News

June 2024

April 2024

March 2024

October 2023

July 2023

June 2023

February 2023

January 2023

December 2022

October 2022


August 2022


July 2022

July 202

June 2022

 

May 2022


April 2022

Ph.D. student Hongsheng has a deep learning paper accepted at Journal of Petroleum Science and Technology.  Congratulations!   The paper is entitled: "Deep-learning-based workflow for boundary and small target segmentation in digital rock images using UNet++ and IK-EBM". 

Segmentation of boundary and small targets in 3D X-ray CT images is challenging due to the partial volume blurring effect, which is caused by the fact that the volume of a boundary or small object is close to the size of a single CT voxel. In this paper, we focus on the improvement of training data quality and development of a novel UNet++ deep learning model.   For training data quality improvement, we use the entropy-based-masking indicator kriging (IK-EBM) method, which is based on the covariance matrices between data and data and between data and unknowns, to improve the labeling accuracy for the pixels near boundaries and small targets.  Based on the high-quality training datasets generated by IK-EBM, the UNet++ is a state-of-the-art deep learning model that integrates multiscale features across different depths of the encoder-decoder network, which enables a smooth semantic transition and consequently faster convergence and more accurate extraction of fine-scale features near boundaries and small targets.  

March 2022

January 2022

Zihao has a new paper accepted by Journal of Petroleum Science and Technology.  Congratulations!!   The paper is entitled: "Experimental Investigation of Non-monotonic Fracture Conductivity Evolution in Energy Georeservoirs".

This work is the first study that uses well-controlled laboratory experiments to comprehensively investigate non-monotonic fracture conductivity evolution as a function of increasing proppant concentration under various effective stresses, proppant particle sizes, rock types, and water soaking time.  The findings from this experimental study will advance the fundamental understanding of proppant embedment and compaction and will contribute to the development of workflows for optimizing proppant placement and maximizing productivity in hydraulic fracturing.  

December 2021

Dr. Chen gave an invited presentation at ExxonMobil’s Corporate Strategic Research Center in New Jersey.  The talk was entitled: "Integration of Experimental, Numerical, and Machine Learning Methods for Subsurface Energy and Environmental Systems".    

November 2021

Long-term storage of carbon dioxide (CO2) in geological formations, such as deep saline aquifers, is a promising solution to mitigating the impact of anthropogenic CO2 emissions on global climate change. CO2 dissolved in formation water increases the solution density and can lead to miscible density-driven downward convection, which accelerates the dissolution of CO2 in formation water and thus improves the long-term security of the system. However, investigations of the critical system parameter and critical time scales for triggering downward convection have relied heavily on numerical simulations because of the challenges associated with laboratory experiments. In this study, we used experimental methods to find an empirical linear correlation between reflected visible light intensity and solute concentration, which enabled in situ measurements of solute concentrations in the spatial and temporal domains. Using these novel experimental techniques, we determined the critical Rayleigh-Darcy number and critical time scales for the onset of density-driven instability and convective dissolution.

October 2021

September 2021

August 2021

It has been our honor and privilege to have the opportunities working with the outstanding faculty at Virginia Tech.  In the meanwhile, we look forward to starting the new journey at Stevens Institute of Technology!


July 2021

June 2021




This work involved laboratory experiments, theoretical model development, and numerical simulation.   Specifically, we used a pressure pulse decay permeameter (PDP) to measure shale's apparent permeability evolution under comprehensive combinations of pore and confining pressures.   We then developed a novel multi-physics multi-scale multi-porosity shale transport (M3ST) mode to fit and interpret the PDP measurement data, which provides deep insight into the fundamental mechanisms that regulate shale's apparent permeability, such as geomechanics, fluid dynamics and transport, and the Klinkenberg effect. 

April 2021

February 2021

September 2020

August 2020

June 2020

May 2020

February 2020

      Two fundamental research projects aim to reduce the impact of carbon emissions on climate



December 2019

      Two Research Projects Aim to Reduce Carbon Emissions


October 2019

      Study has wide-reaching impact on safe storage of nuclear energy waste

August 2019

July 2019

     Critical Study has Wide Reaching Impact on Security of Nuclear Energy

July 2019

June 2019

April 2019

January 2019



December 2018

August 2018



November 2017

August 2017

June 2017

May 2016

August 2015

August 2015