You are here
Project Reporting and Progress for Rounds 1-3
The first round of UCFER-funded proposals began work in 2016, and researchers in five out of the six projects completed their research and submitted final reports in 2018.
UCFER funded six projects in the second round, which kicked off shortly after the first round. One of those projects has submitted a final report. Researchers taking part in the other projects from round 2 began to submit their final reports during fall 2019. The six projects funded in the third round are currently in their third quarter of research.
To view quarterly reports for every round of funding, visit the “Members Only” section of the UCFER website at energy.psu.edu/ucfer. Listed below are summaries of the completed projects:
Round One Projects
Shunde Yin, University of Wyoming—A Low-Cost Technique for in-Situ Stresses and Geomechanical Properties Measurement Based on Leak-Off and Caliper Logs
Knowledge of in-situ stress and geomechanical properties is essential to understanding the potential wellbore instability and induced fracturing in the injection zone and confining zone as a result of carbon dioxide (CO2) injection in a carbon storage site. Traditional in-situ measurement of stress fields is costly, and traditional laboratory measurement of geomechanical properties is affected by the disturbed core samples. It is therefore desirable to develop a method that can keep the measurement “in-situ” while in the meantime reducing the cost and enhancing the accuracy.
To accomplish this, researchers developed an in-situ technique to measure in-situ stress and geomechanical properties at low costs and with enhanced accuracy and demonstrated the feasibility of such a technique for in-situ stress measurement by comparing them with different field data from oil fields. To achieve those objectives, the researchers took advantage of the low-cost in-situ measurements of multiple sources in the borehole such as the pressure from the leak-off test (LOT) and deformation from the caliper log.
In this work, researchers derived an analytical solution of elliptical boreholes to better relate the in-situ stresses and geomechanical properties to borehole pressure, deformation, and fracturing. An inverse of the analytical solution of elliptical boreholes was derived and solved for in-situ stresses, based on the borehole pressure measured from LOTs. They also developed a 3-D borehole simulator based on finite element methods to simulate the borehole deformation and breakouts. An artificial neural network model was developed to represent the geomechanics model and then embedded in a genetic algorithm. Based on borehole deformations measured by caliper logs, the genetic algorithm was used to search for in-situ stresses and geomechanical properties.
Researchers also developed models for the inversion of geomechanical parameters and in-situ stresses. The results were compared with the literature and the geomechanical properties obtained from NETL’s Geomechanics and Flow Laboratory and showed similar conclusions. In all, the scientists produced five papers outlining the research that were accepted by peer reviewed journals.
Behnam Jafarpour, University of Southern California—A Novel Point Process Paradigm for Stochastic Modeling and Inversion of Microseismic Monitoring Data for CO2 Storage
The objective of this project was to develop a novel geomechanically-driven point process modeling approach for accurate representation and inversion of microseismic monitoring data, as a key monitoring technology, during bulk CO2 injection into geologic formations. Accurate representation of the distribution and attributes of discrete microseismic monitoring events is critical for characterization of rock flow and mechanical properties. These properties, in turn, determine the changes in the subsurface stress and strain distributions due to CO2 injection. The proposed method established physical correlations among rock mechanical properties and discrete microseismic events.
The project developed a new framework for modeling and assimilation of microseismic monitoring for application to geologic CO2 storage. The prediction model for microseismic data is based on coupled flow and geomechanics simulation with uncertain flow and mechanical rock properties. The data assimilation algorithm developed combines predicted and recorded microseismic data to update and characterize important rock properties and to enable more accurate predictions and risk assessment. The project was performed in collaboration with researchers at the NETL.
Bruce Koel, Princeton University—Converting CO2 and Methane to Fuels by Enhanced Plasmonic Effects in a Nanotemplated Catalyst Plasma Reactor
The primary objective of this study was to demonstrate low temperature operation of a novel nanoplasma catalysis reactor characterized by an intensive and uniform discharge.
Another objective was to address the most significant challenge in the plasma-catalysis hybrid system, which was how to achieve the strongest synergistic interaction between the plasma and catalyst and to increase the selectivity for producing methanol and alkanes/alkenes. This enhanced synergy was enabled by using bimetallic nanoparticles synthesized on nanorods present on surfaces of a nanotemplated oxide support (e.g., TiO2).
To accomplish those objectives, the researchers designed and built a novel plasma catalysis hybrid reactor loaded with nanotemplated catalyst plates; constructed hierarchical catalysts to promote synergistic effects between plasma and catalysts and to increase the catalytic reactivity; synthesized novel bimetallic nanoparticles and loaded them on the nanorod support to achieve high selectivity for coupling reactions; and developed a hybrid nanosecond and direct current (DC) discharge to optimize the plasma electric field and plasma chemistry for efficient reforming.
