When Dave Saltzman prepares for a business trip, he charges up the main battery in his notebook computer, removes the CD-ROM drive and fills the bay with a second battery, and then packs a third one in his bag. That's sufficient for long trips, says Saltzman, systems manager at United Parcel Service in Atlanta.

Like many users, Saltzman wants to be able to work continuously during extended flights, but he also wants to use power-hungry features such as wireless networking while travelling. These changing usage patterns and the demand for faster notebooks have created a power gap between what batteries can provide and what systems can deliver.

While notebooks continue to benefit from Moore's Law (you know Moore's Law? If not read here), batteries haven't kept up. The future of disconnected computing depends on century-old electrochemical technology that has improved only gradually.

It's not that batteries haven't improved. "If we were to put today's battery on a notebook built five years ago, you'd get eight hours of battery life," says Carl Pinto, director of product development for notebooks at Toshiba in Irvine California. The problem is that mobile devices are demanding more power, he says.

Until recently, investment in battery technology has been relatively small. "In the last 100 years, there hasn't been enough work put into batteries. It's just not exciting stuff," says Rob Enderle, an analyst at Enderle Group in San Jose.

But battery life has risen to become one of the top three purchase criteria for notebook computers, says Mike Trainor, chief mobile technology strategist at Intel, which produces logic boards and chip sets used by the majority of notebook makers. "IT shops want more performance, more wireless and slimmer systems, which cuts down the room for batteries," he says.

All-day computing needs 100Watt-hours
Intel's Centrino mobile chip set has reduced power consumption, extending projected operating times from two to three hours into the five-hour range, which is still short of the all-day battery that users want. Eight hours of life would require 100 watt hours (Wh) of power, but the best available battery technology - lithium ion - delivers less than 60Wh.

Trainor is confident that Intel can "give Moore's Law's worth of features" through the end of the decade while keeping consumption at the 100Wh mark. But that still leaves a power gap. "The other side of the equation has become equally important: How do we get more energy into the system?" he says.

Fuel-cells swallow research dollars
Vendors have recently awakened to the problem, but government and private research and development dollars have poured into fuel-cell research rather than into basic battery designs. The direction of investment away from batteries has contributed to today's power gap, contends Donald Sadoway, a professor of materials engineering at MIT. "They put all of their eggs in one basket. Here we are, seven or eight years later, and fuel-cell applications are nowhere near to realisation," he says.

Users are feeling the pain. Tony Scott, chief technology officer at General Motors, says Centrino-based notebooks have improved battery efficiency 20 percent to 30 percent, but actual operating times remain under three hours. That's not always enough when people bring such computers to meetings, says Scott.

"If you get two back-to-back one-hour meetings and you're making any significant use of the machine at all, you can start running into problems. And three meetings in a row -- forget it," he says. "We need eight-plus hours of use, and that's a struggle with a lot of the devices we have today."

For Saltzman, the battery issue goes beyond notebooks. He manages 200,000 battery-powered devices at UPS, including radio-enabled handhelds used in delivery trucks. The batteries don't charge well in hot or cold weather, so charging must be done at the dispatching location. And because drivers are on the road for up to 10 hours, UPS must use bigger batteries, which adds weight to the devices. Saltzman would like to see higher energy densities to reduce weight.

Incremental improvement
Battery manufacturers have made incremental improvements in lithium ion batteries since they were introduced in the early '90s, says Kurt Kelty, director of business development at Panasonic Energy Solutions Lab, a unit of Panasonic Technologies. Over the past five years, lithium ion batteries have replaced nickel cadmium (NiCd) and nickel metal hydride (NiMH) technology in mobile computing devices.

Lithium ion offers a higher volumetric energy density. It also doesn't suffer from the memory effects that shorten the life span of NiCd batteries. And it's environmentally superior to NiCd, which faces a gradual phaseout because cadmium is toxic, making it a hazard in the waste stream.

While nickel-based chemistry has reached its capacity limit, lithium ion continues to make small gains. In recent years, capacity has increased at a rate of about 10 percent per year, while competition has reduced prices at 10 percent to 20 percent annually, Kelty says.

Although lithium ion hasn't yet hit the theoretical capacity limit, the industry consensus is that future gains will be unlikely to close the power gap. That conclusion has spurred renewed interest in battery research.

Can alternative chargeables fill the gap?
Companies such as Mississauga, Ontario-based Electrovaya Inc. use lithium ion polymer, which uses a gel-like electrolyte. Despite early promise, the technology remains more expensive than lithium ion and hasn't improved energy density. But it does have one advantage: Polymer-based cells can be formed into flat shapes that fit into small devices, while lithium ion is limited to cylindrical cell designs.

Pionics in Shiga, Japan, has shown a prototype battery with an energy density of 600Wh/litre. Most of today's lithium ion batteries fall into the 200Wh to 250Wh/litre range, says Atakan Ozbek, principal analyst at ABI Research. Still other vendors are working on designs based on materials such as lithium sulphur and lithium phosphate.

Even the traditional alkaline battery may get back into the game. With zinc-based alkaline batteries, it has been difficult to get more than 10 recharge cycles, says Robert Zeiler, president of Zinc Matrix Power in Santa Barbara, Calif. The company has replaced the alkaline battery's traditional electrolyte solution with a polymer-based formula that extends the number of recharge cycles. "We can get hundreds of cycles," with an energy density of 600Wh/liter, he says. The company has an agreement with Intel and says it will have a commercial notebook battery in production by 2006.

While Intel has invested in the more promising companies, Trainor is realistic about early claims. "What we have not seen is anyone manufacture these cells in high volume," he says.

Next: What are the prospects for fuel cells?