Fossilized worm burrows influence the resource quality of porous media

Transcription

1 Fossilized worm burrows influence the resource quality of porous media Murray K. Gingras, Carl A. Mendoza, and S. George Pemberton ABSTRACT Burrow-associated, selective dolomitization in the Yeoman Formation limestone (Ordovician, Williston basin) is characterized by distinct textural heterogeneity. Physical parameters such as permeability, porosity, tortuosity, and dispersivity are therefore difficult to assess. This study compares the relative dispersivities of three geologic media: homogeneous sandstone, fractured limestone, and burrowed dolomitic limestone. Results show that the flow paths present in burrow-associated dolomite are tortuous, and that the interaction between the flow paths and the matrix is extensive. Such rocks act as dual-permeability systems in the subsurface. Hydrocarbon production from such deposits will be strongly influenced by burrow-related heterogeneity, and its influence should be carefully considered before secondary recovery schemes are implemented. INTRODUCTION GEOLOGIC NOTE The fabric of carbonate strata can be dominated by chaotically distributed mottles in which the most common mineral is dolomite. Selective dolomitization of calcareous rock may occur adjacent to body fossils, in association with sedimentary lamination, or around the fossilized burrow passages of animals (such as worms) that colonized the sediment before it passed into the rock record (Kendall, 1975; Zenger, 1992; Gingras et al., 2002). This paper focuses on fabrics where dolomite has formed in association with ( burrow) trace fossils. Burrow-associated dolomite represents an important textural element of many Paleozoic carbonates. Exceptional North American examples include the Devonian Palliser Formation in the Copyright #2004. The American Association of Petroleum Geologists. All rights reserved. Manuscript received June 24, 2003; provisional acceptance September 4, 2003; revised manuscript received December 29, 2003; final acceptance January 26, AUTHORS Murray K. Gingras Department of Earth and Atmospheric Sciences, 1-26 Earth Sciences Building, University of Alberta, Edmonton, Alberta, Canada, T6G 2E3; mgingras@ualberta.ca Murray Gingras received his diploma in mechanical engineering technology from the Northern Alberta Institute of Technology in 1987, his B.Sc. degree from the University of Alberta in 1995, and his Ph.D. from the University of Alberta in Gingras has worked professionally in the hydrocarbon industry, at the Northern Alberta Institute of Technology, and as an assistant professor at the University of New Brunswick. His research focuses on applying sedimentology and ichnology to sedimentary rock successions, as a paleoecological tool, a reservoir development tool, and in process-driven sedimentology. Carl A. Mendoza Department of Earth and Atmospheric Sciences, 1-26 Earth Sciences Building, University of Alberta, Edmonton, Alberta, Canada, T6G 2E3 Carl Mendoza received a B.A.Sc. degree in geological engineering from the University of British Columbia in He worked for Shell Canada Resources for more than two years, before obtaining his M.Sc. degree and his Ph.D. in hydrogeology from the University of Waterloo. He has been at the University of Alberta since 1992 and is currently an associate professor. His research focuses on reactive gas and vapor transport in the unsaturated zone, ground-water/surface-water interactions at both natural and oil-sands impacted sites, and flow and transport in heterogeneous porous media. S. George Pemberton Department of Earth and Atmospheric Sciences, 1-26 Earth Sciences Building, University of Alberta, Edmonton, Alberta, Canada, T6G 2E3 S. George Pemberton is a professor in the Department of Earth & Atmospheric Sciences at the University of Alberta. He is a Fellow of the Royal Society of Canada and holds a Canada Research Chair in Petroleum Geology (Natural Sciences and Engineering Research Council). George s field of research and expertise are in the field of ichnology, the investigation of animal-sediment interactions in both recent and ancient environments. Current research activities include the AAPG Bulletin, v. 88, no. 7 (July 2004), pp

