Tuesday, January 31, 2023

Fission Power

Evidence continues to mount that fossil fuel combustion is causing climate havoc. Floods and droughts, damage to cities and farms: it is becoming clear to more and more people that we need to wean ourselves off fossil fuels somehow. This is, however, an enormous challenge. We humans live very large on the earth nowadays, in our combinations of large populations and comfortable lifestyles. We consume energy globally at a rate of about 20 TeraWatts. We cook, heat our homes, drive our cars, run our factories... energy is fundamental to our modern way of life. Most of this energy comes from fossil fuels: coal, petroleum, and methane. To avoid enormous difficulties from any total change to our way of life, we need to substitute non-fossil sources to continue to provide energy at the required scale. Maybe in the future we will develop new sources, but in the next few decades at least we will need to rely on existing technology. Renewable sources such as solar, wind, and hydro are already in widespread use. Energy storage systems can help bridge the gap between fluctuating supply and fluctuating demand. But how to scale up renewable sources to meet the requirements of our modern society remains a daunting challenge. Nuclear fission is another existing technology that already provides steady reliable power at large scale. It is an very real option on the table for addressing climate change.

When your credit card bill is due and your checking account is empty, it is tempting to pay one credit card bill by borrowing from a different credit card. The general temptation is to solve short term problems by creating even larger long term problems. It's not an entirely invalid approach, but it's definitely smart to go down that road with eyes wide open. If we do choose to ramp nuclear power up by the factor of about 25x that would be needed to meet our energy needs, how might that move fit into a longer term strategy?

The long term strategy for modern society is rather cloudy but still worth considering. There is not going to be any kind of consensus possible, but that shouldn't stop a person from thinking about it. Some of the main options:

  • The world is due to end quite soon, so a long term strategy has no application.
  • We cannot have any idea about the future. Long term planning is an absurd pretense.
  • Technology will continue to advance at an ever more astounding pace. Any problems we create now will easily be fixed by the people of the future with their capabilities that will be almost miraculous by our present standards.
  • Maybe after a few thousand more years of expanding population and increasing comfort, humanity will start to bump up against actual planetary limits, but there is no point in worrying about that now.
  • We are clearly hitting real planetary limits already. But it takes time for us to shift our various systems, such as agriculture, to more sustainable patterns. We cannot continue to consume energy at today's rate, but we need a few decades to shift. The immediate dangers of climate change mean that we need to shift to non-fossil sources sooner than we can reduce our energy consumption. Nuclear power can provide a bridge from today's unsustainable way of living to a future sustainability.
It's worth thinking through what nuclear power would look like under these various scenarios. To ramp up nuclear power by 25x over the next decade or two is already a daunting prospect. If energy consumption continues to double every 50 years or so... what this would mean exactly in terms of uranium mining, waste management, fuel transport, etc. - I don't have answers, but it would be worth exploring such possibilities.

To flesh out such visions of how nuclear power could be scaled up in the future, perhaps the baseline assumption might be that everything goes according to plan. But effective engineering requires us to think about what might go wrong. If we are considering the option of walking down a tightrope to get to our destination, we'd be wise to understand how high off the ground that rope is!

Some of the unpleasant surprises worth considering:

  • Natural disasters such as earthquakes can cause radioactive material to escape containment.
  • Safe management of nuclear material can require a somewhat advanced level of industrial capabilities to make available the necessary equipment and materials. Even with scaled up nuclear power, other factors could cause our industrial capabilities to be significantly reduced.
  • All kinds of human bungling are not just possible but unavoidable. People are not perfect - not even close to perfect.
  • It's not just that people make mistakes. People will quite deliberately act to benefit themselves at whatever cost to others. It may be possible to build a very safe reactor, but it will cheaper to build one that is less safe.
  • People are always involved in conflicts at every scale. Nuclear technology can be weaponized in any number of ways. Of course we have very many nuclear explosive devices already built and ready for action. But the more we have fissile material circulating and the machinery for refining it etc., the easier it will be to build more explosive devices.

    Weaponization is not limited to nuclear explosives. Depleted uranium is already in widespread use in various types of bullets and other projectiles, just because of its metallurgical properties. Easy availability of radioactive materials will make them attractive for all sorts of uses. Various sorts of dirty bombs, conventional explosives coupled with radiactive shrapnel, are also straightforward possibilities. We have seen in the Ukraine where Russian troops occupied nuclear power facilities, because Ukrainian forces would not likely attack them there because of the risk of releasing radiative materials into the environment.

  • Nuclear technology can be a source of conflict. A nation might be developing nuclear technology for entirely peaceful purposes, but this unavoidably also increases their ability to build nuclear weapons. Their enemies will be motivated to attack and destroy their nuclear facilities, to cut off that nuclear capability.
It's also important to think about how we should evaluate consequences. We could just decide that it is too difficult to wean ourselves off fossil fuels, and just accept the ensuing climate change. We could cut our energy consumption dramatically to avoid climate change, and just accept the ensuing disruptions to our way of life. Or, if we decide to scale up nuclear power and some of the possible negative consequences arise, how bad could they be? Nowadays I see folks arguing that nuclear war wouldn't be so bad. Perhaps any cost short of human extinction should be considered acceptable. Even if ramping up nuclear power leads to human extinction... well, humans will surely go extinct sooner or later anyway, and if nuclear power improves our lives before that point, maybe it is a worthwhile bargain.

