Research paperAn integrated workflow for stress and flow modelling using outcrop-derived discrete fracture networks
Graphical abstract
Introduction
Natural fracture networks are multiscale systems that develop through a combination of mechanisms that are only partly understood (Olson et al., 2009, Philip et al., 2005). Understanding the multiscale distribution of fracture networks in the subsurface allows for optimisation of fractured reservoir development (Nelson, 2001). However, limited observations from seismic and wells do not provide the complete fracture network geometry and associated flow properties, particularly of the subseismic fracture network (Fabuel-Perez et al., 2010, Martinez-Landa et al., 2016). Outcrops are the only source that provides realistic descriptions of fracture networks, as no models exist that can create realistic fracture networks on the scale of fractured reservoirs. To derive lessons for fractured reservoirs, we need outcropping datasets that contain at least several hundred fractures covering several orders of magnitude in spacing and length, without suffering from censoring and truncation artefacts, over an area that covers at least several grid blocks in conventional reservoir flow models (Bonnet et al., 2001, Ortega et al., 2006). Such datasets are not easily obtained from conventional outcrop photographs or scanline studies, as these methods capture only a limited number of scales (Bisdom et al., 2014).
Photogrammetry, particularly Structure from Motion (SfM) Multiview stereo (MVS), is an inexpensive and fast method to accurately map 3-D structures from 2-D images taken from different positions (Harwin and Lucieer, 2012, Ullman, 1979). In recent years, this method has been embraced by geologists to create digital outcrop models as an alternative to the more expensive and less flexible LiDAR (Light Image Detection And Ranging) methods (Hodgetts, 2013, Mahmud et al., 2015, Reif et al., 2011, Rotevatn et al., 2009, Tavani et al., 2014, Wilson et al., 2011). Partly overlapping images are aligned by identifying and extracting common points, which can be positioned in 3-D space to reconstruct the outcrop geometry (Bemis et al., 2014, James and Robson, 2012). The resulting models provide a complete and unobstructed viewpoint of the outcrop that can be changed and adjusted for any purpose (Tavani et al., 2016).
As this approach requires that the outcrop is fully covered by images with an overlap of at least 50%, Unmanned Aerial Vehicles (UAVs or drones), equipped with a camera and positioning sensors, are best suited to acquire the images required for photogrammetry modelling (e.g. Bemis et al., 2014; Bond et al., 2015; Hodgetts, 2013; James and Robson, 2012; Tavani et al., 2014; Vasuki et al., 2014; Vollgger and Cruden, 2016). Fracture geometries can be extracted from the resulting georeferenced models in 2-D or 3-D (Duelis Viana et al., 2016, Hardebol and Bertotti, 2013, Tavani et al., 2014). Extraction of 2-D data from a 3-D photogrammetry model is more accurate than fracture interpretation from conventional 2-D images, as the photogrammetry model is accurately orthorectified and the multiple viewpoints allow for more precise digitisation of fracture geometry. Irrespective of whether the fracture data is used for 2-D or 3-D analysis, 3-D outcrop models provide a higher accuracy.
The second challenge is to obtain realistic aperture predictions from outcropping geometries. At depth, permeability is a function of aperture, which is partly controlled by the in-situ stresses (Baghbanan and Jing, 2008, Lei et al., 2015, Tao et al., 2009, Zoback, 2007), but pressure relief during exhumation and weathering dissolves cements and changes aperture. Outcropping apertures are therefore not representative, unless it can be proven that fractures have not been reactivated during exhumation. This is typically assumed to be the case for veins (e.g. Hooker et al., 2014), but preserved veins are relatively rare. Alternatively, aperture is modelled as a function of stress, using subcritical crack growth as defined by Linear Elastic Fracture Mechanics (LEFM) or conductive shearing defined by Barton-Bandis (Barton, 1982, Barton et al., 1985, Barton and Bandis, 1980, Lawn and Wilshaw, 1975, Olson, 2003, Pollard and Segall, 1987, Vermilye and Scholz, 1995). These models require the local stress state, which is typically derived from Finite Element (FE) models with explicit fracture representations (Barton, 2014, Bisdom et al., 2016b, Lei et al., 2014, Lei et al., 2016, Nick et al., 2011).
The third challenge is modelling permeability through fractured rocks, taking into account the coupled flow through fractures and matrix (Belayneh et al., 2009, Geiger et al., 2013, Lang et al., 2014). Conventional reservoir simulation tools scale up fracture density, porosity and permeability to effective grid properties in dual-porosity dual-permeability grids, resulting in a significantly simplified flow model (Cottereau et al., 2010, Geiger and Matthäi, 2012, Jonoud and Jackson, 2008). Methods exist to model flow through discrete fracture-matrix models without requiring upscaling, making use of a Finite-Element Finite-Volume (FE-FV) approach, but the use of these methods is often limited to relatively small-scale synthetic fracture networks (Lei et al., 2014, Matthäi and Belayneh, 2004).