The researchers mainly focused on constructing the plasma reactors and studying the synergistic effect of plasma and catalysts, which supported the main objectives of this proposal. They also initiated the fabrication of nanotemplated catalyst plates in the Micro/Nano Fabrication Laboratory (MNFL) at Princeton University. This was one of the main goals of this project, which is aimed at creating enhanced local fields to promote the plasma discharge.
The researchers successfully built and operated two types of dielectric barrier discharge (DBD) reactors (with a coaxial tube and parallel plates) that were used to screen catalytic effects of different catalytic metal surfaces and supported catalysts. In these studies, the researchers discovered strong synergistic effects between non-equilibrium plasma and catalysts for both the methane coupling reaction and ammonia (NH3) synthesis. Researchers compared the catalytic performance for several catalysts containing active metals (Pd, Pt, or Fe) or less active metals (Au, Ag, or Cu) or their alloys. These studies found new results on the catalytic properties of the catalysts in the presence of plasma.
The plasma-enhanced catalytic conversion of methane and CO2 is expected to have a great impact on the environment along with energy and fuel production. Since the advancement and maturity of renewable energy technologies may take a long time, transitional energy sources such as natural gas are required to maintain global economic growth and development. The advantage of this approach is that it converts methane to high-value chemicals and fuels in a one-step process and involves the conversion of CO2, a problematic greenhouse gas. Results obtained from this work contribute to a better understanding of selectivity and efficiency in a plasma promoted process for converting methane and CO2 to fuels. Additional work on the catalytic trend of metal and metal alloys in non-equilibrium plasma also contributes to a better understanding of the role of the catalyst in the plasma catalysis field. The development of a nanosecond and DC hybrid discharge provides a new capability to optimize the plasma discharge for efficient fuel reforming. In addition, the diagnostics for the electric field and electron temperature provided insights for understanding the physics and chemistry of non-equilibrium plasma.
Michael Hickner, Penn State—Efficient Reduction of CO2 in a Bipolar Electrochemical Cell
In this project, researchers developed a new type of bipolar membrane electrochemical cell with the goal of reducing CO2 into useful products. Advanced membranes and catalysts synthesized at Penn State enabled this technical accomplishment. The researchers tested membrane and catalyst components for their function and then integrated them into a device that demonstrated the reduction of CO2 to carbon monoxide (CO) and hydrogen (H2). Ultimately, the researchers achieved a Faradaic efficiency of more than 30 percent at 50 mA/cm2. The Faradaic efficiency could be tuned by the cell potential. This demonstration full cell can be further optimized by the design of the electrodes and reactant flow conditions. Additionally, this process can be weighed against other means of CO2 conversion.
This research demonstrates the ability to successfully reduce CO2 to CO and H2 with reasonable current densities and Faradaic efficiencies in a bipolar electrochemical cell. This full cell was the result of integrating new membranes and catalysts developed at Penn State. Through this project, researchers studied the performance properties of new bipolar membranes with an anion fourteen exchange layer, graphene oxide as the interfacial catalyst, and a cation exchange layer. New catalyst compositions were effectively screened using combinatorial techniques and identified for their CO2 reduction activity. The full cell performance can be improved in future projects by optimizing the electrode design and flow conditions of the cell.
To realize a high-performance CO2 bipolar electrolysis cell, the work was subdivided into bipolar membranes, catalysts, and cell testing.
The development of new bipolar membranes and catalysts resulted in a working cell that can be used for further developments in this area.
In this project, the researchers accomplished the major goals of developing a CO2 bipolar electrolysis cell based on custom bipolar membranes and anode and cathode catalysts developed at Penn State and engaged researchers at NETL to test their catalyst in a working cell and to evaluate the product distribution from this technology. Researchers are now ready to evaluate the viability of this approach versus other alternatives and examine the potential for large scale testing of this type of cell.
Ahmed Ghoniem, Massachusetts Institute of Technology—Grid Independence and Uncertainty Quantifications in Gas-Solid Flow Simulations
Through this project, researchers developed frameworks for multivariate sensitivity analysis and uncertainty quantification and machine learning integration in computational fluid dynamics (CFD) simulations.
In this project, the researchers conducted an extensive multivariate sensitivity analysis, using the Morris One-At-a-Time (MOAT) sensitivity approach, to determine the impact of model parameters on the accuracy of 3-D computational fluid dynamics-discrete element methods (CFD-DEM). Moreover, the team used a combination of physical insights and machine learning techniques to build a deep convolution network capable of mimicking the underlying dynamics of multiphase flows by accelerating CFD simulations. Our results showed that a smaller subset of the model parameters had a stronger impact on the predictions, including normal spring stiffness, normal restitution, inter particle, and particle wall friction. Low normal spring stiffness, while popularly used for reducing the computational time, leads to unphysical and highly compacted dense phase by affecting the particle velocities. Moreover, the impact of contact dissipation parameters is tightly coupled and sensitivity to any one parameter hinges on the choices of others. The researchers showed that the choice of these parameters impacts the predicted stability of bubble patterns, where nearly elastic collisions and strongly dissipative dynamics lead to inaccurate predictions.
