Tuesday 28 July 2015

Interesting online discussion I unearthed on the AREVA EPR

The question of why the EPR is turning out to be so much more expensive than other reactors is interesting. It looks like the EPR has overspecified safety requirements. For example: rather that just be certain the core could not melt in the event of an accident, the designers went one step further and built a core catcher in as well. Some people blame German Greens for the design complexity: (here:)

lion :
The EPR must meet the security demands of the ASN and, above all, the demands voiced by the German Green politician, Jürgen Trittin, who made the restrictions applying to the design and build so severe that it is almost unfeasible. So I wonder whether the EPR is not “over secure” in terms of design and build, a fact that would not even mean that it is optimally secure when it comes to operations.
Jürgen Trittin is a German Green politician. He was Federal Minister for the Environment, Nature Conservation and Nuclear Safety from October 1998 to November 2005 in Germany. Angela Merkel was the Minister before him: From November 1994 to October 1998. [ Merkel was and is a pro-nuclear power person. ]

It seems to me that the design was substantially complete before 1998 when the Greens took over control of the German energy ministry. See this PDF report, which has some history and these NRC slides. The design continued from 1997 to 1999 with an updated "Basic Design Report" for the EPR issued in February 1999.

Friday 24 July 2015

Edward Calabrese talking about hormesis

Youtube video. Time: 40:37

I wanted to validate the hormetic model in some sort of way and test its frequency in the population and in studies. And so I said how did the government and how did the scientific community validate the threshold model. Because I'll just copy how they did that, because I never validated a model before. So I'll just find what they did, follow what they did. Maybe I'll do it right. So I looked and I looked and looked and I looked and I looked and looked and, guess what, I never found that any government, any scientist, any industry, any expert committee, ever ever published a paper in which they attempted to, or showed validation of the threshold model for making accurate predictions below the threshold. Never. And we live below the threshold. All of our standards were based upon a model that was never validated by anyone. OK. All the drugs, all the environment contaminants that are based upon the threshold model have a model that was never a validated. Never thought to be validated.

So I decided to try to validate it. And so what happened is I got 3 or 4 very large datasets. Looking at a wide variety of compounds. Applied statistical analysis with a team of researchers to test the linear threshold and hormetic dose responses. And guess what happened. In every single case the threshold and the linear model fail to make accurate predictions in the low-dose zone. And the hormetic model was the only one that actually did it. We submitted it to the leading toxicology journals went through some tough peer reviews and all three made it through.

And basically what we showed, for the very first time, was that in fact the models that are used by the regulatory agencies fail when put to the test. The only one that actually passed the test was the one that the regulatory agencies actually ignored, marginalized, and basically would never consider. And so what you have is, you have a world that is governed by non-science really. And it's hard to tell them, it's hard to tell the teacher that they're wrong.