Groundwater Flow, Late Cementation, and Petroleum Accumulation in the Permian Lyons Sandstone, Denver Basin1

Ming-Kuo Lee2 and Craig M. Bethke3


The gray diagenetic facies of the Permian Lyons Sandstone, which is associated with all known petroleum accumulations in the formation, formed late in the history of the Denver basin as an alteration product of the formation’s red facies. The red facies that makes up most of the sandstone contains iron oxide coatings, quartz overgrowths and calcite cements. The gray facies, which occurs locally in the deep basin, is distinguished by pore-filling dolomite and anhydrite cements and by a lack of iron oxide and calcite. The dolomite and anhydrite cements overlie bitumen that was deposited by migrating oil, and hence formed after oil was first generated in the basin, late in the Cretaceous or early in the Tertiary. The isotopic composition of oxygen in the dolomite ranges to such light values that the cement must have formed deep in the basin in the presence of meteoric water. The gray facies likely formed in a regime of groundwater flow resulting from Laramide uplift of the Front Range during the Tertiary. In our model,

saline groundwater flowed eastward through the Pennsylvanian Fountain Formation and then upwelled along the basin axis, where it discharged into the Lyons Sandstone. The saline water mixed with more dilute groundwater in the Lyons, driving a reaction that dissolved calcite and, by a common-ion effect, precipitated dolomite and anhydrite. The facies’ gray color resulted from reduction of ferric oxide in the presence of migrating oil or the Fountain brine. Underlying source beds by this time had begun to generate petroleum, which migrated by buoyancy into the Lyons. The association of the gray facies with petroleum accumulations can be explained if the Fountain brines discharged across aquitards along the same fractures that transmitted oil. As petroleum accumulated in the Lyons, the newly formed cements prevented continued migration, as is observed in shallower strata, by sealing oil into the reservoirs from which it is produced today.

©Copyright 1994. The American Association of Petroleum Geologists. All rights reserved. 1Manuscript received, August 5, 1992; revised manuscript received, September 13, 1993; final acceptance, September 24, 1993. 2Department of Geology, University of Illinois, Urbana, Illinois 61801. Present address: Department of Geology, Temple University, Philadelphia, Pennsylvania 19122. 3Department of Geology, University of Illinois, Urbana, Illinois 61801. We thank Roger Burtner of the Chevron Oil Field Research Company for supplying Lyons core samples, Keith Hackley and Jack Liu of the Illinois State Geological Survey for helping make isotopic measurements, and Richard Hay, Randy Cygan, Wendy Harrison, and Steve Johansen for advice and helpful discussions. We appreciate assistance in the laboratory from Tom Anderson, Steve Altaner, Jay Matthews, and Eric Daniels, and help in computer programming from the Hydrogeology staff. Jessie Knox and Joan Apperson drafted the figures. This study was supported by National Science Foundation grants EAR 85-52649 and EAR 8601178, and the generosity of Amoco Production Research, Arco Oil and Gas, British Petroleum Research, Chevron Oil Field Research, Conoco, Du Pont, Exxon Production Research, Illinois State Geological Survey, Japan National Oil Company, Lawrence Livermore National Laboratory, Marathon Oil Company, Mobil Research and Development, Sandia National Laboratory, Texaco, and Union Oil of California.

Geologists have long puzzled over the nature of diagenetic alteration in the Permian Lyons Sandstone of the Denver basin because the occurrence of diagenetic cements in the formation mirrors the distribution of petroleum reservoirs. Most of the Lyons consists of a red diagenetic facies. The red Lyons is familiar to many geologists because it crops out at the basin’s western margin on the Flatirons, a series of hogbacks along the Front Range of the Rocky Mountains, and because of its use as a distinctive building stone in the area. The facies, which is barren of petroleum, is characterized by hematite stain, clay minerals, quartz overgrowths, and calcite cements. A gray diagenetic facies occurs locally deep in the basin. The gray Lyons, distinguished by anhydrite and dolomite cements that overlie bitumen stains and by a lack of calcite and hematite, hosts all known oil fields in the formation. There has been no convincing explanation of the

AAPG Bulletin, V. 78, No. 2 (February 1994), P. 221-241.


Lyons Sandstone, Denver Basin

origin of the gray facies nor the reason for its association with petroleum accumulations. The anhydrite and dolomite cements of the gray facies might have formed soon after the Lyons was deposited, as suggested by the high intergranular volume preserved in the sandstone. Anhydrite occupies up to 25% and dolomite cements occupies up to 15% of the rock’s pre-cement pore volume. For this reason, Dimelow (1972) suggested that these cements might have precipitated from formation water deposited with the sediment. Dolomite and anhydrite might also have precipitated in an environment open to flow of saline groundwater. For example, they could have formed where evaporated seawater infiltrated carbonate platforms in sabkha environments (Butler, 1969; Patterson and Kinsman, 1977; Achauer, 1982); some geologists working in the Denver basin have suggested such an origin to us. Levandowski et al. (1973), on the basis of the cement distribution in the sandstone, argued that the cements precipitated when underlying strata compacted during burial, driving saline groundwater upwards through the Lyons. Anhydrite and dolomite cements of the gray Lyons Sandstone might also have formed in an environment open to regional groundwater flow, long after the formation was buried. Such an origin is not unusual: dolomite cements can form where fresh and saline groundwaters mix (Badiozamani, 1973; Land, 1973; Harrison, 1991), where migrating groundwater reacts with host sediments (Banner et al., 1988; Gregg and Shelton, 1989; Bethke and Marshak, 1990), and where carbon dioxide escapes from migrating brine (Leach et al., 1991). In this study, we use hydrologic and geochemical models to develop an explanation of the origin of Figure 1—Stratigraphic column for the Denver basin gray Lyons facies that is consistent with petrographic near the Front Range (after Weimer, 1973). observations as well as the stable isotopic compositions of the Lyons cements. The results of our modeling suggest a strong relationship between diagenetic (e.g., Hoyt, 1962; Martin, 1965; Izett, 1975; Trimble, alteration of the sandstone, the basin’s paleohydrolo- 1980; Sonnenberg and Weimer, 1981) . The Denver basin was a marine shelf that subsidgy, and the accumulation of petroleum in the Lyons. ed slowly through most of the Paleozoic and Mesozoic. Figure 1 shows the stratigraphic column near GEOLOGIC SETTING the Front Range. Pennsylvanian sediments consist mainly of carbonate and shale, although sandstones The present-day Denver basin is an elongate, predominate in the west along the uplift of the asymmetric structure whose axis runs north-south in ancestral Rocky Mountains. During the Permian, a Colorado and Wyoming, just to the east of the Front broad sea intermittently covered a surface of low Range uplift. The Laramide and Front Range uplifts relief where sediments, including the Lyons Sandbound the basin to the west. Basin strata extend stone, were deposited in environments ranging from eastward across parts of eastern Colorado, southeast- fluvial to normal marine to hypersaline. During the ern Wyoming, southwestern Nebraska, and north- Mesozoic, Triassic and Jurassic sediments consisting western Kansas. The deepest part of the basin lies mainly of sandy shale buried the Paleozoic section. near Denver, Colorado, where more than 4 km of The basin downwarped rapidly in the Cretaceous, sediments are preserved. Because of the basin’s when an interior seaway covered much of western hydrocarbon potential, its structure, stratigraphy, and North America. During this period, more than 3 km tectonic history have attracted considerable study of shale and shaly sandstone were deposited at the

but pinches out into carbonate and shale facies as it grades eastward. Orogeny continued into the Tertiary and reached its peak during the Eocene. The Lyons. Figure 3 shows the relationship of the Lyons Sandstone to aquitards and other aquifers within the basin. the thickest in the basin. (4) basal Cretaceous sandstones. a regional aquifer. is predominantly shale. is absent in our study area. Black Hollow. providing the main water supply for consumption and irrigation in the central Great Plains area. sandstone. Figure 4 shows that the zone of greatest cementation lies close to all known Lyons oil fields. For purposes of hydrologic analysis. Line AA' shows the cross section considered in this study. These sandstones grade eastward to dolomite and shale. The Pennsylvanian-Permian unit consists primarily of interbedded carbonates and shales. the Permian Lyons Sandstone lies between anhydritic siltstones. 1973). however. marine. In the western basin.. strata in our study area can be divided into five hydrostratigraphic units (Belitz and Bredehoeft. Walker and Harms. because of extensive secondary porosity.Lee and Bethke 219 basin’s depocenter. the basin assumed its present configuration. Fluvial. which constitute aquifers in the southern basin. serves as a supply of potable water where it is shallowly buried and as a petroleum resource in the deep basin. and New Windsor fields. (2) Permian-Triassic. Adams and Patton. DEPOSITION AND DIAGENESIS OF THE LYONS SANDSTONE Figure 2—Isopach and lithofacies map of Permian sediments in the Denver basin (after Martin. which runs from Big Thompson Canyon to the Keota oil field. Mississippian carbonates form a sixth unit that. 1965). which include the Dakota Sandstone. 1980. Blood. 1982. the area west of the presentday basin began to be uplifted. The unit acts as a regional aquitard. 1983). but sandstones predominate toward the basin’s western margin. Morse. With this final adjustment. 1970. The Cretaceous shale unit confines the aquifer system. This lowermost unit. 1979). constitutes an important aquifer system in the northern Great Plains (Downey. The Pennsylvanian Fountain Sandstone underlies the Lyons to the west. sandstones of the Fountain Formation lie at the base of the unit. Back et al. where the Lyons was deposited eastward about 150–200 km from the present-day position of the Front Range. 1980). and chalk. The Denver basin in the Permian contained two major subbasins where evaporites accumulated: the Alliance basin to the north and the Sterling basin to the south (Figure 2). The most recent uplift. 1981). This unit. and eolian processes transported sand to the coast (e. With the onset of the Laramide orogeny near the end of the Cretaceous. Interbedded red beds and evaporites dominate the strata of the Permian-Triassic unit. The basal Cretaceous sandstones. Uplift resumed during the middle Oligocene (Trimble. In eastern Colorado. Secondary porosity and permeability are less well developed in carbonates of this unit than they are in Mississippian limestones.. Two diagenetic facies of the Lyons (one red and the other creamy gray) can . caused about 450–600 m of Tertiary sediments to be eroded from the basin (Trimble. The overlying Triassic-Jurassic unit contains shale interbedded with local sandstone layers. 1976. 1988): (1) Pennsylvanian-Permian. and (5) Cretaceous shales. The Lyons Sandstone formed as a nearshore deposit along a band between the slightly emergent ancestral Rocky Mountains to the west and the evaporite basins to the east.. Lyons sediments have been extensively altered since deposition in a pattern suggesting that the process of diagenesis was closely related to oil accumulation (Levandowski et al. about 5–6 million years ago. (3) Triassic-Jurassic. but contains minor amounts of limestone. The basin tilted eastward as the Front Range began to emerge. Anhydrite-rich siltstones bounded the Lyons above and below. including Pierce. make up the most important aquifer system within the basin.g.

