Static Conceptual Fracture Modeling

Preparing for Simulation and Development

Ronald A. Nelson

• Publisher: Wiley-Blackwell
• Year: 2019
• Print ISBN:
  9781119596936
• eBook ISBN:
  9781119596950
• Edition: 1
• Topic: Physical Sciences
• Subtopic: Geology & Earth Sciences

Modelling of flow in naturally fractured reservoirs is quickly becoming mandatory in all phases of oil and gas exploration and production. Creation of a Static Conceptual Fracture Model (SCFM) is needed as input to create flow simulations for today and for prediction of flow into the future. Unfortunately, the computer modelers tasked with constructing the gridded fracture model are often not well versed in natural fracture characterization and are often forced to make quick decisions as to the input required by the software used to create these models. 

Static Conceptual Fracture Modelling: Preparing for Simulation and Development describes all the fracture and reservoir parameters needed to create the fracture database for effective modelling and how to generate the data and parameter distributions. The material covered in this volume highlights not only natural fracture system quantification and formatting, but also describes best practices for managing technical teams charged with creating the SCFM. This book will become a must on the shelf for all reservoir modelers.

1
Purpose and Scope

The purpose of this manuscript is to provide a guide for the construction of a quantitative Static Conceptual Fracture Model (SCFM) from predominantly physical descriptive rock data, which along with a Dynamic Conceptual Model (DCFM) constructed from predominantly fluid and reservoir engineering data, can be used for reservoir flow simulation (concept from Trice 2000). These simulations constrain the current reservoir behavior, as well as predict how it will perform in the future.

This manuscript will discuss the various parameters that are needed to constrain the SCFM, and later, populate computer models used to generate gridded fracture models as input to simulation. The various parameters will be detailed along with techniques I have used to gather the needed data and populate the computer models. The parameters discussed are the same regardless of the simulation modeling style used or the computer programs used to house the data and make the needed reservoir calculations. In addition, I will comment on what to do and not do in technical planning when acquiring some of the needed data sets.

An important aspect of the modeling process is, in my mind, innovation. The rock and fracture data needed for the models can come from different sources and different scales of measurement. For example, constraining fracture corridor width in the subsurface can come from core, image logs and geophysical data. All three are measured at different scales and levels of precision, therefore, giving different values and accuracy. The important thing in the modeling is to make the measurement however you can with the data you have. The variation within the distribution of the measures is perhaps more important than the actual values, as values can be shifted in bulk during history matching to obtain credible results with respect to reservoir response. The guiding principle in my mind is innovation. Get the parameter distributions however you can with the data available.

Several of the topic areas described herein have previously been documented in the two editions of a previous textbook (Nelson 1985, 2001). However, this author is not a computer modeler. Rather, I have more than 40 years' experience studying fractured reservoirs and providing real‐time assistance to the experienced computer modelers during the process.

The computer modelers know very well the ins‐and‐outs of inputting the data and successfully getting appropriate results in the proper format. However, a team of multidisciplinary workers (geologists, geophysicists, engineers, petrophysicists, etc.) are needed to generate the basic fracture and reservoir data. This team is intimately familiar with the input data and knows its strengths and weaknesses which is especially important during history matching of simulation results. They, or their representative (a Fracture Champion), are the appropriate people to assist the computer modeling expert(s) in evaluating and selecting appropriate input of the natural fracture‐related data. An excellent example of the process is given in Richard et al. (2017).

What is presented here is a procedure for completion of a well‐constrained SCFM. Of course, individual reservoir studies may lack many of the data types needed for the complete model, especially early in the history of a project. In most studies, the SCFM generation occurs in steps over time. We move from early model descriptions with little hard data to later models with richer data bases to draw from, therefore, leading to a better constrained model with lower associated risk.

The procedure detailed in this book has been created using several guiding principles. These include the following in Figure 1.1.

Figure 1.1 A compiled list of guiding principles for constructing a usable Static Conceptual Fracture Model (SCFM).

