The exploratory projects of hydrocarbons in the Parnaíba Basin have primarily targeted Poti and Cabeças Formations. With the rich geological knowledge obtained from the drilling of wells, the Longá Formation is viewed as a potential new exploratory play. This formation, which some studies reckon that can act as seal or source-rock, is characterized by the intercalation of shales, siltstones, and sandstones. During the drilling of a well, with the subsequent detection of gas, three 18 m long whole-cores were extracted for geological and petrophysical studies. In addition, a complete set of conventional and nuclear magnetic resonance (NMR) logs were obtained along with laboratory analyses of routine core analysis (RCA), capillary pressure, NMR, X-ray diffraction (XRD), and rock mechanics, for a complete petrophysical evaluation. The Longá reservoir is a complex reservoir with millimeter-thick laminations and reservoir layers with conductive minerals that suppress the resistivity curve. As a result, the log data had to be integrated with core data and ultimately a Domain-Transfer analysis model in uncored wells to correctly estimate petrophysical properties and make development decisions. The integration of core-log data made it possible to obtain important information about the depositional environment, lithology, reservoir characterization, calibration of the main petrophysical parameters, and mechanical properties of rocks , which can help realize hydraulic fracturing, thereby contributing to production optimization and risk reduction in exploratory projects. The productivity of the well increased by approximately 500% after stimulation of reservoir. Furthermore, the subsequent drilling of a few more exploratory wells revealed the first commercial field of the Longá Formation in the Parnaíba Basin. INTRODUCTION Petrophysical evaluation of thinly laminated reservoirs presents great complexity, especially regarding the estimation of hydrocarbon volume in place. Conventional well logging tools have a vertical resolution, which is larger than the size of the laminations in thinly laminated reservoirs, and thus fail to solve the petrophysical properties of these small layers. In addition, the presence of clay minerals generates an excess conductivity that affects the resistivity curve. The above- mentioned effects are known well in the petrophysical technical literature as complicating factors for the generation of reliable models. Additionally, this case study presents a greater difficulty due to the presence of complex mineralogy that contains metallic, heavy, and conductive minerals, thereby corroborating the need for complementary studies on core-log integration as a way of calibration of the main petrophysical parameters. Geology is strongly related to the in situ measurements performed by well logs, and thus helps in deeply understanding the spatial distribution of petrophysical properties and geometry of the different lithologies. Given the type of reservoir that this work presents, special emphasis will be given to understanding the relationship between geology and petrophysics.

Borehole images provide many different texture features for facies analysis and natural fracture identification. However, classification of most of these features is achieved manually. The workflow proposed here is to implement geological "facial recognition" from borehole images and other petrophysical measurements. The image segmentation is the first step to split the geological "facies" from continuous borehole images. Then the clustering based on texture similarity and petrophysical measurements is the second step to major facies categories. The major facies categories are labeled manually, and a deep learning model is trained to recognize geological facies on new borehole images in the same reservoir. A borehole image can be visually recognized as a composition of successive zones; different zones have different statistical properties, which can be used to characterize the image and generate the zonation. The continuous histogram and variogram derived from image data are used for image segmentation. From the highresolution borehole images, the segments obtained are numerous enough to perform what is known as unsupervised classification. Among various methods of unsupervised classification, we choose to use the mean shift algorithm for the automatic clustering. It is a deterministic process, which is suitable in determining the number of clusters. The segments are assigned as facies with a local geological setting, then structured and formatted to build a library of multimodal data (image data and petrophysical log data) for a given facies. A deep learning model is trained to associate multimodal data to a given facies. This model is used to identify automatically the image features of another borehole, for continuous facies analysis in similar depositional environments. We demonstrated this workflow in different depositional environments. Twenty-four facies were recognized from a water-based mud image in a braided river environment from China compared with 14 with core description Twelve facies identified from an oil-based mud image in a lacustrine system from the United States were then applied to another water-based mud image in the same reservoir with the deep learning model. The results from this approach were verified after comparing with a manual interpretation from cores. INTRODUCTION Microfacies identification is the fundamental information for sedimentary analysis. The drilling core data is commonly used for microfacies classification. Because of the high cost of coring, the high-resolution borehole image prevails for facies or facies association analysis over that with core calibration. There are lots of successful case studies on facies modeling or deposition environment identification from high-resolution borehole images in combination with other logging technologies or seismic images (Lawrence et al., 2003; Blount, 2017). However, for most of these studies, the sedimentary facies identification was achieved manually and required significant effort.

