reservoir gas compositions and saturations even in older cores, bridging a gap caused by the scarcity of desorption data from fresh canister core. Chemical and isotopic composition of the released gases is an indicator of thermal maturity in subsurface reservoirs, and retained gas yield is directly related to variations of lithology and pore connectivity. This new method can provide useful information to evaluate gas storage and deliverability when old cores are available.
The gases that fill nanoscale pores in organic-rich shales can be released during crushing. Our results show that both thermal maturity and gas desorption contribute to changes in the CH4/CO2 ratio of gases released during rock crushing. CH4/CO2 ratios of these gases are lower at low thermal maturities and higher at high thermal maturities because more CH4-rich gas is generated at higher maturity levels. However, no obvious compositional fractionation occurs among C1, C2, and C3 of crushed-rock gas, and C1/C2 and C2/C3 ratios remain nearly constant during crushing although these ratios are greatly increased overall when the level of thermal maturity is high. Gas geochemical parameters (C1/C2, C2/C3, and i-C4/n-C4) of released gas are good indicators of thermal maturation of organic-rich shales. Gas yields from a set of samples of varied TOC content but uniform thermal maturity, vary directly with TOC and porosity, suggesting greater pore connectivity for high TOC samples because of the development of organic matter-hosted pores critical to gas storage and deliverability. Trends in released gas yield and gas chemistry during rock crushing relate to gas storage states and pore connectivity. The d13C1, d13C2 and d13C3 values of gas released from particles of coarser size (> 250 µm) are similar to values of gas produced from Barnett shales after hydraulic fracturing. CH4-dominated gas appears to be stored in larger connected pores and is therefore released during the initial stages of crushing. The carbon-isotope values of C1, C2, and C3 are heavier in the more thermally mature samples, suggesting that this released gas is representative of the gas chemistry of the subsurface reservoir. Released gases provide a basis for understanding reservoir gas compositions and saturations even in older cores, bridging a gap caused by the scarcity of desorption data from fresh canister core. Chemical and isotopic composition of the released gases is an indicator of thermal maturity in subsurface reservoirs, and retained gas yield is directly related to variations of lithology and pore connectivity. This new method can provide useful information to evaluate gas storage and deliverability when old cores are available.
Panel_14929
Panel_14929
8:30 AM
5:00 PM
8:30 a.m.
Application of Computed Tomography and Scanning Electron Microscopy in the Determination of Pore-Space Characterization of a Potential Unconventional Reservoir: A Case Study in Mowry Shale in the Powder River Basin, WY
Exhibition Hall
By E. C. Melville, Z. O. Baran, L. Stetler
The Cretaceous Mowry Shale in the Powder River Basin (PRB), WY is a well-known hydrocarbon source for overlying Frontier and underlying Muddy sandstones. Recent studies have focused on the hydrocarbon potential of the Mowry formation as an unconventional reservoir. It consists of primarily siliceous black shale which can be subdivided into three intervals: a fissile, weakly cemented, bioturbated shale at the bottom; highly siliceous, thinly laminated, organic-matter-enriched shale in the middle; and a coarsening upward, bioturbated silty-shale in the upper section. Twelve primary bentonite beds are distributed throughout the Mowry section and can be correlated across the PRB. Investigation of cores, drill stem test, production data, rock evaluation and sequence stratigraphy indicates two ideal drill intervals within the middle Mowry section. The lateral continuity of these two zones and the correlation between bentonites provide further work on the facies changes and deformational structures across the PRB to find ideal drill locations. In this study, we utilize computed tomography (CT) and scanning electron microscope (SEM) to better understand the heterogeneity of pore space and fracture networks at varying depths in the PRB. CT and SEM analysis of available core samples within the potential drilling intervals provide critical information about micron-to-nanometer scale characteristics of the Mowry shale. A detailed fracture analysis yields a structural model to explain the potential tectonic effects on pore space evolution. Sampling across facies changes within the drilling interval will help develop an understanding of the effect that variations in depositional minerals or material have on pore space. The results of this study will show changes in granular and organic-matter-hosted pore space, pore throat size and effective permeability due to both structural deformation and variations in sedimentation. The major contribution of this study is to provide a better understanding for successful drilling practices in unconventional black shale and clay rich hydrocarbon resources.