These individual problems have been studied extensively, focusing on 3-D outcrop modelling (Hodgetts, 2013, Tavani et al., 2014, Vasuki et al., 2014), meshing (Karimi-Fard and Durlofsky, 2016, Nejati et al., 2016, Nick and Matthäi, 2011a, Paluszny et al., 2007) and flow modelling (Lang et al., 2014, Nick and Matthäi, 2011b), but integrating these components remains a challenge. Our aim is to present an integrated workflow for modelling the complete permeability tensor of large-scale fracture networks with apertures representative of in-situ stress conditions by combining fast data acquisition using a UAV with outcrop modelling using photogrammetry (Fig. 1). This workflow builds upon the stress-aperture modelling approach presented in Bisdom et al. (2016b), making use of the geometrical aperture approximation from Bisdom et al. (2016d), and the modelling of permeability for a range of aperture definitions presented in Bisdom et al. (2016d). The 3-D outcrop models are used to accurately digitise fracture patterns in 2-D, which form the basis for stress, aperture and equivalent permeability (i.e. combined matrix and fracture permeability) models. The main result is a discrete fracture-matrix model consisting of an unstructured mesh with discrete fractures, from which the full permeability tensor is calculated. The aim of this workflow is to improve the representativeness of outcrops as a proxy for flow in naturally fractured reservoirs, by capturing larger-scale high-resolution fracture patterns covering distances comparable to well spacing in fractured reservoirs, followed by modelling of aperture and flow representative of subsurface conditions. We illustrate the effectiveness of the workflow using an example of 2-D fracture patterns in outcropping carbonates in the Potiguar Basin, NE Brazil (Bisdom et al., 2017, de Graaf et al., 2017).
Section snippets
Image acquisition with a UAV
We use a multi-rotor UAV (Fig. 2) to acquire images of multiscale fracture patterns over an area that covers several reservoir simulation grid blocks, which are subsequently merged into 2-D georeferenced outcrop models. To ensure that an area is fully covered by images with constant overlap, flight paths are programmed prior to flights (Fig. 3). The programmed flights are automatically executed and controlled using a GNSS sensor (2 m accuracy) for horizontal positioning and a
Finite Element meshing and stress modelling
The 2-D fracture networks are meshed for mechanical and flow modelling, using unstructured FE meshes with explicit fractures. The meshing and the subsequent geomechanical simulations are done using ABAQUS CAE® (Dassault Systèmes®). Compared to other meshing tools, we find that this tool can handle meshing of more complex geometries, with minimal pre- and postprocessing.
Flow modelling
To construct the flow model, we use the workflow from Bisdom et al. (2016b) summarised below. Here, we extend this workflow from calculating only equivalent permeability parallel to the edges of the model to calculating the full permeability tensor to derive the principal maximum and minimum permeability values.
Flow is modelled using the same FE mesh used for the geomechanical models, where the seams in the mesh have been replaced by lower-dimensional elements to which modelled fracture
Application
The integrated workflow is applied to model permeability through an outcropping network of fractures in the Jandaíra Formation, which is a carbonate formation outcropping in large parts of the Potiguar Basin in NE Brazil. Large-scale fracture networks were formed predominantly during burial in a compressional setting (de Graaf et al., 2017). The sub-horizontal position of the rocks provides excellent exposures of multiscale fracture patterns covering areas of several hundred by several hundred
From outcrops to representative subsurface flow models
Contrary to other studies, the presented workflow uses only the outcropping network geometry as input for deterministic flow models, not taking into account outcropping apertures. Instead, we use geomechanical FE models to solve the stress state around the fracture network, based on estimates of subsurface stress conditions and rock properties. These geomechanical parameters can typically be derived from subsurface datasets, albeit with uncertainty ranges. However, the applied methodology is
Conclusions
Outcrops provide a wealth of data for studying and modelling of fracture networks, which cannot be fully captured with 1-D scanlines, as these only capture spacing and aperture of one orientation set. LiDAR on the other hand captures entire outcrops at a high resolution, but this method has limited flexibility in terms of the type of outcrops it can be applied to and in terms of processing (Hodgetts, 2013). The presented workflow enables fast generation of highly detailed realistic fracture
Acknowledgements
Total S.A. is thanked for sponsoring the PhD of the first author. The fracture patterns from the Potiguar basin were acquired with financial support from the National Petroleum Agency (ANP) of Brazil, Petrobras (through the Porocarste Project) and the Brazilian Research Council (CNPq) project “The syn-to post-rift evolution of the NE Brazil passive continental margin: implication for sedimentary systems and deformation structures” (no. 406261/2013-0, PVE), with additional support in the field
References (90)
- et al.
Optimisation of decision making under uncertainty throughout field lifetime: a fractured reservoir example
Comput. Geosci.
(2016) - et al.
Stress effects on permeability in a fractured rock mass with correlated fracture length and aperture
Int. J. Rock Mech. Min. Sci.
(2008) - et al.
Orthogonal cross joints: do they imply a regional stress rotation?
J. Struct. Geol.
(2002) - et al.
Fundamentals of rock joint deformation
Int. J. Rock Mech. Min. Sci. Geomech. Abstr.
