Stochastic Inversion of Fracture Networks Characterization via Deep Generative Model

Stochastic Inversion of Fracture Networks Characterization via Deep Generative Model

  • Date

    April 12,2022

  • Time

    4:00PM - 4:30PM

  • Venue

    Zoom only

  • Speaker

    Mr. Guodong CHEN Department of Earth Sciences, HKU

The distribution of fracture networks is crucial to characterize the behavior of flow field, as fractures provide preferential flow path. However, estimating the parameters and quantifying the uncertainty based on observing data is a nontrivial task because inverse modeling of fractured model is strongly nonlinear and non-Gaussian distributed. To address this issue, a novel inverse modeling framework is proposed for estimating the fracture field parameters. The hierarchical parameterization method is adopted: for small number of large fractures, each fracture is characterized by length, azimuth and coordination of the fracture center; for large number of small fractures, fracture density and fractal dimension are utilized to characterize the fracture networks. Moreover, variational auto-encoder and generative adversarial network (VAE-GAN) are combined to capture the distribution of the parameters of complex fracture networks, thereby mapping the high-dimensional complex parameter distribution into low-dimensional continuous parameter field. Afterwards, relying on the Bayesian framework, ensemble smoother is adopted based on the collected data from hydraulic tomography to reduce the uncertainty of the fracture distribution. Numerical cases with different complexity are conducted to test the performance of the proposed framework. The results show the proposed algorithm can estimate the dominant distribution of the fracture fields after providing sufficient prior constraint information and hydraulic measurements.

Additional information: Mr. Guodong CHEN,