Abstract This volume grew out of a symposium held at the 27th Annual Meeting of the Clay Minerals Society in Columbia, Missouri, during October, 1990. The symposium was designed to present a current synthesis of research devoted to the origin, diagenesis, and petrophysics of clay minerals in sandstones. Appropriately, this meeting of the Clay Minerals Society was held in honor of Walter D. Keller in his 90th year. Keller's pioneering work in sedimentary petrology, geochemistry, and clay mineralogy, including the application of scanning electron microscopy (e.g., Borst and Keller, 1969), has contributed much to our understanding of the origin of clay minerals in diverse rock types, including sandstones. This introductory paper presents a brief historical perspective of research that has molded our understanding of clay minerals in sandstones, and provides perspective for this collection of papers, which represents current trends in research activities.

Abstract: Authigenic illite generated during diagenetic events occurring in clastic reservoirs is datable by isotopic methods. Questionable analytical procedures adopted by the investigators can lead to doubtful meanings of the data. To avoid these ambiguities, precautionary steps should be systematically followed in selection, preparation, and characterization of the illite-enriched clay material. These steps include specific techniques, such as gentle disaggregation of the rocks instead of routine crushing, and extensive use of electron microscopy (TEM and SEM). Potential exists for 40 Ar/ 39 Ar-age determinations of these illites, which would avoid the time-consuming separation of the clay materials by size fractionation, but would not avoid its careful characterization. Basic tests for the systematics have still to be determined before routine application of this technique can be expected and investigator factor controlled.

Abstract: The oxygen isotope compositions of diagenetic minerals from sandstones and conglomerates have been determined for a Permian to Upper Cretaceous section through the Alberta deep basin. These data have been used to reconstruct variations in the δ 18 O values of pore water during diagenesis of intercalated sandstones and shales. The results confirm earlier observations on the oxygen isotope evolution of pore water in the western Canada sedimentary basin. First, meteoric water was abundant during early diagenesis, even in sediments deposited in shallow marine environments. Second, the maximum δ l8 O values attained by pore waters during burial diagenesis were generally < + 3‰, much lower than in other shale-dominated basins (e.g., Gulf Coast). However, pore waters in sandstones located adjacent to, or intercalated with, carbonates and shales of the underlying Paleozoic section had substantially higher δ 18 O values (+ 7 to +9‰) at or near maximum burial. Third, late diagenesis was dominated by low- 18 O meteoric waters. Saturation of the sedimentary section by meteoric water during early diagenetic processes probably resulted from infiltration during subaerial exposure associated with sea level fluctuations of the inland sea. Influx of meteoric water probably continued episodically throughout burial diagenesis. Contribution of meteoric water during early diagenesis is reflected in the low maximum porewater δ 18 O values that characterize the peak of burial diagenesis. These results also indicate that the smectite-to-illite reaction in Mesozoic shales did not dominate oxygen isotope evolution of sandstone pore waters in the western Canada sedimentary basin. Any increase in the δ 18 O value of pore water during burial probably resulted from variable mixing between early diagenetic, meteoric-dominated pore waters and 18 O-rich pore waters that had equilibrated with the underlying Paleozoic carbonate platform. The δ 18 O values of late diagenetic minerals and present pore waters reflect both this process and the subsequent, substantial influx of surface-derived meteoric water during post-Laramide uplift and erosion.

Abstract: The Middle Ordovician St. Peter Sandstone (dominantly quartzarenites and feldsarenites) occurs throughout the central Michigan Basin at depths ranging from 1.5 to 3.5 km and produces condensate and/or natural gas from several horizons in more than 36 fields. The primary mineralogy, textures and reservoir characteristics have been dramatically modified by a complex diagenesis. The main diagenetic features, generally observable throughout the basin, are, chronologically: early marine carbonate cement; quartz overgrowth and pressure solution; burial dolomite and anhydrite; dissolution of framework grains and early cements; and pervasive authigenic illite and chlorite cementation. This late clay cement commonly occurs in secondary porosity formed after the dolomite cement. A K-Ar study of the fine-grained authigenic illite indicates that this regionally significant episode of clay cementation in the Michigan Basin occurred during Late Devonian-Mississippian times. Sixteen samples from 11 wells located mainly in the central part of the basin yielded consistent ages ranging from 367 Ma to 327 Ma with an average of 346 ± 11 Ma. Combined with burial-history reconstructions for the central basin, these ages indicate that illite formed at depths of approximately 3 km. Fluid-inclusion temperatures for the dolomite and quartz cements associated with the diagenetic illite suggest temperatures of formation for the latter on the order of 150°C or higher. Combined with the depth estimate, these temperatures imply the existence of elevated geothermal gradients, i.e., 38°C/km or greater, in the central Michigan Basin at the time of illitization. Because illite and/or hydrocarbons typically fill secondary pores in the St. Peter Sandstone, the K-Ar age data also place constraints on the timing of development of secondary porosity and hydrocarbon emplacement.