Marine carbonate reservoir formation evaluation is typically not concerned about the presence of clays, provided that the deposition of good quality carbonate platform facies is normally limited to clear waters. Conversely, the South Atlantic lacustrine Pre-Salt deposits contain abundant magnesian clays (e.g., kerolite, stevensite, mixed-layer kerolite-stevensite, sepiolite, and saponite), associated to the bioclastic and chemical carbonate reservoirs. These clay minerals precipitated under extreme alkaline environmental conditions, and are peculiar in terms of composition (e.g. kerolite - (Ca 0,03 Sr 0,02 Na 0,01 ) (Mg 2,88 Al 0,01 ) Si 4.02 O 10 (OH) 2 .nH 2 O) and occurrence, exhibiting laminated, massive, ooidal, peloidal, and coating habits. They are very distinct from the conventional detrital or common diagenetic clays that occur in siliciclastic reservoirs, requiring different petrophysical interpretation models. For instance, the conventional approach for clay content evaluation using gamma rays is not adequate, considering that Mg-clays are poor in radioactive elements such as potassium. Moreover, the density versus neutron cross-plot does not display a clear contrast pattern between clean reservoirs and clay-rich rocks. On the other hand, NMR logs exhibit a highly distinctive clay bound water relaxation time (<3ms) in Mg-clays-rich intervals, similar to a shale pattern, even though the proportion of such clays in relation to carbonates (calcite and dolomite) is rarely higher than 30%. In addition, Mg-clays strongly affect sonic logs, decreasing both shear and compressional velocities, which can be useful to identify them in cross-plots of density versus interval transit time. Mg-clays are quite unstable minerals, what resulted in their dissolution and/or replacement by other minerals. A reasonable amount of porosity within Pre-Salt reservoirs has been interpreted as secondary, formed by dissolution of these clays. They are absent in many wells. However, in some areas, Mg-clays-rich carbonates may be thicker than 200 meters. The intervals with abundant preserved Mg-clays are not considered reservoirs, as, despite their fair porosity, they have very low permeability (<.1mD). In contrast, preserved Mg-clays are scarce in the reservoir facies, indicating that either Mg-clays were not deposited in those areas and/or periods, or that they were dissolved soon after deposition, due to changes in the chemistry of the lacustrine fluids, or later during burial diagenesis. Therefore, the understanding, evaluation and prediction of Mg-clays occurrence are of paramount importance for the petrophysical interpretation in the exploration and development of Pre-Salt reservoirs. INTRODUCTION The presence of clays in marine carbonate reservoirs typically is not a problem, considering that environments favorable for the generation and accumulation of carbonate sediments are usually clear waters with low solid suspension content (James and Jones, 2016). Additionally, the authigenesis of clay minerals on those rocks is limited, since the interaction of interstitial waters with carbonates normally does not provide ions ( e.g. Al and SiO 2 ) for their formation. However, continental carbonates may be composed of complex mineral associations, which precipitation is controlled by local environmental conditions (Renaut et al., 1986; Buchheim and Awramik, 2013). This is the case of the extensive South Atlantic Pre-Salt sag deposits, which precipitated directly from lacustrine waters in association with Mg-clay minerals and silica under extreme alkaline environmental conditions (Wright and Barnett, 2015). The underlying rift reservoirs are composed of bioclastic deposits (Carvalho et al., 2000), which are also associated to Mg-clays in complex patterns (Armelenti et al., 2016; Goldberg et al., 2017).

The quantitative interpretation of Sand/Shale thin beds is a well-known problem with a multitude of solutions. However, the Thin Bed Analysis (TBA) is still perceived as overly complicated. The TBA goal is often restricted to the identification of an additional pay in shaly intervals with low resistivity, but an equally important objective is to provide quantitative consistency between the results of the log analysis (shaliness, porosity) and the core derived relationship (Permeability and Saturation functions). This paper discusses the lessons learned in several fields with significant stakes in Sand/Shale thin beds and presents a pragmatic workflow. The first step is to identify the intervals where the TBA is both applicable and useful. The methodology effectively applies to thin beds of Sandstone and Shale with sharp transition and very distinct petrophysical properties. Permeable Sandstone Beds are prone to contain hydrocarbons while Shale Beds contain nonmovable water. The identification of Thin Beds Facies must be performed using all available data including, but not limited to, mud gas, borehole images, NMR, electrical and acoustic anisotropy. A preliminary simplified analysis must outline the potential stakes associated to sandstone thin beds and evaluate the opportunity to progress with a full study. The second step is the computation of the Sandstone Fraction and Sandstone Porosity that constitute the key deliverables of the analysis. These products can be obtained with different methodologies. Most of the time, Sandstone Fraction and Sandstone Porosity are inferred from Porosity and Shaliness using the "Thomas and Stieber" Xplot that provides the Laminated Sandstone Fraction along with Dispersed or Structural Shale. A simplified in-house methodology was introduced to deal with the common turbiditic environments characterized by a low amount of clay in Sandstone beds but variable properties of both Sandstone and Shale. The third step is the computation of the Sandstone Water Saturation. Although this is considered the main goal of the TBA it cannot be always performed. In fact, a robust computation requires the acquisition of the vertical resistivity. In absence of Triaxial resistivity, the use of NMR proved crucial in most of the studies and a methodology to use the NMR data as a standalone tool for providing the three TBA deliverables was introduced. The fourth and fifth step are the validation of the results using core data and the definition of cutoffs to be used for summation.

Performing high-resolution electrofacies analysis greatly contributed to the understanding of the connectivity and to the volume estimation of the recently discovered deep-water Calumbi Formation play of Sergipe Basin, Northeast of Brazil. It is considered the most important discovery in the country after the Pre-salt carbonate reservoirs and consists of turbiditic channel and lobe complexes of Campanian-Maastrichtian age. In the present study, unsupervised classification based on the Multi-Resolution Graph-Based Clustering (MRGC) method was used to the electrofacies analysis. Available rock data were initially classified into lithofacies according to Mutti facies. For the interpretation of the electrofacies, a reclassification was made based on the lithofacies described in the Bouma sequence, resulting in 5 clusters correlated to conglomerate and sandstone to interlaminated and shale slump facies. The quality control was performed for all available logs and the petrophysical core data were incorporated into the analysis for rock data depth calibration. After log processing and validation, the correlation between core and well-log data was performed. Textural facies were extracted from resistivity image logs through the generation of self-organizing maps (SOM), which worked as the input that would guarantee the high resolution of the electrofacies analysis. The use of SOM with MRGC method presented excellent results, with a good correlation between the electrofacies and the rock data and showing a good resolution as required to better understand the reservoirs of the Sergipe Basin. INTRODUCTION Northeast of Brazil has the most important discovery of the country after the Pre-salt carbonate reservoirs. It is a play of Calumbi Formation in the deep waters of the Sergipe Basin. The first well of this area was drilled in 2010. After six years of exploratory campaign the new oil province had more than 20 wells drilled, in water depths of almost 3,000 m and well final depths sometimes greater than 6,000 m. The discoveries constitute accumulations with significant heights of hydrocarbons, and the drilling campaign represents an exploratory success rate of over 70%.

