EAGE Workshop on Naturally and Hydraulically Induced Fractured Reservoirs – From NanoDarcies to Darcies

The objective is to assist seismic interpreters to efficiently map complex faulted areas with good details, for future fault and fracture modelling purposes. Accurate seismic dips are very important in this context, primarily because dips are the mandatory input to the curvature attribute, which is a generally accepted fault/fracture indicator. Therefore, accurate dip estimation becomes extremely important if we want to apply automated techniques for fault extraction. We will here present preliminary results of the application of a novel algorithm for the estimation of dips, which employs a set of global dip consistency constraints. The examples are from a carbonate reservoir in the Danish part of the North Sea.

A key element in performing an effective flow simulation in a fractured reservoir is construction of a well-constrained Static Conceptual Fracture Model for the Reservoir. The Static Fracture Model Along with the Dynamic Fracture (fluid distribution fracture flow properties) and Geomechanical Model for the reservoir are merged to create the final gridded simulation grid.

Integrating dynamic data into reservoir numerical models is essential for capturing the actual dynamic flow behaviour and the true performance of a reservoir. History matching efforts could prove to be fruitless without including such data into the reservoir model. Fractures and super-high permeability streaks are two examples of important flow features that open-hole logs and core sampling do not provide a satisfactory characterization and can be overlooked. Thus a log and core-based models will not be able to predict the actual performance of the field. It is not uncommon to measure dynamic permeability values, through pressure transient testing, at magnitude(s) of order higher than those obtained from core samples and core-log correlations, indicating presence of un-interpreted high permeability intervals within the reservoir. This integration of dynamic data, and the identification of reservoir fractures, in their different forms, poses a challenge for the simulation engineer during the history matching process. Variable impact of the different types of fractures (due to their varying size, shape, proximity and continuity) on the flow performance and water cut development forms an additional difficulty during. The presentation discusses some of these challenges associated with modeling of a giant carbonate Middle East reservoir, aiming as starting a brainstorming discussion of what options and alternative approaches are there for incorporating the dynamic permeability into the simulation model, and for identification of fractures presence, type, size and continuity in the reservoir.

A newly developed technology analyses the readily available fluid production histories from existing wells to develop statistical correlations in rate fluctuations. Results of the application of this novel technology to naturally fractured fields in the North Sea will be presented. The results reveal surprising characteristics: many of the correlated well pairs are very long-range; they also appear to be stress-related and fault-related. The postulated mechanism is that faults and fractures are reactivated due to the stress perturbations brought about during field development. In one field, comparison with independent data, particularly microseismic recordings, provides very encouraging calibration. Identification of major reservoir pathways is of substantial advantage to efficiency in reservoir management, leading to benefit for practical issues of well placements and configurations, injectivities, productivities, sweep efficiencies, short-term and longer-term forecasting. The means of integrating the technology with other reservoir management processes will be outlined. In particular the techniques can be used in a time-lapse fashion in order to monitor changes in reservoir behaviour and provide real-time updating of reservoir models.

Fracture models, including connected fracture networks, fracture corridors, unconnected fractures and nonconductive fractures, are presented, using different dual-permeability models that have been developed. To simulate the dynamic response of fractured reservoirs, the opening of nonconductive fractures is simulated based on either fracture dilation due to shearing or fracture opening due to tensile failure. Once the opening of nonconductive fractures occurs, the permeability of all fractures (both conductive and nonconductive) changes dynamically according to the changes in the effective stresses that occur due to reservoir depletion and/or injection. Dual-permeability reservoir models developed on the basis of the proposed fracture models improve reservoir simulations. The newly developed dual-permeability modeling was applied to produced-water reinjection scenarios in a field in Southeast Asia and realistically simulated the dynamic behavior of the reservoir during injection. Such modeling reveals the ongoing interaction among pressure, rate, permeability, and deformation in the reservoir. This approach assists in the design of an optimal water-injection schedule to maximize production with a reduced risk of cap rock damage and water leakage into overburden.

The main causes of azimuthal anisotropy in hydrocarbon reservoirs are anisotropy of the horizontal stresses and presence of vertical or dipping fractures. The aim of this paper is to analyse and compare anisotropy patterns associated with these two causes of anisotropy. Stresses affect elastic properties of rocks due to presence of discontinuities such as cracks and compliant grain contacts.

Fracture distribution in a naturally fractured reservoir is related to many different parameters. These parameters include fracture drivers such as tectonic phenomenon, fracture controllers such as lithology and porosity and fracture indicators such as mudloss and production data. Each of these parameters is related to fracture distribution in a specific point of view. This study is focused on Asmari Formation in Gachsaran oil field is which is the most famous naturally fractured reservoir in Iran. The main goal of this work is comparison and integration of all of the fracture signs and controller parameters to generate a comprehensive fracture intensity model. 3D strain is calculated and integrated in fracture intensity modeling as a tectonic fracture driver. Distance to fault was also contributed as a second driving factor to the modeling process. On the other hand, the mudloss was normalized and converted to fracture permeability and its 3D model was generated to be added in the intensity modeling. Productivity Index map was another indicator that was involved in modeling. The result fracture intensity model includes the effect of all of the fracture controllers, drivers and indicators and this is the main advantage of integration approach for modeling.

This Middle East case study addresses the modelling of a Type 2 (Nelson, 2001) green carbonate reservoir which production is acknowledged to be largely controlled by fractures. In the case of an undeveloped reservoir the main challenge is posed by the limited amount of hard data and production history. The ultimate objective was to minimize production forecast uncertainties. This was achieved by means of a customized workflow that integrates 3D seismic, well and production data to predict fracture distribution and flow control in the reservoir. The calibration against dynamic data indicated overestimation of the original values of fracture properties. Changes were applied thorough an iterative workflow that aimed to preserve both original distribution and contrast in the seismically constrained fracture model.

This paper aims to present the fracture permeability aspect of an integrated workflow whose goal was to predict reservoir permeability in a carbonate brownfield using available dynamic data. The dynamic data was used to estimate a permeability height product (kh) which was tested in dynamic simulations, where both pressure match and production index match drove the selection of a suitable fracture permeability distribution. Well tests indicated reservoir permeability >> matrix in a significant number of tests. Fractures are assumed responsible for the excess permeability after a basic comparison of the Fracture Production Index (Reiss, 1980) where khwelltest/khmatrix values indicated a fracture signature in some wells. Our analysis of the data led us to interepret that the production data could be reasonably matched with a model where 33% of the reservoir had an excess fracture permeability of 372 mD +/- a standard deviation of 626 mD.

This presentation addresses the relationship between fracture network modeling and well test interpretation. Well test provide information on the geometry and properties of fracture networks. Network models can reproduce behaviors similar to conventional well test models, but with different geometries that can have significant impacts on reservoir behavior. Well test interpretations are non-unique, but nonetheless constrained by geometric elements. Further constraints on the well test interpretation must come from geologic and geophysical information, hence well test analysis should be integrated with other disciplines.