Tying rock properties from core to depositional processes and examining the relationship through forward seismic reflection modeling in the kaiparowits plateau, utah

Nearshore fluvial to tidal transitional depositional systems are becoming increasingly important due to the large number of global hydrocarbon reserves held in such deposits. These deposits are inherently complex due to their heterolithic nature and therefore, interpreting facies and facies relationships in seismic reflection profiles is problematic. The fluvial and tidally influenced nearshore deposits of the late Cretaceous John Henry Member (JHM) of the Straight Cliffs Formation, located in the Kaiparowits Plateau of southern Utah, offers an excellent opportunity to improve our understanding of how the fluvial to tidal transition impacts subsurface petroleum reservoirs and their expression in seismic reflection profiles. The focus of the first chapter is to investigate the impact of heterogeneous depositional environments and their rock properties to model amplitude versus offset (AVO) using a single core. Core EP-25 exhibits lithofacies from a progradational succession, from shoreface through tidal to fluvial. In order to model the most likely lithofacies stacking patterns present in the core, Markov Chain analysis was conducted. Benchtop measurements performed on 1 inch core plugs obtained rock properties (Vp, Vs, density, permeability, and porosity) for each lithofacies. Average rock properties for each lithofacies were used to generate synthetic seismic reflection models of the different upward fining facies associations documented directly from the core, in order to model variations in amplitude versus offset responses as a function of variable tidal influence. The focus of the second chapter is to capture probable 3-dimensional geobody distributions with a particular focus on coal geobody distribution using previously studied cores and outcrops on the plateau. Three different seismic forward models were created ranging in complexity, using cores EP-25, EP-07, density logs, and the nearby outcrop study Left Hand Collet. The rock properties obtained from the benchtop measurements were used to populate the three models based on different depositional environments at the separate depth slices capturing multiple geomorphic rather than stratigraphic models. A seismic survey was acquired on the plateau using 80 Hz frequency; this produced a high-resolution seismic profile. Comparing the forward seismic model to the acquired seismic profile allows for a conceptual understanding between predictive models of what is expected and what is captured in seismic reflection profiles.

TYING ROCK PROPERTIES FROM CORE TO DEPOSITIONAL PROCESSES AND EXAMINING THE RELATIONSHIP THROUGH FORWARD SEISMIC REFLECTION MODELING IN THE KAIPAROWITS PLATEAU, UTAH by Karenth Dworsky A thesis submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Master of Science in Geology Department of Geology and Geophysics The University of Utah August 2015 Copyright © Karenth Dworsky 2015 All Rights Reserved The University of Utah Graduate School STATEMENT OF THESIS APPROVAL The thesis of Karenth Dworsky has been approved by the following supervisory committee members: Lisa Stright Chair 3/25/2015 Date Approved Cari Johnson Member 3/25/2015 Date Approved Tiziana Vanorio Member Date Approved and by John Bartley , Chair/Dean of the Department/College/School of Geology and Geophysics and by David B. Kieda, Dean of The Graduate School. ABSTRACT Nearshore fluvial to tidal transitional depositional systems are becoming increasingly important due to the large number of global hydrocarbon reserves held in such deposits. These deposits are inherently complex due to their heterolithic nature and therefore, interpreting facies and facies relationships in seismic reflection profiles is problematic. The fluvial and tidally influenced nearshore deposits of the late Cretaceous John Henry Member (JHM) of the Straight Cliffs Formation, located in the Kaiparowits Plateau of southern Utah, offers an excellent opportunity to improve our understanding of how the fluvial to tidal transition impacts subsurface petroleum reservoirs and their expression in seismic reflection profiles. The focus of the first chapter is to investigate the impact of heterogeneous depositional environments and their rock properties to model amplitude versus offset (AVO) using a single core. Core EP-25 exhibits lithofacies from a progradational succession, from shoreface through tidal to fluvial. In order to model the most likely lithofacies stacking patterns present in the core, Markov Chain analysis was conducted. Benchtop measurements performed on 1 inch core plugs obtained rock properties (Vp, Vs, density, permeability, and porosity) for each lithofacies. Average rock properties for each lithofacies were used to generate synthetic seismic reflection models of the different upward fining facies associations documented directly from the core, in order to model variations in amplitude versus offset responses as a function of variable tidal influence. The focus of the second chapter is to capture probable 3-dimensional geobody distributions with a particular focus on coal geobody distribution using previously studied cores and outcrops on the plateau. Three different seismic forward models were created ranging in complexity, using cores EP-25, EP-07, density logs, and the nearby outcrop study Left Hand Collet. The rock properties obtained from the benchtop measurements were used to populate the three models based on different depositional environments at the separate depth slices capturing multiple geomorphic rather than stratigraphic models. A seismic survey was acquired on the plateau using 80 Hz frequency; this produced a high-resolution seismic profile. Comparing the forward seismic model to the acquired seismic profile allows for a conceptual understanding between predictive models of what is expected and what is captured in seismic reflection profiles. iv TABLE OF CONTENTS ABSTRACT................................................................................................................ iii LIST OF FIGURES................................................................................................... vii LIST OF TABLES...................................................................................................... ix ACKNOWLEDGEMENTS....................................................................................... x 1 CORE DESCRIPTION, MARKOV CHAIN ANALYSIS AND AVO MODELING OF A TIDAL TO FLUVIAL TRANSITION ZONE IN THE CRETACEOUS STRAIGHT CLIFFS FORMATION, SOUTHERN UTAH, USA.............................................................................................................. 1 1.1 Abstract................................................................................................... 1 1.2 Introduction............................................................................................. 3 1.3 Geologic Background.............................................................................. 5 1.4 Methods.................................................................................................... 12 1.5 Results....................................................................................................... 17 1.6 Discussion................................................................................................ 24 1.7 Conclusions................................................................................................ 31 2 TYING ROCK PROPERTIES FROM CORE TO SEISMIC REFLECTIVITY IN THE KAIPAROWITS PLATEAU, UT, USA................................................... 49 2.1 Abstract................................................................................................... 49 2.2 Introduction........................................................................................... 50 2.3 Geologic Setting....................................................................................... 53 2.4 Methods................................................................................................... 59 2.5 Results....................................................................................................... 62 2.6 Discussion................................................................................................ 66 2.7 Conclusions............................................................................................. 73 FUTURE WORK......................................................................................................... 91 Appendices A. CORE EP-25 ........................................................................................................ 93 B. RAW CORE PLUG LABORATORY DATA 111 C. ROCK PROPERTIES LABORATORY MEASUREMENT METHODOLOGY..................................................................................................... 113 REFERENCES............................................................................................................ 115 vi LIST OF FIGURES Figure Page 1.1. Map of Kaiparowits Plateau and previous studies ..... 34 1.2. Regional stratigraphy and stratigraphic columns of study area ....35 1.3. Lithofacies and core plugs ...36 1.4. Core EP-25 logs including depositional environments, architectural elements, and lithofacies ..37 1.5. Markov Chain transitional statistics ...38 1.6. Lithofacies thickness box and whiskers plot ..39 1.7. Facies associations columns ..40 1.8. Rock plots................................................................................................. .........41 1.9. Architectural element plot with averages ...42 1.10. Channel sandstone porosity versus Ip plot and the relationship to marine influence............................................................................................................ 43 1.11. Facies associations AVO modeling ..44 1.12. 1-dimensional AVO modeling of core EP-25 ....45 2.1. Map of Kaiparowits Plateau including study area ....75 2.2. Regional stratigraphy and stratigraphic column of study area ......76 2.3. Lithofacies with core plug photographs and locations .....77 2.4. Core EP-25 and EP-07 depositional environment, architectural element, and lithofacies logs............................................................................................78 2.5. Modern analogs for building the architectural element 3-dimensional model............................................................................................................ 79 2.6. Seismic acquisition profile and core locations........................................ 80 2.7. Rock plots of depositional environments, architectural elements, and lithofacies.......................................................................................... 81 2.8. 2-dimensional cross-sections between EP-25, EP-07, and Left Hand Collet................................................................................................................ 82 2.9. Coal geobody cross-section........................................................................ 83 2.10. Three 3-dimensional models..................................................................... 84 2.11. Architectural element 3-dimensional model modern analogs................... 85 2.12. Seismic profile and interpreted depositional environments and seismic features......................................................................................... 86 2.13. Three 2-dimensional seismic forward models at seismic acquisition profile............................................................................................................. 87 viii LIST OF TABLES Table Page 1.1. Facies associations, environmental description, architectural elements and lithofacies................................................................................................ .......... 46 1.2. Markov chain analysis of lithofacies....................................................... ........47 1.3. Rock properties table........................................................................................48 2.1. Geometries from Left Hand Collet ..... 88 2.2. Acoustic impedance of EOD and architectural elements .... 89 2.3. Impedance contrast and seismic expression of EOD and architectural elements ...90 ACKNOWLEDGEMENTS I would first like to thank my advisor, Lisa Stright. Not only did she believe in me from the very beginning, but he has provided so many fantastic opportunities that have enabled a more than successful graduate school experience, in addition to helping build the foundation for a successful future. I would also like to thank my committee members, Cari Johnson and Tiziana Vanorio; Cari for being a solid foundation of geologic knowledge concerning not only the Kaiparowits Plateau, but sedimentary geology in its entirety, and Tiziana and the rest of the Stanford Rock Physics Laboratory for support and donated time to making sure I understood the equipment I worked with and the theory and concepts behind them. The Rocks to Models consortium have also provided invaluable discussion and ideas throughout the course of my work, including William Gallin, Patrick Dooling, Brenton Chetnik, Julia Mulhern, Wassim Bellinham, Alexandre Turner, and Tyler Szwarc. I would also like to acknowledge the companies who support the consortium without whose support the work could not have been accomplished. I would also like to express gratitude towards the helped I received from undergraduate students from the University of Utah, Utah Valley University, and Brigham Young University during the seismic acquisition. I would like to especially thank Pier Paolo Bruno for designing the seismic acquisition survey and the INGV for providing processing facilities. I would like to thank the UGS, in particular Peter Nielsen, for all the technical help digitizing logs, and being incredibly helpful when it came to the core description and core plug extraction process. I am appreciative to the BLM for allowing me to conduct fieldwork in the beautiful Kaiparowits Plateau. Lastly I would like to personally thank my partner Gavin Dixon for his comfort, patience and unwavering support throughout my entire graduate school experience. xi CORE DESCRIPTION, MARKOV CHAIN ANALYSIS AND AVO MODELING OF A TIDAL TO FLUVIAL TRANSITION ZONE IN THE CRETACEOUS STRAIGHT CLIFFS FORMATION, SOUTHERN UTAH, USA Abstract Deposits within tidal successions are generated by a complex mixture of processes from fluvial input to tidal reworking, both impacted by shoreline transgressions and regressions. The resulting subtle changes at the bed scale in the transition from tidal to fluvial deposits make lithofacies differentiation from subsurface wireline log and seismic reflection data problematic. Because reservoirs comprised of tidally influenced deposits account for a significant portion of petroleum reserves, forward seismic reflection modeling coupled with a predictive facies model framework derived from core and outcrops can lead to invaluable insights for interpreting subsurface data. The goal of this research is to assess to what extent the fluvial and tidal facies associations can be distinguished in amplitude versus offset (AVO) forward models using lithofacies stacking patterns, depositional interpretations, and direct rock properties. The Cretaceous John Henry Member (JHM) is exposed along the edges of the Kaiparowits Plateau in southern Utah, revealing excellent outcrop exposures of fluvial and tidally influenced deposits. These exposures, coupled with core and wireline log data 2 in the center of the plateau, present an exceptional opportunity to improve our understanding of wireline log interpretation and seismic imaging in similar subsurface petroleum reservoirs. The focus of this study is centered on core EP-25 (240 m), located in the north-central Kaiparowits Plateau. The core captures a progradational succession from shoreface through tidally influenced lagoon to fluvial. Transition probabilities between lithofacies observed in the core were quantified with a Markov Chain analysis, resulting in seven complete upward fining packages (facies associations), and 34 incomplete packages. These facies associations represent end members from tidally influenced to fluvially influenced deposits and transitional packages between the two end members. Benchtop measurements were performed on 1 inch core plugs to obtain rock properties (Vp, Vs, density, permeability, and porosity) for each lithofacies. The resulting rock property exhibits a wide range of values as a direct result of the highly heterolithic nature of these deposits. As expected, significant overlap between fluvial and tidal rock properties is observed due to the transitional nature of the depositional environments. However, there is a distinguishable difference between clusters of the tidal and fluvial groups. Average rock properties were assigned to lithofacies comprising the seven facies associations and 1-dimensional amplitude versus offset (AVO) predictions were generated to elucidate near- to far-offset amplitude changes as a function of variable tidal influence. A near-offset amplitude decrease is observed in the transition from more fluvial- to tidally influenced depositional environments. 