Sand injectites on the Norwegian Continental Shelf have proven their commercial significance. Some are already producing e.g. Volund, Viper, Balder, Ringhorne and Kobra fields, while others like in production licenses (PL) 340 and 869 have recently been discovered and appraised. Extensive literature on the geology of sand injectites have been published (e.g. Jenssen et al. 1993; Jolly et al. 2002; Huuse et al. 2003; Hurst et al. 2005). However, few references are available on the petrophysical and geophysical aspects of sand injectite reservoirs. In this paper, the petrophysical properties of sand injectite facies; dykes, sills and brecciated sands are discussed, along-with their identification from seismic data. A perception that volumetrics of sand injectite reservoirs cannot be reliably evaluated is assessed. Sand injectites in PL340 & 869 were interpreted as remobilized sands from the Hermod and Heimdal formations of Paleocene age injected into the overlying Balder formation and Hordaland group mudstones of Eocene age. The mudstones acted as a seal, forming an intrusive stratigraphic trap. The trap geometry varied locally depending upon the dyke and sill geometries of the sandstone. Dykes had large vertical reach with corresponding high hydrocarbon column, while sills had low vertical relief with large lateral extent. Intervals of brecciated sands were also observed within the injectite complex, especially where sands were thin. These brecciated sands contained large amounts of angular mudstone clasts of different dimensions suspended in an overall sandy matrix. Close examination of cored dykes made it possible to observe this, while it might not be as obvious when looking at bulk well logs. Petrophysical log responses for clean sills and dykes behaved the same way as they would in a clean sandstone reservoir. If sills and dykes were very thin, they would also risk not being counted as net or pay (Dromgoole et al. 2000; Flølo et al. 2000; Suau et al. 1984). Such errors can impact in-place volumes in a significant way. Sills appeared as blocky clean sand on logs, but it was difficult to differentiate a dyke from a sill or thin sands using logs. Dykes are high angle features and are identified either by core studies or borehole images when intersected by a well, or if large enough observable on seismic. Brecciated sand intervals appeared with cm-to-dm scale mudstone clasts suspended in sand with approximately 40 to 60% net-to-gross. Log responses over these intervals indicated shaly-sand or thin-sands. Resistivity and thermal neutron porosity logs were highly affected by the shale clasts. For this reason, a fractional Net/Gross interpretation technique was used to evaluate the sand-content and hydrocarbon-pore-volume. To further verify these results, they were compared to observations directly on core. To qualify to what extent petrophysical logs and interpreted products thereof can be relied on to evaluate hydrocarbon volumes of sand injectite reservoirs, a high resolution petrophysical interpretation was generated using a computerized tomography (CT) scanned core image. Core image sand counting, and image derived high-resolution bulk density logs with shale corrected resistivity were used. Results of this high-resolution interpretation featured an excellent match with routine core analysis data and manual core observations in core laboratory. The fractional Net/Gross method used is the modified Thomas-Stieber method (Johansen et al. 2018). Its results compared well to the high-resolution CT-Scan image results and better evaluated hydrocarbon-pore-volume of sand facie as compared to conventional bulk formation evaluation approach. This result confirms that Thomas-Stieber method can be used for brecciated rocks, which leads to some useful recommendations on how to best log and perform petrophysical evaluation in such reservoirs.

ABSTRACT Measurements of rock dielectric properties can help to quantify porosity, mineralogy, clay content, texture and especially water saturation of reservoir rocks. Permittivity in the range of MHz to GHz, can be measured rapidly using downhole tools and validated in the laboratory, offers insights into rock flow and mechanical properties not only for conventional reservoirs, but also for tight reservoirs and overburden shales. One barrier to the wider acceptance and adoption of dielectric logging has been the scarcity of laboratory measurements on a range of standard rock types at controlled saturation levels that are verified with independent methods. A second limitation is the lack of well understood and calibrated models for interpreting and inverting advanced, multi-frequency dielectric logs. CSIRO is embarking on a program to collect standardized dielectric spectroscopy data on a library of different rock types obtained from sources worldwide. We present examples of the data collected so far on typical, (mainly clean, i.e., low clay content) sandstones and carbonates as a function of water saturation, for air-brine and oil-brine systems. We also show how this data can be tied into standard core analysis results and mineralogy, and critically, how it can be cross-matched with NMR T 1 , T 2 and D distributions measured in the same rocks at the same saturations. These are the initial steps in building up a comprehensive database of measurements for pore volume- and saturation-controlled dielectric permittivity and effective conductivity at frequencies from < 10 MHz to 2 GHz in combination with the NMR response. The combined dataset can be used to build, test and improve upon, dielectric rock physics models based either on mixing laws, effective medium theory or on solutions of electrodynamic equations in 3D digital rocks. Calibrated measurements and interpretation models will maximise the value of advanced petrophysical methods and broaden their application. INTRODUCTION Several petrophysical methods can be used to determine the water and hydrocarbon saturation of rocks in the laboratory and downhole, but each is affected differently by matrix properties and mineralogy (especially clays), and each has different sensitivity to water salinity, and hydrocarbon type. In a conventional reservoir where clay content is low or well characterized, and formation water salinity is known, then a standard combination of density-neutron porosity with resistivity is routinely successful (Archie's method, and a plethora of shaly sand models). More "advanced" petrophysical methods, including nuclear magnetic resonance and dielectric logging have found applications where standard methods fail and where particular characteristics such as hydrocarbon type, wettability and mobility are required to discriminate pay from non-pay. They can help to predict key well performance characteristics such as production rate and water cut. Even in "Archie" reservoirs, dielectric tools, sometimes combined with NMR have been used to reduce uncertainty in residual water saturation (Bean et al., 2013; Pillai et al., 2015) .