which is bitumen deposited by migrating petroleum.220 Lyons Sandstone. The section shows the relationship of the Lyons Sandstone to other hydrostratigraphic units within the Denver basin.. 1960). is distinguished (Levandowski et al. For this reason. The gray Lyons facies. Some calcite formed even earlier than the hematite.. covers quartz grains and their overgrowths. the anhydrite and dolomite cements must have formed after oil first migrated into the Lyons. 1973) by anhydrite and dolomite cements that overlie organic matter (Figure 5B). shows that the Lyons contains at least two generations of calcite cements that precipitated at different times. Textural relations in the facies suggest a general paragenetic sequence of hematite and clay overlain by secondary quartz and then calcite. Together. The red Lyons is nonpetroliferous. 1973). be distinguished by mineralogy and color. which is commonly cited as the first phase to form in the diagenetic history of the sandstone. (1973) and in this study. however. Open circles show locations of wells where samples were taken for isotopic analysis by Levandowski et al. Hematite and clay rim the grains of quartz sand and much of the pore space is filled by quartz overgrowths and calcite cement (Figure 5A). Figure 4—Relationship of areas of the most intensive anhydrite cementation (ruled pattern) to Lyons oil fields (solid pattern). The organic matter. the anhydrite and dolomite occupy as much as . Detailed petrographic study of samples from outcrops along the Front Range (Hubert. which is associated with oil fields. Denver Basin Figure 3—Cross section AA' (see Figure 2 for location) through the Denver basin from Big Thompson Canyon to the Keota oil field (after Levandowski et al.

in contrast. These values are significantly lower than would be expected in marine carbonates. whereas the anhydrite and dolomite overlie the quartz overgrowths and later bitumen. seems to have consumed little quartz. groundwaters with δ18O values in the range –26 to –13‰ would be required. or about 8% of the formation’s total volume. Alteration. Hematite (arrows) coats the original surfaces of quartz grains (Q) beneath overgrowths of secondary quartz (O). the alteration process must have consumed the calcite. (C) View under a scanning electron microscope at high magnification showing the pre-cement surface of a detrital quartz grain. . (B) Photomicrograph showing (under crossed nicols) the gray facies of the Lyons Sandstone. and suggest that the cements formed at depth in the basin. Sample 13398 from Horsetooth Reservoir. At 25°C. shows shallow solution pits indicating that quartz surfaces beneath the late cements were only slightly etched.2‰ relative to SMOW. The low oxygen ratios implicate meteoric water in the emplacement of the late dolomites. The photomicrograph of the detrital quartz grains (Figure 5C).Lee and Bethke 221 Figure 5—(A) Photomicrograph showing (under crossed nicols) the red facies of the Lyons Sandstone.. If the gray facies is an alteration product of the red facies. Organic matter coats quartz grains (Q) and overgrowths. made by scanning electron microscope after carbonate and sulfate in the samples had been chemically digested. Calcite cements are ubiquitous in the red Lyons facies. Scale bar = 100 µm. especially seawater concentrated by evaporation. but have yet to be observed in the gray facies. Calcite cementation almost certainly preceded deposition of anhydrite and dolomite because some calcite in the red facies underlies the early hematite cement (i. especially those samples strongly depleted in 18O. Using the equation of Northrop and Clayton (1966). Sample 14435-21 from the Troy 1 Jones well. The cements range in δ18O from +8. Hubert. and are difficult to reconcile with the likely compositions of meteoric water in the latitude of about 40°N. their results are shown in Table 1. ISOTOPIC COMPOSITIONS OF LATE CEMENTS Oxygen in Carbonate Cements Levandowski et al. 1960). Calcite cement (C) fills some pore spaces. The analyses did not distinguish among carbonate minerals. we can estimate at any temperature the fluid composi- tion that would be required to explain dolomite of a given composition. but the dominant carbonate mineral from gray facies samples is dolomite. Dolomite (D) and anhydrite cements (white arrows) fill pore spaces that are coated with organic matter (black arrows). which normally fall in the range +30 to +32‰ (Figure 6). Scale bar = 100 µm. Sample 13390-6 from the Calco 1 Ferch well. (1973) analyzed the δ18O and δ13C compositions of carbonate cements. which in turn underlies the quartz overgrowths. The quartz is lightly pitted. 40% of precement porosity.8 to +21.e. Such fluids are much lighter than any seawater.

5 –2. The 18O contents of the cements appear incompatible with an early diagenetic origin. mixing it with meteoric water having a δ18O value of about –13‰ would produce fluids spanning the required range in isotopic composition. ‡Outcrop samples. dolomite cements of the gray facies span a smaller range and are enriched in l3C compared to many of the calcite samples. The composition is also inconsistent with oxidation of organic matter. If a brine had a δ18O value of about 0‰.7 +8.222 Lyons Sandstone.y.1 +6. Isotopic Composition (‰) of Carbonate and Anhydrite Cements in the Lyons Sandstone Sample Number* Gray Facies 11326-11 13390-6 14385-200 14385-236 14431-14 14433-25 14435-21 14435-23 14441-13 Red Facies 13398 14674-6 13452-2 14883-23 14883-5 14875-1 14875-3 14875-7 Well Calco 1 Winder Calco 3 UPRR Calco 1 Ferch Calco 1 Ferch Calco 1 Brunner Priddy 4 Troy 1 Jones Troy 1 Jones Venable Fee-1 Horsetooth Reservoir‡ Owl Canyon Outcrop‡ Hummel 1 Allison Calco 1 Hayes Calco Hayes Nebraska Drillers Ball Nebraska Drillers Ball Nebraska Drillers Ball Depth (ft)** 8903 9073 9136 9206 8997 9219 9279 9235 8997 δ18OSMOW** — +9.5 +11. for example.5 +8.7 –2. Carbon in Carbonate Cements The small CO 2 contents of many groundwaters lead many investigators to assume that the δl3C composition of a carbonate cement simply reflects the δl3C of its precursor mineral (Mattes and Mountjoy. whereas the heavier l3C compositions suggest that these calcite samples could have formed from seawater or from isotopically heavy fluids in the deep basin.4 +6.4 +12.6 +12. Denver Basin Table 1.6 +9.5 +19.5 –3.9 +11. In the red facies. †This study. Especially in sandstone. Lander and Anderson. ††Sample contains insufficient anhydrite for analysis. The dolomite can be .9 –15. Because Tertiary rainfall in this area could have δ18O values down to –14‰. 1980.4 +15.2 –7.5 –3.1 –5. On the other hand. 1979). however. Formation temperatures of about 100°C would be expected in the Lyons Sandstone at present burial depths of about 3 km (Table 1). infiltration of meteoric water deep into the basin could explain even the lightest 18O compositions determined. Meyers and Lohmann.6 +11. This assumption may not be true. **Levandowski et al. where carbonate minerals occupy a small fraction of the rock volume.8 +16. and therefore.5 –4 0 –11.2 +21.7 +15. 1989). These samples likely represent calcite from different origins and generations. 1974.1 –1.3 –4. High-altitude runoff of the Rocky Mountains in the Tertiary and present-day have δ18O values as low as –14‰ (Taylor.3 –0. as already discussed.8 +9. 1985).4 δl3CPDB** — –1.5 +9.0 –16.2 +18. Groundwater having δ 18O values between –13 and about 0‰ would be required to precipitate dolomite at this temperature. preclude (barring the possibility of later isotopic reequilibration) the possibility that the cements formed in a sabkha environment.6 –3. carbonate dissolved in the fluid can represent a second important reservoir of carbon. assuming a normal geothermal gradient of about 30°C/km. The depleted δl3C values (< –20‰) might result from oxidation of organic matter.0 –26.4 +11.7 †† †† †† †† †† †† †† †† 7829–7834 9148–9149 9089–9090 9395–9399 9405–9410 9425–9430 *Numbered by Chevron Oil Field Research.5 +l2.6 +15. It is difficult to argue from these data that the dolomite reflects the carbon isotopic compositions of a calcite precursor. (1973). Since this part of the North America craton was not in far northern latitudes during the past 300 m.6 +11. (Irving. it is unlikely that δ18O values of meteoric waters in this area could range as low as –26‰.6 δ34SCDT† +11. if the sediment is open to flow of a groundwater charged with CO2. the δ l3 C of calcite cements ranges widely from –26 to –2‰ (Figure 6).5 +3. The scatter in the δ18O data (Figure 6) could arise from mixing in varying proportions of meteoric water with a basin brine.

