Fractured Reservoir Discrete Feature Network Technologies

FRDFNT Home

Project Overview

Research Team

Scope of the Project

Technology Transfer Database

Site Map

Search Site

In this task, Golder Associates will develop a new generation of geological and advanced mathematical discrete fracture data analysis technologies. These technologies will be based upon Golder’s FracMan computer software

Task 2.1.1: Fracture Sets Analysis

In this task, we will develop and demonstrate a neural net technology for identifying conductive fracture sets from typical fractured reservoir data.

Task 2.1.2: Spatial Location Analysis

In this task, quantitative technologies for evaluation of the spatial pattern of discrete feature heterogeneity will be extended to include rule-based evaluation of fracture patterns due to tectonic processes

Task 2.1.3: Hydraulic Parameter Analysis:

The existing “OxFilet” approach within the FracMan package provides a method for the derivation of conductive fracture intensity and transmissivity distributions from hydrologic packer tests (Dershowitz, 1992). In this task, the “OxFilet” approach will be extended to allow direct derivation of conductive fracture frequency and intensity from oil field DST results and production logs. The analysis will take transient DST results as input and will derive conductive fracture frequency simultaneously for multiple fracture sets and for the rock matrix. It will use curve fitting of fractional dimension well response to derive fracture hydrologic properties. This will incorporate a more sophisticated rough fracture/fracture generation flow simulation combined with simulated well test interpretation.

Task 2.1.4: Compartmentalization Analysis

In this task, we will develop compartmentalization technologies appropriate for fractured reservoirs. We will use these technologies to analyze the vertical fault population in the relevant portions of the Yates Field to determine the connective geometry of these faults, the real extent of fracture networks, and the matrix block size distribution formed by the fractures using the approaches developed in Tasks 1.3.1 and 1.3.2

Discussion:

Since 1985, Golder Associates has developed practical discrete fracture data analysis, integration, and reservoir modeling technologies. In this task, Golder Associates will extend these technologies for direct application to analysis of fracture orientation, size, spatial structure, and hydraulic properties. The extensions will be based upon Golder’s FracMan computer software (Figs. 2-1, 2-2, 2-3). FracMan provides an extensive platform for analysis of discrete fracture orientation, size, spatial structure, and flow geometry (“dimension”) data, along with the calculation of fracture connectivity, matrix block dimensions, and fluid flow geometry. In this task, a new generation of geological and advanced mathematical discrete fracture data analysis technologies will be developed, and demonstrated using fracture data from the Yates field. Specific features to be developed and tested include the following:

Fracture Sets: Most methods for characterizing fracture sets rely exclusively on cluster analysis of fracture orientations. This method often fails to identify sets or subpopulations that are meaningful from a fluid-flow perspective. Recently, Thomas and La Pointe (1995) have demonstrated a neural net approach for identifying significant conductive fractures incorporating a host of geological information. This process is essential for identifying and characterizing the conductive fractures in an oil field, since only the conductive fractures can be considered in most practical scale problems. Golder Associates will develop and demonstrate a neural net technology for identifying conductive fracture sets from typical fractured reservoir data.

Spatial Location: Another key to understanding the heterogeneity of fractured rock reservoirs is the spatial pattern of fracturing. Golder Associates has developed statistical procedures for evaluation of fracture lineaments as seen on outcrop maps, borehole TV (BHTV) and formation microscanner (FMS) logs (e.g. La Pointe et al., 1993). These procedures include fractal methods based on Levy-flight, box-counting, and spectral dimension; geostatistical methods for spherical, exponential, and powerlaw variograms; as well as a number of additional procedures based on statistical methods and empirical heterogeneity measures. As part of this research project, these quantitative measures of heterogeneity will be extended to include rule-based evaluation of fracture patterns due to tectonic processes.

Fracture Fluid Flow Parameters: Drill stem hydraulic tests (DST) are conventionally used to derive continuum hydraulic properties of permeability and specific storage. However, for fractured rock formations, these values must be transformed to transmissivity distributions for fracture populations, and rock matrix permeabilites. Existing methods for the derivation of conductive fracture intensity and transmissivity distributions from hydrologic packer tests (Dershowitz, 1992) assume steady-state, fracture-only flow. As part of this research project, this approach will be extended to allow direct derivation of conductive fracture frequency and intensity from oilfield DST results and production logs. The analysis will take transient DST results as input and will derive conductive fracture frequency simultaneously for multiple fracture sets and for the rock matrix. It will use curve fitting of fractional dimension well response to derive fracture hydrologic properties. This will depend on use of a more sophisticated rough fracture/fracture generation flow simulation combined with simulated well test interpretation.

Compartmentalization Analysis: Reservoir compartmentalization may arise from a variety of factors, including structural offset of permeable reservoir units, mienralization of faults and fractures, development of a fracture system that tends to form isolated networks, or highly permeable vertical faults which create barriers for lateral movement of steam or fluids to or from wells. At the Yates Field, compartmentalization has been demonstrated. The compartmentalization is probably due to highly permeable vertical faults, and perhaps to isolated fracture network development.

All of the above analytical tools will make it possible to analyze a fractured reservoir, such as Yates Field, to extract the salient data for discrete fracture flow models and make it possible to build realistic models.