Daniel Lemire's blog

, 3 min read

Science and Technology links (January 25th 2020)

3 thoughts on “Science and Technology links (January 25th 2020)”

  1. RAD says:

    With all due respect, Daniel, your assessment of nuclear power generation is not rational. You, and others, are quite correct when you point out that the assumptions about the health impact of nuclear accidents have proven to be wrong. The cost equation, therefore, has shifted away from health costs as the key factor to just plain old cost per kilowatt-hour over the lifetime of the plant. Your province has famously implemented a hydro-electric centric strategy. My province (Ontario) has famously implemented a nuclear-power centric strategy that has proven to be safe (CANDU reactors), though finicky, but has proven to be uncompetitive cost-wise.

    What Three Mile Island, Chernobyl, and Fukushima has taught us is that the worst case nuclear accidents (core meltdown) cause very little health costs but that the likelihood of a core meltdown is orders of magnitude worst than any worst case estimates. Ontarian’s woke two Sundays ago to an emergency alert on our phones informing us of a problem at the Pickering nuclear plant. The alert was sent out in error but do a quick Fermi estimate of the cost of an accident that spreads low-grade radioactive material over a swath of dense urban land, like that around the GTA, and I think your opinion about nuclear power would change. This is the risk of all boiling water reactors until an alternative to zirconium clad fuel rods goes mainstream. They are prone to core meltdowns and it is not only due to communist incompetence. CANDU reactors have proven safe but at enormous capital cost.

    In an era of cheap natural gas, it is crazy to promote nuclear energy unless you place an irrational cost on greenhouse emissions. It is also irrational to shut-down nuclear plants prematurely but that is a different argument.

    1. CANDU reactors are decade-old technology. We can do much better today.

      1. RAD says:

        The same argument has been made about SQL databases being decades old. I don’t think you would make the same claim without a careful analysis of the underlying data structures; the kind you are famous for. Bang-for-the-Buck is the metric to use (hopefully minus the hydrogen gas bang) not health costs. I have seen a technical analysis that amounts to more than marketing rhetoric. Natural gas fired power generation is the system to beat, as far as I can tell. Nuclear power has a sweet-spot: submarines. Beyond that I’m not so sure. I’m not against nuclear power ideologically, just economically.