Abstract: Model arkoses containing K-feldspar (or Na-feldspar) + kaolinite + quartz (or silica glass or boehmite) were reacted in solutions of a variety of compositions in the KCl-NaCl-H 2 O system, and in sea water at 200 to 350°C and 500–1,000 bars in rapid-quench, cold-seal pressure vessels. Experiments reveal that: (1) mixed-layer illite/smectite (I/S) with or without discrete illite are the major neoformed phases for all runs with low fluid/rock ratio; (2) the precipitation rate for neoformed clays is similar for near-neutral solutions of varying compositions, but the expandability of the I/S depends strongly on solution composition at the same temperature and pressure; (3) discrete illite appears only in solutions rich in K*; (4) sea water dramatically retards the formation of illite layers in I/S; (5) illitic clay with fibrous habit grows more effectively in solutions with silica oversaturated with respect to quartz; illitic clays formed in initially alkaline and acidic solutions tend to be platy; and (6) illitic clays form in near-neutral solutions at much slower rates than those in alkaline solutions. The experimental study verifies that the nonequilibrium mineral assemblage feldspar + kaolinite + quartz serves as a control for fluid composition, which favors precipitation of illitic clays in rock-dominant systems. The diagenetic products of a feldspar-bearing sandstone are determined primarily by the presence or absence of kaolinite and by the flow rate, and secondly by the initial fluid composition or temperature. The initial fluid composition becomes crucial only in the fluid-dominant system, whereas temperature is more important in controlling the kinetics than in shifting the thermodynamics of illitic-clay formation. The commonly observed illitization of kaolinite-bearing arkosic sandstones during burial diagenesis is attributed more to the reduction of the flow rate of the existing fluids than to the influx of new illitization fluids, or to the increase of temperature alone.

