A system of exceptionally well-exposed, early-burial fractures in outcrops of Devonian carbonates on the Lennard Shelf, Canning Basin, Western Australia, serves as an analogue for fractures in deep reefal carbonate petroleum reservoirs in the Precaspian Basin of Kazakhstan. Data from early fractures were collected at Windjana Gorge, which crosscuts such a reefal depositional system. Less extensive evaluations were also made at three auxiliary sites along the Lennard Shelf. These early-formed, near-vertical fractures strike parallel or normal to the strike of the Devonian depositional margin. Fracture density correlates with the textural character of the host rock, with the highest values in boundstone-rich rock, lowest values in grain-rich rocks, and intermediate values in rocks of mixed assemblage, including breccia. Fracture density is also controlled in part by the mechanical character of bedding at the time the fracture system developed, syndepositionally or soon after. The environment of deposition (EoD) shows a weaker correlation with fracture density. At Windjana Gorge, fracture height correlates with the EoD, with the largest values in the reef core, progressing serially to shorter average height in the upper slope, middle slope, reef flat, and platform top/reef flat (a transitional EoD). To better understand the fracture size distribution within a reservoir that can influence fluid drainage, we introduce a new parameter, the intersected fracture height density (IFHD). This describes the aggregate fracture height connected directly to a traverse line or borehole. Because IFHD requires the measurement of fracture heights, it is measurable only at outcrop analogue exposures.

Using image logs from horizontal wells in the asymmetric East Painter Reservoir anticline in Wyoming, scanlines and petrographic observations from nearby outcrops, and methods that allow statistical quantification of fracture spatial arrangement patterns (including clustering), we show that in early Jurassic Nugget Formation sandstone degree of clustering and intensity of fractures varies with structural position. In the backlimb, fractures are markedly more clustered than random, and hierarchical and regularly spaced clusters are present, but intensity is low. In the forelimb, arrangements are mostly indistinguishable from random, but intensity is high. Based on image log response, the backlimb contains a higher proportion of open fractures, although the forelimb has greater numbers of open fractures. Spatial arrangement and intensity patterns are similar in a forelimb outcrop. Petrography reveals that quartz seals fractures less than 0.15 mm wide. Wider fractures are open with quartz rinds or are quartz lined and sealed with calcite or filled by quartzose cataclasite. The proportion of open fractures depends on quartz and calcite deposits and closure by reactivation (shearing) of pre-existing quartz -lined fractures. Based on increased abundance and greater strike dispersion, results suggest that in forelimbs, high fracture intensities reflect shear on preexisting fractures, possibly resulting in more numerous but less spatially correlated open fractures. The most prolific gas and water production wells are in the backlimb, sug-gesting that clustered but sparse quartz-lined open fractures are more effective fluid conduits than closely spaced partly damaged/sheared arrays.

SummaryWe present here a practical method for estimating the directional permeabilities in anisotropic reservoirs. The method uses pressure transient-analysis results from at least three sets of interference/pulse tests among wells offset at different azimuths. Knowledge of the maximum/minimum permeability directions in anisotropic reservoirs helps to optimize injector/producer locations and is important for reservoir management, especially under secondary/enhanced recovery of hydrocarbons. The proposed method uses transient-test data rich with dynamic information to provide fieldwide permeability distribution at well-spacing scale, which is relevant for estimating fluid movement and recovery. Its application in a carbonate oil field in Kazakhstan is also discussed.The proposed method uses well coordinates and multiple sets of analysis results of interwell transient tests through mathematical matrix operations. It is straightforward to use and works efficiently. The algorithms to calculate directional permeabilities in anisotropic homogeneous reservoirs from interference tests were first introduced by Ramey (1975) and extended to pulse tests by Kamal (1983). Our new approach can use the analysis results of any type of interwell transient test directly in heterogeneous reservoirs. Any valid modern methods can be used to analyze each interwell test, and all analysis results can be integrated to generate the field permeability-tensor map.The proposed method was validated using synthetic cases. Its application in a large set of multiple-well tests in a naturally fractured reservoir illustrated its practicality and efficiency. Extensive interwell transient data have been collected and analyzed from carefully designed and conducted tests among 12 wells in Korolev Field, a carbonate field in Kazakhstan. Of the 12 wells, 10 have interwell tests at three different directions, which allows the calculation of directional permeabilities. The permeability-tensor map is generated for the entire field and compared with the fracture orientations derived from geological-structure and image-log interpretation. Both static and dynamic data resources indicate that fracture orientations vary substantially throughout the field. In some areas, the dominant permeability directions from interwell transient data are consistent with those from image-log interpretation. However, they differ in other areas, emphasizing the need for using dynamic measurements at well-spacing scale for better understanding of fracture distributions/orientations and their effects on flow communication among wells.The novelty of this method of estimating directional permeabilities is that it uses well coordinates and analysis results of individual interwell transient tests directly in heterogeneous reservoirs. It is convenient and efficient. It can be easily used to generate a fieldwide permeability-tensor map using dynamic transient data. Its application in a large carbonate reservoir demonstrates its practicality, even in fields with complex varying anisotropy. Integrating the results from this method with those from geological and petrophysical analyses reduces uncertainty in reservoir characterization. This method has already been implemented in some commercial well-test-analysis software.

