In August of 2015 the first Mountjoy Carbonate Conference, co-hosted by the Society for Sedimentary Geology (SEPM) and Canadian Society of Petroleum Geologists (CSPG), took place in Banff, Alberta. As the approaches to characterization and modeling of carbonate reservoirs are undergoing rapid changes, this was the theme of the meeting. This Special Publication, following the inaugural meeting, contains nine state-of-the art papers relating to the (1) characterization of carbonates and advances in analytical methods, (2) controls on carbonate reservoir quality and recovery factors, and (3) reservoir distribution, the modeling of dolostone geobodies, and reservoir prediction. The Introduction includes an overview of Eric Mountjoy’s career and his many contributions to the science. The contents of this Special Publication should be useful to those engaged in the characterization and modeling of carbonate reservoirs, including unconventional carbonate reservoirs, and is highly recommended as one of the most impactful recent publications for those working in this area of sedimentary science.

Abstract Numerous field and laboratory studies over the past two decades claim that microbes catalyze nucleation and growth of dolomite at temperatures common in low-temperature geologic environments (25–60°C). However, a critical reexamination of the X-ray diffraction (XRD) data presented by these studies indicates that the laboratory products are not dolomite but rather a mixture of minerals, including very high-magnesium calcite (VHMC). Because VHMC can be “compositionally” indistinguishable from dolomite (i.e., 50 mol% MgCO3), the positions of the principal (104) XRD reflection for VHMC and dolomite can be identical. Nevertheless, published XRD patterns of products derived from microbial experiments lack convincing evidence of cation ordering, which is a unique characteristic of carbonate minerals exhibiting (dolomite) symmetry. The lack of cation ordering in laboratory precipitates instead indicates that the products are VHMC, which possesses (calcite) symmetry. Hence, previous laboratory studies have misidentified VHMC for dolomite. Despite the failure to synthesize dolomite in microbial experiments, the low-temperature laboratory results remain interesting. High-temperature (60–300°C) dolomitization experiments have long shown that ordered dolomite is invariably preceded by disordered VHMC precursors that recrystallize to dolomite over time. Although recrystallization from VHMC to ordered dolomite has not been documented in the low-temperature microbial experiments, it may be common in natural settings where higher surface temperatures and longer time periods overcome kinetic barriers to dolomite formation. Mineralogical arguments aside, petrological observations show that VHMC products from microbial laboratory experiments are dissimilar to both natural dolomites and high-temperature synthetic dolomites. First, the published microbial experiments produced VHMC or other carbonates as cements via direct precipitation from solution rather than by replacement of a CaCO3 precursor, whereas the latter is demonstrated in high-temperature synthetic dolomites and inferred for most natural dolomites. Second, these precipitates tend to be spheroidal and/or dumbbell shaped, and as such they are fundamentally different from both the dominant rhombohedral form and the mimetic replacement textures observed in natural and high-temperature synthetic dolomites. Thus, the microbial products are not only mineralogically unlike natural dolomites, they also differ with respect to their mode of formation and their morphological characteristics.

