11: Pore-Scale Imaging of Solid Bitumens: Insights for Unconventional Reservoir Characterization
Published:January 31, 2020
- PDF LinkChapter PDF
Tim E. Ruble, Kultaransingh (Bobby) Hooghan, William Dorsey, Wayne R. Knowles, Christopher D. Laughrey, 2020. "Pore-Scale Imaging of Solid Bitumens: Insights for Unconventional Reservoir Characterization", Mudstone Diagenesis: Research Perspectives for Shale Hydrocarbon Reservoirs, Seals, and Source Rocks, Wayne K. Camp, Kitty L. Milliken, Kevin Taylor, Neil Fishman, Paul C. Hackley, Joe H. S. Macquaker
Download citation file:
Characterizing unconventional reservoirs involves the investigation of a wide range of potential source-rock targets at various stages of thermal maturity. These samples may contain a mixture of kerogen, bitumen, oil, and pyrobitumen within their fabric. Thus, it is critical that we properly identify and examine each organic phase to better understand reservoir properties. In the present study, we have selected samples of gilsonite from a naturally occurring solid hydrocarbon deposit to serve as an analog for characterizing the bitumen phase of generation.
Gilsonite is an aromatic-asphaltic solid bitumen found in vertical veins along the eastern portion of the Uinta Basin, Utah. It is thought to be an early generation product from oil-prone Green River Shale source beds and is similar to low-maturity crude oil in composition. It has a high nitrogen content, low sulfur content, and high melting point (fusibility) and is soluble in organic solvents. We have used a variety of analytic methods to characterize this material, including standard optical organic petrology and scanning electron microscopic imaging to examine the occurrence of organic porosity.
Optical organic petrology analyses using both air and oil immersion objectives show that the polished gilsonite surfaces are typically dark grey and featureless. Optical evidence for the presence of macerals and inorganic constituents is absent. Visual estimates suggest that fractures make up approximately 1% of the conchoidal fracture plane, whereas the pencillated variety contains approximately 2% fractures along with 5% shallow pits. Scanning electron microscopic images also show the occurrence of fractures within gilsonite, but the matrix contains no evident organic porosity.
The results of our analyses suggest that, unlike pyrobitumen, pre-oil solid bitumen represented by gilsonite was found to contain no significant occurrences of organic nanoporosity within its matrix. Gilsonite does have minor pitting and fractures, but these do not represent an effective interconnected pore network and are probably artifacts of weathering/sampling. Thus, this material would not represent a potential candidate for in-situ petroleum storage capacity. Whether this is typical of all naturally occurring solid bitumen is debatable, considering that gilsonite has undergone some secondary alteration via devolatilization and limited biodegradation. Nevertheless, the pore-scale imaging of this solid bitumen provides potentially important new insights for unconventional reservoir characterization.
Solid bitumens have been reported in oil fields worldwide and their occurrence can have a detrimental effect on production (Lomando, 1992). In most conventional reservoirs, solid bitumens tend to reduce porosity and permeability by filling or lining pore spaces and by restricting and closing pore throats (Walters et al., 2006). In unconventional reservoirs, similar processes can occur, but in many instances solid bitumens may actually contribute to the in-situ organic porosity and provide storage for generated petroleum (Loucks et al., 2009; Laughrey et al., 2010; Han et al., 2017). Organic pore development is believed to be largely caused by the thermal cracking of kerogen (Jarvie et al., 2007; Loucks et al., 2009) and/or bitumen (Bernard et al., 2012), although primary organic pores have also been observed within immature organic matter as well (Löhr et al., 2015; Pommer and Milliken, 2015). The ability to predict the porosity evolution of shales as a function of thermal maturation is critical for successful unconventional resource development, yet a comprehensive literature review reveals that the formation mechanisms of organic pores are only poorly understood (Hackley and Cardott, 2016).