ABSTRACT Digital rock physics (DRP), via both direct numerical simulation and pore-network modeling, holds great promise in terms of probing such pore-scale controls on transport, particularly with multiphase flow and sensitivity analysis of time-intensive measurements such as relative permeability. However, despite advances in micro-computed tomography (microCT) and scanning electron microscopy (SEM) techniques, obtaining cost-effective representative elementary volumes (REV) at sufficient resolution that capture dual-scale porosity and surface textures remains a formidable challenge in establishing digital rock physics as a predictive toolset. Furthermore, implementers are faced with several options of numerical solvers such as finite element modeling, lattice Boltzmann method, and mass balance-based pore-network modeling. This paper reviews the current status of establishing an REV and upscaling techniques for DRP in tight and/or diagenetically-altered rocks, highlighting successful and unsuccessful pore-to-core data benchmarking examples by the authors and the greater literature in terms of static and dynamic properties. The review finds that performing DRP on a single image modality is not sufficient, even for many conventional rocks, and that it is crucial to interface with experimental data, be it core analysis deliverables or subpore-scale and Darcy-scale microfluidics. In unconventional rocks, the majority of work does not leverage mesoscale simulations, instead zooming in to a discrete pore-scale scenario that is often not benchmarked with SCAL data. Even when a simulation domain is benchmarked, the matching of a discrete case with a multivariable situation is non-unique. Benchmarking with dynamic or pseudo-dynamic core data such as MICP and single phase permeability will greatly help reduce variables. Finally, this paper offers a technical roadmap for the robust application of unconventional DRP for the petrophysics and general subsurface community. INTRODUCTION Rock diagenesis can generate complex pore-lining and pore-filling textures beyond the idealized sedimentary "spherical grain pack" that greatly influence pore size distributions and transport properties including permeability, capillary trapping, diffusion, and relative permeability. Compaction, cementation, dissolution, and microporosity are examples of such geometric complexity. Meanwhile, mineralogical composition and organic matter content can lead to multiple surfaces of potentially varying wettability. Petrophysically-speaking, dispersed shale, laminated shale, and structural shale grains are categories of complexity as well. These various configurations often necessitate the need for visualization of rock pore systems, a practice that has been done for years via thin section and SEM imaging as well as computed tomography. Traditionally, imaging techniques have been used for validation of a model or assumption (such as laminated sands in shaly sand analysis), but, as computing power and microscopy technologies have increased, many researchers and vendors have leveraged these technologies to create digital laboratories where petrophysical properties can be directly calculated. This intriguing field of study is called digital rock physics (DRP) and is a potential addition to the petrophysical toolkit.

ABSTRACT Conventional petrophysical evaluation techniques are unreliable to assess individual bed properties in laminated depositional sequences with beds thinner than the vertical resolution of standard logging tools. The main cause of this limitation is that well logs average formation properties across multiple interbedded intervals. Commonly used solutions are limited to shaly sandstone models to account for either presence of graincoating clay in sandstones or laminated shale-sandstone systems. Other solutions rely on volumetric techniques which require subjective interpretation of volumetric concentration of shale and total porosity. Likewise, it is typically assumed that both sandstone and shale properties remain constant within siliciclastic reservoirs, which is not always the case in heterolithic bedding or in laminated sequences with strong diagenetic alterations. To address this technical challenge, we introduce an inversion workflow that reproduces measurements via analogues of thinly-laminated reservoirs. We use a Markov-Chain Monte Carlo inversion algorithm to generate independent realizations of each petrophysical property. All petrophysical properties are combined to estimate probability histograms rather than attempting to obtain a single value for each petrophysical property. The method is applied to a deepwater heterolitic clastic sequence of grain-coating clay sandstones where bed thickness varies from 3 to 4 inches. In addition to conventional well logs, high-resolution borehole images are used to detect bed boundaries. The statistical method is used to estimate total porosity, water saturation, and permeability based on the earth-model-derived properties. Finally, net-to-gross and hydrocarbon pore volume are estimated using the calculated statistical properties. Compared to conventional interpretation procedures, the formation evaluation method developed in this paper enables the incorporation of non-constant matrix and shale properties in the sandstone-shale laminated sequence, and estimates individual layer properties and their uncertainties, thereby reducing subjectivity in the interpretation of static and dynamic petrophysical properties of heterolithic clastic sedimentary sequences. INTRODUCTION Complex depositional systems generate spatially heterogeneous reservoirs with petrophysical properties difficult to quantify. The main challenge in studying thinly-laminated formations arises when the vertical resolution of traditional logging devices is smaller than the thickness of the beds. Conventional logging tools average formation properties in such sedimentary systems. In addition, in these environments the electric conductivity of shale dominates the response of conventional electric resistivity logs (Shray and Borbas, 2001). Hence, a typical petrophysical assessment of thinly-bedded shaly-sandstone systems leads to underestimation of hydrocarbon saturation (Pritchard et al., 2003) and poorly characterized permeability.

