. . .
Everybody loves the idea of using water falling by gravity to cool a nuclear reactor containment. It’s what many people think of when they think of passive nuclear power plant designs, especially of the AP1000. The ideas are not unfamiliar: Locating small water tanks to be used in an emergency to quench fires on a building’s roof is quite common. You don’t have to count on the fire truck coming and getting its pumps going; so long as you succeed in getting the valves positioned correctly, and keep things from blocking the flow so it gets to the right places, the water drains from the tank to quench fires, just by gravity. Hence the appeal.
The idea of removing heat from a building’s interior by spraying water on its roof is also appealing: so much more efficient than running air handling equipment. And the use of a domed roof makes it even better. In fact, in a domed roof building, even water pools inside the building’s courtyards and water spray on the interior walls are effective means of passive cooling of the building’s interior when the sun beats down on it. These are all considered smart designs in passive solar building, especially in hot, sunny, dry climates (http://www.lightningcanyon.com/pages/passive-cooling/domes.html) .
But something ironic, tragically ironic, happened on the way to actually implementing this beautiful idea on the AP1000. During the design process, nobody remembered to include the effects of the radiant heat from the sun in their calculations. I brought this up to the Nuclear Regulatory Commission during the public comment period when the AP1000 was undergoing the design certification process in 2005. Surely sunlight falling on a building will be a source of heat input and must be included in the calculations for a method relying on passive methods of heat removal via passageways in buildings, I argued then. The NRC is required to respond to public comments on proposed rules, in writing, in the Federal Register, and it did so. Its response was unsatisfactory, in my estimation. However, the rulemaking process by which design certification is granted had just been revised so as to eliminate mandatory hearings for design certifications, so I had no voice in the matter to respond to the NRC’s unsatisfactory response. I laid this story out in an earlier blog post, here: https://mattersofproportion.wordpress.com/energy-the-environment/nuclear-energy/ap1000-new-nuclear-power-plant-design-making-the-nrc-accountable-for-doing-its-job-in-regulating-new-reactor-designs/
Now, a later version of the AP1000 (revision 19) is undergoing design certification. To follow up on the issue of including the heat of solar radiation in the safety analyses of the AP1000, over the summer I looked more deeply into where solar radiation should have been, but wasn’t, included in calculations for the safety analyses of the lastest revision of the design. How much did it matter? Did it make a difference to anything important to public safety? I pored over the publicly available materials about the latest design on the NRC’s website.
What I found was nothing short of shocking (to me, at least). Not only did the structural analysis of the shield building in the latest revision assume that concrete surfaces never got hotter than the ambient air nor cooler than it, which is contradicted by our everyday experience, but the error of neglecting solar radiation hadn’t been corrected anywhere else, either.
The much-vaunted experimental tests of the process by which the containment is cooled so as to keep the pressure in check (and, thereby, to keep the steel containment intact) were performed outdoors, in the sun — whereas the wetted steel surface will actually be indoors, in the dark. Evaporation from the surface of a fluid by direct radiation of the sun is an important factor — in fact, pan evaporation is the method by which solar radiation is often estimated. That means the conclusions that were drawn from the experimental tests aren’t valid. These are the tests that were used to “validate” the calculational methods of the computer code used to compute peak containment pressure. The values of peak containment pressure computed by that computer code are the basis on which the design is certified. What this means is that now we really don’t know whether the safety systems would be able to keep the containment intact in the event of an accident.
I dropped everything to prepare letters to explain the situation to the Nuclear Regulatory Commission. On Tuesday (August 16th, 2011) I gave a brief slide presentation to the Advisory Committee on Reactor Safeguards (ACRS) summarizing the points in my letters, explaining their significance to public safety, and urging the ACRS not to let the opportunity to correct this error before it is too late slip by.
I wanted to hear what the AP1000 designers might have to say, before posting my letters and presentation on this blog. So I waited until after the meeting. The meeting is public, and I listened in. Was there something I had not noticed? No, there wasn’t. Did the NRC staff or the applicant have good answers? No, they did not. So, now it is time to post my letters and presentation.
Here are the slides of my talk (“Forgetting about the sun”) on August 16, 2011, in pdf format: SterrettSlidesACRS16August2011
Here is the text of my talk keyed to the slides: SterrettTalkTextACRSMeeting16August2011
UPDATED September 30, 2011: The following URL links to a single pdf document of the slides integrated with the text of the talk ( < 5 minute): SterrettACRSPresentation16August2011TalkTextWithSlideImages
Here is the letter about the heat of solar radiation on the concrete shield building: SterrettCommentShieldBuildingExteriorSurfaceTemperaturesAP1000Meeting30June2011
Here is the letter about the heat of solar radiation and the experimental tests used to validate the computer simulations of peak containment pressure: SterrettCommentAP1000WGOTHICvalidationPCSLSTsolarinsolationissue30June2011Meeting
Given that ratepayers are already being assessed fees to construct this plant, I think what I have found out deserves to be well known. Given the political pressure to grant design certification, the NRC is in a tight spot and it will probably be only a brave few at the NRC who can garner the courage to do the right thing. Let’s support the brave few.
I’ve never asked readers to retweet and repost and share before. This time, I do: please share the link to this blogpost widely.