The Open Thread 21 has passed 500 comments and is getting a little bloated, so time for a new one.
The Open Thread is a general discussion forum, where you can talk about whatever you like — there is nothing ‘off topic’ here — within reason. So get up on your soap box! The standard commenting rules of courtesy apply, and at the very least your chat should relate to the general content of this blog.
The sort of things that belong on this thread include general enquiries, soapbox philosophy, meandering trains of argument that move dynamically from one point of contention to another, and so on — as long as the comments adhere to the broad BNC themes of sustainable energy, climate change mitigation and policy, energy security, climate impacts, etc.
You can also find this thread by clicking on the Open Thread category on the cascading menu under the “Home” tab.
There was quite a bit of discussion in the previous OT on radiation levels and the Fukushima evacuation zone. Relevant to this is the recent announcement that Japan will lift the entry ban on some cities within the prefecture. To quote:
In areas where annual radiation measurements are below 20 millisieverts per year, a government safety guideline, residents will have free access to their homes during the day and will be allowed to return permanently at the earliest opportunity post-decontamination. Where readings are between 20 to 50 millisieverts annually, evacuees will also have unrestricted access during the day although their permanent return will come later. In areas where measurements top 50 millisieverts, residents will not have free access and they will not be allowed to return for a minimum of five years.
A past BNC guest poster, engineer Chris Uhlik, analysed the situation a private email distribution list, and I thought his summary with respect to LNT (linear no-threshold hypothesis of radiation damage to living organisms) was very useful. With Chris’ permission, I reproduce it below:
The official position of every regulatory agency & scientific body, and even the people who will tell you “we don’t know what’s going on under 50 mSv”, the weight of the evidence favors LNT.
Here’s what I think is going on:
Under 50mSv/year we can’t find any epidemiological data to support LNT. There is simply too much noise and other effects to see sub-0.5% changes in cancer rates in populations where the variations from other effects (smoking, stress, chemical exposures, etc) are in the range of 20–45%.
The rates of different kinds of cancers are affected differently by radiation. Some kinds appear to increase while others decrease. Some kinds of cancer are more treatable than others and thus result in different mortality rates, even if the occurrence rate increases. Simple statements like “cancer death rates show a LNT response to radiation exposure” are way too simplistic to be true, but such statements are easy to base regulations around. When regulators feel a need to support a regulation with some math, they’d rather choose simple math than more-correct, but difficult to understand and explain math.
We can find biological data from cell culture experiments that DNA disruptions are linearly related to exposure. However, most of these experiments are not with healthy, normal, human cell cultures. Bacteria and yeast might have different DNA repair mechanisms than humans. Some human cell culture experiments show hormesis. (example)
In the absence of unambiguous scientific evidence for a simple dose response model, regulators choose a conservative, simple model. They (and the scientists) agree that the model is simple and conservative, i.e. over-estimates the number of deaths. But what gets me riled up is that we ignore the opportunity cost of being excessively conservative. For example, we’ll spend $billions to avoid tens of theoretical deaths counted by the conservative model while not spending similar amounts on things that would much more reliably save thousands of lives. And, at the same time, we take the opposite point of view with global climate change. There, we have good models that show massive disruption, but we take business-as-usual actions because changing would be inconvenient. We are totally inconsistent about what sort of inconvenience is acceptable.
All risk-avoidance regulation should take a years-of-life-lost approach where the best available model (not simplest model) of years of productive life lost are counted against a standard value for a year of productive life. If we did this consistently, we’d spend lots of money developing cures for disease and less money treating disease because treating saves just one person’s life while a cure saves thousands or millions. Likewise, coal air pollution takes thousands (maybe millions) of years of life from asthmatic children while an accident like Fukushima requires extreme assumptions to reach ~1000 years of life lost and where the evacuation has already claimed >500 lives which is at least 5000 years of life lost.
Local optimization results are often extremely sub-optimal relative to global optimization, especially for complex systems. These piecemeal regulations that ignore the greater context can be extremely harmful. The conservative LNT assumption is one such unfortunate local optimization that protects the regulator while harming the populace.
Footnote: More here from Depleted Cranium blog: Evacuation Policy Versus Radiation Level Measurements In Japan