ENHANCED RESERVOIR CHARACTERIZATION IN A DEEP WATER TURBIDITE SYSTEM USING BOREHOLE IMAGES AND SPECTROSCOPY LOGS

ENHANCED RESERVOIR CHARACTERIZATION IN A DEEP WATER TURBIDITE SYSTEM USING BOREHOLE IMAGES AND SPECTROSCOPY LOGS Indrajit Basu 1, Nigel Machin 1, Anil Tyagi 2, Kamlesh Saxena 2, Raphael Altman 1, Alex Mathew 1 and Rajesh Kumar 2 1 Schlumberger, 2 Reliance Industries Ltd Copyright 2007, held jointly by the Society of Petrophysicists and Well Log Analysts (SPWLA) and the submitting authors. This paper was prepared for presentation at the SPWLA 48 th Annual Logging Symposium held in Austin, Texas, United States, June 3-6, 2007. petrophysical characteristics of the reservoir units and provide important input for lithological and textural heterogeneity during advanced formation evaluation. ABSTRACT It is commonly accepted that deep marine reservoirs are dominated by submarine channels and channelized lobes. Submarine channels can be broadly classified as gravity-flow channels and leveed channel complexes. While gravity-flow channels are characterized by strong textural heterogeneity, thin-bedded channels and associated levees show rapid lithological changes at cm scale, well beyond the vertical resolution of standard logging techniques. This paper highlights a technique for enhanced lithofacies prediction in a deep-water clastic turbidite reservoir, offshore India, principally using resistivity images and spectroscopy logs. We also show how an integration of the micro-resistivity images together with other well log data enabled a more accurate estimation of gas in place compared with using only conventional logs in this low resistivity pay, thin-bedded reservoir. The micro-resistivity images, together with the spectroscopy log derived dry weight lithologies were used to generate an enhanced lithofacies using a novel lithofacies processing and classification scheme. The system uses the dry weight mineralogical output from the spectroscopy data, and with a corresponding set of rules to create dry-weight mineralogy-based lithofacies. This output is further sub-divided to a finer scale using calibrated micro-resistivity image data. Advanced image processing and interpretation enabled an estimation of reservoir heterogeneity. A heterogeneity index is calculated from the percentile resistivity distribution of the image spectrum. Since the heterogeneity index is related to a distribution of resistivities around the borehole, it is independent of the absolute resistivity values. The integration of these results also reinforce the fact that lithology and texture are equally important as the 1 As continuous coring is not always feasible due to its innumerous cost, borehole image logs and spectrometry data, when calibrated with core, provide sufficient resolution and quantification for a continuous and detailed facies description. INTRODUCTION Deepwater plays increasingly are the focus of hydrocarbon exploration and development in recent years, in India. The principal exploration targets in such a geological setting are submarine channel systems and channelized lobes (Fig.1). The Krishna-Godavari (KG) basin is located on the east coast of India in the State of Andhra Pradesh, and covers an approximate area of 100,000 sq. km. The stretch of sedimentary tract comprises a range of depositional environments comprising coastal plain, delta, shelf-slope apron, deep-sea channel and deep marine channel and fan complex (Bastia et al., 2006). The depositional systems resulting from the interplay of dynamics of Krishna and Godavari delta progradation, gravity- induced processes and relative sea level changes gave rise to mixed deepwater sand mud system. These Tertiary deepwater clastic sediments are characterized by a wide variety of mass flow processes, resulting in complex, irregular and often disorganized lithofacies distribution. While gravity-flow channels in this area are characterized by strong textural heterogeneity, the thin-bedded leveed channels and associated levees show rapid lithological changes at cm scale, well beyond the vertical resolution of standard logging techniques, and reservoirs often remain underestimated. This paper outlines the strength in the integrated analysis of image logs and geochemical logs to account

