ISSN 1018-5593

EUROPEAN COMMISSION

Interim guide to fracture interpretation

and flow modelling in fractured

reservoirs European Commission

Interim guide to fracture interpretation

and flow modelling in fractured

reservoirs

E. S. Aarseth, B. Bourgine, C. Castaing, J. R Chiles, N. R Christensen,

M. Eeles, E. Fillion, A. Genter, R A. Gillespie, E. Håkansson,

K. Zinck Jørgensen, H. F. Lindgaard, L. Madsen, N. E. Odling, C. Olsen,

J. Reffstrup, R. Trice, J. J. Walsh, J. Watterson

JOULE II

Contract No CT93-0334

Research funded in part by the European Commission

in the framework of the JOULE programme

Sub-programme 'Advanced fuel technologies:

security and supply of hydrocarbons'

Directorate-General XII

Science, Research and Development

1997 EUR 17116 EN Published by the

European Commission

Directorate-General XII

Science, Research and Development

B-1049 Brussels

LEGAL NOTICE

Neither the European Commission nor any person acting on

behalf of the Commission is responsible for the use which might be made of the

following information

Cataloguing data can be found at the end of this publication

A great deal of additional information on the European Union is available on the Internet.

It can be accessed through the Europa server (http://europea.eu.int)

Luxembourg: Office for Official Publications of the European Communities, 1997

ISBN 92-827-8812-1

© ECSC-EC-EAEC Brussels · Luxembourg, 1997

Reproduction is authorized, except for commercial purposes, provided the source is acknowledged

Printed in Italy CONTENTS

PREFACE (NH) 1

1 INTRODUCTION (FAG) 3

2 SAMPLING AND ANALYSIS OF FRACTURES 8

2.1 Introduction (FAG)

2.2 Systematics of fractures in outcrop (FAG)

2.3 Sampling of Well Data (EO & FAG) 15

3 GEOMETRICAL FRACTURE MODELLING (BRGM) 21

3.1 Introduction 2

3.2 Step 1. A deterministic and statistical 1-D fracture description of the well 2

3.3 Step. Selection of One (or More) Relevant Templates

3.4 Step 3. Simulation of a 3-D Conceptual,Near-Well Fracture Model 27

3.5 Step 4. Upscaling Of The Fracture Model9

4 FRACTURE SYSTEMS, FLUID FLOW AND EFFECTIVE PERMEABILITY (UiB) 3

4.1 Fracture system connectivity 31

4.2e apertures

4.3 Fracture orientation and permeability anisotropy5

4.4 Semi-quantitative estimates of permeability6

4.5 Discrete fracture flow modelling - the method

4.6 Usinge data in discrete fracture flow models 38

5 RESERVOIR SIMULATIONS ON THE BASIS OF FRACTURE PATTERNS. (GEUS) 4

5.1 Introduction 4

5.2 Determination of permeabilities and matrix block size from a fracture pattern.. 4

5.3 Results of simulation5

6 SUMMARY (UiB & FAG)8

6.1 Overview

6.2 Future Work

References9

GLOSSARYOFTERMS 51

A1 FRACTURE DATASETS5

A1.1 VERTICAL PERSISTENCE AND SCALING OF FRACTURES AT CAPPANAWALLA, THE

BURREN (FAG)

A1.1.1 Introduction2 Stratigraphy

A1.1.3 Mapping Technique 54 Fracture Sets8

A1.1.5 Persistence6 Cliff Section 64

A1.1.7 Joint Architecture6 8 Bedding Thickness and Spacing

A1.1.9 Spatial Distributions9

A1.1.10 Length Populations 71 Vein Thickness Populations7

A1.1.12 Connectivity3 Discussion

A1.2 FAULTS AND JOINTS IN CHALK, DENMARK (GEUS) 81

A1.2.1 Introduction 82 Fracture Data

A1.2.3 Structural Setting ·. 83 4 Concluding Remarks8 A1.3 ANALYSIS OF THE SAUDI ARABIAN FRACTURE DATASET (BRGM) 89

A1.3.1 Introduction 82 The Multiscale Fracture Datasets Of The Western Arabian Sedimentary Cover

Rocks

A1.3.3 Tectonic Interpretation Of The Multiscale Fracture Datasets Related To Red Sea

