Stardate 20020925.2305

(Captain's log): Several people have written to me about something called biodiesel. It represents a process whereby things like waste cooking oil and excess animal fat from slaughterhouses can be converted into a fuel which can be burned in existing diesel engines. It evidently works now.

It doesn't help. The problem here is scale. There isn't a sufficiently large source from which to make this stuff so that it could actually produce a total quantity of energy per year large enough to even begin to offset our petroleum use.

It's actually a rather indirect form of solar energy. What's going on is that the two primary sources, animal fat and vegetable oil, both come out of agriculture. But unlike direct burning of biomass, as I described in the last article on this subject, it only represents a very small proportion of the total solar energy per acre which was captured by the original plants. (In that regard, animal fat is far worse. The animal consumes only part of the plant matter, and only converts a small part of that into fat.)

As a form of recycling it's probably a good way to get rid of that kind of stuff. (Certainly better then burying it or dumping it in a river to be carried out to sea.) But as a source of energy, it barely generates a blip on the scale.

There are a heck of a lot of ideas like this one. The problem with all of them isn't that they can't be made to work, it's that the amount of power (energy per time) they can provide us is several orders of magnitude too small to make any real difference if our goal is to significantly reduce our consumption of petroleum. What you find is that most of them can generate really substantial amounts of power in short bursts, but the average power generation is tiny on the scale we're discussing.

Small sources of energy are easy. Big sources of energy aren't, and the small sources of energy can't be scaled, and there aren't enough of them so that even when added up they would be enough to matter. One person asked if I had researched new sources of energy; I don't need to, because the physical reality is that there aren't all that many natural sources of energy which are large enough to make any difference, and the process of utilizing all of the large ones is reasonably straightforward, or else it's damned near impossible. There are probably all sorts of weird and obscure new energy technologies out there. Each may have its place; each might be able to solve some kind of problem. But not this problem.

For instance, we use solar cells now. The highway between here and Las Vegas is lined with emergency telephones, and each of them has a solar cell which trickle charges a storage battery to power the phone. But they're being used because they don't have to be wired to the power grid, not because the amount of energy they generate is very large.

You've got to think big. I've run into this before. Most non-engineers (and even a lot of engineers) don't actually have an intuitive understanding of large numbers. (That's why people play the lottery.) For most people, any number above about a thousand is the same size. I am, perhaps, exaggerating but only a bit; people know that a billion is larger than a million but don't really understand how much larger. Maybe the reason I have some idea of the kinds of scales is that as an engineer I'm used to dealing with really vastly different time scales. I'm used to dealing with things in nanoseconds, and over the years I've come to internalize just how small a nanosecond truly is. It can be demonstrated with factoids, but that's not the same. But let's give it a try:

KHz: A millisecond is to a second as a second is to 16 minutes and 40 seconds.
MHz: A microsecond is to a second as a second is to 11 days, 13 ¾ hours.
GHz: A nanosecond is to a second as a second is to 31 years and 8 months.

My new workstation's processors are running 2.4 GHz. One clock cycle to one second is as one second to 76 years. And when you're talking about energy use at the level of big industrialized nations, the range in the scale is even bigger. By the standards of power engineers, we're wimps.

When you're talking about energy for the kinds of stuff I was designing, 100 watts is pretty large. (When you're working with batteries, a hundred milliwatts is pretty large.) But when you're talking about power sources for major nations, you've got to think REALLY big. And there aren't a lot of actual sources of energy, real or speculative, which are at that kind of scale.

My dad was an electrical engineer and he worked on power generation. (He spent most of his career on the hydro projects on the Columbia river.) He lived in an entirely different world than I did, a world where units like kilofarads and kilohenries were actually useful. That's the kind of numbers you see when you're describing long distance transmission lines. In my world, a microfarad is huge. In his world, a farad was tiny. (If you don't know what that means, just let it pass.)

You've got to start thinking really, really big.

