Spent fuel pools found earthquake resistant

The March 11, 2011, earthquake and tsunami that resulted in significant damage to the Fukushima Dai-ichi nuclear power station in Japan increased the interest of the U.S. government, the nuclear power sector, and all stakeholders in the ability of spent fuel pools (SFPs) to withstand earthquakes without a radioactive release.  Although the SFPs and the used fuel assemblies stored in the pools remained intact at the Fukushima plant, the event led to questions about the safe storage of spent fuel and whether the Nuclear Regulatory Commission (NRC) should require the expedited transfer of spent fuel from SFPs to dry cask storage containers at U.S. nuclear power plants.

Following the accident, the U.S. NRC launched a study to determine if more rapid transfer of spent fuel rods from SFPs to dry casks would reduce the risk to the public should a strong earthquake affect a nuclear power plant.  (Sabotage events were excluded from the scope of the study.)  In the first draft of the study, which is now available, the NRC concludes that SFPs are inherently robust structures that are likely to withstand severe earthquakes without leaking.  More specifically, the draft calculates the risk of a release from an SFP from the design seismic event used in the study at about once in 10 million years.  Similar earlier studies also indicated that SFPs present low risks of radioactive releases following an earthquake, but the current draft study indicates an even lower risk.

Further, even in the unlikely event of a release, the NRC draft predicted no early fatalities attributable to acute radiation exposure.  Also, the study showed that the risk of an individual within 10 miles of the plant dying from cancer because of radioactive poisoning following a release is between about two in a trillion and five in a hundred billion per year.  However, extensive mitigative actions may be needed to ensure the local population is not exposed, says the NRC; these are similar to the large-scale evacuations of local populations that occurred near the Fukushima power station.  Also, the draft noted that configuration of the spent rods in the SFPs can affect risk, particularly in cases of high-density loading (i.e., placing more fuel rods in the SFP than the SFP was designed to accommodate).  According to NRC’s modeling, a high-density loading configuration in which the hotter fuel is dispersed throughout the pool generally prevents or reduces the size of potential releases.

Dry storage preferred by some

The draft report is an important development in the evolution of our understanding of the safety of nuclear power plants.  Many public advocacy groups believe that packing SFPs with spent fuel rods beyond the design capacity of the pools is an invitation to disaster should the facility be affected by an earthquake; the groups have urged the NRC to mandate extensive use of dry cask storage.  While dry cask storage is occurring, industry is strongly opposed to any requirement to accelerate use of this option.  The main reason is cost.  For example, in a 2010 study, the Electric Power Research Institute (EPRI) estimated that the expedited movement of spent fuel rods into dry casks and construction of the needed housing would carry a price tag of $3.6 billion for the industry.  The EPRI moreover found that early movement of spent fuel into dry storage would have “significant radiological impacts” affecting workers.

Overall, the draft indicates that the NRC and industry are on the same page in believing that the reduction in risk that would result from speeding up transference of spent fuel rods into dry cask storage is too small to warrant the cost.  The NRC requested comments on the draft and said the final study will be used to “inform a broader regulatory analysis of SFPs at all operating nuclear reactors in the U.S.”

Mark I BWR

The study focuses on a typical SFP at a Mark I Boiling Water Reactor (BWR), the type of reactor affected by the Fukushima event.  SFPs at Mark I BWRs are well above grade.  During earthquakes, higher structures tend to shake more than structures closer to ground level when other factors are equal.  Most inputs the NRC used in its modeling were based on plant-specific design and operational information from the Peach Bottom Atomic Power Station near Harrisburg, Pennsylvania.   The study considered two SFP configurations:

• A relatively full pool where the hottest spent fuel assemblies are surrounded by four cooler fuel assemblies in a 1×4 pattern throughout the pool (referred to as the high-density loading scenario).
• A minimally loaded pool where all spent fuel with at least 5 years of pool cooling has been removed so the hottest fuel assemblies are surrounded by additional water (referred to as the low-density loading scenario).

The study found that the likelihood of an SFP release was equally low for both high- and low-density fuel loading configurations. This is because high- and low-density fuel loading contains the same amount of new, hotter spent fuel recently moved from the reactor to the SPF.  In the event of an earthquake-induced SFP leak, the likelihood of fuel heatup leading to a release was more strongly affected by the fuel loading pattern than the total amount of fuel in the pool.  In other words, the use of favorable fuel patterns such as the 1x4 pattern promotes natural circulation air coolability and reduces the likelihood of a release from a completely drained pool.  NRC’s analysis also shows that for the scenarios and SFP studied, spent fuel is  susceptible only to a radiological release within a few months after the fuel is moved from the reactor into the SFP.  After that time, the spent fuel is coolable by air.

