Fine Structure

Recently Bloom Energy has been getting a lot of attention for their fuel cell technology they officially launched Wednesday morning and the whole pitch seems very magical at first glance. With a little digging and an understanding of what to look for, I think we can find out just how magical this technology is and what the potential pitfalls are.

Fuel cell technology isn't new by any means. It was originally discovered in the 1830s (yes, the 1830s) though the modern fuel cell is a little more recent (like the 1950s) but limitations have forced it into a niche as essentially a toy for scientists to play with beyond a couple ultra-specialized applications like the Gemini and Apollo programs.

Scientists and engineers love fuel cells because they can have decent efficiencies, generally above 50%. The internal combustion engine, for reference, sits around 15% efficient. The reason they're so limited in use is that their components can be very expensive. Rare and valuable materials, such as platinum and palladium, are used in the production of certain types of fuel cells and this drives up the cost of producing the cell in the first place.

Typically you hear about fuel cells as a powerplant for cars and they generally run on hydrogen because pressurized hydrogen is portable (though heavy - the containers at least). These fuel cells all work the same way; take in hydrogen in one end, insert gaseous oxygen in another end and with the right mixture of cathode, anode and electrolytic materials, you can produce power and emit only water.

Bloom's fuel cell is a particular type of cell called a solid oxide fuel cell (SOFC) which isn't a new technology in itself and is useful for its lack of expensive materials used in production. Not only does the solid oxide process require heat but it produces huge amounts of heat (upwards of 1000˚C), sustaining its own fuel cycle. In all the mystery surrounding the Bloom fuel cell, perhaps the greatest space for innovation is either a SOFC that operates at a lower maximum temperature or materials that better withstand the high temperatures.

The biggest source of mystery in Bloom's product is the green and black "ink" (yes, even they use the "quotes") that form the cathode and anode of the fuel cell. If Bloom has really invented a cheap way to manufacture the cathode and anode portions of the fuel cell, they only have one more hurdle to clear - lifetimes. Bloom doesn't mention the lifetime for their product very prominently but apparently it's around 10 years. That's primarily for the enclosure equipment as I've heard that the fuel cells themselves have to be replaced twice during the lifetime of the product. That's about a three year lifespan for the cells, a pretty decent span of time but also not without waste. Presumably the price of the replacement cells and labor to switch them out is built into the $700-$800k that you would pay for a commercial box.

Between interviews and press coverage, a lot of things have been mentioned about Bloom, the most misleading, I think, are the idea that it can "run on solar" and that it produces "no emissions". "No emissions" could be one of two scenarios, the first being a "run on solar" option as well. Since this type of solid oxide fuel cell can be run on hydrogen alone, a potential configuration is a solar array producing energy that is directly used to electrolyze hydrogen. The hydrogen is then pumped through the fuel cell and produces electricity with oxygen through the process mentioned earlier. The only emission in this case is water. The second is probably more likely with natural or biogas being pumped through the cell. The gas isn't used in the same way that combustion uses it so there aren't any "emissions" so to speak, but there is still used gas to dispose of or refine so it can be used again.

Considering that solar panels aren't generally enough to power an entire house, could they really be enough to generate enough hydrogen to power a house via fuel cells? Probably not. More conversion means more losses to the inefficiencies of systems and going from sunlight to electricity to hydrogen to the fuel cell and back to electricity doesn't seem like it's going to be more efficient than solar alone. Can it potentially use hydrogen as a natural energy storage container, reducing or removing the need for expensive batteries? It's definitely possible, but if hydrogen via solar was all that efficient to begin with, we'd probably be producing more of it already.

It sounds like Bloom is a little more realistic now but there are a lot of mysteries still in play. The cost of manufacturing could be much larger than suggested, Bloom could be burning through their hundreds of millions of dollars in startup funding because the boxes are more expensive to produce then they sell them for. Keep in mind the fuel cell replacement costs twice over ten years. Maybe the cost of developing the inks used in the fuel cell releases lots of carbon dioxide despite the fact that the box itself could produce "no emissions". Or perhaps the fuel cells can't be adequately recycled afterwards and they just build up in some landfill somewhere.

We really won't know the implications until the technology has been on the market for longer and more details about the inks are released. No doubt, this will all come with time as more and more people sign up to get Bloom boxes installed in their backyards.

Everyone who publishes anything (even if they don't know it) loves being mentioned on other websites. The longest time ago when I wrote a personal blog for friends about whatever came across my mind we had link sharing; you link to my blog and I'll link to yours. The practice amongst friends or readers conveyed a sense of trust in the content of the other site which was picked up on by Google and became the backbone of PageRank. So, forward ten years into the future, spammers try to increase their rank on searches by having others link to their site (or just creating lots of sites themselves and linking them all together).

People still love to be mentioned on the web and it surely says something for your own personal brand when other people see you mentioned on the web. Calling someone out on your blog is a good way to get their attention and maybe even a link back to your site. Add in some psychology that says that if you do something nice for someone else, they feel compelled to do something nice for you and you have an exploitable situation. Enter bad guys. They're thinking that if you linked to enough people from some semi-spammy site, you could get enough people to link back to you to gain some search cred.

Okay, enough theory. There's a site out there called Accredited Online Colleges (I won't link to them because I don't trust them) and they're essentially a front for another site that pushes online degree programs¹. There's lots of repetitive links and they're clearly optimized for convincing search robots that this is the place to go for an online college search. Unfortunately, they have no valuable content.

This is where our ideas behind linking to others come in. They do have a blog where they (you guessed it) link to twitterers and others sites on bombastic lists such as "100 Amazing Scientists You Should Follow On Twitter".

Yes, I am listed as an "Amazing Scientist" on that list. No, I don't think that it's a sincere appreciation of anyone on twitter.

The blog is designed to get people to link to it because people love being linked to. And it works! The bit.ly aggregate link has been clicked more than a thousand times. From all these links, they get a better search position and make more cash referring search users to the actual appropriate site. It's the classic scenario of stacking the space between manufacturer and consumer with middlemen that drive up the price (or, in this case, the clicks). It's not OK because it pollutes the web.

The scientists on twitter make for a really interesting community of people and lots of great chatting and trading of ideas goes on there. @flyingjenny is a really fascinating look into how NASA works on the inside, even moreso now that we know the fate of the Shuttle. @BadAstronomer is, well, the Bad Astronomer and you probably need no more reason than that to follow him on twitter. People that you could never even hope to communicate with on a personal level are out there, responding to tweets from regular people (@neiltyson, @garrettlisi and @seanmcarroll come to mind).

So do yourself a favor and check that link before they you link back to it... especially if it's just a passing mention of you. We're not link whores yet.

¹ Do I universally condemn online degree programs? Of course not, I'm railing against the method which this site uses to gain links and popularity, not the content itself.