Frequently Asked Questions About Reactor Vessel Head Degradation
Index to All Frequently Asked Questions Pages
The NRC has stated that the type of degradation discovered at Davis-Besse in March 2002 was unforeseen. Had the NRC been aware of it, what would have been the NRC's actions?
Upon discovery of any leakage through the reactor coolant system pressure boundary, the plant would have been in violation of the technical specifications of its operating license (which allows no known primary coolant system pressure boundary leakage, except for the separate requirements for steam generator tube integrity), and the licensee would have had to initiate a controlled shutdown. Absent the licensee taking such action on its own, the NRC staff would have required such action based upon plant violation of the technical specification requirement. Thus, if the existence of RCS pressure boundary leakage through the reactor vessel head had been identified prior to shutdown, the NRC staff had both justification and means to require shutdown. In addition, had the NRC known of the severe degradation that was later discovered on the RPV head at Davis-Besse, it would have required that the plant be shutdown. The basis for such action would have been that the designed safety margin was lost due to the severe degradation of the carbon steel reactor pressure vessel head.
Does the safety analysis for nuclear power plants consider cracks or holes specifically located in the reactor vessel head?
Cracks or holes specifically located in the reactor vessel head are considered in safety analyses in nuclear plants as the consequences of a break in a single control rod drive mechanism (CRDM) housing. The CRDM housing fully penetrates the reactor pressure vessel head. The break of a single CRDM housing which would produce a hole in the upper reactor vessel head is evaluated in each plant's final safety analysis report (FSAR). Analyses of this event include the additional conservative assumption that a control rod assembly is ejected from the core with an accompanying increase in power at that portion of the core. Pressurized water reactors (PWRs) operate at power with very few control assemblies inserted within the core, so these analyses are conservative. The offsite radiological consequences are evaluated and are required to be within the limits of the Commission's regulations. Only the early portions of control rod assembly ejection accidents are evaluated in the plant's FSAR, since the long-term recovery phase of this event, which involves the operation of the emergency core cooling systems, is bounded by the analyses of the larger breaks in the coolant piping that are also described in the plant's FSAR and are part of the plant design basis.
Failure of the vessel head as a result of boric acid corrosion of the carbon steel is not explicitly postulated in the design basis for all PWRs. Such a failure is not considered acceptable, and as a result of the corrosion found at Davis-Besse, the NRC is taking action to assure such failures do not occur.
Never the less, had the as-found corroded area in the Davis-Besse head failed, the consequences were within regulatory limits.
The staff concludes that the consequences from a crack or hole in the reactor vessel head of a nuclear power plant would be bounded by the consequences of a break of equivalent size in the coolant piping. The consequences from breaks in the coolant piping are routinely analyzed in the plant's FSAR.
If the stainless steel clad had failed, what assurance does the NRC have that the safety systems at Davis-Besse or any other plants would have mitigated the event?
If the stainless steel clad in the upper head of the Davis-Besse reactor had failed, a hole approximately 8 inches in diameter with an equivalent cross-sectional area of about 100 square inches (in2) would have been created in the reactor vessel head. The high pressure and low pressure safety injection systems are designed to mitigate the effects of breaks up to and including a double-ended rupture of the largest pipe in the reactor coolant system. The largest pipe in the Davis-Besse reactor coolant system is the hot leg, which measures 36 inches in diameter. In the case of a double-ended rupture, coolant discharges through both ends of the pipe. Therefore, the Davis-Besse safety injection systems can mitigate the consequences of a break with a cross-sectional area equal to approximately 2000 in2, which is 20 times the size of the degraded area of the head. Even in the case of a double-ended rupture of the hot leg, the high and low pressure injection systems would keep the temperature of the fuel rod cladding and the cladding oxidation below the limits established in 10 CFR 50.46. The requirements for maintaining a coolable geometry and long-term cooling would also be met.
If the clad had failed, what would the consequential damage to the control rod drive system have been?
Control rod drive mechanisms (CRDMs) are electromechanical devices which rely on electric power to engage the control rod and raise or lower it in response to signals from the reactor control panel. In response to an emergency, the electric power to the CRDMs is quickly interrupted causing the CRDMs to detach from the control rods, which drop into the reactor core. This is commonly called a "scram." The electric power to the CRDMs is turned off by opening electrical breakers (switches), which are located outside the reactor containment. The scram breakers open when they receive a signal from various safety sensors, which detect a variety of abnormal operating conditions, including loss-of-coolant accidents (LOCAs).
Damage to the CRDM electrical cables above the reactor vessel could cause the electrical power to be interrupted directly, in which case the control rods would drop into the core. Alternatively, an ejected housing could cause a short in the electrical power system, which would be detected and then open the scram breakers, which would cut off power to the control rods, dropping them into the core and shutting down the reactor. In both of these cases, pressure sensors in the reactor coolant system are designed to detect the occurrence of the LOCA and they would send their own separate signal to the scram breakers to open, allowing the control rods to drop into the core.
