Nuclear power plants are complicated systems. But energy startups don't have to be. Photo of a nuclear control room mockup: John Grabowski/Flickr
Picture a classic software startup. Two people sit in some dark room, faces lit by their computer screens. From these humble beginnings, the world can change, we tell ourselves. And it has been true often enough to become a part of Silicon Valley’s mythology.
Now picture a nuclear reactor. Hundreds of people build it. Dozens operate it. Whole walls are covered with gauges. If you could build one nowadays, it might cost you $5 or $10 billion, which is a lot of domain names and processor cycles at EC2. A lean little startup does not seem like the way to reinvent the nuclear reactor, or any of the rest of the massive, centralized energy system.
Yet that may be exactly what happens in the next few years. That’s because energy problems are, in some important ways, software problems. And the companies that get built to solve them might be as lean, capitalist and competitive as any Peter Thiel investment.
There have long been energy startups. California’s first solar rush occurred in the early decades of the 20th century when people realized that if you left water under glass out in the sun, it could get really hot.
An entrepreneur named William Bailey improved on the original Climax design and opened up shop as the Day and Night Solar Hot Water Heaters. Business was good for years before the natural gas boom in the state took away the cost advantages of those early heaters.
Later, around World War II, a small, star-studded group including Vannevar Bush, FDR’s science advisor, banded together to build a wind turbine 10 times larger than any before it.
“The wind turbine is notable as the physical result of a project conceived and carried through by free enterprisers,” Bush wrote, “who were willing to accept the risks involved in exploring the frontiers of knowledge, in the hope of ultimate financial gain.”
But throughout the 1900s, power plants kept getting bigger and bigger. For a long while, that scale led to electricity price drops, too, wiping out most small-timers.
The problem that had to be solved was more power, not less. More plants, not fewer. More coal, not less. For that problem, what you really needed, or more precisely, what people thought they needed, was more huge plants.
There wasn’t a lot of room for the startup in that world. And utilities — for some good reasons — were heavily regulated and averse to the kinds of risk presented by working with small companies. Not only that, some high-profile energy innovators from the 1990s, like Enron and wind-power company Kenetech, failed pretty spectacularly.
We face a very different set of problems now. Building power plants is expensive, it turns out, and in any case, we want to reduce the amount of energy we use, both because it makes us vulnerable to fuel-supply price fluctuations and because of climate change.
On the other hand, we have a whole new set of tools and ideas since we were last trying to make structural changes to our energy system in the late 1970s.
For one, we’ve got unbelievable and easily accessible computing power. We’ve also got the ultrafast and wide communications and organizing platform of the internet. And lastly, we’ve got an increasing amount of data about what’s going on in our electric grid, and we’ll only be getting more as smart-grid investments continue.
So, you’ve got a newish set of problems and a newish set of tools. Despite the up and down scribbles in green-technology venture funding, oil prices and belief in climate change, there’s a huge and long-term opportunity to create a startup that uses data and the internet to change the way the energy system works.
What kind of startup? Let’s look at few signals from what I think the future may look like.
Story continues…
Physorg | 
Google | Email this | Comments