The development and validation of CFD-DEM simulations are critical for the fundamental investigation of complex particle-scale phenomena and their coupling with reactor-scale transport. There continues to be considerable uncertainty in the selection of model parameters because of the limitations in experimental measurements. To address these issues, the researchers conducted a multivariate sensitivity analysis using thirteen model input parameters and 3-D CFD-DEM simulations with almost 170,000 glass bead particles (0.4 mm diameter) for a small-scale rectangular pulsating fluidized bed. Bubble statistics and particle dynamic metrics in more than 250 simulations showed that: choosing exceedingly low normal spring stiffness has strong implications on particle velocities; the impact of contact dissipation parameters is tightly coupled and sensitivity to any one parameter hinges on the choices of others; and stability of predicted bubble patterns is contingent on the choice of contact parameters.
Next, the researchers investigated particle dynamics inside and around bubbles to derive an optimal criterion for the selection of normal spring stiffness. They identified dilute areas in the bed as an ideal data set for the validation of the tangential damping coefficient, a model parameter which has received considerably less attention in the modeling community.
The use of machine learning techniques is gaining popularity in the CFD community because of their ability to learn complex flow features without a prior understanding of the functional form of the underlying dynamics. In this work, the research team used a combination of physical insights and machine learning techniques to build a deep convolution network capable of mimicking the underlying dynamics of gas-solid multiphase flows. The developed model made excellent predictions for bubbling fluidization. Researchers can generalize the results to other operating conditions by training on diverse data sets and/or dynamic, in-situ optimization for re-tuning network weights. However, several aspects, such as the architecture and learning parameters, must be explored further in order to make this framework more robust and capable of direct coupling with CFD.
This work on multivariate sensitivity analysis was the first of its kind and establishes a robust framework for the statistical analysis of complex, highly coupled simulations. By applying the developed methodology to CFD-DEM simulations, the researchers identified critical simulation parameters, established their coupled impact on dissipation dynamics, and proposed key guidelines for their selection. These findings have been published in peer-reviewed articles and directly contribute to the development, validation, and verification of NETL’s MFiX simulation suite. Additionally, the research team demonstrated that deep convolution neural networks can successfully capture the underlying dynamics in computational fluid dynamic simulations. Although further study is required to optimize the network architecture and the learning process, the established framework and developed tools will be extremely valuable for the simulation community.
Fred Aminzadeh, University of Southern California—Integration of Geophysical and Geomechanical Modeling
The major goals of the project were to model and identify effective and low-cost monitoring techniques for CO2 carbon capture and storage; to derive geophysical techniques (seismic) and attributes for an accurate and robust CO2 monitoring system; and to evaluate geophysical monitoring ideas for safe CO2 storage and identify any geohazard risks.
Carbon capture and storage through CO2 injection continues to exhibit high costs and is associated with risks of leakage and induced seismicity during and after injection. There are several gaps in understanding the migration of CO2 in the subsurface and the associated changes in mechanical stresses within and outside the injection layers as well as on the faults. To address these concerns, numerical modeling of multiphase flow and mechanical deformation in real-world storage systems based on rock-physics models derived from representative experiments is necessary. The researchers conducted such a study in this project on the Farnsworth Unit oilfield dataset. To achieve the above objectives, the project scope focused on fluid flow and geomechanical simulation, rock physics and 4-D seismic modeling, and validation of rock physics models with the field data.
The project was successful in developing a multiphase flow geomechanical simulation model of the Farnsworth Unit field that honors major faults, stratigraphy, and well data in the field, a Petrel database of the field data that can be used in other DOE/NETL projects, results from geomechanical experiments regarding the effect of CO2 on elastic properties and acoustic velocities at different temperatures and pore pressures, and a rock-physics model for 4-D seismic modeling. The project benefited from the research of one graduate student and a postdoctoral researcher who worked on different tasks. Researchers are preparing manuscripts of conference and journal publications based on the research findings, reports and the final project report submitted to NETL.
Round Two Project
Michael A. Hickner, Penn State—Designing Polymer/2D MOF Composite Membranes with Enhanced CO2 for CO2/N2 Separation
Overview and Achievements: The researchers developed new classes of CO2 separation membranes using a mixed matrix membrane approach. The membranes were composed of cationic polymers and 2-D metal organic frameworks. The researchers modeled these systems using density functional theory and grand canonical Monte Carlo methods to probe intermolecular interactions between the materials and gases. The research team demonstrated that the addition of metal organic frameworks with 2-D structures to polymer membranes increases the CO2 and nitrogen gas separation performance of the mixed matrix membranes compared to pure polymer membranes. This approach shows that researchers can optimize mixed matrix membranes for high membrane performance using the classes of polymers and metal organic frameworks detailed in this project.