2 application of trace-fossil studies in sequence stratigraphy and the exploration and exploitation of hydrocarbons. Recent research activities involve emphasis on the Cardium and Viking formations, the Athabasca and the Cold Lake oil sands of Alberta, as well as the offshore Hibernia, Ben Nevis, Terra Nova, and Venture fields. ACKNOWLEDGEMENTS Many of the members of the Ichnology Research Group at the University of Alberta were instrumental in generating ideas for dispersivity testing equipment. Special thanks to Floyd (Bo) Henk, Tom Saunders, John-Paul Zonneveld, Demian Robbins, Arjun Keswani, and Brian Jones for their insight and comments. Vivienne Robertson provided editorial assistance. Funding for this research was made available by the Natural Sciences Engineering and Research Council (NSERC) Research grants to Murray Gingras, Carl Mendoza, and George Pemberton, and an NSERC equipment grant to Murray Gingras. Western Canadian sedimentary basin (Gingras et al., 2002), the Upper Ordovician Bighorn Dolomite in the Bighorn basin of Wyoming (Zenger, 1992), and the Ordovician Yeoman Formation in the Williston basin (Kendall, 1975, 1977). These visually striking rocks form an important class of geological media because their physical characteristics, such as permeability and porosity, may vary widely between burrowed and unburrowed zones. In the above examples, dolomite formation is typically focused in and around generations of interpenetrating ichnofossils. Thus, the heterogeneity of such rocks is invariably high. These networks provide tortuous pathways for fluid transmission and result in complex storage characteristics and interactions. Tortuous, heterogeneous media present a notable complication for reservoir development. We propose that unlike fractured rock, which can be viewed as a dual-porosity system with distinctly mobile and immobile zones, burrow-mottled dolomite forms a heterogeneous dual-permeability system, where zones range from being mobile or less mobile. That is, burrows may provide flow conduits that interact extensively with the surrounding porous matrix. Their tortuous nature implies that dead-end pores may be common. An understanding of how burrow-associated heterogeneities control fluid flow in dolomitized units is necessary if petroleum production from such reservoirs is to be optimized. Several fluid-flow parameters are expected to depend on the degree of burrow dolomitization. These include the permeability, porosity, and dispersivity. Dispersivity is a coefficient that represents the degree of mechanical mixing that occurs during flow through a porous medium. It includes the effects of variable path lengths (tortuosity), varying pore size, pore frictional effects, and heterogeneity (Fetter, 1999). In a dual-permeability medium, gross dispersivity also reflects the interaction between the mobile and less mobile domains (de Marsily, 1986). Capillary pressure vs. saturation relationships could also provide information about the connection between the two hydraulic domains; such static functions only elucidate the range of pore sizes (Dullien, 1992). Dispersivities, however, provide insight into the dynamic interactions between domains. Thus, we focus on dispersivity relationships here. We use simple dispersivity measurements to compare relative tortuosities and hydraulic interactions for Tyndall stone (Red River Group, Yeoman Formation), which contains dolomitized burrows, to homogeneous sandstone and fractured limestone. The laboratory parameters are not necessarily the same as subsurface values, but they serve to demonstrate that burrow-mottled carbonate rocks should be represented and exploited as a dual-permeability system. The bulk permeability of the geological media is also evaluated. Previous Work Although a small body of literature recognizes that bioturbation can influence reservoir quality (summarized in Gingras et al., 1999), 876 Geologic Note

3 essentially no research characterizes the dispersivity of burrowed media. Some recent papers (Dornbos et al., 2000; Fong et al., 2001; Gingras et al., 2002) do recognize the overall complexity of burrow-mottled sedimentary media, but do not attempt to assess flow and transport behavior at a local scale. This is unfortunate because dispersivity is related to the overall tortuosity of the flow paths, thus controlling the overall amenability of resource extraction to different production techniques. More importantly, the interpretation of dispersivity data can lead to an understanding of the fluid interactions between primary flow paths and the matrix in dual-permeability systems. Enhanced knowledge of dual-permeability systems is crucial for implementing secondary or enhanced recovery techniques in such reservoir