The Cretaceous Mowry Shale in the Powder River Basin (PRB), WY is a well-known hydrocarbon source for overlying Frontier and underlying Muddy sandstones. Recent studies have focused on the hydrocarbon potential of the Mowry formation as an unconventional reservoir. It consists of primarily siliceous black shale which can be subdivided into three intervals: a fissile, weakly cemented, bioturbated shale at the bottom; highly siliceous, thinly laminated, organic-matter-enriched shale in the middle; and a coarsening upward, bioturbated silty-shale in the upper section. Twelve primary bentonite beds are distributed throughout the Mowry section and can be correlated across the PRB. Investigation of cores, drill stem test, production data, rock evaluation and sequence stratigraphy indicates two ideal drill intervals within the middle Mowry section. The lateral continuity of these two zones and the correlation between bentonites provide further work on the facies changes and deformational structures across the PRB to find ideal drill locations. In this study, we utilize computed tomography (CT) and scanning electron microscope (SEM) to better understand the heterogeneity of pore space and fracture networks at varying depths in the PRB. CT and SEM analysis of available core samples within the potential drilling intervals provide critical information about micron-to-nanometer scale characteristics of the Mowry shale. A detailed fracture analysis yields a structural model to explain the potential tectonic effects on pore space evolution. Sampling across facies changes within the drilling interval will help develop an understanding of the effect that variations in depositional minerals or material have on pore space. The results of this study will show changes in granular and organic-matter-hosted pore space, pore throat size and effective permeability due to both structural deformation and variations in sedimentation. The major contribution of this study is to provide a better understanding for successful drilling practices in unconventional black shale and clay rich hydrocarbon resources.
Panel_14921
Panel_14921
8:30 AM
5:00 PM
8:30 a.m.
Pore Networks in Niobrara, Piceance Basin, Western Colorado Exhibit Minimal Regional Variability as a Function of Thermal History
Exhibition Hall
By C. C. Burt, D. A. Budd
The Niobrara member of the Mancos Shale is an unconventional gas reservoir in the Piceance basin, western Colorado. In general, burial and thermal histories were less on the western side of the basin and greater in the northeastern side of the basin. As a result, vitrinite reflectance values for the full Niobrara interval range from ~0.5 Ro to ~1.2 Ro. This study investigates how nano- and micron-scale pore systems differ as a result of this variability in thermal histories. Core material was sampled from 6 wells that cover a 6,750 ft range in subsea burial depths and Ro values of 0.5 to 0.9. To control for lithologic heterogeneity, all samples were taken from the same ~50 ft-thick stratigraphic interval that comprises the primary horizontal landing zone across the basin. Lithologies in that zone are marls to marly shales with reasonably high (up to ~55%) carbonate content. The primary constituents are quartz silt, peloids, and argillaceous clays. Elemental mapping shows that peloids are almost exclusively calcite and contain disseminated carbon. Detrital carbon (kerogen) and siliciclastics are in the matrix. Pore networks were identified, characterized, and analyzed using focused ion-beam scanning electron microscopy (FIB-SEM) and Avizo Firetm image analysis software. A variety of pore types were observed, with the dominant forms being intercrystalline pores between recrystallized calcite in the peloids and clay-related plus interparticle pores in the intervening argillaceous matrix. At one point in time the average amount of porosity in peloids (13%) was 6 times greater than the average amount in matrix (2.2%). But many of those pores (70% in peloids, 60% in matrix) are now filled with residual migrated hydrocarbons that contain organic matter bubble pores (~7% of all residual hydrocarbon is now bubble pore). As a result, total imaged porosity now averages only 3.5% in peloids and 0.9% in matrix with OM pores comprising 10% to 20% of those values. Cumulative area and size distributions (equivalent circular diameters average 257.6 nm in peloids and 266.2 nm in matrix) for peloids and matrix pore networks show no trend related to burial depth or thermal history. The lowest maturity well (~0.5 % Ro) shows a much lower abundance of organic matter porosity relative to all other wells, but there is no trend in organic matter porosity above a Ro of ~0.7.
