Second EAGE Workshop on Naturally Fractured Reservoirs

Gas-oil gravity drainage (GOGD) is perhaps the most effective recovery mechanism for those naturally fractured reservoirs that are lithologically stratified, the oil is stored in rock matrix, and vertical fractures cut through (multiple) low-permeability layers, providing the necessary permeability in the direction of gravity-driven flow (eg., Gilman 2003, 2011). The conceptual process model for GOGD is that fracture saturation with gas that is much less dens than the oil creates a pressure differential that drains the oil into the fractures, where it flows downward and can be produced from wells placed directly above the oil water contact. Prerequisites are that saturated conditions can be obtained, the fractures are sufficiently tall, and the oil is mobile enough (e.g., Kazemi et al., 1976). Since Gilman and Kazemi (1983) and Saidi (1983) GOGD has predominantly been simulated using the oil industry’s adaptation of the dual continuum (DC) approach (e.g., Warren and Root, 1963),using separate fracture and matrix grids coupled via source terms for fracture-matrix transfer (FMT),see Abushaika and Gosselin (2009). However, it is challenging to derive DC input parameters such as matrix-block size, shape factor or fracture permeability from field data because they tend to obey power-law or multi-modal distributions (e.g., Bonnet et al., 2001). Recently, new numerical fracturematrix upscaling methods have been proposed to overcome this DC model parameterization problem (Sarda et al., 1997, Karimi-Fard et al., 2006), however, they only address aspects of pseudo steadystate FMT and entail other restrictive assumptions. In many cases, history matching of DC models is only possible through over-calibration, raising doubts about the predictive capabilities of such models (e.g., Gilman, 2003). Since fractures have no discrete representations in DC models, these cannot be used to test long-standing hypotheses about the details of FMT (cross flow, capillary continuity, reimbibition of oil, see Fung, 1991; Saidi, 1993), or evaluate the balance between gravitational – viscous and capillary forces and the mutual influence of respective flow processes within the reservoir. Assumptions about how saturation changes in matrix blocks and the pathways that the oil takes on its way to the production wells cannot be tested. Discrete-Fracture and Matrix models (DFM) provide an alternative for such purposes (e.g., Kim and Deo 2000). Using a spatially adaptive poly-element type unstructured grid, complex fracture geometries can be represented in the model explicitly, permitting the modelling of realistic fracture geometries and property statistics (e.g., Matthai et al., 2007). In 2011, Bazrafkan and Matthai,presented a new hybrid FEM node-centered FVM methods where finite elements also serve as control volumes, permitting the correct representation of saturation jump discontinuities at material interfaces and also the realistic simulation of capillary FMT. Here, we apply this novel DFM approach to simulate GOGD using an outcrop analog model that has a fracture geometry that is simple and regular enough to construct an equivalent dual permeability model. The latter is run using a commercial simulator and here we present, for the first time, a comparison of DFM and DCM results, using the same saturation functions and parameters. The DFM results further allow an investigation of the flow patterns in individual fractures and matrix blocks so that the validity of the assumptions that underpin the DC approach can be evaluated.

A large amount of world’s hydrocarbon reserves lies in reservoirs where one of the key features is the presence of a system of fractures spanning different length scales. Fractures may provide ways to drain the matrix, but they also drive gas and water towards wells. Fractures intensity, geometry and conductivity are characterised in conjunction with matrix properties using all possible data, such as production logging, mud-losses, image log, well test, geophysics and outcrops. This may lead to a geometrical characterisation in terms of a discrete fracture network (DFN), with the definition of a set of geometrical 2D objects embedded in the matrix domain. A DFN is a computational challenge for conventional dual-porosity simulators, where a dual medium formulation is cast in a corner point geometry grid (CPG). In this context small, with respect to CPG grid spacing, fractures can be easily incorporated using some practical recipes, see Dershowitz et al. (2000), in single porosity or in dual-porosity models. On the other hand, long range, interconnected fractures cannot be integrated in the simulation model unless some crude approximation is implemented in order to fit the fracture network geometry into the Warren and Root (1963) dual media framework, at least as it is available in conventional commercial simulators, see e.g, Eclipse (2011) . A more accurate solution is the implementation of unstructured gridding where the matrix is discretized by tetrahedral or more general polyhedra and the fractures are discretized using 2D polygons. These grids can be used with finite volume, connectivity based reservoir simulators (see Karimi-fard et al. (2004)), but the approach is computationally inefficient for most of the commercial simulators, and this motivated the development of the EDFM by Li and Lee (2008). In this approach the intersections between fractures and matrix blocks define the degrees of freedom (DOFs) for the high connectivity medium. Then, 2D flow between fracture DOFs and 1D flow between fractures and matrix can be integrated with 3D flow in the CPG grid representing the matrix. In a nut-shell, an unstructured grid for the fracture network is comined with a structured grid for the matrix. Differently from Li and Lee (2008), our implementation is not based on the customisation of the reservoir simulator. Rather, we exploit the capability of most commercial simulators to define non neighbouring connections across grid cells to implement EDFM in a non-invasive manner. Our results confirm that EDFM can be as effective as fully unstructured gridding but much more computationally efficient.