settings. In contrast, studies of dispersivity in fracture systems are comparatively common: this body of research aims to (1) assess the degree of mechanical mixing induced by flow through fractures and (2) quantify interactions between fracture flow paths and diffusiondominated matrix. Fractured media studies have depended on numerically, experimentally, and field-based research, thus providing an understanding that is conceptually applicable to dual-permeability systems induced by fossil bioturbation. For instance, Becker and Shapiro (2003) used dye tracers to assess the dispersivity of in-situ fractured bedrock, a methodology conceptually similar to that outlined later. Unlike burrow fabrics, however, fractures are amenable to numerical modeling (McKenna et al., 2003; Reimus et al., 2003). Because of the complex and chaotic nature of ichnofossil geometry, similar endeavors are still anticipated for bioturbated sedimentary media. Finally, studies in fractured media (e.g., Esposito and Thompson, 1999) have shown that simulation of multiphase flow in dualporosity systems is complex, yet tractable. Multiphase flow through bioturbated media is not considered in this study, but it is certainly a direction that future research should take. Methods Porosity (f) estimates were made from blue-epoxy impregnated petrographic sections. Point counts were used to estimate the proportion of void space vs. total volume. Point counts used 300 line-intersection points for each domain, i.e., matrix dolomite and burrow dolomite in sample 1 and the matrix of samples 2 and 3 (samples described later). Spot permeability (Dreyer et al., 1990) was measured using a portable probe permeameter (CoreLabs Model PP-250), with nitrogen as a pore fluid. Permeabilities between 0.01 md and 3 d are considered to provide optimal accuracy for the device ( ±1%). Permeability measurements were made on selected zones deemed to represent textural domains of the rock. Bulk permeability and dispersion characteristics were determined using a dispersometer (Figure 1). Each sample was cut into a 10-cm (4-in.) cube and bonded to Figure 1. Process and instrumentation schematic of the dispersometer. Inlet and outlet gas streams are controlled with pressure regulators. The tracer gas, O 2, is measured at the outflow of the device. Gingras et al. 877

4 Figure 2. Results from the dispersometer flow experiments: relative tracer-gas concentration (C/C 0 ) vs. gross pore volumes exchanged (V *). The inset shows a broader range for pore volumes exchanged to better illustrate the leading exhibited by the fractured and the burrowed carbonate rocks. the inside of the chamber with polyester resin. First, steady-state flow of N 2 gas was established across the sample, and volumetric flow rates were measured at the discharge end. The effective permeability was calculated using Darcy s law, modified for compressible gas transport. Next, O 2 tracer gas, at a constant concentration, was introduced into the steady-state flow field. Effluent O 2 concentrations were measured through time using a hand-held probe with a resolution of 0.1 ppm and accuracy of ±1.5%. These data were used to compare relative degrees of dispersion between samples. Dispersometer results are shown as breakthrough curves (Figure 2) of tracer concentration ( y-axis) vs. pore volumes filled (x-axis). Relative concentration (C/ C 0 ) is the measured effluent concentration (C) normalized by the constant, applied influx concentration (C 0 ). Thus, the breakthrough curves show the rate that tracer gas (O 2 ) concentration increases at the O 2 probe. Breakthrough occurs because O 2 displaces the N 2 that was used to initially flood the porosity. The shape of the breakthrough curve is related to the efficiency of the tracer s sweep and can be related to the nature of the permeability system in the rock (see below). One pore volume corresponds to the complete, uniform exchange of gas contained in the void space of the sample. Accounting for end effects, pore volumes for our dispersometer (Figure 1) are defined by V ¼ QP exit t fv P total V in P in V out P out in QP exit QP exit where P in and P out are the inlet and outlet gas pressures; P exit is the pressure at which the volumetric flow rate, Q, is measured; V in and V out are the volumes of the 878 Geologic Note