Fracture pattern development has been a challenging area of research in the Earth sciences for more than 100 years. Much has been learned about the spatial and temporal complexity inherent to these systems, but severe challenges remain. Future advances will require new approaches. Chemical processes play a larger role in opening‐mode fracture pattern development than has hitherto been appreciated. This review examines relationships between mechanical and geochemical processes that influence the fracture patterns recorded in natural settings. For fractures formed in diagenetic settings (~50 to 200 °C), we review evidence of chemical reactions in fractures and show how a chemical perspective helps solve problems in fracture analysis. We also outline impediments to subsurface pattern measurement and interpretation, assess implications of discoveries in fracture history reconstruction for process‐based models, review models of fracture cementation and chemically assisted fracture growth, and discuss promising paths for future work. To accurately predict the mechanical and fluid flow properties of fracture systems, a processes‐based approach is needed. Progress is possible using observational, experimental, and modeling approaches that view fracture patterns and properties as the result of coupled mechanical and chemical processes. A critical area is reconstructing patterns through time. Such data sets are essential for developing and testing predictive models. Other topics that need work include models of crystal growth and dissolution rates under geological conditions, cement mechanical effects, and subcritical crack propagation. Advances in machine learning and 3‐D imaging present opportunities for a mechanistic understanding of fracture formation and development, enabling prediction of spatial and temporal complexity over geologic timescales. Geophysical research with a chemical perspective is needed to correctly identify and interpret fractures from geophysical measurements during site characterization and monitoring of subsurface engineering activities.

SummarySimple but geologically reasonable and calibrated 2D stochastic models are useful to quantify significant risks and uncertainties associated with alternative-development-well trajectories, particularly when statistical relationships can be established to help quantify those risks and uncertainties, and when the geologic features that create the risks and uncertainties are not adequately addressed within reservoir-flow models.Our example stochastic 2D model considered the naturally fractured depositional-slope region of an isolated carbonate buildup, and the model was populated with relevant features including distributions and geometric details of natural fractures, natural-fracture clustering, and intraformational slope clinoforms that define a mechanically layered sequence. The model was calibrated by use of well-production results and production-logging data so that it reproduced observed well results for cases where the lower sequence boundary does not occur above the oil/water contact (OWC), adding confidence that the model could be used to represent the statistical effect of various alternative trajectories for future wells.Experimental design (ED) was used to determine the significant uncertainties and well-path decisions. Heel and toe elevation and the number of clinoforms encountered by the well were the only significant variables for modeling the frequency of water production. For modeling the frequency of direct well communication to the gas cap, the same variables were significant, in addition to well direction, completion length, and fracture density. The amount of fracture clustering applied in the model was also significant. For our example case, changing the well-elevation profile was effective in managing gas or water risks; however, tradeoffs were evident—and quantified—in attempting to simultaneously address both risks. Minimizing drawdown was not an effective strategy because productivity was low and rarely resulted in economic water-free production if any open fracture connected the well with the aquifer.