1973). so any sulfate derived from the oxidation of sulfide would be depleted in 34S. but is depleted in 18O. 1969).Lee and Bethke 223 Figure 6—Isotopic compositions of calcite (open squares) and dolomite cements (solid squares) in the Lyons Sandstone. PETROLEUM GENERATION AND MIGRATION We analyzed the sulfur isotopic composition of anhydrite cements from nine Lyons core samples from various wells. which ranges from –4 to +4‰. Unlike overlying strata. anhydrite cements in the Lyons likely formed long after burial in an environment open to groundwater flow. even seawater evaporated to the point of anhydrite saturation would contain only about 14 g/kg of dissolved sulfate to account for 0. Dolomite falls near the normal range of δ13C for marine carbonate. most samples can be traced by geochemical correlation to source beds. Several theories explaining the source of the Lyons oils have been proposed: the oil formed from (1) adjacent Permian and Pennsylvanian shales (Levandowski et al.. The anhydrite is more likely a late diagenetic phase. is a Paleozoic variety that is chemically distinct from oils common in younger rocks (Clayton and Swetland. the composition of sulfate from modern seawater is about +20‰. or from sulfide minerals that oxidized to form sulfate. the carbon isotopic composition of dolomite that forms from such a mixture will largely reflect the δl3C values of species in the brine.6 to +12. because the cement precipitated after oil first migrated into the Lyons. 1980). and Mowry shales and the Greenhorn Limestone. rather than carbon derived from early calcite cements or from oxidation of hydrocarbons in the Lyons Sandstone. Sulfur in Anhydrite Cement cements suggests that the anhydrite could have formed from seawater during the Permian or from sulfate that was derived from evaporite rocks in the basin and imported into the Lyons Sandstone. in a fluid mixture. The cements have sulfur isotopic compositions (Table 1) ranging from +9. organics. the Lyons Sandstone is not known to have served as a carrier bed for long-distance migration. Of this oil. the dissolved carbon derived from a CO2-charged brine overwhelms that from meteoric waters. Some oil in the Cretaceous section has migrated as far as 150 km from the deep basin toward the basin’s eastern flank. consequently basin sulfide reservoirs are typically depleted in 34S (Thode et al.0‰ (Claypool et al. which is +9. including the Carlisle. 1980). which also lie in the Cretaceous section (Clayton and Swetland. rather than precipitated from seawater or as pore fluid deposited with the sediment. The “normal” carbon-isotopic composition of the dolomite cements likely reflects in large part the composition of marine carbonate that dissolved into a brine deep within the basin. The anhydrite cements are too enriched in 34S to be consistent with such an origin. much less than is present in the formation. For reference. Sulfide that reoxidizes does not fractionate significantly. The sulfur isotopic composition of the anhydrite Reservoirs in Cretaceous strata provide most of the oil and gas produced from the Denver basin. For these reasons. Graneros. 1951). In addition. Oil produced from the Lyons. The 34S values of reduced sulfide dissolved in natural waters normally range from –32 to 0‰ on the CDT scale (Hartmann and Nielsen. long after the sandstone was deposited. hence.. We show in the next section that most Paleozoic oil in the Denver basin did not form until the Late Cretaceous or early Tertiary. Sulfur isotopes are strongly fractionated between sulfide and sulfate. The anhydrite is unlikely to have formed from sulfur derived from H2S.5‰ relative to the CDT standard. on the other hand. interpreted to reflect the composition of a deep CO2charged groundwater that has δl3C values similar to that of seawater. using the method of Westgate and Anderson (1982) for extraction and analysis. The carbon composition of dolomite is compatible with a fluidmixing origin suggested by the oxygen composition. Oil usually forms late in a basin’s history because source beds must be deeply buried long enough to become thermally mature.6 to +13. 1980).7% of the sandstone’s compacted volume. These values span almost exactly the range in isotopic composition of sulfate derived from Permian evaporites worldwide. (2) the Permian Satanka . mostly as a result of the bacterial reduction of sulfate in the near surface.

against the maturity in Cretaceous source beds determined by Clayton and Swetland (1980) by vitrinite reflectance measurement. (2) (1) Index n varies from the lowest to the highest temperature interval encountered by the sediment. Clayton. According to kinetic theory. models of thermal maturation depend on the choice of thermal conductivity and basal heat flow into the basin. Figure 7 shows how the thermal maturity of Pennsylvanian source beds from the deepest basin evolved through geologic time. the index value n in equation 1 increases suddenly by 1. ∆tn is the length of time (in million years) that the sediment spends within each temperature interval. Momper. Dolomite and anhydrite cements in the gray Lyons Sandstone. According to the calculation. in percent reflectance in oil) from 0. 1984. Middle Pennsylvanian black shales. which corresponds to vitrinite reflectance measurements (Ro. 1978). respectively) from basement. 1989). in cal/cm·s °C) increases during compaction according to the correlation with porosity φ: which is taken from the data of Sclater and Christie (1980) for North Sea shales. The TTI of a sediment is given as the sum of maturities developed in each temperature interval Figure 7—Evolution of thermal maturity in Pennsylvanian source beds beneath the Lyons Sandstone.224 Lyons Sandstone. 76 mW/m 2 . however.8 HFU (50. For each 10°C increase or decrease in temperature.. .65 to 1. even the deepest strata did not become thermally mature until late in the basin’s history. until thick shale sequences were deposited in the Cretaceous.2 to 1. 1980). 1980). Oil forms when TTI falls between 15 and 160.4φ) × 10–3. Much of this range.5. however. near the average for continental crust. The time-temperature index (TTI) was calculated using Lopatin’s method (Waples. are rich in organic material and are considered to have excellent source potential over much of the northern Denver basin (Clayton and Ryder. When the TTI exceeds 160. For this reason. 1978). the Lyons oils could well have been derived from these beds. According to the method. The oil window is the area between the two dashed lines. For a given burial history. source beds are immature until they attain TTI values of about 15. as calculated for differing heat flows. In a Cretaceous reservoir. K= (5. We assume that sediment thermal conductivity (K. 63.5 HFU. time and temperature control the generation and preservation of oil in sedimentary basins. As the buried sediment passes from one temperature interval to the next.8 HFU (1 HFU = 10–6 cal/cm2·s). likely reflects the uncertainty in estimating the depth of past burial and the timing of uplift and exhumation for samples taken from outcrop in the western basin. calculated assuming differing heat fluxes of 1. The third theory seems unlikely because the oils differ in carbon isotopic composition from that of Phosphoria oils in the Big Horn basin (Clayton et al.35 – 4. We can place a limit on the age of the cements by calculating when oil generation began in Pennsylvanian rocks. Clayton (1989) found oil from a Middle Pennsylvanian source. or (3) the Permian Phosphoria Formation near the Idaho-Wyoming border (Dimelow. 1. We find that a heat flow of about 1. the deepest sources began to generate oil between 78 and 50 Ma. Heat flow values. oil begins to break down to form natural gas. the index value increases or decreases by 1. as already discussed. reproduce individual measurements made near our study area. ranging from 1. 1. gives the best fit to the pattern of maturity basinwide in Cretaceous strata.2. formed after oil first migrated into the formation. Waples. We calibrate heat flow. Basin strata were rather shallowly buried during most of the Paleozoic and Mesozoic. during the Late Cretaceous or the early Tertiary. Denver Basin Formation (Berman. assuming that it remained constant through time. 1987). which causes the small kinks on the TTI curves. In the model. The interval from 100–110°C was chosen as the base interval and given the index value n = 0. 1971. source-bed maturity is described by a time-temperature index (TTI). For our calculations. 1972.3. we use the simple model of Lopatin (Lopatin. There is no reason to believe that oil could not have migrated vertically from a Pennsylvanian source to Permian reservoirs. demonstrating that oil has migrated vertically upward through the Paleozoic and into the Mesozoic section. The factor 2n assumes the doubling of the maturation rate with each 10°C rise in temperature.

