Intel proposes hybrid fuel cell for notebooks

Intel Corp. is proposing a concept for a hybrid fuel cell that could drive a thin-and-light notebook computer for eight hours. The chipmaker is trying to rally notebook, battery and component makers to turn the idea into a viable product by 2007.

The 12W battery could fit into an existing DVD drive bay in a notebook and would measure about 160cc for the fuel cell and 100 cc for a fuel cartridge. Intel hopes the Smart Battery System Implementers Forum will drive the so-called Swap Bay proposal forward.

"I am working with key stakeholders in the industry on this," said Don Nguyen, a senior power architect at Intel, speaking at the Intel developer Forum Wednesday (February 18, 2004).

Fuel cells face a host of challenges before they can be commercialized, Nguyen said. Individual cells typically are only rated at 0.4V per cell, thus relatively large and hard-to-make stacks of up to 15 cells are required to get to a notebook-class 6V battery. Associated mechanisms such as air and fuel pumps, a cartridge, a dc-dc voltage regulator and a water reservoir take up more space than the cell stack itself.

"It's actually quite complicated if you look at it," he said.

Although fuel cells reliably deliver 10- to 12W of power, they do not respond well to demand spikes of 30- to 50W notebooks routinely generate. A super-capacitor or conventional on-board battery would be required to meet power demands during spikes or the fuel cell's warm-up time.

"This is where the hybrid model makes sense," Nguyen said. The Intel architect mentioned just in passing regulatory and logistics hurdles fuel cells face. Transportation authorities would need to approve use on planes of pressurized cans of fuel, and retailers would need to make standard fuel cartridges broadly available at a reasonable cost.

Despite the challenges, a hybrid fuel cell is the best option for hitting the goal of creating a notebook with an eight-hour battery life by 2007, Nguyen concluded. Indeed, because fuel cells are refillable, the system's battery life would only be capped by the amount of fuel on hand.

In a presentation, Nguyen briefly discussed other alternatives for getting to the eight-hour notebook by about 2007 including advanced lithium polymer, silver-zinc and solid-state batteries. "You will need to adopt some form of emerging battery technology to reach this goal," he said.

Advanced lithium polymer holds promise for production availability by 2006. However, it has the lowest energy density by weight of all the alternatives and faces challenges with cell swelling. Intel has measured current prototypes at 550W-h/L but believes they could reach 800W-h/L by 2006.

Silver-zinc batteries have slightly higher energy density by weight. However they could have an unknown cost premium and currently lack backing from multiple vendors.

Solid-state batteries based on a lithium cobalt oxide offer the highest energy density by weight, offer significantly faster recharge times and support three times more recharge cycles than the alternatives. Batteries made by companies such as Front Edge Technology also appear to be safer and environmentally cleaner because they use ceramics, rather than a toxic chemical as an electrolyte.

However, finding ways to stack the 20- to 30?m thick cells without losing efficiency, still appears problematic, and the batteries are expected to carry a significant cost premium, Nguyen said.

He also cautioned against using two traditional 6V lithium-ion cells in a notebook. That approach can double the charge time for a system, he said.

A Hewlett-Packard notebook engineer attending the presentation said he thought Nguyen's estimates for a 2007 hybrid fuel cell were much more realistic than estimates made a year ago that such power sources could be commercialized by 2005. Today's prototype fuel cells are far too large for notebooks, he said.

- Rick Merritt

EE Times