for the high degree of mineralogical and textural variability, which is imperative of the gravity and mass flow related deposits. Borehole image logs, among other information, provide azimuthal distribution of high-resolution resistivity data in a near-borehole environment to capture the details of internal fabric and texture of different lithofacies. The technique described here illustrates how the textural heterogeneity can be extracted; quantified and used to differentiate sedimentary facies that can then be classified on the basis of their textural variability. Integration of image logs with spectroscopy logs brings out high-resolution mineralogical changes existing within the mixed lithologies. The only ideal alternative of such a reservoir description is extensive coring, which is impractical due to economic constraints. Therefore, integration of borehole images and spectroscopy logs described in this study is an effective approach for the geological interpretation of deep-water sediments and thereby significantly improving petrophysical evaluation and reserve estimation. and contorted thin bed turbidite, slide blocks and chutes infilled with slope related mudstone, thick and thin bedded, discontinuous turbidities and debris flows. CHALLENGES & ADVANCEMENTS Very thinly (less than a few inches) bedded sandstones with little distinguishable textural heterogeneity are commonplace in deepwater environments. The massive (thick) sand bodies, on the other hand are often characterized by large textural variability depending on the nature of mass-flow deposit. The main objective of this study is to develop highresolution facies and geological models that take into account mineralogical and textural heterogeneity, which can also be upscaled for property distribution and simulation purposes during the field development phase. The high-resolution data usually comes from well logs, which controls the optimum level of vertical resolution that can be achieved. GENERAL GEOLOGY The Krishna-Godavari Basin, formed during the breakup of the Gondwana supercontinent, is situated on the eastern passive continental margin of India. The basin trends NE-SW parallel to the Precambrian Eastern Ghat trend. The basin extends from the onshore to the offshore deepwater. The Stratigraphy of the deep-water ranges from Mesozoic to Recent sediments with a thick Tertiary package (5 km). The Three Lithostratigraphic units namely Vadaparu Shale, Ravva Formation and Godavari Clay comprise the Tertiary stratigraphy (Fig.2) and constitute the proven petroleum system in Krishna-Godavari offshore. The Vadaparu Shale primarily represents a shale sequence with minor sand layers and was deposited in an outer shelf regime. The Ravva Formation comprises alternate sand and shale/ clay with minor siltstone and was deposited in an outer shelf environment (Rao, 2001). The subsurface information obtained so far through drilling has indicated that mixed sand mud system may be used to interpret the depositional process in the deepwater Krishna-Godavari offshore basin. The depositional system is dominated largely by slump packages composed of deformed hemipelagic shales 2 However, the evaluation of deepwater deposits utilizing conventional open hole logs may often be difficult as many of these thin-beds are below the resolution of traditional open hole logs. Borehole images have become an established tool in identification and delineation of these thinly laminated sand-shale sequences. However, lithological characteristics of thin beds largely remain unknown as image logs describe only a resistivity anomaly at higher resolution. The facies characterization is therefore primarily dependent only on petrophysical logs and results in petrophysical facies. The limitation of this approach includes: Limited log resolution due to an average sampling rate of 6 inches to 1foot Not all core facies can be recognized due to overlap and inconsistency in log attributes of mass-flow deposits Mismatch between petrophysical correlation and geological correlation due to similar log responses for varied depositional elements and vice-versa This advance methodology involves image logs, which accounts for high resolution geological interpretation and quantification of its texture. The high-resolution resistivity information can also be effectively used to enhance petrophysical evaluation. Elemental