Abstract: The most abundant clay-mineral cements in North Sea reservoir sandstones are kaolinite, illite and mixed-layer minerals. Authigenic chlorite commonly is present but rarely is abundant. Permian and Triassic sandstones were deposited in an arid to semiarid climate and contain mostly illite/smectite with subordinate kaolinite. Isotopic and textural evidence suggests that authigenic kaolinite in these sandstones formed after tectonic uplift and meteoric-water flushing. In fluvial and shallow marine Lower and Middle Jurassic reservoirs, kaolinite is the dominant clay mineral and feldspar dissolution forms abundant secondary porosity even in the shallowest reservoirs. Stable-isotope analyses of authigenic kaolinite suggest crystallization at relatively low temperature and from pore water of meteoric origin. Upper Jurassic sandstones representing shallow marine facies also contain secondary porosity, due to feldspar leaching, and authigenic kaolinite. In turbiditic sandstones and sandstones interbedded with the main source rock (Kimmeridge Clay Formation), however, diagenetic kaolinite is rare. The sandstones also show little evidence of feldspar leaching, probably because this distal facies was not effectively leached by meteoric water. Organic acids or carbon dioxide that may have been released from source rocks seem to have had little effect on adjacent sandstones in terms of feldspar leaching and precipitation of kaolinite. Dissolved feldspar and authigenic kaolinite are also relatively uncommon in the Cretaceous and Tertiary turbiditic sandstones. The main control on feldspar leaching and distribution of authigenic kaolinite appears to have been the degree of meteoric-water flushing, which depends on climate, depositional environment and continuity of sandstone beds. Later tectonic uplifts may also have caused meteoric-water recharge from exposed areas into the basin. The geochemistry of formation water from reservoir rocks suggests that pore waters in North Sea reservoirs are mostly in the stability field of illite during burial diagenesis. In Jurassic reservoirs, illite can be observed to replace kaolinite in the deeper wells if K-feldspar is available as a source of potassium. A strong increase in the degree of illitization can be observed below 3.7 to 3.8 km burial depth in several oil fields. K/Ar dating of illites from reservoirs in the North Sea and Haltenbanken areas give a wide range of estimated ages, many between 30 and 50 Ma. Many of these sandstones would only have been buried to about 2 km during those early Tertiary times. If these dates for illite precipitation are correct, they might indicate high geothermal gradients at that time. The present depth distribution of illite suggests, however, that illitization of kaolinite takes place at greater depth (3.5–4 km). illite may also form from a smectite precursor and illite of this origin is particularly abundant in Triassic and Permian reservoirs. Chlorite may replace kaolinite starting at about 90 to 100°C, but the amount is limited by the supply of iron and magnesium from dissolving mafic minerals and rock fragments. Co-existing illites and chlorites appear to follow regular compositional trends with increasing burial temperatures. These trends are in covariance with the stability of the endmember components of the clay minerals. Crystal-size distributions may be explained in terms of Ostwald-ripening mechanisms. Chlorite crystals show compositional variations from core to rim. Burial diagenesis in the North Sea basin is interpreted to be relatively isochemical and our diagenetic models do not require large-scale transport of solids in solution. Theoretical models of corapactional poreawater flow and observed compositional stratifications of the pore water in this basin also constrain large-scale mass transfer by advection.

Abstract: Jurassic plays in the North Sea are sandstones deposited in fluvial, shallow marine or submarine-fan environments located in the footwalls or hanging walls of rotated fault blocks. After two decades of exploration, a large data base is available on the diagenesis of Middle Jurassic Brent and Upper Jurassic Piper-Claymore-Brae-Fulmar plays. There is still no consensus, however, on whether the diagenesis of these sandstones is dominated by meteoric-water flushing or by the influence of thermobaric waters released during burial. Diagenetic assemblages in mudstones and sandstones of the basin-margin sequences currently exposed on the mainland UK can be used as modern analogs for the inferred Mesozoic subaerial exposure of rotated fault-block crests in the North Sea. In this respect, mudstones are particularly sensitive indicators of subaerial exposure and rapidly develop soil profiles. Back-scatter electron microscopic (BSEM) observation of mudstones from outcrops indicates that kaolinite is largely authigenic. Basin-margin sandstones commonly preserve the early diagenetic clay-mineral assemblage (kaolinite, smectite and chlorites) but are generally dominated by vermiform authigenic kaolinite. Analytical transmission electron microscopy (ATEM) analyses of these clay minerals indicate large variations in their chemistries. Each Jurassic reservoir in the North Sea is characterized by a particular diagenetic-mineral assemblage, of which authigenic clays constitute an important component. Early clays in sandstones may coexist with later authigenic clays, although the earlier formed clays are often replaced. In general, authigenic smectites are present only in the shallower reservoirs; kaolinites typically occur at the crest of structures above 4 km depth and close to faults; illite is most abundant in the deeper reservoirs exceeding 4 km depth; authigenic chlorite is rare. Detailed petrographic observation, supported with SEM and TEM investigation, indicates that clay-mineral authigenesis was typically multiphase. K-Ar dating of illites supports this interpretation with youngest ages in the Tertiary. Oxygen isotope analyses of illites and kaolinites remain problematical. ATEM analysis of authigenic clays in mudstones and sandstones documents considerable uniformity of chemical composition at depth. The authigenic mineralogy evolves toward a widespread and uniform phengitic illitequartz-albite-ankerite assemblage at depths below 4 km. Authigenic clay-mineral assemblages support the argument for Mesozoic flushing with meteoric waters in some reservoirs but not in others. No single 'diagenetic model' can thus be applied to Jurassic North Sea fault-block plays; there is probably a spectrum of scenarios from reservoirs that have been extensively flushed with fresh water to those that have not experienced meteoric-water ingress. The potential for freshwater flushing depends upon a combination of depositional environment (fluvial and barrier-beach sandstones having greater potential than shallow and deep marine sandstones) and structural location (footwall crests have a greater potential for flushing than the hanging walls). Meteoric-water flushing is most likely where footwall erosion is greatest at the site of maximum displacement along a fault. Not all footwall crests, however, need have experienced subaerial exposure; the width of an individual fault block controls height of the scarp, whereas rate of erosion dictates fault-scarp persistence.

