Quantifying Topological Uncertainty in Fractured Systems using Graph Theory and Machine Learning

Fractured systems are ubiquitous in natural and engineered applications as diverse as hydraulic fracturing, underground nuclear test detection, corrosive damage in materials and brittle failure of metals and ceramics. Microstructural information (fracture size, orientation, etc.) plays a key role in governing the dominant physics for these systems but can only be known statistically. Current models either ignore or idealize microscale information at these larger scales because we lack a framework that efficiently utilizes it in its entirety to predict macroscale behavior in brittle materials. We propose a method that integrates computational physics, machine learning and graph theory to make a paradigm shift from computationally intensive high-fidelity models to coarse-scale graphs without loss of critical structural information. We exploit the underlying discrete structure of fracture networks in systems considering flow through fractures and fracture propagation. We demonstrate that compact graph representations require significantly fewer degrees of freedom (dof) to capture micro-fracture information and further accelerate these models with Machine Learning. Our method has been shown to improve accuracy of predictions with up to four orders of magnitude speedup.

[1]  A. O'Hagan,et al.  Bayesian calibration of computer models , 2001 .

[2]  J. Thovert,et al.  Effective permeability of fractured porous media with power-law distribution of fracture sizes. , 2007, Physical review. E, Statistical, nonlinear, and soft matter physics.

[3]  David D. Nolte,et al.  Approaching a universal scaling relationship between fracture stiffness and fluid flow , 2016, Nature Communications.

[4]  Gowri Srinivasan,et al.  Predictions of first passage times in sparse discrete fracture networks using graph-based reductions. , 2017, Physical review. E.

[5]  Benoit Noetinger,et al.  A quasi steady state method for solving transient Darcy flow in complex 3D fractured networks , 2012, J. Comput. Phys..

[6]  Satish Karra,et al.  dfnWorks: A discrete fracture network framework for modeling subsurface flow and transport , 2015, Comput. Geosci..

[7]  Satish Karra,et al.  Shale gas and non-aqueous fracturing fluids: Opportunities and challenges for supercritical CO2 , 2015 .

[8]  Aric Hagberg,et al.  Exploring Network Structure, Dynamics, and Function using NetworkX , 2008, Proceedings of the Python in Science Conference.

[9]  Sergio Gómez,et al.  Ranking in interconnected multilayer networks reveals versatile nodes , 2015, Nature Communications.

[10]  Andrew Gelman,et al.  Data Analysis Using Regression and Multilevel/Hierarchical Models , 2006 .

[11]  Satish Karra,et al.  Evaluating the effect of internal aperture variability on transport in kilometer scale discrete fracture networks , 2016 .

[12]  S. P. Neuman,et al.  Trends, prospects and challenges in quantifying flow and transport through fractured rocks , 2005 .

[13]  Satish Karra,et al.  Learning on Graphs for Predictions of Fracture Propagation, Flow and Transport , 2017, 2017 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW).

[14]  D. Gunopulos,et al.  Scaling up the Naive Bayesian Classifier : Using Decision Trees for Feature Selection , 2002 .

[15]  Jean-Raynald de Dreuzy,et al.  Hydraulic properties of two‐dimensional random fracture networks following a power law length distribution: 2. Permeability of networks based on lognormal distribution of apertures , 2001 .

[16]  Thomas J. Santner,et al.  Design and analysis of computer experiments , 1998 .

[17]  N. Odling,et al.  Scaling of fracture systems in geological media , 2001 .

[18]  G. B. Hudson,et al.  Trace gas emissions on geological faults as indicators of underground nuclear testing , 1996, Nature.

[19]  Cesare Marchetti,et al.  The future of natural gas: A Darwinian analysis , 1987 .

[20]  Sonja Kuhnt,et al.  Design and analysis of computer experiments , 2010 .

[21]  Steven A. Gabriel,et al.  The Future of Natural Gas , 2010 .

[22]  Jeffrey D. Hyman,et al.  Conforming Delaunay Triangulation of Stochastically Generated Three Dimensional Discrete Fracture Networks: A Feature Rejection Algorithm for Meshing Strategy , 2014, SIAM J. Sci. Comput..

[23]  A. Munjiza,et al.  Computational Mechanics of Discontinua: Munjiza/Computational Mechanics of Discontinua , 2011 .

[24]  Javier M. Buldú,et al.  Competition among networks highlights the power of the weak , 2016, Nature Communications.

[25]  Vito Adrian Cantu,et al.  Machine learning for graph-based representations of three-dimensional discrete fracture networks , 2017, Computational Geosciences.

[26]  野村栄一,et al.  2 , 1900, The Hatak Witches.

[27]  S Karra,et al.  Modeling flow and transport in fracture networks using graphs. , 2018, Physical review. E.

[28]  S Karra,et al.  Where Does Water Go During Hydraulic Fracturing? , 2016, Ground water.

[29]  Stephen B. Seidman,et al.  Network structure and minimum degree , 1983 .

[30]  Dimitrios Gunopulos,et al.  Selective Bayesian Classifier: Feature Selection For The Naive Bayesian Classifier Using Decision Trees , 2002 .

[31]  D. Higdon,et al.  Computer Model Calibration Using High-Dimensional Output , 2008 .

[32]  A. O'Hagan,et al.  Curve Fitting and Optimal Design for Prediction , 1978 .

[33]  H. S. Viswanathan,et al.  Understanding hydraulic fracturing: a multi-scale problem , 2016, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[34]  Satish Karra,et al.  PFLOTRAN User Manual A Massively Parallel Reactive Flow and Transport Model for Describing Surface and Subsurface Processes , 2015 .

[35]  Satish Karra,et al.  Fracture size and transmissivity correlations: Implications for transport simulations in sparse three‐dimensional discrete fracture networks following a truncated power law distribution of fracture size , 2016 .