spectroscopy log helps determining the broad elemental composition and therefore quantitative lithology Image log attributes, when integrated with elemental spectroscopy log, can distinguish lithological variation within a heterolithic reservoir. The main advantages of this borehole image - spectroscopy log combination are: Enhanced petrophysical evaluation in thinbedded reservoirs Integration of micro-resistivity, borehole image and spectroscopy data to generate highresolution, mineralogy based lithofacies Textural description and quantification of formation heterogeneity Core-like lithofacies interpretation from logs The ultimate goal is to move one step ahead in the direction of formulating a workflow to generate a highresolution lithofacies scheme, which should be looked upon in conjunction with the petrophysical thin beds. DATA USED The present study is carried out using the following data from one of the deepwater wells that includes Borehole images and high resolution resistivity data Quantitative mineralogy from spectroscopy log Other conventional open hole logs measurements, e.g., gamma ray, bulk density, neutron porosity and resistivity Core photos and core descriptions for correlation and calibration METHODOLOGY Calculation of reserves in thin, laminated beds Thinly laminated sequences in deepwater channel facies and channelized lobe type of deposits can often contain considerable volumes of hydrocarbons. Standard resolution logs are often not appropriate to characterize reserves in these formations because of their inability to resolve fine hydrocarbon sand laminae, and as a result the shale laminae have a disproportionate effect on measurements, suppressing the resistivity and increasing neutron, and ultimately underestimating reserves (Fig. 3a). The image logs shown in this paper, together with conventional logs were used to calculate hydrocarbons in place more accurately. The technique begins with the squaring of the high resolution resistivity image curve to define the discrete boundaries of a high resolution layer-cake model. This model is populated by five distinct lithofacies that are generated from the discrete binning of high resolution and standard resolution logs. For each lithofacies over an entire zone in the high resolution model, average standard resolution log responses are calculated statistically. The various five lithofacies that occur in the high resolution model are therefore populated with these statistically averaged standard resolution log responses. The initial high resolution model is therefore populated with density, neutron, resistivity and gamma ray properties. By applying a corresponding vertical response filter to each of the logs in the high resolution model, a synthetic log is created which can be compared to the original. A minimization technique is used to ensure the synthetic and standard resolution logs agree, by adjusting the property values of the lithofacies in the high resolution model subject to intelligent constraints defined for each lithofacies. The resulting model contains a set of high resolution sharpened logs (gamma, resistivity, density, and neutron) that can be used directly in the calculation of hydrocarbons in place (Fig. 3b). High resolution lithology identification Besides accurate pay determination, borehole microresistivity images are routinely used for geological interpretation. The high resolution of these images is critical for interpretation. However, an aspect lacking in these images is the mineralogical information. A borehole spectroscopy tool generates a continuous output of mineralogical data. This tool measures the elemental yields of the principle elements in rocks (Grau et al., 1989). The principal strength of the borehole spectroscopy data is the determination of total clay (WCLA), total QFM or Quartz-Feldspar-Mica (WSAN), and total carbonate (WCAR). A major strength of the borehole micro-resistivity images is the high vertical resolution. The spectroscopy data, with a vertical resolution of 1.0-1.5 ft (30-46 cm), lacks this level of detail. Combining the strengths of these two types of data creates a very powerful interpretative solution that provides mineralogical data at a high resolution. Classifications based on three components are most effectively done using a ternary diagram (see Hurlbut et al., 1977). Therefore, a ternary diagram was used to create a 100 % inclusive system of classification of 3

* rocks based on the selected spectroscopy outputs (see Fig. 4). Using the three end points of the ternary diagram, twelve lithofacies were defined (Fig. 4) namely: Marl, Claystone, Shale, Sandy Shale, Shaly Sand, Sand, Clean Sand, Calcareous Sand, Calcareous Shale, Carbonate, Sandy Carbonate, and Shaly Carbonate. The mineralogical classification of the spectroscopy data, as mentioned above, lacked vertical resolution, hence a high vertical resolution input, inherent in the micro-resistivity images, was integrated with the outputs of the ternary diagram classification. Using a calibrated micro-resistivity value, two classes of rocks were assigned: shale and sand. Any bed too thin to be detected by the spectroscopy tool (a value of 12 inches (30.5 cm) was designated) would be described as shale or sand, based on selected cut-off values using micro-resistivity data from image logs. In this way, a detailed lithofacies log is rapidly generated for the entire interval where both sets of data are available. These lithofacies are called icore* lithofacies (Kumar et. al., 2003). The numerical output can subsequently be used as input into other interpretation and modeling software tools. Determining textural heterogeneity Borehole Electrical images have been utilized to visually evaluate textural changes for some time. However, visual understanding by an interpreter remains subjective and is difficult to quantify. The technique described addresses this problem. The workflow consists of multiple steps involving image log and spectroscopy log processing and interpretation followed by correlation with core data at different stages of data integration. The measure of rock texture and heterogeneity is carried out using a sandstone textural analysis program. The first step of this sandstone textural analysis involves the calculation of resistivity image spectrum circumferentially around the wellbore over a short interval (few inches to fraction of an inch). Then, an image sorting index or heterogeneity index (Fig.5) is calculated from the percentile distribution of the resistivity image histogram. The variation in image resistivity is similar to that seen in grain size sorting in clastic rocks and the resistivity distribution can vary from well sorted, poorly sorted, bimodal, and skewed either high or low. The heterogeneity index calculation * Mark of Schlumberger 4 used is a relatively simple function of the percentile distribution (Newberry et. al., 2004). In this form, the heterogeneity index is independent of the absolute resistivity values and will have a similar response in low resistivity and high resistivity formations. The heterogeneity or sorting calculation of the resistivity measurements are the same methods used in distinguishing sorting from a core grain size distribution. Finally, the high-resolution image data acquired in the steps above is merged with the available log data to generate a facies description (Fig.6), which captures much of the textural content of the images. The external data is typically a type of shale indictor of the users choice or a combination of indicators. Examples of these indicators are an externally calculated Vshale, Neutron Density curves, or a Gamma ray measurement and utilized to distinguish the up to 4 major facies, i.e., sand (yellow), shaly sands tone (orange), sandy shale (green) and shale (grey). Each of these major facies is then subdivided into as many as six sub-categories defined, computed and displayed based on cut-offs applied to the textural information extracted from the images. These major lithological divisions are then divided into textural subdivisions based on the resistivity variation, e.g., well sorted, moderately sorted, poorly sorted, and chaotic sorted etc. A clear distinction of heterogeneity is displayed between two successive sandstone beds in figure 6. The sandy zone between XX96m-XX97m is more homogeneous (higher matrix fraction) and the lower sandy zone from XX97m-XX98m have suspended mud clasts and shows higher heterogeneity. The outputs from the sandstone textural analysis calculation may be compared with core results to calibrate the image-based textural facies for specific geological environments. ENHANCED RESERVOIR DESCRIPTION - AN EXAMPLE FROM OFFSHORE INDIA In this study, we show an example of how oil-based micro-resistivity images were used to more accurately estimate hydrocarbons in place. Figure 3a shows standard resolution logs over a 7 meter interval. These logs are some of the highest resolution standard logs