Solid bitumens have been classified on the basis of their physical properties using a generic naming scheme (Figure 1). In such classifications, two major groups have been defined on the basis of solubility in organic solvents, with “bitumens” being soluble in common organic solvents and “pyrobitumens” being insoluble (Abraham, 1945; Hunt, 1979). Other more recent classification schemes (Figure 2) recognize that solid bitumens can form from a variety of processes, and thus a genetic scheme was developed that incorporates these processes (Curiale, 1986). In this scheme, solid bitumens are assigned a position on a continuum between pre-oil and post-oil maturity levels (Curiale, 1986). Pre-oil solid bitumens are early generation products of rich source rocks, probably extruded as very viscous fluids that migrated the minimum distance necessary to reach fractures and voids (Curiale, 1986). Various geochemical parameters were used to determine the classification of pre-oil or post-oil of a select group of solid bitumens. The bitumens from the Uinta Basin (which included gilsonite) were considered to be immature, pre-oil type bitumens (Curiale, 1986; Ruble, 1990). It is important to note that all of these solid bitumen classification schemes were developed based upon the examination of samples found away from their sources as large veins in surrounding rock and typically present in mineable quantities. These “macro-solid bitumens” are not identical to the “micro-solid bitumens” commonly recognized in unconventional reservoirs; however, in some circumstances, they may be suitable analogs for evaluating the occurrence of organic porosity.
Lewan (1997) described the processes thought to occur during petroleum generation in both nature and hydrous pyrolysis laboratory simulation experiments (Figure 3). These processes happen in three stages: (1) bitumen generation, (2) immiscible oil generation, and (3) pyrobitumen/gas generation. Kerogen decomposition to bitumen involves cleavage of weak noncovalent bonds at low maturity, which produces a substance enriched in high molecular-weight hydrocarbons and heteroatom components (Lewan, 1993). At higher maturity, this pre-oil bitumen that is saturated with dissolved water undergoes partial decomposition, which produces aliphatic moieties that form an immiscible oil phase (Lewan, 1994). This immiscible oil phase is expelled from the source rock during primary migration and is collected from the surface of the water at the completion of each hydrous pyrolysis experiment. In natural systems, this expelled oil travels out of the source rock into adjacent carrier beds and moves upward to more shallow depths because of buoyancy effects. At higher maturity, the expelled oil and retained bitumen eventually decomposes into gas and pyrobitumen (Lewan, 1993). Although most geochemists have adopted this model of petroleum generation, others have criticized it for its use of archaic species such as “bitumen” and “oil” that confuse the role of chemical kinetics and phase equilibria in the observed phenomena (Burnham et al., 2016). Specifically, the use of expelled oil and extracted bitumen as chemical species in a kinetic reaction network is deemed inappropriate because the distinction is heavily influenced by phase-separation transport processes that differ considerably from experimental to natural conditions (Burnham and Braun, 1990; Burnham, 2017).
Experimentally derived hydrous pyrolysis products have been compared to naturally occurring petroleum products from the Uinta Basin, which include crude oils and various types of semisolid and solid native bitumens (Ruble et al., 1999, 2001). Low-maturity aromatic-intermediate and aromatic-asphaltic oils associated with semisolid gilsonite tars (Figure 4) are depleted in saturates compared to the immiscible oils generated by hydrous pyrolysis (Ruble et al., 2001). The fractional compositions of these samples more closely correspond to hydrous pyrolysis bitumens than to immiscible oils, and Ruble et al. (2001) concluded that these samples are not true oils because they do not represent a natural phase separation between bitumen and oil. Instead, samples such as these are thought to represent extruded pre-oil bitumens, which arise as a consequence of bitumen generation in extremely organic-rich source rocks (Pepper, 2017), where bitumen formation exceeds the capacity of the rock to retain this organic phase before immiscible oil formation. Such a process may explain early bitumen extrusion, previously reported in some organic-rich lacustrine sequences (Qiang et al., 1997; Katz and Xingcai, 1998; Qiang and McCabe, 1998). Early bitumen generation is also likely to be responsible for some of the unusual native bitumen deposits in the Uinta Basin (Hunt, 1963; Ruble, 1990). Aromatic-asphaltic bitumens, such as gilsonite, show greater proportions of polar constituents compared to hydrous pyrolysis bitumens (Ruble et al., 2001). However, this is most likely a consequence of secondary alteration effects associated with devolatilization and possibly limited biodegradation (Ruble, 1990, 1996; Ruble and Philp, 1998).
