The New and Improved CTL Scheme
“In theory, there is no difference between theory and practice. But, in practice, there is.” – Jan L. A. van de Snepscheut
A very short paper (two pages) appeared in the latest Science detailing an improvement on coal-to-liquids (CTL) technology. A couple of people have e-mailed me to ask for my take on it. You can get the executive summary from the quote at the top, or from my version, which is:
“In the world of energy, people sometimes have trouble distinguishing make-believe from reality.” – Me
Wired Magazine weighed in on the (subscription-only) Science report – Producing Transportation Fuels with Less Work – a few days ago:
If oil prices rise again, adoption of the new coal-to-liquid technology, reported this week in Science, could undercut adoption of electric vehicles or next-generation biofuels. And that’s bad news for the fight against climate change.
The new process could cut the energy cost of producing the fuel by 20 percent just by rejiggering the intermediate chemical steps, said co-author Ben Glasser of the University of the Witwatersrand in Johannesburg, South Africa. But coal-derived fuel could produce as much as twice as much CO2 as traditional petroleum fuels and at best will emit at least as much of the greenhouse gas.
“The bottom line is that there’s one fatal flaw in their proposed process from a climate protection standpoint,” Pushker Karecha of NASA’s Goddard Institute for Space Studies wrote in an e-mail to Wired.com. “It would allow liquid fuel CO2 emissions to continue increasing indefinitely.”
The Wired story has spread like wildfire. I have seen numerous references to it. It has spawned lots of debate over whether it is desirable to increase our usage of coal. But one important detail seems to have been overlooked: It’s not real.
So what’s the story? Here is a clue, from early in the paper:
We outline reaction chemistry and processing designs that could dramatically reduce these energy inputs and minimize the amount of CO2 emissions that would be emitted or mitigated by other costly strategies, such as carbon capture and sequestration.
This is a ‘process’ only in that it has been drawn up on paper. This isn’t even at the stage of lab scale. This is Step 1: You have an idea. Next comes the step where you try to economically evaluate the implications, and then if they are favorable you start doing some laboratory experiments. This is the stage where probably upwards of 90% of all ideas that were thought to be good fail because the theory overlooked something that turns out to be a problem in the lab.
However, if the lab studies look good, you go on to build a pilot plant to demonstrate the concept at a larger scale. Again, the majority of the ideas that make it past the lab stage get weeded out at this stage. You have situations like CWT building a pilot plant only to find out that they have an odor problem that didn’t seem all that bad in the lab. Or you find out that while the front end of the process works well in the lab, and while the back end works well, when the front is connected to the back there is a problem. Maybe the front end produces a trace impurity that is a real problem for the catalyst in the back end. In the lab, this wasn’t a problem because you were simulating the back end with gases from a cylinder. However, those gases were of a higher purity than what you are producing now in the pilot plant. The list of land mines is endless.
None of this suggests that this is an idea without merit. But an idea is not a process, and so Wired got a little carried away with their description. In fact, the Science paper heavily references Sustainable fuel for the transportation sector (PDF) by Rakesh Agrawal et al. at Purdue. Theirs is another proposed process. It is called the H2CAR ‘process,’ and is another idea in need of some laboratory testing. For a layman’s explanation of H2CAR process, see:
For a more detailed technical (and skeptical) view see:
In that essay Engineer-Poet (of The Ergoshpere) concludes that this idea is doomed to die once the economics are considered fully.
But the bottom-line here is that it is important to distinguish between what is real and what is make-believe. A process that has survived lab tests or that is successfully being piloted is real. A process that only exists on paper is make-believe, and should be recognized as such. That’s not to say that the make-believe process won’t some day turn into a real process, but the lab has an uncanny ability to kill a lot of seemingly great ideas.