The Niobrara member of the Mancos Shale is an unconventional gas reservoir in the Piceance basin, western Colorado. In general, burial and thermal histories were less on the western side of the basin and greater in the northeastern side of the basin. As a result, vitrinite reflectance values for the full Niobrara interval range from ~0.5 Ro to ~1.2 Ro. This study investigates how nano- and micron-scale pore systems differ as a result of this variability in thermal histories. Core material was sampled from 6 wells that cover a 6,750 ft range in subsea burial depths and Ro values of 0.5 to 0.9. To control for lithologic heterogeneity, all samples were taken from the same ~50 ft-thick stratigraphic interval that comprises the primary horizontal landing zone across the basin. Lithologies in that zone are marls to marly shales with reasonably high (up to ~55%) carbonate content. The primary constituents are quartz silt, peloids, and argillaceous clays. Elemental mapping shows that peloids are almost exclusively calcite and contain disseminated carbon. Detrital carbon (kerogen) and siliciclastics are in the matrix. Pore networks were identified, characterized, and analyzed using focused ion-beam scanning electron microscopy (FIB-SEM) and Avizo Firetm image analysis software. A variety of pore types were observed, with the dominant forms being intercrystalline pores between recrystallized calcite in the peloids and clay-related plus interparticle pores in the intervening argillaceous matrix. At one point in time the average amount of porosity in peloids (13%) was 6 times greater than the average amount in matrix (2.2%). But many of those pores (70% in peloids, 60% in matrix) are now filled with residual migrated hydrocarbons that contain organic matter bubble pores (~7% of all residual hydrocarbon is now bubble pore). As a result, total imaged porosity now averages only 3.5% in peloids and 0.9% in matrix with OM pores comprising 10% to 20% of those values. Cumulative area and size distributions (equivalent circular diameters average 257.6 nm in peloids and 266.2 nm in matrix) for peloids and matrix pore networks show no trend related to burial depth or thermal history. The lowest maturity well (~0.5 % Ro) shows a much lower abundance of organic matter porosity relative to all other wells, but there is no trend in organic matter porosity above a Ro of ~0.7.
Panel_14917
Panel_14917
8:30 AM
5:00 PM
8:30 a.m.
Completely-Cemented Natural Fractures in Mudrocks: Flow Barrier or Highway?
Exhibition Hall
By C. J. Landry, P. Eichhubl, M. Prodanovic, A. Tokan-Lawal
It has been postulated that hydraulic fractures reactivate natural fracture networks, resulting in greater access to the host rock and increased rates of production. However, many of these natural fractures are completely-cemented, and currently there is very little evidence that completely-cemented natural fractures are anything but impermeable, and thus would block production. Among natural fractures observed in core of Eagle Ford Shale, Texas, tall sub-vertical calcite-cemented fractures were the focus of this investigation. Similar sub-vertical, completely-cemented, opening-mode fractures are very common in mudrocks. We used SEM imaging on broad-ion-beam (BIB)-milled samples of a calcite-cemented fracture to study the microstructure of the calcite for any indication that completely cemented fractures are permeable. In the fracture calcite cement, we observed open flow-paths between calcite grains that are generally well-connected with an average aperture between ~25 and ~100 nm. The permeability of these flow-paths was determined by lattice Boltzmann methods to be between 60 to 660 µD. These flow-paths have a spacing between 200 and 400 µm; therefore, a square centimeter (length*height) of fracture cement will contain on average more than 500 flow-paths. Using simple effective medium upscaling the fracture cement studied here is found to have a permeability in the range of 30 to 630 nD. Although this is a very low permeability, it is within the range of the permeability of typical mudrocks; therefore, these calcite cements would have almost no effect on flow orthogonal to the plane of the fracture. These flow-paths are also connected within the calcite cement creating a flow-path network along the fracture. Although the flow-path network has a bulk permeability close to that of typical mudrocks, due to the much lower porosity of the calcite cement (< 0.05%) in comparison to the host rock (2% < typical mudrock porosity < 15%) the actual velocities in the flow-path network are much greater than the host rock. During production, the significantly greater actual velocities of flow in the cemented natural fracture results in the lower pressure of the wellbore-hydraulic-fracture being quickly translated into the natural fracture. This will effectively increase the hydraulic fracture/host rock interfacial area, and production rates. Therefore, completely-cemented natural fractures in mudrocks can act as flow highways.
