Join researchers from Pacific Northwest National Laboratory at the ARPA-E 2022 Summit. The conference showcases technologies and people from different technical disciplines working on new ways to overcome America’s energy challenges. At this year’s conference, five Pacific Northwest National Laboratory projects will be on display. A complete list of the projects, team leads, and booth numbers is below. We hope to see you there.
Machine learning for natural gas to electric power system design
Team Lead: Jie Bao
Booth Number: 634
Decarbonizing the nation's power grid and solving the most critical energy and environmental challenges requires robust computational capabilities. Traditional research and development methods won't get us there fast enough. To this end, a team of researchers are leveraging artificial intelligence to accelerate the design of chemical and energy systems.
The team developed an open-source library of machine learning tools that automate the entire conceptual system design process. These tools enable industry analysts and power system engineers to:
- Automatically build reduced order models using physics-informed machine learning.
- Explore relationships and connect units for chemical and energy systems, such as solid oxide fuel cell power island, reformers, heat exchangers, and more, using reinforcement learning and graph network methods without requiring existing draft system designs for training.
- Automatically adjust connections to achieve key performance targets, such as energy efficiency, product purity, or revenue.
The toolset was created using the Process Systems Engineering framework from the Institute for the Design of Advanced Energy Systems. This ARPA-E project is a collaboration between Pacific Northwest National Laboratory, National Energy Technology Laboratory, and the University of Washington.
Oxide Dispersion Strengthened Steels for Fusion Environments
Team Lead: Dalong Zhang
Booth Number: 1105
Fusion can provide abundant, carbon-free, base-load energy to power the world and curb climate change. But finding the right structural materials that can maintain mechanical strength and reduced activation at high temperatures is a challenge. Oxide dispersion strengthened steel is a promising material for fusion reactors, but conventional methods for steel fabrication, including prolonged ball milling and thermomechanical processing, are time-consuming, labor-intensive, and expensive.
Faster Fabrication, Stronger Steel
Microstructure Optimization and Novel Processing Development of Oxide Dispersion Strengthened Steels for Fusion Environments , or MONDO-FE, leverages novel metal powder processing and advanced manufacturing methods, including shear assisted processing and extrusion, or ShAPE, and laser-powder bed fusion to enable scalable and cost-effective manufacturing of reduced activation oxide dispersion strengthened steel. ShAPE and laser-powder bed fusion processes will generate high density of nano-size oxide particles, leading to high tensile and creep strength.
This novel method of steel fabrication will enable power conversion cycles that are 40 percent more efficient than traditional methods at operating temperatures beyond 900 K (623 °C).
This ARPA-E project is a collaboration between Pacific Northwest National Laboratory, Ames Laboratory, and North Carolina State University.
High-Performance Adaptive Deep Reinforcement Learning-Based Real-Time Emergency Control
Team Lead: Yousu Chen
Booth Number: 915
On any given day, about half a million utility customers experience power outages lasting two hours or more. Studies estimate that annual economic loss from all power outages in the United States is between $30–50 billion. Existing emergency control technologies lack the real-time capabilities needed to respond quickly to emergency conditions as they happen.
High-Performance Adaptive Deep Reinforcement Learning-Based Real-Time Emergency Control, or HADREC, is a decision support tool for grid operators that leverages artificial intelligence. HADREC features real-time, validated, and predictive emergency control capabilities needed to effectively safeguard the grid against costly disturbances from extreme weather and other disruptive events.
Using HADREC, grid operators can confidently respond to conditions nearly six time faster than available methods and reduce system recovery time by at least 10 percent. This results in a more resilient and secure grid, reduced operating costs for power companies, and prevents billions of dollars lost each year due to outages.
- HADREC has been successfully validated in a Texas-sized test system with more than 56,000 real-world scenarios.
- The decision-making time for one step is less than 0.01 seconds, while average control is 20 percent more effective than existing solutions.
- For test scenarios requiring emergency control, total load shedding amount is reduced by 20 percent on average and generator tripping time is reduced by 25 percent on average.
HADREC is a collaboration between Pacific Northwest National Laboratory, Google Research, PacifiCorp, and V&R Energy.
Grid Optimization (GO) Competition
Team Lead: Stephen Elbert
Booth Number: 936
ARPA-E’s Grid Optimization (GO) Competition comprises a series of prize challenges to accelerate the development and comprehensive evaluation of new software solutions for tomorrow’s electric grid. Key areas for development include, but are not limited to, optimal utilization of conventional and emerging technologies, management of dynamic grid operations (including extreme event response and restoration), and management of millions of emerging distributed energy resources. Innovative solutions and technologies identified by the GO Competition will enable greater flexibility, reliability, security, energy efficiency, and resilience for the power grid while substantially reducing costs.
Since the start of the GO Competition, ARPA-E has awarded $5.8 million to 19 winning teams. Challenge 2: Monarch of the Mountain and Challenge 3 are open for submission. Could your team be next?
Find out how to compete by visiting gocompetition.energy.gov.
ARPA-E Go Competition is a collaboration between Pacific Northwest National Laboratory, Arizona State University, Texas A&M University, Georgia Tech, Los Alamos National Laboratory, and National Renewable Energy Laboratory.
Selective Thermal Emission Coatings for Improved Turbine Performance
Team Lead: Peter McGrail
Booth Number: 107
Improving the energy efficiency of gas turbines, aviation turbines, and other industrial turbines can reduce carbon emissions and get the nation one step closer to meeting ambitious clean energy goals. One way to do this is by increasing the operating temperatures of the gas turbines. However, current materials used in turbine components, such as nickel or cobalt-based superalloys, cannot withstand ultra-high heat.
Enabling Heat Resistant Materials
To solve this problem, a team of materials researchers are developing a new type of high-performance thermal coating that will act as a barrier to conventional heat transfer, while also altering the wavelength of light radiated from the hot turbine blade surface. The coating turns wasted heat energy into a wavelength range that can be absorbed in the turbine exhaust, thereby producing additional power or thrust, with simulations showing a potential to increase turbine power output by as much as six percent.
This ARPA-E project is a collaboration between Pacific Northwest National Laboratory, Praxair Surface Technologies, the University of Minnesota, and the University of Washington. The team is currently experimenting with different materials and manufacturing technologies to develop selective emitter coatings that can withstand operating temperatures up to 1800°C.