that are available; however they are still unable to resolve thin hydrocarbon bearing laminae that occur between XX73m and XX75.5m. Figure 3b shows the output over the same interval from the high resolution methodology described in this paper using the image logs. The thin laminae in the interval XX73m XX75.5m are very well defined. The image in track 3 shows these laminae. The corresponding synthetic resistivity log is shown in track 5 as a red sharpened log and is a more accurate representation of these thin laminae. The resulting interpretation shows an enhancement in the calculated gas in place. For example, track 7 shows that the thin sand laminae (XX73m XX75.5m) have gas saturation of 0.8. This can be compared with a gas saturation of 0.15 calculated using the standard resolution log curves in the same interval (Fig. 3a). clay-rich facies (i.e., sandy shale, claystone and shale in a downward sequence). This description is not an accurate representation of the sedimentary facies. Sandstone textural analysis complements the lithofacies results by adding the critical textural component. In this example, textural attributes from sandstone textural analysis show that this shaly zone is highly heterogeneous and shows chaotic to poor sorting, which can be clearly seen on both the core photograph and image log. When these bulk lithological facies and textural facies are integrated, the integrated facies can be classified as chaotic and poorly sorted sandy shale facies (Fig.9). Textural analysis therefore gives an insight and quantification of the level of heterogeneity existing within a massive sand unit. Such debris-flow channels with specific heterogeneity index, when correlated across a prospect can give a better understanding of the lateral variation of the reservoir. Borehole image-spectroscopy log integrated lithofacies acts as the missing link between image facies (which is primarily a resistivity measurement) and core lithofacies by adding the lithological attributes to the petrophysical facies. It is observed that icore* lithofacies are not only able to distinguish the thinly bedded sand-shale interbeds or heterolithics but also gives a measure whether the heterolithics are sand-rich or shale rich and lithology sub-classes from sand to shale (Fig.7). Typically, the heterolithics are observed in channel-levee complexes, lobe or sheet type of deposit or as reworked sediments. Once the thin-bedded lithofacies are put within the specific architectural elements of a deepwater system, identification of sandshale facies and their relative distribution at cm. scale is critical for accurate calculation of net pay determination,. In the example shown in figure 7, thin beds correspond to channel system. Heterolithics near XX78m is shale-rich, while those near XX79m are sand-rich, which is also evident from core photographs. This example suggests that significant variation can be expected within a specific reservoir. The lithofacies from icore* is found to give excellent results in case of massive and featureless sandstones or planar laminated lithofacies. However, icore* cannot incorporate the textural variation from the image facies due to irregular, discontinuous bedding and presence of clasts and/or nodules. This is complemented by the incorporation of sandstone textural analysis outputs that define the textural variation of the respective lithofacies. Overbank splays of a debris-flow are shown in figure 8. It is apparent that litho classes alone are unable to account for the presence of large clay-rich clasts suspended within a predominantly sandstone matrix and classifies the shale rich zone as continuous 5 The mineralogical and textural facies are combined and calibrated with core data in an attempt to create a facies matrix (Fig.10), which could clearly distinguish 18 texture and mineralogy based lithofacies. These lithofacies can be correlated to specific depositional units, with some overlap in various mudstone lithologies. This integrated facies scheme encapsulates and quantifies some major characters of the lithofacies that can closely correlated to specific depositional units and associated environments as mentioned above. The details of depositional interpretation using this facies scheme are beyond the scope of this paper. STRENGTHS AND LIMITATIONS Image logs have the unique strength of detailed visualization and more accurate reservoir characterization. Petrophysical results and net pay are greatly enhanced by using micro-resistivity as a guiding factor in determining hydrocarbon saturation. However, these thin-bedded reservoirs often lack proper lithological description and thus leaving a gap between log derived reservoir facies and lithofacies described from core. The main application and strength in mineralogytexture combination is a replication of core-like lithofacies using wireline logs derived information. The integrated facies may not replace but definitely minimize the need of core in the non-cored wells and provide similar information. icore* gives highresolution lithology that conventional logs are unable to generate and sandstone textural analysis identifies reservoir and formation heterogeneity. Image logs have sufficient resolution to identify small-scale textural and