ABSTRACT Estimation of the petrophysical properties using imaging techniques and image analysis has been an active area of research in recent years. This has been primarily focused on 3D imaging and subsequent modeling of formation properties, termed Digital Rock Physics. This work requires an intact sample of material. When sample type or quality are not sufficient for 3D imaging, or where cost is prohibitive, modeling of petrophysical properties using thin section images is a viable, and for some samples, a superior alternative. This study involves the use of transmitted light image tiles from the light microscope. Each thin section is scanned in plane, cross-polarized, and fluorescent light. Each of these provide information about the framework and authigenic phase mineralogy. The advantages of thin section imaging for estimating petrophysical properties, include the ability to identify framework grain types for provenance and reservoir quality modeling studies, differentiation of detrital and authigenic phases, and establishing paragenesis. We apply a Carmen-Kozeny type model to thin section images of samples for which laboratory measurements of porosity and brine permeability were made. This has allowed us to model absolute permeability to within a factor of two for sandstone samples having laboratory-measured permeabilities ranging from 175 mD to 8D, and builds on a calibration data set of approximately 30 samples. For the original data set, modeled permeability agreed with measured permeability to within a factor of two, over a range of permeability from 8mD to 3D. A key input parameter to this model is an estimate of specific surface area. The 2D estimate of specific surface area, defined in Hathon, et al., (2003), is the ratio of the total porosity perimeter to the total area analyzed. The normalization to the total area analyzed, allows the number of image tiles analyzed to be varied for each sample. The estimated specific surface area is a function of image magnification, down-sampling and filtering of the image are applied during analysis. This work extends the previous model by adding a fractal analysis to the estimate of specific surface area. This allows us to account for the impact of magnification and processing when modeling permeability from 2D image data. Total porosity from 2D image segmentation agrees to within +/- 2% of laboratory measured porosity. Future work includes developing a model for converting 2D pore body size distributions to 3D equivalents for comparison to NMR based estimates of pore body size, together with calibration of the surface relaxivity term. We also intend to investigate the relationship between the estimated tortuosity in thin section and Formation Factor, and estimating pore throat size distributions for modeling Mercury Injection Capillary Pressure (MICP) data.

ABSTRACT Clastic rocks with variable grain sizes exhibit different flow properties depending on how the various grain sizes are geometrically arranged within the grain pack. Grain arrangement must be assessed to quantify dynamic petrophysical properties when rocks exhibit non-unimodal grain-size distributions. A non-unimodal throat-size distribution will be observed, for instance, in cases where different grain sizes are arranged in laminated form (e.g., aeolian sandstones). Clastic rocks that have been subject to extreme diagenesis and recrystallization, such as tight-gas sandstones, often exhibit bimodal grain and pore-throat size distributions. This paper investigates the impact of grain arrangement on permeability and capillary pressure in clastic rocks that exhibit multiple grain sizes. Two extreme cases are studied for grain packs that include variable grain sizes: when grains are (i) randomly dispersed, and (ii) laminated in the grain pack. Equations are derived to calculate permeability in each case. Additionally, shale concentration is accounted for in the calculation of permeability for both laminated and poly-dispersed grain arrangements. A three-dimensional chart is constructed to illustrate the behavior of permeability with respect to rock type fraction and shale concentration. The assessment includes the calculation of permeability anisotropy resulting from grain-size laminations. Synthetic samples of grain packs are also constructed and subject to pore-scale fluid flow simulations to calculate permeability and throat-size distribution and to examine how these properties change with different grain-size arrangements. Finally, a new rock classification method that considers grain arrangement, capillary pressure, shale concentration, and permeability is introduced and verified with measurements acquired in a Carboniferous tight-gas sandstone from northern Germany. Our method of rock classification yields improved permeability calculations compared to widely used classification methods such as Winland R35, which implicitly assume a unimodal throat-size distribution. The new rock classification method can be readily adapted to calculate more specialized fluid-transport properties such as relative permeability. It can also be modified to account for capillary pressure during both imbibition and drainage and their consequence of saturation-height behavior.

ABSTRACT In complex geological environments, interpretation methods based on fast modeling and inversion procedures deliver better estimates of petrophysical properties than conventional methods. Neutron measurements are affected by high-order formation and borehole effects. Their depth of investigation is also very sensitive to porosity, borehole size variations, and fluid and rock properties. Consequently, reliable petrophysical interpretation of neutron logs under complex rock and geometrical conditions requires fast modeling methods. We develop a fast-forward algorithm for a commercial LWD neutron tool. The algorithm is based on perturbation theory, flux sensitivity functions (FSFs), and diffusion flux-difference (DFD) approximations. The DFD method interpolates between Monte Carlo (MC)-derived, base-case FSFs using one-group diffusion models and a Rytov approximation. Diffusion approximations successfully capture sensitivity flux perturbations: neutron porosities simulated using DFD-perturbed and MC-derived FSFs agree within one porosity unit (p.u.) in highly deviated wells, enlarged boreholes, and wells with invasion. They significantly outperform linear interpolation approaches, reducing errors in estimated porosity by as much as 10 p.u. Even when diffusion approximations are applied, perturbation theory may still yield inaccurate results when compared to neutron porosity estimated from MCNP counts for regions exhibiting contrasting properties, and primarily enlarged boreholes. We introduce a new two-step algorithm to improve neutron modeling accuracy to MCNP counts in the presence of standoff. The algorithm is based on two sets of base cases: one for detector counts, and the other for sensitivity functions. On one hand, diffusion flux-difference approximations are used to compute perturbed sensitivity functions for any tool, borehole, and formation configuration. On the other hand, count base cases obtained for 21.6-cm (8.5-in), 24.1-cm (9.5-in), and 26.7-cm (10.5-in) boreholes extend the validity of the Taylor series expansion by minimizing the size of the borehole perturbations. Compared to neutron porosity calculated from MCNP counts, the new algorithm yields relatively low errors in enlarged boreholes. Comparison benchmarks against synthetic and field cases in vertical and deviated wells confirm good agreement with Monte Carlo simulations and field logging data, respectively. For the synthetic cases, vertical and horizontal wells were assumed with the tool penetrating several bed layers with contrasting hydrogen index. Monte Carlo simulations were carried out to simulate neutron detector counts and to compare against results obtained from our fast-forward algorithm. The fast-forward algorithm was then applied to a field case featuring a high-angle well penetrating a Norwegian sandstone with light hydrocarbons; good agreement was obtained with field logs.