1-dimensional AVO model was then performed on the full length of the core using velocities and densities from the core plug measurements. Not only does the core also distinguish between reflectors in tidal and fluvial in the offset plot, but exhibits offset in the far angle based on increasing 3 marine influence with depth. Introduction A number of studies have recognized the importance of understanding the impact of the fluvial-marine depsitional processes and their intricate link to reservoir quality amidst increasing tidal influence (Shanmugam et al., 1993; Nordahl et al., 2005; Longhitano et al., 2012). Core studies can not only be used to capture vertical stacking patterns (Powers and Easterling, 1982; Ahmad et al., 2012), but they also aid in the recognition of heterolithic lithofacies as a key indicator for interpreting depositional environment and stratigraphic architecture (Yoshida et al., 1999; Martinius et al., 2001). A significant amount of work has been done on describing tidal depositional environments and the direct link between repeating stacking patterns of lithofacies (termed parasequence by some authors) and their seismic-reflectivity character in both modern (Fenies et al., 1998; Yoshida et al., 2001; Chakrabarti, 2005) and ancient environments (Tanavsuu-Milkeviciene and Plink-Bjorklund, 2009; Feldman et al., 2014). These studies have focused primarily on the 1- and 2-dimensional seismic expression of geobodies. Although these studies are useful in interpreting geobody distribution, what they lack is the ability to identify the impact of depositional processes on the deposits within the fluvial to tidal transition zone, and their effect on seismic reflectivity, and more specifically, amplitude response with offset (AVO). Forward seismic-reflectivity modeling is a valuable tool, particularly when used to predict seismic-reflectivity responses as a function of lithofacies changes in varying depositional environments. There have been a number of forward seismic-reflectivity modeling studies that focus on offshore sedimentary rocks (Christensen and Szymanski, 1991; Vernik and Nur, 1992; Falivene et al., 2010; Tetyukhina et al., 2010; Howell et al ., 2014; Stright et al., 2014), as well as tidal sedimentary rocks (Wen et al., 1998; Yoshida et al., 1999; Hodgetts and Howell, 2000; Yoshida et al., 2001; Tetyukhina et al., 2014), although very few have concentrated on the fluvial to tidal transition zone (Martinius and Gowland, 2011). A majority of these studies have used seismic rock properties either from analog core plug measurements, wireline logs from analogs, or theoretical rock property modeling. The first rock property models were derived from empirical relationships between acoustic and elastic wave velocities as a function of pressure, clay content, and porosity, and calibrated to benchtop measurements (Han, 1986; Eberhart-Phillips et al., 1989). Following the anaysis of empirical rock property relationships, a number of studies investigated the rock properties from specific environments of deposition using either direct or modeled rock properties (Biddle et al., 1992; Falivene et al., 2010). Although these studies have characterized and modeled rock properties based on instrinsic properties (pressure, clay content, porosity, and acoustic and elastic wave velocities), there are few studies that couple the rock properties tightly with bed-scale depositional processes and none specifically in the fluvial to tidal transition zone. The goal of this research is to assess to what extent the fluvial and tidal facies associations can be distinguished in bed- to log-scale AVO forward models using rock properties measured directly from core, lithofacies stacking patterns, and depositional environment interpretations. In order to accomplish this, core description within the John Henry Member of the Straight Cliffs, Utah is presented within a hierarchical framework. 4 Stacking patterns of fundamental bed-scale observations (lithofacies) are quantified using Markov Chain Analysis to calculate transition probabilities and characterize lithofaci e s stacking patterns (facies associations) to aid in characterization of depositional environments. The resulting facies associations represent deposits from tidal and fluvial end members and the intermediate transitional packages. AVO forward modeling was generated to reveal near- to far-offset amplitude changes as a function of variable tidal influence by employing direct rock property measurements. Coupling geological observations from core to their seismic response, by placing them within a hierarchical framework, enables core plug rock property measurements to be tied to depositional processes and upscaled in order to analyze the seismic amplitude in fluvial to tidal transition zones. Geologic Background Regional Geology Sediments preserved in the Upper Cretaceous Straight Cliffs Formation of the Kaiparowits Plateau were deposited in an asymmetrical foreland basin formed by loading from the Sevier fold and thrust belt to the west (Kauffman and Caldwell, 1993; Allen and Johnson, 2010a) (Fig. 1.1). The siliciclastic sediments that form the Kaiparowits Plateau were sourced from three locations: the Sevier fold and thrust belt to the west, the Mongollon Highlands to the south, and the Cordilleran Volcanic Arc to the southwest (Peterson, 1969a; Eaton and Nations, 1991; Hettinger, 1995a; Szwarc et al., 2014). Paleoshorelines were oriented northwest-southeast (Peterson, 1969a; Shanley and McCabe, 1991; Shanley et al., 1992; Hettinger, 1995a; Allen and Johnson, 2010a). The 5 6 easternmost outcrop exposures along the plateau comprise offshore through intertidal facies and represent the paleoshoreline of the Western Interior Seaway in southern Utah (Shanley and McCabe, 1991). Stratigraphy Peterson (1969) divided the Straight Cliffs Formation in the Kaiparowits Plateau into four members: the lower cliff-forming Tibbet Canyon Member, the slope and ledge forming Smoky Hollow Member and John Henry Member (JHM), and the upper cliff-forming Drip Tank Member. The JHM, the focus of this study, is the thickest (200-500 meters) and early Coniacian to late Santonian in age (~88 to 83.5 Ma) (Fig. 1.2) (Peterson, 1969a; Eaton, 1991; Hettinger et al., 1993; Szwarc, 2014). The base of the JHM is marked by a landward shift in facies recording a transgression that occurred after deposition of the Smoky Hollow Member (Shanley and McCabe, 1991). In the southern, western, and northeastern Kaiparowits Plateau, the upper JHM consists primarily of multi-story and single-story fluvial channel belts interbedded with carbonaceous floodplain mudstones and coal (Gooley, 2010; Pettinga, 2012). The center of the plateau contains paludal deposits with thick coal beds and are interstratified with tidal deposits (Vaninetti, 1979). Seven shoreface sandstone packages were identified at Left Hand Collet in the central part of the plateau along the paleoshoreline, ranging from wave dominated lagoon to tide dominated barrier island system (Peterson, 1969a; Dooling, 2012). There are three coal zones present in the JHM: Lower Christensen, Rees, and Alvey coal zones (Fig. 1.2). These zones were interpreted by Shanley and McCabe (1992) as being deposited in steep raised mires. The preservation of transgressive lagoon 7 and complete progradation (regressive) shoreface successions indicate that accommodation and sedimentation rates were moderately high throughout depositi on of the entire JHM (Allen and Johnson, 2010b). The northeastern part of the plateau has been examined to a lesser degree than the southern part of the plateau. Left Hand Collet and highway 12 outcrops are the only outcrop exposures that have been studied thus far (Shanley et al., 1992; Dooling, 2012; Chentnik et al., 2014). In the northern-central section of the plateau, data are limited to 23 core that were extracted by Utah Power and Light (UP&L) in the early 1970s in order to investigate coal in the region. One particular core intersects the entire JHM, core EP-25. Gallin (2010) logged and interpreted this core to provide valuable information regarding the extent of depositional environments across the central Kaiparowits Plateau. Gallin's (2010) core analysis offered a benchmark on which to support further work in the northeastern part of the plateau. Core Description Core EP-25 is a 240 m thick cored interval, logged in the northeastern Kaiparowits Plateau, Utah that intersects the entire John Henry Member (Table 1.1). The smallest scale of observation, lithofacies, was subsequently used to interpret the different depositional environments and their associated architectural elements. Within core EP- 25, eight different lithofacies were identified and described at a core plug scale (1-20 cm) (Gallin, 2010) (Fig. 1.3). The lithofacies are defined by grainsize, bed features, bioturbation, laminations, and organic content. Gallin also noted an overall trend that reveals a transition from marine to fluvial influence moving up in the core. Based on 8 fossils assemblages, coal, amount of bioturbation, and lithofacies distribution, three depositional environments were identified: coastal plain (fluvial), tidal, and shoreface (Hettinger, 1995a; Gallin, 2010). The lithofacies are described below by Gallin's observation, categorized into their interpreted architectural elements, and then grouped into depositional environments (Fig. 1.4). Core Description: Lithofacies, Architectural Elements, and Depositional Environments Carbonaceous Shale This lithofacies is dark gray to black colored shale. It is often laminated but can be structureless. Abundance of organic material is present, comprised primarily of plant fragments; however, brackish water corbulids and other small gastropods are also present. Vertically, carbonaceous shale often grades into coal. This lithofacies is interpreted as being low energy lagoonal mudstone within a tidal depositional environment. Bioturbated Mudstone This lithofacies consists of horizontally laminated mud or mud with no laminations and varying degrees of bioturbation. The bioturbation index used to quantify the amount of bioturbation uses a scale which ranges in amount of disturbance to original bedding structures from 0-100%, with 100% being complete obliteration of original sedimentary structures (Droser and Bottjer, 1988). The bioturbation in the tidal mudstone tends to be filled with very fine to fine grained sandstone. Organic material in this 9 lithofacies consists of plant fragments, leaf imprints, bivalve, and gastropod shells. Vertically, this lithofacies often grades into carbonaceous shale and coal. This l ithofacies is interpreted as being low energy, represented in both the tidal and coastal plain depositional environments. In the tidal depositional environment, bioturbation is heaviest and there is an abundance of tidal environment trace fossils (Teichichnus, Thalassinoides, Ophiomorpha, Asterosoma). This lithofacies represents the architectural elements tidal mudstone and overbank fines in the respective depositional environments. Flaser/W avy/Lenticular Bedded Sandstone This lithofacies reveals alternating laminations of sandstone and mudstone. Grainsize ranges from very fine to medium, while the mud fraction consists of siltstone, shale, or carbonaceous shale. Flaser mud drapes sometimes occur as accumulations of carbonaceous material. This bedding style is indicative of bi-directional flow and tidal influence (Finzel et al., 2009). Intervals comprised of flaser/wavy/lenticular bedforms are interpreted as deposits from tidal and coastal plain depositional environments. Although this lithofacies is associated mainly with tidal influence, it can be found seen in the fluvial environment as likely being located on the edge of the transition zone between fluvial and tidal environments. The architectural elements include tidal mudstone and overbank fines. Planar Bedded Sandstone This lithofacies is composed of fine to medium grained sandstone. Laminations observed in the sandstone dip < 5°. This lithofacies is associated with high-energy flow 10 regimes and is present in all three depositional environments. In the fluvial depositional environment, this lithofacies represents part of upward fining, fluvial channel architectural element. In the tidal depositional environment, this lithofacies represents highly sinuous distributary tidal channel architectural elements. In the shoreface depositional environment, this lithofacies represents well-sorted shoreface sheet sandstone clean on mud draping and may constitute hummocks or swales. In this environment, large robust bivalve fragments are preserved including Granocardium. Trough Cross-bedded Sandstone This lithofacies is comprised of fine to coarse grained sandstone. The trough and planar tabular sets range between 1- 5 mm thickness and the foresets dip between ~5- 40°. This lithofacies represents unidirectional flow and is present in all three depositional environments. In the fluvial depositional environment, this lithofacies represents part of upward fining, low sinuosity fluvial channel architectural element. In the tidal depositional environment, this lithofacies represents the landward end o f a highly sinuous distributary tidal channel architectural element. In the shoreface depositional environment, this lithofacies represents well-sorted shoreface clean sheet sandstone ripples and hummocks with no mud draping as the shoreface architectural element. The trough cross-stratification may constitute hummocks and swales in the shoreface environment. 