It has been postulated that hydraulic fractures reactivate natural fracture networks, resulting in greater access to the host rock and increased rates of production. However, many of these natural fractures are completely-cemented, and currently there is very little evidence that completely-cemented natural fractures are anything but impermeable, and thus would block production. Among natural fractures observed in core of Eagle Ford Shale, Texas, tall sub-vertical calcite-cemented fractures were the focus of this investigation. Similar sub-vertical, completely-cemented, opening-mode fractures are very common in mudrocks. We used SEM imaging on broad-ion-beam (BIB)-milled samples of a calcite-cemented fracture to study the microstructure of the calcite for any indication that completely cemented fractures are permeable. In the fracture calcite cement, we observed open flow-paths between calcite grains that are generally well-connected with an average aperture between ~25 and ~100 nm. The permeability of these flow-paths was determined by lattice Boltzmann methods to be between 60 to 660 µD. These flow-paths have a spacing between 200 and 400 µm; therefore, a square centimeter (length*height) of fracture cement will contain on average more than 500 flow-paths. Using simple effective medium upscaling the fracture cement studied here is found to have a permeability in the range of 30 to 630 nD. Although this is a very low permeability, it is within the range of the permeability of typical mudrocks; therefore, these calcite cements would have almost no effect on flow orthogonal to the plane of the fracture. These flow-paths are also connected within the calcite cement creating a flow-path network along the fracture. Although the flow-path network has a bulk permeability close to that of typical mudrocks, due to the much lower porosity of the calcite cement (< 0.05%) in comparison to the host rock (2% < typical mudrock porosity < 15%) the actual velocities in the flow-path network are much greater than the host rock. During production, the significantly greater actual velocities of flow in the cemented natural fracture results in the lower pressure of the wellbore-hydraulic-fracture being quickly translated into the natural fracture. This will effectively increase the hydraulic fracture/host rock interfacial area, and production rates. Therefore, completely-cemented natural fractures in mudrocks can act as flow highways.
Panel_14924
Panel_14924
8:30 AM
5:00 PM
8:30 a.m.
Investigating Controls on the Transport Properties of Mudstone: Implications for Shale-Gas Production
Exhibition Hall
By R. McKernan, J. Mecklenburgh, E. H. Rutter, K. G. Taylor
Knowledge of mudstone permeability and sensitivity to stress is required to enhance interpretation of well logs and subsurface variability. This project provides an integrated study of both lab-measured permeability and mechanical properties, and sedimentological, diagenetic and burial histories of mudstones. Samples tested include a Jurassic mudstone (Whitby Mudstone Formation), a clay-bearing, silt-rich mudstone with 6-9% porosity, 1.5% TOC and an anisotropic texture. Samples from the carbonate-rich Eagle Ford Formation and clay-rich Marcellus Formation are also being tested to examine the relationship between permeability and porosity and the microstructural arrangement and elasticity of the component mineral phases in these different mudstones. Permeability was measured as a function of effective pressure (Peff ) for flow of argon across 25 mm diameter cylindrical samples, using the oscillating pore pressure method. Alongside experiments, petrographical characterization of samples is used to explore the geological controls on fluid transport properties of mudstones. A large number of experiments have been performed that demonstrate the effect of pressure cycling on mudstone permeability and as a result the intrinsic sensitivity of permeability to variations in effective stress. Further, the validity of the effective pressure concept as defined by Terzaghi (1923) as simply the difference between total confining pressure and pore pressure has also been investigated by comparing variations of pore pressure under constant confining pressure with variations of confining pressure under constant pore pressure. Crucial to gas reservoir evaluation, results show that after reconditioning, variation of permeability with Peff is reproducible, and can be used to model the reduction in gas flow to be expected as a result of the increased Peff that results from in-situ pore pressure decay during gas extraction. Additionally, the influence of (hydrostatic) effective confining pressure on permeability, defined by the traditional effective stress law, must be modified by an additional term that depends only on pore pressure, of the form log k = A + Peff + b Pp where k is permeability, Peff is effective pressure, Pp is pore pressure and A and b are parameters. Such an approach is essential to the realistic application of laboratory-determined permeability data to gas reservoir evaluation.