Organic matter-hosted pores, rather than mineral-hosted pores, are considered to be the dominant contributors to total porosity and petroleum storage in many organic-rich unconventional reservoirs (Loucks et al., 2012; Milliken et al., 2013; Löhr et al., 2015). The common working model of organic pore genesis holds that the formation of organic pores is largely a function of thermal maturity and infers that porosity increases with thermal maturity (Loucks et al., 2012). Organic pores are thought to develop during thermal maturation through the oil window as generated petroleum products are expelled from the kerogen, which leaves behind interconnected nanopores. Organic pores can also develop as secondary features because of the thermal cracking of residual oil into gas and pyrobitumen at elevated thermal maturities within the gas window. However, the occurrence of organic pore development at relatively low thermal maturity within the pre-oil bitumen generation window remains poorly understood. The objective of the current study is to examine selected samples of gilsonite from a naturally occurring solid hydrocarbon deposit to serve as an analog for characterizing the bitumen phase of generation and to document the quantity, size, and morphology of organic pores within these samples.
The Uinta Basin encompasses an area of ca. 24,000 km2 (9266 mi2) (Osmond, 1964) in northeastern Utah. From the Late Cretaceous through the Middle Eocene, more than 2200 m (7218 ft) of siliciclastic and carbonate lacustrine sediments were deposited in this asymmetrical basin, including the oil shales of the Green River Formation (Picard and High, 1968; Fouch et al., 1994). The Uinta Basin is the location of the world’s largest deposits of the solid bitumen gilsonite and is the only place where gilsonite is economically produced in large quantities.
Gilsonite is classified as asphaltite bitumen (Figure 1) according to the generic scheme of Abraham (1945). The physical properties of gilsonite are black color, conchoidal fracture, bright to fairly bright luster, brown streak, hardness on Moh’s scale of 2, specific gravity of 1.05–1.10, and a fusing point of about 120–175°C (248–347°F) (Abraham, 1945). Gilsonite occurs as vein deposits located in the northeast part of the Uinta Basin just south of its synclinal axis (Bell and Hunt, 1963). The veins are found in distinct groups or systems that are nearly parallel to each other and have a general strike of northwest to southeast (Pruitt, 1961). The gilsonite veins in the Uinta Basin are distinctive for the uniformity of their features (Figure 5). In outcrop, they extend across the landscape in essentially straight lines parallel to adjacent veins. The walls of the vein are surprisingly straight and smooth and do not deviate from the vertical by more than about 5° (Bell and Hunt, 1963). Gilsonite veins maintain their uniform widths for remarkable horizontal distances and vertical depths. Veins a few inches wide can be traced for miles in many places and vertical width changes are usually very gradual, rarely “pinching out” (Pruitt, 1961). The rock walls of the vein are seldom impregnated with gilsonite and a sharp line of demarcation typically exists in all shale and most sandstone zones (Davis, 1957). The vein material is generally pure bitumen, although occasional fragments of wall rock occur in some veins, often only a few feet from where they were broken from the wall (Bell and Hunt, 1963). Very detailed descriptions of these systems are given by Pruitt (1961), Verbeek and Grout (1992; 1993), and more recently by Boden and Tripp (2012).