11 Structureless Sandstone This lithofacies is composed of very fine to coarse grained sandstone. Sedimentary structures are not apparent. Occasional bivalve shell fragments are present. This lithofacies is present in all three depositional environments. In the fluvial depositional environment, this lithofacies represents part of the fluvial channel architectural element. In the tidal depositional environment, this lithofacies represents the tidal channel architectural elements. In the shoreface depositional environment, this lithofacies represents well sorted clean shoreface architectural element. Mud Rip-up Sandstone This lithofacies reveals grainsizes that range from very fine to coarse. Within the sandstone matrix, angular to subangular, flat to moderately rounded mud rip-up clasts are present. The clasts are thin (< 1 cm) flakes. Vertically, the grainsize in this lithofacies coarsens upward over 10's of cm. This lithofacies is interpreted as high energy channel lag deposits within a fluvial depositional environment. More specifically, mud rip-up clasts are interpreted as being located at the base of the fluvial channel architectural element that cut through the overbank fines and tidal mudstone. Coal This lithofacies is black with dull vitreous reflectance and lacking distinct laminations. It is often accompanied by surficial and interstitial sulfur precipitates. Typically, this lithofacies is underlain by bioturbated mudstone or carbonaceous shale. The coal seams are < 1 m thick. This lithofacies is interpreted as being present 12 exclusively in the tidal depositional environment. Although the coal is associated with the tidal depositional environment, it is believed to have formed in raised coal mire s that persisted for long periods of time with limited input from either fluvial or marine realms (Hettinger, 1995). Vertically, the coal tends to become thinner and less frequent upward. This decreasing frequency of the coal and increase in grainsize of sandstone lithofacies upward in the core is tied to an overall decrease in marine influence from the base of the core to the top. Detailed core description and observations at the lithofacies core plug scale provide a hierarchical framework which can be subsequently upscaled and tied to depositional processes. The upscaled lithofacies and their rock property measurements are analyzed in order to gain a better understanding of the seismic amplitude in fluvial to tidal transition zones. Methods Prediction between scales, from lithofacies to stacked geobodies (core to reservoir scale), is accomplished by placing the observations within a hierarchical framework. A Markov Chain Analysis (MCA) is used to determine the 1-dimensional transition probabilities of lithofacies to quantify larger scale stacking patterns as facies associations. Rock property measurements from core plugs are used to generate AVO forward models of the facies associations to test if there is a near- to far-offset amplitude signature that can aid in predicting the amount of tidal influence, and therefore, reservoir quality from seismic reflectivity. 13 Facies Associations Stacking Patterns In order to determine whether the lithofacies making up the facies associ ation show a preferred stacking or regular vertical arrangement, a MCA was performed on the core observations. The MCA quantified the probability of transitioning from one lithofacies into another and therefore captures probable stacking patterns (Ahmad et al., 2012). The rationale of calculating transition probabilities is that an unbiased analysis of the repeatable stacking patterns can be captured from the core. The MCA was performed over the entire core with a transition count matrix where all possible vertical lithological transitions were tabulated at a 0.1 ft. sample rate (Table 1.2). These were converted to transition probabilities by normalizing the counts for each of the lithofacies by the total counts. To characterize fully preserved depositional packages, only complete facies associations were extracted from the core based on three criteria. The three criteria of a full facies package were: 1) The facies associations had to include at least three consecutive, statistically related lithofacies. No facies association was selected if it contained only two lithofacies, for example, only carbonaceous shale and coal. The rejection of packages that did not have at least three stacked lithofacies takes into account the fact that the association may not be directly penetrating full architectural elements. This eliminates the packages that occur in the marginal position of the architectural elements and therefore do not represent a full depositional sequence. 2) Each of the lithofacies had to be at least as thick as the average bed thickness for each lithofacies. For example, if a facies association was comprised of trough cross-bedded sandstone, planar bedded sandstone, bioturbated mudstone, carbonaceous shale, and coal, but all or one of the thicknesses was less than the lowest standard deviation, the association was disregarded. 3) The bases of each facies association were assumed to be comprised of either mud rip-up sandstone or trough cross-bedded sandstone. Using the three criteria for building the facies association, in conjunction with the MCA results, seven facies associations were ultimately selected to conduct AVO modeling. Rock Properties To investigate the rock properties of the eight lithofacies 60, 1 inch diameter by 1 inch vertical, perpendicular to bedding core plugs were collected from the EP-25 core. The only lithofacies not represented by core plugs was the coal. Measurements for rock properties of coal were not attempted due to its friability and many of the coal intervals were missing from the core as a result of sampling from historical coal exploration work. Core plug benchtop measurements (at room temperature (68°F) and atmospheric pressure) were conducted for porosity, density, compressional wave (Vp), and shear wave velocities (Vs). The samples were prepared by first sanding the top and base to get a flat surface to ensure 1) good contact with the electrodes, which uses piezoelectric crystals to generate a high voltage electric pulse into the core plug to measure velocity and 2) accurate length measurements for bulk volume calculations. Prior to testing the samples were dried for 72 hours to remove all moisture. Dry rock density was calculated from 14 15 bulk volume and mass measurements. Porosity was measured using a porosimeter. Vp and Vs velocities were measured using an ultrasonic velocimeter. Laboratory-derived measurements of Vp, Vs, and density were used to generate average P-impedance (Ip) and S-impedance (Is) for each lithofacies. These relationships were interrogated to elucidate the relationship between lithofacies and the measured rock properties. Clay content was calculated by estimating a percentage of shale (VShale) from the preexisting gamma ray log. Average values for each lithofacies within each depositional environment were used for the forward modeling to compare the impact of packaging of the lithofacies (facies associations) without introducing variability (noise) from rock property measurements. Cross plots of Ip versus Vp/Vs colored by lithofacies, architectural elements, and depositional environments were used to elucidate the impact of depositional processes and environment controls on the seismic response. In order to analyze net-to-gross (NTG) and grainsize, the core plugs were visually examined with a hand lens. The core plug values were then plotted on the Ip versus Vp/Vs cross plot to identify possible patterns and ties to lithofacies, architectural elements, and depositional environments. 1-Dimensional Synthetic Seismic Modeling 1-dimensional synthetic seismic models were generated using the seven facies association columns obtained from the MCA and populated with the average rock property values of lithofacies from the different environments of deposition (Table 1.2). The background rock property used was the overall average values for the bioturbated mudstone in the respective depositional environments (7022 Ns/m for coastal plain and 16 3 7600 Ns/m for tidal). Angle-dependent reflectivities were then calculated for each facies association column using the Zoeppritz equation (Zoeppritz, 1919) to capture sei sm ic reflectivity as a function of incidence angle (Castanga, 1992). The reflectivity series was subsequently convolved with a 25-Hz, zero phase, + 90° rotated, Ricker wavelet with an offset range from 0 to 30 degrees. To better understand the imaging of the stacked facies association packages at the well log to seismic scale, a 1-dimensional synthetic trace was generated of the full core EP-25. All 60 measurements of Vp and Vs were used for this modeling instead of the average values. The only logs associated with core EP-25 available were gamma gamma density and gamma ray logs. Vp and Vs logs were generated using sequential Gaussian simulation (sGs) for each lithofacies leveraging values measured from core plugs as hard data. The coal lithofacies used a single value for the Vp (2000 m/s) and Vs (1200 m/s) rather than a simulation (Morcote et al., 2010). Once the seven different lithofacies logs were combined using values extracted from each lithofacies simulation at their exact locations in the core at 0.5 ft. steps. The logs were spliced together to create full 1D logs that represents the eight different lithofacies throughout the core. A pre-existing density log from when the core was originally drilled was corrected to match the density values of sandstone bodies from the rock property measurements because the logs were initially calibrated to the coal values when originally cored; however, since the rock property measurements demonstrate the accurate density values, the log was recalibrated to the sandstone bodies. Using the velocities (Vp and Vs), and density log, Ip, and Is were calculated. The reflectivity series was convolved with a 25-Hz zero phase +90 rotated Ricker wavelet with an offset range from 0 to 30 degrees. 17 Results Facies Association Stacking Pattern The combination of the MCA with the cutoff criteria produces seven complete facies associations based on the three criteria of containing at least three lithofacies, lithofacies thickness that falls within the standard deviation and underlain by mud rip-up and trough cross bedded sandstone. These facies associations were subsequently categorized as either tidal or fluvial. One key indicator of depositional environment was the capping lithofacies of the packages. The facies associations that are capped by carbonaceous shale and coal are typical of tidal depositional environment, whereas facies associations that are capped by bioturbated mudstone exclusively, without carbonaceous shale or coal present, is indicative of the overbank fine architectural element, pointing to a fluvial depositional environment. Another indicator of depositional environment is the base lithofacies of each package, where, when present, mud rip-up sandstone is interpreted to be associated with fluvial channels (Table 1.2). These interpretations corroborate Gallin's (2010) interpreted depositional environments. The lithofacies stacking patterns observed within the facies associations indicate that there is a range of distinct stacking patterns in both the fluvial and tidal environments of deposition (Fig. 1.5). The MCA results illustrate which lithofacies are most likely to stack on top of one another in a statistically significant pattern. These stacking patterns and their measured bed thicknesses (Fig. 1.6) comprise the facies associations that were observed within interpreted environments of deposition both of complete and incomplete packages (Fig. 1.7). The results for one of the fluvial facies association (fluvial 1) is comprised of a 18 base of mud rip-up sandstone to trough cross-bedded sandstone (55 %), capped by bioturbated mudstone (14%). The most likely fluvial facies association based on (fluvial 2) stacking pattern that was extracted from the core is a mud rip-up base followed by trough cross-bedded sandstone (55%) followed by planar bedded sandstone (50%), flaser/wavy/lenticular bedded sandstone (52 %), capped by bioturbated mudstone (84 %). There are five MCA results for the tidal facies association likely due to the fact that it is a much thicker depositional environment in the core. The tidal facies associations do not exhibit mud rip-up sandstone in this environment; the base in each of the five facies associations is trough cross-bedded. Tidal 1 has a base of trough cross­bedded sandstone transitioning to planar bedded sandstone (45 %), carbonaceous shale (12 %), and capped by coal (100 %). Tidal 2 has a base of trough cross-bedded sandstone transitioning to planar bedded sandstone, bioturbated mudstone (42 %), and capped by coal (22 %). Tidal 3 has a base of trough cross-bedded sandstone transitioning to flaser/wavy/lenticular bedded sandstone (36 %), bioturbated mudstone (84 %), carbonaceous shale (34 %), and capped by coal (100 %). Tidal 4 has a base of trough cross-bedded sandstone transitioning to planar bedded sandstone (50 %), flaser/wavy/lenticular bedded sandstone (46 %), carbonaceous shale (16 %), and capped by coal (100 %). Tidal 5 has a base o f trough cross-bedded sandstone transitioning to planar bedded sandstone (50 %), flaser/wavy/lenticular bedded sandstone (46 %), bioturbated mudstone (84 %), carbonaceous shale (38 %), and capped by coal (100 %). Tidal 5 facies association is comprised of lithofacies that represent the most likely stacking based on percentages. This association is assumed to represent the most representative end member in a tidal environment. Based on the most likely stacking 19 pattern in the fluvial environment, the representative end member would be something in between fluvial 1 and fluvial 2. The facies associations capture not only likely statistical transition between lithofacies, but also show distinct patterns linking each package to depositional processes. All of the facies associations contain the trough cross-bedded sandstone lithofacies, which indicates the presences of either fluvial or tidal channels. The planar bedded sandstone points to a high velocity flow, the flaser/wavy/lenticular bedded sandstone illustrates bi-directional flow which suggests tidal influence, bioturbated mudstone indicates a quiet environment with little disturbance from flow, and both coal and carbonaceous shale represent organic input and stagnant depositional conditions representative of paludal depositional environments. Between all the facies association, the columns can be differentiated by the absence or inclusion of the following lithofacies; 1) facies associations containing planar bedded sandstone (in 5 of 7 facies associations, both in tidal and fluvial); 2) facies associations with flaser/wavy/lenticular bedded sandstone (in 4 of 7 facies associations, both in tidal and fluvial) and; 3) facies associations capped by carbonaceous shale and coal (in all tidal facies associations). Rock Properties The rock properties show a wide range of values as a direct result of the highly heterolithic nature of these deposits. The distribution of rock property data for each lithofacies sorted by depositional environment is presented in Table 1.3. The values derived from laboratory measurements including Vp, Vs, density, and porosity, are 20 cataloged into the three depositional environments. Three samples with a Vp values less than 1500 m/s were removed from further calculations because they were believed to be bad samples due to these unrealistically low values. The error in the remainder of the measurements is less than 1% for all tests. The standard deviation for Vp, Vs, and density is low for the mud rip-up clast sandstone. Although there are not enough data to calculate a standard deviation for the planar bedded sandstone, values for this lithofacies has the largest range the maximum and minimum values for Vp and Vs. The trough cross-bedded sandstone in both fluvial and tidal depositional environments exhibit similar values. In the tidal environment, the highest standard deviation for Vp and Vs is the planar bedded sandstone 940 m/s and 618 m/s, respectively. The lowest standard deviation for Vp is flaser/wavy/lenticular bedded sandstone; however, it also has the highest standard deviation for porosity (275 m/s and 6.2 %, respectively). The lowest standard deviation for Vs and porosity is carbonaceous shale (185 m/s and 2.7 %, respectively). The other lithofacies all have similar standard deviations in the middle for both velocities. The wide range in values emphasizes the complexity in interpreting the seismic expression of these highly heterolithic depositional environments. It is difficult to identify velocity and density as separate variables from seismic profiles; however, Ip and Is