Lab Home  Phone  Search  


Discrete fracture networks (DFN) are often used to accurately model flow and transport in fractured porous media.Highfidelity flow and transportsimulation on a large DFN involving thousands of fractures is computationally expensive. This makes uncertainty quantification studies of quantities of interest such as travel time through the network computationally intractable, since hundreds to thousands of runs of the DFN model are required to get good bounds on the uncertainty of the predictions.In this context, we present a systemreduction technique for DFNs using supervised machinelearning via a Random Forest Classifier. Theinsample errors (in terms of precision and recall scores) of the trained classifier are found to be very accurate indicators of the outofsample errors, thus exhibiting that the classifier generalizes well to test data. Moreover, this systemreduction technique yields subnetworks as small as 12\% of the full DFN that still recovertransport characteristics of the full network such as the peak dosage and tailing behaviour for late times. Most importantly, the subnetworksdo not get disconnected, and their size can be controlled by a single dimensionless parameter.Furthermore, measures of KLdivergence and KSstatistic for the breakthrough curves of the subnetworks with respect to the full networkshow physically realistic trends in that the measures decrease monotically as the size of the subnetworks increase. Host: David Métivier 