The gilsonite deposits in the Uinta Basin appear to originate from the mahogany oil shale zone of the Paleogene Green River Formation (Parachute Creek Member) and are hosted primarily in the overlying Uinta and Duchesne River Formations. Verbeek and Grout (1993) concluded that the veins formed in two stages associated with thermal maturation of the Green River Formation in the deeply buried source kitchen along the synclinal axis of the basin. First, overpressuring deep in the Uinta Basin expelled large quantities of thermal water from the source rocks and induced hydraulic fractures in the overlying and adjacent strata (Uinta and Duschesne River Formations). With increased thermal maturity, a second phase of viscous liquid bitumen was extruded from the source rocks, which forced open the existing fractures and filled them with bitumen. This liquid bitumen precursor later solidified in these veins to form gilsonite, primarily through cooling, devolatization, polymerization, and/or limited biodegradation.
Gilsonite has a dull, black, coal-like appearance on weathered surfaces and a shiny, black, obsidian-like appearance on fresh surfaces. Fracture surfaces vary from conchoidal to columnar (pencillated) to flaky or scaly (Figure 6). Occasionally, gilsonite in deep parts of excavated veins occurs in a semisolid state, and this viscous fluid is called “gilsonite tar.” Major subdivisions of gilsonite are historically based on appearance and fusing (melting) temperature. The “select” grade of gilsonite tends to occur in the center of veins and is characterized by a higher luster, conchoidal fracture, and generally lower fusing point of 149–168°C (300–334°F) (Abraham, 1945; Kretchman, 1957; Boden and Tripp, 2012). The “seconds” tend to occur along vein margins or at the surface and have a somewhat duller luster, pencillated texture and a higher fusing point of 152–183°C (306–361°F) (Abraham, 1945; Kretchman, 1957; Boden and Tripp, 2012). Pencillated texture forms at right angles to vein walls and typically penetrates about 15 cm (6 in) into the ore (Verbeek and Grout, 1993). Gilsonite is reported to contain small quantities of authigenic quartz and barite, and occasionally 1–3-mm (0.04–0.19 in.) vesicles that originally contained water or gas (Monson and Parnell, 1992).
Helms et al. (2012) used advanced one- and two-dimensional solid state and solution 1H, 13C, and 15N nuclear magnetic resonance spectroscopy and electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry to investigate the structural characterization of gilsonite. They proposed a structural model in which five-membered pyrrole rings and small clusters of fused six-membered rings are connected to, and linked by, alkyl chains with an average length of 10 carbons (Figure 7). Clearly, not every molecule of gilsonite resembles this particular structure, but molecules exhibiting similar structural features are likely to be very common. Gilsonite is a complicated mixture containing a wide range of fragments with varying molecular weights. The model structure is depicted as a polymeric structure because a significant number of the pyrrole and/or aromatic subunits appear to be connected in an amorphous polymer or network structure, but the degree of interconnection and length of the connecting aliphatic chains are highly variable (Helms et al., 2012). Macrocycles such as petroporphyrins and refractory biomolecules (e.g., steranes and terpanes) are thought to constitute a small fraction of the gilsonite (Quirke and Maxwell, 1980; Ruble et al., 1994; Schoell et al., 1994).
Further information regarding the locations, uses, classification, and physical and chemical characteristics of Uinta Basin gilsonite is available in Ladoo (1920), Abraham (1945), Crawford (1949), Hunt et al. (1954), Bell and Hunt (1963), Hunt (1963), Monson and Parnell (1992), Verbeek and Grout (1992; 1993), Boden and Trip (2012), and Helms et al. (2012).
MATERIALS AND EXPERIMENTAL METHODS
The samples of gilsonite used in this study were collected in the Uinta Basin in Uintah County near the town of Bonanza, Utah (SW¼, SE¼, Sec. 15, T9S., R24E.; lat. 40.031911, long. −109.203189). These samples were collected from the Bonanza 48 mine (B-48), which is owned and operated by the American Gilsonite Company. The samples of gilsonite and gilsonite tar were collected by Tim Ruble and Christopher Laughrey (Weatherford Laboratories) during a field trip on September 26, 2013. They were collected directly from the operational mine at a depth of 305 m (1000 ft) below the surface from excavated areas within sandstones of the Eocene Uinta Formation.