can be easily obtained. Therefore, a common way to identify velocity and density is to plot Ip and Is as a ratio to derive Vp/Vs. The general trends observed in the rock properties cross plot of Ip versus Vp/Vs is increasing Ip and decreasing Vp/Vs as a function of decreasing grainsize, decreasing net-to-gross, and decreasing porosity (Fig. 1.8D, 1.8E, and 1.8F, respectively). Although there is an inverse relationship between grainsize and Ip, the siltstone, and mudstone are difficult to distinguish from one another (Fig. 1.8D). A general trend between decreasing n et-to - gross and increasing Ip is clear; however, there is a considerable amount of overlap in the high Ip with some higher net-to-gross exhibiting higher Ip (Fig.1.8E). In addition, there is a clear inverse relationship between porosity and Ip (Fig. 1.8F); however, where the porosity values have values of less than < 5 %, the porosity plots show a distinct positive slope. The inverse trend is also evident in the percentage of clay illustrating an increase in Ip with an increase in clay content (Fig. 1.8G). Poorly-sorted samples tended to have a high Ip, while the well-sorted tended to have a lower Ip (Fig. 1.8H). Furthermore, sandstone lithofacies in general are characterized as high net-to-gross, high porosity, large grainsize, and low Ip (Fig 1.8A). One lithofacies in particular that is an outlier to these general trends is planar bedded sandstone which exhibit larger grainsize and net-to-gross, but high Ip and low porosity. Bioturbated mudstone and carbonaceous shale have low net-to-gross, low porosity, small grainsize, and high Ip. Tidal and coastal plain environments are clearly differentiated with the tidal elements having higher average values of Vp/Vs and Ip (Fig. 1.8B). There is a v-shape apparent in the plots showing low Ip from fluvial and tidal channels, demonstrating a poorly correlated negative slope, and high Ip architectural elements which correspond to the tidal and fluvial mudstones and exhibit a distinct positive slope (Fig. 1.8B). The positive slope, high Ip trend in the rock properties plot represents the tidal mudstone and overbank fines. Tidal mudstone and the fluvial overbank fines are distinguishable as shown by an offset in the averages which indicates that the tidal mudstone has a higher Vp/Vs and Ip compared to the overbank fines at, 0.08 21 22 and 11100 (Ns/m ), respectively (Fig. 1.9). Although the tidal mudstone and overbank fine elements have low porosity, grainsize, and net-to-gross, the tidal mudstone in particular have varying amounts of bioturbation while the circled points (overbank fine), illustrate no bioturbation (Fig. 1.8C). The plot colored by architectural elements illustrates a strong distinction between the different types of sandstone beds (interpreted as channels) in the Vp/Vs, where the tidal mudstone and overbank fines are more easily distinguished by the Ip spread (Fig. 1.9). In the channel sandstone, the tidal Vp/Vs is 0.04 and the Ip is 2000 (Ns/m ) which is greater than the fluvial sandstone which has a Vp/Vs of 0.35 and Ip of 5000. In order to investigate different ways to distinguish the channel sandstone elements, the lithofacies that comprise the sandstone elements were plot as Ip versus porosity (Fig. 1.10). The lithofacies that comprise channel sandstone are planar bedded sandstone (D, Fig. 1.3), trough cross-bedded sandstone (E, Fig. 1.3), structureless sandstone (F, Fig. 1.3) and mud rip-up sandstone (G, Fig. 1.3). The core plugs are separated into two groups based on marine influence moving up in the core: tidally influenced channels, and fluvial channel sandstone at the top of the core. Employing tidal and fluvial channel sandstone lithofacies investigates the differences in the channels based on the channel architectural elements and core depth (Fig. 1.10). The end member, fluvial channel sandstone has high porosity values and low Ip while the tidally influenced channel sandstone demonstrates low porosity and high Ip values (Fig. 1.10a). Based on the depth plot, the tidal channel samples that are clustered in the low porosity zone correlate to the upper most section of the tidal environment (310-340 m). The mid-range porosity and Ip samples in the tidal environment correlate to the middle section of the 3 23 tidal environment (340-370 m), while the base of the tidal environment correlat es to high Ip and low porosity channels (370-400 m) (Fig. 1.10B). Recalling that core EP-25 displays increasing marine influence moving down in depth the plot Ip versus porosity aids in highlighting the distinction between the more marine influenced channelized sandstone and the fluvial influenced channel sandstone (Fig. 1.10C). Forward Seismic Modeling The maximum positive amplitude at zero-phase was used as the location across which to analyze offset amplitudes (Fig. 1.11). In the near-far amplitude versus offset plot, fluvial facies associations have higher amplitude for the full range of offsets in comparison to the tidal facies associations which have lower amplitudes. The key difference between the tidal and fluvial facies associations is the presence of coal and carbonaceous shale in the tidal facies association and the mud rip-up clast sandstone at the base of the fluvial. In contrast to the modeling of single facies association packages with a mudstone background, AVO modeling of the full core EP-25 represents multiple facies associations stacked on top of one another with the measured and interpolated rock properties rather than averages (Fig. 1.12). The AVO model of the entire core illustrates some similarities and differences to the facies association AVO models. In the offset plot, the maximum amplitude fluvial coastal plain amplitude reflection is also larger than the tidal zone amplitude reflection. Facies associations containing planar bedded sandstone are observed in both fluvial and tidal facies associations, and have a large spread in values which is muted in the individual facies association modeling due to the use of average 24 values. The impact is a large range of resulting AVO responses. The Drip Tank Member, at the top of the core, is a gravelly sheet comprised of high Ip sandstone. When juxtaposed against the low Ip overbank fines and fluvial channels in the upper JHM, an amplitude reflection is generated that is unique compared to the other wavelets in the core; this wavelet has a more significant decrease in amplitude in the far offset. The second pattern that emerges from the AVO analysis of the facies associations is observed in the amplitude reflection coefficient versus offset plot (Fig. 1.12). The fluvial facies associations have higher amplitude reflections compared to the tidal facies association in the near and the far angles. However, there is more obvious distinction between the fluvial and tidal facies associations in the near angle; the distinction is much less apparent in the far angle. The difference between the tidal and fluvial is due to the fluvial associations having a base of low Ip and capped by bioturbated mudstone that has a lower Ip than the tidal association bioturbated mudstone. The inclusion of the high Ip carbonaceous shale and low Ip coal is also a significant distinguishing feature between the tidal and fluvial associations that is revealed in both the rock properties and subsequently the AVO analysis. Discussion Interpreting Degree of Tidal Influence from Lithofacies Stacking Patterns Predictive, probabilistic stacking patterns at the lithofacies scale using MCA is an advantageous way of recognizing repeatable and statistically significant stacking patterns in vertical successions. Due to the complexity of tidally influenced deposits, stacking patterns predicted from a MCA indicate nuances which help better constrain interpretations and place measured rock properties within an architectural framework. The fluvial and tidal facies associations illustrate variable terrestrial versus mari ne influence evident in both the stacking patterns and patterns in the rock properties. These trends can be used as a basis to improve seismic interpretation in tidal environments. All seven of the resulting facies association columns have a slightly different stacking pattern. These variations in stacking patterns signify small but important differences in the depositional processes. The most likely stacking pattern in the fluvial succession is a basal mud rip-up sandstone, trough cross bedded sandstone, planar bedded sandstone, and capped by bioturbated mudstone. This is interpreted as an erosional fluvial channel transitioning to overbank fines in a coastal plain depositional setting with no marine influence. However, the second facies association located within what has been interpreted as a fluvial package (fluvial facies association 2) includes flaser/wavy/lenticular bedded sandstone. This section is likely located within a coastal plain because of the lack of a carbonaceous shale and coal cap. However, this package sits above a tidal package and contains bedforms consistent with tidal influence (carbonaceous shale and coal). This section could be exhibiting some transitional properties associated with a conformable change from tidally influenced to coastal plain, more seaward than fluvial 1, but still more inward than the tidal facies associations. The five different tidal facies associations show even more variation. The most likely stacking pattern for a tidal facies association is a basal trough cross-bedded sandstone, planar bedded sandstone, flaser/wavy/lenticular bedded sandstone, bioturbated mudstone, carbonaceous shale, capped by a coal. However, there is only one facies association column that represents these statistics and is interpreted as an end member 25 26 representation of the furthest seaward tidal package (tidal 5). This full stacking pattern is characterized by a high flow regime tidal channel transitioning to low energy tidal mudstone and capped by coal. Tidal associations 1 and 2 do not contain flaser/wavy/lenticular bedded sandstone which points to a lack of direct tidal influence. However, the inclusion of coal indicates these two facies associations as being deposited in a more paludal setting transitioning from coal mires to channels more inland compared to the other tidal facies associations. Because tidal 2 contains bioturbated mudstone rather than carbonaceous shale, it is likely that this package was deposited the furthest from the coast line, possibly in the transition between paludal and coastal plain depositional environments. Tidal 1 and 2 may also correspond to a more lateral transition or deposition off axis from the tidal and fluvial channels. In tidal associations 3 through 5, flaser/wavy/lenticular bedded sandstone is present, thus placing them closer to the shoreline. The variations in AVO in these three associations are linked to the inclusion and exclusion of bioturbated mudstone and planar bedded sandstone. Those facies association that include bioturbated mudstone correlate to a longer period of quiescence with low energy and tied more closely with a paludal depositional environment. The facies associations with planar bedded sandstone are related to a high energy flow regime. This can be interpreted as tidal channels that have an increased flow compared to those tidal facies associations that lack the planar bedded sandstone lithofacies which represent a lower flow regime. While the stacking patterns reveal the type of processes and location within the depositional environments, the rock property characteristics indicate how the lithofacies properties are expressed collectively as a depositional package and revealed in AVO modeling. 27 Interpreting Degree of Tidal Influence in Rock Properties The relationship between Vp/Vs and Ip show the anticipated trend of an i ncrease in Ip with increasing mudstone and decreasing porosity (Fig. 1.8 C, D, E, G). Although the v-shape appears to correlate to clay content, when separating the trends by architectural elements, several additional patterns begin to emerge. For example, fluvial and tidal mudstone can be distinguished in the cross plot clusters; however, the acoustic responses within the tidal mudstone exhibit a much larger spread in properties (Fig. 1.8). This spread is possible due to a larger range of bioturbation, and a wider range of organic material. Fluvial overbank mudstone is often homogeneous and does not exhibit bioturbation compared to its fine-grained tidally influenced counterpart (Fig. 1.8C). The second major trend that appears is the distinction between tidal and fluvial channel elements (Fig. 1.8 B). There is a significant amount of overlap in rock properties across these two channel sandstone types; however, the averages illustrate there is a small amount of offset between the two architectural elements (Fig. 1.9). When examining the relationship between Ip and porosity for the sandstone beds within these channelized architectural elements (Fig. 1.10). The fluvial end member channel sandstone exhibits the lowest Ip of all the channel sandstone. The tidal facies association end member is comprised of channel sandstone with the most marine influence which exhibits the highest Ip values. The low fluvial channel Ip juxtaposed against the mid-range overbank fines Ip should produce a high amplitude reflection. Although the tidal mudstone can be variable in Ip, it is generally high; the high Ip channel sandstone juxtaposed against high Ip tidal mudstone are expected to produce lower amplitude reflection. 28 AVO Analysis The AVO analysis clearly shows the impact of lithofacies associated with tidal influence, the consequence of averaging rock properties rather than using exact property results, and a discrete difference between the fluvial and tidal facies association. The individual facies associations reveal a distinction between the tidal and fluvial associations in the AVO plot. There is not a clear trend within the tidal associations but the implication of the offset between the fluvial and tidal in the amplitude versus offset plot is a useful predictor to distinguish between the better reservoir channel sandstone in the fluvial environment and tidal channel sandstone which exhibit lower quality reservoir sands. The full EP-25 core AVO illustrates, on a much larger scale, the increase in marine influence with increasing depth of the core and the ability to discern the tidal fluvial packages from tidal in the AVO response (Fig. 1.12). There are two trends when examining the full EP-25 core AVO modeling results. The first trend is based on the rock properties within the JHM, and the link to overall increase in rock properties values of both channels and mudstone with increasing marine influence (Fig. 1.10). Similar to the individual facies association models, the amplitude of reflectivity in the near and far angles decreases with increasing marine influence (Fig. 1.12). This decrease is caused by the increase in Ip with increasing marine influence in the channel sandstone, but also the increase in coal with increasing marine influence. The AVO modeling suggests that the relatively high Ip channel sandstone contrasting with the frequent alternation between coal and carbonaceous shales produces lower amplitude than the facies associations with more fluvial input. The fluvial and tidal channel sandstone with little marine influence 29 has much lower relative Is in contrast to the overbank fines and the tidal mudstone. The decrease in the amplitude reflection coefficient steadily downward in the core is inextricably linked to increase in marine influence. In addition, it is more difficult to distinguish between the tidal and shoreface reflectors in the near angle as they are clustered around the 0.2 amplitude reflection coefficient; however, with increasing marine influence moving downward in the core the reflectors appear to separate in the far angles, with the far offset amplitude of