Abstract: Mass transfer of aluminum is investigated on a thin-section scale by comparing volumes of dissolved plagioclase and authigenic kaolinite in quartzofeldspathic sandstones from the San Joaquin Basin. Samples include Oligocene marine-shelf sandstones, which have been infiltrated by meteoric water, and Late Miocene turbidite sandstones, which contain diluted sea water. Other aluminum sources and sinks are volumetrically minor in these sandstones. Dissolved plagioclase and kaolinite presently appear from 600 m to depth of sample control (30–70°C present temperature) in the meteoric zone and from 2,100 m to depth of sample control (75–130°C present temperature) in the marine zone. Leached plagioclase and kaolinite are rare in the matrix-rich or carbonate-cemented sandstones, but appear in more than 80% of the uncemented turbidite sandstones and in up to 60% of the uncemented sandstones in some meteoric-zone reservoirs. Point-counted volumes of plagioclase porosity and kaolinite in all sandstones are compared with relative volumes calculated from a mass-balance reaction in which aluminum is conserved between An 30 plagioclase and kaolinite (25 to 50% microporosity). Aluminum is conserved on a centimeter scale in shale-encased turbidite sandstones exposed to limited fluxes of marine pore water, despite enrichment in organic-acid anions, which potentially may mobilize aluminum in soluble complexes. The average marine-zone sandstone has a volume of kaolinite approximately equal to that calculated for plagioclase porosity, based on relative volumes of the mass-balance reaction. In contrast, the average sandstone with meteoric pore water has a kaolinite shortfall of 0.4 ± 0.3 volume percent of total rock relative to plagioclase porosity. This average aluminum loss is 0.2 ± 0.1 gm/100 cm 3 rock volume. Complementary zones of aluminum import are not found in the meteoric zone. A small amount of aluminum is mobilized beyond a centimeter scale in shelf sandstones flushed by low-temperature, dilute waters. Plagioclase dissolution and kaolinite precipitation in sandstones from both porewater settings result in compositional shifts of less than 0.4 weight percent Al 2 O 3 , too small to discriminate from the natural bulk chemical variation of 1 to 2 weight percent Al 2 O 3 . Data from this study do not support models proposed for transfer of large masses of aluminum over significant distances.

Abstract: Twenty-nine sets of sandstone and stratigraphically adjacent, fine-grained, organic-rich samples representing diverse depositiona! environments (fluvial, coastal, paludal, and marine) have been obtained from cores of the Mesaverde Group within the Piceance Basin, Colorado. These units have been buried to a sufficient depth that organic maturation has progressed to an advanced state (vitrinite-reflectance values range from 1.0 to 2.2 in the study area). The authigenic-mineral paragenesis and organic data are integrated with burial and thermal-history modeling to place diagenetic events and hydrocarbon generation into a temporal framework. Three phases of diagenesis (early, late, and post-hydrocarbon) are characterized on the basis of petrography, XRD, and geochemistry of authigenic phases. Clay-mineral distributions of sandstone/mudstone pairs indicate that mixed-layer illite/smectite dominates mudstone mineralogy, although subordinate amounts of chlorite are observed in the marine interval. Sandstone mineralogy also includes mixed-layer clays and chlorite, with the addition of kaolinite. Additional aspects of sandstone diagenesis examined include feldspar albitization and dissolution, and carbonate mineralogy. Organic analyses indicate the presence of a type III kerogen component in all samples. The paludal interval contains the most oxygen-rich organic material and exhibits the highest sandstone porosities. Water-soluble organic compounds released from organic material as burial progresses have been invoked as agents affecting the course of mineral diagenesis in clastic sediments and sedimentary rocks. Our results indicate the potential for the field evaluation of the effects of thermal maturation of organic matter on the diagenesis of closely adjacent sandstones.