Knowledge of mudstone permeability and sensitivity to stress is required to enhance interpretation of well logs and subsurface variability. This project provides an integrated study of both lab-measured permeability and mechanical properties, and sedimentological, diagenetic and burial histories of mudstones. Samples tested include a Jurassic mudstone (Whitby Mudstone Formation), a clay-bearing, silt-rich mudstone with 6-9% porosity, 1.5% TOC and an anisotropic texture. Samples from the carbonate-rich Eagle Ford Formation and clay-rich Marcellus Formation are also being tested to examine the relationship between permeability and porosity and the microstructural arrangement and elasticity of the component mineral phases in these different mudstones. Permeability was measured as a function of effective pressure (Peff ) for flow of argon across 25 mm diameter cylindrical samples, using the oscillating pore pressure method. Alongside experiments, petrographical characterization of samples is used to explore the geological controls on fluid transport properties of mudstones. A large number of experiments have been performed that demonstrate the effect of pressure cycling on mudstone permeability and as a result the intrinsic sensitivity of permeability to variations in effective stress. Further, the validity of the effective pressure concept as defined by Terzaghi (1923) as simply the difference between total confining pressure and pore pressure has also been investigated by comparing variations of pore pressure under constant confining pressure with variations of confining pressure under constant pore pressure. Crucial to gas reservoir evaluation, results show that after reconditioning, variation of permeability with Peff is reproducible, and can be used to model the reduction in gas flow to be expected as a result of the increased Peff that results from in-situ pore pressure decay during gas extraction. Additionally, the influence of (hydrostatic) effective confining pressure on permeability, defined by the traditional effective stress law, must be modified by an additional term that depends only on pore pressure, of the form log k = A + Peff + b Pp where k is permeability, Peff is effective pressure, Pp is pore pressure and A and b are parameters. Such an approach is essential to the realistic application of laboratory-determined permeability data to gas reservoir evaluation.
Panel_14922
Panel_14922
8:30 AM
5:00 PM
8:30 a.m.
Guided Sampling of Pore-Scale Imaging for Heterogeneous Mudrocks
Exhibition Hall
By D. Unrau, K. G. Lagarec, G. Lesniak, P. Such, M. Dohnalik, R. Cicha, M. Mroczkowska, M. Marsh*, S. Bhattiprolu, A. Steinbach, M. Phaneuf
High-resolution microscopy enables the investigation of petrophysical properties of reservoir rocks at the pore scale, even in the case of mudrocks with pores as small as 1-10 nm across. However, tying these nanometer scale findings to longer length scales is challenging. Because of the heterogeneity of micro- and nano-scale rock properties for mudrocks, uninformed sampling can give results that are not representative at longer scales. A reasonable workflow strategy is to image samples at a course scale and then perform higher resolution imaging of representative zones. This has been proposed, and executed in limited cases, but doing so requires painstaking effort to track sample orientation and position in order to locate and image the targeted features manually. We have developed and demonstrate here a software platform that simplifies this critical task. Chiefly, the software platform is designed to use image data from any prior experiment to inform and guide targeted imaging for a subsequent experiment. We have identified and then implemented four key components to streamline this workflow. First, the system must manipulate and display disparate image data, which will likely come from a variety of microscopy techniques. Second, the platform must enable the user to properly register disparate image data to unify them in one common frame of reference. Third, the platform must be stage position-aware and capable of driving the stage to user-targeted areas of interest. Finally, the platform should present the high-resolution imagery in the context of the coarser data. Some of these requirements have been addressed individually, but they have never been integrated such that the coarse data can guide the high-resolution data collection. Prior to this work, an existing software platform could visualize various image types, and furthermore, it already had stage awareness. We demonstrate the addition of a highly flexible image registration tool that permits registration of 2D images, allowing translation, rotation, scaling, and shearing. We extended this to 3D with a common plane identification tool; once that plane is determined, the remaining alignment is carried out as a straightforward 2D registration problem. We demonstrate here this platform on various mudrock samples. We show integration of multiple imaging techniques for mudrocks spanning many orders of length scale (from mm to nm) and incorporating both 2D and 3D imaging and analysis.
High-resolution microscopy enables the investigation of petrophysical properties of reservoir rocks at the pore scale, even in the case of mudrocks with pores as small as 1-10 nm across. However, tying these nanometer scale findings to longer length scales is challenging. Because of the heterogeneity of micro- and nano-scale rock properties for mudrocks, uninformed sampling can give results that are not representative at longer scales. A reasonable workflow strategy is to image samples at a course scale and then perform higher resolution imaging of representative zones. This has been proposed, and executed in limited cases, but doing so requires painstaking effort to track sample orientation and position in order to locate and image the targeted features manually. We have developed and demonstrate here a software platform that simplifies this critical task. Chiefly, the software platform is designed to use image data from any prior experiment to inform and guide targeted imaging for a subsequent experiment. We have identified and then implemented four key components to streamline this workflow. First, the system must manipulate and display disparate image data, which will likely come from a variety of microscopy techniques. Second, the platform must enable the user to properly register disparate image data to unify them in one common frame of reference. Third, the platform must be stage position-aware and capable of driving the stage to user-targeted areas of interest. Finally, the platform should present the high-resolution imagery in the context of the coarser data. Some of these requirements have been addressed individually, but they have never been integrated such that the coarse data can guide the high-resolution data collection. Prior to this work, an existing software platform could visualize various image types, and furthermore, it already had stage awareness. We demonstrate the addition of a highly flexible image registration tool that permits registration of 2D images, allowing translation, rotation, scaling, and shearing. We extended this to 3D with a common plane identification tool; once that plane is determined, the remaining alignment is carried out as a straightforward 2D registration problem. We demonstrate here this platform on various mudrock samples. We show integration of multiple imaging techniques for mudrocks spanning many orders of length scale (from mm to nm) and incorporating both 2D and 3D imaging and analysis.