05 0. expressed as a fraction. according to these results.40 0.55 0. The basin surface in our calculations is held at a constant temperature of 10°C and atmospheric pressure. which we take as a barrier to fluid flow and the source of a constant heat flux of 1. Model results provide quantitative estimates of various parameters needed to evaluate diagenetic models.5 10 10 *φ = φoexp(-bZ) + φ1. Bethke et al. kX ≤ 1 µm2. We calculate the evolution of porosity and permeability as each rock type is buried. 1988. near the end of the Cretaceous. migrated from source to reservoir in the Tertiary.05 0. This timing provides a further argument that the cements of the gray Lyons Sandstone formed late in the basin’s history. including past rates and directions of fluid migration as well as pressure and thermal gradients along flow paths. Table 2 shows the hydrostratigraphy assumed in the simulations.5 HFU. the evolution of porosity and permeability.85 φ1 0. BASIN PALEOHYDROLOGY Table 3. Jelm Formation lower Lykins Formation Lyons Sandstone Owl Canyon Formation. Most Paleozoic oil in the basin.50 0. 1 µm2 ≈ 1 darcy. The model calculates the groundwater flow that arises from sediment compaction and topographic relief.. 1985.55 Porosity* b (km–1) 0. Each stratigraphic unit in the calculations is composed of varying fractions of four rock types: sandstone. Corbet and Bethke. and cementation by migrating groundwater. Ralston Creek Formation upper Lykins Formation. Satanka Shale Ingleside Formation Fountain Formation Cretaceous Jurassic Triassic Triassic Permian Permian Permian Pennsylvanian Assuming our best-fit value for heat flow of 1. **log kx (µm2) = Aφ + B. Carlile Shale. We show the results of two reconstructions of the basin’s paleohydrology. The mathematical basis of the calculation technique is described in detail elsewhere (Bethke.85 0. The first simulation depicts how the basin’s groundwater flow regime evolved as the basin subsided and infilled with sediments. the maturity of petroleum source beds through time. and evaporite-rich siltstones. The inferred timing also suggests that the cements did not precipitate from fluids remobilized by compacting sediments. Graneros Formation Dakota Group Morrison Formation. shale. Correlations used in the Hydrologic Models to Calculate Porosity and Permeability φ0 Sandstone Carbonate Shale Evaporite 0.05 A 15 6 8 8 Permeability** B kx/kz –3 –4 –7 –7 2. carbonate.40 0. after infilling of the basin was complete. the first oil probably formed about 68 Ma. The bottom of the cross section is the contact with crystalline Precambrian basement. Hydrostratigraphy Assumed in the Hydrologic Models System Cretaceous Shales Lower Cretaceous Sandstones Triassic-Jurassic Permian-Triassic (1) Permian-Triassic (2) Permian-Triassic (3) Pennsylvanian-Permian Cretaceous Rock Stratigraphy Pierre Shale. using correlations shown in Table 3. 1992).05 0. Sundance Formation.5 2. the second portrays flow resulting from Eocene uplift of the Front Range along the basin’s western margin. because sediment compaction was complete or nearly so by the time oil generation began. thermal conductivity is set accord- .55 0. to simulate fluid migration through the Denver basin in the geologic past. the transfer of heat by conduction and advecting groundwaters. The sides of the basin remain at hydrostatic pressure and are open to groundwater flow. Niobrara Formation. Z is burial depth (km). as some have suggested.5 HFU.Lee and Bethke 225 Table 2. 1993. Greenhorn Limestone. We used BASIN2. a numerical model of groundwater flow in sedimentary basins.

Hydrostratigraphic units are described in Table 2. Denver Basin Figure 8—Cross section used in simulations of the compaction-driven groundwater flow shown at the end of Cretaceous.226 Lyons Sandstone. Section extends 400 km across the northern Denver basin (BB'). .

which represent the drive for fluid migration.. primarily into the Cretaceous aquifer complex and. as shown by the scale bar. temperatures in the calculation fall along a conductive gradient that corresponds to the assumed thermal conductivity correlation and heat flux from basement. and fluid expulsion from compacting sediments are common to basins of the cratonic interior (e. The role that fluids displaced by sediment compaction could have played in precipitating cements is limited by the modest flow rates predicted by the calculation. because of a lack of data. 1991). The simulation follows cross section BB' (Figure 8). correlations for the hydrologic properties of shales were assigned to the evaporitic siltstones. 1991). The estimated flow velocities (true rather than Darcy) through the Lyons Sandstone are everywhere less than about 2 cm/yr or 20 km/m. which traverses the northern Denver basin.y. fluids migrate from compacting shale and carbonate strata upward and downward. Figure 9 shows the results of the calculation for the final time slice in the calculation. are taken from previous studies of basins from the cratonic interior (Bethke et al. at a rate less than 100 m/m. which infilled over 30 m. except the evaporitic siltstones. Dashed lines show predicted temperature distributions. at the end of the Cretaceous. but it seems unlikely that sediments in this basin were impermeable enough to allow genuine overpressures. The correlations for porosity and permeability of all rock types. such as those observed in the Gulf of Mexico basin (e. Contours are calculated excess pressures (atm) in the basin. Such slow rates of subsidence. about 100 times more rapid than rates of burial in the Denver basin.. Lyons Sandstone. These aquifers act as drains that carry fluid laterally from deep strata to the basin margins. Vertical components of cross section are exaggerated by about 40:1.000 m/m. In the calculation. . Bethke et al. infilling. near the basin’s depocenter.g. 1986) to have developed.. Arrows indicate the flow directions in the Dakota Sandstone.g. (Harrison and Summa.y. into the Lyons and Fountain sandstones. Overpressure in the Gulf basin results from burial rates sometimes exceeding 10. The equipotentials in this case (where there is no topographic relief and the reference elevation is sea level) also show the extent to which pressure exceeds hydrostatic. The small flow rates reflect the slow rates of burial and compaction in the basin. Because groundwater flow here is too slow to alter the thermal structure of the basin. Greater pressures result when permeability is set smaller. to a lesser extent. The calculation predicts that excess pressures developed in the Cretaceous shales were very small.1 MPa). about 1 atm (0.y. groundwater migrates toward areas of lower hydraulic potential.. Bethke. ing to equation 2. 1991). Contours in Figure 9 are equipotentials. Flow Driven by Sediment Compaction The first paleohydrologic model simulates groundwater flow resulting from sediment compaction as the basin infilled through most of the Paleozoic and Mesozoic.y.Lee and Bethke 227 Figure 9—Calculated groundwater flow driven by sediment compaction at the end of Cretaceous. and Fountain Formation.

l. Section extends 350 km across the northern Denver basin (CC'). s. = sea level. Hydrostratigraphic units are described in Table 2.228 Lyons Sandstone. Denver Basin Figure 10—Cross section used in simulations of the past groundwater flow during the Eocene. .

Calculation shows relatively rapid groundwater migration in response to the uplift of the Front Range. and the Fountain sandstone grades into dolomite and shale near the basin axis. which represent materials eroded and deposited during the Eocene. The Lyons pinches out near the eastern margin of the cross section. Groundwater recharged at high elevation along the basin’s uplifted western margin and discharged toward the eastern margin. a hypothetical well drilled into the aquifer in the Eocene would have been somewhat . The greatest hydraulic potentials in the deep basin develop within the Fountain aquifer (Figure 11). Groundwater flow continues in this regime today (Belitz and Bredehoeft. Present-day flow rates in these aquifers are not known. In Figure 10. Flow in this gravity-driven regime is much more rapid than in the compaction-driven system shown in Figure 9. but a model calculated by Belitz and Bredehoeft (1988) predicted flow velocities of about 3 m/yr. The predicted flow velocities are somewhat slower (<20 m/yr) in the Lyons and Fountain sandstones because these aquifers are more deeply buried and hence more compacted and less permeable. Trimble (1980) estimated that the Front Range was uplifted during the Laramide orogeny to between 2500 and 6000 m above sea level. On the basis of the thicknesses of the Laramie Formation and the Dawson sandstone. the predicted velocities shown here are perhaps optimistically high. basin fluids migrated eastward in response to the hydraulic gradient created by the slope on the water table. As the Front Range emerged during the Laramide orogeny. however. Hydraulic potential is greater deep in this aquifer than along the land surface directly overhead. The velocities predicted. and because neither aquifer is hydraulically continuous across the basin. therefore.Lee and Bethke 229 Figure 11—Calculated Eocene groundwater flow and the hydraulic potential distribution (gray contours. reflect two poorly constrained assumptions: the past elevation of the western basin and the aquifer permeabilities. we assume that the Front Range in the Eocene was uplifted 2000 m above the elevation of the basin’s eastern margin. in atm) in the Denver basin. but modern flow rates are presumably somewhat smaller now than in the past because the Front Range has eroded over time. The topographic relief drives groundwater through the Cretaceous aquifers at velocities of tens of meters per year (Figure 11). Flow Driven by Topography The second paleohydrologic model simulates groundwater flow driven by Eocene uplift of the Front Range in the western basin (Figure 10). By this standard. 1988).

Appendix 1 gives the mathematical basis for the calculation. but downward over much of the basin relative to fixed elevation. The present-day underpressuring in these shallow aquifers likely reflects erosion of the uplifted basin margin and poor hydraulic continuity across Front Range faults. the aquifer is also a zone of dispersive mixing. Therefore. where the fluid is descending or ascending across temperature gradients. It is possible.1% of the formation’s volume. because the solubility of this mineral varies more strongly with pH than with temperature or pressure. We refer to this type of calculation as a simple model of cementation because it accounts only for the effects of changing temperature and pressure along flow paths. Denver Basin overpressured. The cumulative volume is the volume of the mineral that formed or dissolved from the Permian to the Late Cretaceous per unit volume of the formation. the complex is underpressured over broad areas in the present day (Belitz and Bredehoeft. is not consistent with the observation that the greatest enrichment of anhydrite cements occurs within the Lyons oil fields (Figure 4). expressed in percent. however. whereas it dissolves quartz as it migrates toward higher temperatures. the drive for cross-formational flow is provided by the hydraulic potentials in the Fountain which. The cumulative amount of anhydrite precipitated in this regime. The Lyons Sandstone serves as an aquifer for eastward flow. not just in deep strata where the gray facies occurs. We did not attempt to calculate the distribution of dolomite cementation. groundwater in the deep basin warms with time and. mix with Lyons groundwaters. This simple model of cementation poorly predicts the volume percentage of anhydrite cements in the gray Lyons sandstone. Quartz. Fluids migrating along the underlying Fountain Formation discharge upward across stratigraphy where the sandstone grades into less permeable facies near the basin axis. for example. In the hydrologic models. groundwater flowed upward relative to the subsiding strata. Simple Models of Cementation Figure 12—Calculated cumulative volumes (X) of anhydrite and quartz across the Lyons sandstone at the end of Cretaceous. Groundwater saturated with quartz precipitates quartz cement as it migrates along a path of decreasing temperature. because of the retrograde solubility of anhydrite. Potentials in the Cretaceous aquifer complex are slightly overpressured to the west and underpressured to the east. and (2) flow driven by uplift of the western basin during the Eocene. We applied this type of calculation to predict the amounts of anhydrite and quartz that could form and be consumed as a result of (1) compaction-driven groundwater flow occurring as the basin subsided during the Paleozoic and Mesozoic. are greater than the potentials in overlying strata. The model predicts that the most extensive cementation or dissolution occurs along the flanks of the basin. This predicted cementation pattern. is increasingly soluble with rising temperature. Cementation rates near the Lyons oil fields calculated for groundwater flow in the Eocene (Figure 13) are small (<1%/m. the simple . precipitates anhydrite cement.). along the basin axis. In the calculated flow regime. to calculate the rates at which various minerals dissolve and cements precipitate throughout a basin from knowledge of mineral solubility and the groundwater flow pattern. and then continue to migrate eastward. therefore. As shown in Figure 11. but the cement is distributed across the sandstone. Groundwater migrating through a sedimentary basin alters the sediments through which it flows because mineral solubility varies with temperature and pressure along the flow path. For this reason. is less than about 0. again.y.230 Lyons Sandstone. Appendix 2 gives the correlations we used to compute mineral solubility. Along the basin axis. The calculation results for the compaction-flow regime (Figure 12) show the cumulative amounts of anhydrite and quartz formed or dissolved within the Lyons Sandstone in this hydrologic regime by the end of the Cretaceous. 1988). Not only is this volume too small to account for the origin of the sandstone’s gray facies. Fountain fluids along the basin axis discharge upward across the lower Satanka Formation and into the Lyons.

