Earthquakes and Nuclear Power Plants- This Chat is closed
My name is Dr. Annie Kammerer and I am a senior seismologist and earthquake engineer at the NRC. I oversee research on a broad range of seismic topics. I also have 15 years of experience in the private sector, and have authored many publications, regulatory guidance, technical reports and journal articles. I hold a BS in Civil Engineering, an MS in Geotechnical Engineering and a PhD in Geotechnical Earthquake Engineering.
I look forward to your questions about how the NRC makes sure plants can withstand any earthquakes they may experience. I can also talk about the Mineral, Va., earthquake in 2011 that affected the North Anna nuclear power plant. (Just for the record, I won’t be able to address specific questions about designs or risk at specific sites.)
Moderator, 
Erica Gray, 
Moderator, and 5 others are discussing. 
NRC Chat 
Moderator 2:01 pm on August 13, 2013 Permalink
Hi, thanks for joining our chat. We’ll be discussing how the NRC makes sure U.S. nuclear power plants can withstand any earthquakes they may experience. My work deals with broader seismic topics, so I won’t be able to address detailed questions about designs or risk at specific sites. I can, however, also talk about the Mineral, Va., earthquake in 2011 that affected the North Anna nuclear power plant.
Please remember to refresh the page regularly for the latest content. And even if you’re replying to a comment or response, please use the comment box at the bottom of the Chat. That way your comment stays in chronology order.
AK
Moderator 2:03 pm on August 13, 2013 Permalink
Here’s a question we received via e-mail:
Does the NRC currently envision changing the Safe Shutdown Earthquake for any of the Western plants as a result of the 50.54(f) process?
The Safe Shutdown Earthquake is the original ground motion level used for design of the plant. As such, the SSE doesn’t change. However, if the new ground motion estimates exceed the original design level, the plant will have to conduct a seismic probabilistic risk assessment (seismic PRA). Based on the outcome of that assessment, the NRC may determine that the plant must perform modifications to strengthen equipment or anchorage based on the new higher ground motion. At that point, a new ground motion level would be added to the licensing basis of the plant.
AK
Moderator 2:05 pm on August 13, 2013 Permalink
Another question that came in before the chat asked:
To what extent will the Western plants be required to apply the 10 CFR 100.23 criteria — originally developed for license applications filed after January 1997 — in developing their 50.54(f) responses?
The 50.54(f) responses require that probabilistic seismic hazard assessments (PSHA) be performed using the NRC guidance described in NUREG 2117 and NUREG/CR-6372. The 50.54(f) letter outlines a process similar to that used for new plants. The plants in the West must perform site-specific PSHA using processes similar to new reactors, which is in line with 10 CFR 100.23.
Plants covered by the new Central and Eastern US Seismic Source Characterization (CEUS SSC) model described in NUREG 2115 can use that new model along with existing geotechnical information. The CEUS SSC model was developed using the guidance in NUREG 2117. The NRC is allowing plants to use existing site data; although the state of California has required new investigations for the two California plants.
AK
Mike 2:07 pm on August 13, 2013 Permalink
Would these modifications be considered backfits and be subjected to all of thoses requirements?
Moderator 2:11 pm on August 13, 2013 Permalink
I think you’re referring to the modifications in response to the new risk re-evaluation studies? That’s a regulatory question that can be answered after the plants finish their re-evaluation. It will be answered by the Japan Lessons Learned directorate once the data is available.
AK
Moderator 2:08 pm on August 13, 2013 Permalink
Earthquake Fact: The Japan 2011 earthquake was caused by a “subduction zone” event, which is the type of earthquake that can produce the largest magnitudes. A subduction zone is a tectonic plate boundary where one tectonic plate is pushed under another plate. In the continental US, the only subduction zone is the Cascadia subduction zone which lies off the coast of northern California, Oregon and Washington.
Bill K 2:10 pm on August 13, 2013 Permalink
Hi–just joined, a few minutes late. Thanks for doing this.
On an NRC seismic issues meeting that was webcast last week, two geologists were disagreeing about what plate, or what geological zone, Florida should be classified as part of. If there is professional disagreement about such a basic question, how can models for specific plant sites be considered reliable and accurate at the levels of precision that are supposed to be achieved?
