The question “is nuclear safe” comes up regularly in discussions.  It’s a loaded question that is much harder to answer than it may first appear.  Say yeah….it is safe….. and you set yourself up to be proved wrong with references to Chernobyl and Fukushima.  But if you say anything that accepts that it is not totally completely utterly safe it opens the door to attack from any anti-nuclear protagonists and will probably alienate colleagues who want to hear you say clearly that it is safe.  It’s a rock and a hard place.

The problem, once again, is not a technical one. It is semantics. “Safe” is one of those words that is not very well defined so that it is used differently in different circumstances and means different things to different people.  Dictionary definitions of the word safe tend to be absolutist and include

“free from harm or risk”   (Merriam-Webster)

“not in danger”  (Cambridge Dictionary)

“If a person or thing is safe from something, they cannot be harmed or damaged by it”. (Collins)

Safety in these definitions is black and white; something must pose no risk whatsoever in order to be safe. 

But in everyday language we don’t use the word safe to mean that at all.  Safe in common language is analogue with issues of relativeness.   My mother used to think my passion for rugby was unsafe, I considered it safe because it was much safer than the other things I di which I did not tell her about.

Dictionaries often also have these grey area alternatives.  The Collins dictionary, as well as saying “cannot be harmed or damaged”, has a definition of a safe place as one “where it is unlikely that any harm, damage, or unpleasant things will happen” contradicting the original absolutist definition. 

To add to the problem, safety, like risk, includes two issues that cannot be compared with each other, magnitude and frequency.  Something that happens frequently and has a significant adverse consequence is clearly considered unsafe, but it may be considered safe if either the frequency or the consequence decreases even if the other remains unchanged.

Because people (incorrectly) see nuclear as something different and because people are scared by things that are different, the public perception is that nuclear must be “absolutely safe” in order to be described as “safe” even though they do not expect anything else they describe as safe to be absolutely safe.  The field is quite substantially tilted against nuclear power as a result.

This absolutism also appears in oft quoted anti-nuclear memes such as “there is no safe level of radiation”.  When an absolutist approach to the Linear No-Threshold model  is combined with an absolutist approach the meaning of the word safe then this idea that there is no safe level of radiation can make sense and that is why it has got so much traction even though it is an obvious nonsense.

So how do we answer the questions “is nuclear safe” and is there no “safe level of radiation” as the meme tries to suggest. 

The industry “go to” for the safe level of radiation issue is to question LNT either by suggesting there is a threshold or by invoking the concept of hormesis.  This, in my experience, is a complete waste of time, as the LNT issue has not been definitively resolved and probably never will be.  You just end up having to agree to disagree. 

Actually, this approach may even have the unhelpful effect of confirming any risk no matter how small makes something unsafe.

I have a different approach that is much more compelling.  I just ask the questioner whether they feel safe from radiation now, as they speak to me.  If they say yes, I simply point out that they have confirmed there is a safe level of radiation, it is the level of radiation they presently live with.  I have not defined it, they have.  It is then possible to talk about whether they would feel safe from radiation living in Winnipeg or Denver and point out that they would be exposed to much higher levels of radiation than they are now if they did live there.   They then have to agree that the level of radiation they consider safe is a lot higher than the one they currently experience (unless they happen to live in Winnipeg or Denver). 

The generic “is nuclear safe” question is harder.  One of the things I consider important is to avoid getting sucked into satisfying the apparent demand by doubling down on the absolutism with statements like “it’s absolutely safe” or “it’s totally safe”.  This sets you up for failure and removes the one degree of freedom that is needed to win the argument.

Descending into absolutism hands the anti-nukes the higher ground by once again accepting that absolute safety is possible and should be expected.  It also creates an opportunity for them to undermine trust in what is being said by listing the nuclear incidents I mentioned earlier and others that may be specific to the circumstances.  

I also try to avoid saying that it is safe because anti-nukes will exploit the ambiguity and seize on the absolute interpretations of the word.  Instead, I try to refocus the discussion onto what people mean by the word and ask the questioner what they mean when they say safe.  This moves them away from their initial absolute expectations into a more rational comparative analysis.  It reminds them of their normal expectation of the word and their normal expectation of the way the world works.   It forces them to think about how the safety of nuclear power compares with other things.  It allows them to come to their own conclusion without me having to convince them of anything.

When I am no longer on the back foot and have leveled up the playing field it becomes possible to talk about the unfortunate stuff we wish had never happened such as Chernobyl in a rational way.  Chernobyl, even in the most pessimistic of the reality-based estimates, gave rise to less loss of life than the original Zaporizhzhia dam break or that of the Banqiao or Shimatan dams and actually locates itself in amongst many other industrial accidents.

In turn this opens up the potential for a discussion about the data on loss of life arising from various forms of power generation.  This data shows nuclear to be better than most including things the public believe are completely benign like wind and solar.  Note however that this data does not “prove nuclear is safer” as many people will say.  Safety, as previously mentioned, includes an assessment of risk, so that something can have done no harm in the past but could have the potential to do harm in the future.  What has happened in the past is important evidence about safety, but it is not proof of safety.

This approach prevents me from being led into a fruitless discussion of whether something is safe or not, when no one has established what is meant by the word “safe”.  Instead, it enables a rational, defined, discussion about the real issues and that enables the person you are having the discussion with to come to their own conclusion.

Don’t get me wrong, this approach will not persuade the ardent anti-nuke.  Nothing will do that.  In fact, as they feel the ability to argue based on ambiguity being pulled out from underneath them, they will probably become increasingly aggressive and after punching air for a bit will default to the fallacious arguments I wrote about in my last blog.  And that’s when I start listing the fallacies they are using.