5 Figure 3. A28 17-cm (11 7-in.) (12 cm [5 in.] thick; not shown) block of quarried Tyndall stone, a piece of which was used in this study. Lighter areas are composed of limestone and represent the original limestone matrix. Darker mottles are dolomite and are mostly associated with trace fossils. Minipermeameter measurement values (millidarcy) are indicated. inlet and outlet components of the chamber and tubing; V is the sample volume; and t represents the elapsed time for the flow experiment. V* is the number of pore volumes of gas passed through the sample given quantification of the aforementioned variables. The relationship between effluent tracer concentrations and pore volumes exchanged depends on the heterogeneity and so the dispersivity and dual-permeability nature of the rock. For one-dimensional transport through a perfectly homogeneous medium, C/C 0 = 0.5 when V* = 1.0, and the breakthrough curve is compact and nearly symmetric (Fetter, 1999), with a final C/C 0 = 1.0. Henceforth, V* for C/C 0 = 0.5 is denoted as V* 0.5. A compact breakthrough curve is characterized by a rapid rise of C/C 0. Symmetry suggests the curve has a similar (but mirrored) profile about C/C 0 = 0.5. Symmetrical breakthrough curves typically show a somewhat linear rise of C/C 0 with similar leading and tailing characteristics. Leading refers to the rate C/C 0 increases during the initial flooding of the pore space. The term first arrival, which is referred to below, refers to the first appearance of tracer gas at the probe and is indicative of the onset of leading. Prolonged leading suggests that C/C 0 rises slowly during the early part of the experiment, possibly because of flow-path tortuosity. Tailing refers to the how rapidly C/C 0 approaches 1.0 near the end of the experiment. Long tailing appears as slow ascent to C/C 0 = 1.0 and can be related to dual-permeability systems. The sandstone curve of Figure 2 demonstrates compactness and, to a lesser degree, symmetry; the curve for the burrowed limestone is not compact, and it shows significant tailing. Increased rock heterogeneity leads to more spreading of the breakthrough curve. Dual-permeability regimes are further manifest by early breakthrough (leading) at low concentrations and long tailing; that is, the C/C 0 will only approach 1.0 after many more pore volumes (de Marsily, 1986). Such processes result in a pronounced asymmetry of the breakthrough curve. In general, higher values for V* are associated with greater dispersion and interaction with the matrix. Geologic Material Used The sample database consisted of three rock samples. These included a burrow-mottled, dolomitic limestone (sample 1), a laminated sandstone (sample 2), and a fractured, massive-appearing limestone (sample 3). The focus of this study, sample 1, is burrowmottled, dolomitic limestone (Figure 3) procured from outcrop (building-stone quarry) of the Ordovician Tyndall stone in Manitoba. This unit is formally the Selkirk Member of the Red River Group (Cowan, 1971; Kendall, 1977). Core studies show that burrow-mottled, dolomitic limestone occurs primarily in the lower half of the Selkirk Member (Kendall, 1977). Similarly mottled strata comprise the subsurface-equivalent Yeoman Formation, which is present in southeastern Saskatchewan (Kendall, 1975). Age-equivalent, subsurface strata provide various exploration targets in the Williston basin of Saskatchewan, Manitoba, and North Dakota. The distinct sedimentologic and ichnologic signature of Tyndall stone is similar to other Paleozoic epicontinental carbonates, including the Devonian Palliser Formation of the Western Canadian sedimentary basin Gingras et al. 879

6 (Beales, 1953) and the Upper Ordovician Bighorn facies from Williston basin deposits in the United States (Zenger, 1996 a, b). Minipermeameter measurements show that the average matrix permeability is 1.65 md, and the average burrow permeability is 19.2 md. Although the matrix permeability is relatively uniform, burrow permeability varies widely, ranging between 1 and 100 md (Figure 3). Based on five steady-state flow experiments, bulk permeability of the sample is 17 md. The porosity of the calcite matrix ranges between 2 and 5%. The porosity of the dolomite mottles ranges from 6% near the mottle margins to 20% near the mottle cores. Bulk porosity is approximately 5%. Sample 2 is an outcrop-derived sandstone from the Cretaceous Horseshoe Canyon Formation in central Alberta. Although this formation is not a typical resource target in Alberta, it is host to notable coal-bed methane and shallow natural gas. This rock was chosen for its overall homogeneity, which contrasts with the heterogeneity