Why We Should Phase Out Nukes
Posted on July 10, 2017
Nuclear power has always been too dangerous to mess with, but the risk just took an upward quantum leap.
The cyberattack on Kansas’s Wolf Creek reactor and other reactors has failed (so far), but that is not especially reassuring.
FACT 1: using existing computer science, no computer in industrial use anywhere can be made acceptably safe from cyberattack.
FACT 2: no existing nuclear power plant is acceptably safe from a major nuclear accident, even ignoring cybernetic risks.
FACT 3: a successful attack on a nuclear power operating system could trigger a major nuclear accident by simply crashing its computer control systems and backups. No special operating knowledge would be needed.
FACT 4: nuclear energy cannot make a significant dent in the global warming problem.
FACT 5: causing a nuclear meltdown would be the holy grail for terrorists.
FACT 6: the expertise needed to hack a nuclear site with some likelihood of success is not limited to state-level actors.
Evidence on Fact 1:
–Even an isolated system is subject to attacks that cross the air gaps, using already published techniques. One critical link is that even airgapped systems must occasionally accept new code that has partly outside origins.
–The main risk is most system is human-engineering that exploits mistakes made by technicians. That risk can be reduced but not eliminated.
–Granted that perfect security is impossible and too high a standard. What standard should we set?
I’m guessing most folks who live near a reactor would say a 0.1% risk of a meltdown is far too high. Yet 99.9% safety against hacking does not look possible at present.
Evidence on Fact 2:
–There is a long track record of incidents, some them terrible, with no hard upper bound on future casualties and property loss.
–Nuclear engineers cannot be trusted on safety claims. They have a long track record of failed assurances, and they have an institutional bias.
–Existing nuclear power generates plutonium as a side-effect. Plutonium itself constitutes a major social risk we do not need (e.g. for building terrorist bombs).
Evidence on Fact 3:
–While there is talk of failsafe reactors that shut themselves down as the natural result of any failures, no one claims that is what we have now.
Evidence on Fact 4:
— Building new nukes has a startup time in the decades. That won’t save us.
–Existing nukes are a relatively small share of the energy base.
–Absent a carbon tax we are in deep trouble. With a carbon tax and the kind of rapid investment we need in renewables, nukes are practically irrelevant.
Evidence on Fact 5:
Are you kidding me?
–Terrorists desperately want an event significantly bigger than 9/11. The possibilities are few: break a major dam, crash the electric grid, cause a nuclear meltdown, build a nuclear bomb.
–I’m guessing that causing a reactor incident is technically the easiest. It is cool even if the casualties are low, and the possible upward payoffs are immense.
Evidence on Fact 6:
–In several cases, secure military sites have been hacked by private thrill seekers.