Futurists really never learn, do they?

[jducoeur]

As I desperately attempt to catch up on LJ (I seem to be perpetually four days behind at the moment), I came across jikharra's reference to this story. Most of which is interesting and reasonable, but I was quite struck by this paragraph:

If all goes well with the experimental reactor, officials hope to set up a demonstration power plant in Cadarache around 2040. Officials project that as much as 20 per cent of the world's energy could come from fusion by the century's end, said Raymond L. Orbach, the U.S. Department of Energy's office of science director.

"Century's end"? Oh, come on, guys -- that's 95 years out! Predicting subtle changes a century off never works. I am reminded of those articles from the 1940s about computers, and how by the year 2000 there would be literally dozens of these things, serving all the world's computing needs.

Honestly, 20 percent sounds completely silly to me as an estimate for century's end. I would instead project two likelier scenarios:

• The technology works, and proves to be economically viable. In this case, it accounts for something more like 75% of the world's energy needs. If they have a demo reactor really working by 2040, then 60 years is plenty of time for it to supplant traditional fuel sources, most of which are in decline due to overexploitation.


• The technology doesn't work, or doesn't make economic sense. In which case it accounts for essentially 0% of world energy. (And in 2100, everyone is still saying that fusion power is just a few decades off.)

It's far from certain, of course, but this looks like a disruptive technology to me. If it ever really works, it's going to change the landscape rapidly and dramatically, displacing existing tech over the course of just a few decades...

Comments

OT: How was dinner?

[new-man.livejournal.com]

Where'd you end up? And will we see ya'll on the Esplanade today?

Re: OT: How was dinner?

[jducoeur]

As we started driving, msmemory pointed out that we hadn't been to Shilla in an eon, and that was our favorite susherie for many years. So we headed over there, and it was just as good as we remembered it. Not only was the yellowtail and unagi exceptionally tender, even the squid was tooth-perfect. And their Spicy Tuna appetizer is still the best anywhere. So that proved to be the right choice.

As for Esplanade: yep, we're expecting to show up sometime around 5ish...

[cvirtue.livejournal.com]

Thank you for posting the fusion vs. fishes article link! Most fascinating.

(Long ago some article I read said that some techs call "fission," "fishes" to be, well, an in-joke. I've found it useful to help me remember which one we actually use at present. Too many SF books have fogged my memory!)

[corwyn-ap.livejournal.com]

Agreed. Estimates for the crossover point for oil (i.e. when demand exceeds supply) vary from 2005 to 2050. Certainly by 2100, we won't be burning much of it in electric plants. I am unsure whether the production of coal can pick up the slack (presumably, deconstructing mountains isn't technically difficult). New nuclear fission plants are going to be a tough sale no matter how bad things get (they are a tempting terrorist target if nothing else). Hydro just doesn't have much room to expand. Wind has plenty of room, but will have trouble getting a large percentage. Solar can reasonably pick up household requirements, but industrial will be tough (even with cheap cells the required acreage gets problematic (unless we put them in space of course, but that requires much cheaper launch vehicles)).

Next question is consumption. The increase in consumption of energy over the past century was I don't know how many orders of magnitude. Are we expecting similar jumps in this century?

I suspect the 20% is taken from a usage viewpoint. In other words, planes aren't going to have a fusion reactor on-board (Mr. Fusion not withstanding). Batteries will need to improve markedly to achieve are reasonable W/kg ratio. This is just lack of imagination in my opinion. We are NOT going to be burning oil in planes in 2100.

[learnedax.livejournal.com]

I suspect the 20% is taken from a usage viewpoint. In other words, planes aren't going to have a fusion reactor on-board (Mr. Fusion not withstanding).

I don't know, 95 years is a long time. Something like Pyroelectric Fusion (but with actual power-generation potential) might well be developed which allowed small-scale fusion. Even the 35 years until this project is supposed to be up and running has the potential for major shifts in the scientific landscape. By then we might have pebble bed reactors viable for use in large vehicles, or even consumer vehicles although that seems unlikely with this particular technology. Besides, there's little reason to work on that when you can use fuel cells in smaller power-consumption applications.