the shoreface approaching zero. Although there is a more noteworthy offset between the fluvial and tidal amplitude reflections in the near angle, the offset decreases in the far angles, making it more difficult to distinguish between the fluvial and tidal amplitude reflections. This indicates that the AVO modeling of the Ip and Is does not significantly improve interpretation between fluvial and tidal, and in order to distinguish the tidal and fluvial reflectors, only Ip is necessary; however, it can be used to potentially differentiate tidal and shoreface. One caveat is the first fluvial package shows a dramatically different AVO character than the other packages, with a much higher amplitude reflection coefficient in the near offset angle and lower in the far offset angle. This variation is tied to the difference between using average rock property values and true rock property values for the lithofacies. Those lithofacies that exhibit a large standard deviation in the rock properties have a much more significant impact on the full model. The gradient difference between the fluvial 1 facies association and the fluvial 2 facies association in the AVO plot may be tied to the difference between the high Ip Drip Tank Member juxtaposed against the JHM (Fig. 1.12). This gradient difference between fluvial 1 and fluvial 2 illustrates the difference between the sheet gravel Drip Tank Member and the 30 paralic JHM. Therefore, the AVO modeling appears to be useful not only when distingui shing between tidal and shoreface environments and the association with increasing marine influence, it may also be useful when detecting sharp changes due to unconformities. Similar to the AVO modeling of the individual facies associations, the full EP-25 core forward model highlights in a more comprehensive way the importance of location in the depositional system by the variations in the near- to far- offsets in the AVO plot by the distinction in the fluvial and tidal amplitude reflections. Interpreting the Proximity to Shoreline from Seismic Reflectivity Based on the MCA and rock properties results, a landward shift upward in the core is corroborated. Marine influence in the lower portion of the core is shown in the lithofacies stacking patterns by the inclusion of different lithofacies that represent tidal influence (flaser/wavy/lenticular bedded sandstone, carbonaceous shale, and coal) (Fig. 1.9). A decreasing marine influence upward in the core is shown by higher Ip and lower porosity in more marine influenced channels and lower Ip and higher porosity with less marine influence (Fig. 1.10). The overlap in the high Ip as a function of net-to-gross supports the interpretation of high impedance tidally influenced channel sandstone (Fig.1.8E). Stacking patterns predicted from a MCA help better constrain interpretations and illustrate the impact of variable terrestrial and marine influence. The implication of variable marine influence and proximity to shoreline is inherently linked to reservoir quality. Based on the results from the rock properties, the high porosity channel sandstone is located in the environments with the least amount of marine influence, while the channel sandstone that are located in the tidal depositional environment and m ore seaward have much lower porosity. Consequently, the channel sandstone that has the least amount of tidal influence correlates to the highest reservoir quality. The impedance contrasts between the different channels and the overbank fines and tidal mudstone facilitate in distinguishing the high quality reservoir sandstones from the poor quality reservoir sandstone. The facies associations that contain little to no tidal influence exhibit a stronger Ip contrast and therefore likely a higher amplitude reflection. The effect of reservoir quality due to the variability of marine influence is best understood by modeling how these variations in seismic reflection. Not only do the rock properties show a wide range of values as a direct result of the highly heterolithic nature of the deposits, the overlap and the fluctuating marine influence demonstrates the heterogeneity of tidal zones. Although AVO modeling did not show a strong impact on differentiating the tidal influence, it did show a strong difference between tidal and shoreface deposits. The forward seismic modeling did, however, show a strong difference in the zero-incidence amplitudes that would help to differentiate coastal plain from tidal, and a clear gradational change between the two end members. Conclusion Tidally influenced reservoirs are significant global sources of petroleum reserves. The complex nature of coastal depositional processes results in highly heterolithic deposits which make reservoir prediction difficult. A Markov Chain Analysis performed 31 32 on lithofacies observations, based on bedforms generated by different depositional processes, from core revealed statistically significant stacking patterns which were interpreted as either tidally influenced or fluvial packages. These stacking patterns corroborate previous depositional environment interpretations and help to clarify where there was ambiguity by simple core observation alone. In particular, the MCA analysis assisted in interpreting facies associations located in the system based on the depositional processes that the stacking patterns signified. Stacking patterns and the inclusion of tidally influenced lithofacies such as flaser/wavy/lenticular bedded sandstone, carbonaceous shale, and coal enables an interpretation of where the package was deposited relative to the shoreline. Analysis of rock properties measured from core further aided in distinguishing tidal and coastal plain depositional environments as well as their related architectural elements. By examining rock properties cross plots, channel sandstone lithofacies show an increase in Ip for both tidal channel sandstone and mudstone which made these deposits distinguishable from coastal plain overbank fines and fluvial channels. In particular, the two resulting patterns emerged from the MCA and rock property analyses. These were clear indications of proximity to shoreline from the MCA stacking pattern results including flaser/wavy/lenticular bedded sandstone, carbonaceous shale, and coal which point to a more seaward depositional environment. The impact that proximity to the shoreline has on the resulting tidal channel sandstone is a decrease in porosity with increasing tidal influence. Using the MCA facies association results and the rock properties, the tidal and fluvial facies associations were differentiated in the AVO plot. The AVO, however, did not add any additional information. The zero-incidence 33 reflectivities are enough to differentiate fluvial and tidal end member. The implication of this discovery is the ability to denote high porosity fluvial packages from low poro sity tidal packages in seismic reflection profiles for improved understanding of reservoirs. The full EP-25 core AVO and rock property results also indicate that the increase in marine influence of being closer to shoreline results in a higher the Ip in the tidally influenced channel sandstone lithofacies. The direct impact on the seismic response is a lower amplitude reflection than the low Ip fluvial channels and overbank fines. Higher fluvial input suggests high porosity and low Ip in both the mudstone and the channels compared to the tidal environments. The information from the rock properties and AVO modeling demonstrates the extent that the depositional environment geology impacts the seismic expression. Small seaward or landward shifts of the tidal environment have significant effects of the rock properties including porosity of the channel sandstone porosity. 34 Figure 1.1. Map of Kaiparowits Plateau and previous studies modified from Chentnik (2014). The black dots indicate the location of previous studies throughout the plateau. The main focus of this study is on core EP-25 located in the north central portion of the plateau. The western edge o f the plateau is dominated by fluvial deposits, the arrows are indicative of fluvial paleocurrent flow direction. The center of the plateau is tidally influenced deposits, and the western edge along Fifty Mile Mountain is the paleoshoreline extent o f the western Cretaceous seaway. 35 A. Regional Stratigraphy _________ B. Straight Cliffs Formation, Kaiparowits Plateau Lithostratigraphy Coastal Sequence Age Kaiparowits Age Onlap Curve Stratigraphy Plateau e .2 '£njO. Kelly Core F.P-25 Rogers Canyon Kaiparowits Fm Grade (This Study) & Left Hand Collet Allen (2010) Shanley & McCabe (1991) u Drip Tank Men ber Land Basin Drip Tank Seq. Wahweap Fm / / 1 1 Fluvial- I dominated deposition/ Fluvial-dominated deposition G shoreface Drip Tank S.B. C3 / CQc3 . 1 t deposition F ) fCcQ (j 1 / / Fluvial-dominated c / / N.Vv \ R e e s * ^ ^ E V J Drip Tank Mbr < A c A Sequence \ Tide- J y C £ .C/] is Jo h n Henry M b r .3 A dominated deposition Christensen \ coal zone co cca U -.ii oc a C/5 / Tide­/ dominated n. B shoreface C/5 \deposition Lower B shoreface (' Ml a in coal zone \ c03 Smoky Hollow Tibbet Canyon V\ / Tide-dominated deposition A shoreface ( A S B. Calico Sequence C E3H Tropic Shale V V \ Gallin (2010) \ Allen (2010) \Dooling (2013) c S 6 .3 - Calico Bed c C3 £ o Dakota Fm Coniacian •8 9 .8 - J ____ 1 Smoky Hollow Member Cahco S.B. Tibbet Seq. <u U Turonian i Tibbet Canyon Member __________________1__________________ ^ f i b B e t ^ B ^ Figure 1.2. Regional stratigraphy and stratigraphic columns of study area. (A) Stratigraphic column of the Coniacian to Santorian deposits throughout Kaiparowits Plateau, and (B) the study area. Kelly Grade to Left Hand Collet to EP-25 represents the south to north deposits of the John Henry Member. Seven marine sandstone packages were named "A-G" by Peterson (1969a). The packages pinch out landward into coal zones and coastal plain facies. Core EP-25 intersects the three coal zones through the B sandstone. The coastal onlap curve is derived from the Rogers canyon study on the southwestern part of the plateau (Allen and Johnson, 2011). 36 Core EP-25 Depth Sample (m) locations Lithofacies Description Number of samples Interpretation Core plugs A Carbonaceous shale Dark gray to black colored shale, often laminated but sometimes structureless; abundant organic material; often grades in and out of coal facies 10 Low energy paludal and lagoonal deposits ■ B Bioturbated mudstone Horizontally laminated mud or mud with no apparent laminations, heavy bioturbation 12 Low energy lagoonal and fluvial overbank muds ■ C Flaser/wavy/lenticular bedded sandstone Alternating laminations of sandstone and mudstone; sand grain size ranges from very fine to medium sandstone 12 Bi-directional flow associated with tidal deposits ■ D Planar bedded sandstone Laminations dip at < 5*; grain size ranges from fine to medium sandstone 4 High flow regime, proximal shoreface, tidal and fluvial channels ■ E Trough cross-bedded sandstone Trough and planar foresets dip between 5-40'; grain size ranges from fine to coarse sandstone 11 Unidirectional flow associated with fluvial and tidal channels F Structureless sandstone Sedimentary structures are apparent; grain size ranges from very fine to coarse sandstone 4 Rapid deposition from high energy suspension in proximal shoreface deposits ■ G Mud rip-up clast sandstone Angular to subangular mud rip-up clasts in a sandstone matrix; matrix grain size ranges from very fine to coarse sandstone 4 High energy fluvial channel lag deposits ■ H Coal Black with dull vitreous reflectance and lacking distinct laminations; coal beds are < 1 m thick 0 Low energy paludal and lagoonal deposits Figure 1.3. Lithofacies and core plugs. Lithofacies observed in core EP-25 with a representative core plug photograph. The number of core plug samples from each lithofacies is a statistical representation of the proportion of that lithofacies present in the core. 37 Figure 1.4. Core EP-25 logs including depositional environments, architectural elements, and lithofacies. Gallin (2010) logged core EP-25 producing a grainsize log, identifying the three different environments of deposition and eight lithofacies. The environment of deposition was further broken down into architectural elements. 38 Figure 1.5. Markov Chain transitional statistics. Transition probability tree showing frequency and probability (%) with which lithofacies are overlain by other lithofacies. Arrows point towards the lithofacies, and boxed numbers indicate the probability of occurrences of the lithofacies that are succeeded directly by another lithofacies. The blue arrows indicate the most likely stacking. 39 1.5 1.2 "g 0.9 _ lI//)l ---- QJ c U - jE 0.6 - 0.3 - 0 Figure 1.6. Lithofacies thickness box and whiskers plot. The graph illustrates the average thickness in meters of each of the lithofacies. The bar shows the standard deviation in thickness. The number on top of each bar represents the number of beds of each of the lithofacies. Coal Carbonaceous Bioturbated Flaser/wavy/ shale mudstone lenticular bedded sandstone Planar Trough Mud rip-up bedded cross-bedded sandstone sandstone sandstone 40 Figure 1.7. Facies associations columns. Facies associations extracted directly from core EP-25. There are two fluvial facies associations and five tidal facies associations. Each of the facies associations represents different lithofacies stacking patterns. The facies associations also exemplify different thickness present in the core. 41 A. I L i t h o f a c i e s B ,7S Architectural Elements 0 Carbonaoeous Siate BoturbateO Mudstone 0 RMsw/Wwy/lflftlicular bedded wodiione % Ranar bedded sandstone % Trough crossbedded sandstone 1.7 IB S % Huvial ctianret sandstone CVerbank mudstone 0 Tidat channel sardslone # lidal mudstone # 9iorefaoe % 9 ructuretess sandstone % Mud rip up sandstone 1j6 • • • A v „ * V V i W • c. Ip (Ns/m3) Boturbation index Tidal mudstone 4000 € 0 X 3000 1X00 12000 Ip (Ns/m3) E. Net-to-gross > 15 • • • v G. oooo eooo Ip (Ns/m3) Sorting Ffcorty sorted Mocwatrfy sorted WbI sorted • * / > V 6000 eooo lp (N s/m3) 5C l > /V v D. eooo aooo lp(Ns/m3) Grainsize 9 <10% . 10% 17 # 1040% # 40€0% 186 % greater than 60% I , • ) Overbank fines 0 Coarse graned sandslooe Medium yarned sandal one 0 Ftnegraned sandstone # Sttitonc # Mudslooc f * V v V 6000 8000 lp(Ns/m$) Porosity 175 166 16 ^ 1 5 6 Q. > 15 • • • • 145 14 *y 136 V 6000 9300 10000 Ip (Ns/m3) Clay content (%) /V V Y ■ 100 1 90 Iso 170 •60 ^50 •40 130 I20 110 r j000 8000 Ip (Ns/m3) Figure 1.8. Rock plots. Vp/Vs versus Ip cross plots showing the relationship between measured rock properties and core plug attributes: A) lithofacies, B ) architectural elements, C) bioturbation index of tidal mudstone and overbank fines, D) grainsize, E) net-to-gross, and F) porosity. The black circle in the bioturbation plot corresponds to the overbank fines. The remaining points in the bioturbation plot correspond to the tidal mudstone. 42 £>Q. 1.75 1.7 - 1.65 - 1.6 - 1.55 - • • 1.5 - • m • • 1.45 - # • • • • 1.4 - • 1.35 - 1.3 Fluvial channel sandstone Overbank mudstone Tidal channel sandstone Tidal mudstone Shoreface • & 2000 4000 6000 8000 Ip (Ns/m3) 10000 12000 Figure 1.9. Architectural element plot with averages. Plots of Vp/Vs versus Ip directly from core plug measurements and colored by architectural elements. The stars represent the averages for each of the architectural elements, highlighting the distinction between the sandier elements and the finer grained elements. Both the tidal elements have higher Vp/Vs and Ip compared to the fluvial elements. 