Jaczko reportedly disagreed with the stakeholder recommendation and believed that Borchardt’s attachment was improperly presented based on procedural grounds. The chairman unilaterally intervened to remove the Borchardt attachment before the task force report was formally transmitted to the commissioners and to Congress. That action, among others, contributed to the internecine conflict that further isolated Jaczko within the agency and ultimately led to his resignation in June 2012.
Although back in July 2011 Jaczko had won the battle over the initial transmittal of the task force report, in subsequent votes he lost the war. Rather than vote directly on the task force’s recommendations, the majority ordered the staff to first produce a series of papers analyzing the recommendations and ranking them in priority. But the commissioners made one decision immediately: the task force’s highest priority should become the NRC’s lowest priority. Recommendation 1, to revise the regulatory framework, was put at the bottom of the list, and the staff was given eighteen months to come up with a proposal for it. In the interim, the commissioners said, the staff should use the NRC’s existing regulatory processes to assess the other recommendations.
From one point of view, this move made sense: it could take years to impose safety fixes if they had to await a new regulatory framework. But on the other hand, imposing new requirements under the old framework would only add to the patchwork of regulations and exacerbate the inconsistencies identified by the task force as part of the problem.
Given that the majority of the commissioners did not believe urgent safety upgrades were needed, the necessity for speed was not their motivation for bumping Recommendation 1 off the priority list. Rather, it was driven by the shared view, expressed in a vote by Commissioner William Ostendorff, that the current system was not “broken.”6
After putting Recommendation 1 aside, the staff proceeded to rank the others in terms of priority by putting them into three tiers. Reevaluating seismic and flood hazards, addressing station blackout risks, and improving containment vents were among the Tier 1 actions that the staff believed “should be started without unnecessary delay.”
However, the decision to proceed without first fixing the regulatory framework meant that the new proposals had to navigate all the old impediments to improving safety. In particular, the backfit rule hung over the task force recommendations like the sword of Damocles. Would the NRC buck tradition and acknowledge that U.S. nuclear plants were not adequately protected in light of Fukushima? If not, then most of the changes recommended by the NTTF would be considered “backfits” and would have to meet the NRC’s convoluted “substantial safety enhancement” and cost-benefit tests. In that case, there was a good chance that none of the recommendations would ever become a requirement.
But the old system threw up even more hurdles for change. Although Fukushima had proven that beyond-design-basis accidents were a real threat, the NRC’s obsolete guidelines still ranked them as very low-probability events, meaning that the calculated benefits of reducing the risk would also be very small. And the benefit could even be zero if the calculation addressed an event that had been left out of the risk assessment models used for the cost-benefit analyses.7
The NRC staff ultimately sided with the task force and recommended the commission approve the safety upgrades on the basis of “assuring or redefining the level of protection . . . that should be regarded as adequate.” But the majority of the commissioners needed more convincing. Early on, Commissioner Kristine Svinicki warned members of the task force that they were intruding into areas where they could quickly find themselves “swimming in the waters of backfit.” In an October 2011 vote, Commissioner Ostendorff asserted that “decisions on adequate protection are among the most significant policy decisions entrusted to the Commission and are not impulsive ‘go’ or ‘no-go’ choices.” This view, endorsed by the commission majority, led to yet more delay.
However, in a March 2012 decision timed for the first anniversary of Fukushima, all five commissioners approved three orders imposing new regulatory requirements without the need to pass backfit tests. Although the outcome was unanimous, each commissioner had gotten to the destination via a different circuitous route. Most contended that their actions were not expanding adequate protection but merely “ensuring” it. It all added up to a lot of confusion. But in the end, except perhaps for courts of the future that may be called on to parse these distinctions, it didn’t really make a difference.
What did matter was how comprehensive and stringent the orders were in addressing Fukushima’s lessons. And in those respects, the orders fell short on specifics.
Part of the problem was that the commissioners had directed that the requirements be “performance based.” The concept behind performance-based regulation is that requirements should not be too prescriptive. The regulator should specify the desired outcome and let the plant owner figure out the best way to achieve it. While advocates of this approach consider it more efficient because it gives owners more flexibility, it actually makes requirements harder to interpret and enforce. The industry is free to write its own playbook, leaving regulators with the burden of figuring out whether or not it meets the regulatory intent of the safety rules.
As a measure of how popular this approach is with the nuclear industry, over the past couple of decades the NEI has actually drafted guidance documents on complying with performance-based regulations, which the NRC, after some negotiation, then approved.
The B.5.b measures, introduced after 9/11 to help workers cope with the aftermath of an aircraft attack, were one case in which performance-based requirements hadn’t worked very well. The industry argued against imposing measures it viewed as too prescriptive or specific, contending that there were simply too many potential disaster scenarios to contemplate. Through their lobbying group, plant owners insisted that they needed maximum flexibility to come up with their own solutions. The NRC relented, imposing only very general requirements for the B.5.b equipment and giving the industry a great deal of leeway to design its own strategies.