Panel_14927
Panel_14927
8:30 AM
5:00 PM
8:30 a.m.
Interpreting Permeability From Mercury Injection Capillary Pressure Data
Exhibition Hall
By A. A. Brown
Laminar flow theory predicts a strong correlation between permeability and pore-throat distribution as revealed by Mercury Injection Capillary Pressure (MICP) data. Previous studies have developed relationships between MICP data and permeability; however, the permeabilities predicted by different methods can differ substantially from the measured permeabilities and from each other, especially in low permeability samples of interest for unconventional reservoirs. The purposes of this study are to evaluate why there is such large scatter, identify algorithms that best predict permeability over a wide range of permeabilities, and evaluate what type of permeability is actually measured by MICP data. Precision of permeability predictions is low due to insufficient MICP pressure measurements, assumption of MICP curve shape, permeability anisotropy of geological samples, and low precision and accuracy of permeability measurement of tight rocks. Four methods for estimating permeability from MICP data are found to have small bias and reasonable precision over a wide range of permeability: the modified Purcell, the Katz-Thompson Lc, Katz-Thompson Lh, and the Swanson methods. A weighted average of these permeability estimates corrects for accuracy problems and increases permeability estimate precision. However, this MICP-predicted average permeability still varies from measured Klinkenberg-corrected steady permeability by an average of a factor of 2. This mismatch may be more apparent than real. Restoring reservoir stress prior to conventional permeability measurement fails to remove completely the core damage caused by microfractures created during extraction, preparation, and storage of tight rock samples from deep boreholes. MICP permeabilities are estimated from the pore-throat distributions, which do not include the significant flow contributions from microfractures. Difference between MICP permeability estimates and measured permeability of tight samples may be caused by the inability of conventional permeability analysis to remove damage effects by stress restoration. If so, MICP permeability estimates are as good as or better than permeability measured from tight, subsurface samples. MICP permeability is either the ambient matrix permeability or a stressed matrix permeability, depending on the relative magnitude of in situ reservoir stress and Hg pressure at threshold saturation.
Laminar flow theory predicts a strong correlation between permeability and pore-throat distribution as revealed by Mercury Injection Capillary Pressure (MICP) data. Previous studies have developed relationships between MICP data and permeability; however, the permeabilities predicted by different methods can differ substantially from the measured permeabilities and from each other, especially in low permeability samples of interest for unconventional reservoirs. The purposes of this study are to evaluate why there is such large scatter, identify algorithms that best predict permeability over a wide range of permeabilities, and evaluate what type of permeability is actually measured by MICP data. Precision of permeability predictions is low due to insufficient MICP pressure measurements, assumption of MICP curve shape, permeability anisotropy of geological samples, and low precision and accuracy of permeability measurement of tight rocks. Four methods for estimating permeability from MICP data are found to have small bias and reasonable precision over a wide range of permeability: the modified Purcell, the Katz-Thompson Lc, Katz-Thompson Lh, and the Swanson methods. A weighted average of these permeability estimates corrects for accuracy problems and increases permeability estimate precision. However, this MICP-predicted average permeability still varies from measured Klinkenberg-corrected steady permeability by an average of a factor of 2. This mismatch may be more apparent than real. Restoring reservoir stress prior to conventional permeability measurement fails to remove completely the core damage caused by microfractures created during extraction, preparation, and storage of tight rock samples from deep boreholes. MICP permeabilities are estimated from the pore-throat distributions, which do not include the significant flow contributions from microfractures. Difference between MICP permeability estimates and measured permeability of tight samples may be caused by the inability of conventional permeability analysis to remove damage effects by stress restoration. If so, MICP permeability estimates are as good as or better than permeability measured from tight, subsurface samples. MICP permeability is either the ambient matrix permeability or a stressed matrix permeability, depending on the relative magnitude of in situ reservoir stress and Hg pressure at threshold saturation.