[umbran.livejournal.com]

...then 60 years is plenty of time for it to supplant traditional fuel sources, most of which are in decline due to overexploitation.

Upon what do you base the statement that traditional fuel sources are "in decline"? Let's be clear on something - at no time in history have we had greater output from our traditional fossil fuel sources. Not a single one of them has gone into a decline of output.

People have been saying that the end is nigh for fossil fuels for decades. And for decades, they've been wrong. Technology has so far more than kept pace with our need to find and retrieve fuel.

We are not in danger of running out of fossil fuels. Even fields that we consider "dry" are still loaded with hydrocarbons we can't extract right now. We may be in danger of running out of oil and gas we can reach economically. But, at this point, nobody has a solid idea of when we might cross the line. We can only estimate what might happen based upon current technology and knowledge. But our technology and knowledge haven't been static in the least.

That's the root of the actual problem - the trend has been that we always have the right technology to get more fuel. So, we wind up with two camps - those who are comfortable and see no reason to change policy, and alarmists who say we're doomed. Nobody is willing to take the middle road, and support a gradual shift in policy.

[jducoeur]

Fair point (and heaven knows, one the Economist makes regularly). But the trends do make me nervous. It's so *easy* to continue to push for ever-greater exploitation, so long as the tech keeps making it easy to exploit. The problem, of course, is that this tends to breed acceleration: despite well-intentioned measures like Kyoto, world consumption of fossil fuels continues to skyrocket. And with China developing rapidly, that's only likely to increase.

And do bear in mind that we *are* hitting some bumps in the road. There was a major debacle in the oil industry last year, as several of the major oil producers revealed that their "proven reserves" were nothing more than loose and hopeful speculations. IIRC, Shell in particular got hammered, as it was revealed that their reserves were *much* lower than they had been claiming. So some of the numbers that are bandied about to demonstrate that we continue to have plenty of fossil fuels have to be considered a bit suspect: the petro companies have *considerable* incentive to lie, due to stock market pressures. (Since stock price correlates pretty directly to reserves.)

That said: yes, it ultimately comes down to economics. The crux of my point is that I don't think a middle ground is likely with fusion, at least in the long run. My *suspicion* (admittedly ungrounded in any sort of fact) is that the economics of fusion will prove to be qualitatively different from those of fossil fuels -- either simply not feasible (that is, the cost of infrastructure plus production never gets down to reasonable levels), or it winds up cheap enough after amortization that it renders fossil fuel an economically poor choice in most circumstances.

Of course, the history of fission shows that this question is a difficult one. Fission is cheap on the hypothetical level, but the political and waste-disposal costs have proven much higher than anyone ever expected. And it's worth noting that it has played out very differently in different countries, precisely related to how those countries have dealt with those costs.

So my argument is mainly an economic one. I don't expect a slow-and-steady transition, precisely for economic reasons. I think fusion will be a tiny niche market until the amortization passes a critical point, and then I expect it to take over *rapidly*, precisely because the cost of production of a unit of energy is likely to be significantly lower. I don't see a gradual switchover: I see a transition from 2% to 60% in the course of 20 years. *When* those 20 years happen is an uncertain combination of economics and technology...

[umbran.livejournal.com]

The crux of my point is that I don't think a middle ground is likely with fusion, at least in the long run.

There's a middle ground on development, which is clearly the more immediate issue. Honestly, economics will handle adoption if and when the time comes. Right now we have to worry about how we sell spending on research, rather than deployment. But talk about replacing everything quickly scares people.

But even leaving that out for the moment, I think you are highly underestimating the capital costs required for a changover. Power plants are massively expensive things, and they take years to build, even if there aren't massive regulatory issues. We'd expect fusion plants to be even more complicated, and thus more costly and time consuming to build. One doesn't change a major nation's energy infrastructure overnight.