1970. we used REACT to correct it to the temperature of the calcula- . we use quantitative modeling techniques to predict which chemical reactions occurred and how stable isotopes fractionated during the mixing process. precipitates where the fluid is descending and dissolves where the fluid is ascending. could have driven a reaction by which calcite dissolved and dolomite and anhydrate precipitated. REACTIONS ACCOMPANYING FLUID MIXING From the results of paleohydrology modeling.. The cementation rate is expressed as a percentage of the formation’s bulk volume per unit time (%/m. such as the mixing of groundwaters of varying composition or common-ion effects in which species produced by dissolution of one mineral might drive the precipitation of another. Calculations here are based on version R46 of the thermodynamic database compiled at Lawrence Livermore National Laboratory (Delany and Lundeen. the Lyons Sandstone at depth contained a zone of fluid mixing. The location of the mixing zone corresponds generally to the area in which rocks of the gray facies are found.. 1991). do not account for other potential causes of cementation in the Eocene flow regime. we interpret that. Johnson et al. any minerals that become supersaturated precipitate. 1992) to solve for the overall reaction that would accompany fluid mixing in the deep Lyons Sandstone.y. The conceptual basis of our calculation is shown in Figure 14. 1983) that trace the chemical evolution of systems open to mass fluxes. The second fluid is more saline.. 1982. is initially dilute. The modeling results show that fluid mixing We used the computer program R EACT (Bethke. Wolery. Figure 13—Calculated cementation rates (dX/dt) of anhydrite and quartz across the Lyons Sandstone during the Eocene. A packet containing 1 kg of dilute fluid migrates along the Lyons sandstone. Since McConaghy et al. however. having reacted with evaporite beds in the Pennsylvanian and Permian section. As the fluids mix over the course of the calculation. A second fluid discharges upward into the Lyons and mixes into the fluid packet. REACT is one of a class of reaction path models (Helgeson.’s (1964) analysis was made at 51°C. Reaction Modeling ideas of cementation by flow through temperature and pressure gradients cannot explain the origin of the gray facies. creating the gray Lyons facies as an alteration product of the red facies. this simple model of cementation poorly predicts the distribution of anhydrite cement in the gray Lyons Sandstone. in the Eocene. We assumed that the mixing occurred at 100°C. These simple models of cementation. which originated as meteoric recharge along the Front Range.. 1964. In this section.Lee and Bethke 231 Figure 14—Conceptual model for tracing chemical reaction during fluid mixing in the Lyons Sandstone. maintaining equilibrium with calcite and quartz in the formation. which has a retrograde solubility. Reed. The packet. Plummer et al. We used an analysis of modern Lyons groundwater to set the fluid’s initial composition (McConaghy et al. Anhydrite. Helgeson et al. Again. 1979. 1969) for calculating activity coefficients. and employ the extended Debye-Hückel method (Helgeson. 1968. sample C5-69-18bbcc). Quartz shows an opposite pattern due to its prograde solubility.). 1990.

Chemical Compositions (mg/kg) of Endmember Groundwaters used in Mixing Calculations Lyons Na+ Ca2+ K+ Mg2+ SiO2 HCO– 3 Cl— 2— S04 pH (100°C) 110 19 4. however. where Q is the activity product for the dissolution reaction. and we found that mixing the Fountain brine into the Lyons Sandstone causes dolomite and anhydrite to precipitate at the expense of calcite. dolomite is a more voluminous reaction product than anhydrite in the calculation results (Figure 15). anhydrite. We assumed that the composition of the Fountain groundwater reflects equilibrium with minerals in the evaporite strata that lie beneath the Lyons. The plot shows changes in mineral volume as brine mixes into the formation by dispersion. there is no petrographic basis for determining the pH of the Fountain brine.6 49 310 55 36 6. we do not know whether magnesite occurs in the evaporite strata.000 85. This apparent 2– . the Ca2+ and HCO3– added to solution drive precipitation of anhydrite and dolomite by a common-ion effect. We fixed pH by setting the CO2 fugacity to reflect a partial pressure of 50 atm. The reaction model (Figure 15) traces the chemical consequences of progressively mixing the Fountain brine into the Lyons aquifer at 100°C. the mineralogic results of the fluid mixing reaction vary little over a broad range of activity ratios. the slight predominance of anhydrite over dolomite cements observed in the facies. The fluid has equilibrated with dolomite and magnesite. Fortunately. → CaCO3 ← CaMg(CO3)2 – MgCO3 calcite dolomite magnesite As a result.000 13. calcite has a saturation index (log Q/K.5 tion by heating the analyzed fluid in the presence of calcite and quartz. As long as the Mg 2+ /Ca 2+ activity ratio is greater than about 0. As calcite dissolves into the undersaturated brine.7 Fountain 55.232 Lyons Sandstone. The predicted reaction explains the origin of anhydrite and dolomite cements. whereas the Fountain brine is mostly a sodium-chloride-bicarbonate solution (Table 4).6 0. so the saturation state is fixed by the reaction. The choice of NaCl concentration represents the upper limit of validity of the activity coefficient correlations we used. so the actual brine from these rocks could have been undersaturated with respect to this mineral and hence have a smaller activity ratio than predicted. and negative volume changes show dissolution. magnesite.800 4. It does not explain. regardless of pH or CO2 fugacity. The choice of CO2 fugacity for the Fountain brine reflects our interpretation of the isotope data. Positive volume changes indicate precipitation. as previously explained. as well as the lack of calcite in the gray facies. The sandstone is initially filled with the infiltrating meteoric water. and quartz. In fact. Denver Basin Table 4. but the degree to which calcite is undersaturated in the brine.000 510 3400 22 28. the critical variable in the calculation. Hence. We might have chosen a different fugacity. is independent of pH or CO 2 fugacity.3. In the model. The overall reaction predicted by the model is 5 CaCO3 + 2 SO4 + 5⁄2 Mg2+ calcite → 2 CaSO 4 + 5⁄2 CaMg(CO 3 ) 2 + 1⁄2 Ca 2+ anhydrite dolomite This reaction reduces porosity in the sandstone because the volume of anhydrite and dolomite produced is greater than the volume of calcite dissolved (Figure 15). and fixed the fluid’s Mg2+/Ca2+ activity ratio to an endmember value of about 20 by specifying saturation with magnesite. Figure 15—Predicted overall reaction resulting from mixing of the ascending Fountain brine into the Lyons Sandstone. The resulting Lyons groundwater is predominantly a sodium-bicarbonate solution. as would be expected in a saline groundwater from evaporite beds. calcite is undersaturated in the brine. and K is the equilibrium constant) of about –1. The Fountain groundwater in our calculations is a 3-molal NaCl solution that has equilibrated with dolomite.06. we set the Ca2+ concentration in the brine by assuming equilibrium with dolomite.