Moderator 2:14 pm on August 13, 2013 Permalink
In developing models, we take into account differing explanations for the available data, and the models do capture “alternative technically defensible interpretations,” which means that we capture a broad range of opinions in the model. It is explained in detail in NUREG 2117.
AK
Moderator 2:11 pm on August 13, 2013 Permalink
Another question submitted before the chat asks:
Are dams upstream of nuclear power plants analyzed for earthquake protection adequacy? An example would be those dams along the Missouri River upstream the Fort Calhoun and Cooper nuclear stations. A catastrophic failure of these dams would result in a tsunami for these plants.
The NRC does consider the potential impact of flooding from dam failure on the nuclear plants, although the design criteria for the dams themselves are the jurisdiction of other agencies. Seismic dam failure is addressed in NUREG 0-800 (the Standard Review Plan) as well as other NRC guidance documents. Most recently, the NRC developed Interim Staff Guidance (JLD-ISG-2013-01, “Guidance For Assessment of Flooding Hazards Due to Dam Failure,” July 2013, ML13151A153) on how the assessment is to be performed.
AK
Moderator 2:12 pm on August 13, 2013 Permalink
From e-mail:
Hello,
My question is:
After the 5.8 earthquake in VA 2011, Dominion installed new seismic equipment as part of the NRC allowing them to restart. Dominion restarted 11/15/11.
What equipment did they install? Seems the new equipment did not function and is not sensitive enough.
2/1/12
“None of the station’s seismic instrumentation actuated during the event,” the NRC report said. A new seismic monitoring system installed since last summer’s quake–a free field seismic instrumentation unit that can alert operators to a potentially damaging quake–also did not activate.
http://fredericksburg.com/News/FLS/2012/022012/02012012/680309
North Anna’s design basis was exceeded by the original earth quake. How are aftershocks (over 450) accounted for in analyzing safety … what are the cumulative effects?
Thank you,
Erica Gray
It’s very common for seismometers to have a certain triggering ground motion threshold. The newspaper article you cite is about an aftershock of magnitude 3.2. This is a very small earthquake. The instrumentation that alerts operators in the control room is designed to only go off if the ground motion exceeds the operating basis earthquake ground motion, which wasn’t the case in this aftershock.
But yes, you are correct that since the 2011 earthquake, the plant installed new seismometers, including a new instrument in the free field i.e. away from the plant.
The capacity of safety-related equipment is unaffected by aftershocks that have ground motions that are not at a damaging level.
AK
Erica Gray 2:15 pm on August 13, 2013 Permalink
Please state what the seismic equpment is at North Anna.
And where can I find the studies about cumulative/after shocks not having an impact?
Moderator 2:19 pm on August 13, 2013 Permalink
The seismic equipment is new seismographs installed at the site.
Safety-related structures and components are designed to remain damage free up to and beyond the design basis ground motion. Even repeated small ground motion does not have a detimental affect.
AK
Moderator 2:15 pm on August 13, 2013 Permalink
Earthquake Fact: Ground motion is a function of both the magnitude of an earthquake and the distance from the fault to the site. Nuclear plants, and in fact all engineered structures, are designed based on ground motion levels, not earthquake magnitudes. The existing nuclear plants were designed in a way accounts for the largest earthquakes expected in the area around the plant, with an additional margin added in.
Erica Gray 2:21 pm on August 13, 2013 Permalink
Please state the name/brand of the equipment in use at North Anna.
Moderator 2:24 pm on August 13, 2013 Permalink
I’m afraid you’ll have to pose that question to the plant. I don’t recall the brand off the top of my head.
AK
Erica Gray 2:33 pm on August 13, 2013 Permalink
Why did the NRC not know about North Anna’s missing seismic sensors and that they were not in the proper locations?
Moderator 2:34 pm on August 13, 2013 Permalink
There was a failure of one sensor due to lack of power. The remainder of the equipment in the plant’s licensing basis was available and we have the data.
AK
Jeff Clark 2:16 pm on August 13, 2013 Permalink
If a new ground motion level is established (after the Risk Assessment) for a plants licensing basis, does that new ground motion level need to be developed under 10CFR50 App B requirements?