Abstract Numerous field and laboratory studies over the past two decades claim that microbes catalyze nucleation and growth of dolomite at temperatures common in low-temperature geologic environments (25–60° C). However, a critical reexamination of the X-ray diffraction (XRD) data presented by these studies indicates that the laboratory products are not dolomite but rather a mixture of minerals, including very high-magnesium calcite (VHMC). Because VHMC can be “compositionally” indistinguishable from dolomite (i.e., 50 mol% MgCO 3 ), the positions of the principal (104) XRD reflection for VHMC and dolomite can be identical. Nevertheless, published XRD patterns of products derived from microbial experiments lack convincing evidence of cation ordering, which is a unique characteristic of carbonate minerals exhibiting R 3 (dolomite) symmetry. The lack of cation ordering in laboratory precipitates instead indicates that the products are VHMC, which possesses R 3 c (calcite) symmetry. Hence, previous laboratory studies have misidentified VHMC for dolomite. Despite the failure to synthesize dolomite in microbial experiments, the low-temperature laboratory results remain interesting. High-temperature (60–300°C) dolomitization experiments have long shown that ordered dolomite is invariably preceded by disordered VHMC precursors that recrystallize to dolomite over time. Although recrystallization from VHMC to ordered dolomite has not been documented in the low-temperature microbial experiments, it may be common in natural settings where higher surface temperatures and longer time periods overcome kinetic barriers to dolomite formation. Mineralogical arguments aside, petrological observations show that VHMC products from microbial laboratory experiments are dissimilar to both natural dolomites and high-temperature synthetic dolomites. First, the published microbial experiments produced VHMC or other carbonates as cements via direct precipitation from solution rather than by replacement of a CaCO 3 precursor, whereas the latter is demonstrated in high-temperature synthetic dolomites and inferred for most natural dolomites. Second, these precipitates tend to be spheroidal and/or dumbbell shaped, and as such they are fundamentally different from both the dominant rhombohedral form and the mimetic replacement textures observed in natural and high-temperature synthetic dolomites. Thus, the microbial products are not only mineralogically unlike natural dolomites, they also differ with respect to their mode of formation and their morphological characteristics.

Abstract The majority of carbonate reservoir rocks have been developed using conventional development schemes, owing to the presence of macropores that are the product of depositional textures modified by diagenesis. Carbonate reservoir heterogeneity is complex, due to ternary porosity distributions composed of matrix, vugs, and fractures. Recently, matrix-related microporosity has been recognized as an important control on storage capacity and hydraulic conductivity of hydrocarbons. With the advancement of completion technologies for low-permeability reservoirs, quantifying the matrix-related microporosity, understanding pore size and pore throat distributions has become increasingly important. Matrix porosity contribution is often overshadowed by the relative contribution from vugs and fractures, yet it is the matrix pore network that effectively “feeds” the vugs and fractures. The main focus of this research has been on carbonate reservoir mudrocks that lack macropores and contain pores that are less than a micrometer in size. Examples come from both conventional mudrocks from the Arabian Peninsula and unconventional mudrocks from the Bakken-Three Forks reservoirs of the Williston Basin. These mudrocks have porosities that range from <5% to >20%, and permeabilities that are most commonly ≪1 mD. Porosity is quantitatively estimated by petrographic image analysis and QEMSCAN® analysis. Estimated porosities are compared with measured porosity from a Core Measurement System (CMS)® 300 automated permeameter. Porosity and pore throat distributions are determined by mercury porosimetry and nitrogen gas adsorption experiments to capture both micropore and nanopore distributions. Results show distinct differences in porosity, permeability, surface area, and tortuosity among different facies. Pore size distributions indicate bimodal pore systems that are in the microporosity to nanoporosity range and that vary across different lithofacies. These variations are related to subtle differences in physical rock properties. Effective fluid flow requires a significant volume of larger micropores to access and connect nanopores.