Anything which, when fully deployed, generates less than ten gigawatts average (10¹⁰ joules per second) is useless for our purposes in terms of actually making a meaningful contribution to the total amount of energy we consume. For scale purposes:

Hoover Dam = 1.5 gigawatts.
Grand Coulee Dam = 6.5 gigawatts
Small coal or nuclear plant = 300 megawatts
Large coal or nuclear plant = 1 gigawatt

Average US electrical power consumption = 400 gigawatts

Peak US electrical power consumption is probably above 1 terawatt

And that's just electricity. We also use a heck of a lot of fuel for vehicles, and burn fuel for heating and other purposes, and I don't just mean in cars and home furnaces. Do you have any idea how much coal a big steel plant uses? Refining steel, casting it, shaping and working it consumes truly mammoth amounts of energy, and it isn't coming from electricity. And you might be surprised by how much fuel the airline industry uses: Seventy five million gallons per day in the US alone. The railroads use huge amounts of diesel fuel, and so do all the ships which are moving bulk cargo worldwide and the tugboats moving barges up and down our rivers and all the trucks on our highways.

I don't have any specific reference for the total energy average that this nation consumes, but if it isn't a terawatt average, it will be soon. A million megawatts. You've got to start thinking really big. (Have I made the point yet that you've got to think big?)

[Update: Actually, it turns out that the US is using 3.29 terawatts. We were using more than a terawatt fifty years ago.]

If any proposed energy source can't be scaled up to generate 10 gigawatts average (1% of that), it won't be large enough to make any significant difference in the grand scheme of things even if it works and is really, really cool and clever and innovative and nifty.

Which is why windmills aren't interesting, for example. The Irish windmill project will, once completed, utilize every reasonable site in Ireland and will generate 500 megawatts when the wind is blowing. For us to be interested within the context of this discussion, an American windmill effort would need to be at least 20 times larger, and that's unlikely.

Things like biodiesel, or harvesting wave power with underwater flappers, or big floating devices in the ocean which utilize the temperature difference between the surface and the deep, just aren't in the ballpark. They can generate energy, but not enough. If biodiesel ever exceeds ten megawatts, I'd be surprised, and that's three orders of magnitude too small. While these things might well be practical ways to generate small amounts of energy, especially from waste products which are otherwise hard to dispose of, they're still miniscule overall. (I think what I'm trying to say is, please stop sending me letters suggesting other ways. There are no other ways which can scale large enough to matter.)

In terms of really large future sources of energy, I know of only four even theoretically, and none of them are practical now.

The first is called a "core tap". It's sort of like an artificially-created source of geothermal energy. You drill a hole somewhere between 10 and 30 kilometers deep, down to where a substantial amount of heat from the mantle becomes available, and then inject water into it and use the resulting steam to drive a turbine. Unlike geothermal, which relies on existing vulcanism, a core tap could be placed anywhere that the crust of the earth is acceptably thin, which on continents mainly means away from mountain ranges. The site also has to have access to a substantial source of non-salty water. Beyond the simple problem of drilling that deep, there are other technical issues involved, like making it so that the entire hole you've drilled can hold the resulting massive steam pressure without rupturing.

There is also the solar satellite. It's basically a big mirror in geostationary orbit, which uses the light of the sun to generate electricity which is converted to microwaves and beamed down to a fixed reception station on the ground. Because it's in freefall the structure can be amazingly flimsy, but it's also got to be huge and even with a flimsy structure it's going to involve a huge amount of mass. The most likely generation process is to focus the light on a boiler that runs a standard turbine, but the combination of actually putting something that big (more mass than everything the human race has ever orbited combined) and keeping it working (it will need a permanent staff) plus the problem of actually getting the energy down to earth is nontrivial. It won't be practical for decades.

Another is fusion, but I'm skeptical as to whether it will ever be possible to make that work with a capital cost per megawatt of capacity that's even remotely reasonable. Every existing proposal for fusion, not just tokamaks, involve truly amazing amounts of high tech equipment. Though the likelihood is that the power yield from these kinds of plants will likely be similar to existing coal or fission plants (assuming they can ever make them work at all), the equipment cost to make one may be stunning. Tokamaks involve mammoth cryogenically cooled magnets and world-class vacuum pumps, among other things. Laser or particle beam implosion requires a large number of very high power lasers or non-trivial particle accelerators (and world-class vacuum pumps). It's hard to see how such a plant could be built at any kind of reasonable price, not to mention how one avoids immense operating and repair costs. Acceptable MTTF's in a research facility won't cut it in an operational environment. Even assuming free fuel, the amortized capital cost and ongoing operating cost may well make the power generated unfeasibly expensive. I'm not willing to pay $10 per kilowatt-hour for electricity; the seven cents per KwH I'm paying now is already among the highest rates in the US.