Moderator 2:21 pm on August 13, 2013 Permalink
Ground motion assessment studies are to be conducted consistent with guidance in NUREG 2117 and NUREG/CR 6372 as noted in the 50.54f letter.
AK
Bill Horstman 2:16 pm on August 13, 2013 Permalink
Hello Dr. Kammerer
Since the ground motion recorded at North Anna during the August 2011 Mineral VA earthquake exceeded the SSE-equivalent for the site (I believe that they use the term DBE), does the utility need to add a similar earthquake ground spectrum to their design basis and perform deterministic evaluations of the plant for this new ground spectrum, or is this SSE exceedance being addressed through a probabilistic seismic hazards analysis/probabilistic risk assessment?
Moderator 2:23 pm on August 13, 2013 Permalink
As part of the restart of the North Anna plant, the ground motions experienced at the site were added to the plant’s licensing basis. Additionally, North Anna, like all operating U.S. reactors, is undergoing a seismic re-evaluation, which includes probabilistic seismic hazard analysis. It is also expected they will perform a risk re-evaluation.
AK
Moderator 2:18 pm on August 13, 2013 Permalink
Here’s a question from our e-mail:
1.) How is the corporate data collected at a facility correlated to the USGS data? What is the frequency of these certification activities and is there any feedback from existing certification data?
2.) Are the spent fuel pools in the containment buildings part of the design basis for the study. These were modified after base design and their safety risk is of prime concern.
3.) What are you doing to improve the science behind these studies?
Fukushima is evidence of poor design basis and poor planning, I expect every government employee to perfrom due diligence and challenge their own rules. To do otherwise borders on criminal or moral negligence.
Sincerely,
Richard C. Wood
Diablo Canyon is the only facility that submits its seismograph network information directly to the UGSG monitoring system. The decision to do this is up to the facility.
Can you give me more information about what you’re asking in regards to certification activities?
Spent fuel pools use the same ground motion design basis as the reactor building itself. I think that’s what you’re asking. Currently the NRC is re-evaluating risk of spent fuel pools in several ways as part of our post-Fukushima re-evaluation.
I’m not sure what studies you mean in Question 3.
AK
Anonymous 2:18 pm on August 13, 2013 Permalink
What frequency range was the OBE exceedance at North Anna in during the Mineral, Va., earthquake in 2011 ?
Moderator 2:24 pm on August 13, 2013 Permalink
It exceeded a wide range of frequencies of engineering interest. I can’t recall specifics off the top of my head.
AK
Erica Gray 2:43 pm on August 13, 2013 Permalink
Hows that tritium leak that was reported after the earthquake?
Moderator 2:44 pm on August 13, 2013 Permalink
I’m not aware of a tritium leak at the site related to the earthquake.
AK
Erica Gray 2:49 pm on August 13, 2013 Permalink
Yea I guess the tritium leak was just a coincidence
http://news.fredericksburg.com/newsdesk/2012/02/22/north-anna-nuclear-plant-reports-elevated-tritium-reading/
Moderator 2:21 pm on August 13, 2013 Permalink
Earthquake Fact: The ground motions used as seismic design bases at US nuclear plants are called the Safe Shutdown Earthquake ground motion (SSE). In the mid to late 1990s, the NRC reviewed the potential for ground motions beyond the design basis. From this review, the we determined that seismic designs of operating nuclear plants in the US have adequate safety margins for withstanding earthquakes. Now, we’re again assessing the resistance of US nuclear plants to earthquakes. Based on NRC’s preliminary analyses, the mean probability of ground motions exceeding the SSE over the life of the plant for the plants in the Central and Eastern United States is less than about 1%.
Erica Gray 2:24 pm on August 13, 2013 Permalink
Based on what data?
The Usgs stated only a handful of seismometers were available.
Dominion VA ~ North Anna was missing sensors and others were not in proper locations.
Moderator 2:30 pm on August 13, 2013 Permalink
I’m sorry, but I’m not sure what you’re asking in terms USGS and available seismometers. As to the location of sensors at North Anna prior to the earthquake, they were in the locations specified by their license.