Abstract The Great American Bank, deposited along the Cambro-Ordovician coast of Laurentia, consists of over 3000 m of carbonate deposits. Middle Cambrian Ledger Formation dolomitized ooid shoals and partially dolomitized microbial reefs, exposed in the Magnesita Refractories quarry (York, Pennsylvania), formed on the shelf margin [de Wet, Dickson, Wood, Gaswirth, Frey, 1999 , “A new type of shelf margin deposit: rigid microbial sheets and unconsolidated grainstones riddled with meter-scale cavities,” Sedimentary Geology vol. 128, pp. 13-21; de Wet, Frey, Gaswirth, Mora, Rahnis, Bruno, 2004 , “Origin of meter-scale submarine cavities and herringbone calcite cement in a Cambrian microbial reef, Ledger Formation (USA),” Journal of Sedimentary Research vol. 74, pp. 914-923; de Wet, Hopkins, Rahnis, Murphy, Dvortetsky, 2012 , “High energy shelf margin carbonate facies: microbial sheet reefs, ooid shoals, and intraclast grainstones: Ledger Fm. (Middle Cambrian), Pennsylvania.” In Derby, Fritz, Longacre, Morgan, Sternbach (Editors), The Great American Carbonate Bank: The Geology and Economic Resources of the Cambrian–Ordovician Sauk Megasequence of Laurentia , Memoir 98: American Association of Petroleum Geologists, Tulsa, Oklahoma, p. 245a–251a (extended abstract) and p. 421–450]. This depositional setting meant that Ledger strata were well positioned to be bathed in updip migrating burial fluids during shallow burial. Shallow subsurface fluids precipitated dolomite with different types of textural preservation: fabric retentive (mimetic) and fabric obscuring. Ooid shoals were pervasively dolomitized due to high primary porosity and permeability, but, because adjacent microbial reefs were syndepositionally cemented, they formed local aquitards that funneled porefluids into shoal and grainstone channel deposits. Local to regional faulting, associated with Paleozoic burial, created additional permeable conduits for dolomitizing fluids to infiltrate reef strata. Both fabric retentive and fabric obscuring dolomite types have overlapping geochemical and δ 18 O and δ 13 C signatures, interpreted as representative of a single porefluid origin. The similarity in isotopic and geochemical results between the calcite reef rocks and dolomites suggests that the porewaters were primarily buried marine water, mixed with Mg 2+ enriched porewater associated with diagenetic stabilization of high-Mg calcite to low-Mg calcite. The δ 18 O values provide evidence that diagenetic porewaters re-equilibrated with higher temperature burial fluids, but δ 13 C values and trace elements reflect at least a partial Cambrian seawater signature. Primary fabric preservation is interpreted as a function of the rate of dolomitization rather than different porefluid compositions. Baroque (saddle) dolomite cements precipitated from higher temperature fluids associated with deeper burial. Mesozoic uplift and regional faulting accompanied Pangean rifting, producing a karst Cambrian-Triassic unconformity. Ledger Formation deposits are patchily dedolomitized, forming coarse calcite lenses, typically containing red Triassic sediment. Petrographic and geochemical data (trace elements and stable isotopes) show that the diagenetic fluids responsible for dedolomitization were primarily low temperature meteoric waters associated with karstification.

Abstract The Niagara-Lower Salina reef complex reservoirs of the Michigan Basin host significant hydrocarbon volumes and have recently been identified as promising targets for enhanced oil recovery and carbon sequestration. Although these carbonate buildups have been studied extensively since the late 1960s, there is still wide uncertainty and disagreement concerning their morphology and internal stratigraphic and facies architecture. The prevailing paradigm depicts the reef complexes as tall, symmetric “pinnacles” with heterogeneous internal facies distributions that are patchy and unpredictable. The current study challenges this model of the reefs by examining four Silurian reef reservoirs with abundant core and petrophysical wire-line logs. New and existing subsurface data show that Silurian reefs in the Michigan Basin are highly asymmetric with internal facies distribution patterns that are strongly influenced by east-northeast paleowind direction. Six major depositional environments are identified during the main stage of reef complex growth based on sedimentological characteristics observed in core, as well as the vertical progression (stacking) of facies observed both in core and wire-line log signatures. A central reef core environment is identified based on interspersed coral-stromatoporoid boundstone and skeletal wackestone facies consisting of frame-building organisms such as tabulate corals and stromatoporoids, as well as intrareef faunal assemblages of bryozoans, brachiopods, crinoids, and rugose corals. Environments to the east (windward) of the central reef core are steeply inclined to the east (~40°) with narrow facies belts characterized by coarse reef talus. In contrast, environments to the west (leeward) of the central reef core have shallower slopes that dip to the west (< 15°) and are characterized by wide facies belts composed of carbonate mud and skeletal debris that become finer and thinner in the leeward direction. Application of this new Silurian reef model to reef complexes throughout the basin demonstrates remarkable consistency with respect to the overall asymmetric shape of the reef complexes, as well as the windward-leeward internal facies architecture. The asymmetric architecture and windward-leeward facies distribution patterns described in the new model offer a significant improvement upon preexisting models for Silurian reefs in the Michigan Basin and more accurately reflect our modern understanding of how environmental controls affect reef development and architecture. Furthermore, this new reef model can be used to more accurately predict the shape and internal facies distributions for other Silurian reef complex reservoirs within the Michigan Basin, particularly those that lack abundant well control.