of sample 1. The laminated sandstone comprises fine- to medium-grained sandstone (approximately 0.25 mm [0.01 in.] mean grain diameter), with very little interstitial silt and clay. The sample is calcite cemented, with a net porosity of 12%. Spot permeability assessments range between 15 and 90 md. Bulk permeability (three experiments) averaged 27 md. Sample 3 comprises a comparatively homogeneous (subsequently fractured) limestone taken from Devonian Palliser Formation outcrop in Alberta. Palliser strata are equivalent to the Wabamun Formation, which is exploited for natural gas hosted in fractured reservoirs and diagenetic traps. The fine-grained, calcareous media had a low permeability (<1 md). One fracture was mechanically induced, and the sample was set into epoxy resin with about 0.1 mm (0.004 in.) fracture aperture. Postexperiment inspection showed that epoxy resin intruded the fracture, from the sides, to a depth of 2 9 mm ( in.). The porosity of the calcite matrix was dominated by micropores, which occupied from2to6%ofthesamplevolume.basedonthemeasured fracture aperture, porosity was about 0.1%. Bulk permeability (two experiments) averaged 140 md. RESULTS Figure 2 shows breakthrough curves for all three samples. The breakthrough curve for the unburrowed sandstone (sample 2) shows that the tracer gas arrives as an abrupt front. Little tracer leading is observed; the first arrival of the tracer gas occurs at approximately V* = 0.1. Tracer gas concentrations increase linearly, and V* 0.5 = 1.2. Tailing is minimal, and C/C 0 = 0.9 occurs by V* = 2.9. The fractured limestone (sample 3) shows similar behavior, except that strong tailing is observed. Minor leading occurs as much as C/C 0 =0.1atV* = 0.5, but tailing, where C/C 0 slowly exceeds 0.8, is evident after V* = 3. Between these fronts, the tracer concentration increases rapidly, and V* 0.5 = 2.1. In contrast, the dispersivity curve for the dolomite-mottled limestone (sample 1) is characterized by a delay in the first arrival and long tailing. That is, first arrival of the gas occurs at V* = 0.5, but tailing is prolonged, i.e., V* < 0.95, until after V* exceeds The breakthrough curve is distinctly asymmetric about C/C 0 = 0.5. Tracer concentrations increase slowly compared to the sandstone and fractured limestone, and V* 0.5 = 6.6 is almost six times greater than that of the sandstone. DISCUSSION The data for the unburrowed sand (sample 2) represent continuous, uniform flooding of the pore space by the tracer gas. The absence of strong leading or tailing indicates a well-defined front. These data point toward relatively homogeneous flushing of the pore space. Such behavior is indicative of transport through a comparatively homogeneous media (Sauty, 1980; Fetter, 1999). Fractures are commonly characterized by permeabilities several orders of magnitude higher than porous matrix permeabilities. Thus, fractured porous media commonly represent the end-member, dual-porosity, dual-permeability permeability system. The breakthrough data produced for the fractured limestone (sample 3) illustrate this. As with the burrowed sample (Figure 2; discussed below), tailing is prolonged, but this follows abrupt flooding of the fracture. The tailing is inferred to result from diffusion-controlled interaction between the fractures and the matrix. Early arrival of the tracer and rapid increase of C/C 0 suggests that the fracture porosity efficiently delivers gas. In contrast, the data for dolomite-mottled limestone (sample 1) are typical of a heterogeneous flow media (Sauty, 1980). The early first arrival (i.e., exaggerated leading) of the tracer represents preferential flow of gas through burrow-dolomite conduits. Slow increase of the tracer-gas concentration is attributed to 880 Geologic Note

7 slow gas migration from the burrow flow paths into the lower permeability matrix. It also reflects the effect of the tortuous flow networks afforded the gas. The strong tailing (i.e., asymmetry) is best explained by gas exchange between the burrow-dolomite and the matrix and between flow paths and any ineffectively connected porous zones. This is because gas flows comparatively freely through the burrowed zones, but passes only slowly into and out of the limestone matrix. Thus, these data are interpreted to represent the response to a heterogeneous, dual-permeability flow system. Similar interaction with matrix fluids (via diffusion) has been demonstrated in fractured media (Neretnieks, 1980). Significance