Rifting 95

A1.3.4 Scaling Analysis Of Fracture Systems 99 5 Discussion 110

A1.3.6 Conclusions2

A1.4 DATASET FROM HORNELEN, WESTERN NORWAY (IBM/UiB) 117

A1.4.1 The Devonian Basin Of Hornelen, Western Norway

A1.4.2 The Data :3 Data Resolution 126

A1.4.4 Data Analysis5 Connectivity9

A1.4.6 Conclusions 13

A1.5 OVERVIEW OF FRACTURE ANALYSIS (FAG) 134

A1.5.1 Introduction

A1.5.2Geological Setting

A1.5.3Sampling

A2 AN ATTEMPT TO SIMULATE FRACTURE NETWORKS FROM BOREHOLE IMAGERY IN A

CLASTIC RESERVOIR (BRGM) 138

A2.1 Introduction

A2.2 Simulation Method For Fracture Networks From 2-D High-Quality Datasets 13

A2.3 Application Of The Simulation Method To A Buried Clastic Reservoir 151

A2.4 Conclusions 157

A3 FLUID FLOW IN FRACTURED ROCKS AT SHALLOW LEVELS IN THE EARTH'S CRUST

AN OVERVIEW (IBM/UiB) 161

A3.1 Introduction

A3.2 Flow In Single Fractures

A3.3 Flow In Fractured Rock Masses And Fracture System Connectivity 16

A3.4 Flow In Fault Zones9

A3.5 Flow In Fractured Rock Masses Away From Fault Zones 172

A3.6 Flow Modelling In Fractured Rocks 178

A3.7 Discussion 18

A4 RESERVOIR SIMULATIONS (GEUS)

A4.1 Simulation Models.....

A4.2 Results Of The Simulations 190

A4.3 Input Parameters For The Reservoir Simulations 193

A5 DERIVATION OF A FRACTURE APERTURE MODEL AND HIERARCHICAL FLOW MODELLING

APPLIED TO THE FRACTURE SYSTEM IN HORNELEN, WESTERN NORWAY (IBM/UiB) 197

A5.1 Resolution And Fracture Length

A5.2 A Fracture Aperture Versus Trace Length Model8

A5.3 Discrete Fracture Flow Modelling

The institutions primarily responsible for each section are indicated by the following abréviations:

FAG Fault Analysis Group

BRGM Bureau de Recherches Géologique et Mineres

UiB Universitet i Bergen

GEUS Geological Survey of Denmark & Greenland (formerlyDGU)

NH Norsk Hydro

BD Enterprise Oil

IV PREFACE

Enterprise Oil Ltd and Norsk Hydro A/S joined this project in 1993 as industrial partners. The

project aims concerned problems which we were facing when producing from high porosity, low

permeability fractured chalk reservoirs and other fractured limestone reservoirs. The aims, goals

and methods put forward in the proposal and later outlined in interim reports, could give very

valuable input to how we build geological models, and how we produce and manage these

complicated reservoirs in which the understanding of 3-D fracture interconnectivity and spatial

distribution is the key to a successful field development.

The dual porosity/permeability system poses two main problems: the siting of production wells

and of injection wells. In this type of reservoir the producers must intersect several of the major

fractures to secure acceptable production rates from the rock volume. Injection wells, if wrongly

placed can result in rapid advance of injected water towards producers, isolating potentially

recoverable hydrocarbons in the fracture-bounded rock matrix. These hydrocarbons are then

almost impossible to recover.

To be able to model/simulate the subsurface fracture systems, there is a need for a better

understanding of the development and spatial distribution of fractures. How these vary with

respect to lithology, bed thickness and tectonic setting (doming, extension or compression) can

only be understood by examination of field analogues. This approach has been used in this

project, where five areas have been studied to try to unravel the significant relationships.

Based on the above considerations, the project has addressed a series of problems and made

good progress in trying to solve them. The results and methods, as presented in this interim guide

are relevant to the industry as they will improve our geological models with respect to how we

model/simulate the spatial distribution of fractures and how we "parameterize" the fracture

properties.