As previously noted.K. Isotope Fractionation Factors Assumed in the Calculations. as we previously hypothesized. Sheppard and Schwarcz (1970).3 17. the value chosen for CO2 fugacity has little effect on the mineralogic results of the reaction path. as implemented in R EACT . We assumed that the initial system contained a meteoric water having an initial δ l8 O of –13‰ (SMOW). Malinin et al. the solubility of anhydrite varies with temperature and fluid salinity. however. dissolved species. as is commonly done in this type of model. The δ13C values predicted by the model depend on the CO2 fugacity assumed for the Fountain brine. Chiba et al. Figure 16 shows how the isotopic compositions of dolomite cements (assuming varying CO2 fugacities for the Fountain brine) evolved as the Fountain brine mixed into the Lyons Sandstone.5 — *Fractionation factors calculated from regression curves compiled by J. has very little effect on the results since the amount of oxygen in the system provided by water overwhelms the small amount obtained by dissolving calcite. A reaction model . and minerals. and hence is not constrained by petrographic observations. Bottinga (1968). The l8 O composition of calcite assumed in the model. (1977). the dolomite δ13C more closely reflected the isotopic composition of carbon species in the brine than carbon from the precursor calcite. (1972). not precipitate. Only the increments of minerals precipitated during a reaction step are in isotopic equilibrium with the fluid. 1969). the calculation predicted dolomite compositions similar to the δ13C of the precursor calcite. the mean 18O composition of this cement.38 1. salinity curves are convex upward (see Figure 6 of Blount and Dickson.5 21. Factors Describe Fractionation at 100°C of 18O Relative to Water. The flaw in this reasoning.1 28. The calculation technique employs mass balance equations similar to those derived by Bowers and Taylor (1985) and Bowers (1989). however. O’Neil et al. anhydrite 2– Table 5. The brine has δl8O and δ13C values of 0‰. Table 5 shows the fractionation factors assumed in the calculation. Bohlke from the following sources: Northrop & Clayton (1966). Clayton et al. indicating that mixing a dilute and a saline fluid should dissolve. → CaSO4 ← Ca2+ + SO4 . Expressed as 1000 ln α. accounts for fractionation among the solvent. The δl8O of calcite was set to +11‰. **Assumed. gases. which migrates into the Lyons during the reaction process (Figure 14). Assuming CO2 fugacities in the range of about 25 to 100. the calculation segregates minerals from isotopic exchange. as already discussed. Richet et al. In this way. which was set to reflect the mean isotopic composition of calcite cements (–11‰) found in the red facies of the Lyons Sandstone. which would seem to negate the analysis presented above. are quite useful because they trace the effects among all of the reactions that occur in a system in order to predict the overall reaction. (1967). Isotope Fractionation To test whether the reaction model presented in the previous section can explain the isotopic compositions of cements in the gray facies of the Lyons Sandstone.91 — 0 3. We further assumed that the Fountain brine. The solubility vs. Reaction models. For example. and 13C Relative to CO2* 18O 13C Anhydrite Calcite Dolomite Quartz CO2(aq) HCO– 3 2— CO3 SO2— 4 18. or a brine with a higher SO42–/Mg2+ ratio than assumed in the calculation. whereas the reaction model correctly accounts for the effects of other reactions occurring in the system. When we assumed small values for the CO 2 fugacity of the Fountain brine. (1969).Lee and Bethke 233 discrepancy might be accounted for by slower precipitation rates of dolomite. (1981). The technique. is that the solubility curve or function represents the single reaction. The δl3C of the initial fluid was –12‰ (PDB).1 — 3. in isolation. An analysis that considers only anhydrite solubility would incorrectly predict that fluid mixing could not account for the cementation observed. we traced how the stable isotopes 13C and 18O fractionate during the predicted reaction. anhydrite. the calculation procedure models the fractionation that results from the reaction process in the absence of isotope exchange. gives dolomite compositions near the observed range. as might be expected in a sedimentary brine. and only the mass of minerals dissolved over a step affects the fluid’s isotopic composition.5 3. which matches estimates for Tertiary rainfall in the region.0 0** 0** 17. is isotopically heavier. Treating reactions individually without concern for their interactions can be misleading. but so little of this mineral reacts that its isotopic effect can be neglected. We further assumed that quartz was in equilibrium with the initial fluid. Instead of assuming that minerals maintain isotopic equilibrium with the fluid. In this case. such as the one we used. Results show that the range in δ l8O values observed for the dolomite cements can be explained by mixing in differing proportions of Lyons groundwater with Fountain brine.1 21.

1989). We argue that the gray facies of the Lyons formed as an alteration product of the sandstone’s red facies in zones where fluids. Oil in the Lyons Sandstone probably derived from source beds in underlying Pennsylvanian strata. Tick marks along the center curve show waterrock ratios (by mass). the brine has δl8O = 0‰ and δ13C = 0‰ (×). before the peak of the Laramide orogeny. dolomite and evaporite minerals. therefore.. In a common-ion effect. indicating that CO 2 pressures at about 120°C could range up to at least 40 bars. and CO 2 carried in the brine to form the dolomite and anhydrite cements of the gray facies. The association in time and space of oil migration . Oil generation began toward the Late Cretaceous or the early Tertiary. likely along the same fractures that transmitted the oil. The fugacities suggested by the above analysis are somewhat greater than those measured in the Gulf Coast. The slowly flowing Fountain groundwater became saline by reaction with minerals in the evaporite beds. for a given CO2 fugacity of brine (on an atmosphere scale). with meteoric water. 13C in the dolomite is consistent with deriving the CO2 by dissolving Paleozoic marine carbonate into the brine. The CO2 contents in fluid inclusions there are greater than 0. By this process of fluid mixing. and 34 S in the anhydrite matches the composition of sulfate from Permian evaporite beds. the brines migrated upward across confining aquitards. The regime of groundwater flow set up by the orogeny drove brines along the Fountain sandstone. sulfate. calcium from the dissolving calcite reacted with magnesium. This reaction accounts for the isotopic compositions of the cements. Calcite dissolved as the brine mixed into the Lyons because the upwelling brine was undersaturated with this mineral after having reacted with Figure 16—Calculated isotopic composition of dolomite resulting from the mixing of ascending brines into the Lyons Sandstone. Denver Basin that calls on migration of a CO2-charged brine into the Lyons. rocks of the red facies of the Lyons Sandstone were transformed locally along the basin axis into the gray facies. such as the Fountain brine. and discharged into the Lyons. and perhaps by the brine itself. Meteoric water recharged into the Lyons and moved deep into the basin. migrated into the formation from underlying strata. we integrated quantitative models of groundwater flow and chemical reaction with petrographic and isotopic observations to explain the origin of the petroliferous gray facies of the Lyons Sandstone. 1991). including groundwater and petroleum. Hematite stains of the red facies were also consumed in the mixing zone by reduction in the presence of the migrating petroleum. how the isotopic composition of dolomite evolves over the course of the reaction. Initial isotopic composition of the Lyons groundwater is δl8O = –13‰ and δ13C = –12‰. In our model. which reached its peak in the Eocene. The orogeny likely produced fractures along the basin axis along which the oil could migrate upward by buoyancy across aquitards and into the Lyons Sandstone. The precursor calcite cement (×) has δ l8 O = +11‰ and δ13C = –11‰. The broad range of δ18O in dolomite requires mixing of an isotopically heavy fluid. Each curve shows. Where Fountain sandstone facies pinch out. Flow in the underlying Fountain Formation was much more restricted because sandstones in this formation pinched out to carbonate and shale beds along the basin axis. set up a regime of eastward groundwater flow.35 molal. Kilgore and Elmore (1989) documented a similar reaction in which hydrocarbons removed hematite cements from red beds of the Triassic Chugwater Formation in Montana. Calcite and dolomite compositions (boxes) are from Levandowski et al. reflecting the mineral’s mean composition. where CO2 pressures are less than 5 bars at 100°C (Smith and Ehrenberg. as shown in Figure 15. DISCUSSION In this study. explains both the petrography and isotopic compositions of cements in the Lyons gray facies. Laramide uplift of the Front Range along the basin’s western margin. The fugacities suggested for the Denver basin might better be compared to CO2 concentrations and temperatures measured in fluid inclusions in late dolomite cements throughout the Ozark region (Leach et al.234 Lyons Sandstone. (1973). Our results show how past groundwater flow in the Denver basin drove diagenetic reactions in the Lyons that are intimately related to the accumulation of petroleum into present-day reservoirs.