SummaryA systematic work-flow to integrate pressure-transient data collected from single-well buildup tests in numerical reservoir-simulation models for a fracture/matrix system is presented. The results of its application in a sector model in the southeast region of Tengiz field in Kazakhstan are also discussed.The procedure starts with a selected numerical-simulation model, either a discrete fracture/matrix (DFM) model or a dual-porosity dual-permeability (DPDK) model, and follows with the analysis of a numerically generated buildup test to calculate the fracture spacing and shape factor of the model. Then, following the correlations between pressure-transient-analysis results and the average or representative values of the model-input parameters near the well, which contain the previously obtained fracture spacing and shape factor, the numerical-model parameters are adjusted in each iteration to match the pressure-transient behavior observed in the buildup test including the interporosity flow between matrix and fracture and the radial flow in the total system.Before field application, the numerical-simulation results from both DFM and DPDK models were validated against analytical pressure-transient solutions for a dual-porosity system. The gridding and time-steps were calibrated to reproduce the analytical transient behavior. Finally, the new work-flow was applied to a sector model of Tengiz field in the southeast region focusing on two wells. Following the developed work-flow, a DFM model was constructed, and its fracture and matrix properties were adjusted to honor buildup-test data at both wells and the transient data collected during a pulse test conducted between them. The study results show that the key factors of a DFM model on buildup transient response are the fracture permeability, fracture aperture, and matrix permeability in the well-drainage area, and the dominant parameters on pulse-test response are the fracture permeability and matrix porosity in the influence area between the two wells. Using the correlations quantitatively for each simulation step could reduce the total number of iterations needed to converge to the numerical solution. The modified model also generated flow distribution along the wellbore, consistent with production-logging data at one well. The resulting sector map of pressure change during buildup test indicates the area with well-connected fracture network. Dynamic transient data contain rich information about reservoirs, and the effective integration of dynamic and static data would have a big impact on reservoir management by potentially minimizing the number of wells to be drilled, maximizing the production, and optimizing recovery.The novelty of this study is the quantitative use of the correlations between pressure-transient-analysis results and the representative values of the input parameters in a numerical model to reduce the number of simulation iterations. Its application in Tengiz is also one of the rare examples in which single-well and multiple-well transient data, production logging, and image-log data are all available.

The determination of petrophysical properties in carbonate rocks is strongly affected by depositional heterogeneity at different scales, complex mineralogy, and diagenetic modification. Superimposed complex depositional and diagenetic processes typical in carbonates lead to a very wide range of pore sizes involving many orders of magnitude of length scales. Pore types are a primary control on fluid movement in the reservoir and constitute the primary variable in determining petrophysical rock types. Multimodal pore systems and irregular water distribution across the field caused by complex charging history present challenges in water saturation modeling using conventional saturation-height functions. Bitumen quantification presents another challenge in carbonate reservoirs, especially in fields without nuclear magnetic resonance logs. Fractured reservoirs entail uncertainties related to characterization, modeling, and enhanced oil recovery aspects that involve matrix oil recovery via higher permeability fracture networks.These challenges exist in Tengiz Field, a Paleozoic isolated carbonate platform reservoir in the Precaspian Basin in Kazakhstan one of the world's deepest and most prolific supergiant fields with more than 25 billion barrels original oil in place. The buildup includes: productive platform-top settings composed of grainstone and packstone with an average porosity of 8%; and fractured margin and slope environments predominantly consisting of boundstone, breccia, and grain-dominated deposits with an average porosity of 4%. Tengiz exhibits primary stratigraphic and depositional heterogeneity and has an extensive and complex diagenetic history that greatly modified original pore systems and affected present-day reservoir quality. This paper describes how the above challenges were addressed in petrophysical characterization of Tengiz field. The solutions include special core and log acquisition programs, bitumen modeling with Multimin calibrated with TOC data, comprehensive rock typing, hybrid water saturation modeling using pore based saturation-height functions and bulk volume water approach, total permeability from production logs, baffles identification from pressure gradients, and integrated fracture characterization.