This report outlines a method for modelling fractures in 2-D and 3-D from a 1-D dataset, and how

we can incorporate this in our flow simulation model. This method will assist in planning the

development of such reservoirs and in optimising the reservoir performance during production. 1 INTRODUCTION

This guide is intended to assist the non-specialist reservoir geoscientist or engineer to optimise

the application of limited data to characterisation of a fractured reservoir. A procedure for

simulating fluid flow in fractured reservoirs is outlined in the guide, and flow diagrams are

provided in each chapter in order to clarify the steps which are to be undertaken. A glossary of

terms is included at the end of the guide which defines the terms printed in bold in the main text.

The authors of each chapter are indicated on the contents pages and the guide was collated

and edited by the Liverpool Fault Analysis Group. Technical details of work on which this guide is

based are given in a separate volume of appendices which contain' supplementary information.

Fractured reservoirs are becoming of increased importance as increasing numbers of such

reservoirs are discovered and exploited. The behaviour of the reservoir depends on many factors

including the relative permeabilities of the fractures and of the matrix. Typically, the effects on

the production history of fractures intersected by the production well are instantaneous, but

this reserve of hydrocarbon is quickly depleted. At this point the hydrocarbon may be drawn

from the matrix at flow rates limited by the matrix permeability. The matrix hydrocarbon may

be voluminous and therefore represent the bulk of the storage capacity of the reservoir. In

order to produce a fractured reservoir efficiently it is necessary carefully to control the

production rates and so encourage the transfer of hydrocarbons from the rock matrix to the

fracture system and on into the production well. Therefore, accurate reservoir simulations are

required. A review of the literature on fluid flow in fractured rocks is given in Appendix A3.

Fractured reservoirs represent a major challenge to reservoir geologists. Firstly, the fracturing

is frequently complex and difficult to predict and may occur on a variety of scales. Secondly,

once fractures are defined, they are usually too numerous to be incorporated individually into a

reservoir simulation, and so indirect means have to be devised for bringing this geological

information into the simulation.

The general method which we are suggesting is outlined in the flow diagram (Fig. 1.1) and in the

flow diagrams given at the beginning of each chapter which indicate the break-down into sub-

tasks. Conventional reservoir simulators cannot model large numbers of fractures explicitly

and so the discrete fracture flow model is used to provide values for the permeabilities of

the matrix plus fractures for input to the reservoir simulator. Well fracture data and knowledge

of fractures in outcrop are combined in order to produce a fracture model. This fracture model

is then used as input to an upscaling flow simulation in which the fractures are explicitly defined.

This discrete fracture flow model is used to produce upscaled permeability values. The results

given are for single phase flow only and their application to multiphase flow is uncertain.

The fractures which are dealt with here are opening-mode fractures, such as joints and veins:

faults and other shear fractures are not considered. Well fracture data are the most

important direct source of information about reservoir fractures (Chapter 2). However well data

are not sufficient to define the 3-D fracture geometry and so cannot be used alone to predict

features important for flow, such as the connectivity and the size of unfractured blocks. It is

proposed, therefore, that characteristic end-member fracture geometries defined from

outcrops should be used as an additional source of information. Thus, well data can be used to

choose between various fracture templates, which then provide the required information

about the sizes, spatial distributions and connectivities of the fractures. Because of the

importance of outcrop data, large fracture databases have been collected and are presented in

Appendix 1. It has been found that individual fractures tend either to be confined to individual

strata, in which case the spacing between thes is a simple function of the thickness of

the layer or, alternatively, that in the absence of layering they are vertically unconstrained. In

this second case the fractures typically occur on a wide range of scales and have fractal

properties. This scaling behaviour causes problems in creating fracture models and flow

simulations because the fractures are not characterised by a single scale.

As every fracture in a reservoir cannot be known explicitly, it is necessary to create a

geometricale model (Chapter 3 & Appendix A2). This model is created by combining well

data with fracture templates, and the model is conditioned to the wells so that the model

fractures and the known fractures coincide. The resulting fracture model is then used as input

to the discrete fracture flow model, which simulates flow in the modelled fractures FLOW MODELLING IN FRACTURED RESERVOIRS

Outcrop Well Fracture

Templates Data

(1-D) (2) (2-D) (2)

Fracture Modelling (3)

Discrete Fracture Flow

Row Modelling (4) Templates

t

Permeabilities

J

Fteservoir Simulation (5)

( Calibration J

Good

Well Tests

Comparison?

T

End

Figure 1.1 Organogram showing overview of procedure for flow modelling in

fractured reservoirs. Numbers refer to chapter numbers of this volume.