AK
Bill K 2:23 pm on August 13, 2013 Permalink
An emerging theme in NRC discussions of plant safety is “excessive conservatism.” For example, it;s now being said that the results of SOARCA indicate that some standards can be relaxed a bit. How is this topic playing out in the realm of seismic standards?
Moderator 2:28 pm on August 13, 2013 Permalink
Generally, the NRC is moving towards risk-informed decision making in a wide variety of ways including seismic risk assessment and design. One of the benefits of a risk-informed framework is its ability to target the most risk important stuctures, components and systems so that you reduce excessive conservatism in places where it’s not needed.
AK
Moderator 2:24 pm on August 13, 2013 Permalink
Earthquake Fact: The only nuclear plant near the Cascadia subduction zone is the Columbia Generating Station in Washington state. But the plant is 300 miles from the zone and 225 miles from the coast, unlike the plants at Fukushima, which were nearer the zone and on the coast.
Erica Gray 2:29 pm on August 13, 2013 Permalink
Earthquake Fact ~ They can not be predicted!
North Anna is on a fault line….Do you think that is a good idea?
Moderator 2:33 pm on August 13, 2013 Permalink
North Anna sits in a known moderate seismicity region. It is not on an active fault line. What’s important to understand is that earthquakes can occur anywhere and that’s why our regulations take into account background seismic activity that has occurred in the region of the plant as well as any known active faults. What’s important is getting a good assessment of the hazard at a site and assuring the plant has the capacity to withstand the ground motion it may experience.
AK
Moderator 2:27 pm on August 13, 2013 Permalink
Earthquake Fact: According to the USGS, 90 percent of the world’s earthquakes occur along plate boundaries where the rocks are usually weaker and yield more readily to stress than do the rocks within a plate. The remaining 10 percent occur in areas away from present plate boundaries — like the great New Madrid, Missouri, earthquakes of 1811 and 1812, felt over at least 3.2 million square kilometers. They occurred in a region of southeast Missouri that continues to show seismic activity today.
Moderator 2:30 pm on August 13, 2013 Permalink
Earthquake Fact: Do any plants have special design considerations associated with seismic design? Many plants do, but perhaps the most notable are the automatic reactor trip systems in Diablo Canyon and San Onofre.
Moderator 2:34 pm on August 13, 2013 Permalink
Earthquake Fact: The most commonly used magnitude measurement today is the Moment Magnitude, Mw. It is based on the strength of the rock that ruptured, the area of the fault that ruptured and the average amount of slip. It is a direct measure of the energy released during an earthquake. The Richter magnitude scale was developed in 1935 by Charles F. Richter of the California Institute of Technology and was based on the behavior of a specific seismograph that was manufactured at that time. The instruments are no longer in use although the scale so commonly used by the public that scientists generally just answer questions about “Richter” magnitude by substituting moment magnitude without correcting the misunderstanding.
Moderator 2:36 pm on August 13, 2013 Permalink
Earthquake Fact: Can animals predict earthquakes? According to the USGS, the earliest reference to unusual animal behavior before a quake is from Greece in 373 BC. Anecdotal evidence abounds of animals, fish, birds, reptiles and insects exhibiting strange behavior just before an earthquake. But there’s no scientific proof or explanation of the possible phenomenon.
Bill Horstman 2:38 pm on August 13, 2013 Permalink
In response to the earlier question regarding the OBE exceedance at North Anna Power Station:
I have seen a PowerPoint presentation on this event, which was given to the NRC on September 8, 2011 by Dominion Power. Two of the slides show comparisions of the response spectra associated with the Mineral VA earthquake with the OBE and DBE for NAPS. This presentation should be available on the NRC website.
Moderator 2:40 pm on August 13, 2013 Permalink
The plot that was shown by Dominion in that meeting has been also shown a number of other times. Anytime anyone presents to us at a public meeting, it should be captured in the meeting minutes and in ADAMS.
AK
Anonymous 2:44 pm on August 13, 2013 Permalink
Thanks Bill, this is very interesting!
Bill K 2:38 pm on August 13, 2013 Permalink
How would you characterize the seismic environment at the Virgil Summer site, not far from the site of the big Charleston quake and not all too far from the New Madrid region?