Although multiple samples of gilsonite were collected during field operations at different locations within the Bonanza mine, for the purposes of the present study, single samples that are considered to be representative of conchoidal and pencillated gilsonite varieties were chosen for all subsequent analyses. These samples were selected on the basis of visual inspection and may not be representative of the entire gilsonite deposit within this particular vein or to gilsonite from other locations. As noted previously, gilsonite grade varieties with different fusing temperatures are recognized and the limited scope of the present study was not intended to thoroughly investigate pore-scale features in all gilsonite types. However, previously reported SEM imaging of gilsonite samples from other veins (Monson and Parnell, 1992) document pore-scale features that are generally consistent with observations made in the present study.
The accompanying map snapshot (Figure 8) was modified from the UGS Special Study 141 (Boden and Tripp, 2012) and shows the location of the Bonanza (Independent) vein. A red triangle has been added for the best estimate of the current sample collection location detailed above. The Bonanza vein, also known as the Independent Vein, is the northern branch of the Tabor vein west of the point where the Tabor vein splits. The Bonanza, Tabor, and Little Bonanza vein system contains the second-widest gilsonite vein in the Uinta Basin and has supported mining operations since the late 1800s (Pruitt, 1961; Boden and Tripp, 2012). Striking N61°W, this vein runs nearly 12 km (7.5 mi), has its maximum width of 4.3 m (14 ft) near where State Highway 45 crosses the vein, and has an estimated vertical extent of 1100 m (3609 ft) (Cashion, 1957).
Equipment consisted of a Zeiss MRc5 digital camera mounted on top of a Leitz Aristoment microscope and Zeiss AxioVision software for image capture and measurements. Observations were carried out in incident nonpolarized white light with a 546 nm band pass filter inserted in the light train. Macro examination of the bitumen samples was carried out using Leica 10×, 20×, and 40× air objectives on the sample material “as received,” with no preparation.
For higher magnification observations, a 50× oil objective was used on mounted and polished sample material. The sample material was mixed with cold setting epoxy resin in a 32 mm (1.25 in) circular mold and placed under vacuum to remove any trapped air and improve resin impregnation. The resulting hardened “plugs” were ground through a series of silicon carbide grinding papers and polished with 0.3 and 0.05 μm alumina suspension in accordance with ASTM D2797/2797M-11a sample preparation procedures.
Scanning Electron Microscopy (SEM)
A single conchoidal gilsonite sample was used for SEM examination. As per standard operating procedures in our laboratory, we proceeded to subject this sample to argon ion milling (AIM) using a JEOL SM-09010 AIM system for this preparation.
Ion-milling samples of gilsonite presented considerable operational difficulties. It was very challenging to prepare a sufficiently large piece of intact gilsonite onto the mount for AIM because of sample breakage and friability. Multiple attempts were required to obtain a single sample with sufficient integrity. Ultimately, the sample was placed directly into the SEM while still mounted on the mount for AIM. The risk of losing the sample was quite high if any attempt was made to demount from the AIM mount.
The sample was ultimately loaded in the SEM (FEI Helios 650 Small Dual Beam) for further examination. Imaging in the SEM was carried out at 1–5 kV to offset any surface charging, and to obtain better images on the surface of the sample.
Macroscopic examination of the pencillated (columnar) variety of gilsonite reveals a shiny black surface typically associated with fresh, unweathered samples (Figure 6). Hand samples are very friable and appear to be highly fractured and flakey. These samples often leave behind a residue of fine particulate gilsonite “dust” on the fingers when handled. Samples of pencillated gilsonite tend to have less competency and crumble quite readily along the preferred axis of the columnar fractures. This property causes considerable difficulties in preparing such samples for SEM analyses and, as a consequence, all SEM imaging reported in the current study was performed on samples of conchoidal gilsonite.