structural features and nature of bed contacts. The geological details may prove critical alongside upside pay determination using enhanced petrophysical evaluation for thin beds, especially in determining the extent of such reservoirs. This lithofacies approach may be further enhanced by incorporating log derived porosity, permeability, pore size distribution and interpreted grain-size data based on petrophysical analysis, which is beyond the scope of this paper. However, the resolution of this Mineralogical-Textural facies is a function of tool resolution, which may be as high as fraction of an inch in many cases. This kind of interpretation and facies analysis scheme also requires key wells for core calibration to determine facies relationships. Another underlying assumption in sandstone textural analysis is that the resistivity distribution is considered to be analogous with sorting / heterogeneity and therefore texture. The resultant lithofacies can be used along with other geological elements from different sources for a more complete picture of the subsurface. CONCLUSION The methodology described in this study uses microresistivity based textural interpretation and element based lithological information in conjunction with advanced petrophysical evaluation for thin bedded pays. Borehole image and spectroscopy log data, when combined utilizing the methodology discussed in this paper, allows high-resolution mineralogy and texture based lithofacies scheme to be derived. The results are then calibrated with core results for validation and critical assessment of the resultant facies scheme. The successful execution proves the validity of this workflow, which can be applied elsewhere with special interests in deepwater clastic sedimentation. The outputs of this study can be used as inputs for geological modeling and static reservoir modeling with lesser uncertainty compared to other existing lithofacies classification techniques. The high resolution information derived from micro-resistivity images can also be used to estimate more accurately hydrocarbons in place since many of the fine laminae that contain hydrocarbons are below the resolution of conventional logs. The goal of this integrated study is to reinstate the importance of high-resolution geological outputs to understand the nature of thin bedded pays and asses the variability of massive sandstone reservoirs. 6 ACKNOWLEDGMENTS The authors thank the Operator and Schlumberger Oilfield Services for their support and for granting permission to publish this study. REFERENCES Bastia, et al., 2006, Linking Shelf Delta to Deepwater; Krishna-Godavari Basin, Journal of Geological Society of India, Vol.67, pp. 619-628. Grau et. al., 1989, A Geological Model for Gamma-ray Spectroscopy Logging Measurements, Nuclear Geophysics, Vol.3, No.4, pp. 351-359. Gupta, S. K., 2006; Basin architecture and petroleum system of Krishna Godavari Basin, east coast of India, The Leading Edge, pp. 830-837. Hurlbut, C. S., Jr et al., 1977, Manual of Mineralogy (after James D. Dana), 19th Edition, John Wiley & Sons. Kumar et. al., 2003, Lithofacies classification based on Spectral Yields and Borehole Microresistivity Images, GCAGS/GCSSEPM, Vol 53, 2003, pp. 351-359. Newberry et. al., 2004, A method for analyzing Textural changes within Clastic environments utilizing Electrical Borehole Images, GCAGS Convention, pp.1-10. Rao, G. N., 2001, Sedimentation, Stratigraphy, and Petroleum potential of Krishna-Godavari Basin, East Coast of India, AAPG Bulletin, Vol.85, No.9, pp.1623-1643. Reading at al., 1994, Turbidite systems in deep-water basin margins classified by grain-size & feeder systems, AAPG Bulletin, Vol.78, No.5, pp.792-822. ABOUT THE AUTHORS Indrajit Basu is a reservoir geologist for consulting services with Schlumberger in India. He holds an M.Tech. degree in Geo-Exploration from Indian Institute of Technology, Mumbai, India. He is a member of Geological Society of India and AAPG. Nigel Machin is a senior geologist and Geology Domain Champion with Schlumberger in India. He holds a B.S. degree in Applied Geology, University of Leicester, United Kingdom. He is a member of Geological Society of London, IAS and SEPM.