Moderator 2:43 pm on August 13, 2013 Permalink
I haven’t worked on that particular site, so I can’t speak to that. However, the source characterization model that the NRC recently published in NUREG 2115 would be used for that site for the ongoing re-evaluation efforts. The model in 2115 was developed cooperatively with the USGS, DOE and EPRI.
AK
Moderator 2:43 pm on August 13, 2013 Permalink
Earthquake Fact: Earth scientists believe that most earthquakes are caused by slow movements inside the Earth that push against the Earth’s brittle, relatively thin outer layer, causing the rocks to break suddenly. This outer layer is fragmented into a number of pieces, called plates. Most earthquakes occur at the boundaries of these plates.
Moderator 2:44 pm on August 13, 2013 Permalink
The Chat ends at 3 p.m. I really appreciate your interest and have tried to get to most of your questions. Thanks for joining us today!
AK
Erica Gray 2:50 pm on August 13, 2013 Permalink
It’s not 3 pm yet.
Moderator 2:51 pm on August 13, 2013 Permalink
We’re still here!
AK
Erica Gray 2:52 pm on August 13, 2013 Permalink
Does the NRC check to make sure the sensors are working?
Moderator 2:54 pm on August 13, 2013 Permalink
Inspection of all equipment is the responsibility of the licensee. But the resident inspectors (our NRC boots on the ground) generally make sure that the inspections occur consistent with the regulations.
AK
Erica Gray 2:55 pm on August 13, 2013 Permalink
Inspection of all equipment is the responsibility of the licensee. But the resident inspectors (our NRC boots on the ground) generally make sure that the inspections occur consistent with the regulations.
Then why were North Annas sensors not working and not properly placed?
Moderator 2:58 pm on August 13, 2013 Permalink
We have data from all the sensors, which were in the plant per the licensing basis. The non operation sensor was an annuciator only and was not intended to record data.
AK
Moderator 2:47 pm on August 13, 2013 Permalink
Just a reminder that if you have questions or comments unrelated to today’s Chat, you can post them here: http://public-blog.nrc-gateway.gov/category/open-forum/
AK
Moderator 2:51 pm on August 13, 2013 Permalink
Earthquake Fact: In the U.S., according to the USGS, Florida and North Dakota have the fewest earthquakes, while California and Alaska have the most.
Erica Gray 2:52 pm on August 13, 2013 Permalink
Florida has sinkholes!
Moderator 2:55 pm on August 13, 2013 Permalink
Generally any place that has limestone deposits has sinkholes in karst topography. They’re not related to earthquakes.
AK
Erica Gray 2:57 pm on August 13, 2013 Permalink
well NPP should not be in areas prone to sinkholes either!
Moderator 2:54 pm on August 13, 2013 Permalink
Earthquake Fact: Is there such a thing as “earthquake weather”? Not according to the USGS. Statistically, there is about an equal distribution of earthquakes in cold weather, hot weather and rainy weather. That said, very large low-pressure changes associated with major storm systems like typhoons and hurricanes are known to trigger episodes of fault slip and may play a role in triggering damaging earthquakes. But the numbers are small and not statistically significant.
Erica Gray 2:58 pm on August 13, 2013 Permalink
It just takes one bad event!
Bill K 2:56 pm on August 13, 2013 Permalink
It seems to me that the epistemic uncertainty in the field of geology far exceeds the error bars in most engineering designs. Doesn’t this make risk-informed analysis rather a risky way to proceed?
Moderator 3:00 pm on August 13, 2013 Permalink
Over the last few decades, the field of seismology has significantly matured to the point where epistemic uncertainty has been greatly reduced, and many people would say well captured. An example is the new seismic source characterization model in NUREG 2115.
AK
Moderator 2:59 pm on August 13, 2013 Permalink
Earthquake Fact: In Washington State, the small Juan de Fuca plate off the coast of Washington, Oregon, and Northern California is slowly moving eastward beneath a much larger plate that includes both the North American continent and the land beneath part of the Atlantic Ocean. Plate motions in the Pacific Northwest result in shallow earthquakes widely distributed over Washington and deep earthquakes in the western parts of Washington and Oregon.
Erica Gray 2:59 pm on August 13, 2013 Permalink
One word…Hanford!