However, the industry had not thought through its plans to ensure that they would actually work under real-world conditions, such as high radiation fields, excessive heat, and infrastructure damage. These were the very conditions that contributed to the failure of the Japanese severe accident management measures at Fukushima. Yet in a vote on the Fukushima proposals, the commissioners endorsed the B.5.b process as a model for dealing with beyond-design-basis accidents.
Two of the NRC’s new orders were relatively uncontroversial. One required the installation of “reliable hardened vents” at Mark I and Mark II boiling water reactors so that they could be used under station blackout conditions. Previously plant owners had been allowed to install these vents voluntarily, so that NRC inspectors had no power to review their adequacy or require improvements. This situation, a legacy of the NRC’s timid approach to dealing with severe accidents in the 1980s, provides little confidence that vents at U.S. BWRs would have worked any more effectively than those at Fukushima. The second order called for reliable instrumentation in spent fuel pools, something that could have helped workers at Fukushima better understand the unfolding situation.8
The third order directed plants to develop “mitigation strategies” for beyond-design-basis external events. But the title promised more than the order actually delivered. Plants were required to be “capable of mitigating a simultaneous loss of all alternating current (AC) power and loss of normal access to the ultimate heat sink” for all units at a site. As the task force had recommended, the strategies would involve three phases: using installed equipment, using on-site portable equipment, and using off-site equipment. But where the task force had recommended that specific durations be set for the first two phases—namely, installed equipment should be able to maintain cooling for eight hours and portable equipment for the next seventy-two hours—the order did not specify minimum durations for these phases. Each plant site would propose it
s own.9
IT’S RISKY WHEN INDUSTRY WRITES THE RULES
The NRC’s acquiescence to the FLEX program of rapidly deployable emergency equipment was typical of the way the agency and the industry’s advocacy group, the NEI, have interacted for many years in interpreting regulations and developing compliance documents.
In the early days, when a regulation or order needed interpreting, the NRC would develop a “regulatory guide” and submit it for comment by the affected industry. But over time the roles were often reversed: the industry, coordinated by NEI, would write the first drafts of guidance documents and the NRC would comment. This shift gave the industry far greater power to shape the conceptual basis for regulatory compliance, leaving the NRC to tinker around the edges. The NEI guidance documents would then serve as boilerplate text for use by all applicants and licensees.
The NEI guidance for the FLEX program was such a template. The NRC required all plant owners to submit plans by the end of February 2013 to demonstrate how they would implement the program to comply with the mitigation strategies order. On February 28, Exelon Corporation, which owns seventeen reactors, submitted its plan for the Peach Bottom plant in south central Pennsylvania. Exelon’s proposal was typical of all the plant owners’ responses, closely hewing to the NEI’s guidance. It clearly demonstrated the shortcomings of the FLEX approach.
The Peach Bottom plant, situated beside the Susquehanna River, has two Mark I BWRs closely resembling Fukushima Daiichi Units 2 and 3. Exelon’s plan for dealing with a beyond-design-basis accident there assumed from the get-go that batteries and electrical distribution systems for both AC and DC power would be available. It assumed that off-site personnel called to duty would be able to reach the plant in as little as six hours, not necessarily a realistic assumption in the event of the type of major natural disaster that the emergency plans were being designed for.
The Peach Bottom Atomic Power Station, located on the Susquehanna River in south-central Pennsylvania, has two Mark I boiling water reactors, similar to Units 2 and 3 at Fukushima Daiichi. Emergency response plans for the plant depend upon many assumptions now proven to be unrealistic in the aftermath of events at Fukushima Daiichi. U.S. Nuclear Regulatory Commission
The plan also set time frames for coping with an extended station blackout that fell short of the periods recommended by the NTTF. The task force wanted plants to be able to keep fuel cool for the first eight hours of a blackout using permanently installed equipment and thereafter using portable equipment (such as FLEX items) that was available on-site for the next seventy-two hours. After that, one could assume that additional equipment and supplies delivered from off-site would be available.
For Peach Bottom, Exelon estimated that the batteries needed to run critical equipment, including the RCIC, would last no longer than five and a half hours—but it also asserted that the FLEX generator could be hooked up and ready to start recharging the batteries within five hours. Therefore, there was no need for Exelon to extend the battery capacity.
Exelon’s plan assumed that additional backup equipment from the closer of two Regional Response Centers, located nearly one thousand miles away in Memphis, would arrive at Peach Bottom within twenty-four hours. (Exelon contended that it could keep the plant stable forever without any off-site assistance, with repeated torus venting, but conceded that it would be better to eventually have an alternative, less radioactive method.)
Consistent with NEI’s guidance, Exelon’s plan allowed FLEX equipment to be stored below flood level on the assumption that workers would have time to move it to a safer place “prior to the arrival of potentially damaging flood levels.”
As questionable as these timelines are, they don’t take into account the worst case. Exelon found that if a station blackout occurred during refueling, when the entire core of a unit was in the spent fuel pool, there would be no time for the first coping phase at all: FLEX equipment would have to be set up immediately. Exelon asserted that this was not a problem because even though the pool would start to boil after two and a half hours, the spent fuel would not become uncovered until eight hours had elapsed.
