Civilian nuclear power: A safe but expensive necessity for one more generation

What comes to mind when you hear the words "nuclear power"?

Do you think of technological progress? Radioactive horrors? Virtually unlimited cheap energy? Virtually unlimited horrifically expensive energy? Cutting-edge engineering? Dangerously outdated junk?

I think we can all admit that the first couple of generations of nuclear power weren't quite what the marketing teams had made them out to be. Virtually unlimited, virtually free electricity turned out to have a rather hefty price tag attached. Cutting-edge engineering suffered more than a few compromises in the name of cost-cutting. And, despite the early promises of impeccable safety, a few high-profile accidents have raised hell among the media and the general public.

It's important, though, to keep everything in perspective.

In the history of civilian nuclear power, there have been only three accidents classed as "serious" or "major" (INES levels 6 or 7). They are:

  • Chernobyl, Ukraine (USSR), 1986. A graphite-moderated reactor, of a type that was known to have dangerous design flaws, had been built without a proper containment building. The plant operators shut down several backup systems before trying, and repeatedly failing, to perform an emergency core cooling test. With the reactor now in a highly unstable state, they went ahead and did the test anyway, causing a steam explosion and fire that blew the lid off the reactor and scattered 5200 PBq worth of its contents over three thousand square kilometres of Ukranian countryside. 64 people died; credible estimates of eventual cancer deaths that may be attributable to Chernobyl range from 4,000 to 27,000 (although Greenpeace claims this figure could be as high as 200,000).
  • Kyshtym, USSR, 1957. A cooling system failure at a nuclear waste processing facility led to a chemical explosion in a storage tank, releasing roughly 800 PBq of assorted nuclear wastes. A few dozen to a few hundred deaths are believed to be linked to this accident and to previous radiation releases from this shoddily run, notoriously spill-prone site. It's debatable whether this even qualifies as "civilian", given that the site's main purposes were military.
  • Fukushima, Japan, 2011. A complex of early-generation boiling water reactors had been built in a location known for centuries to be vulnerable to tsunamis, and had had its operating licence extended well beyond its intended design life. The plant was hit by a beyond-design-basis earthquake and tsunami, which it survived, but grid power and emergency backup generators were lost. By the time the cooling water supply was re-established, four reactor cores had melted down; the total radiation release is not yet known but will likely be around a thousand petabequerels. Only two deaths were, or will be, directly attributable to the accident at the plant; in addition, there may be a roughly 1% increase in lifetime cancer risk (mostly thyroid cancer, which is almost always completely curable) among infants who lived within 30 km of the site.

Pretty much all other nuclear accidents in history have been either military in nature, or sufficiently well contained to have no appreciable effect outside of the facility in question. The famous Three Mile Island accident of 1979 has yet to yield a single radiation-related injury or illness, and statistically speaking, probably never will.

What virtually all nuclear power accidents – major, serious or otherwise – have in common is remarkably clear:

  • Poor planning, whether in the original design or in the disaster response plans, for contingencies which were clearly foreseeable.
  • Cutting corners on safety-critical systems and procedures. Had the operators at Chernobyl waited a day or two for their core to stabilize, rather than forging ahead with the test when conditions were clearly contrary to the procedure, the reactor would not have blown up. (And, had there been an actual containment building on that reactor, the accident would not have been nearly so devastating.) Had the owners of the Fukushima facility read the centuries-old markers on the hill above the plant that said something like "It can flood up to here on occasion", they wouldn't have put the plant's only backup power supply at sea level.
  • Someone doing something stupid, like just about everything that happened in the control room before the Chernobyl accident, or the Fukushima plant's owners refusing to allow seawater to be used for emergency cooling just in case they might still have a chance of using the damaged reactor again.
  • Early generation technology that was missing many, if not most, of the fail-safes and backup systems that more modern designs have.

Should we reject nuclear power, then, based on its safety record? No. The record speaks for itself; in terms of deaths per terawatt hour generated, civilian nuclear is safer than anything else we've ever created. And virtually all of those deaths are from one piece of Soviet junk that should never have made it off the drawing board. Nuclear beats hydroelectric dams (they kill a few people with flash floods every year, and they occasionally take out whole towns when they break.) And it's even safer than solar photovoltaics (people fall off the roof while installing them).

Don't think for a moment, though, that I'm a cheerleader for the nuclear industry. I'm not. I pay taxes and hydro bills here in Ontario, and I hate the thought of paying for these things.

Nuclear technology's Achilles heel is its capital cost. Doing it properly and safely is expensive. There's no way around that, unless you offload all the risks of the project onto the general public – and that just won't fly these days.

If we don't build new nuclear plants to replace our existing old ones, though, we'll be in a bit of a pickle. Fossil fuels are out of the question as far as anyone who plans to live past mid-century is concerned, and we've already dammed almost every major river we can. Replacing the 3.5 GW nuclear plant at Darlington, Ontario with renewables would mean covering almost a hundred square kilometres – or 20% of Toronto – completely in solar cells, plus eight pumped storage centres the size of the Marmora project to keep things running when the sun's not shining. This one mid-size nuclear plant, which runs nearly 24/7, produces as much electricity in a year as 10% of all the solar panels in the world.

We'll get to a renewables-dominated power system someday, with more efficient (and more plentiful) storage, a smarter grid, and more efficient solar cells. We're not there yet, though, and without a concerted international effort, it'll likely be several decades before we are. And so I think we're going to have to rely on one more generation of nuclear fission reactors for baseload generation; there is simply nothing else that can take their place when the current ones wear out in the next few years.




Nuclear power

I submit for discussion the technology known as LFTR. Liquid Fluoride Thorium Reactor.
Seems to answer many of the existing objections to nuclear.

Add new comment