Abstract The aim of this report is to describe a Middle Devonian dolostone reservoir in western Canada that has produced more than 57 million barrels of oil and water from moldic-pore reservoir facies with low permeability. The Slave Point Formation consists of six carbonate depositional facies, the relative proportions of which change in response to location on the basement paleotopographic surface. The most significant porosity in Slave Point dolostones is moldic porosity that formed by leaching of fossil fragments; not all Slave Point facies contain fossils. The distribution of fossiliferous carbonate facies, and moldic pores, is ultimately controlled by basement paleotopography. The main conclusion is that there is not a good correlation between permeability and porosity in these rocks. Permeable zones are restricted to dolostones that have touching moldic pores—a constraint that has limited oil production to fossiliferous-carbonate belts that fringe Precambrian granite-basement highs. In contrast, expanses between basement knolls accumulated mostly carbonate mudstone. Mudstone lacks fossils and consequently does not contain moldic pores. The simplest description of this reservoir is that it contains pods of permeability that are surrounded by less-permeable rock, a description suited to many carbonate reservoirs. The desired impact of this work is to draw attention to methods for estimating the probability of finding more-permeable or less-permeable facies types at different locations within these types of reservoirs. This approach would lead to design of more efficient exploration and production programs in complex moldic-pore carbonate reservoirs. Both oil and water are produced from these dolostones. Understanding the effects of high fluid flow in high-permeability zones can enhance oil recovery and reduce water production from similar vuggy-dolostone reservoirs.

Abstract Following the early successes of subsurface platform and pinnacle reef exploration in central Alberta, the Upper Devonian Leduc Formation of eastern Alberta has over the last two decades seen sporadic exploration and production, as well as limited research. In the heavy oil belt south of Lloydminster, the uppermost several meters of the Leduc are oil bearing within erosional karst remnants formed by sub-Cretaceous erosion. The highs have been rendered producible by horizontal drilling and the presence of reservoir quality dolomite. The Leduc Formation in east-central Alberta is composed of an impressive 200 m platformal accumulation that typifies the rapid carbonate growth during the Frasnian Stage. From bottom to top, the Leduc becomes increasingly more restricted, as indicated partly by the decrease in skeletal carbonate and increase in restricted peritidal facies. The lower Leduc contains a series of prograding stromatoporoid reefal and interreefal facies, which grade upward into back-reef facies, then finally into restricted lagoonal and peritidal facies of the upper Leduc. Leduc members have been thoroughly correlated in the area, and the youngest two from which oil is produced were sampled across the region for petrographic and geochemical analyses. Contrary to most other models for dolomitization of the inner Leduc platform of the Western Canada Sedimentary Basin, we interpret pervasive dolomitization to have replaced the original mineralogy of the upper Leduc early, during shallow burial, as a result of intraformational brine generation and reflux within peritidal facies. The upper Leduc dolomites recrystallized with burial to produce nonferroan dolomites, with an average δ 18 O value of −4.1‰ Vienna Pee Dee Belemnite (VPDB), and an average δ 13 C value of +1.0‰ VPDB. The dolomites are nearly stoichiometric, with cloudy cores and in some samples, clear rims. After initial burial, laterally extensive pre-Cretaceous erosion created an approximate 260 million year gap between Upper Devonian carbonates and upper Lower Cretaceous siliciclastics. The proximity of the sub-Cretaceous Unconformity to the upper Leduc dolomites is interpreted to have led to local dedolomitization in some lithologies, yielding low-magnesium calcites with very negative, meteoric δ 18 O compositions. The erosion of the overlying seal exposed areas of the upper Leduc to replacement calcite associated with dedolomitization, and an influx of Cretaceous clays, both of which are a detriment to reservoir quality. Proximity of the sub-Cretaceous Unconformity to the Leduc reservoir could increase the likelihood that these processes took place. Resolving the position of the unconformity relative to the upper Leduc reservoir is thus a critical tool in predicting reservoir quality.