Natural fractures bear significant influence on productivity in Tengiz field, which is one of several giant light-oil accumulations trapped in isolated carbonate platforms in the Pricaspian Basin of Kazakhstan. Outcrop analogs are particularly important for understanding reservoir fracture systems because many aspects of fracture character (e.g. height, length) are impossible to measure with subsurface data. The Devonian margin of the Canning Basin in NW Australia presents a well-exposed outcrop analog for steep margin and slope deposits of Tengiz field. Fracture data gathered from Tengiz core and image logs suggest affinity to fractures in the Canning outcrops in terms of origin, orientation, and range of density. Inclusion of additional information - gained through outcrop study - into reservoir fracture description leads to improved understanding of stratigraphic influence on their occurrence and character. Shallow-burial fractures – those formed in carbonate strata prior to significant burial, including neptunian fractures – are the most important for reservoir productivity at Tengiz field. These fractures dip steeply and strike dominantly parallel and/or normal to the local orientation of the depositional margin. They are most well-developed in brittle, boundstone-dominant facies of the outer-platform to upper slope environment. Dissolution by corrosive fluids following burial led to enlargement of fracture apertures, which range from small to cavernous. In Tengiz field, cavernous fractures pose both high lost-circulation risk, as well as the reward of highly productive wells. Outcrop data from the Canning Basin show fractures in the mid- to upper-slope facies and reef core are, on average, not limited by bedding and hence much taller than fractures in the reef flat and outer platform areas. Fracture size cumulative distributions are mainly exponential, and they differ between stratigraphic settings. We expect such size differences will have important effects on fracture connectivity and permeability in a reservoir. Fracture density, which is measured routinely in the Tengiz reservoir, was measured with long pseudowells (i.e. scanlines) “drilled” along Canning outcrops. Fracture density shows significantly less variation among facies than does fracture height. Thus, outcrop-based data can add substantially to our understanding of key fracture system characteristics that are unavailable from well data alone.

Tengiz Field is a steep-sided, isolated carbonate platform in the Precaspian Basin, Kazakhstan, with hydrocarbon production from Carboniferous platform and slope facies. Systematic differences in reservoir pressure decline during production indicate that this reservoir consists of three subcompartments or material balance regions: (1) a central "platform reservoir" made up of cyclic platform-top facies that acts like a single, stratified, multistory reservoir; (2) a "wedge reservoir" formed by a prograding margin containing upper slope microbial facies; and (3) an "apron reservoir" containing allochthonous facies deposited in deep water around the base of the buildup. The facies in the apron reservoir accumulated during an early depositional stage and were subsequently partly to fully buried by prograding microbial slope facies of the wedge reservoir. The wedge and apron reservoirs together form a succession 800 to 1000 m thick within the Tengiz oil column.The wedge reservoir shows uniform pressure decline with time and is well connected. Field data (cores and well logs) are insufficient to determine internal continuity of lithofacies and depositional environments or to quantify the pore network responsible for the high connectivity. An outcrop analog (Asturias, Spain) with facies matching those observed in Tengiz cores was used to predict that the microbial lithofacies form a distinct and continuous mechanical unit within the wedge reservoir. Tengiz microbial facies contain a high concentration of solution-enlarged, syndepositional and other early fractures oriented parallel and normal to depositional strike. Borehole image logs provide data on enlarged fracture apertures and local fracture density, but no data related to fracture height or length. An outcrop analog with early fractures in similar facies (Windjana Gorge, Australia) was used to obtain large-scale height and spacing data for solution-enlarged syndepositional fractures. Dissolution processes in the outcrop are different from those of Tengiz, but the fracture aperture and cavern sizes are comparable to their known counterparts in the Tengiz wedge reservoir, and application of the outcrop height data to geologic models of the Tengiz wedge subcompartment can account for its dynamic behavior. The apron reservoir shows a nonsystematic pressure decline with time and is less depleted than the wedge reservoir. The irregular decline indicates reduced internal connectivity within the apron reservoir, which is corroborated by core and borehole image data indicating high lithofacies heterogeneity and the absence of continuous microbial facies responsible for reservoir continuity in the wedge reservoir. A reservoir pressure increase of 1700 psi from the wedge reservoir to the apron reservoir observed in a single well penetration suggests reservoir communication between them may be reduced across a stratigraphic baffle.The wedge and apron reservoirs both contain a late burial matrix diagenetic overprint represented mainly by co-precipitated bitumen and calcite cement and local development of matrix microporosity. Enlargement of the early fractures in the wedge reservoir also occurred during burial diagenesis based on the presence of diagenetic halos containing the burial overprint around the fractures and based on the presence of co-precipitated bitumen and calcite in the fractures. Scenarios and mechanisms for fracture enlargement are evaluated against the observations from field data and the outcrop analogs.

