During a recent discussion thread on Atomic Insights, a frequent commenter cited a 1997 vintage Nuclear Regulatory Commission document as indicating — in his opinion — that there was always a chance that a spent fuel pool could experience an event that might release dangerous amounts of radioactive material.
Besides the fact that the report actually stated that the risk went to zero as long as the most recently operating fuel had been in the pool for at least 7 months (BWR) or 17 months (PWR), there were other simplifications associated with the study due to the state of computer modeling in 1997.
Aside: As an indication of the fact that the 1997 vintage document — designated as NUREG/CR-6451 — is considered to be obsolete by the NRC, it is no longer available from the web page that provides access to currently valid NUREG/CR reports. End Aside.
In response to the comment indicating a misunderstanding of the results of the report and the results of more recently conducted analysis, an additional comment was posted that led me to believe that the author, who had not revealed his name or much of his educational or professional background knew a lot about the topic.
When contacted directly, the commenter agreed to have the comment giving some details about spent fuel pool modeling promoted to the front page and agreed to reveal his name.
Here is his comment, which has been slightly revised to be more appropriate as a stand alone post.
There has recently been an increased amount of interest in the safety of on-site storage of spent fuel at nuclear reactors, particularly in what could happen if there is an accident involving a spent fuel storage pool. The behavior of spent fuel is a complex field that I have only touched the edges of, but I might be able to help explain some of what’s going on and why there are conflicting predictions about what might happen in an accident.
Traditionally, spent fuel received a lot less attention than the reactor itself. This started to change in the 1990’s. An example of this was NUREG/CR-6451, a project designed to identify a worst-case event. Some of the assumptions it makes are reasonable, such as for the fuel loading in the pools. Others are very conservative. In particular the temperature chosen to represent failure has been set at 565°C.
This is the point where the cladding can start to swell, potentially resulting in a rupture that would allow the release of most volatile or gaseous radionuclides trapped between the cladding and the fuel itself. This isn’t a good thing, but it will only have significant consequences if there is still a large amount of iodine 131 in the spent fuel. As iodine 131 has an 8-day half-life this concern only lasts a few months after the fuel was removed from the reactor.
More serious failures may start to occur once the fuel reaches about 800°C. This is when the cladding can begin to rapidly oxidise and large quantities of fission products may be released. Spent fuel will obviously be able to reach 565°C for a much longer length of time after removal from the reactor than it will be able to reach 800°C.
NUREG/CR-6451 used SHARP (Spent-fuel Heatup: Analytical Response Program) to model the behavior of the spent fuel – it was actually the first time that SHARP was used. SHARP is useful for quick modeling as it is relatively easy to use but it is basic and lacks key aspects required for an accurate model.
In particular SHARP is basically incapable of modeling at higher temperatures (say above about 500 C or 600 C), as it does not take radiation heat transfer or materials changing (oxidizing, melting etc.) into account. Calculations performed with SHARP are generally not acceptable for use in making safety related decisions. This doesn’t mean that the SHARP is useless, just that it needs to be used with care at high temperatures.
As part of the post September 11 NRC review much more serious modeling of spent fuel accidents was undertaken using MELCOR rather than SHARP. MELCOR is fully validated for modeling severe accidents and includes almost everything you can think of – but it is difficult and time consuming to use.
The post September 11 modeling was further refined after the Fukushima accident, particularly to take into account higher burnup from modern reactor operation. This advanced modeling is substantially more accurate than modeling using SHARP and matches physical experiments well.
From this we can see how our understanding of spent fuel in an exposed storage fuel has changed: Assuming it will be fine -> basic modeling indicating a potential issue in a worst case situation -> detailed modeling showing the there really isn’t a problem, including calculating actual limits that need to be observed and how long it will all take.
Whenever I see things like this I think that the progress the industry has made is astounding even without the opportunity to make major changes by building new reactors. I also gain even more respect for the engineers and scientists involved in the early days of the nuclear industry who had to make so many decisions without modern tools and knowledge but still managed to get so much right.
Andrew Yule is an Australian health physicist. He has a background in high energy physics and computer modeling. He has recently increased his involvement with nuclear physics, with a special focus on modeling and simulation.
Andrew’s detailed explanation is a good companion to a post published yesterday on ANS Nuclear Cafe titled Spent fuel pool fire risk goes to zero a few months after reactor shutdown.
Between the two articles, I hope that a growing number of people recognize that this particular issue has been severely thrashed to the ground. The Nuclear Regulatory Commission took the correct action when it voted 4-1 to stop any additional generic analysis on the topic of spent fuel pool fire risk at decommissioned plants.
It was time — probably way past time — to stop spending more valuable time and money on additional analysis. It is also time to reassure people who live anywhere near such facilities that they have no reason to worry about the spent fuel pools, no matter how often activists try to induce fear, uncertainty and doubt.
Politicians like Senators Boxer, Sanders and Markey should be encouraged to stop bringing up the issue in an attempt to force nuclear power plant operators to spend money on unnecessary movements of used nuclear fuel into expensive dry casks. Affected citizens should remind their representatives via cards, phone calls and letters that resources wasted on trivial issues that have nothing to do with improved economics, safety, reliability, or cleanliness are resources that are not available for more productive uses.