ABSTRACT Reservoir petrophysics is the most important study that defines qualitatively and quantitatively a reservoir's performance. The reservoir itself is characterised as clean, heterogeneous or poor in quality based on its mineralogical composition. Of the crucial minerals that decide reservoir quality are clay minerals. Not only do the clays affect the reservoir quality but also play a key role in understanding source rock and hydrocarbon-generation, as a tool for depositional environment characterisation, stratigraphic correlation and identification of exploration targets. Hence, understanding the clays is essential. The Fulmar Formation is the principal reservoir within the North Sea Central Graben. The Upper Jurassic Kimmeridge Clay Formation is the source rock, although coals of the Middle Jurassic Pentland Formation are also locally mature for oil (Isaksen 2004). Besides gas and black oil, many of the high-pressure, high-temperature (HPHT) Fulmar reservoirs comprise condensate. The association of overpressure and clay diagenesis appear to be the main factors in controlling reservoir quality and the migration of pore-fluids and hydrocarbons. The reservoir consists of thick highly bioturbated sandstones, which were deposited on a subsiding shallow-marine shelf under the influence of tectonic movements. The formation has been cored a lot, and is mostly homogeneous in appearance. No preserved primary sedimentary structures can be seen in the fine to very fine sandstones due to rigorous bioturbation. Although Kaolinite, Illite, mixed-layer Illite/Smectite, Chlorite, Smectite, and mixed-layer Chlorite/Smectite have all been identified in the Fulmar reservoirs, it is observed that Illite is dominant. The clay mineralogy of the Fulmar sandstones differs between wells, and may vary with depth within a well. Since clay mineralogy leads to significant variation of porosity, permeability, and wireline log response there is need for a systematic study of clays during the exploration and appraisal phases of a field. Clays seem to have both a positive and negative effect on the reservoir quality of the deep Fulmar sandstones. The deep HPHT Fulmar reservoirs have quite high porosities which have been attributed to conversion of smectite to illite, inhibiting macro-quartz cementation, thereby preserving porosity (Osborne & Swarbrick, 1999). Thus, over-pressuring coupled with early clay coatings has prevented compaction at depths >10,000ft. Impact of overpressure on reservoir quality will be active only when the reservoir is sealed suitably for long duration by very fine-grained and ductile rocks. Notably, the distinct relationships between facies and reservoir quality appear to be closely related to the bulk-rock total clay content. Porosity and permeability are functions of depth, clay content and grain size. Increasing clay content means increasing ratio of microporosity to macro-porosity. For reserve calculations, the ability to determine from wireline logs the total clay content in hydrocarbon-bearing reservoirs is very important. Distribution of these clays in the sandstones may be dispersed, laminated, or structural and these variations result in different facies, which if not understood could lead to misinterpretations. This study demonstrates how integrated analysis through petrophysical and geological methods could help interpret clay type, volume, distribution and morphology for reservoir characterisation from well data in the Fulmar reservoirs.

ABSTRACT RCAL/SCAL characterization as other geoscientific modeling methods exhibits the inherent lack of ubiquity and sampling bias due to the finite nature of experimental resources, samples and time. Lab based X-Ray technology can be applied in any rock sample from plugs to cuttings and radiographic inspection can be performed in seconds per sample helping to diminish such sample/time bias. X-Rays interacts with matter based on sample density and atomic number depending on the material under inspection and the accelerating voltage used to generate the X-Rays. Reservoir properties like porosity, permeability and acoustic response are also causally related to these scalars (density and atomic number). The purpose of this work is to study the predictive potential of radiographic information quantified by X-Ray attenuation statistical cumulants for some RCAL/SCAL properties. Radiographic screening can be performed in seconds per sample providing a cost/time efficient method to obtain sample data. 15 samples were chosen from a well with several siliciclastic intervals exhibiting different rock qualities. A RCAL/SCAL protocol was performed in these samples to measure porosity, permeability and acoustic response using conventional methods. A fast radiographic screening was then applied to the complete universe of samples. Each sample radiograph was around 5 seconds of exposure time which gave a total time of less than one hour for the complete X-Ray experiment. A linear multivariate scheme was applied to this X-Ray attenuation response. The local rock properties (porosity, permeability and acoustic velocities) were considered as independent variables while several statistical cumulants from the registered X-Ray attenuation were used as a the linear vector base. The results showed a significant predictive power of those X-Ray attenuation statistical cumulants for porosity, permeability and acoustic velocities. Porosity and Permeability exhibited the higher possibility to be predicted using the derived linear coefficients (R 2 >0.9 and R 2 >0.79 respectively) while acoustic velocities showed a lower level of prediction (R 2 >0.7). The acoustic property prediction could be improved by multi-energy acquisitions protocols which will be studied somewhere else. Cross validation was applied to the complete set with errors within a reasonable geological uncertainty in the interval. The constructed linear model could allow faster/cheaper data population along the complete prospect in this well for a more precise earth model built by adding pseudo-data points from X-Ray information where no RCAL/SCAL was obtained. Other well/lithology models are needed to verify the robustness of the procedure; also multi-energy acquisitions are suggested for more accuracy in acoustic response estimation.