Abstract Improved and enhanced oil recovery (IOR/EOR) processes, such gas injection and waterflooding, are difficult to model accurately in fractured reservoirs due to complicated multiphase physics and the strong property contrast between permeable natural fractures and tight matrix rock. Traditionally, fractured reservoirs are modeled using coarse dual-porosity (DP) simulation models that idealize the fracture system. Consequently, important recovery physics can be missed, causing misleading flow predictions. We present an unstructured discrete-fracture-matrix (DFM) approach for three-phase, fully compositional simulations of IOR/EOR processes. We generate detailed DFM grids that represent several thousand realistic discrete fractures and different well configurations. These grids are then discretized for IOR/EOR flow simulations. We apply the flow-based multiple subregion (MSR) technique with two subregions to generate unstructured coarse-scale fracture models that are directly based on the DFM sector models. We next employ MSR upscaling to create corresponding structured DFM-based dual-porosity, dual-permeability (DPDK) models, computing parameters like shape factors and effective fracture permeabilities. We lastly generate conventional DPDK models by invoking simplifying assumptions and analytical equations. Using DFM results as reference solutions, we find the following order of decreasing accuracy: MSR, DFM-based DPDK, and conventional DPDK models. The DPDK models result in significant inaccuracies at individual wells when important fluid movement (e.g., gas breakthrough) is missed. The errors at sector scale are lower due to cancellation of errors at wells. The simulation results indicate that oil recovery and injected fluid breakthrough behavior are a complicated function of the reservoir heterogeneity, well configurations, and phase behavior. Large modeling errors can potentially lead to the selection of a sub-optimal injectant. One to two orders of computational speedup are achieved from the upscaling. These findings point to the need for systematic upscaling of realistic DFM representations to generate fast and accurate coarse fracture models for the assessment of field-development IOR/EOR schemes.

In naturally fractured reservoirs, determining fracture properties, such as size and permeability, is difficult due to limited data about the fractures. The primary information that is available, mainly from image logs or core, is known only at the wellbore; however, while fractures can be local to the wellbore region, they often extend hundreds of feet from the well. If fractures can be characterized more accurately, then flow paths and flow behaviors in the reservoir can be better delineated. The goal of this work is to combine dynamic data with static data to better characterize fracture properties, in particular, the size of fractures away from the wellbore.Production log (PLT) information is used in conjunction with borehole image logs (BHI) to more accurately define fracture properties. A single-well flow-simulation model is constructed, and discrete fractures are included at each location where wellbore inflow (observed on PLT logs) is correlated with fractures (observed on BHI logs). An optimization method is developed to automatically perturb the lengths and heights of model fractures until their flow response matches the actual measured response. This method is designed for wells that intersect multiple fractures.The technique is applied to two different examples where fracture sizes are estimated. Fracture lengths range from a few feet to thousands of feet. To provide parameter sensitivity for the fracture sizes, the examples are matched with multiple models. Different options are used for the different models to demonstrate the various capabilities of the program. The optimization is efficient as it generally finds solutions in about 10-20 flow simulations. Knowledge of fracture size can be used to estimate potential for fracture interconnectivity among wells in a field, reservoir drainage patterns, and implications of injection and production patterns, all of which can impact reservoir management decisions. Furthermore, fracture size distributions can be used as an input for flow-simulation modeling, thus providing an improved physical basis for modeling overall reservoir performance. (C) 2012 Elsevier B.V. All rights reserved.