43 A. MD (m) Core EP-25 0 Fluvial channels 0 Tidally influenced channels 0 Shoreface • • B. 10 15 20 25 30 35 40 Porosity V ♦ 10 IS 20 25 30 35 40 Porosity Environments o f d ep o s ition Coastal Plain Tidal Figure 1.10. Channel sandstone porosity versus Ip plot and the relationship to marine influence. Ip versus porosity o f channel sandstone only from the core plugs to illustrate the effect of reservoir quality (e.g., porosity) as a function of either fluvial to tidal processes on impedance. A) Ip versus porosity divides the plugs into tidally influenced channels and fluvial channels. B) Ip versus porosity plots the plugs based on core depth, illustrating the lower channels in the core have higher Ip and lower porosity. C) Core EP- 25 shows the core plug with depth correlating to the colored points in plot B. 44 Fades Depth Association (m) AVO (degrees) Depth (m) Facies Association AVO (degrees) kPas/m 1,500 9,000 K Plana bedded 1 Trough cicus bedded Xwndtfcrx \ Mod np up dnlt Tidal 2 ^ Fluvial 1 ■ Fluvial 2 Depth (m) 44 c ? I Figure 1.11. Facies associations AVO modeling. Modeled AVO responses for tidal and fluvial facies associations using Zoeppritz equation and a 25 Hz rotated + 90° Ricker wavelet from 0 to 30 degrees. Each lithofacies Ip average was used as the input for the AVO modeling. The background lithofacies used is the average Ip of bioturbated mudstone. The AVO plot is analyzed at the maximum amplitude of the full stack. The reflection coefficient along the line at the maximum amplitude in the full stack at zero degrees is plotted at the 0, 5, 10, 15, 20, 25, and 30 degree. 45 O Fluvial 1 ^ F lu v ia l 2 ■ Tidal 1 Tidal 2 A Shoreface 0 10 20 30 Offset angle ^ Figure 1.12. 1-dimensional AVO modeling of core EP-25. The full core EP-25 is presented by the grainsize log which shows the location of the seven facies association and the environment of deposition log. Vp and Vs logs were calculated from a sequential Gaussian simulation based on core plug values and combined with a density log to obtain Ip and Is. The wavelet used for AVO modeling was a 25 Hz +90 rotated Ricker wavelet. 46 Table 1.1 Facies associations, environmental description, architectural elements and lithofacies Environment T T of deposition Description Architectural elements Lithofacies Facies Association 1 - Coastal Plain 1.1 - Laterally extensive channel belts Fine to coarse grained sandstone. Channel sandstone fines upward. Vertical increase of channel belt amalgamation. Low sinuosity fluvial channels. Individual channels 0.5 - 3 m thick. Channel complexes 10 m thick. Overall lack of fossils present. Fluvial channels D, E, F, G 1.2 - Mudstone floodplain and overbank fines Floodplain mudstone are silt and shale rich with lack o f carbonaceous shale and coal, weakly bioturbatcd with some visible laminations. Facies Association 2 - Tidal 2.1 - Highly sinuous tidal channels, isolated distributary channels 2.2 - Lagoonal and estuarine mires with heavy bioturbation Very fine to medium grained sandstone. Grain size coarsens upward in the core along with an increase in marine influence. Tidal rythmites present. Individual channels I - 5 m thick. Channel complexes 15 m thick. Prevalence of brackish water and marine bivalves. Grainsize ranging from silt to mudstone. Bioturbatcd mudstone often grades into carbonaceous shale and terrestrial coals. Heavy bioturbation with some orginal sedimentary structures completely obliterat­ed. Abundance of tidal environment trace fossils. Facies Association 3 - Shoreface 3 .1 - Shoreface sheet sandstone Fluvial overbank fines B, C Tidal channels C, D, E Tidal mudstone A, B.H Proximal shoreface D, E, F Fine- to coarse grained well-sorted sandstone. Hummocky cross-stratified sandstone. Large robust marine bivalves are present. Lithofacies rock properties averages in the three depositional environments. The descriptions are by grainsize, stacking patterns, bedform, geometry, and fossils. The descriptions represent the smallest scale of observation of lithofacies. Each of the architectural elements exhibits multiple different lithofacies. Within each environment of deposition are architectural elements. There are three different environment of deposition interpreted in core EP-25. 47 Table 1.2 Markov chain analysis of lithofacies upper lithofacies 1</D1 "w 0 JZ <D 1 Coal Carbonaceous shale Bioturbated mudstone Flaser/wavy/ lenticular bedded sandstone Planar bedded sandstone Trough cross­bedded sandstone Mud rip-up sandstone Coal X 0.41 0.09 0.04 0.03 0.43 Carbonaceous shale 0.72 X 0.08 0.01 0.04 0.05 Bioturbated mudstone 0.18 0.30 X 0.02 0.06 0.32 0.12 Flaser/wavy/ lenticular bedded sandstone 0.15 0.78 X 0.05 0.02 Planar bedded sandstone 0.12 0.39 0.44 X 0.04 0.01 Trough cross­bedded sandstone 0.35 0.13 0.5 X 0.02 Mud rip-up sandstone 0.45 0.55 X EOD Lithofacies core plugs (# ) Vp (m/s) Vs (m/s) Density (g/cm3) Porosity max min average std. dev. max min average std. dev. max min average std. dev. max min average std. dev. z Bioturbated mudstone 3 3531 2738 3053 296.7 2425 2022 21% 149.6 2.429 2.16 2.308 0.099 11.71 7.003 9.005 1.752 <_ i cl Flaser/wavy/ lenticular ss 2 2926 2917 2921 n/a 2119 2063 2091 n/a 2.314 2.230 2212 n/a 15.06 11.91 13.49 n/a S Planar bedded ss 2 4042 1909 3110 n/a 2874 1318 2223 n/a 2.428 1.840 2.171 n/a 30.33 9.381 18.70 n/a <o Trough cross­bedded ss 5 2521 1533 1963 393.34 1785 1045 1386 297.9 2.184 1.656 1.931 0.201 37.22 16.04 27.03 7.418 Mud rip-up clasts ss 4 1747 1685 1716 31.80 1275 1136 1206 69.47 2.057 1.946 2.002 0.055 25.98 21.13 23.56 2.435 Coal n/a 2400 1800 2000 200 1200 1200 1200 0 1.2 1.2 1.2 0 n/a n/a n/a n/a Carbonaceous shale 9 3985 2667 3117 401.5 2575 2083 2204 183.7 2.925 2.207 2.404 0.217 13.54 3.019 8.061 2.721 _ l <4 Q Bioturbated mudstone 8 3709 1929 3235 455.1 2556 2148 2353 318.1 2.445 2.035 2.325 0.103 20.07 6.16 10.13 3.966 ( - Flaser/wavy/ lenticular ss 9 3147 2320 2645 272.4 1887 1515 1739 243.0 2.318 1.975 2.163 0.126 25.93 10.73 17.79 6.206 Planar bedded ss S 4236 1836 2699 940.2 2861 1250 1859 618.4 2.220 1.848 2.049 0.141 30.04 18.02 23.48 4.529 Trough cross­bedded ss 8 2699 2031 2255 474.4 1929 1178 1546 355.3 2.128 1.765 1.982 0.146 33.17 18.97 26.36 5.347 1 LUo H Trough cross­bedded ss 1 2150 2150 2150 n/a 1519 1519 1519 n/a 1.931 1.931 1.931 n/a 27.04 27.04 27.04 n/a i < {/) LL. Structureless ss 1 2031 2031 2031 n/a 1399 1399 1399 n/a 1.892 1.892 1.892 n/a 28.66 28.66 28.66 n/a -P5> 00 Rock properties table TYING ROCK PROPERTIES FROM CORE TO SEISMIC REFLECTIVITY IN THE KAIPAROWITS PLATEAU, UT, USA Abstract Fluvial-shoreline transitional deposits are becoming increasingly important to understand due to the large number of global hydrocarbon reserves held in such deposits. As the resolution of seismic reflectivity profiles increases, concepts of sequence stratigraphy offer insight into reservoir quality through interpretation of seismic reflection profiles in these transitional successions. The Cretaceous John Henry Member (JHM), located in the Kaiparowits Plateau of southern Utah, reveals numerous exposures of a transition zone which captures fluvial and tidally influenced paralic deposits capped by prograding fluvial deposits, and offers an excellent opportunity to improve our understanding of imaging similar deposits in the subsurface. Previous work along the plateau has been focused on outcrops along the edge, as well as 4 dispersed cores. Fluvial-shoreline transitional deposits are typically associated with coal; however, coal geobody mapping can be particularly problematic in these depositional environments due to its exiguous nature. Four supplementary density logs from cores in the JHM were utilized to map coal and gain a more accurate understanding of the coal thickness, distribution, and frequency, as well as their impact on the seismic reflectivity responses. In this study, we link a newly acquired seismic line to laboratory derived rock 50 properties from a core (~ 500 meters from the line) and outcrop exposures (~ 1500 m from the line). The goal of this study was to use these new data combined with previou s studies to provide insight into reservoir quality interpretation from seismic reflection data in fluvial-shoreline transitional deposits, with particular focus on the interpretation and imaging of coal beds within the tidal sequence. In order to accurately correlate rock properties to seismic, 60 core plugs were extracted representative of the eight lithofacies exhibited in a single core (EP-25). Bench-top measurements were conducted on the core plugs to obtain compressional-wave velocity, shear-wave velocity, density, permeability, and porosity. Three forward reflectivity models were generated and analyzed using these rock properties derived from the core plugs combined with a detailed core description, 1) simple zone-average properties, 2) simple zone-average properties with coal beds, and 3) complex model capturing representative geobody shapes and sizes. These models provide a template for interpretation of a high resolution seismic line (7 km in the north-south direction, 80 Hz frequency source). Interpreting the difference between environments of deposition between the models and seismic profile were clear due to the difference in averaged impedance contrasts. However, individual architectural elements distinguishing reservoir quality when including architectural elements were not clear due to the overlap in rock properties between these zones. Introduction As cheap and easy sources of oil have been exploited, new exploration plays increasingly target more complex reservoirs such as those contained in fluvial-shoreline transitional environments of deposition. Due to the considerable amount of oil reserves held in transitional deposits and the recognition of their inherent complexity, signi ficant effort has been put into the study of analog outcrops for improved subsurface prediction (O'Bryne and Flint, 1993; Weimar and Posamentier, 1993; Ainsworth and Pattinson, 1994; Tetyukhina et al., 2014). The resulting observations have been used to develop predictive models based on lithofacies and architectural elements at a subseismic scale (Martinius et al., 2001; Essam et al., 2013). Advances in seismic acquisition and processing have also yielded higher quality and better resolution seismic data (Thomas and Anderson, 1994; Kendall, 2006; Duarte, 2014). However, even given higher quality data, informed interpretation of reservoir quality from seismic-reflectivity is a function of the seismic rock properties of the reservoir and non-reservoir facies. Insights into the interpretation of these heterolithic depositional environments can be garnered from forward seismic models of 3-dimensional deterministic interpolations derived from core and log observations coupled tightly with nearby outcrop studies. Early rock property measurements in the laboratory focused on effects of clay content and porosity on compressional-wave velocity (Vp) and shear-wave velocity (Vs) as a function of pressure and found a decrease in velocity with an increase in porosity and a decrease in pressure and clay content (Han, 1986; Eberhart-Phillips et al., 1989). These types of laboratory-derived empirical relationships can be used as models to aid in reservoir quality predictions from subsurface seismic reflectivity data. Before acquiring subsurface data, forward model studies of outcrop exposures can lend insight into the anticipated subsurface seismic expression of geobodies. One of the first seismic forward models used large-scale carbonate basin profiles with acoustic 51 measurements from a single core plug in each depositional environment (Biddle et al., 1992). Seismic forward modeling studies of outcrops have been conducted in deep wate r channel systems (Campion et al., 2000; Sullivan et al., 2004; Schwab et al., 2007; Falivene et al., 2010; Stright et al., 2014), and large-scale shallow marine nearshore and shallow marine depositional environments (Hodgetts and Howell, 2000; Tetyukhina et al., 2010; Tetyukhina et al., 2014). Many o f these studies have relied on rock properties from local wireline log data or laboratory measurements of core plugs from areas of similar depositional environment to populate seismic properties in models. However, few studies have focused on direct rock property measurements applied to small scale (m) architectural element geobody distribution in a coal-rich transition zone to generate forward synthetic seismic models (Christensen and Szymanski, 1991; Marion et al., 1992; Vernik and Nur, 1992; Hodgetts and Howell, 2000; Schwab et al., 2007; Falivene et al., 2010; Tetyukhina et al., 2010; Stright et al., 2014). The fluvial-shoreline transition zone is particularly problematic to forward model because the seismic rock properties are not only poorly understood, but they are highly variable. In particular coal deposits, which are prolific in these depositional environments, vary significantly in both rank and distribution. Coal characterization is a key indicator of depositional environment, and therefore, a critical component to predicting and interpreting the seismic expression of nearshore successions (Van Riel, 1965; Ruter and Schepers, 1978; Gochioco, 1992, 2000; Morcote et al., 2010). The focus of this study is in modeling the geobody distribution in a prograding shoreline succession, from shoreface to lagoonal to fluvial depositional environment in the Cretaceous JHM of the Kaiparowits Plateau (Fig. 2.1). Due to the complicated 52 53 depositional setting of coal, their geobody distribution are still not well understood. Because the center of the plateau has the thickest coal deposits, it is an ideal setting to gain a more in-depth understanding of coal distribution in a nearshore fluvial-tidal transition zone. Two cores in the center of the plateau, a data-poor region, are leveraged to document the transition from observed outcrop at Left Hand Collet to the highly heterogeneous, coal-rich transition zone. A high resolution seismic reflection survey (7 km), acquired in order to further fill the data gap in the center o f the plateau and coupled with 3-dimensional models generated from a few core and laboratory-derived rock properties, is interrogated to better understand reservoir prediction from seismic reflectivity in a highly complex coal-rich depositional setting. The two key questions this research aims to answer are 1) are coal deposits in the fluvial-tidal transition zone mapable geobodies in seismic reflection profiles and 2) what insights into subsurface seismic interpretation can be garnered from a 3-dimensional model of a highly complex depositional system. Three 3-dimensional models are presented varying in complexity, populated by laboratory derived rock properties from core plugs, and compare the results to the field acquired seismic reflection profile. Geologic Setting Regional Geology The Kaiparowits Plateau is located in the southwestern part of the Colorado Plateau province in southern Utah (Fig. 2.1). The Kaiparowits Plateau is an uplifted plateau comprised o f upper Cretaceous foreland basin sediments (Peterson, 1969a; Shanley et al., 1992; Hettinger, 1995a). Paleoshorelines were oriented northwest- 54 southeast (Peterson, 1969a; Shanley and McCabe, 1991; Hettinger, 1995a). Recent detrital zircon analysis reveals sediment transport to the shoreline via an axial c hannel system sourced from Mongollan Highlands, Sevier fold and thrust belt (STFB), and Cordilleran volcanic arc, as well as sediment transport via longshore drift from the SFTB (Szwarc et al., 2014). Peterson (1969) divided the Straight Cliffs Formation on the plateau into four members: the lower cliff-forming Tibbet Canyon Member, the slope-and ledge-forming Smoky Hollow and JHM, and the upper cliff-forming Drip Tank Member. This study focuses on the Coniacian to Turonian in age (~88 to 83.5 Ma) JHM in the northeastern central plateau where facies preserved are within a fluvial-tidal transition zone (Gallin, 2010) (Fig. 2.2). Stratigraphy The northeast-central section of the plateau is comprised of fluvial channels and overbank fines, tidal channels, tidal mudstone, interspersed coal, and shoreface deposits (Shanley et al., 1992; Gallin, 2010; Dooling, 2012). In the southwestern part of