[tptacek]

I feel like you're not getting his point. The coal industry in its steady state reliably kills more people every year than every nuclear accident put together, even when you include imputed civilian cancer fatalities from Chernobyl.

I think you are going to have a difficult time coming up with an evidence-based argument that (a) supplies electricity to everyone in the world currently depending on it, (b) doesn't use nuclear, and (c) kills fewer people as a result of (b). I am ready to be surprised, though.

[dredmorbius]

Coal and petroleum are responsible for raising the population of earth from the 800 millon or so in 1800 to over 7.3 billion alive today. Nuclear alone wouldn't be capable of that.

Nuclear at the scale of present fossil fuel energy provision would be on the order of 15,000 plants, simultaneously, with a lifetime of about 40 years. There are fewer than 400 nuclear power plants operating today. We'd be looking at commissioning nearly as many per year (15,00 plants, 40 year life, 1.03 per day, or 375 per year).

Each of which would be creating at least some long-term waste.

There's the prospect of advanced reactor designs, with thorium being the darling of some, despite little actual experience and significant technical challenges (glowing hot highly corrosive radioactive salts, one test reactor run briefly 50 years ago for which cleanup is still not complete).

For uranium or plutonium fuel cycles, there's a very real concern over total fuel availability.

And even with nuclear you don't have liquid fuels without a heck of a lot of trouble. Some form of synfuel seems too be the best bet, with hydrogen from electrolysis combined with carbon from... Well, that's tough, limestone would still be carbon-positive, carbon recovery from the atmosphere or seawater is posssible but one heck of a challenge at scale. Ships and planes have few options other than hydrocarbons, and a lot of ground uses favour it.

Solar, wind, hydro, and some form of storage pencil out for raw scale, though my general sense between energy and other resource constraints is that a high-energy, abundant future on Earth will require a vastly smaller population. Likely achieved relatively quickly.

Or you could go the low-energy, non-abundant lifestyle. Which would likely see a similar population reduction.

Bit of a Hobson's choice there, in terms of misery.

Which do you choose?

[saulrh]

Now this is interesting. The really good power sources really cant be achieved unless you already have a bunch of power. Nuclear, for example, absolutely requires purification industry, which both directly requires energy and indirectly requires a large industrial base. About the simplest you can get is concentrating solar, and getting useful quantities of power out of those still requires industrial-scale fabrication.

Fossil fuels, in contrast, need almost literally nothing. You can build a decent coal-fired steam engine on an anvil and evolutions of that same coal-fired steam engine stay relevant well into the information age. Fossil fuels are a spectacular boostrap technology.

But! Fossil fuels are only good at bootstrapping. We're on a forum for startups and venture capitalists - we know better than anybody about the prototype-MVP-refactor cycle. Once you've launched your MVP, you need to turn around and deal with all the technical debt you've built up before it overwhelms you. There are a ton of products out there that share zero common code with their MVP. Bootstrapping technologies tend to be awful options for long-term use and you want to get away from them as soon as you've finished bootstrapping.

So here's what I say: build a prototype. Fossil fuels. MVP: build fast, lean on the cheap easy fossil fuels. And, finally, stabilize on your Once you've got enough basic infrastructure, use it to figure out a better long-term option and switch as fast as you can. Even if it's not as good immediately, it'll improve with time, and in the worst case (alien invasion?) you can bring the coal plants out of mothballs for a few years while you industrialize again. I won't say that it's easy, but I can easily imagine a less-intellectually-crippled species pulling it off.

[dredmorbius]

Fossil fuels are seed capital, yes. Modern civilisation has quite the burn rate.

I like to consider the technological stack height for various energy systems. For nuclear, as you note, it's quite high. One thing this creates is systemic risk of sheer system failure and inability to reboot. One of the consequences of the Fukushima disaster and following shutdown of much of Japan's nuclear power sector was that the entire nation faced a shortfall of electrical capacity. Not helped, incidentally, by its dual incompatible power grids, one operating at 60hz, the other at 50hz, limiting inter-tie capabilities.

While fossil fuels have a fairly low tech stack requirement, it's not zero. For a long time, plant-based sources of equivalents: olive oil and fuelwood, principally, were preferable. They could be obtained locally, processed readily, and renewed themselves.

For local use, all you need is a coal face or oil seep. Gas, without industrial means to capture, transport, and contain it is quite problematic, though the Chinese actually did just this, nearly 2,000 years ago, using bamboo pipes and hide bladders[1]. I hate to ask what the fire incidence was....

Oil didn't take off until an existing coal-based industry provided iron, and later, steel, for drilling, storage, pipes, railroads, and ships. This industry developed incrementally from roughly 1860 to 1900, though even _quite_ early in the process, oil equipment took on much of its modern appearance.