Abstract: Two cores from the Permian (Guadalupian) Yates Formation in Winkler County, Texas, were analyzed using thin-section petrography, scanning electron microscopy/energy dispersive X-ray, stable-isotope geochemistry, and 40 Ar/ 39 Ar laser step heating. The Yates was deposited in a coastal-sabkha environment. The sandstone facies is the hydrocarbon reservoir; the dolostone and anhydrite facies are impermeable. Sandstones are very fine- to fine-grained arkoses, subarkoses and lithic arkoses. The major authigenic phases in the sandstones are corrensite and dolomite. Mg-rich sabkha-based pore fluids were responsible for their formation. Other modifications to the sandstones include K-feldspar and quartz overgrowths, and unstable-grain dissolution. The clay-mineral suite in the Yates Formation consists of corrensite, illite and chlorite. Corrensite (Rl ordered, trioctahedral chlorite/smectite, with approximately 50% smectite layers) is found in the three major facies; however, it is most prominent in the sandstones and is absent from the interbedded black shales. Chlorite-to-smectite ratios do not vary with changes in stratigraphic position or lithology. Clay-mineral suites in dolostone and anhydrite layers, where present, are similar to those in the sandstones. Illite is more prevalent in dolostones than in sandstones, however. Elemental analysis indicates that Mg is a major component in the corrensite, although Fe is also present. The persistence of corrensite in the sandstones is interpreted as the result of relatively uniform porewater salinity and Mg levels. Stable-isotope values were determined on dolostones and dolomite cement in sandstones. In well #269, dolostone values range from δ 13 C +4.97 to +5.94%o and δ 18 O from –1.74 to +1.94‰ PDB. In well #270, δ 13 C ranges from +4.46 to +5.97‰ and δ 18 O from +0.15 to +2.39‰. Dolomite cements in sandstones from well #269 range from δ 13 C +0.67 to +4.92‰ and δ 18 O from –1.68 to +0.91‰. In well #270, δ 13 C ranges from +0.97 to +3.26‰ and δ 18 O from –6.09 to +1.23‰. The 40 Ar/ 39 Ar laser step-heating method was used to determine absolute ages on three clay-size mineral separates. In these samples, the clay-mineral suite consists of mixtures of corrensite, illite, and chlorite. Feldspars and mica are also present. The analyses revealed a low-retentivity phase with an apparent age of 275 to 250 Ma, which may represent the age of corrensite formation.

Abstract: Microcrystalline intergranular material of varying compositions occurs as well-developed geopetal structures, cutans, and massive pore fillings within subsurface samples of the fluvial, volcanogenic sandstones of the Comodoro Rivadavia Formation. These intergranular materials occur either as mixtures of chlorite, smectite and iron oxides, or as mixtures thereof that contain abundant quartz. Electron microprobe analyses of the material show that the quartz-rich material contains from 71 to 91% SiO 2 . The textures are interpreted to be the result of infiltration. It is suggested that the microcrystalline quartz-rich material is the alteration product of originally fine-grained volcanic glass, which was infiltrated into the sands from suspended loads of the "Comodoro Rivadavia" rivers during the Cretaceous. This inclusion of fine-grained glassy material is attributed to the erosion of airfall ashes that intermittently blanketed the drainage basin. The formation of the quartz-rich material requires both a siliceous source and additional silica derived from hydration reactions of adjacent volcanogenic sediment. The erosion of volcanic-ash deposits provides for intermittent, concentrated, suspended loads, resulting in a variable source of solutes during early burial. These variations of suspended-load composition are potentially overlooked features of arid and semiarid floodplain deposits, which are proximal to contemporaneous volcanism (e.g., rift, intermontane, forearc, and intraarc settings). Infiltrated glassy materials are a potential source of compositional and textural heterogeneity in volcanogenic sandstones that influences the path of later diagenetic alteration and the change of permeability and porosity of those sandstones.