and Application to Reservoir Rocks Compared to fractured media, burrow-associated heterogeneities typically occupy a far greater volume of the rock. In the above example, dolomite volume (associated with trace fossils) is approximately 45%. Such a significant volume represented by heterogeneous aspects of the dolomite-mottled limestone results in a very high interfacial surface area between the two permeability systems present in the sample. This interface is locally sharp but commonly gradational; thus, it is difficult to model. Moreover, fractures represent planar elements that are to a degree geometrically predictable (i.e., fractures are conceptually simpler to model compared to biogenic structures.) Gingras et al. (1999) attempted to demonstrate the effects of burrow connectivity in three dimensions. The results of that study indicated that permeability in bioturbated substrates was most sensitive to (1) the degree of permeability contrast between the matrix and the burrowed zone, (2) the degree of three-dimensional connectivity between burrow systems (commonly by interpenetration), and (3) the density of higher permeability burrow structures observed in the rock. Notably, the degree to which connectivity occurs is typically high for rocks that are moderately to highly bioturbated at the core scale. Regarding dispersivity, the overall orientation of the trace fossils is probably not crucial to understanding the burrow-matrix interactions. This is because such interactions are primarily determined by the surface area separating the two permeability systems, the permeability contrast between those permeability regimes, and the overall three-dimensional connectivity of the flow network. However, meandering flow paths with numerous branches afford more zones of null flow than simple burrow geometries and are deemed to influence the dispersivity of the flow media. The above examples demonstrate that dual-permeability systems strongly influence resource quality in porous media; this is a function of both the bulk permeability and rock dispersivity. The presence of permeability streaks in an overall tortuous medium suggests that complex flow behavior is expected from a reservoir dominated by permeable burrow-related fabrics. Where the burrows are characterized by higher permeability than the matrix, early production will primarily drain the burrow networks (Figure 4). As production from a reservoir matures, the resource will increasingly be derived from the matrix. Where hydrocarbons are being produced, secondary recovery techniques should be carefully evaluated because some methods, such as waterflooding, will potentially isolate the matrix by occupying the burrow flow conduits with water. In the example of the Yeoman Formation (i.e., sample 1), 40 60% of the initial resource would be bypassed by waterflood and thus isolated. However, it is worth suggesting that dual-permeability systems (such as burrow-mottled sample 1) have better reservoir potential than dual-porosity systems (like fractured sample 2), because slow, constant production from reservoirs dominated by dual permeability can ensure better recovery compared to draining the primary flow network of a dual-porosity system. All in all, flow media that are strongly influenced by a burrowed fabric represent a difficult challenge to predict, model, and characterize. Nevertheless, understanding these fabrics will contribute to better reservoirmodeling techniques in bioturbated reservoir rocks. The application of this technology to reservoir settings will depend on laboratory, numerical, and field-based techniques to determine the influence of complex biogenic fabrics on the deliverability of fluids, recoverability of fluids, and methods of sweeping similar fabrics for enhanced recovery. SUMMARY Burrow-related, selective dolomitization results in distinct textural heterogeneity, where permeable conduits are chaotically distributed, and flow paths are tortuous. Consequently, interaction between the flow paths and the matrix may be extensive. Furthermore, bioturbated media can act as dual-porosity dual-permeability systems in the subsurface. Production strategies for bioturbated units should consider these factors because Gingras et al. 881