Panel_14931
Panel_14931
8:30 AM
5:00 PM
8:30 a.m.
Full-Physics Thinking in Unconventional Plays
Exhibition Hall
By G. D. Couples
Unconventionals, perhaps more than other plays, demand consideration of process interactions. Geomechanical interactions occupy a central role in Unconventionals: geohistory, and the mechanical processes that operate, creates pre-cursor conditions; manufacturing the reservoir is a dominantly mechanical activity; and during reservoir production, mechanical interactions play a governing role. Classical methods of geomechanical interpretation and analysis fail to address the physics interactions, and can lead to incorrect deductions and decisions. These difficulties arise because the classical approaches assume that rock stress is an independent parameter and can be assigned a value. That view is physically impossible. The key point is that the concept of stress can be expressed in multiple ways –the most important one is that stress is the specific (mass/volume-related) elastic energy. Using this “take” on stress, we examine some important aspects of Unconventional reservoirs, focusing on hydrofracture stimulation. We assess some notions that inhibit understanding and interfere with the discovery of better practices. The hydrofracture process involves injecting a medium (usually water-based) into perforations, aiming to create new openings in the rock mass that will allow better hydrocarbon flow. The injected fluid pressure (an energy measure) and volume define the energy input. Some energy is consumed in making new discontinuities, and in shifting rocks. Where is the rest? As discontinuities open, the adjacent rocks become strained, typically in ways that lead to local contractions and volume loss, so their stress (elastic energy) state increases. In poro-elastic terms, the pre-existing pore fluids gain some of this added energy, so have higher pressures. Calculations show that injected fluids do not invade the pore system of the matrix rocks, and therefore, those not yet recovered in flowback must be located in newly-created (or enhanced) openings – typically fracture-like features. After one hydraulic fracture stage, the subsurface state is considerably altered, with impacts on subsequent stages. After multiple stages, the state is characterised by high energy levels that work against the maintenance of the permeability created by the stimulation activities. The full-physics interactions, expressed in terms of energy components and partitioning, lead to new insights, and provide a framework within which new operational practices can be contemplated.
Unconventionals, perhaps more than other plays, demand consideration of process interactions. Geomechanical interactions occupy a central role in Unconventionals: geohistory, and the mechanical processes that operate, creates pre-cursor conditions; manufacturing the reservoir is a dominantly mechanical activity; and during reservoir production, mechanical interactions play a governing role. Classical methods of geomechanical interpretation and analysis fail to address the physics interactions, and can lead to incorrect deductions and decisions. These difficulties arise because the classical approaches assume that rock stress is an independent parameter and can be assigned a value. That view is physically impossible. The key point is that the concept of stress can be expressed in multiple ways –the most important one is that stress is the specific (mass/volume-related) elastic energy. Using this “take” on stress, we examine some important aspects of Unconventional reservoirs, focusing on hydrofracture stimulation. We assess some notions that inhibit understanding and interfere with the discovery of better practices. The hydrofracture process involves injecting a medium (usually water-based) into perforations, aiming to create new openings in the rock mass that will allow better hydrocarbon flow. The injected fluid pressure (an energy measure) and volume define the energy input. Some energy is consumed in making new discontinuities, and in shifting rocks. Where is the rest? As discontinuities open, the adjacent rocks become strained, typically in ways that lead to local contractions and volume loss, so their stress (elastic energy) state increases. In poro-elastic terms, the pre-existing pore fluids gain some of this added energy, so have higher pressures. Calculations show that injected fluids do not invade the pore system of the matrix rocks, and therefore, those not yet recovered in flowback must be located in newly-created (or enhanced) openings – typically fracture-like features. After one hydraulic fracture stage, the subsurface state is considerably altered, with impacts on subsequent stages. After multiple stages, the state is characterised by high energy levels that work against the maintenance of the permeability created by the stimulation activities. The full-physics interactions, expressed in terms of energy components and partitioning, lead to new insights, and provide a framework within which new operational practices can be contemplated.
Panel_14926
Panel_14926
8:30 AM
5:00 PM