Minutes - 20020502 - Laserfiche WebLink

Browse Search

which are essentially exercises in real time where a mock attacking force would challenge the security force and the security force would have to prevent the attacking force from causing enough damage to cause a meltdown. The way these were carried out were first tabletop exercises and were carried out where the NRC and contractors could observe the plant security strategy sort of like a chess game. That was meant to simulate the passive insider that I talked about before — someone supplying information about the plant security strategy. Then the NRC would choose the particular scenarios that would be run, and they would run the exercises in a matter of a couple of days. The type of adversary that was chosen ranged from the so called Farmer Brown, which is the lone, paranoid, right-wing nut who decides that he wants to threaten the U. S. government by attacking the nuclear power plant with a pitchfork; ranging all the way to a team that had most of the characteristics of the design basis threat. But the full design basis threat was never tested. In fact, one loophole in the OSREs was active insiders they never considered. So the attacking force was never helped by someone in the control room who turned on the other operators and shot them and forced them to manipulate the controls. That is an unrealistic assumption, but that is the state of things. Now, the goal of the OSRE force was destruction of the target set. And target set was defined in a way, which means that of every element of the targets that was taken out, the plant would proceed toward a meltdown and eventually radioactive release. The only way that that could be stopped if the targets that were taken out was if the operators were able to respond heroically in a situation of great uncertainty and try to make repairs in an environment where there are still terrorists around, where there are high radiation levels, doing things that they have not been trained to do. Just a heroic intervention is really the only way that they could respond. So it is assumed that if a target set is taken out, the plant would meltdown realistically. What is interesting to note is that in some plants a target set can be all of the equipment in a single room, so if you set off an explosive in the right place, that in itself is enough to cause a meltdown. That is a scary thought. Now, the goal of the defending force in an OSRE exercise is to prevent all of the targets in a target set from being taken out. That means basically making sure that at least one of them is protected by enough guards who reach the target in time who are protected and who have the appropriate weapons. All those are non- trivial acts, especially taking into account the realism of things like actually getting the weapons you would need in an attack because they are usually locked in storage and they have to be retrieved before the guards can actually use them. So all of these things have to be realistically timed to demonstrate whether or not the guards would succeed. Now, there are a lot of artificial aspects of the OSREs, for instance, they were announced six to ten months in advance, they were rehearsed and trained like a stage play pretty intensively, and a lot of money was spent preparing for them. Even so, the miraculous fact that is pretty well known and has been well documented, is that 46% of the exercises that were run in a ten-year period, and some plants had more than one OSRE since there were only on the order of 65 on average sites during this time period, so some have had more than one. Still, 46% failed, and in at least one of these exercises, the attackers were able to reach and destroy a target set. And in a number of those, the containments were also breached, meaning that the attackers were able to blow open an airlock, and therefore, the last barrier between the radiation resulting from the fuel melting and the environment was also breached, so there was no uncertainty. If there was a meltdown, there would have been radiation released into the environment. Performance did not improve over the life of this program. In fact, in the last two years, prior to September 11`h, the percentage of failures was exactly the same as it was over the whole life of the program, 46%. In fact, the very last exercise that was held in August 2001 was one of the worst that was publicly documented and received the worst score by the NRC since they started scoring and using the current color coding. Lest there be any confusion whether OSREs pass/fail or not, I could quote Chairman Meserve that these are not pass/fail exercises, which is particularly irksome, because I think that we have passed an education law