(First let's get the objection to bringing up Chernobyl out of the way. We do not use the technology that was used at Chernobyl so any failure ON THE SCALE of Chernobyl would have a different cause. However if we have a failure on that scale, given the proximity of many nuclear reactors in this country and the broad geographic distribution of the anticipated damage, such an occurrence here would be far more damaging both to life and the economy than what happened in the Soviet Union.)Does the diminishingly small statistical chance of a Chernobyl scale nuclear accident *effectively* equal ZERO chance of such an accident occurring?
If it doesn't, what does it mean?
The basic argument being made regarding safety is that the statistics show such a low probability of this occurring that it is effectively ZERO. They never outright assert zero risk, but the implication is unmistakable. That is a claim to have PERFECTED, in the sense of operational safety, the extremely complex systems that are nuclear power plants.
The following article is about an event that occurred in 2002 which demonstrates that the claim and the supporting statistics are not accurate mathematical representations of reality.
Davis-Besse Retrospective
On March 6, 2002, workers repairing a cracked control rod drive mechanism (CRDM) nozzle at the Davis-Besse Nuclear Power Station in Ohio discovered a football-sized cavity in the reactor vessel head.<1> Their finding is linked to two other discoveries 15 years earlier. On March 13, 1987, workers at Turkey Point Unit 4 in Florida discovered that a small leak of borated water had corroded the reactor vessel head. Their revelation prompted the Nuclear Regulatory Commission (NRC) to require all owners of pressurized water reactors, including Davis-Besse, to take specific measures to protect plant equipment from boric acid corrosion.<2> On March 24, 1987, the NRC learned that control room operators at the Peach Bottom Atomic Power Station in Pennsylvania had been discovered sleeping while on duty. That revelation prompted the NRC to issue an order on March 31st requiring Peach Bottom Unit 3 to be immediately shut down.<3>
The three findings spanning 15 years are intertwined. Turkey Point demonstrated that a small amount of boric acid leaking onto the reactor vessel head corrodes carbon steel at a high rate. Had the FirstEnergy Nuclear Operating Company, the owner of Davis-Besse, remembered Turkey Point's lesson, the serious damage at Davis-Besse would have been averted. Peach Bottom demonstrated that a pervasive safety culture problem creates unacceptable conditions for operating a nuclear power plant. Had NRC remembered either Turkey Point's or Peach Bottom's lesson, they would have issued the order they drafted to shut down Davis-Besse. It would have been the first shut down order issued by the agency since the Peach Bottom order. But both FirstEnergy and the NRC forgot the past and relived the wrong event from March 1987 by having yet another reactor vessel head damaged by boric acid corrosion.
Many individuals, from both within and outside the NRC, have accused the agency's move towards risk-informed decision-making as the reason for its failure to issue the order to shut down Davis-Besse. On the contrary, the NRC's handling of circumferential cracking of control rod drive mechanism (CRDM) nozzles as reported by the Oconee nuclear plant in February 2001 was a successful demonstration of proper application of risk-informed decision-making ¾ with the sole and significant exception of its mistake in not issuing the shut down order for Davis-Besse. But even that mistake, as bad as it was, does not impugn the risk-informed decision-making process for the simple reason that the NRC deviated from that process. Had the NRC adhered to its risk-informed decision-making process, it would have issued the shut down order for Davis-Besse and capped off a stellar example of how this process can and should be used.
In February 2001, the NRC learned of a new aging mechanism, the circumferential cracking of stainless steel CRDM nozzles based on inspection results from Oconee. The NRC properly reacted to this finding by revisiting the nuclear industry's inspection regime for CRDM nozzles. It determined that the existing inspection regime did not provide adequate assurance that circumferential cracks would be identified and repaired. The NRC did not require all plant owners to immediately address this inspection shortfall, which would have imposed an unnecessary regulatory burden on those plants with low susceptibility for the problem. Nor did the NRC allow all plant owners to address the shortfall at their next regularly scheduled refueling outage, which would have imposed an unnecessary challenge to safety margins at those plants with high susceptibility. Instead, the NRC applied risk-informed decision-making by issuing Bulletin 2001-01 in August 2001 to all owners of pressurized water reactors. This Bulletin required the high susceptible reactors to resolve the inspection shortfall by December 2001, the medium susceptible reactors to resolve the inspection shortfall at their next regularly scheduled outage, and merely collected information from the low susceptible reactors.
Only two reactors with high susceptibility for circumferential cracking of CRDM nozzles did not conform to the inspection requirements of Bulletin 2001-01: D.C. Cook Unit 2 and Davis-Besse. The NRC applied risk-informed decision-making to these two exceptions. The NRC determined there was adequate assurance that D.C. Cook Unit 2 could safely operate until its scheduled refueling outage beginning January 19, 2002. The NRC determined it lack adequate assurance that Davis-Besse could safely operate past December 31, 2001, and sent a proposed shut down order up to its Commissioners.
At this point, the NRC abandoned its risk-informed decision-making process. The FirstEnergy Nuclear Operating Company, the owner of Davis-Besse, was prepared to contest an order. Rather than fight, the NRC switched to a compromise where Davis-Besse would be allowed to operate until its rescheduled refueling outage beginning February 16, 2002. To balance the increased likelihood of a meltdown, the NRC required FirstEnergy to implement a number of compensatory measures at Davis-Besse.
Had the NRC issued the shut down order, it would not have prevented the gaping cavity in the reactor vessel head at Davis-Besse. It would simply have meant that the cavity would have been found sooner. More importantly, it would have meant that the reactor would have been operated for less time with that serious problem. Most importantly, it would have been abiding by the risk-informed decision-making process under all conditions, not just when it is convenient and popular.
Notes
<1> Nuclear Regulatory Commission, Preliminary Notification of Event of Unusual Occurrence, PNO-III-02-006, "Significant Metal Loss Observed n Reactor Vessel Head," March 8, 2002.
<2> Nuclear Regulatory Commission, Generic Letter 88-05, "Boric Acid Corrosion of Carbon Steel Reactor Pressure Boundary Components in PWR Plants," March 17, 1988.
<3> Letter dated May 12, 1987, from William F. Kane, Director – Division of Reactor Projects, Nuclear Regulatory Commission, to J. W. Gallagher, Vice President – Nuclear Operations, Philadelphia Electric Company, "Combined Inspection 50-277/87-10; 50-278/87-10."
Open access article available for download here:
http://www.ucsusa.org/nuclear_power/nuclear_power_risk/safety/davis-besse-retrospective.html#Note2You can also find more detail with photos at the other download on that page: Davis-Besse: The Reactor with a Hole in Its Head.