8 Figure 4. (A) Petrographic section demonstrating (1) the flow path in burrowed carbonate, in this case Tyndall stone; (2) the difference in porosity between the matrix and the mineralogically altered burrowed zone; (3) the direction of diffusion into and out of the burrow fabric; and (4) the complex nature of such fabrics. (B) Conceptualized production from a strongly heterogeneous reservoir. This diagram suggests that resource production will consist of two phases. First, the main phase of production will be derived from the most permeable zones. Second, slower, prolonged production will permit better exploitation from the less accessible matrix. they can strongly influence the efficiency of primary and secondary recovery. Future research should focus on the characterization of similar units in the subsurface, production history from selected pools, and geochemical analysis of the dolomitic and calcitic components of the reservoir rock. REFERENCES CITED Beales, F. W., 1953, Dolomitic mottling in Palliser (Devonian) Limestone: Banff and Jasper National Parks, Alberta: AAPG Bulletin, v. 37, p Becker, M. W., and A. M. Shapiro, 2003, Interpreting tracer breakthrough tailing from different forced-gradient tracer experiment configurations in fractured bedrock: Water Resources Research, v. 39, p. 13. Cowan, J., 1971, Ordovician and Silurian stratigraphy of the interlake area, Manitoba, in A. C. Turnock, ed., Geoscience studies in Manitoba: Geological Association of Canada Special Paper 9, p de Marsily, G., 1986, Quantitative hydrogeology: Orlando, Florida, Academic Press Inc., 440 p. Dornbos, S. Q., W. Phelps, D. J. Bottjer, M. L. Droser, and B. Anderson, 2000, Effects of bioturbation on reservoir sandstone porosity and permeability: Studies of outcrop samples from the Upper Cretaceous, Book Cliffs, Utah (abs.): AAPG Annual Meeting Program, v. 9, p. A40. Dreyer, T., A. Scheie, and O. Walderhaug, 1990, Minipermeameterbased study of permeability trends in channel sand bodies: AAPG Bulletin, v. 74, p Dullien, F. A. L., 1992, Porous media fluid transport and pore structure, 2d ed.: San Diego, California, Academic Press, 574 p. Esposito, S. J., and N. R. Thomson, 1999, Two-phase flow and transport in a single fracture-porous medium system: Journal of Contaminant Hydrology, v. 37, p Fetter, C. W., 1999, Contaminant hydrogeology, 2d ed.: Upper Saddle River, New Jersey, Prentice Hall, 500 p. Fong, G. W., S. G. Pemberton, M. K. Gingras, and B. Henk, 2001, Assessing the reservoir quality of burrow-mottled carbonates in the Devonian Wabamun Group, Pine Creek area, northwest 882 Geologic Note

9 Alberta, Canada (abs.): AAPG Annual Meeting Program, v. 10, p. A65. Gingras, M. K., S. G. Pemberton, C. Mendoza, and F. Henk, 1999, Modeling fluid flow in trace fossils: Assessing the anisotropic permeability of Glossifungites surfaces: Petroleum Geoscience, v. 5, p Gingras, M. K., B. MacMillan, B. J. Balcom, and S. G. Pemberton, 2002, Visualizing the internal physical characteristics of carbonate sediments with magnetic resonance imaging and petrography: Canadian Society of Petroleum Geologists Bulletin, v. 50, p Kendall, A. C., 1975, Anhydrite replacements of gypsum (satin-spar) veins in the Mississippian caprocks of southeastern Saskatchewan: Canadian Journal of Earth Sciences, v. 12, p Kendall, A. C., 1977, Origin of dolomite mottling in Ordovician limestones from Saskatchewan and Manitoba: Bulletin of Canadian Petroleum Geology, v. 25, p McKenna, S. A., D. D. Walker, and B. Arnold, 2003, Modeling dispersion in three-dimensional heterogeneous fractured media at Yucca Mountain: Journal of Contaminant Hydrology, v. 62, p Neretnieks, I., 1980, Diffusion in the rock matrix, an important factor in radionuclide migration: Journal of Geophysical Research, v. 85, no. B8, p Reimus, P. W., M. J. Haga, A. I. Andrews, T. J. Callahan, H. J. Turin, and D. A. Counce, 2003, Testing and parameterizing a conceptual solute transport model in saturated fractured tuff using sorbing and nonsorbing tracers in cross-hole tracer tests: Journal of Contaminant Hydrology, v. 62, p Sauty, J. P., 1980, An analysis of hydrodispersive transfer in water: Water Resources Research, v. 16, p Zenger, D. H., 1992, Burrowing and dolomitization patterns in the steamboat point member, Bighorn Dolomite ( Upper Ordovician), northeast Wyoming: Contributions to Geology, v. 29, no. 2, p Zenger, D. H., 1996a, Dolomitization patterns in widespread Bighorn facies ( Upper Ordovician), western craton, U.S.A.: Carbonates and Evaporites, v. 11, p Zenger, D. H., 1996b, Dolomitization of the C zone, Red River Group (Upper Ordovician) in deep core, Williston basin, Richland County, eastern Montana: Contributions to Geology, v. 31, no. 1, p Gingras et al. 883