Flagship Ventures, VantagePoint Venture Partners, the Massachusetts Technology Development Corporation, and an Oman-based industrial conglomerate called Omzest Group.
Ze-gen’s overall story is an illuminating one, not just for other clean energy startups, but for technology startups in general, many of which run into the same general types of problems that Davis has had to tackle over the past two years. Here’s my take on a few of the lessons:
“Invention” is not a discrete event—it’s a process that stretches out over years, and often involves sharp turns.
When I visited Ze-gen in 2007, the centerpiece of the demonstration plant was the company’s prototype gasification furnace, which would later be half-filled with molten iron. Under the company’s original design, waste was fed into the middle of the iron bath through a tube that was partially submerged in the liquid metal. This tube also supported a rotating coil that constantly stirred the iron to speed gasification.
The company quickly found that this system didn’t work. The torque on the feed tube from the stirring, along with the drastic thermal gradient between the bottom of the feed tube and the top, inevitably caused it to weaken and break off after a few minutes or hours of operation. (During my tour this week I got a look at an old photo of the remnants of one feed tube, floating atop the molten iron as partially melted slag.) After each failure, the tube would have to be replaced, causing great expense and delay.
“Our first feed tube cost $12,000 and lasted about 12 minutes, and four months later they cost $700 and lasted 7 hours, so there was progress made,” Davis recounts. “But ultimately we decided that having a feed tube was not the way to go. We needed to get the feedstock into the gasifier without the use of feed tubes.”
That led to what Davis calls “a major engineering modification”—essentially, a complete redesign of the the furnace. In the new design, the tube that brings feedstock into the furnace has no stirring function; in fact, it ends just above the surface of the copper bath, meaning it acts mainly as a shroud, carrying the feedstock to the zone of highest temperature before it gasifies. Also, Ze-gen switched from molten iron to molten copper, which has a lower melting point, and therefore doesn’t congeal on the tube and the inside walls of the furnace when it splashes (a major problem with the first prototype).
All of this reengineering cost Ze-gen some time, but it wasn’t necessarily a surprise, Davis says. “When you start off, you know what you want to create, and you have an idea of how you want to get there, but we also knew there were a lot of things we would have to figure out along the way,” he says. “We were confident that we could assemble a team to do that, and that if we hit a roadblock, we’d find a workaround. That is what we have been able to do.”
Scaling things up is rarely a smooth process.
There’s little way that Ze-gen could have anticipated the problems with the original feed tube and the splashing iron bath without actually building a barn-sized prototype, filling it up with molten metal, and trying it out.
“There are some technologies where you could build a version that fits on a table, and be reasonably confident that you could replicate it at a much larger scale,” Davis says. “But when you get into stuff like this, you can’t prove the chemistry in a coffee pot. If you want to prove commercial viability, you’ve got to be meaningfully scaled to do that.”
Ze-gen’s next step—after it has collected enough data from continuous operation of the demonstration plant—will be to