Optical organic petrology analysis of the pencillated gilsonite using an air objective shows that the polished gilsonite surfaces are typically dark grey and featureless (Figure 9). These samples appear to consist almost entirely of solid bitumen; discrete macerals and inorganic materials are absent. Visual estimates suggest that fractures make up approximately 2% of the pencillated fracture plane (Figure 9). On occasion, these fractures appear to be partially or wholly filled with the epoxy resin used to prepare the sample plug. Micropore- to mesopore-size shallow pits make up approximately 5% of the pencillated fracture plane (Figure 9). These pits are not thought to be attributed to the sample polishing process as they also occur on freshly broken surfaces that were not subjected to any polishing preparation and there was no visual evidence of “comet tails” formed by plucked material being dragged across the polished face. The pitting does appear to occur in somewhat random locations within the samples. In some instances, the pits do form a more linear pattern that appears to be cross-cut by fractures (Figure 9).
The conchoidal variety of gilsonite was also examined in this study to provide a more comprehensive documentation of the pore-scale characteristics of solid bitumen. Macroscopic examination reveals a shiny black surface typically associated with fresh, unweathered samples and an obvious conchoidal fracture surface (Figure 6). Joints and shear fracture patterns are typically present on the surfaces of the conchoidal gilsonite and are easily observed in hand samples without the aid of magnification. Hand samples of conchoidal gilsonite are much more competent than pencillated gilsonite and tend to not leave behind a residue of particulate gilsonite “dust” when handled. Despite their greater competency, even conchoidal gilsonite samples create considerable difficulties in SEM preparation.
Optical organic petrology analysis of the conchoidal gilsonite using an air objective shows that the unpolished gilsonite matrix is typically black and featureless with the exception of the diagnostic conchoidal fractures (Figure 10). These samples appear to consist entirely of solid bitumen; macerals and inorganic materials are absent. Visual estimates suggest that fractures make up approximately 1% of the conchoidal fracture plane and pitting is generally absent. Unlike the pencillated gilsonite, fractures in the conchoidal gilsonite samples do not appear to have any epoxy fill and appear to be freshly formed as a consequence of breakage. The fractures are composed of twist hackle marks commonly associated with brittle sample breakage (Figure 10).
Scanning electron microscopic images of conchoidal gilsonite show that the matrix contains no evident organic porosity. The notch of the ion-milled sample shows a somewhat irregular rugosity on the surface of the un-milled portion, with evidence of fine particulate gilsonite “dust” on the surface (Figure 11). The ion-milled surface is essentially featureless, with no evident nanoporosity at any level of magnification (Figure 12). Rare micropore pits were observed on the ion-milled surface of the conchoidal gilsonite sample, although it is likely that these pits would be much more abundant within pencillated gilsonite based upon the organic petrology results reported in this study. These pits are discontinuous and their internal surfaces appear to have a high degree of rugosity. High-magnification (up to 100,000×) examination of the subtle textural features associated with these pits confirms that they do not appear to contain any evident organic nanoporosity within the limits of SEM imaging. However, other analytical techniques such as transmission electron microscopy (TEM) and helium ion microscopy (HIM) would be required to further investigate down to the micropore scale (i.e., <2 nm) (e.g., King et al., 2015; Smith et al., 2016). The rare and occasional voids that are present in this gilsonite sample appear to simply represent highly folded rugosity along the surface walls of the pits (Figure 12).
Scanning electron microscopic images of the edges of the ion-milled gilsonite surfaces show the occurrence of fractures that are typical of conchoidal gilsonite (Figure 13). Conchoidal fracture planes exhibit twist hackle marks even at high SEM magnifications (Figure 13). One concern regarding the preparation of the ion-milled surface was whether the sample preparation process may have altered the surface of the solid bitumen material. Although previous investigations have demonstrated this is unlikely (Hooghan et al., 2017), one method to further examine such potential experimental artifacts was to examine a tilted view image of the ion-milled contact (Figure 14). Image results in 45° tilted view show similar conchoidal fracture surfaces and an absence of nanoporosity, providing additional circumstantial evidence that the ion-milling process is unlikely to have altered gilsonite to any significant extent.