Abstract: The clay-mineral assemblage of Middle Miocene sediments from the E2.0 Reservoir (3,590–3,655 m) in the Kolo Creek field is correlated with the color of palynomorphs in order to understand the level of diagenesis experienced by the sediments. The clay minerals in the mudstones are predominantly detrital kaolinite with significant amounts of illite/smectite, whereas in the sandstones, clays occur in minor amounts as mainly authigenic kaolinite. Quantitative measurements of the color of palynomorphs show that the sediments are immature to marginally mature and this level of maturity can be related to early diagenetic conditions before the emplacement of oil in the reservoir. A plot of palynomorph carbonization (luminance) versus percent illite in illite/smectite shows a fairly good correlation (statistically significant at a 95% level of confidence) between the two diagenetic parameters. The low level of maturity is not obvious from Kubler's illite crystallinity index for the mudrocks, which show a considerable variation over a short depth range.

Abstract: The Oligocene Sespe Formation has undergone differential burial resulting from complex tectonic evolution of the Ventura Basin. Samples collected from more than 5 km of burial depth allow documentation of the origin and depth-dependent diagenesis of clay minerals in texturally variable sandstones. Clay minerals can be distinguished as depositional matrix, infiltrated clays, or authigenic clays on the basis of pétrographie and SEM criteria. "Clean" (<20% ductile components) and "dirty" (>20% ductile components) sandstones reacted differently during burial diagenesis, resulting in contrasting clay-mineral assemblages. In "clean" sandstones, the clay-mineral assemblage changes from smectite plus subordinate kaolinite in shallow samples to chlorite in deep samples, suggesting that the major clay-mineral reaction is the transformation of smectite to chlorite. In "dirty" sandstones, the clay-mineral assemblage changes from smectite plus minor kaolinite at shallow depths, to smectite plus chlorite at intermediate depths, to mixed-layer illite/smectite plus minor chlorite at greatest depths. Formation of illite at the expense of smectite appears to be the predominant clay-mineral reaction. The origin of chlorite and the role that it plays in illite formation are unclear. In both types of sandstones, the presence of kaolinite in shallow samples may reflect meteoric diagenesis following tectonic uplift. Original texture significantly influenced clay-mineral diagenesis in Sespe sandstones. Low-permeability "dirty" sandstones behaved as closed diagenetic systems, whereas higher permeability "clean" sandstones behaved as relatively open diagenetic systems.

Abstract: The Sergi Formation, a Jurassic pre-rift sequence composed mostly of fluvial sandstones, is one of the major hydrocarbon reservoirs of the Recôncavo Basin in northeastern Brazil. Interstitial clays are important components of sandstones and exert significant control on reservoir properties, including permeability, irreducible water saturation and residual-oil saturation. These clays can be grouped into three types: (1) depositional clays; (2) mechanically infiltrated (MI) clays; and (3) authigenic (neoformed) clays. Each type shows a characteristic petrographic aspect that permits recognition and quantification using thin sections. Depositional clays were incorporated into the rocks as mud intraclasts resulting from reworking of overbank fines by fluvial processes. Early mechanical compaction crushed the mud clasts among more rigid grains, forming a compaction matrix. Mechanically infiltrated (MI) clays occur chiefly as coatings of tangentially accreted particles (cutans) or, locally, as complete pore fills. MI clays appear to be concentrated within the upper part of the formation. These clays can modify the pore geometry of sandstones. Shrinkage porosity, developed by diagenetic transformation of clays, is the dominant porosity type in the upper part of the Sergi Formation. Authigenic clays are kaolinite and chlorite. Kaolinite occurs as pore fills in large secondary pores and, where present in large amounts, may generate high microporosity in the reservoirs. Chlorite occurs as pore linings and, locally, as pore fills. In the reservoirs, chlorite causes permeability reduction and is related to the presence of low resistivity in water-free, oil-producing zones. These authigenic clays show a distinct distribution within the basin. Kaolinite dominates in the western portion, where Sergi reservoirs are found at shallow depths (above 1,000 m), whereas chlorite is dominant in the eastern portion, where Sergi reservoirs are found at greater depths. The distribution of these clay minerals is the result of differences in burial/temperature histories, which are still reflected by present depths.