When developing a fractured reservoir, it is beneficial to learn from analogue fields about the fracture parameters which significantly impact fluid flow and hydrocarbon recovery under various recovery schemes. The lack of an industry standard workflow for the modelling of naturally fractured hydrocarbon reservoirs and the numerous ways in which naturally fractured rock volumes can be parameterized for input to modelling precludes a robust comparison of results from published studies. This paper presents a retrospective analysis of the results of a number of fracture modelling studies sharing a common workflow for parameterization, modelling and uncertainty handling. The objective is simple: to better understand which parameters of a natural fracture system are of key importance in estimating recovery for different development options and how parameter sensitivity varies over asset lifecycle.

The geomodel of a fractured basement reservoir model is realized in order to integrate all available data, particularly 3D seismic cubes and attributes, and flowmeter logs. Considering the geological context of this field, a particular attention has been given to capture its flow behaviour. To do so, it has been necessary to use a dual media approach, thus to characterize and model the matrix and the fracture network. We intend to describe the implemented approach. Starting with the description of the generated Discrete Fracture Network DFN calibrated towards dynamic data which allows reducing the uncertainty on key production characteristics of the field and production forecasts for the field development. Focus will be set on the synergies arising while combining leading edge fracture modelling techniques and uncertainties management through history matching.

A case study of a stacked, multi reservoir carbonate oil field in Central Onshore Oman is presented here to highlight the need for ongoing appraisal and integrated field development studies even after 25 years of production. Subsurface parameters such as abundance of high-k layers, and connectivity of faults, fault related fractures, distributed back ground fractures, and layer bound fractures are the key geological uncertainties that impact the selection of the field development options. Data integration is the key to a better understanding of such a complex fractured field. The proper and timely integration of reprocessed seismic (fault imaging, attribute interpretation), interference testing (cooperation with asset and field operations team) and iterations with dynamic reservoir simulation resulted in further improved history matches and reducing uncertainties, which enabled taking a development concept decision. This presentation is demonstrating the value of applying a fully integrated decision-driven static and experimental design dynamic modeling workflow to progress the ongoing FDP study.

Multiscale fracture networks frequently affect carbonate reservoirs and play a crucial role in controlling permeability. Network geometries are here derived from the Latemar platform outcropping analog (Dolomites, N Italy) integrating satellite images, LiDAR airborne enriched Digital Elevation Models and outcrops. Our observations cover lengths from kms to cm. Two sets of fractures are recognised in the Latemar, both formed by mode I features and displaying traces insensitive to facies boundaries. The older set is visible in all length classes. Fractures heights and spacing change from one platform domain to the other. Fractures of the second set are generally confined between those of the first set and are, therefore, <50-80m long. Their heights and spacing distances are larger than those of the first, older fracture set, as a result of increased lithification between the two episodes of deformation. The existence of two fracture sets with significantly different geometries impacts platform-scale permeability patterns.

in Carbonate reservoirs. These two natural phenomena can be acting as positive reservoir contributors, in terms of reservoir efficiency and production, and also can be acting as negative contributors. In the last years, the carbonate reservoir remains the most challengeable reservoirs in exploration and developments where their reserves estimation and production amounts remains undefined precisely. This is directly related to the nature of the Carbonate reservoirs depositional processes formation syn-depositional and post-depositional. Carbonate reservoirs account for 40% of today’s hydrocarbon production, and because of several giant fields in the Middle East they are expected to dominate production through the next century. In Libya, carbonate reservoirs are mainly concentrated in Sirte Basin and offshore where the first exploration in 50’s where in reef carbonate reservoir. Therefore, understanding carbonate reservoirs and producing them efficiently have become industry priorities and are likely to remain so.

Fracture data from outcropping analogs are often acquired but rarely used to improve permeability predictions in buried reservoirs. To improve this we present a systematic workflow which includes i) the building of a geometric model of the investigated structure, ii) the execution of mechanical simulations to determine state of stress and strain field, iii) the use of outcrop data to populate the model and iv) develop a reservoir-scale permeability model To make full use of outcrop data, we present new acquisition and processing tools allowing able to provide a full characterization of the fracture field even when fractures are not bed-confined. A few case studies are presented to discuss this workflow in sedimentologically and structurally heterogeneous settings.

The studied block is the largest and most faulted within the carbonate field which is currently under water-flood FDP. The study kicked off with extensive borehole image interpretation review of existing and new fracture picks. In parallel, several high resolution seismic inversions and spectral imaging attributes were generated as drivers to geological and fracture modelling. High resolution seismic was used to highlight subtle faults. Facies changes were also visible from seismic as seen in cored wells. Sequential geological modelling of GR, density, porosity and SW was constrained by seismic attributes. The derived fracture frequency logs were compared against geological, structural and seismic drivers in a process called driver ranking. The results confirmed the role of faults as well as facies being primary causes of fracturing. Subsequently, the screened and cross-correlated potential drivers were carried forward to constrain the fracture models. Multiple stochastic realizations were derived through NNT training and testing and an average model was kept. Final models were validated against hidden BHI data. New BHI was used to confirm model prediction. Different types of dynamic data in non-BHI wells were also used to validate the fracture models as specific production/injection related issues could be directly linked to presence of fractures. These data include PLT, PTA and tracer tests from which injectivity issues and short circuiting were explained by higher fracture densities and corridors. Through dynamic calibration, the fracture model was converted to fracture permeability. The fracture permeability is the product of fracture density and a scaling factor derived from history matching. Subsequently, the addition of matrix permeability and fracture permeability will determine the effective permeability. This Keffective was directly used in Daleel’s simulator without upscaling since it was part of the same grid hosting the fracture models. The results were encouraging as the simulation was smooth and error-free.