Do Bloom Fuel Cells Really Get $8.25 a Watt in Subsidies?

Rumors also swirl that Bloom raised another $100 million. With half a billion dollars raised, this is going to have to be a big IPO or acquisition.

On the heels of Bloom's announcement yesterday that it would offer the Bloom Box in a no-money-down service contract structure comes news that Bloom has raised $100M in venture capital from Kleiner Perkins, NEA, and Morgan Stanley, according to Dow Jones VentureWire as reported by Xconomy. Last year, PEHub reported that Bloom was looking for $50 million. The new funding might exist to expand production -- Bloom says it currently only makes around one fuel cell a day -- or it could be to provide financing for their new fuel-cells-as-a-service sales model.

The company has not commented on the new funding.

More interestingly, we examined the company's subsidy situation in greater depth with Sam Jaffe of IDC. Bloom customers (or Bloom itself when the company is setting up power plants on its own) gets a 30 percent tax credit on equipment. California also gives a $2.50 per watt subsidy if the Bloom Servers burn natural gas, and $4.50 if that gas comes from a renewable source like landfills on projects up to 1 megawatt. Adobe, which has the biggest Bloom installation at the moment, uses biogas.

A source told us recently that Bloom Servers prior to subsides sell for $12.50 a watt: $10 a watt for the box and $2.50 for the warranty. A 30 percent tax credit comes to $3.75 a watt. Add the $4.50 tax credit from California and the total comes to $8.25 a watt. "Whew!" exclaim solar and wind executives. If you don't count the warranty and switch to a fuel cell that runs on regular methane, the subsidy drops to $5.50 a watt, or more reasonable but still somewhat cozy.

Subsidies drop further as projects expand in size. Projects in the 1 to 2 megawatt range get $2.25 (and $1.25 per megawatt in California) and get halved again for 2 to 3 megawatt projects.

Bloom Servers individually create 100 kilowatts of power. Adobe, however, installed 1.2 megawatts worth of them and Caltech plans to put in 2 megawatts, one megawatt at a time. (How these will get counted in the California program is beyond us.) If someone erected a 3-megawatt Bloom farm and ran it on natural gas and the warranty didn't qualify for a federal credit, the subsidy might drop to $3.63 a watt.

Still, some clever accountant is likely playing the angles to maximize the subsidy. A feed-in tariff exists too, but most use the fuel cells to generate local power for their own operations.

These numbers underscore how the company remains locked in a race to reduce costs. The California program has a finite amount of money in it. Jaffe estimates that Bloom might be getting gas at 7 cents a kilowatt hour for its fuel-cell-as-a-service program. The deal could accelerate demand by eliminating the barriers to resistance -- it's a new technology and these things cost a lot -- but further improvements will be needed.

Our article on the fuel-cell-as-a-service program from yesterday follows:

****

Bloom Energy is taking a page out of the solar playbook and starting to market its solid oxide fuel cells as a service.

Under the Bloom Electrons program, Bloom installs its Bloom Energy Servers on a customer's premises and then sells the power to its customer the same way a utility would. Bloom buys the natural gas to run the Energy Servers, conducts maintenance, and takes care of all of the tax credits and permits. (The Energy Servers convert natural gas to electricity via ceramic plates coated with zirconium oxide inks.)

The customer, meanwhile, sees its power bills decline by 5 percent to 20 percent. The customer, perhaps more importantly, doesn't have to pony up millions of dollars of capital. A 100-kilowatt Energy Server costs $700,000 to $800,000, including federal and state incentives. (Without incentives, the price jumps to $1 million or more.)

It also avoids the risk that the fuel cells might start to become less efficient over their ten-year run or require more repairs than anticipated, a valid concern in the new and unknown world of solid oxide fuel cells.

"All they do is lower their bills," said Asim Hussain, director of product marketing.

Can the company make money off this? It's hard to say, but rising power prices, low gas prices and declining prices for manufacturing these boxes will certainly help. In California, commercial customers pay 14 cents a kilowatt hour for conventional electricity, he said. Apply a 5 to 20 percent discount to that and Bloom is likely selling power to these customers at 13.3 to 11.2 cents a kilowatt hour.

Customers that are buying Bloom Servers say the boxes generate power for around 7 to 10 cents a kilowatt hour. Adobe specifically said that its boxes generate power at 8.5 cents per kilowatt hour (down from the usual 13 cents a kilowatt hour) and Adobe feeds it more expensive biogas. These numbers include the 30 percent federal tax credit and California state incentives of around $2.50 a watt for regular gas and a whopping $4.50 a watt for biogas.

Bloom thus has a potential margin of around 3 to 6 cents a kilowatt hour, including the incentives. There are worse ways to make a living. Note, however, that California has some of the highest rates in the country and many states do not have the same local-level incentives.

Nationwide, commercial/industrial power on average sells for 10.55 cents a kilowatt hour, according to the Energy Information Administration, and it sells for under 10 cents a kilowatt hour on average in the Southeast, Mountain West and in the Plains states. In those regions, Bloom likely isn't economical just yet.

Thus, don't expect to see a lot of these in Minot soon. Bloom for now, in fact, will concentrate on California, Hussain said. 

"The long-term challenge for Bloom is in bringing down the cost of the fuel cell as the subsidy subsides," said Sam Jaffe, research manager for renewable and distributed energy strategies at IDC Energy Insights. Bloom, he added, likely pays around 7 to 8 cents per kilowatt hour for its gas, if the gas comes under long-term contracts for $7 per million BTUs and the fuel cell is 50 percent efficient.

Will the concept take off? If solar is any indication, the answer is yes. Sungevity, the retail solar installer, says that its customer base has almost completely converted from solar buyers to solar-as-a-service customers within the space of a year. The business has also grown tremendously. SolarCity and others report similar results.

Bloom is currently in the process of installing a 2-megawatt fuel cell facility on the Caltech campus under the Electrons program, said Hussain. Since each Energy Server generates 100 kilowatts, the Caltech facility will have 20 boxes, making it the largest installation in the world. Companies and institutions participating in the Bloom Electrons program include Kaiser Permanente, Walmart and Coca-Cola.

"Nobody wants to spend $800,000 to say they produce their own electricity," Jaffe added.

Hussain added that production at Bloom is up to one Bloom Server a day, up from one per month two years ago.

For now, Bloom will market its servers and services to provide baseline power. Conceivably, the boxes could serve as storage devices or provide jolts of peak power. Utilities are currently scrambling for energy storage and grid balancing technologies. Some companies such as Xtreme Power are installing battery banks at wind farms that do double duty as grid-balancing assets. Like batteries and unlike traditional peaker plants, the Energy Servers can also provide power quickly.

It's an interesting idea, Hussain noted, but maybe for a later day. The company is also keeping an eye on ways to couple its fuel cells with electric vehicle charging stations. Some flow battery companies are looking at EV charging too. By installing a fuel cell, flow battery or other storage device, the fears about EVs spiking peak power consumption can decline. Cars charge from the storage device and the storage device gets recharged overnight.