The fourth and last future source I can envision is direct conversion of mass to energy, and some of my younger readers may live to see it. With the work in cosmology going on now, they're getting near to actually having an explanation of exactly what mass truly is and how the interconversion of mass and energy actually happens. It's more complicated than just particles appearing and disappearing. The energy released by fission and fusion doesn't come from a change in the number of particles; rather, the hadrons are changing weight, and the energy release comes from that. The reason that fusing hydrogen into helium releases energy is that the protons and neutrons in helium weigh slightly less than the ones in protium and deuterium, and the excess mass is released as energy (which is why the Sun shines). We know that's true, but no one knows why. No one can explain why it is that the hadrons in iron weigh less than for any other element, so that below that fusion releases energy and above that fission does. Why iron, instead of cobalt or carbon or gold? Why isn't it a single slope curve, so that fusion of everything would release energy and fission would always consume it? No one knows.

Once the theoreticians actually figure that out, it may turn out that there are ways totally unsuspected by us now to convert mass into energy that don't involve elaborate silliness like plasmas and toroidal magnetic fields and fissionable materials. What I'm talking about is a theory at the level of subatomic physics as comprehensive and important as quantum mechanics was at the level of atoms.

Our nuclear technology now is about like 19th century chemistry: we have a lot of recipes but we don't really know why they work. It took quantum theory to tell the chemists what they actually were doing, and once they had it they began to produce miracles that made 19th century chemistry look lame. Quantum mechanics also taught us how to make field effect transistors to replace vacuum tubes; a completely different approach to the same result which was vastly smaller, far more reliable, and far more efficient, dropping size and power and manufacturing costs by something like 10 orders of magnitude. Once the nuclear engineers have an equivalent theory and actually know what they're doing, they will almost certainly make all existing nuclear technology totally obsolete, and they may well figure out a straightforward way to produce energy directly from any mass. For example, it might become possible to create a system which took ordinary hydrogen, crashed the electrons into the protons to produce neutrons, and then annihilated the neutrons to produce quite large amounts of energy leaving behind only an ash of neutrinos (or antineutrinos; I can never remember which). Or it might turn out that there's an easy way to directly convert matter into antimatter, which is then a twofer in terms of energy production.

I'm not sure I actually want us to learn to do that kind of thing that efficiently. It also opens up the possibility of weapons which would make an H-bomb look like a firecracker. Once you've got direct conversion, you might conceivably be able to build weapons that could destroy planets. I'm worried enough about Saddam developing 15 kiloton fission bombs. What if some lunatic creates the first multigigaton weapon?

Regardless, of the four I actually think core taps are the most feasible. The problem is nontrivial, but it's also reasonably straightforward and doesn't require any breakthrough in theory (though it might require advances in materials science or laser technology). The process of digging the hole might well be slow but it might not be as expensive as all that, especially if it gets dug with a high power laser instead of a physical drill.

In the mean time, we're not going to substantially decrease our consumption of petroleum by converting waste animal fat into diesel fuel, or putting a million gerbils on treadmills, or by capturing the air turbulence caused by migrating birds, or using the light from fireflies, or anything like that. Irrespective of whether they can be made to work, they just won't generate enough energy to make any difference in terms of actually significantly reducing the amount of petroleum we consume, which is where this entire discussion started. Remember, the original question was whether we could send a message to our Arab friends by reducing the amount of their oil we buy. We would not only need truly substantial amounts of alternate energy, we also would need it soon and cheap.

If nifty obscure new energy source du jour can't reasonably produce at least twice as much power as Grand Coulee Dam, it isn't going to make any difference.

My new blog: Chizumatic