ABSTRACT Conventional permeability models based on Mercury Injection Capillary Pressure (MICP) curves, such as the Katz-Thompson model, require estimation of microscopic length parameters from core MICP experiments. In wells lacking core data, these models cannot be applied to estimate permeability. On the other hand, conventional Nuclear Magnetic Resonance (NMR) permeability models can provide consecutive permeability assessment along the wellbore, but they are often unreliable in complex formations. We herein propose a new model for depth-by-depth directional permeability assessment in complex formations, by combining borehole NMR and directional electrical resistivity measurements. The objectives of this paper are (a) to introduce a new NMR characteristic T 2 value to represent the characteristic length in real pore size distribution that controls rock permeability, (b) to establish a new directional permeability model combining NMR and directional resistivity measurements, and (c) to test the reliability of the new model for permeability assessment in formations with multi-modal porosity (e.g. carbonate and tight sandstone formations). We numerically simulate the directional permeability, directional formation resistivity factor (FRF), and NMR T 2 relaxation curves, in 202 digital rock samples using our state-of-the-art pore-scale simulation software. The digital rock samples are obtained by multi-scale imaging on real rock samples, including dolostone, limestone, chalk, (tight) carbonate, (tight) sandstone, and sandpack samples. The simulated directional permeability is treated as the target reference for each digital rock sample. Next, we define a new NMR characteristic T 2 value to represent the microscopic length parameter in real pore size distribution. The new NMR characteristic T 2 value is minimally influenced by diffusional coupling effect that distorts NMR T 2 distribution in multi-modal pore systems, and we numerically and theoretically demonstrate this property. Then we develop a new NMR-Resistivity model to estimate directional permeability by combining NMR characteristic T 2 and directional FRF. The new permeability model is calibrated on 32 digital rock samples and successfully tested on the other 170 samples, with permeability ranging from 8.4x10 -5 mD to 48000 mD, spanning ten orders of magnitude. Results show that the model-estimated permeability is in good agreement with the target permeability for all rock types. Furthermore, we compare the model-predicted permeability with lab permeability on 16 carbonate core samples. We upscale the NMR characteristic T 2 and FRF from pore-scale to core-scale, and then estimate the core-scale permeability by the proposed NMR-Resistivity model. Excellent agreement between the model-estimated and lab permeability demonstrates the reliability of the new model on the core scale. Our proposed permeability model can be applied to wells with NMR and directional resistivity logs, overcoming the requirement for core measurements. It provides accurate and consecutive directional permeability assessment along the wellbore. It also shows a potential in estimating relative permeability in hydrocarbon-bearing formations. The outcomes of this research can significantly improve permeability assessment in complex reservoirs with multi-modal porosity, including carbonate, tight sandstone, and organic-rich source rocks.

ABSTRACT Electromagnetic formation evaluation currently relies on low-frequency resistivity and high-frequency dielectric measurements that are typically not interpreted jointly. In consideration that formation electromagnetic responses in different frequency ranges are controlled by different physical phenomena, analysis of a wideband electromagnetic response can provide new and complementary sensitivities to formation petrophysical parameters. The frequency-dependent complex conductivity of ion-conductive sediments in the range from millihertz to kilohertz exhibits the spectral induced polarization (SIP) effect, in which the impedance phase has a near-resonance peak at a characteristic frequency due to a strong polarization response. In this study, SIP spectra were measured on a collection of quarried clean sandstones saturated with brines. The influence of other factors on the SIP effect, such as the presence of clays, was minimized by carefully selecting samples. The dielectric dispersion was measured to characterize a subset of twin samples in the megahertz to gigahertz range. The combination of these methods provided core electromagnetic responses over 12 decades of frequency. We established a wideband rock model based on a differential effective medium approach that accounts for both the Maxwell-Wagner interfacial polarization related to the rock texture and the electrochemical polarization due to the presence of charged grains. The model is based on first principles and uses a minimal number of parameters to describe the essential electromagnetic properties of well-sorted clean sandstones in the millihertz to gigahertz range. We investigated the relationship between the SIP effect and the dominant grain size of our sandstone collection. The dominant grain size was determined using a digital image analysis of scanning electron microscope (SEM) images obtained on thin sections. SIP spectra were inverted with the rock model to obtain the dominant grain size. The model was shown to be capable of reproducing well the experimental SIP spectra, with the inverted dominant grain size comparing favorably with values determined from image analysis. We analyzed the wideband electromagnetic measurements by applying the rock model in the full frequency range. The wideband data inversion enabled the estimation of five formation parameters: water-filled porosity, water salinity, cation exchange capacity, dominant grain size, and cementation exponent. Our analysis also demonstrated that the use of only low- or only high-frequency data subsets is not sufficient to reliably invert for the full set of formation parameters.

ABSTRACT Clastic laminated reservoirs have historically posed difficulties for subsurface groups engaged in formation evaluation. Difficulties are largely due to convoluted log responses which preclude accurate assessment of key petrophysical properties such as thin sand bed porosity and water saturation. In South East Asian (SEA) basins the abundance of silt in reservoir and non-reservoir rocks adds another layer of complexity to the formation evaluation and directly affects the design of appropriate data acquisition programs. This paper describes the silty thin bed problem by assessing the efficacy and uncertainties of various log measurements to arrive at the correct petrophysical solution. A review of rock physics literature is presented to highlight the underlying reasons for unusual log behavior in silty facies. Generally, laminated rocks are evaluated from two different approaches: high-resolution or bulk rock volume. High-resolution approaches include borehole image logs, de-convolution, and digital core imaging analysis. Bulk rock (or volumetric) approaches generally use Thomas-Stieber, multi-component resistivity, and nuclear magnetic resonance (NMR) techniques. Three wells drilled in different sedimentary basins in SEA were selected to demonstrate the theory, challenges, and pitfalls of the most common approaches and techniques. The Thomas-Stieber approach is often regarded as the most suitable for a binary sand-shale system and if conditional assumptions are met, results in a linear trend from which net-to-gross can be calculated. Adding a third component, such as silt, violates the assumptions and distorts this trend into a non-linear "boomerang" shape. Resistivity anisotropy, i.e., vertical/horizontal resistivity ratio (Rv/Rh), provides further necessary input for accurate formation evaluation in laminated sand-silt-clay systems. Rv is a key measurement because it is very sensitive to hydrocarbons in thinly laminated sands or silts. Additional information, like borehole image and NMR data, contribute to reducing net-to-gross uncertainty or understanding the reservoir geometry. Where available, the saturation height function results are compared to multi-component resistivity results. One very silty to fine-grained sand reservoir in Vietnam, displays anisotropy due to grain-size variation on a very fine level, in addition to the laminar sand-shale phenomenon. In this example, the relevance of shale laminar estimations is questioned and can only be justified with detailed core studies. It is, however, argued that reliable identification of hydrocarbon-bearing silt-rich sequences is only possible with multi-component resistivity data. In addition, quantification of sand lamina resistivity, Rsand, is still possible in these silty sands with variable amounts of irreducible water. Although many papers discuss the thin-bed formation evaluation problem, very few publications address issues related to laminated sand-silt-clay reservoirs. This paper partly addresses this literature scarcity. Identification and accurate quantification of silt-rich laminated reservoirs is of critical importance in basins with a high proportion of silt.