Understanding the various risks is very difficult. Many of the numbers involved are simply unknown, especially when the time scales involve many thousands of years. But there are also more complicated sources of uncertainty. Government inspectors will help prevent dangerous cost-cutting in nuclear facilities, but then government inspectors are themselves corruptible too. Nuclear advocates will point out that there have been no documented fatalities due to plutonium toxicity. But of course the people that handle plutonium employ many safety measures. Is plutonium safe because we know how dangerous it is? It's a bit like how the Mutually Assured Destruction provided by nuclear weapons has made the world a safer place, in some sense or other.

How can we decide what to do, in a game with such high stakes, with such high uncertainty, faced with such paradoxical logic? At least if we can get some common understanding of the predicament, that might be a start!

Tuesday, January 24, 2023

Steady Growth

There is a notion around that humanity requires steady growth to be healthy and happy. Steady growth clearly cannot continue for long on a finite planet. So there is another related notion around, that interplanetary colonization is required for humanity to be healthy and happy. Even the solar system is finite of course, so interstellar colonization is a natural next step. Why not intergalactic?!

But there are other physical limits that will constrain growth. Of course it could be that we will discover that our notions about physical limits are not accurate. But then our notions about the need for growth could be wrong, too. Any and all of our ideas could be wrong, but still, we're thinking beings; if we expect to succeed with interstellar colonization, we'd better hone the precision of our thinking!

One of the most fundamental physical limits in our theories today is the speed of light. Perhaps we'll find a way to colonize other galaxies, but it will take us a very long time to get to any of them!

Steady growth generally means exponential growth. Over a generation, the growth in whatever segment of the population will grow in proportion to the size of that segment. If health and happiness is to be equitably distributed, and if health and happiness requires growth, then growth will be exponential.

Physics comes in because humans, whatever else they might be, are also physical objects. The disciples of Ray Kurzweil might quibble: perhaps humans, in essence, are actually information. But even information requires some minimal physical substrate to be stored and processed! In any case, I am certainly not proposing that the specific numbers of my back-of-envelope calculations here should be taken with any seriousness. My point here is that steady growth will eventually bump up against the physical limit of the speed of light. I invite everyone to run the numbers as they see fit.

Suppose humanity's domain is some large sphere, centered on the earth presumably, and stretching out through interstellar space toward the distant galaxies. Since humanity is steadily growing, its domain is also growing. If humanity is growing exponentially, the volume of its domain will also be growing exponentially. Of course humanity can grow, to some extent, while in some fixed domain. That's what we've been doing on earth so far.

What exactly the carrying capacity of earth is, that's difficult to say. But, again, there are physical limits. The earth's mass is about 10^13 times the total mass of humanity. If the population grows at a steady 1% per year, then in about 3000 years, the total mass of humanity will exceed the total mass of the planet earth. Obviously we will run into serious trouble long before that; it is difficult to predict the exact course of our battle against limits to growth. The point of my quick calculations here is that they set some quite hard bounds. If humanity is to continue to grow at a steady 1%, certainly before 3000 years have gone by, we will need to be well down the road of interplanetary colonization.

It's easy to run similar numbers for the solar system. In less than 5000 years, the steadily growing mass of humanity will exceed the total mass of the solar system. Probably we will not find a way to digest the sun, so we will need to be colonizing distant stars well before then.

So let's say that we have spread out in the galaxy out to some radius R. If humanity is growing at 1% per year, the volume of its domain must also be growing at 1% per year, and then the radius will need to grow at 0.3% per year. Once that radius hits 300 light years, that steady growth will require the radius to grow more than one light year per year, i.e. faster than the speed of light!

So a reasonable bound on steady growth of 1% per year is that the domain of humanity will hit a hard physical limit at radius 300 light years. That's a volume of about 3 x 10^61 cm. Given the rough density of galactic matter, the total mass in that volume would be about 3 x 10^40 grams. A human weighs about 10^5 grams, so that would be a maximum population of about 3 x 10^35... assuming humans have incorporated all material into their bodies! Today's population is about 10^10, so that's a population growth of a factor of 3 x 10^25. At a steady 1% growth rate, we'll hit the speed of light in about 6000 years.

Of course these rough calculations involve many very unrealistic assumptions. There is no way that humanity will absorb into their bodies the entire mass of galactic matter inside a sphere of radius 300 light years. But even if they could, we'd hit the speed of light in 6000 years, given a steady 1% growth rate. 6000 years is already not an absurdly long time - it's roughly our historical horizon. Absurdly generous assumptions about the success of humanity's battle against the limits to growth already run into limits that are not absurdly far away.