2
What Is a Static Conceptual Fracture Model and Why Do We Build It?

Current practices in the numerical simulation of fractured reservoirs rely on the creation of both static and dynamic conceptual models to construct an integrated reservoir model to be simulated. The Static Conceptual Fracture Model (SCFM) focuses on the physical description of the fracture system present and how it varies throughout the reservoir; while the Dynamic Conceptual Fracture Model (DCFM) focuses on the reservoir fluid properties and the fluid‐flow characteristics of the fractures and matrix and their variation in 3D. The first usage of the terms or model sub‐types that I know of is from Robert Trice in 2000 at an SPE Research Forum on Naturally Fractured Reservoirs in Nice, France, Figure 2.1. These conceptual models are an elemental volume representation of the reservoir representing the entire reservoir with all its variability; Davy et al. (2018), Richard et al. (2017).

Figure 2.1 Depicted is a schematic of how static and dynamic fracture data are generated and combined to form the final simulation model; after Trice (2000), SPE Research Forum, Nice, France. This was a very early introduction to the idea of parallel Static Conceptual Fracture Model (SCFM) and Dynamic Conceptual Fracture Model (DCFM) creation that are merged into the final product for simulation.

There is great variation in how SCFMs are created due to data limitations and scales of observation. Ideally, the model includes a quantitative description of all elements of the fracture system and their variation in 3D.

In my studies, I developed an organized series of steps I use to create an SCFM.

To complete the process for the creation of the SCFM, there are 13 major topic areas that must be constrained in the reservoir, Figure 2.2. This list highlights major study areas, each of which may contain multiple data sets, analysis techniques and computer applications. A similar list can be constructed for creation of a DCFM, but that is not included in this volume.

Figure 2.2 The 13 key topic areas needed to fully constrain an effective SCFM. The topic areas highlighted in green are typical for most standard fractured reservoir studies. The topic areas highlighted in blue are the additional topic areas needed for an effective quantitative fracture model.

Source: from Nelson (2011b).

Highlighted in green in Figure 2.2 (first four parameters) are the standard approaches performed in a typical qualitative fractured reservoir study with focus on...

Table of contents

• Cover
• Table of Contents
• Foreword
• Acknowledgments
• 1 Purpose and Scope
• 2 What Is a Static Conceptual Fracture Model and Why Do We Build It?
• 3 Fracture Model Creation Workflow
• 4 Gathering Natural Fracture Orientation and Intensity Data Directly
• 5 Gathering Natural Fracture Orientation and Intensity Data Indirectly
• 6 Analyzing the Natural Fracture Data Once Gathered
• 7 Gathering and Analyzing Structural Data
• 8 Gathering Constraints on Fracture Aperture
• 9 Creation of Natural Fracture Scaling Laws
• 10 Gathering and Analyzing Mechanical Property Distribution Data
• 11 Locating Fracture Corridors
• 12 Rock Anisotropy and its Importance in Determining Dominant‐Fracture Orientation and Relative Intensity
• 13 Determine the In‐situ Stress Directions and Magnitudes and their Variation
• 14 Production Calibration
• 15 Determining the Fractured Reservoir Classification and, Therefore, Which Simulation Style Is Most Appropriate
• 16 Use of Reservoir Analogs
• 17 The Importance of 3D Visualization in Data Integration and Static Fracture Model Creation
• 18 Thoughts on History Matching of Simulation Results
• 19 Preparing the Fracture Data for Input to the Gridded Model
• 20 Discussion of Error and Uncertainty in the Modeling Process
• 21 Published Examples of the Process
• 22 Final Comments
• Appendix A: Detailed Static Fracture Modeling Workflow
• Appendix B: How we Use Various Seismic Attributes to Predict Natural Fracture Intensity in the Subsurface, After Nelson (2006)
• Appendix C: How I Learned to Interpret Natural Fractures in Core
• Symbols and Abbreviations
• References
• Index
• End User License Agreement