The other thing oil wanted for was a practical application: the internal combustion engine, which required: roads, tires, rubber, steel, electricity....

Another challenge is that fossil fuels hugely distort prices. We price fossil fuels at extraction effort, but the long-term cost is the replacement cost. Since replacing fossil fuels with more fossil fuels isn't viable (we don't have 100-500 million years to wait), that generally means existing biomass. Humans appropriate roughly 40% of that now, and it produces at net efficiencies of 1-3% of incident sunlight converted to fuel (algae might boost that to 10%), which has highly specific land, nutrient, (generally fresh) water, and climate requirements.

Nuclear seems to offer a possible out, but for any fission-based process is also* a nonrenewable option. Estimates of fuel abundance range from 80 years to a few thousand, at present rates of consumption. Relying on uranium for all present human energy needs would exhaust reserves in 6 years. Breeders offer a 100x improvement in fuel utilisation, but that pushes us from 6 years to 600 (without growth), or 80 to 8,000 at present rates -- a goodly interval, but only slightly longer than our current history. Suggestions of recovery of uranium from seawater might extend lifetimes further, but with another expansion of the the tech stack requirement. And there's the near-term immediacy of the problem to boot.

I'm leaning to the side that says that reducing our overall throughput (smaller population, decrreased per-capita affluence, possibly both) is going to have to be part of the solution. That's going to be unpopular. Possibly thermodynamically improbable without massive systemic changes, possibly catastrophic.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Notes:

1. http://www.computersmiths.com/chineseinvention/natgas.htm

[idlewords]

Not everything is a startup, bro.

[13years]

No, I do get it. I will restate my point again. "My point is about the disregard of the significance of the Fukishima accident".

Why is that important. Because nuclear is not a panacea of risk free energy. The public should demand technology and safety to be continually improved and outdated reactors to be revamped or decommissioned.

If Chernobyl possibly killed 100k, then Fukishima will likely in the end kill far more. http://www.greenpeace.org/international/en/news/features/che...

This is not a plea to abandon nuclear, it is that in order to prevent future disasters there has to be an honest assessment of potential risk impact so that there is incentive to make better designs.

Otherwise, all of these counter arguments could be interpreted as saying "Coal kills 1 million, so what if nuclear only kills 100k every once and a while".

[tedunangst]

"Coal kills 1 million, so what if nuclear only kills 100k every once and a while".

That sounds like a pretty compelling argument for nuclear to me.

[saulrh]

> Otherwise, all of these counter arguments could be interpreted as saying "Coal kills 1 million, so what if nuclear only kills 100k every once and a while".

Well, yes. We are, unfortunately, stuck in a pretty awful region of the configuration space of possible existences. We can't be perfect. And that, unfortunately, means that perfect is the enemy of good enough. Right now we only have a few choices either we stick with coal or oil and kill a lot of people and maybe everybody, or we switch to nuclear/solar/whatever and only kill a few people. I acknowledge that nuclear kills people. I fully understand that Chernobyl and Fukushima were catastrophes. And I'm not going to disregard them. I've thought about them carefullyh, and considered the options, and even if I accept the most pessimistic evaluations I come to a single conclusion: We need to switch. Now.

I can give you credit for wanting to be sure. But there's a point where you have to stop wanting to be sure and start making choices. And it's well past the point where we needed to make that choice, and as a result the conversation needs to be different. Right now our biggest obstacle is the public, and that means we need to present a united front. We can't tolerate dissenting opinions on a topic where the decision is so critical and the public is so pathetic. When you say "don't disregard Fukushima", people hear "radioactive scientist-men will eat your babies". So stop fearmongering.

Also, stop bullshitting about Chernobyl. "The total global collective dose from Chernobyl was earlier estimated by UNSCEAR in 1988 to be "600,000 man Sv, equivalent on average to 21 additional days of world exposure to natural background radiation."". And that's an early, pessimistic estimate that's since been revised to half that. Maybe half a million people total were exposed in any measurable way. The few thousand people who worked on the reactor itself lost maybe ten years each. Everybody else lost... a few hours? Fukushima, similarly, is estimated based on data from Chernobyl (which, remember, happened thirty years ago! we have data from it! lots of data!) to kill a few hundred people. Pessimistically, a few thousand. Indirectly, ten years before they'd have died of old age anyway, and with a relatively treatable cancer. Seriously, thyroid cancer? Five-year survival rate 98%, and that's if you aren't watching for it already and you catch it late.

And, really, greenpeace? They haven't been reasonable about nuclear since ever. If they ever realized how radioactive their bananas were their heads would explode.