and thus would precipitate and dissolve in an opposite pattern. but depends only weakly on pressure and fluid salinity over the temperature range considered in this study (Kitahara.Lee and Bethke 235 and groundwater mixing explains the fact that all known petroleum reservoirs in the formation are found within the gray facies.82 × 10–5T2 – 7. P. v. C. 47–57. Handford. account for the effects of groundwater advection. p. Wyoming. and fluid salinity. 1982. The dorag dolomitization model—application to the Middle Ordovician of Wisconsin: Journal of Sedimentary Petrology. Plummer. The final terms. is lnmanh = –2. Rightmire. the correlation log mqtz = 1. R. The dimensionless solubility Vi. eds.05 × 10–8P2 + 3. that is. φ is porosity. v. taken in the reference frame of the sediments as they are buried or uplifted. W. 965–984. it is sufficient for our calculations to carry ρf as 1 and TDS as 0 in equation 6. as occurred higher in the section along the Cretaceous aquifers. Back. which gives molal solubility as a function of temperature T (°C) and pressure P (bars). 1973. 43... Rubin. mi. Davies. Quartz solubility increases rapidly with increasing temperature. which are the most significant even at modest flow rates. 1983. v.881 – 2. Anhydrite has a retrograde solubility. and TDS is the total dissolved solids (in mg/kg) in the fluid. p. Walther and Helgeson. 1977). To calculate quartz solubility. 1979. and M. the rate at which the mineral precipitates or dissolves can be computed from the velocities of groundwater flow and the temperature and pressure fields.028 × 10–3Tk—1560/Tk (5) of Rimstidt and Bames (1980) gives the molal solubility of quartz as a function of absolute temperature Tk (in kelvins). p. especially at lower temperatures. (4) By this correlation. For simplicity. 94. and the fluid can be assumed to remain in local equilibrium. T. as would occur when a formation is buried or exhumed. Adams.1 × 10–4P + 1. C. and that the species has an activity coefficient of about 1. increases with temperature. Dx and Dz are coefficients of hydrodynamic dispersion (cm2/yr). we assume that the mineral is soluble as SiO2(aq). and ν'x and ν'z are lateral and vertical groundwater velocities (cm/yr) in curvilinear coordinates. Depositional and diagenetic spectra of evaporites: SEPM Core Workshop 3. can be calculated from a mineral’s molal solubility. the volume of a mineral soluble per unit volume of groundwater. 16. Under these conditions. If a mineral’s solubility depends on temperature and pressure alone. the cements probably began to seal the oil into reservoirs as it accumulated. Powder River basin. we represent solubility by the dimensionless ratio Vi of the volume of mineral i that can be dissolved in per unit volume of groundwater. N. anhydrite solubility decreases with increasing temperature but increases with increasing pressure. pressure. Process and rate of dedolomitization: mass transfer and 14C dating in a regional carbonate aquifer: Geological Society of America Bulletin. as is generally the case except under alkaline conditions. p. Terms on the next line account for the effects of diffusion and dispersion of the dissolved mineral within the basin. R. the cementation rate dXi/dt for mineral i is given directly by APPENDIX 2: SOLUBILITIES OF ANHYDRITE AND QUARTZ The equilibrium solubility of anhydrite depends on temperature. Badiozamani. Quartz dissolves where groundwater migrates toward higher temperatures because the fluid must acquire silica to remain in equilibrium with quartz. 1960. preventing farther migration. The solubility of quartz. To calculate the role that fluids migrating across temperature and pressure gradients play in anhydrite precipitation. we calculated the amounts of anhydrite and quartz that would have precipitated from groundwaters migrating along temperature and pressure gradients. Similarly. and J. in C. Hanshaw. Sabkha-dune deposition in the Lyons Formation (Permian) northern Front Range. 1415–1429.. G. H. and G.. as (6) where MVi is the mineral’s molar volume in cm3/mole. B. The first terms on the right side of equation 3 give the cementation rate caused simply by changing temperature and pressure within a sediment. Since the migration and cementation processes were so closely linked. The resulting regression. W. we correlated solubility to values measured by Blount and Dickson (1969) for a 2-molal NaCl solution. APPENDIX 1: SIMPLE MODEL OF CEMENTATION BY GROUNDWATER FLOW To test the relationship between regional groundwater flow and cementation in the basin. we represent cement volume by the ratio Xi of the volume of mineral i present per unit volume of the formation. Rahn. which account for diffusion of the solute as well as mechanical mixing that occurs in a groundwater flowing through a porous medium. K. and the observation that some dolomite and anhydrite cements overlie bitumen stains left behind by migrating oil. L. Patton. . Blount and Dickson’s measurements span the range of about 100 to 250°C and 5 to 1350 bars. B.59 × 10–2T + 6. 193–209. Sabkha anhydrite: the supratidal facies of cyclic deposition in the Upper Minnelusa Formation (Permian) Rozet fields area. for example. The model also explains why oil in the Lyons apparently did not continue to migrate eastward through the sandstone. Colorado: The Mountain Geologist. where. In this case. so the correlation serves to extrapolate the data to temperatures found in shallow strata. Quartz cements form where groundwater flows in a direction of decreasing temperature and hence decreasing solubility.. Because the largest uncertainty in calculating cementation rates is the estimation of groundwater flow rates. The time derivatives d/dt in equation 3 are Lagrangian. Loucks. R. J.14 × 10–7TP. T and P are temperature and pressure. REFERENCES CITED (3) Achauer.33 – 1. ρf is the fluid density in g/cm3. The net effect is that anhydrite is less soluble at depth in a basin than at surface conditions. on the other hand.

221–230. 3057–3067. Permian stratigraphy and paleotectonics.-K. and L. H. A. 261–267. and GTPLOT: University of Illinois. A.. v. and Wyoming: U. Oelkers. and W. Geological Survey Open File Report 82-914.. M. Disequilibrium fluid pressures and groundwater flow in the Western Canadian sedimentary basin: Journal of Geophysical Research. v. Supercomputer analysis of sedimentary basins: Science. 18. K. 227–245. REACT. M. J. and K. 1980. C. 34. R. Shelton. W. northern Denver basin: AAPG Bulletin. 46–59. Ryder. Mayeda. 1613–1633. 1986. Hanson. Harrison. Elmore. 7203–7217. 195–231. 1972. 39. 70–89. 1968. p. Claypool.. northern Denver Basin— applications for exploration: AAPG Bulletin.. C. J. 46. Colorado: The Mountain Geologist. Colorado. J. S Matsuo. King. 242 p. The geochemist’s workbench. M. Front Range. Modeling fluid-rock interactions in sedimentary basins. 1989. v. Petroleum generation and migration in Denver basin: AAPG Bulletin. Upson. Petroleum source-rock potential of Pennsylvanian black shales in Powder River basin. Trucial Coast. p. Urbana. TACT. Kreiling. J... southern Montana: Geological Society of America Bulletin. 1991. Clayton. Modern evaporite deposition and geochemistry of coexisting brines. p. and J. J. W. Geohydrology of the Madison and associated aquifers in parts of Montana. T. The solubility of anhydrite (CaSO 4 ) in NaCl-H 2 O from 100 to 450°C and 1 to 110 bars: Geochimica et Cosmochimica Acta. Colorado: M. Bredehoeft. and B2 VIEW : University of Illinois. D. T. v. H. Chiba. p. C. F. Late Cenozoic sedimentation and deformation in northern Colorado and adjoining areas.. 1992. 1988.. 68. p. C. Clayton. 1987. Jr. J. T. 1979. p. Hydrodynamics of Denver basin: explanation of subnormal fluid pressures: AAPG Bulletin. gases. Kaplan. eds. 81. Thermodynamics of hydrothermal systems at elevated temperatures and pressures: American Journal of Science. 86 p. A. Reed.. Kilgore. G. v. v. G. Geochemical correlation of Paleozoic oils. 935. p. Hydrologic constraints on the genesis of the Upper Mississippi Valley mineral district from Illinois basin brines: Economic Geology. 1. L. a guide to using B ASIN 2. Vuletich. Gregg.. J. B. p. F. Berman. ed.. 1990.236 Lyons Sandstone. 73. W. Johnson.. F. F. D.. 12583–12606. Hubert. Nebraska (abs. A study of the relationship between hydrocarbon migration and the precipitation of authigenic magnetic minerals in the Triassic Chugwater Formation. An integrated chemical and stable-isotope model of the origin of Midocean Ridge hot spring system: Journal of Geophysical Research. Helgeson. the sabkha. Blood. v. 977–988. 33–48. Swetland. p. Bethke.. J. Nigrini. Bethke. 1969. E. I. 1988. 53. v. p. Bottinga. v. C.. C. 1981. Somiya. 1989.): AAPG Bulletin. O’Neil... v. Helgeson. 1991. S. and H. p. p. J. v.. p. Golden. 127 p.. 150 p. Dickson. Stable isotope signatures of water-rock interaction in mid-ocean ridge hydrothermal systems: sulfur. and S. Baker. N. 729–804. 32. v. Colorado. aqueous species.. p. 267. 1960. 800–808. Altaner. 55–62. 1970. and P. J. 174 p. B2 VIDEO . A. 1984. v. 233–249. Shukla and P. Lyons Sandstone. v. E. Quantitative dynamic stratigraphy: New York. 58. 621–655. δ34S-Werte in rezenten Meeressedimenten und ihre Deutung am Beispeil einiger Sedimentprofile aus der westlichen Ostsee: Geologische Rundschau. Larimer County. Nielsen. p. oxygen. Evaluation of irreversible reactions in geochemical processes involving minerals and aqueous solutions—I.. v. H. v. Hirano. and S. 1972. and hydrogen: Journal of Geophysical Research. E. 103–109.. Paleopoles and paleolatitudes of North America and speculations about displaced terrains: Canadian Journal of Earth Science... L. Meyers. 1962. p. H.. p. and W. Clayton. 1985.. 101. and C. W. Blount. Clayton. Oxygen isotope fractionation factors between anhydrite and water from 100 to 550°C: Earth and Planetary Science Letters. 64. and S. L. Y. Water-rock interaction history of regionally extensive dolomites of the BurlingtonKeokok Formation (Mississippian): isotopic evidence. 1975. L. T. Pennsylvanian and Lower Permian of northern Denver basin. The age curves of sulfur and oxygen isotopes in marine sulfate and their mutual interpretation: Chemical Geology. 1978. E. Curtis. 1280–1288. Colorado: Colorado School of Mines Quarterly. Stratigraphy and petroleum. W. Bethke. v. Bethke. Colorado School of Mines... 1989. 77. 1991. C. T.. p.. Colorado School of Mines.. 109–176. p. 179–209. Izett. and S. Colorado. 1970. H. Bethke. The LLNL thermochemical database: Lawrence Livermore National Laboratory Report UCRL21658. Brown. 101 p. L. P. v. Helgeson. 5775–5786.. D... M. Wyoming. v. E. p. Bethke. M. Golden. 1990. 28. R. W. Corbet. 97. 16. Brine migrations across North America—the plate tectonics of groundwater: Annual Review of Earth and Planetary Science. and A. Hartmann. Sakai. P. Downey. and northern Denver basin. Calculation of mass transfer in geochemical processes involving aqueous solution: Geochimica et Cosmochimica Acta. T. Belitz. 97– 114. Hydrogeology Program. 101. 33. 94. 1334–1359. 291. and reactions from 1 to 5000 bars and 0°C to 1000°C: Lawrence Livermore Laboratory Earth Sciences Department. Holser. 1985. Cenozoic history of the southern Rocky Mountains: Geological Society of America Memoir 144. Hydrogeology Program. Petrology of the Fountain and Lyons formations. in V. G. and T. 72. v. J. Bethke. M. ACT2. S. D. p. L. and H. 925–1945. Helgeson. v. and D. Minor and trace-element distribution in the Bonneterre Dolomite (Cambrian). in T. T. Oltz. L. T. C. 6817–6828. J. 55. Dimelow. Taylor. p. v. v. 1969. A. R. M. Threlkeld. 1980. A. a users guide to RXN.-I. p. thesis. G. J. 1988. C. C. H. Zak. H.. Summa. Bowers. Clayton. v. Calculation of fractionation factors for carbon and oxygen exchange in the system calcite–carbon dioxide–water: Journal of Chemical Physics. 1969. 1991.S. p. SUPCRT92: A software package for calculating the standard molal thermodynamic properties of minerals. J. Oxygen isotope exchange between quartz and water: Journal of Geophysical Research. Elsevier. M. 124 p. Hoyt. 1992. p.S. southeast Missouri: evidence :for possible multiple basin fluid sources and pathways during lead-zinc mineralization: Geological Society of America Bulletin. 75. 1982. 71. 90. Long-range petroleum migration in the Illinois basin: AAPG Bulletin. Quinodoz. H. 199–260. v.. S. M. H. N. 853–877. M... Bethke. Delany. and R.. and F. M.. B2 PLOT . 1960. Bellvue-Livermore area. J. 239.. and R.. S. S. The solubility of quartz in the aqueous sodium chloride solution at high temperatures and high pressures: The . M. A numerical model of compaction driven groundwater flow and heat transfer and its application to the paleohydrology of intracratonic basins: Journal of Geophysical Research. N. Jones. W. Cross. Sedimentology and geochemistry of dolostones: SEPM Special Publication 43. 225 p. Lee. v. S.. J. Geochemical evidence for Paleozoic oil in Lower Cretaceous O sandstone. p. Kitahara. in B. 1968. p. 1989. N. thesis. M. Upper portion of the Fountain Formation and Lyons formation at Morrison.. Wyoming and Nebraska: AAPG Bulletin. Arabian Gulf: Journal of Sedimentary Petrology. South Dakota. v. A. C. Denver Basin Banner. v. v. C. and T. P. C. H. 669–694. no. p. Paleohydrology of the Gulf of Mexico Basin: American Journal of Science. v. northeastern Colorado: M. and I. J. p. Harrison. J. Marshak. Harrison. Irving. 1969. K. 569–592. W. 90. Kusakabe. C. R. ed. Butler. Basin modeling with B ASIN 2. M. 72. 7. Thermodynamic relations: Geochimica et Cosmochimica Acta. Urbana. and H. Lundeen. p. 287–315. p. 1993. Bowers. J.