Abstract: Inherited grain-rimming clays are common in sandstones from eolian dune and marine-shelf environments. Inherited clay rims are defined as clay coats that form on framework grains prior to their deposition. Development of such clays requires that these clays become attached to framework grains at some other location and that they then be recycled with the host grain to form the present deposit. Such clays have been recognized in a number of major eolian-reservoir units in North America and the North Sea. Inherited clay rims also are extensive in major shelf-sandstone reservoirs. The relatively high levels of reservoir quality of many of these dune and shelf sandstones are at least partially attributable to the presence of inherited clay rims. The distribution and composition of inherited clay rims allow them to be distinguished from other types of detrital clays and from neoformed clays. Characteristics most useful in identifying inherited clay rims include: (1) presence at points of contact between framework grains; (2) widely varying rim thickness; (3) increased thickness in embayments; (4) absence on the surfaces of diagenetic components; (5) clay mineralogy similar to that of clay interbeds; (6) enhanced development in finer grained sandstones where clay-filled depressions are more abundant; and (7) tendency to be developed and preserved only in selected environments. Recognition can be difficult where significant regeneration has occurred, where rims are extremely thin, or where rims are masked by detrital matrix or by authigenic cements. In eolian settings, clay rims initially form in sabkha and interdune environments. The coatings are interpreted to form by infiltration of clay-charged waters or by adhesion to wetted sand-grain surfaces in eolian-soil and sabkha environments. Mild reworking allows for preservation of complete clay rims, whereas extended transport results in the removal of rims from projections on grain surfaces. Inherited clay rims are absent in water-washed environments or in coastal dunes derived directly from beach sands. In shelf environments, inherited clay rims are generated when sand grains passing through the digestive tracts of organisms, or disturbed by burrowing, become coated with clay. Later reworking may reduce the thickness of clay rims or remove them altogether. Clay rims are destroyed in high-energy nearshore environments. Because the presence of inherited clay rims can severely limit quartz-overgrowth development, high levels of intergranular porosity may be preserved to depths where associated clay-free sandstones exhibit extremely low porosities. Inherited clay rims are not entirely beneficial. The presence of clays at grain contacts may significantly accelerate the process of pressure-solution suturing. Commonly, clays in the rims are regenerated at depth to forms having a much higher surface area and containing a large amount of microporosity.

Abstract: Strata of the lower and middle Atoka Formation in the Arkoma Basin comprise submarine-fan and marine-slope facies that display a variety of primary and secondary sedimentary structures, formed by sediment gravity-flow depositional processes and dewatering, respectively. Primary sedimentary structures are most common in beds deposited by unconfined sediment gravity flows on submarine-fan lobes, whereas secondary sedimentary structures are most common in beds deposited by channelized sediment gravity flows in fan channels and slope channels. Primary sedimentary structures display horizontal fabrics, whereas secondary sedimentary structures display deformed and vertical fabrics. Abundance and distribution of clay minerals in Atoka sandstones are related to sedimentary structures. Beds that display primary sedimentary structures contain little detrital clay that is sparsely disseminated through the sandstone. In contrast, beds that display secondary sedimentary structures contain more detrital clay that forms pervasive grain coatings, bridges between grains, and consolidation laminae. Other beds lack sedimentary structures and display abundant detrital clay that forms a matrix-supported fabric. The abundance and distribution of detrital-clay minerals exerted significant influences on diagenesis and reservoir quality of Atoka sandstones. Among sandstones with grain-supported fabrics, those that display primary sedimentary structures and contain little detrita) clay were pervasively cemented by quartz overgrowths and are characterized by poor reservoir quality. Those that display secondary sedimentary structures and contain more abundant detrital clay retained primary porosity because quartz-overgrowth nucleation was inhibited by clay coatings on detrital grains. Porosity was enhanced in these sandstones by dissolution of framework grains, and the sandstones are characterized by good reservoir quality. Sandstones with matrix-supported fabrics apparently had little original porosity, which was reduced by compaction of the pervasive matrix; they are characterized by poor reservoir quality. These observations suggest that channelized turbidite facies have greater potential to retain good reservoir quality than unconfined turbidite facies, because the former have detrital-clay minerals emplaced within sand during dewatering and those clay minerals inhibit destruction of porosity by quartz cementation.