Fractures within steep rimmed carbonate platform reservoirs can have significant impact on hydrocarbon migration, storage and permeability of the reservoir. This paper focuses on the syn-sedimentary fractures, which form contemporaneous with the development of the platform. We present 2D geomechanical models to investigate the effect of compaction-induced differential subsidence on the development of syn-sedimentary fractures, and attempt to predict their occurrence and distributions. A significant effort of the mechanical modeling is the development of proper constitutive laws for different carbonate facies of steep rimmed carbonate platforms as their rock types strengthen through time and burial. We calibrate the outcome of our models with field observations from the Devonian carbonate reef of the Canning Basin, W. Australia, and the Permian Capitan reef of the Guadalupe Mountains, NM, USA, both of which are field analogues for major carbonate platform reservoirs like Tengiz and Karachaganak in Kazakhstan.

Abstract Two discrete-fracture models (DFMs) based on different, independent numerical techniques have been developed for studying the behavior of naturally fractured reservoirs. One model is based on unstructured gridding with local refinement near fractures, while in the second model fractures are embedded in a structured matrix grid. Both models capture the complexity of a typical fractured reservoir better than conventional dual-permeability models, leading to a more accurate representation of fractured reservoirs. The accuracy of the DFM approaches is confirmed by their match with a structured, grid-aligned, explicit-fracture model in tests involving capillary imbibition during water flooding and gravity drainage in oil-gas systems. The DFMs are insensitive to grid orientation. Simulations also show consistency and agreement of results of the DFM methods in synthetic models with complex fracture patterns. Our simulations indicate that conventional dual-permeability approaches are appropriate when the fracture system is very sparse relative to the grid spacing. In these situations a DFM can be used as the basis for defining dual-permeability model parameters. However, conventional dual-permeability approaches are inadequate in the presence of high localized anisotropy and preferential channeling. When used with general purpose reservoir simulators, both DFMs show computational performance that is comparable to that of dual-permeability models.

AbstractTengiz Field is the world's deepest developed supergiant oil field, with an oil column of nearly 1600m. The reservoir consists of Devonian and Carboniferous platform/slope carbonates, divided into three stratigraphically-defined units. Production of over 500,000 BOPD is mainly from the upper unit. A significant portion of this production is controlled by natural fractures. A new reservoir model has been constructed to support a future growth project, including miscible gas injection, and to guide reservoir management strategies, development planning, and oil-in-place estimation.Our new model represents a significant change from previous models in that a dual porosity, dual permeability flow formulation (fracture and matrix) is being applied. This change to dual porosity, which is necessary to effectively characterize the fracture-matrix flow, requires a significant modification of the overall reservoir modeling workflow. Matrix porosity and permeability are distributed into the fine-scale model and fracture properties are populated into the model after upscaling, because representative fracture attributes are only available at a large scale.Fracture porosity is high in the depositional slope (about 0.4%) due to the cavernous character of portions of the fracture system. This fracture porosity represents a significant volume, and has a large impact on reserves. Fracture permeability is also a critical factor, and has a large affect on flow behavior and recovery. Although fractures are important, their intersection by wellbores is relatively rare, due mainly to under-sampling of sub-vertical fractures by vertical wells and sparse well control in the slope. Because of this under-sampling, and general ambiguity in image-log data, uncertainty associated with fracture properties is large. To address this uncertainty, significant effort has gone into collecting reservoir surveillance data. In addition, fracture properties from image logs were reconciled with well tests and other logs.

This paper describes grid generation algorithms for simulating flow and transport in fractured porous media. The methods are designed to only capture details of the fracture network geometry larger than the specified grid resolution. The final grids honor fractures approximately while maintaining good cell quality. Improved representations of the input geometry can be obtained by generating a new grid with finer resolution. Several numerical examples are presented in two and three dimensions that demonstrate that the algorithms are robust and practical for industrial applications.

