A lot of guesswork about Bloom Energy’s fuel cells

Fuel cells hold a special place in the hearts of many energy optimists; and clean tech still doesn’t hit the mainstream media that often, so excitement (and scepticism) over start-up Bloom Energy is fairly understandable.

A segment on CBS’s 60 Minutes three days before the company’s first big press launch later today guaranteed at least a couple of days of speculative coverage. Much of the reason for the hype is the involvement of VC firm Kleiner, Perkins, Caulfield and Byers, an early investor in Google. And the less successful Segway.

Grizzled energy-watchers are harder to convince than the tech audience: even if one small company did make a transformative development, they tend to reason, the challenges of scale, grid upgrades, and other infrastructure challenges remain extremely difficult. So some of the excitement about Bloom relates to its ‘off grid’ power potential, with its stated goal to make for affordable fuel cells that could power a household in a decade – although as far as we can tell they would require natural gas.

KPCB partner John Doerr told 60 Minutes his firm invested more than $100m in Bloom, while company chief executive and co-founder KR Sridhar suggested he had raised about $400m in total and has 20 big customers already, including FedEx and Wal-Mart.

The programme said these were mostly in the $700,000 – $800,000 price range and benefited from a 20 per cent Californian state subsidy, and a 30 per cent tax break. The programme also said Google had used four of the devices for 18 months – they use natural gas, but reportedly twice as efficiently as a traditional power turbine. But as the programme notes, a lot of Bloom’s development has so far been fairly secretive – and 60 Minutes’ own report isn’t enough to satisfy the curiosity of those with an energy bent.

One of the headline statements was that Bloom wants to make devices sufficient to power a household for $3,000 in 10 years. How would that be achieved, and how cost-effective will those devices be? In particular, how efficient will they be at converting fuel into electricity, and with lower emissions? Solid oxide fuel cells (SOFCs) after all are nothing new, but so far their application has been limited.

The NY Times has some more specifics: Sridhar told the newspaper its devices can generate electricity at 8c – 10c per kWh, cheaper than some commercial prices today. From this report, it appears that some of Bloom’s research focused on using cheaper materials, such as ceramics, and some kind of ink for the anode and cathode components of the cell.

Another point made in that story by UTC, a bigger fuel-cell maker, is that reliability, particularly in the face of extreme temperatures, is a big challenge for SOFCs. One commenter on PCWeek pointed to this report by Bloom Energy on the DOE website, showing two tests of 25kW units deployed in Alaska for 2,000 hours. Again, they used natural gas as their input. The units both showed AC efficiency of about 46 per cent and there were six outages, mostly due to weather (sub-zero degree temperatures seemed to be a problem). Output in both cases fell to 20kW well before the 2,000 hours was up. The report is dated May 2009 so the final outcome of the wider project, which included testing reliability and cost over the course of a year,  wasn’t clear.

It remains to be seen whether the star-studded Bloom launch today (Arnold Schwarzenegger and Colin Powell will apparently be there) will provide enough detail for more educated guesses about whether the company might have the potential to sell mass-market, high-yielding, low-emission fuel cells.

Related links:

Bloom Energy claims a new fuel cell technology (NY Times)
Secretive fuel cell start-up stirring hopes, doubts (AP)

Energy Source is no longer updated but it remains open as an archive.

Insight into the financial, economic and policy aspects of energy and the environment.

Read our farewell note

About the blog


« Jan Mar »February 2010