IBM is busy looking for ways to make lithium-air batteries with 10 times thestoragecapacity of their lithium-ion cousins, and it has a Swiss supercomputer being cooled with water at the chip level – the technology could find its way into cooling data center servers and other IT equipment.
Both projects announced Tuesday represent moves by the computing giant to get into more energy-efficient pastures, so to speak. In the case of its water-cooled supercomputer, IBM is hoping to find more efficient ways to cool the IT equipment. The equipment takes up about 2 percent of the world's energy.
The project with the Swiss Federal Institute of Technology in Zurich plans to introduce chip-level water-cooling to a 10-teraflop supercomputer made of two IBM BladeCenter servers. Each processor will have a microscale liquid cooler system that IBM expects to save 40 percent on the normal energy required to keep the machines cool enough to operate.
While water cooling at the mainframe or rack level are already available, "this now takes it to the next step," said Rich Lechner, IBM's vice president of energy and environment. Water is about 4,000 times better at removing heat than air is, and cutting heat within the processors themselves is a lot cheaper than using air conditioning systems to cool the racks or rooms the servers are in.
While the supercomputer isn't set to be up and running until next year, and the company isn't saying yet when chip-level water-cooling might be commercially available, observers can, "Think of this as a prototype," he said.
Beyond the energy saved by chip-level water-cooling, the system will apply the heat it collects toward heating the Zurich facility it will be housed in, Lechner said. That will help the building cut its energy use enough to see an 85-percent reduction in its carbon footprint, he said.
IBM has done similar IT waste heat recovery projects, such as heating a swimming pool using heat from another IBM data center in Switzerland, he noted.
IBM has been making a big push into IT energy efficiency, announcing in 2007 that it would spend $1 billion per year on new products and services for data-center efficiency, and working with data center energy monitoring and management companies such as Folsom, Calif.-based SynapSense (see What Big Companies Want From Green Startups).
It's also making a big push into water management, a move that fits in with a string of smart grid-related projects it's involved in (see IBM Dives Into Smart Water Management and Top Ten Smart Grid: IBM).
But what about batteries? IBM's announcement that its Almaden Research Center in San Jose, Calif. will lead a new research effort into lithium-air batteries may seem like an odd decision for a company with little background in energy storage.
But IBM does have experience in membrane technology – something it has applied to water filtration, and could also be applied to lithium-air batteries, Lechner said.
Lithium-air batteries, like their cousins zinc-air batteries, generate power by exposing metal and an electrolyte to oxygen, oxidizing the metal and releasing energy (see Electricity From Air and Zinc? A Growing Chorus Says Yes).
Zinc-air batteries are already commercially available for some uses like powering hearing aids, and developers say they can offer four times the energy storage of lithium-ion batteries.
Theoretically at least, lithium-air batteries could provide about 10 times the energy density – the amount of energy stored per kilogram – than the roughly 200 kilowatts per kilogram that cutting-edge lithium-ion batteries now provide, said Spike Narayan, the functional manager for science and technology at IBM's Almaden research lab.
The main challenge metal-air batteries pose is that it's very hard to reverse the chemical reaction that provides their energy without putting more energy into it than you'd get out of it. Meaning: Short of replacing the chemical components of the batteries, they can't be recharged, Narayan said.
It's a problem researchers have been trying to figure out for decades.
"I think a small startup with some venture funding is not the environment in which to tackle this grand challenge," Narayan said.
Advances will require finding the right catalysts to reverse the chemical reaction at low enough energy levels, as well as advances in nanotechnology to distribute that catalyst close enough to the metals so that it is effective, he said.
While IBM hasn't publicly announced its partners on the project, Narayan said they would be both industry and government research leaders in the field. IBM plans to unveil more about its plans in August.
Eventually, IBM envisions partnerships with battery makers or automotive companies to bring advances to commercial production, he said.
The technology could also have implications for storing energy for the electricity grid, he noted – something that Drew Clark, of IBM's Venture Capital Group, in December said is likely to be one of the hottest markets for venture capital investment this year (see For 2009, It's All About Smart Grid and Storage).