Could the control rod drive mechanism (CRDM) nozzle cracking discovered at Davis-Besse initiate a loss-of-coolant-accident?
Although the potential existence of the CRDM cracking phenomenon did not warrant immediately shutting down the plant based on risk-informed principles, there was some increase in loss-of-coolant accident frequency compared with the base-case probabilistic risk analysis. The NRC staff has concluded that a through-wall circumferential crack of sufficient length to challenge the structural integrity of the primary coolant pressure boundary was very unlikely to occur during the remainder of Davis-Besse's limited operating cycle. If there was a failure of the RPV head due to a CRDM failure, the resulting loss-of-coolant-accident was within the design basis envelop of the plant.
A loss-of-coolant accident during the period prior to the scheduled February 16, 2002, shutdown was deemed very unlikely. Nonetheless, in light of the increased risk in comparison with the base-case probabilistic risk assessment, additional prudent measures were deemed appropriate.
It should be noted that in the absence of the vessel head corrosion, which the staff was unaware of, the inspections performed by the licensee in early March of 2002, found CRDM cracking that was consistent with the staff's assumptions and confirmed the adequacy of the risk-informed decision to allow the plant to shorten its scheduled outage from March 31, 2002, to February 16, 2002.
Would the containment have failed if there was a LOCA and the control rods failed to insert into the core?
Since the event is within the scope of existing analyses for bounding accidents for which the plant safety systems are designed, the containment response would also be within the scope of existing containment analyses. Containment systems would function to cool the containment atmosphere. Passive cooling from condensation of steam on the walls and other internal containment structures would contribute to maintaining the containment atmosphere pressure and temperature within acceptable values.
Therefore, the staff concludes that the containment would not fail.
If the stainless steel clad had failed, how would the containment recirculation sump have performed?
The containment sump performance is considered when evaluating the ability of a plant to mitigate design-basis accidents, including loss-of-coolant accidents (LOCAs), and to establish long term cooling. Long term cooling requires the plant to operate under the mode of recirculation, which involves pumping water from the containment sump to provide inventory for the safety injection and containment spray systems. Recently, the NRC's Office of Regulatory Research identified a credible concern associated with the assumptions used in the licensing of PWRs. This issue is applicable to all PWRs, including Davis-Besse, and is under review as a generic safety issue (GSI-191). The NRC determined that the continued operation of PWRs is justified while the issue is under review because the probability of the initiating event, a large break LOCA, is low. Although more probable, small to intermediate size LOCAs also would not preclude continued operation because they generate smaller quantities of debris and require less ECCS flow, allowing more time for operator intervention.
The NRC did consider the susceptibility of the Davis-Besse plant to sump blockage in its evaluation of the ability to mitigate a LOCA in response to the rupture of the degraded area of the head. Although the probability of a LOCA was elevated by the degraded area, because of its size and the lack of fibrous material in the zone of influence, sump blockage was unlikely to be a significant threat to the ECCS performance of Davis-Besse in the event of such a rupture.
Would there have been sufficient trisodium phosphate (TSP) in the containment to mitigate the effects of the amount of boron found on the top of the reactor pressure vessel head?
With regard to insufficient TSP in the containment, the NRC staff has performed independent calculations using conservative assumptions of boron sources including the leakage from the reactor coolant system and the amount of boron on the top of the reactor pressure vessel head. The Davis-Besse technical specifications require a minimum of 290 cubic feet of TSP. Using conservative assumptions, results from the NRC staff calculations indicate the effectiveness of the TSP would not have been compromised by the presence of the additional boron.
Why was the modification of the service structure at Davis-Besse deferred?
The initial decision to cancel the modification in 1992 was made by licensee management. The decision to fund the modification as proposed again in 1994 was made by the licensee's Project Review Group in November of 1998. After funding approval, the licensee scheduled installation of the service structure modification for the 2002 refueling outage. Prior to the 2002 outage, based on budgetary concerns and the expected replacement of the head in 2004, the modification was deferred to the 2004 outage.
The NRC is not part of the licensee's approval process for plant modifications such as that proposed for the service structure. The licensee is authorized to make certain changes to their facility without prior NRC approval as discussed in 10 CFR Part 50.59 "Changes, Tests, and Experiments." As permitted under 10 CFR 50.59, this particular modification likely would not require NRC approval.
How many other licensees have deferred and/or never undertaken similar modifications to assure access to their service structures?
The service structure is a feature of plants designed by Babcock & Wilcox (B&W). In response to the Davis-Besse situation, the NRC has contacted the B&W licensees and verified that all (except Davis-Besse) have made the necessary modifications to ensure access to their service structure.
Page Last Reviewed/Updated Tuesday, March 10, 2020