ABSTRACT Gassmann equations (Gassmann, 1951) are used to calculate seismic velocity changes that result from variations in reservoir fluid saturation. These equations became predominant in the analysis of a direct hydrocarbon indication from seismic data through their use in analyzing the compressional to shear velocity ratio, Vp/Vs. This Vp/Vs ratio is used in many industry analyses, such as the amplitude variation with offset (AVO) analysis developed by Castagna et al. (1993). Multiple authors have since published a variety of Vp/Vs seismic interpretation techniques that use empirical relationships with Vp, Vs, and porosity terms. Unfortunately, however, there is a gap in the use of Vp/Vs relationships in petrophysical interpretation. The Vp/Vs ratio analysis was expanded in 1995 when Brie et al. proposed the application of a Vp/Vs vs. Vp crossplot for gas trend indication and included a correction for shale effect. The crossplot of Vp/Vs vs. Vp was published in 2015 by Quirein et al. and was applied to organic shale reservoirs for kerogen volume and anisotropy trend indications. This paper explores the use of a crossplot of Vp/Vs vs. Vs for quantitative petrophysical interpretation. A relationship developed in the paper is used to describe water-wet and gas-saturated sandstone trends, and to independently calculate water saturation from a proposed crossplot in low and medium porosity isotropic sandstones. These proposed Vp/Vs vs. Vs crossplot water saturation results are compared to traditional resistivity-based results. This proposed simplified method provides a suitable approach for determining gas saturation when resistivity logs yield inadequate results in, for example, medium porosity or low-resistivity pay formations.

ABSTRACT Permeability is a fundamentally important property of reservoir rocks that governs the flow of the reservoir fluids and production rates. Many methods have been developed to measure permeability, including well established and documented laboratory measurements on whole core and plugs, as well as analysis of formation test data, analysis of production and well tests data, acoustic and Stoneley Wave data analysis. In addition, obtaining permeability from NMR T 1 or T 2 has proven to be a cost effective method that can provide continuous permeability along a wellbore. This method relies on a transform that is applied to NMR well log data to obtain permeability. The NMR permeability transform requires calibration to fit local data and is usually developed based on measured data from representative cores from the reservoir of interest. We found that commonly used NMR permeability transforms do not work for our tight sandstones, and, hence we developed a general method to optimize permeability transform for any given reservoir from laboratory data and applied the method to a tight sand reservoir. Specifically, we report on: A systematic method developed to search for an optimal NMR permeability transform based on NMR and permeability data. This method, coded in Matlab, can be used to obtain an optimal NMR permeability transform for any reservoirs when core data are available. Application of the method to a Middle East tight sand to establish NMR permeability transforms for this reservoir. The role of clay and high-Z minerals on NMR relaxation time and permeability for a Middle East tight sand. The new transforms developed for the Middle East tight sandcan be used to calculate permeability along a wellbore from NMR wireline and/or LWD log data. The permeability results, the permeability profile, could then be used to evaluate the reservoir heterogeneity and select high grade zones for hydraulic fracturing. This may reduce the need for additional extensive coring and core permeability measurement.

ABSTRACT Defining petrophysical and mechanical properties of target and barrier zones are key components of the hydraulic fracture modeling process; subsequently, the selection of the detail necessary to accurately model fracture/reservoir performance is challenging. This work investigates whether using detailed petrophysical and mechanical properties provides fracture design parameters that better represent actual fracture behavior and subsequent well performance than a single-layered model. The approach was to model an existing hydraulic fracture treatment and well performance from a well located in the northern Delaware Basin producing from the lower Brushy Canyon Formation. Models varied from a single layer model with simple-averaged, petrophysical properties to a fine resolution 1-ft model with detailed petrophysical values. Detailed core descriptions were constructed to appropriately represent the thin-bedded and micro-laminated sandstones and siltstones. In addition, point load tests measured values of fracture toughness for specific lithofacies from 600 to 1100 psi-in½. In comparison, the default value for a sandstone system is 1000 psi-in½. Other mechanical properties, e.g., Poisson's ratio and Young's modulus were derived from well logs, and were within typical values. For the fracture modeling phase, the actual treatment volumes, rates and pressures were inputted into the model along with the measured petrophysical and mechanical properties. Model net pressure was matched with the actual values to verify the output. The dimensionless fracture conductivity (FCD) from the various models ranged from 4.8 to 13.6. The range depends on the variation of lithofacies included in the fine resolution models and their associated mechanical/petrophysical properties. Adding micro-laminated and bioturbated siltstones at the expense of clean sandstone in the finer resolution models resulted in higher permeability, fracture toughness and lower stress gradient. For the production history matching phase, simulation pressures were significantly overestimated compared to actual measured bottomhole pressures for all single layer models regardless if actual or default mechanical properties were used. The overestimation reflects a threefold increase in pore volume due to the single layer values. For the finer resolution 1-ft model, the simulation pressure was significantly below measured pressure values using default mechanical properties. However, using actual mechanical properties in the 1-ft resolution model resulted in an increase in the FCD due to the decrease in fracture toughness and stress gradient input values. As a result, a very good match was obtained between simulation and actual pressures; indicating the 1-ft model with the measured mechanical properties is a good representation of the actual reservoir system.