Parsing fact from fiction with the Bloom Energy box
CNET's Martin LaMonica is horrified he missed last week's alternative energy hoopla. Now he's trying to make sense of the excitement.
Boy, did I pick a lousy week to leave the country for a family vacation.
What did I miss? Well, a company seems to have come out of nowhere, raised loads of money, has retired Gen. Colin Powell on its board, and made some audacious claims about reinventing the energy business. As I dug through last week's news, I asked myself: Is Bloom Energy's public relations blitz to be believed?
The short answer, it appears, is "yes." But don't expect miracles.
For people in the green-tech industry, on Bloom Energy has been like the fruitless quest for El Dorado. Finally, last week, Bloom Energy orchestrated a to the world, revealing its and naming a number of companies already using the Bloom Energy Server.
Many viewers of the 60 Minutes segement may have simply been surprised that Silicon Valley's movers and shakers are working so seriously on clean energy. Even for those in the know, though, Bloom Energy's technology is intriguing and brings fuel cells, which many people have dismissed as technically challenged, back into the discussion over the future of energy.
So how do their claims stack up? Based on company-provided materials, they look pretty good. But keep in mind, solid oxide fuel cells are a well known technology with understood benefits and limitations.
Fuel cell apps in action. Fuel cell technology was most recently on the tech media's radar last year when the Department of Energy car solution. Mobile fuel cells have formidable technical challenges, notably storing enough hydrogen for a long driving range, and the lack of a hydrogen distribution infrastructure. Bloom Energy CEO KR Sridar told reporters at least twice during last week's press conference that the technology was not developed for cars, an indication of how much people associate fuel cells with vehicles.for fuel-cell vehicles, which have been touted for many years as the ultimate green
Bloom Energy is using fuel cells for stationary power, which is arguably a better application for the technology. For starters, stationary fuel cells can use the natural gas lines already in place for fuel. The Bloom Energy Server can run on different fuel sources, including biogas, a gas made from organic materials. And they can be strung together, much the way servers are clustered to boost processing muscle. An initial customer, eBay, for example, is using a five 100 kilowatt boxes--each about the size of a parking space, to power 15 percent of its headquarters in San Jose, California.
For companies that need on-site power, fuel cells are already in use because they are very reliable. Fuel cells could power data centers, for example, because of their reliability and the potential to supply DC power directly to electronic equipment. Since they are relatively clean sources of power, fuel cells can receive state subsidies, as they do in California.
Bloom is not alone. In getting so much media attention, Bloom Energy certainly benefited from its connections to high-profile investors, such as John Doerr of famed venture capital company Kleiner, Perkins, Caulfied & Byers. Kleiner's connections probably helped line up initial customers for Bloom, including Google, Walmart, Staples, FedEx--all companies which have invested in alternative energy sources for financial and environmental reasons. But Bloom Energy is not the only company making fuel cells for stationary power.
FuelCell Energy, which is based in Danbury, Connecticut, is already selling fuel cell power systems for commercial customers, which fuel cells that can run range from 300 kilowatts to 2.8 megawatts. Another is start-up ClearEdge Power, which recently introduced a smaller fuel cell for homes or small businesses to make electricity and heat. Panasonic is developing , which also use natural gas to make both electricity and hot water.
Where Bloom Energy stands out is the design and materials it's using in its fuel cells, which offers the potential to lower costs with higher manufacturing volume. The core of Bloom Energy's technology is a solid oxide fuel cell, which takes fuels and oxygen from the air to make an electrical current.
One technical challenge with this type of fuel cell is that they operate at very high temperatures. That allows for greater efficiency in energy conversion but also requires engineers to deal with high heat. Bloom has designed the system to recycle the heat generated from the energy conversion in the process of mixing incoming natural gas with steam, which is needed with this type of fuel cell. So instead of using the heat to make hot water, for example, the heat is fed back into make electricity, according to the company.
Another significant technical achievement is that Bloom Energy's system doesn't use expensive materials, notably platinum which is used as a catalyst in many types of fuel cells. Bloom Energy is cagey on exactly what it uses but says that the fuel cells use a ceramic made from sand and inks. Researchers have been trying to make fuel cells without platinum for years. Another company trying to make ais SunCatalytix, a spin-off from the Massachusetts Institute of Technology, although it's a very different technical approach.