The gilsonite sample in this study was examined in the SEM using both secondary electron and backscattered modes (Figure 15). No major differences on the ion-milled surfaces were noted as a consequence of the SEM imaging mode and nanoporosity appears to be absent.
One observation was noted during the reexamination of the gilsonite sample via SEM. Upon returning to the location of the sample that contained a distinct pit on the ion-milled surface, the previously featureless matrix surrounding the pit was found to exhibit alteration apparently associated with the SEM imaging (Figure 16). Fractures and distortion within the matrix were observed in a clear pattern that was obviously induced by the previous SEM imaging experiments. The susceptibility of gilsonite to this type of SEM image-induced alteration is likely attributable to its brittle nature, although the fundamental physical and chemical alterations that may be responsible for this were not further investigated.
Based upon geochemical evidence, the vein forming gilsonite deposits in the Uinta Basin appear to originate from very organic-rich offshore open lacustrine source rock facies deposited in a stratified lake with predominantly algal and bacterial biomass (Ruble, 1990; Schoell et al., 1994). Previous studies have established the probable source of gilsonite as the bitumen-rich marlstone beds (oil shale) of the upper part of the middle Eocene Green River Formation (Parachute Creek Member) equivalent to the mahogany zone (Hunt et al., 1954; Bell and Hunt, 1963; Hunt, 1963). The extruded bitumen that solidified to form gilsonite was emplaced during the early stages of post-Laramide regional tectonic extension, and emplacement depths are estimated to be 700–2500 m (2297–8202 ft) (Verbeek and Grout, 1993).
There is abundant evidence to support forceful rather than passive intrusion of the gilsonite dikes, which suggests that the dikes propagated as hydraulic fractures owing to overpressures caused by bitumen generation in downdip Green River source rocks (Verbeek and Grout, 1993). The widespread occurrence of gilsonite sills injected along bedding shows that fluid pressures at the time of injection frequently exceeded lithostatic load (Verbeek and Grout, 1993). Pepper (2017) noted one form of bitumen that is transitional with kerogen and apparently represents organic matter whose structure has been mechanically weakened by bond breakage during thermal maturation. Lacking the ability to migrate under its own buoyancy, this form of pre-oil bitumen can only be redistributed under the overburden load within the source bed—and sometimes well beyond in bitumen dykes as Pepper (2017) suggested. A microscopic or macroscopic (outcrop) section of any solid bitumen such as gilsonite that is seen in a load-bearing position must have once been a non-Newtonian fluid whose characteristics change from solid to mobile in response to elevated pressure and temperature (Pepper, 2017). This concept appears to be an extension of the petroleum generation mechanism proposed by Lewan (1997), which implies that petroleum generation induces chemical and physical alterations during the process of thermal maturation. Pepper (2017) further suggested that this bituminized organic matter is extruded into load-protected positions, where open pores can develop within it. In contrast, data presented in the current investigation suggest that bitumens formed in this manner (at least pre-oil bitumens such as gilsonite) do not develop extensive organic pore networks.
The presence of limonite and calcite as early deposits on the sandstone walls of the gilsonite veins and of alteration rinds (bleached zones) suggest that the dike fractures were conduits for the expulsion of significant quantities of formation water before the source beds were sufficiently mature to generate significant amounts of bitumen (Verbeek and Grout, 1993). The viscous bitumen that solidified to gilsonite was emplaced as an immature pre-oil into large hydraulic fractures, which formed as a probable consequence to overpressures generated during an early stage of bitumen generation when the Uinta Basin source kitchen was buried to its maximum depth (Verbeek and Grout, 1993). The bitumen that formed gilsonite likely migrated laterally through the vein-fracture system and gilsonite in the easternmost veins was derived from more mature downdip source beds to the northwest and not from the oil-shale deposits directly beneath (Verbeek and Grout, 1993; Ruble et al., 2001).