The Peach Bottom FLEX plan is a perfect example of the mind-set that led to Fukushima. It represents industry and regulators scripting an accident with little room for improvisation. If a single assumption fails—say, that workers don’t have time to move FLEX equipment to safety in advance of an impending flood—then all the other barriers would collapse like dominoes. Without the FLEX generator, the batteries would fail after five and a half hours and the RCIC could no longer be counted on to cool the reactors. Operators would eventually lose the ability to vent the containment and it would over-pressurize. Backup equipment, located a thousand miles away in Tennessee—and possibly on the other side of massive floodwaters or earthquake destruction—probably would not arrive in time to save the day. Nobody apparently thought those possibilities were worth considering, even after Fukushima Daiichi.
The order also specified that reactor owners must provide “reasonable protection” of the emergency equipment from external events. But what did that mean? The definition was apparently in the eye of the beholder. For instance, the task force had said the equipment should be protected against beyond-design-basis floods, but the commission’s order contained no such requirement.
In any event, by the time the NRC issued the mitigation-strategies order on March 12, 2012, it no longer seemed to matter much. While the NRC commissioners had been ruminating about adequate protection, the industry was creating its own facts on the ground. Plant sites had already ordered or acquired more than three hundred pieces of major equipment under the FLEX voluntary initiative. “Time is of the essence,” said Charles “Chip” Pardee, then head of power generation at Exelon. Arguing that the industry needed to move forward to meet the NRC’s timeline for safety improvements that were still being formulated, Pardee said, “We are proceeding informed by what the NRC is doing, but [we are] out in front of regulations.”
The FLEX approach, however, arguably did not go nearly as far as was needed. For instance, the industry insisted that its purpose was to prevent core damage. Thus, the program did not contemplate how to use such equipment in the highly radioactive and hot environment that would occur after a core started to melt—even though that was one of the greatest challenges at Fukushima. And FLEX was designed with the assumption that a plant’s electrical distribution systems were functional, also at odds with the reality experienced at Fukushima.
The industry’s eagerness to address the lessons of Fukushima even before the NRC could issue requirements was highly ironic at best. After all, plant owners at the same time were dragging their feet—sometimes for decades—on addressing any number of outstanding safety problems, from fire protection to emergency core cooling system blockages. Getting “out in front of regulations” with FLEX was a brilliant move. The more money the industry was spending on equipment that met its own specifications, the more difficult—politically and practically—it would be for the NRC to require substantially different and more robust measures.
The tactic worked. The FLEX program, for all its flaws, did not conflict with the NRC’s ambiguous mitigation strategies order. After many months of deliberation with the NEI on the guidance document it had prepared for the use of FLEX equipment, the NRC largely endorsed NEI’s approach. The tail had wagged the dog.
Over the summer of 2012, the Japanese government solicited views from citizens as it drafted a new energy policy. The results were decisive, noted the survey organizers: “We can say with certainty that a majority of citizens want to achieve a society that does not rely on nuclear power generation.”
Although some believed that the government’s outreach was more appearance than substance, there were signs that the public’s dissatisfaction with the Japanese establishment might be having an impact elsewhere. In September TEPCO announced the creation of an outside review committee to oversee its nuclear operations. With guidance fr
om the committee, TEPCO’s new management hoped to win approval to restart the seven reactors at Kashiwazaki-Kariwa beginning in the spring of 2013, apparently confident that the government’s effort to craft a new energy policy curtailing nuclear power would go nowhere.
On September 14, 2012, the Noda government unveiled its new energy plan, the keystone of which was the elimination of nuclear dependency by the end of the 2030s. The plan was short on specifics. As the Financial Times noted, it was “a somewhat messy compromise that will delight nobody.” It definitely did not delight Japan’s business leaders, who lobbied the Noda government to retain the status quo. “It is highly regrettable that our argument was comprehensively dismissed,” said the head of the powerful business group Keidanren.10
Noda warned that moving toward a nuclear-free Japan would not be easy. “No matter how difficult it is,” he said, “we can no longer put it off.”
It took all of a week before the energy plan was shot full of so many holes that it was barely recognizable.
Noda, like Kan before him, was struggling for political survival. His popularity ratings had not fallen quite as far as Kan’s had a year earlier, but Noda’s party, the Democratic Party of Japan, had an approval rating in the teens. New elections, expected soon, almost certainly would sweep another party and prime minister into office.
Noda’s energy plan originally called for shutting down reactors at the end of their forty-year operating lives and building no new ones. But the day after its release, Yukio Edano, the trade minister, said the ban on new construction would not apply to three reactors already approved for construction. Even more illogically, Japan would continue to pursue its ambitious plan to begin full-scale operation of the Rokkasho Reprocessing Plant, a facility for reprocessing spent nuclear fuel, even though presumably there would be little reactor capacity to dispose of the many tons of plutonium it would produce each year. Operating Rokkasho would increase the massive stockpile of weapon-usable plutonium that Japan had already accumulated, a major security and safety concern.
Fukushima: The Story of a Nuclear Disaster Page 31