Anil Tyagi is a General Manager of Exploration (Logging and Petrophysics) for Reliance Industries Limited (Oil and Gas Division). He holds a M. Tech degree in Applied Geophysics from Indian Institute of Technology, Roorkee, India. Kamlesh Saxena is a General Manager of Exploration (Logging and Petrophysics) for Reliance Industries Limited (Oil and Gas Division). He holds an M.Tech. degree in Applied Geology from the University of Saugar, Sagar, India. He is a member of AAPG and SPE. Raphael Altman is a senior reservoir petrophysicist with Schlumberger Brasil. He holds a Masters degree in Petroleum Engineering, Imperial College, University of London, United Kingdom. He is a member of SPE. Alex Mathew is a wireline Account Manager in Schlumberger, India. He holds a B.S. degree in Mechanical Engineering from Illinois Institute of Technology, Chicago, U.S.A. He is a member of American Society of Mechanical Engineers (A.S.M.E.). 7

Valley Canyon Channel-levee system Slump Basin Plain Depositional Lobes Fig.1 Schematic representation of Deepwater depositional model for a mud-rich submarine system (modified after Reading & Richards, 1994). Fig.2 Generalized Stratigraphy of Krishna-Godavari Basin (Gupta, S. K., 2006) 8

XX70 XX75 Fig. 3a: Standard Resolution logs fail to detect thin laminae that occur between XX73m XX75.5m XX69 XX70 XX71 XX72 XX73 XX74 XX75 XX76 Fig. 3b: High resolution Oil Based Mud Images are able to define the thin laminae that occur between XX73m XX75.5m. Calculation of hydrocarbon in place is enhanced in this interval using the high resolution methodology for which the image plays a central role. 9

Fig.4 Ternary diagram with dry weight input from spectroscopy logs used for lithofacies classification in icore*. Frequency Lower Bound Upper Bound Bimodal Poorly sorted Well sorted Grain Size/Resistivity Generalized Textural Facies Sandstone Fig.5 Principle of Sandstone Textural Analysis and textural classes showing user-defined boundaries from poorly sorted to well-sorted distribution. 10

MD Static Image Fraction Heterogeneity Index Textural Facies XX96 XX97 XX98 Fig.6 Heterogeneity Index and Textural facies as calculated based on image resistivity spectrum. Suspended mud clasts between XX97m-XX98m add to the heterogeneity of the sandstone. A decrease in matrix fraction (lighter portion in track 3) is also indicative of heterogeneity. MD Static LQC Dynamic icore* facies Core Slab XX78 Shale-rich Thin beds Sand-rich Thin beds XX79 Scale Bar - 10 cm Fig.7 Lithological characterization of thin beds using icore*. The icore* lithology and core lithology are strongly in agreement with each other. 11

Core Slab Static MD Dynamic icore* Heterogeneity Index Sorting XX69 Chaotic Sorting Scale Bar 10cm Fig.8 Texture classes complementing icore* lithologies in poorly sorted gravity-flow deposits. Suspended clay clasts and irregular bedding surfaces are visible on core and micro-resistivity images above XX69m. icore* interprets the litho-classes of this zone as sandy shale, claystone and shale (in a downward sequence), while textural heterogeneity indicates a chaotic and poorly sorted interval. Core Slab Static MD Dynamic icore* Het. Index Integrated Facies XX69 Scale Bar 10cm Fig.9 Example of integrated facies classes combining icore* facies and sandstone textural analysis results. Poorly sorted shaly sandstones and sandy shale are better defined combining both lithology and textural facies together. 12

icore Mineralogy Heterogeneity Fig.10 Facies Matrix using icore* mineralogy and sandstone heterogeneity. 13