This paper describes grid generation algorithms for simulating flow and transport in fractured porous media. The methods are designed to only capture details of the fracture network geometry larger than the specified grid resolution. The final grids honor fractures approximately while maintaining good cell quality. Improved representations of the input geometry can be obtained by generating a new grid with finer resolution. Several numerical examples are presented in two and three dimensions that demonstrate that the algorithms are robust and practical for industrial applications.

Abstract In naturally fractured reservoirs, determining fracture properties, such as size and permeability, is difficult due to the limited data about the fractures. The primary information that is available, mainly from image logs or core, is known only at the wellbore; however, the fractures can be local to the wellbore region, or they can extend hundreds of feet from the well. If fractures can be characterized more accurately, then flow paths and flow behaviors in the reservoir can be better delineated. Thus, rates and declines in primary production and breakthrough times of injected fluids can be better estimated, and the field can be developed more efficiently. The goal of our work has been to combine dynamic data with static data to better characterize fracture properties, especially the size of fractures away from the wellbore. PLT information can be used in conjunction with static data to more accurately define fracture properties such as size and permeability. A single-well single-phase flow-simulation model is constructed with discrete fractures at each location where wellbore inflow is observed with a PLT. A program has been created to automatically perturb the lengths and heights of model fractures until their flow response matches the actual measured response (i.e. history matching PLT data by changing fracture properties). This program is designed for wells that intersect multiple fractures. The tool has been used on a number of different examples, and matches are achieved in all cases. Fracture lengths range from a few feet to thousands of feet. The optimization is efficient as it generally finds solutions after iterating through only about 20 flow simulations. To demonstrate the various capabilities of the program, different options are used in different cases. One example is matched with six different models to provide a range of uncertainty for the fracture sizes.

Summary The extent to which fractures affect fluid pathways is a vital component of understanding and modeling fluid flow in any reservoir. We examined the Wafra Ratawi grainstone for which production extending for 50 years, including recent horizontal drilling, has provided some clues about fractures, but their exact locations, intensity, and overall effect have been elusive. In this study, we find that a limited number of total fractures affect production characteristics of the Ratawi reservoir. Although fractures occur throughout the Wafra field, fracture-influenced reservoir behavior is confined to the periphery of the field where the matrix permeability is low. This work suggests that for the largest part of the field, explicit fractures are not necessary in the next-generation Earth and flow-simulation models. The geologic fracture assessment included seismic fault mapping and fracture interpretation of image logs and cores. Fracture trends are in the northeast and southwest quadrants, and fractures are mineralized toward the south and west of the field. Pressure-falloff tests on some peripheral injectors indicate partial barriers, and most of these wells lie on seismic-scale faults in the reservoir, suggesting partial sealing. A few wells show fractured-reservoir production characteristics, and rate-transient analysis on a few producers indicates localized dual-porosity behavior. Producers proximal to dual-porosity wells display single-porosity behavior, however, to attest to the notion of localized fracture response. The spatially restricted fracture-flow characteristics appear to correlate with fracture or vug zones in a low-permeability reservoir. Presence of fracture-flow behavior was tested by constructing the so-called flow-capacity index (FCI), the ratio of khwell (well test-derived value) to khmatrix (core-derived property). Data from 80 wells showed khmatrix to be consistently higher than khwell, a relationship that suggests insignificant fracture production in these wells.

The behavior of naturally fractured reservoirs (NFRs) is typically difficult to predict due to extreme heterogeneity of their flow-related character. This unpredictability affects most aspects of field management; surprises can accompany each well. From our well-based scale of observation, we sample and test reservoir fracture systems below their representative elemental volume (REV). This is the ultimate cause of the unpredictability of their flow behavior. Heterogeneity is caused by the complex and variable geometry and connectivity of the plumbing system of the reservoir (well, fractures, and matrix). Connectivity is a function of: a) Orientation of a well relative to reservoir fracture sets, b) Spacing of fractures in each set, c) Spatial variability of the fracture system, d) Effective length of fractures, e) Fracture height, and f) Fracture-matrix interaction (effective porosity, transfer).