Abstract An innovative methodology for diagenesis characterization and quantification is presented. It includes different geostatistical modeling workflows applied to a partially dolomitized carbonate platform. The case study consists of a Lower Cretaceous (upper Aptian) shallow-water carbonate platform from the Basque–Cantabrian basin (northern Spain), in which a widespread burial dolomitization occurs. Previous studies at basin scale suggested that the flow of dolomitizing fluids through the carbonate succession was channeled by regional faults and that subsequently the dolomite distribution was partially controlled by depositional facies and their modifications after early meteoric diagenesis. Here, at reservoir scale, several carbonate facies were differentiated and grouped in five depositional environments. Two depositional sequences corresponding to transgressive–regressive cycles and three stages of the platform evolution were distinguished. The statistical data treatment indicated that the dolomitization is mainly concentrated in the regressive part of the first sequence, corresponding to the second stage of the platform evolution. The most dolomitized environments are the inner platforms and the shoal. Facies from these shallower/proximal depositional environments were more exposed to early meteoric diagenesis, possibly controlling later dolomitization. The total macroscopic porosity is directly proportional to the degree of dolomitization: pores are most abundant in fully dolomitized portions of the succession, particularly in the rudist-bearing and grain-dominated facies. Abundant aragonitic shells (rudists, corals), easily leached or recrystallized during early meteoric diagenesis, could justify the higher moldic porosity in these facies. For geostatistical modeling purposes, several statistical rules were elaborated in order to associate to each depositional environment, in each of the three platform stages, different proportions of dolomitization and related pore abundance. A direct simulation of the distribution of depositional environments, degree of dolomitization, and pore abundance was achieved using a bi-plurigaussian simulation (PGS) algorithm. A nested-PGS algorithm was used to simulate the same parameters independently: dolomite and pore abundance were distributed within each depositional environment, based on the statistical rules previously defined. These simulations allowed three-dimensional (3D) visualization of the original depositional facies and textures affecting the distribution of dolomitization and pore abundance. Modeling using both bi-PGS and nested simulations accounted for the 3D dolomite body extension: the dolomitized succession is thicker in the north and thins toward the south, in agreement with evidence from mapping of the dolomite geobodies.

Abstract The aim of this study is to evaluate what seismic attributes are best able to highlight porous non-stratabound dolostone geobodies set in low porosity limestone. For this purpose three dolostone geobody volume scenarios were defined using outcrop based three-dimensional models to define the range of dimensions of dolostone geobodies and their association with particular fault populations. Three porosity scenarios were created using a global compilation to assign porosities to three lithologies: host limestone, bulk dolostone geobodies, and dolostone geobodies adjacent to faults. The combination of porosity and geobody volume scenarios yielded nine non-stratabound dolostone geobody scenarios. These include models in which the properties of near-fault dolostones were enhanced or degraded relative to the bulk dolostone geobody values. This allows for the effects of processes such as overdolomitization or dissolution to be implicitly explored, since those processes can degrade or enhance near-fault properties such as porosity, although in all scenarios dolostone porosities are greater than host limestone porosity. Density and compressional velocity ( V p ) were assigned to the scenarios based on a global compilation of the density, porosity, and V p in limestones and dolostones to allow for the calculation of acoustic impedance volumes that are representative of the range of values that could exist at depth. Synthetic seismic cubes and a suite of 14 seismic attributes were generated for each of the nine dolostone scenarios. Each attribute response was evaluated for its potential to highlight porous non-stratabound dolostone geobodies. Attributes that are most sensitive to lateral changes in acoustic properties rank the highest in the evaluation, followed by amplitude attributes, followed in turn by frequency attributes. Continuity attributes rank poorly in this example because fault offset is relatively small and is obscured by dolomitization.