Earthquake risk

Severe earthquakes are rare occurrences, a factor that weighed heavily in NRC’s analysis.  To be conservative, the NRC considered a major earthquake expected to occur once in 60,000 years, which would generate ground motion roughly four to eight times stronger than that used in the model plant’s design.  The study’s structural analysis showed that the SFP would have a 90 percent probability of surviving such an event with no liner leakage (or a 10 percent probability of experiencing damage such that leakage would occur).  The specific conditions for liner failure vary according to site conditions and SFP design.  

The study also evaluated the potential benefits of mitigation provisions that were required under 10 CFR 50.54 (hh)(2) following the September 11, 2001, attacks.  Generally, these provisions require that plant operators develop and implement guidance and strategies intended to maintain or restore core cooling, containment, and SFP cooling capabilities following loss of large areas of the plant due to explosions or fire.  The strategies cover firefighting, operations to mitigate fuel damage, and actions to minimize radiological release.

The NRC said it evaluated successful and unsuccessful mitigation cases and conducted a limited-scope human reliability analysis to estimate the likelihood of an operator successfully implementing 10 CFR 50.54(hh)(2) mitigation measures to prevent fuel damage.  Assumptions used in the evaluation included:

• Postearthquake on-site portable mitigation equipment required by 10 CFR 50.54(hh)(2) is available.
• Minimum plant staffing is available for implementing SFP mitigation.
• The work area is accessible to perform mitigation.

The structural and accident progression analyses show that at least 99 percent of the time, the earthquake would not result in spent fuel overheating even without mitigation for the first 7 days following the accident.  Following the first 7 days, mitigation is needed to prevent fuel damage, and the calculated mitigation success rates range from about 25 percent to 95 percent depending on plant conditions and assuming the refueling floor is accessible.  There are two exceptions where mitigation will be ineffective under the moderate leak scenarios: (1) if the earthquake occurs at the beginning of a refueling outage when the spent fuel is too hot for the assumed mitigation, and (2) if the earthquake occurs when spent fuel is relatively hot and the reactor and SFP are hydraulically disconnected, resulting in insufficient time to deploy mitigation, and natural cooling mechanisms cannot prevent fuel damage.

Dry storage reduces risk

Additional findings and conclusions of the draft study are summarized as follows:

• Expedited movement of fuel from SFPs to dry storage will decrease the amount of radioactive material present if a radioactive release occurred.  This would be “expected to reduce potential health effects, potential land contamination, and economic impacts,” says the NRC.
• Removal of older fuel from SFPs would generally not affect the course of the accident and the off-site consequences because older fuel comprises only 5 percent of the total pool volume.
• A large seismic event is unlikely to result in the loss of structural integrity of the SFP liner; these findings differ from those of past studies. 
• No set of conditions short of a liner failure led to a radiological release in less than 3 days.  In most cases, the available time to prevent a radiological release was much greater than 3 days. 
• Use of the 1x4 pattern has a positive effect on promoting natural circulation air coolability and reducing the likelihood of a release should the SFP become completely drained.
• Absent successful deployment of mitigation, releases could be up to two orders of magnitude larger; releases in such cases are associated with hydrogen combustion events.
• Due to radioactive decay, SFPs tend to have significantly fewer shorter-lived radionuclides than reactors.  Partly because of this, the release is not predicted to be fast and large enough to significantly exceed off-site dose levels necessary to induce early fatalities.
• High-density loading releases without mitigative measures may result in extensive population relocation and land interdiction.  While the amount of land interdiction can be large, the fraction expected to be permanently interdicted is small if a release were to occur.

Ultimately, the NRC concludes that requiring the low-density fuel-storage alternative–i.e., moving older rods to dry casks–is not cost-justified based on the assumptions and analysis performed. 

NRC’s draft, Consequence Study of a Beyond-Design-Basis Earthquake Affecting the Spent Fuel Pool for a U.S. Mark I Boiling Water Reactor, is available at www.regulations.gov in Docket NRC-2013-0136.

William C. Schillaci