(1983) - et al.
Some effects of scale on the shear strength of joints
Int. J. Rock Mech. Mining Sci. Geomech. Abstr.
(1980) - et al.
Strength, deformation and conductivity coupling of rock joints
Int. J. Rock Mech. Min. Sci. Geomech. Abstr.
(1985) - et al.
Ground-based and UAV-based photogrammetry: a multi-scale, high-resolution mapping tool for structural geology and paleoseismology
J. Struct. Geol.
(2014) - et al.
Inter-well scale natural fracture geometry and permeability variations in low-deformation carbonate rocks
J. Struct. Geol.
(2017) - et al.
The impact of in-situ stress and outcrop-based fracture geometry on hydraulic aperture and upscaled permeability in fractured reservoirs
Tectonophysics
(2016) - et al.
Fracturing and calcite cementation controlling fluid flow in the shallow-water carbonates of the Jandaíra Formation
Brazil. Mar. Pet. Geol.
(2017)
Analysis of heterogeneously connected rock masses by forward modeling of fractional dimension flow behavior
Int. J. Rock Mech. Min. Sci.
Algorithms for extraction of structural attitudes from 3D outcrop models
Comput. Geosci.
DigiFract: a software and data model implementation for flexible acquisition and processing of fracture data from outcrops
Comput. Geosci.
Laser scanning and digital outcrop geology in the petroleum industry: a review
Mar. Pet. Geol.
Aperture-size scaling variations in a low-strain opening-mode fracture set, Cozzette Sandstone, Colorado
J. Struct. Geol.
A general gridding, discretization, and coarsening methodology for modeling flow in porous formations with discrete geological features
Adv. Water Resour.
Polyaxial stress-induced variable aperture model for persistent 3D fracture networks
Geomech. Energy Environ.
Effects of geomechanical changes on the validity of a discrete fracture network representation of a realistic two-dimensional fractured rock
Int. J. Rock Mech. Min. Sci.
Use of UAV oblique imaging for detection of individual trees in residential environments
Urban For. Urban Green.
Evaluating the effect of internal aperture variability on transport in kilometer scale discrete fracture networks
Adv. Water Resour.
Circular scanlines and circular windows: new tools for characterizing the geometry of fracture traces
J. Struct. Geol.
A finite element framework for modeling internal frictional contact in three-dimensional fractured media using unstructured tetrahedral meshes
Comput. Methods Appl. Mech. Eng.
Geologic Analysis of Naturally Fractured Reservoirs
Comment on “A note on the scaling relations for opening mode fractures in rock” by C.H. Scholz
J. Struct. Geol.
An improved model for hydromechanical coupling during shearing of rock joints
Int. J. Rock Mech. Min. Sci.
Estimating the mean length of discontinuity traces
Int. J. Rock Mech. Min. Sci. Geomech. Abstr.
Theoretical displacements and stresses near fractures in rock: with applications to faults, veins, dikes, and solution surfaces
The use of topology in fracture network characterization
J. Struct. Geol.
Reply to comments of Jon Olson and Richard Schultz
J. Struct. Geol.
The multiscale restriction smoothed basis method for fractured porous media (F-MsRSB)
J. Comput. Phys.
High precision analysis of an embryonic extensional fault-related fold using 3D orthorectified virtual outcrops: the viewpoint importance in structural geology
J. Struct. Geol.
Building a virtual outcrop, extracting geological information from it, and sharing the results in Google Earth via OpenPlot and Photoscan: an example from the Khaviz Anticline (Iran)
Comput. Geosci.
Semi-automatic mapping of geological structures using UAV-based photogrammetric data: an image analysis approach
Comput. Geosci.
Relation between vein length and aperture
J. Struct. Geol.
Mapping folds and fractures in basement and cover rocks using UAV photogrammetry, Cape Liptrap and Cape Paterson, Victoria, Australia
J. Struct. Geol.
Subcritical crack growth in geological materials
J. Geophys. Res.
Experimental Studies of Scale Effects on Shear Strength and Deformation of Rock Joints
Non-linear behaviour for naturally fractured carbonates and frac-stimulated gas-shales
First Break
The shear strength of rock joints in theory and practice
Rock Mech. Felsmech. Mec. Des. Roches
Comparison of deterministic with stochastic fracture models in water-flooding numerical simulations
Am. Assoc. Pet. Geol. Bull.
A geometrically based method for predicting stress-induced fracture aperture and flow in discrete fracture networks
Am. Assoc. Pet. Geol. Bull.
The impact of different aperture distribution models and critical stress criteria on equivalent permeability in fractured rocks
J. Geophys. Res. Solid Earth
Cited by (94)
Characterization for three-dimensional stress of roadway roof through physics-informed neural network on monitoring data
2023, Tunnelling and Underground Space TechnologyComputed-tomography-based discrete fracture-matrix modeling: An integrated framework for deriving fracture networks
2023, Advances in Water ResourcesAre open-source aerial images useful for fracture network characterisation? Insights from a multi-scale approach in the Zagros Mts.
2023, Journal of Structural Geology