The fluidized bitumen precursor to gilsonite must have moved into the veins as a viscous flow because only minor amounts of this material have penetrated the porous sandstone wall rock of the veins (Bell and Hunt, 1963). However, the minor extent to which gilsonite was able to invade the wall rock has also been attributed to prior precipitation of chlorite cement, which severely restricted the sandstone permeability (Monson and Parnell, 1992). During sample collection associated with the current study, mined areas of the gilsonite veins were noted to have occurrences of “gilsonite tar” seeping from the exposed sandstone wall rock. This material apparently represents continued active migration of the viscous bitumen precursor to gilsonite from a probable active source pod in deeper portions of the Uinta Basin (Ruble et al., 2001). Because of erosion and overburden removal, the present-day bitumen migration process (as opposed to initial post-Laramide hydraulic fracturing and bitumen emplacement) would appear to be more of a passive event rather than a forceful extrusion because the present-day locations of the gilsonite deposits are far outside the current overpressured pod of presumed oil generation (Ruble et al., 2001).
Previous petrographic examination of “gilsonite tar” material (T. Ruble, unpublished data) shows that it is a viscous water-in-oil emulsion that also contains small fragments of completely solidified gilsonite (Figure 17). It is proposed herein that the water droplets contained in this “gilsonite tar” emulsion likely represent the precursors to the remnant and discontinuous pits that were observed in the gilsonite samples examined during the current study. The viscous bitumen that filled the veins subsequently solidified as a consequence of devolatilization and limited but variable biodegradation to form distinct gilsonite types that can be recognized based on their physical (melting point) and geochemical characteristics (T. Ruble, unpublished data). The occurrence of pits within the gilsonite is rare within the conchoidal variety but more common with the pencillated variety. Pencillated gilsonite is generally located nearest to the adjacent wall rock.
If the gilsonite was emplaced in the vertical veins as a water-in-oil emulsion that was inherently unstable and progressively separated, it is possible that the water droplets would preferentially migrate toward the more porous and water-wet sandstone wall rock. This zone could also be a location of preferential biodegradation, which could partially explain the occurrences of different grades of gilsonite even within the same vein and the origin of the pitting in pencillated varieties as documented in the current investigation. Recent SEM imaging of other Uinta Basin solid bitumen samples (wurtzilite) also document the occurrence of similar pitting within the matrix, but some pits contained easily identifiable bacteria (W. Camp, personal communication), which strongly suggests that these areas were sites of active biodegradation.
The results of our analyses suggest that, unlike pyrobitumen, pre-oil solid bitumen represented by Uinta Basin gilsonite was found to contain no significant occurrences of organic nanoporosity within its matrix. Gilsonite does have minor pitting and fractures, but these do not represent an effective interconnected pore network and may be artifacts of weathering/sampling. Alternatively, pitting and the development of pencillated textures within vein deposits of gilsonite may be a remnant artifact of the solidification process associated with a water-in-oil emulsion (gilsonite tar). Regardless, this pre-oil solid bitumen material would not represent a potential candidate for in-situ petroleum storage capacity. Whether this is typical of all naturally occurring pre-oil solid bitumens is debatable, considering that gilsonite has undergone extrusion from the source rock and has been altered by some secondary processes, such as devolatilization and limited biodegradation. Nevertheless, the pore-scale imaging of this solid bitumen may provide potentially important new insights for unconventional reservoir characterization.
The authors wish to acknowledge the generous assistance offered by the American Gilsonite Company in facilitating collection of the samples used in the present study. Andrew Bishop assisted with the acquisition of the unpublished “gilsonite tar” photomicrographs. We would like to thank Joseph Curiale and an anonymous reviewer for their careful reviews and constructive comments that helped improve the manuscript. This manuscript is dedicated to the memory of Kathy Ruble, who passed away in 2017. Kathy was a good friend to all of the co-authors of this manuscript and also my beloved spouse. She participated in the collection of the samples used in this study and ventured down the gilsonite mine shaft along with myself during a field trip after an AAPG meeting in Salt Lake City. Kathy attended the Santa Fe Hedberg conference, joining the spouse program, and met many of the attendees—thus I wanted to acknowledge her in this manner.—Tim Ruble