Mapping natural fracture networks using stochastic and machine learning approaches

A. Chandna, S. Srinivasan

Research output: Contribution to conferencePaper

Abstract

Traditional stochastic approaches for mapping natural fractures in a reservoir have issues representing the complex connectivity of these networks. A new improved Multiple Point Statistics (MPS) algorithm was tested on the field data obtained from Teapot Dome outcrop fracture image. The algorithm was able to successfully capture the orientations of observed different fracture sets. Our ultimate objective is combination of MPS statistical information with prescribed physical geomechanical criteria for fracture propagation. For this we envision an Artificial Neural Networks (ANN) approach for calibrating the probability of fracture propagation from geomechanics. To that end, we implemented an ANN approach for fracture modeling that mimics the MPS approach. Preliminary results indicate that an ANN by itself may have trouble predicting fracture connectivity accurately due to the limited size of the calibration training set. However, the combination of MPS statistics based fracture modeling and ANN for injecting physics into fracture simulation maybe feasible and that is the future direction of our research.

Original languageEnglish (US)
Pages250-254
Number of pages5
StatePublished - Jan 1 2019
Event20th Annual Conference of the International Association for Mathematical Geosciences, IAMG 2019 - State College, United States
Duration: Aug 10 2019Aug 16 2019

Conference

Conference20th Annual Conference of the International Association for Mathematical Geosciences, IAMG 2019
CountryUnited States
CityState College
Period8/10/198/16/19

Fingerprint

Stochastic Networks
fracture network
Machine Learning
artificial neural network
Artificial Neural Network
Statistics
fracture propagation
connectivity
Connectivity
geomechanics
Propagation
machine learning
Dome
modeling
dome
outcrop
physics
Modeling
statistics
calibration

All Science Journal Classification (ASJC) codes

  • Earth and Planetary Sciences(all)
  • Mathematics (miscellaneous)

Cite this

Chandna, A., & Srinivasan, S. (2019). Mapping natural fracture networks using stochastic and machine learning approaches. 250-254. Paper presented at 20th Annual Conference of the International Association for Mathematical Geosciences, IAMG 2019, State College, United States.
Chandna, A. ; Srinivasan, S. / Mapping natural fracture networks using stochastic and machine learning approaches. Paper presented at 20th Annual Conference of the International Association for Mathematical Geosciences, IAMG 2019, State College, United States.5 p.
@conference{a39644519d47491cb4e60f98b10f00d8,
title = "Mapping natural fracture networks using stochastic and machine learning approaches",
abstract = "Traditional stochastic approaches for mapping natural fractures in a reservoir have issues representing the complex connectivity of these networks. A new improved Multiple Point Statistics (MPS) algorithm was tested on the field data obtained from Teapot Dome outcrop fracture image. The algorithm was able to successfully capture the orientations of observed different fracture sets. Our ultimate objective is combination of MPS statistical information with prescribed physical geomechanical criteria for fracture propagation. For this we envision an Artificial Neural Networks (ANN) approach for calibrating the probability of fracture propagation from geomechanics. To that end, we implemented an ANN approach for fracture modeling that mimics the MPS approach. Preliminary results indicate that an ANN by itself may have trouble predicting fracture connectivity accurately due to the limited size of the calibration training set. However, the combination of MPS statistics based fracture modeling and ANN for injecting physics into fracture simulation maybe feasible and that is the future direction of our research.",
author = "A. Chandna and S. Srinivasan",
year = "2019",
month = "1",
day = "1",
language = "English (US)",
pages = "250--254",
note = "20th Annual Conference of the International Association for Mathematical Geosciences, IAMG 2019 ; Conference date: 10-08-2019 Through 16-08-2019",

}

Chandna, A & Srinivasan, S 2019, 'Mapping natural fracture networks using stochastic and machine learning approaches', Paper presented at 20th Annual Conference of the International Association for Mathematical Geosciences, IAMG 2019, State College, United States, 8/10/19 - 8/16/19 pp. 250-254.

Mapping natural fracture networks using stochastic and machine learning approaches. / Chandna, A.; Srinivasan, S.

2019. 250-254 Paper presented at 20th Annual Conference of the International Association for Mathematical Geosciences, IAMG 2019, State College, United States.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Mapping natural fracture networks using stochastic and machine learning approaches

AU - Chandna, A.

AU - Srinivasan, S.

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Traditional stochastic approaches for mapping natural fractures in a reservoir have issues representing the complex connectivity of these networks. A new improved Multiple Point Statistics (MPS) algorithm was tested on the field data obtained from Teapot Dome outcrop fracture image. The algorithm was able to successfully capture the orientations of observed different fracture sets. Our ultimate objective is combination of MPS statistical information with prescribed physical geomechanical criteria for fracture propagation. For this we envision an Artificial Neural Networks (ANN) approach for calibrating the probability of fracture propagation from geomechanics. To that end, we implemented an ANN approach for fracture modeling that mimics the MPS approach. Preliminary results indicate that an ANN by itself may have trouble predicting fracture connectivity accurately due to the limited size of the calibration training set. However, the combination of MPS statistics based fracture modeling and ANN for injecting physics into fracture simulation maybe feasible and that is the future direction of our research.

AB - Traditional stochastic approaches for mapping natural fractures in a reservoir have issues representing the complex connectivity of these networks. A new improved Multiple Point Statistics (MPS) algorithm was tested on the field data obtained from Teapot Dome outcrop fracture image. The algorithm was able to successfully capture the orientations of observed different fracture sets. Our ultimate objective is combination of MPS statistical information with prescribed physical geomechanical criteria for fracture propagation. For this we envision an Artificial Neural Networks (ANN) approach for calibrating the probability of fracture propagation from geomechanics. To that end, we implemented an ANN approach for fracture modeling that mimics the MPS approach. Preliminary results indicate that an ANN by itself may have trouble predicting fracture connectivity accurately due to the limited size of the calibration training set. However, the combination of MPS statistics based fracture modeling and ANN for injecting physics into fracture simulation maybe feasible and that is the future direction of our research.

UR - http://www.scopus.com/inward/record.url?scp=85074989170&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85074989170&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85074989170

SP - 250

EP - 254

ER -

Chandna A, Srinivasan S. Mapping natural fracture networks using stochastic and machine learning approaches. 2019. Paper presented at 20th Annual Conference of the International Association for Mathematical Geosciences, IAMG 2019, State College, United States.