U.S. Supercomputer To Model Renewable Energy Impact

New supercomputing facility will focus on how to integrate renewable energy sources into nation's energy grid.

Patience Wait, Contributor

June 26, 2013

3 Min Read

New York's 32-Story Data 'Fortress'

New York's 32-Story Data 'Fortress'

New York's 32-Story Data 'Fortress' (click image for slideshow)

The National Renewable Energy Laboratory is opening a new supercomputing facility to study ways to integrate electricity from renewable resources into the nation's power grid. The facility will also serve as model of energy efficiency for a supercomputing center.

"The unique thing right now is the coupling of high-performance computing with the energy system integration push," said Steve Hammond, director of computational sciences at NREL. "We're not focused on any one particular [energy source]," he said, "[but we're looking at how multiple energy sources might be] integrated and done in a reliable and cost-effective way."

Access to a supercomputer to tackle renewables issues will allow modeling and simulation of processes too small, too large, or too rapid for scientists to measure directly, Hammond said, as well as evaluating materials in a virtual environment.

[ Google proposes optional business tariff to fund clean energy. Read more at Google Wants To Pay More For Clean Power. ]

"We can put things in circumstances that stress materials in ways that we wouldn't want to try directly, but we can push [them] to the breaking point in simulations," he said. "Modeling and simulation allows us to explore the possible [range] of materials and operating conditions they might find themselves in. We can do life cycle analysis, computational screening, and save lots of experimental time."

Among the multitude of scenarios scientists can examine with the new supercomputer will be how to integrate renewable -- i.e., intermittent -- sources into the transmission grid. And those renewable sources could be industrial-scale, such as solar farms and clusters of wind turbines, or they could be very small but numerous, such as solar panels on residential housing that feed excess electricity back into the grid.

The Energy Systems Integration Facility (ESIF), located in Golden, Colo., is itself designed to be the world's most efficient energy facility, according to Hammond. From the building architecture to the design of the supercomputer, everything has been structured with energy conservation in mind.

Most data centers have a power usage effectiveness (PUE) rating of around 1.8, Hammond explained. "Almost half the electricity used is going to things other than computer power." By contrast, the ESIF will have a PUE of 1.06.

The supercomputer, for instance, will utilize warm water liquid cooling to keep the system cool. As the water gets hot, it will be used as the main heat source for offices and lab space. There are also plans to use the heat in other buildings on the NREL campus, and even to run it through pipes under the sidewalks to keep them free of ice and snow.

The supercomputer at the heart of the center is a new HP model that is scheduled for public introduction this fall, Hammond said.

"It's a fairly typical high-performance computing cluster … The peak capability is just over a petaflop, [and it's] made up of the latest Intel IvyBridge and Xeon processors," Hammond said. "What's distinguishing about this is that it will be the fastest computer dedicated to renewables."

ESIF is scheduled to be operational this fall. Hammond said that NREL will produce a case study of its experience creating the facility, including lessons learned, and share it via Web, white papers, and conferences. "All across the federal government we've hosted lots of visitors," Hammond said. "There are a lot of folks interested in seeing what we're doing."

About the Author(s)

Patience Wait


Washington-based Patience Wait contributes articles about government IT to InformationWeek.

Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.

You May Also Like

More Insights