the plateau near Kelly Grade, the strata consist primarily of multistory-multilateral, meandering channel belts enclosed in thick intervals of rooted mudstone overbank deposits and limited coal that increase with tidal deposition and depth (Shanley et al., 1992; Gooley, 2010; Pettinga, 2012) (Fig. 2.2). Paleocurrent data show that the rivers generally flowed from the southwest to the northeast (Shanley et al., 1992). The southeastern part of the plateau near Rodger's Canyon exhibits seven shoreface packages interspersed with four major coal zones which tend to thin in the southwesterly direction (Vaninetti, 1979; Hettinger, 1995b; Allen and Johnson, 2010a) (Fig. 2.2). The northern 55 part of the plateau reflects a compound incised valley that formed through fluvi al incision and backfilling by estuarine and tidal deposits (Chentnik, in press). The center of the plateau contains the thickest coal deposits in the JHM (Hettinger, 1995b). These thick deposits of coal accumulated on a coastal plain dissected by tidal creeks and estuaries and are in close proximity to the strand plain (Hettinger, 1995a). The clean low-ash nature of the thicker coal deposits indicate that they accumulated in raised mires similar to those described elsewhere in the Straight Cliffs Formation (Shanley et al., 1992). Outcrop Study Outcropping deposits at Left Hand Collet (LHC) provides insight into the geobody architecture distribution, thickness, and vertical stacking patterns (Shanley et al., 1992; Dooling, 2012) (Table 2.1). The JHM at LHC is composed of four main depositional environments: wave-dominated and coastal plain deposits of a regressive shoreface, and tide-dominated and lagoonal deposits in a transgressive barrier island setting (Dooling, 2012) (Fig. 2.2). LHC exposes tidal depositional facies o f a tidal barrier island system with frequent migratory inlets and multiple shoreface deposits with some tidal ravinement (Dooling, 2012). Geobody geometries observed in outcrop at LHC were coupled with core observations to better map 3-dimensional rock body distributions (Table 2.1). Core Study In the 1970s, approximately 23 Utah Power and Light (UP&L) cores were drilled across the northern Kaiparowits Plateau for coal exploration purposes. Six of these cores intersect at least part to all of the JHM. Gallin (2010) logged two of these cores (EP-07 56 and EP-25) (Fig. 2.1). The cores intersect the JHM, and each is greater than 210 m thick. The cores correlate through a unique shared fossil assemblage and are located in the transition zone between marine and fluvial depositional environments. Eight different lithofacies, defined by grainsize, sedimentary structures, bioturbation, and laminations, were classified and described at 0.5 ft. scale in the core (Gallin, 2010) (Fig. 2.3). Three depositional environments comprised of five architectural elements were identified in the JHM and Drip Tank Member: 1) coastal plain consisting of fluvial channels and overbank fine, 2) tidally influenced deposits consisting of tidal channels and tidal muds, and 3) shoreface deposits consists of shoreface architectural element (Gallin, 2010) (Fig. 2.4). Coastal Plain: Fluvial Channels This architectural element consists of lithofacies mud rip-up sandstone, trough cross-bedded sandstone, planar bedded sandstone, and flaser/wavy/lenticular bedded sandstone. The channels are comprised of fine- to coarse-grained sandstone. The channels are represented by fining upward packages. Trace fossils are absent from this element. The channel belts are laterally extensive, and reflect low sinuosity channel belts. The channels vertically amalgamate to form channel belt complexes up to 10 m thick with a vertical increase in channel amalgamation. Coastal Plain: Overbank Fines This architectural element consists of the lithofacies bioturbated mudstone. The floodplain mudstone is mud to silt in grainsize and lacks both carbonaceous shale and 57 coal. They contain some visible laminations and are weakly bioturbated. These beds occur in thin (< 2 m) intervals between fluvial channels. The upper coastal plain of the JHM which includes the fluvial channels and overbank fines constitutes a sheet-like environment between cores EP-25, EP-07 and LHC. Tidally Influenced Deposits: Tidal Channels This architectural element consists of lithofacies trough cross-bedded sandstone, planar bedded sandstone, and flaser/wavy/lenticular bedded sandstone. The channels are comprised of very fine- to medium-grained sandstone that often fine upwards. The channels are comprised o f finer grained sediment at the base of the depositional environment and coarser grained sediment near the top. The occurrence of the channels also tends to increases upward in the core. Brackish water bivalves and trace fossils indicate tidal conditions are prevalent including Teredolites, Ophiomorpha, Thalassinoides, Asterosoma, Taenidium, Planolites, and Teichichnus. The channel belts are laterally restricted and highly sinuous. Channel deposition is represented by relatively thick sandstone beds up to 5 m thick and stack to form sheets up to 20 m thick. Scour surfaces up to 9 m thick separate individual channels. The overall trend in this environment exhibits a transition from marine influence with increasing fluvial influence moving up in the core based on the sedimentation and paleontological observations. Tidally Influenced Deposits: Tidal Mudstone This architectural element consists of lithofacies bioturbated mudstone, carbonaceous shale, and coal. Grainsizes range from silt to mudstone. The bioturbated 58 mudstone lithofacies often grades into carbonaceous shale and terrestrial coal. The base of the tidal environment is characterized by more tidal mudstone, while the top of the environment decreases in mudstone frequency. Heavy bioturbation is present with some original sedimentary structures completely obliterated. There is an abundance of tidal environment trace fossils including Thalassinods and Planolites. These lagoonal and paludal deposits vary in thickness and lateral frequency. The beds range from 5-90 cm and stack to form bedsets between 8-10 m thick. There is a lack of general information about coal body geometries in the Kaiparowits Plateau. This is due to the fact the coal does not exhibit wide sheet-like distribution. The tidal depositional environment, which contains both tidal channels and tidal mudstone in the middle of cores EP-25 and EP-07, pinches out at LHC into multiple shoreface intervals. Shoreface This architectural element consists of lithofacies trough cross-bedded sandstone, planar bedded sandstone, and structureless sandstone. The sandstone in this element is fine to coarse grained, well sorted and clean of mud-draping. The sandstone also tends to coarsen upwards. Marine bivalves and hummocky sedimentary structures indicate shoreface succession. The sandstone forms tabular sheets between 5-25 m thick. The shoreface is a single deposit at the bottom of core EP-25 and EP-07 and correlates to the B sandstone in LHC. 59 Methods Benchtop Core Plug Measurements Sixty core plugs were selected from both the JHM and Drip Tank Member of the EP-25 core to statistically represent rock properties for each lithofacies from the three different depositional environments (Fig. 2.4). Due to the friability and absence due to previous sampling of coal, no core plug samples were extracted from this lithofacies. Core plug measurements were conducted for porosity, permeability, density, (Vp), and (Vs). The core plugs were categorized into a hierarchical framework from lithofacies into architectural elements within the three different depositional environments. 2-Dimensional Correlation Sections Two 2-dimensional cross-sections were created between cores EP-25 to EP-07, which was based roughly on strike with the paleoshoreline, and from core EP-25 to the outcrop study at LHC, in the approximate dip direction of the shoreline (see Fig. 2.1 for locations). The 2-dimensional depositional cross-sections were used as a foundation for the 3-dimensional model by tying together established data points to create accurate stratigraphic intervals. 3-Dimensional Models Three different 3-dimensional models were created ranging in detail and complexity from: 1) average rock properties for each environment of deposition, 2) average rock properties for each environment of deposition with explicit representation of coal, and 3) environment of deposition with mapping of geobodies at the architectural 60 element scale. Correlations from density logs associated with the UP&L boreholes, interpretations from core EP-25, core EP-07, and geometries of geobodies extracted from the LHC outcrop study were used to generate these three interpretive geocellular models. The area of interest (AOI) for the 3-dimensional models is 16 km x 13 km rotated 40.5° NW (Fig. 2.1). There are seven stratigraphic zones in each model for a total thickness of 250 m. The zones correspond to the EODs in core EP-25 (Fig. 2.4). For model 1, each EOD is represented as a single layer with average rock properties. In model 2, a single layer was used for each EOD with the exception of the dominant tidal deposits in zone 5 whose grid size was 2 m thick. This grid size was selected to capture the average coal bed thickness observed from wireline logs in that section. The UP&L cores and core EP-25 and EP-07 were upscaled to the grid in zone 5 and employed to map the coal geobody distribution by hand contouring. In model 3, the same stratigraphic base as model 1 was used; however, the vertical grid cell sizes were defined differently for zone 2 and zone 5 as a function of the average size of the architectural element in each zone. The interpreted architectural elements in cores EP-25 and EP-07 (Fig. 2.4) were upscaled to the grid and used to guide the interpretation of geobodies away from the well locations. On each grid layer, a deterministic and interpretive model of architectural elements (geobodies) was created constrained by the two cores (Fig. 2.4), architectural element geometries from LHC (Table 2.1), and constrained to interpolated coal layers from model 2. The geobody shapes and sizes were loosely based using modern analog from the Ogeechee River, Georgia, USA (Fig. 2.5). 61 Coal Distribution and Rock Properties Coal beds were interpolated using indicator kriging between four UP&L wel l s with density logs (Fig. 2.1 for locations) based on the low density of coal (< 1.2 g/cm ) relative to the surrounding rocks. The velocity value employed for coal was 2000 m/s based on laboratory measurements of coal with similar ranking, age, and location (Morcote et al., 2010). Synthetic Seismic Model Vp and density were assigned to each model at either the architectural element (geobody) scale or as averages for each depositional environment. In model 1, average Vp and density were assigned for coastal plain, tidal, and shoreface depositional environments. Model 2 rock properties are identical to model 1 with the addition of coal Vp and density in zone 5. Model 3 employed the same values as model 2 with the exception of zone 2 and zone 5. In the coastal plain zone, there are two values for fluvial channels and overbank fines, and in the tidal zone, there are two values for tidal mudstone and tidal channels. The resulting 3-dimensional Ip model was convolved with 80 Hz Ricker wavelet using 1-D convolution. This wavelet was selected to not only capture the most realistic amount of heterogeneity in the section, but is a comparable wavelet that was also used for the seismic acquisition profile. Seismic Acquisition, Processing, and Interpretation The seismic profile is approximately 7 km in length, oriented oblique to the shoreline. The profile was shot in between core EP-25 and EP-07. The profile captures up 62 to 500 m depth incorporating the entirety of the JHM. The high resolution reflection seismic data were acquired using dense wide aperture geometry with 5 m spacing (Bruno, 2009; Bruno et al., 2010) (Fig. 2.6). The acquisition was carried out over 2-dimensional profiles using a geophonic array of 12 geodes, 240 vertical 40 Hz geophones, and a high resolution vibrating source (IVI Minivib). The acquisition used three sweeps of 40-250 Hz for 15 seconds. The CDP fold has a maximum of ~200 traces with an average fold of 120-130. The refractor depth model statics were corrected and smoothed to a final datum of 2130 m above sea level. The profile was interpreted by projecting cores EP-25 and EP- 07 and following the reflectors to correlate between the cores at the different environments of depositions. Results Rock Properties Based on the core plug measurements, the resulting range of porosity values from the porosimeter was 3-37%. The ultrasonic acoustic velocimeter measured velocity with resulting Vp ranges from 2500-4500 m/s and Vs from 1200-2800 m/s. The measured densities ranged from 1.8-3.2 g/cm (Fig. 2.3). To visualize the relationship between the rock properties in seismic reflectivity profiles and lithofacies, architectural elements and EODs, values of Vp/Vs are cross plotted against Ip (Fig. 2.7). The lithofacies plot points to a separation between sandier lithofacies and finer grained ones. The lithofacies were divided into architectural elements; the sandstone lithofacies comprise the channels (both fluvial and tidal) and shoreface elements, while the finer grained lithofacies comprise the overbank fines and 63 tidal mudstone. Although there is significant overlap, the architectural elements plot demonstrates a slight offset between low Ip fluvial channels, tidal channels, and shoreface and a slight offset between the high Ip overbank fines and tidal mudstone. However, there is a significant contrast between the sandstone elements and the tidal mudstone and overbank fines. The plots are also differentiated by environment of deposition. The tidal environment tends to have higher Ip overall as well as higher Vp/Vs, whereas shoreface and coastal plain have relatively lower Ip and Vp/Vs. The resulting seismic rock properties were used to populate the three forward models (Table 2.2). Model 1 Ip from the rock properties in all depositional zones were 3 3 averaged; coastal plain Ip was 3880 Ns/m , the tidal 6530 Ns/m , and shoreface 4125 Ns/m . These values are the represent the differences in properties in the environment of deposition rock plot. In model 2, the same rock properties were employed as model 1 with the addition of the coal which was populated using Ip of 2400 Ns/m . Model 3 used the rock properties averages in each individual depositional zone with the exception of zone 2 and zone 5. Zone 2 used impedance from overbank fines: 7475 Ns/m and fluvial channels: 5035 Ns/m . In zone 5, the impedance values employed were from tidal mudstone: 8100 Ns/m3 tidal channels: 5000 Ns/m3and coal: 2400 Ns/m3. 3-Dimensional Model Construction The 1-dimensional cores utilized the lithofacies, architectural elements, and environments of deposition of both cores EP-07 and EP-25 (Fig. 2.4). The cores were used to build the 3-dimensional models through the 2-dimensional cross-section correlations based on the correlated environments of deposition (Fig. 2.8). The cross- 64 section between cores EP-25 to EP-07 consists of an upper coastal plain package, a thick dominant tidal environment with a small coastal plain channel in the middle and shoreface at the base of EP-07. EP-25 is approximately 25 m deeper than EP-07 and exhibits a second small tidal package at the base below the shoreface environment. The cross-section between core EP-25 and outcrop at LHC reveals an upper coastal plain package and the transition between tidal and proximal shoreface in the middle JHM. The B sandstone shoreface package at the bottom of core EP-25 correlates to the B sandstone at LHC. The coal cross-section is in the N-S direction along strike with the shoreline. The density logs are known points of coal locations at specific depths in the JHM (Fig. 2.8). Using the low density assumption, the logs reveal the coal vertical frequency and thickness along strike with the coastline. The architectural elements 3-dimensional model was built using LHC geometries, pertinent analogies, sinuosity, and fluvial paleocurrent direction from southern Kaiparowits Plateau (Gallin, 2010) (Fig. 2.9). 