Cost. So why all the fuss over a well understood technology? Because Bloom Energy has said that it can deliver its electricity at between 8 cents and 10 cents per kilowatt-hour, including the cost of ongoing maintenance. In many parts of the country, that's cheaper that the grid rate. Because fuel cells are their own source of juice, they also offer back up power in the case that there is a grid outage.
According to Bloom Energy's data, companies which purchase this sort of system can earn back the initial outlay of between $700,000 and $800,000 for a 100 kilowatt system in three to five years. A 100-kilowatt system could be enough to power 10 U.S. homes or a small business, such as a Starbucks, according to the company. But keep in mind, that its stated cost per kilowatt assumes subsidies as high as 50 percent of the initial cost and natural gas prices of $7 per million BTUs, according to an interview. Still, Sridhar claims it can drive down the cost steadily and that it can compete without subsidies in the future.
Given that these fuel cell boxes require a hefty upfront cost, the most obvious customers for these mini-power stations are businesses and utilities, not consumers. Companies can afford to make relatively long-term energy purchasing decisions, particularly if there are economic incentives to buy cleaner sources of energy. Unless a company fixes into a long-term contract for gas, the cost of power from a Bloom Energy Server will vary with the cost of natural gas, which has been volatile in recent years, so fuel cells don't have the same predictability that a solar array would have. On the other hand, on-site solar panels cannot deliver electricity at all times of the day as a fuel cell can.
Bloom Energy executives said that utilities could be interested in purchasing relatively large-scale units to, perhaps, power a neighborhood. One advantage is that a utility would not need to build more transmission lines to meet the growing demand in a specific region. On the other hand, utilities are typically very conservative in adopting different technology. And although Bloom Energy has raised some $400 million, it's still a start-up in the eyes of risk-averse commercial customers.
Sridar told 60 Minutes that he would like Bloom to bring down the cost of a 10-kilowatt fuel cell to less than $3,000 in five years, which brings it into the buying range of homeowners. For somebody who needs reliable back up power or wants to promote cleaner energy technologies, that could be interesting, although that price has not been reached. Another open question for all customers is how reliable these systems will be over ten years of use.
How clean is it? Bloom Energy says that its fuel cells convert about 50 percent of the energy in incoming fuel into usable power, which is the equivalent of a 40 percent to 100 percent cut in carbon dioxide emissions compared to centralized power generation. Fuel cells, which operate without combustion, greatly reduce or eliminate smog-producing and other air pollutants as well. But IDC Energy Insights analyst Sam Jaffe points out natural gas power plants can operate near 60 percent efficiency, making the overall efficiency about the same as a Bloom box even after losses of energy in transmission lines are figured in.
That's not to say that Bloom Energy's claims over "clean energy" are bogus. About half of the U.S. gets is electricity from burning coal which produces more carbon emissions and has other environmental problems.
Bloom Energy has also indicated that it can make power using solar energy, which would make it a very clean source of power. During the press conference, Sridar said that the company has technology to convert solar-generated power into hydrogen, which could be used to run its fuel cells to make electricity. In this scenario, the Bloom Energy Server is not only a source of power but also an energy storage device. However, given the cost of the system, he expects the technology won't be commercially viable for eight to 10 years.
Given the technology and the fact that Bloom already has brand-name customers, you can see why it's been so hyped. Having raised hundreds of millions of dollars and not saying a word for eight years also helped build the mystique around what they are doing.
Although company executives say that the Bloom box won't need subsidies forever, the fact is that it does right now. That means its commercial success will rely on those incentives in the short term and companies willing to buy fuel cells because they are relatively clean and reliable sources of on-site power. Also, the company will need sales to scale up from serving a handful of customers to produce higher volumes.
At the very least, Bloom brings fuel cells back into the picture for powering buildings. Given the cost and my needs, I don't expect on installing one in my backyard any time soon. But then again, many people didn't think they'd need PCs at home 25 years ago either.