Abstract The Lower-Middle Ordovician carbonate reservoirs have been recognized as the most important strata in terms of oil and gas exploration in NW Tazhong Uplift, Tarim Basin, China. Intercrystalline porosity, vugs, and fractures were preferentially developed in the dolostone intervals, and understanding the processes of dolomitization is therefore crucial for the prediction of the connectivity and spatial distribution of reservoir quality. Two major types of replacement dolomite (RD1 and RD2) were identified based on fabrics and textures: RD1 dolomite is light gray in hand specimen with porous sucrosic texture and partially preserved precursor lithologic texture. Microscopically, it consists of fine- to medium-crystalline (50-250 μm), euhedral to subhedral dolomite crystals with planar crystalline boundaries. RD1 dolomite has δ 18 O values between −7.0 and −3.2‰ relative to Vienna Peedee belemnite (VPDB; mean −5.9‰ VPDB), either similar to, or slightly higher than, values estimated from Early-Middle Ordovician marine dolomite. The δ 13 C values (−1.9 to −0.1‰ VPDB, mean −0.6‰ VPDB) and 87 Sr/ 86 Sr ratios (0.708021-0.708526, mean 0.708351) of RD1 dolomites fall within the range of coeval seawater values. These features suggest that RD1 probably formed from the reflux of slightly evaporitic (penesaline) seawater that did not reach the salinity required for gypsum precipitation during early diagenesis. RD2 dolomite is dark gray in cores, with the precursor texture completely obliterated. RD2 is microscopically composed of medium- to coarse-crystalline (150-500 μm), subhedral to anhedral crystals with nonplanar/irregular crystalline boundaries and weak undulatory extinction. This dolomite lacks intercrystalline porosity, but vugs and fractures are locally developed and partially filled by minor amounts of saddle dolomite cements, late-stage calcite cements, and fluorite. RD2 dolomite yields depleted oxygen isotopic compositions (−10.7 to −7.0‰, mean −9.1‰ VPDB), indicating that the dolomite formed at elevated temperatures during intermediate to deep burial. Sr isotopes analyses show that the dolomitizing fluids with respect to RD2 dolomite were mainly derived from connate seawater preserved in the Lower-Middle Ordovician carbonate formations. The lower Sr isotopic values (0.707193-0.707723) of some RD2 dolomites suggest mixing of small amounts of hydrothermal fluids derived from basic magmatic fluids that had lower 87 Sr/ 86 Sr ratios, whereas more radiogenic 87 Sr/ 86 Sr ratios (0.709028-0.714287) in RD2 dolomites might have resulted from burial recrystallization and/or were associated with dolomitizing fluids passing through the deeper Precambrian clastic rocks and basement.

Abstract Terminal Ediacaran metazoan reefs ( c. 548–541 Ma) can be locally substantial and the skeletal metazoans Cloudina riemkeae, C. hartmannae, Namacalathus and Namapoikia produced diverse reef types with complex ecologies in association with varied microbialite support or influence. In the Nama Group, Namibia, metazoan reefs grew in three associations with differing dominant frameworks: (1) monospecific aggregations of Cloudina ; (2) Cloudina–Namacalathus –thrombolite assemblages; and (3) thrombolite-dominated metazoan communities. Cloudina hartmannae formed monospecific reefs up to 7 m wide and 3 m high with no microbialite component. The synoptic relief was probably <1 m. Cloudina riemkeae formed densely aggregating assemblages associated with microbialites and thrombolites, each from 30 to 100 mm high, which successively colonized former generations. Isolated Namacalathus either intergrew with C. riemkeae or formed dense, low-relief, monospecific aggregations succeeding C. riemkeae frameworks. Thrombolite-dominated metazoan reefs reached up to 20 m in height and width, with a synoptic relief of up to several metres. Cloudina and Namacalathus grew closely associated with these framework thrombolites and Namapoikia , which was encrusting and modular, reached up to 1 m in size and occupied neptunian dykes and fissures. Cloudina and Namacalathus also grow cryptically, either as pendent aggregations from laminar crypt ceilings in microbial framework reefs, or as clusters associated with thrombolite attached to neptunian dyke walls.