Bloom Energy, fuel cells or fool’s cells?
By now everybody and their mother must have heard about Bloom Energy and their ceramic fuel cell gizmo. I heard about it last week as I hammered down a few beers, ahh, I mean, as I was doing some interdepartmental team-building with several engineers who work at various bay-area companies. They were a little skeptical, as engineers tend to be. We are the ones that have to implement the marketing fever dreams. That is never as easy as it sounds My pals pointed out the huge costs for the Bloom boxes, how it would take hundreds of thousands of dollars to run a house. They mentioned that the box takes a long time to warm up, so it could not just be switched on for a quick energy boost. Then they mentioned that the 800 degree operating temperature meant that it could not be that efficient, and one problems with that is once you did switch it on and got the whole mess up to operating temperature then you would sure want to use it as long as you could.
The Bloom box is a classic case of a scientific curiosity being pawned off as an engineering triumph. Science simply has to work. Engineering has to work and be cheaper than other things that work. There are existing companies doing fuel cells– one using osmotic membranes right here in Menlo Park. And there is even an Australian outfit that uses the ceramic cells.
It shows how much hype counts in today’s world– just get the useful idiots at 60 Minutes to cover it, and all of a sudden, you are the “next big thing”. Just in time for the company going public. And believe me, everything about this hype is related to the fact that this is a venture-funded startup that needs to do a “pump-and-dump”. They want to get the hype all crazy, so when they go public the stockholders are left with the minor detail about getting this thing to work and the VC run off with all the money.
It may be contempt for manufacturing engineers, or it may be wishful thinking, or it may be just profound incompetence, but management, especially financial and scientific management, never understands that getting to high-volume production is no small feat. Just look at NanoSolar, who has continually done all these grand pronouncements about their world-changing cigs solar panels, only to never really deliver them in volume at a cost that makes money. I have a pal in the solar industry startup and he says Nanosolar has poisoned the well for all the solar panel startups since they did so much hype for so long so many times with so little to back it up.
All these huckster visionaries that go into energy “revolutions” lack any appreciation for high-volume manufacturing. One of the million problems that caused my buddy’s solar company to suffer was that the slitting shear that cut the panel would get all the cigs goo on it and that would run the edge of the panel. It may sound stupid to you because it is such an easy-to-understand problem, not like bandgap physics or phonon conversions. But believe me, those easy-to-understand problems keep you from selling panels just the same, something all the scientists have absolutely no respect for. They think that any idiot should be able to make their brilliant idea come to fruition but it is just the opposite. There are a million things that work great in the lab but don’t pan out in high-volume manufacturing. Like my mechanical engineer buddy told a founder of one of the failed startups I have worked at—“Sure it is an easy thing to fix, but it is one easy thing after another easy thing after a thousand other easy things, and then five years has past.”
One “easy thing” I am familiar with is plastic cars. Now I think that Amory Lovins is right on the mark when he tells us to conserve and use our resources wisely. But he is a blithering idiot when he goes on national TV and tells a reporter that carbon fiber body panels will make cars more efficient and would be the savior of humanity if it weren’t for those big dumb old car companies. Sorry, I am a former auto engineer and we looked at plastic cars for, oh since the 1953 fiberglass Corvette. I was also pals with guys that worked on the Pontiac Fiero, a real plastic car fiasco. The primary problem with plastic body panels is cycle time. It takes way too long to hold them in the mold and let them cool off enough so you can open the mold and get it out (carefully) and then start the next mold cycle. One of the beautiful things about sheet metal is how fast you can make it—bang bang bang, out come hoods and fenders and rocker panels. Long cycle times means that you are tying up the capital expense of the entire machine waiting for the plastic to cool off. But even if plastic did not have a profound cycle-time penalty, there are dozens of other reasons why it does not make sense for high-volume automobiles.
- It is mostly made from oil. Duh, I thought that is what we are trying to save.
- It does not conduct electricity, so you have to add ground wires and do all kinds of added cost to prevent EMI and RFI.
- It disintegrates on impact, killing the drivers. Just ask the family of Anton Senna the Formula 1 one driver killed by “second impact”. Once the first impact crash turns the entire body of the car into a fiber bag, there is no structure left to protect the occupants.
- Carbon, kevlar and aramid fiber are very expensive.
- It is hard to paint plastic.
- It is hard to repair plastic
- It is hard to attach any fixture points to plastic, or it costs a fortune to mold in nuts and attachment points.
- Plastic is not dimensionally stable, that is why the Fiero was made in a giant fixture that machined plastic mounting pads so the body panels might align.
- The surface finish of plastic is far inferior to sheet metal, unless you spend a fortune to fill and sand it.
- All the oil and processing makes plastic a rather bad choice relative to the amount of juju CO2 that we have suddenly decided is a poison.
- The price of plastic follows the wild variations in oil prices, so there would have to be a spot price on the car you are selling
- Plastic molds cost more than sheet metal dies, they have cooling channels and slides and all kinds of other expensive features that cost way more than the average stamping die. So when you drag out the cycle time to let the plastic harden, you’re making idle a million-dollar press and a million-dollar mold.
- Since the carbon fiber or aramid panel is made with a thermoset epoxy and not a thermoplastic, you can’t grind up the scrap and re-use it, it is a pure loss, and you have to find a place to dump it.
- Handling, cleaning and storage of thermoset plastics are a major hassle.
No, sorry all you brainiacs, Henry Ford the 1st was right when he said “Plastic is for combs and toothbrushes”. Sheet metal is the ideal material for automobile bodies. Get used to it. Heck we have hardly adopted aluminum, that is how good modern steel auto bodies are.
Now certainly, there are a ton of brilliant things coming from startups. Google, eBay and Amazon come to mind. But remember how the Ginger was going to change the world? At least until we learned the Segway was a scooter that had inherent usage problems in the real world. When something is hyped as hard as the Bloom box I have to feel it is because the backers realize there is no real intrinsic immediate value and they are looking to get the next batch of suckers pulled in so they can bail out.
And when you see some VC-funded startup hype how their lab experiment is going to save the world, you need to take a quick course in snake-oil sales techniques. First watch a Youtube video of the Monorail Song from the Simpson’s. Then watch the movie Boiler Room, so you understand the tactics. Then watch American Psycho so you understand the soul of people willing to take your money and leave you with worthless paper (kind of like the Fed only less official).
Meanwhile, if you see an energy startup staffed with former engineers from GM and Kellogs and Dell who really understand high-volume manufacturing, well invest in that one, not one full of PhDs. I have never met a PhD that could comprehend the problems of high-volume manufacturing and configuration management and I don’t expect to in this lifetime.
[Sept, 2011: Critical update]: To understand sales-type hucksterism, I just found out about the movie Glengarry Glen Ross. You should watch this after Boiler Room and before American Psycho.
Sir Scamalot commented:
Andy T commented:
Critical thinker commented:
wave_man commented:
esmifra commented:
Emrah Unal commented:
formula1 commented:
David commented:
William Ketel commented:
settingitstraight commented:
harikumar commented:
Andy T commented:
Greenhouse editor commented:
Friend of USA commented:
whoa commented:
