Multilevel Monte Carlo methods for computing failure probability of porous media flow systems

We study improvements of the standard and multilevel Monte Carlo method for point evaluation of the cumulative distribution function (failure probability) applied to porous media two-phase flow simulations with uncertain permeability. To illustrate the methods, we study an injection scenario where we consider sweep efficiency of the injected phase as quantity of interest and seek the probability that this quantity of interest is smaller than a critical value. In the sampling procedure, we use computable error bounds on the sweep efficiency functional to identify small subsets of realizations to solve highest accuracy by means of what we call selective refinement. We quantify the performance gains possible by using selective refinement in combination with both the standard and multilevel Monte Carlo method. We also identify issues in the process of practical implementation of the methods. We conclude that significant savings in computational cost are possible for failure probability estimation in a realistic setting using the selective refinement technique, both in combination with standard and multilevel Monte Carlo.

[1]  Rainer Avikainen.  On irregular functionals of SDEs and the Euler scheme , 2009, Finance Stochastics.

[2]  C. R. Dietrich,et al.  Fast and Exact Simulation of Stationary Gaussian Processes through Circulant Embedding of the Covariance Matrix , 1997, SIAM J. Sci. Comput..

[3]  R. LeVeque.  Finite Volume Methods for Hyperbolic Problems: Characteristics and Riemann Problems for Linear Hyperbolic Equations , 2002 .

[4]  Michael B. Giles,et al.  Multilevel Monte Carlo Path Simulation , 2008, Oper. Res..

[5]  A. Niemi,et al.  Effects of permeability heterogeneity on CO2 injectivity and storage efficiency coefficient , 2016 .

[6]  Eric Croiset,et al.  CO2 sequestration in Ontario, Canada. Part I: storage evaluation of potential reservoirs , 2004 .

[7]  Klaus Ritter,et al.  Multilevel Monte Carlo Approximation of Distribution Functions and Densities , 2015, SIAM/ASA J. Uncertain. Quantification.

[8]  Fabio Nobile,et al.  Multi-index Monte Carlo: when sparsity meets sampling , 2014, Numerische Mathematik.

[9]  A. Teckentrup,et al.  Improved Multilevel Monte Carlo Methods for Finite Volume Discretisations of Darcy Flow in Randomly Layered Media , 2015, 1506.04694.

[10]  Michael B. Giles,et al.  Multilevel Monte Carlo methods , 2013, Acta Numerica.

[11]  Elisabeth Ullmann,et al.  Further analysis of multilevel Monte Carlo methods for elliptic PDEs with random coefficients , 2012, Numerische Mathematik.

[12]  K. A. Cliffe,et al.  Multilevel Monte Carlo methods and applications to elliptic PDEs with random coefficients , 2011, Comput. Vis. Sci..

[13]  Daniel Elfverson,et al.  A Multilevel Monte Carlo Method for Computing Failure Probabilities , 2014, SIAM/ASA J. Uncertain. Quantification.

[14]  J. Tinsley Oden,et al.  Practical methods for a posteriori error estimation in engineering applications , 2003 .

[15]  Andrea Barth,et al.  Multi-level Monte Carlo Finite Element method for elliptic PDEs with stochastic coefficients , 2011, Numerische Mathematik.

[16]  R. H. Brooks,et al.  Hydraulic properties of porous media , 1963 .

[17]  Zhangxin Chen,et al.  An Improved IMPES Method for Two-Phase Flow in Porous Media , 2004 .

[18]  C. Doughty,et al.  Simple model representations of transport in a complex fracture and their effects on long‐term predictions , 2008 .

[19]  S.,et al.  " Goal-Oriented Error Estimation and Adaptivity for the Finite Element Method , 1999 .

[20]  R. Tempone,et al.  A continuation multilevel Monte Carlo algorithm , 2014, BIT Numerical Mathematics.

[21]  L. Gelhar.  Stochastic Subsurface Hydrology , 1992 .

[22]  Endre Süli,et al.  Adaptive finite element methods for differential equations , 2003, Lectures in mathematics.

[23]  L. Lake,et al.  Enhanced Oil Recovery , 2017 .

[24]  Sally M. Benson,et al.  Heletz experimental site overview, characterization and data analysis for CO2 injection and geological storage , 2016 .

[25]  Daniel Elfverson,et al.  Uncertainty Quantification for Approximate p-Quantiles for Physical Models with Stochastic Inputs , 2014, SIAM/ASA J. Uncertain. Quantification.

[26]  Shoichi Tanaka,et al.  Possibility of underground CO2 sequestration in Japan , 1995 .

[27]  Robert Scheichl,et al.  Finite Element Error Analysis of Elliptic PDEs with Random Coefficients and Its Application to Multilevel Monte Carlo Methods , 2013, SIAM J. Numer. Anal..

[28]  Auli Niemi,et al.  Upscaling of the constitutive relationships for CO2 migration in multimodal heterogeneous formations , 2013 .

[29]  Andreas Hellander,et al.  MOLNs: A cloud platform for interactive, reproducible and scalable spatial stochastic computational experiments in systems biology using PyURDME , 2015, SIAM J. Sci. Comput..

[30]  Collected Reprint Series.  12. Stochastic Subsurface Hydrology From Theory to Applications , 2014 .

[31]  Stefan Heinrich,et al.  Multilevel Monte Carlo Methods , 2001, LSSC.

[32]  Rüdiger Verfürth,et al.  A posteriori error estimation and adaptive mesh-refinement techniques , 1994 .

[33]  Fabio Nobile,et al.  Optimization of mesh hierarchies in multilevel Monte Carlo samplers , 2014, Stochastics and Partial Differential Equations Analysis and Computations.