We are saying that if energy is the next Industrial Revolution, and if we are going to be competitive in this globally competitive world and we are falling behind right now—gosh, let's go for it. You need hubs for long-term problems, and you need ARPA–e to look for short-term translation of science into technology.
So what are some of the high-risk projects you are working on?
There's a project from Lawrence Berkeley National Laboratory where they've got a new catalyst for hydrogen production that is molybdenum-based as opposed to platinum-based—so it's cheaper. And that was proven in an inorganic setting. They are taking that catalyst and attaching it to a bug [a microbe]. I don't think they have any idea whether the catalyst will actually work when attached to the bug, because around a bug there are salt concentrations, other proteins and all kinds of things. You've got to just try it out. And that's supposed to produce the hydrogen that the bug will consume to produce electrofuels. Now that's risky. But there are slightly lower risk projects as well.
It's taken ethanol 30 years of subsidies to get to 10 percent of U.S. transportation fuel. How long before some of this high-risk research starts to deliver?
They all have different time scales, different industries, different supply chains, different channels of sales, different industrial costs, different economic factors, different regulatory hurdles. It's not going to happen in the next two or three years in terms of full global scale. It's likely to happen beyond 10 years, products actually in the hands of consumers or placed in the energy infrastructure. My most optimistic estimate is 10 years. But frankly, look at how long the Internet took—from 1968 and ARPANET to the late 1980s. That's 20 years. That's the sort of timescale we should be looking at. This is going to take some time.
Which programs can deliver fastest?
Electrofuels is really in an early stage. I don't think we should expect that anytime soon. I don't know, frankly, which ones will reach market first. Maybe power electronics [devices for regulating electricity supply] might. At the same time, power electronics have different power levels. There's consumer power electronics and consumer power supply, and that's very different from power electronics on the grid and transformers.
How does this help with climate change or other energy-related challenges?
It's environmental security. That's not just climate change, that's pollution…. We are doing things that will affect the climate in a positive way. If you can get solar electricity down at 5 cents per kilowatt-hour, and it scales without subsidies, gosh, I think that's pretty good for the climate. If you are doing carbon capture hopefully at a cost of $25 per ton of CO2, that could be good for the climate as well. And if you are creating electrofuels by grabbing CO2 and using CO2 as a feedstock, that could be good for the climate as well. But let's work on things we agree on, let's move. Otherwise it becomes a stalemate and nothing moves, whereas the rest of the world is moving fast.
What about failures? After all, this is high-risk research. How does that work?
They are not meeting the milestones. Something's not working. So our program directors are spending a lot of time on those to help them solve the problems. Sometimes it comes up to me, and I'm thinking—putting my science hat on—and I'm trying to think as to how to help these guys solve those problems. Our primary goal is to help them reach the milestones and move forward. But sometimes the technology just doesn't work for a variety of unforeseen reasons, because we don't know exactly what the issues are going to be. And so those red alerts, if they do not meet the milestones in the next two or three months—we give them a little window instead of just cutting them off—then we will have to terminate some projects. I would rather take that money and put it back in the treasury.