Abstract: Clay coats, which may be continuous or discontinuous, originate from soils as cutans, from infiltration of clay in sand and sandstone, and authigenically as newly formed or regenerated clay minerals. Allogenic cutans and infiltration deposits have a laminar morphology, whereas authigenic-clay coats commonly have a radial morphology. Thick, well-developed, continuous clay coats, regardless of origin, may retard quartz cementation by masking the surface of detrital-quartz grains and preventing the nucleation of quartz overgrowths. Chlorite is the most effective of the clay minerals in preserving intergranular porosity and appears to be important in very deep sandstone reservoirs. The most favorable amount of chlorite to preserve porosity is variable: 4 to 7 volume percent for the Berea Sandstone and 5 to 13 volume percent for the Tuscaloosa Sandstone, for example. Smaller amounts of chlorite permit quartz to nucleate and destroy porosity and greater amounts result in porosity reduction by infill of pores. Clay coats do not retard epitaxial cements (e.g., carbonates and sulfates), which may cover clay coats and occlude porosity. Clay coats may occur in highly lithic (e.g., >35% lithic material) sandstones, but are not important because physical compaction dominates diagenesis and destroys porosity. Experimental growth of clay coats shows that clay flakes are flatly attached to detrital-sand grains and curl upward to form a radial-fibrous morphology. This attached root zone may explain why clay coats are effective at blocking nucleation of quartz cement. Experimental work also shows that mineralogy may provide an initial substrate control over the precipitation of clay coats by providing an in situ source of the cations needed to precipitate the clay. Later, the clay coats nucleate on other framework grains farther from the site of initial nucleation.

Abstract: Quantitative models of the reduction of permeability in reservoir sandstones due to the growth of authigenic-clay minerals in the pore space are based on the ability to estimate the permeability of the original clay-free rock. Simple physical models based on Carman-Kozeny relations are used to calculate permeability for the idealized sandstone pore space. Values for the surface-area parameter in the models are determined from proton NMR longitudinal-relaxation times and area/perimeter ratios extracted by petrographic-image analysis. Although the magnitude of the difference between measured and calculated permeabilities is model dependent, the different models characterize relative behavior for each suite of sandstones. The normalized permeability differences correlate weakly with various measures of total clay abundance. This indicates that permeability reduction is influenced more by clay distribution than by clay abundance. Cation-exchange capacity (CEC) measurements made by flow through the intact rock are lower than values determined by standard methods on powders. As the ratio of flow to bulk CEC values decreases, fewer of the clays in the pore space are accessed by the fluid. Samples with increased fractal dimensions or surface roughness have lower CEC ratios, indicating that increased roughness limits the accessibility of exchange sites. Samples with lower fractal dimensions have more authigenic kaolinite than fibrous illite, in addition to greater differences in measured and calculated permeability. This suggests that physical constrictions caused by clay growth in the throats is more important than surface-roughness effects in reducing permeability in sandstones.

Abstract: Clay minerals in reservoir sandstones may react with drilling and completion fluids. Reactions are influenced also by clay mineralogy, clay form, and location of clay minerals. Smectite and mixed-layer illite/smectite clay minerals may swell to produce blockage, or occur as loose particles that migrate. Illite has a diverse morphology, but commonly occurs as a fibrous "spiderweb" form that may migrate when subjected to high-velocity flow, trap fine particles or water, or be virtually inert. Loose particles of kaolinite may migrate, but when water chemistry is controlled, this problem is minimized. Chlorite may serve as a trap for fine particles, as a habitat for microporosity, and has the potential to react with acids to create a ferric-hydroxide precipitate. Clays in high-permeability "channels," through which the fluid is flowing, are particularly prone to produce formation damage. Laboratory flow-test results show that there may not be a correlation of permeability damage from the flowing fluid with the quantity of clay, even authigenic clay, reported to occur in the rock. Scanning electron microscopy and X-ray diffractometry are not independent methods for predicting clay-reaction problems. However, these techniques are helpful, when combined with flow testing and interpretation experience, for predicting clay reactions.