in N. J.. v. G. Kleerekoper. P. D. 513–528. 843–883. land petroleum potential. 46. McElcheran. p. p. J. v. S.. 1967. 1971. Oil migration limitations suggested by geological considerations: AAPG Short Course Notes 8. J. Y. 277. Morse. 348–351. p. H. growth faulting and early Laramide crustal movement: The Mountain Geologist. 1979. Sclater. J. C. p. Anderson. v. Carbonate cements: SEPM Special Publication 36. 21. Richet. Sheppard. G. Westgate.. Contemporaneous dolomitization of middle Pleistocene reefs by meteoric water. and D. Reed. J. p. 44.. 962. P.. 1977.. N. central Front Range.. 1973. R. Rimstidt. Hydrogeologic data of the Denver basin. v. L. 1970. S. Sonnenberg. v. and M. 5. J. and J. southern Denver basin. Plummer. Fractionation of carbon and oxygen isotopes and magnesium between coexisting metamorphic calcite and dolomite: Contributions to Mineralogy and Petrology.. p. Extraction of various forms of sulfur from coal and shale for stable sulfur isotope analysis: Analytical Chemistry v. P. C. R. 76. v. S. 916–926. 59–69. A review of hydrogen. Ehrenberg. H. v. v. Javoy. 5547–5558. v. 1991. R.. p. Origin of late dolomite cement by CO2-saturated deep basin brines: evidence from the Ozark region. 581–582. v. v. J. 1981. v. 161–198. A. and T. 1989. nitrogen. Isotopic variations in the Tertiary waters of the northern Rockies and Great Plains as recorded by authigenic calcite (abs. Trimble. Parkhurst. Kaley. v. H. J. P. M. H. 1980. 8. G. and T. an explanation of the post–mid-Cretaceous subsiding of the central North Sea basin: Journal of Geophysical Research. Mattes. p. Major. Oxygen isotope fractionation in divalent metal carbonates: Journal of Chemical Physics. v. L. oxygen. F. Weimer. Rowan. S. 64. 1981. R. v.. D. L. v. p. Marine and continental groundwater sources in a Persian Gulf coastal sabkha. 174–196. Land. 115–121. L. 19. J. H. J. 1977. 65. Chase. 49. and D. Grinenko.Lee and Bethke 237 Review of Physical Chemistry of Japan. H. Oxygen-isotope fractionation in systems containing dolomite: Journal of Geology. Harms. J. Lopatin. Schwarcz.. Colorado: Colorado School of Mines Professional Contribution. B1–B60. H.. 1985. 47. Clayton. B. eds. 1980. Bottinga. D. and T. and E. p. N. p. 69. Dunham. eds. 1989. and R. Weimer. V. Levandowski. Boettcher. p.. E. Jr. p. 1683–1699. Weiss. 764–771. and T. A. and R. N. M. A. v. Jasper National Park. W. 1977. 53–97. Colorado (abs. R Silverman. Anderson. K. Momper. 6. C. 1980. G. p. 110–122. v. 26. Alberta. Denver basin. sulphur. Burial dolomitization of the Upper Devonian Myette Buildup. and an aqueous phase: Geochimica et Cosmochimica Acta. D. New Mexico. Mayeda. and J.): AAPG Bulletin. and Nebraska: Colorado School of Mines Quarterly. Continental stretching. A. A. Isotope geochemistry of regional extensive calcite cement zones and marine components in Mississippi limestones. 65–110. 74. Barnes.. A. E. Walker. 665–686. R. Colorado. 3711–3739. 259–297. 2217–2244. Reefs and related carbonates—ecology and sedimentology: AAPG Studies in Geology 4.. 23... Taylor. p. Martin. 1966. p. Landis. F. B.. P. in D.. S. C. and D. 95–106. McConaghy. Hofstra. W. v. 2136–2139. Boulder County. O. M. V. 381–397. F. Correlation of carbon dioxide abundance with temperature in clastic hydrocarbon reservoirs: relationship to inorganic chemical equilibrium: Marine and Petroleum Geology. Plumlee. v. p. G. and H. 1978. T. and R. Time and temperature in petroleum formation: application of Lopatin’s method to petroleum exploration: AAPG Bulletin. p. I. v. Lohmann. A guide to uppermost Cretaceous stratigraphy. F. Isotope fractionation in the bacterial reduction of sulfate: Research (London). J. sedimentation. and V. H.): Geological Society of America Annual Meeting Abstracts. Cementation in Lyons Sandstone and its role in oil accumulation. Northrop. R. 223–239. Wyoming. D. D. p. N. Malinin. Tectonics. D. 64–92. Cenozoic tectonic history of the Great Plains contrasted with that of the southern Rocky Mountains: a synthesis: The Mountain Geologist. p. 1969. L. north Jamaica: Marine Science Bulletin. Schneidermann and P. K. Lyons Sandstone (Permian). 1908–1925. Lander. p. Clayton. Waples. Leach. The development of reaction models for ground-water systems: Geochimica et Cosmochimica Acta. 1976. D. J. Eolian origin of flagstone bed. Calculation of the thermodynamic properties of aqueous silica and the solubility of quartz and its polymorphs at high pressures and temperature: American Journal of Science. Experimental determination of equilibrium constants for carbon isotope exchange ¯(sol) under hydrothermal conditions: in the system CO2(g)-HCO3 Geochemistry International. 1980. 10. Denver basin: AAPG Bulletin. H. Frost. 3. 85. 1315–1351. 1964.. v. Saunders. central United States: Geology. and P. T. p. M. gases. Sedimentology and tectonic implication of Oligocene Castle Rock Conglomerate. S. Thode. 1973. . G. Helgeson. and J. and R. 17. E. M. Kinsman. J. eds. Calculation of multicomponent chemical equilibria and reaction processes in systems involving minerals.. and K.. Effington. Smalley. p.. Mountjoy. A. 57.. 4. Zenger.. 51. 1965. 54. 1982. p. 41 p. 221–222. L. Temperature and geologic time as factors in coalification [in Russian]: Akademiia Nauk SSSR Izvestiya Seriya Geologiches-Kaya.. Meyers. 1974. and chlorine stable isotope fractionation among gaseous molecules: Annual Review of Earth and Planetary Sciences. J. Viets. northern Denver basin. 1980.. Calculation of chemical equilibrium between aqueous solution and minerals: the EQ3/6 software package: Lawrence Livermore National Laboratory Report UCRL-52658. W.. v. A. 1982. J. 1951. 1983. in S. Patterson. Kropotova. M. 1973. p. 4. D. W. 35–36. v. p. carbon. p.. R. G. T.. Harris.. Colorado: AAPG Bulletin. p. A. O’Neil.. and R. Wolery. p. L. L. Walther. Christie. L. 30. 224 p. Smith. and H. Colorado: Colorado Groundwater Basic Data Report 15. Colorado: deltaic sedimentation. and S. v. C. Thorstensen. Concepts and models of dolomitization: SEPM Special Publication 28. p. The kinetics of silica-water reactions: Geochimica et Cosmochimica Acta. W. 129–135. v. p. The application of oxygen and hydrogen isotope studies to problems of hydrothermal alteration and ore deposition: Economic Geology.. N.