3-Dimensional Synthetic Forward Model The forward models were used to understand and interpret the acquired seismic profile based on information from the rocks property trends, depositional environments, geobody geometries, and fluvial sinuosity based on Fig. 2.4, 2.8, 2.9, and Table 2.1. The 2-dimensional forward models were extracted from the 3-dimensional models at the location in the northeastern section of that plateau between EP-25 and EP-07 oblique to the paleoshoreline at the same location of the acquired seismic profile (Fig. 2.6). The location of the three 3-dimensional models extracted as a 2-dimensional trace is illustrated in Fig. 2.9. The three 2-dimensional models are ordered by increasing detail (Fig. 2.10). Table 2.3 catalogues the impedance contrast and seismic response of the three different models. Model 1, the basic environment of deposition, shows a positive reflector in high Ip coastal plain and shoreface, and a negative reflector in the tidal zone. Model 2, the EOD forward model with coal, has relatively dim reflectors in areas with thin coal beds and more distinct negative reflector in the thicker coal zones. In the depositional environment with coal, the reflectors do not exhibit definitively clear locations and dimensions of the coal bodies. Model 3, the environment of deposition with architectural elements, demonstrates the same reflectors as model 2 with the addition of architectural elements (Fig. 2.11). In zone 2 between the overbank fines and fluvial channels is a large impedance contrast that is characterized as bright flat reflectors. In zone 5, where the channels are thicker and there are fewer tidal mudstone, the reflectors are dimmer and less distinctive. However, at the base of the tidal zone, the channels are thin and there is a high degree of frequency between thick packages of tidal mudstone and coal; the larger impedance contrast is characterized with bright reflectors. The result of the seismic responses listed in Table 2.3 is employed as a template for interpretation of the acquired seismic profile. The rock properties assist in recognizing the Ip contrasts that are expected between the environments of deposition as well as between the architectural elements in the seismic profile. The LHC geometries in Table 2.1 aid in identifying the geobodies that are expected in the profile. The resulting acquired seismic profile depicts several interesting features (Fig. 2.12). The first is that channels can be mapped in the coastal plain and tidal depositional environments. Second is the washed out zone in the middle of the profile. The third feature is the negative reflector on the right-hand side of the profile that trends upward. Lastly, the profile does 65 66 not appear to illustrate any particular area where coal bodies can be easily mapped. Discussion Rock Properties The clustering of the rock properties in the plots helps to visualize the relationship from the lithofacies based core plug measurements to the upscaled environments of deposition and architectural elements. The trends in the plots substantiate the usage of average seismic rock property values to populate the 3-dimensional models (Table 2.2). The trends can be further employed to understand the reflection behavior in both the forward models and the acquired seismic profile (Table 2.3). In the lithofacies plot, the distinction between the sandstone elements or between the tidal mudstone and overbank fines is not clear; however, the distinction between the high Ip fine grained elements and the low Ip sandstone elements is unmistakably discernable and can be linked to the architectural elements plot. The distinction allows for the juxtaposed architectural elements to be distinguished in their respective environments of deposition. In model 1, the rock properties were averaged over each depositional environment. The averages are effectively the average of each environment as seen in the EOD plot. The averages revealed the highest Ip in the tidal environment and the lowest in the coastal plain. This is due to the exceptionally low impedance values exhibited in the coastal plain channels and the relatively low impedance values of the overbank fines (Fig. 2.13). The coastal plain is a strong peak, the tidal a strong trough, and shoreface a weak peak. Model 2 has the same average impedance values in each depositional zone as model 1 with the addition of coals. The coals are relatively weak reflectors in the tidal 67 zone 5. The implication of the forward models is that they illustrate the key reflections that are distinguishable are between the coastal plain environment and the tidal environment, while the coal reflections are relatively dim. Model 3 reveals the closest alignment with the architectural element plot and the most accurate portrayal o f detailed Ip. As depicted in the architectural element plot, the fine grained elements such as the tidal mudstone and overbank fines have significantly higher Ip values than the sandier channel elements. In the tidal zone, the coal impedance is fairly close to the tidal channel impedance and thus is not very distinguished. The tidal mudstone is the main impedance contrast to both the tidal channels and coal bodies, producing strong peaks and troughs where they are juxtaposed. The ability to differentiate architectural elements within the depositional environments offers a notably improved ability to interpret seismic profiles in fluvial-shoreline transitional zones. The role the coal plays in the interpretation is relatively low. The coal is evident when placed in the high Ip tidal environment of model 2; however, when the architectural elements are separated out in model 3, the coal become increasingly difficult to distinguish. Table 2.3 aids in possibly differentiating the elements. The tidal mudstone has the highest impedance and the coal is the lowest impedance with tidal channels exhibiting only slightly higher impedance than coal. Therefore, the brightest reflectors are associated with the juxtaposition of the tidal mudstone and coal, slightly less bright reflectors associated with the juxtaposition of the tidal mudstone and tidal channels, while the juxtaposition of the tidal channels and coal would display a very weak reflector if any. The variability of the impedance value is based on how the core plugs are 68 averaged, either by EOD, EOD by zone, or by architectural element, highlights to difficulty in interpreting the seismic behavior in transitional deposits. Although t h e architectural element scale produces the most accurate representation of the three models, the importance of being able to separate out at least individual depositional environments with the Ip information is essential to being able to correctly characterize the seismic response in transitional depositional environments. Seismic Acquisition Challenges The seismic profile was acquired in a relatively remote location; therefore, some issues occurred that would have been difficult to predict and may have affected the resulting profile. The twelve-day acquisition itself also took place during an unfortunate weather pattern that ranged from hail to constant drizzle for the entirety o f the acquisition. Although the direct link between the noise created on the geophones is not clear, there are likely some repercussions from the noise in the resulting profile. There were also multiple cattle guards and a sharp bend in the road in the middle o f the profile that resulted in the minivib skipping large stretches of road in between sweeps. In terms o f the processing, the water table depth was also unknown and therefore could not be taken into consideration. However, it is likely that this also may have affected the resulting profile. Although problematic, the profile does lend itself to some interpretations using the three forward models. 69 Seismic Profile The acquired seismic profile captured several noteworthy features (Fig. 2 . 12) . Firstly, there are several mapable channels in both the coastal plain and tidal depositional environments. The main geometries observed in outcrop that can be seen in the profile are the channel complexes based on the size of complexes from LHC (Table 2.2). The individual channels and the individual overbank fine beds as seen at LHC are not discernable due to the thickness that the seismic resolution captures. Secondly, the coal geometries are also difficult to distinguish. Possible explanations are because the beds are too thin to be captured by the resolution or the discontinuous nature of the coal geobodies. Although the profile has a washed out zone in the tidal environment that may be associated with coal deposits, it is not obvious if that is the case. A few possible causes of the washed out zone are geophysical illumination from the processing or a geological attribute, such as a thick homogenous package of high Ip tidal mudstone which points to a long period of lagoonal setting or low Ip packages such as tidal channels juxtaposed with coal. Thirdly, there is a distinguishable trough reflector on the right-hand side of the profile that trends at a depth of 300 m and 5200 m upward. On the right-side of this particular reflector, the profile is again washed out. Some possible explanations for this reflector are the transition to the shoreface deposit or it could be related to the incised valleys that are observed in the northern part o f the plateau. Fourthly, the profile highlights some small structural features that are not evident from outcrop. Although the structural features are relatively small, they could have a significant impact on processing and interpretation of the seismic profile in the highly heterogeneous tidal transition zone. Comparison of the Forward Model and the Seismic Acquisition Profile Comparing the forward model results to the seismic acquisition survey i lluminat e s some interesting similarities and highlights the difficulty in creating an accurate model in a highly heterogeneous depositional environment (Fig. 2.13). Model 1 illustrates that the overall Ip contrast between the different depositional environments is the same between the forward model and acquired profile. The impedance difference is largest between the tidal and coastal plain environments. The reflector between these two environments is brighter than between the shoreface and tidal which has less impedance contrast. Similar reflector intensities are also seen in the acquired profile. Model 2 illustrates that the low Ip coal is illuminated in the relatively high Ip tidal zone 5 which is not clearly similar to the acquired profile. However, the weak trough reflector character of the shoreface is similar to the shoreface reflector in the acquired profile. Model 3 with architectural elements is the most realistic comparison to the acquired seismic data. The forward model displays the reflections of channels in both the tidal and fluvial zones due to the low Ip contrast against the high Ip contrast of the tidal mudstone which is also visible in the acquired seismic profile. In the coastal plain depositional zone 2, the acquired profile has flat bright reflectors that likely indicate differentiation between the high Ip overbank fines and the low Ip fluvial channels which are evident in the models and acquired profile. In both the forward models and the acquired profile, the coal geobodies are not very distinguishable. Because the coal and tidal channels have relatively similar low impedance elements, it is the juxtaposition of the tidal mudstone that may predict the coal and/or channels. One striking difference between the models and the acquired profile is the negative reflector on the right-hand 70 side of the acquired profile, and a similar shape peak reflector on the forward model that indicated the Ip contrast between the tidal and the shoreface environments. The most notable difference between the acquired profile and the forward model is that the forward model does not capture the same amount of detail. Although heterogeneity is captured in both the acquired profile and the forward models of the tidal zone, the models do not appear to be straightforward interpretation of the acquired profile. There are two possible causes for the dissimilarities between the models and the acquired seismic profile: model error and data error. The difficulty in tying the seismic to field data in the transitional zone is inextricably linked to the fact that this depositional environment is highly heterogeneous and the field data which are being applied to the interpretation are from data points that are far from the acquired seismic profile. The closest information, both seismic rock properties and geological, is derived from core EP-25 which is 500 m from the profile and the outcrop data being used at LHC is approximately 10 km away. This study highlights the complexity of interpreting fluvial-shoreline transitional deposits. The use of multiple cores and outcrop studies (a robust dataset by industry standards) still does a poor job of capturing the amount of detail that is necessary in order to fully model the complex transition zone. As for the possible errors in data, the culmination of misfortunes including acquisition taking place during poor weather, winding roads with cliffs, lack of information about the water table all could have contributed to a flawed data profile. 71 Implications for Interpretation o f Reservoir in a Tidal Transition Zone Table 2.3 can be used as a guideline on how to interpret environments of deposition and architectural elements in seismic profiles. Between environments o f deposition, the strongest reflection is between the high impedance tidal and low impedance coastal plain environments, while the weakest is between the tidal and shoreface. When the environments of deposition are segmented into individual depositional zones, the brightest reflector is still between the coastal plain environments and the tidal environments. The reflector between the tidal and shoreface remains a dim reflector in comparison. The addition o f ancillary architectural elements to the depositional models illustrates the distinctive reflector characteristics in the tidal and coastal plain environments. In the coastal plain environment, the impedance contrast between the overbank fines and the fluvial channels produces bright flat reflectors. In the tidal environment, the high Ip tidal mudstone juxtaposed against the exceptionally low Ip produces bright discontinuous reflectors, while the impedance contrast between tidal mudstone and tidal channels is slightly lower and thus generates slightly dimmer reflectors. Although it is not captured in the models, the gradational changes and sharp contacts o f the coal o f the facies association in the tidal environment (Chapter 1) are likely distinguishing features that would be present in seismic profiles. The shoreface reflector continues to be a moderately bright peak, however not as strong as the architectural elements in the coastal plain and tidal environments. Although the Ip and the seismic expression derived from the core plugs and models provides a template for future interpretation in transitional depositional environments, there is risk of misinterpretation. The risk of misinterpretation significantly 72 73 depends on the wavelet used; the risk increases with lower frequency. In model 1 and 2, the reflectors are clear-cut between the environments: strong contrasts between coastal plain and tidal, lower contrast between tidal and shoreface. Model 3 illustrates the complexity of differentiating e