Bloom Energy has excited techno-watchers. Led by K.R. Sridhar, a graduate of Madras University, who developed comparable technology for National Aeronautics and Space Administration’s space missions, this stealth clean-tech start-up—which raised $400 million in venture capital—hit the airwaves with an awe-struck report on American national TV last month. Its vision is compelling: a solid-oxide fuel cell based on a high-temperature chemical reaction between oxygen and hydrogen or a hydrocarbon fuel, with no combustion.
A box the size of a toaster could eventually power a home at twice the efficiency of a traditional gas-burning system, with 60% fewer emissions. A small-truck-sized server supplies 100kW, enough for an office building. Several have been installed at beta test sites such as eBay and Google. The key technology is a stack of floppy-disk-sized zirconium-oxide electrodes, coated with proprietary inks. The electrodes are not made of precious metals, but from sand using a low-end semiconductor process.
There are many reasons why this idea has great potential. One, it can be “off-grid”, that is, it can be used remotely with no infrastructure. Two, it is “fuel-agnostic”, that is, it can run on any hydrocarbon such as propane, natural gas, ethanol, bio-diesel, or biogas or farm waste (for instance, methane). Three, remarkably, the process is “reversible”, which means it can take in renewable energy from solar or wind and create storable oxygen and hydrogen, which can then be combined to generate electricity at night or when there is no wind (or used in a hydrogen-powered vehicle).
Especially for energy-starved India, the implications are huge: Without expensive transmission networks, and at reduced emission levels, it may be possible to produce power at grid-comparable prices. This could be a technological leapfrog, just as in cellular telephony where India bypassed the expense of copper wire in the ground. India could avoid large infrastructural sunk costs while providing hitherto unreached citizens with electricity.
The current Indian grid is unreliable and is prone to frequent power failures and voltage fluctuations, forcing many to depend on diesel generators. And replacing dirty, noisy generators with Bloom servers could be a good idea. Even better, rural users could harness gobar gas and other local by-products instead of imported hydrocarbons. Bloom also brings the robustness and reliability of a distributed network compared with a centralized one. One nightmare scenario strategic planners worry about is cyber-attackers hijacking the energy grid of a nation and bringing it to its knees.
There are other possibilities: providing electricity on-demand, comparable to “cloud computing” using server farms. Bloom uses the terminology of servers and farms; and Google, cloud computing pioneer and early adopter of Bloom’s technology, has received a licence to sell energy. Google Power may soon compete with national grids.
There is enthusiasm about the possible benefits, but there are also obstacles in Bloom’s path. The most significant is cost, now running at $750,000 for a 100kW server; although Sridhar estimates that larger volumes can bring this down to $3,000 for a home unit in 5-10 years. Even though they have had beta sites for some time, some issues do remain: Can it deal with India’s dust? How do you store generated hydrogen without it exploding?
The proof will be in the pudding. However, there are bad precedents: the fabled Segway was supposed to transform personal transport; Motorola’s 77-satellite Iridium project was meant to revolutionize wireless telephony—both were expensive flops. But scepticism aside, if Bloom can drive down costs, this technology could well be a boon. It could be the proverbial game changer.
Rajeev Srinivasan is a management consultant focusing on innovation and energy. Comments are welcome at firstname.lastname@example.org