Below is a substantial portion of
ISSUES REGARDING THE STORAGE OF SPENT NUCLEAR FUEL
prepared by security expert Gordon R. Thompson as a panelist at a public workshop held by the California Energy Commission in Sacramento, California, 15-16 August 2005
6. Vulnerability of nuclear facilities to attack
Nuclear power plants and ISFSIs have vulnerabilities that arise from intrinsic factors and from design choices. The intrinsic factors derive from the basic processes and structures needed to harness nuclear fission. Notably, spent fuel from a fission reactor necessarily contains biologically harmful and heat-producing radioactive material (e.g., cesium-137). Also, reactor structures and nuclear fuel employ chemically-reactive materials. In US commercial reactors, the fuel cladding is made of zirconium alloy that can react exothermically with air or steam. The latter reaction yields hydrogen that can form aninflammable or explosive mixture.
The intrinsic vulnerabilities have been exacerbated by design choices. Two policy decisions have been especially important in this respect. First, resistance to attack has not been a design goal. Second, the NRC has allowed the nuclear industry to employ cost saving measures that have created vulnerabilities. Four examples illustrate the combined influence of these factors on the vulnerability of present US nuclear facilities, as follows:
• Example #1: Spent-Fuel Pool Fires
Spent-fuel pools are now equipped with high-densit density racks so that they can hold a much larger inventory of spent fuel than was envisioned when the pools were designed. The high-density racks have a closed configuration that is necessary to suppress criticality. As a result, loss of water from a pool would cause the spent fuel to heat up and, across a wide range of scenarios, experience a runaway zirconium-air or zirconium-steam reaction (i.e., a fire). The resulting heatproduction and fuel degradation would release a large amount of radioactive material to the atmosphere.
• Example #2: Cascading Failures
Spent-fuel pools are immediately adjacent to reactors, and share their support systems. At many sites, including SONGS and Diablo Canyon, reactors are adjacent to each other and share support systems. The resulting interdependence means that fires, radiation fields and other effects of an attack could preclude operation of active safety systems or implementation of damage-control measures (e.g., provision of water makeup or spray to a drained spent-fuel pool), leading to cascading failures.
• Example #2: Cascading Failures
Spent-fuel pools are immediately adjacent to reactors, and share their support systems. At many sites, including SONGS and Diablo Canyon, reactors are adjacent to each other and share support systems. The resulting interdependence
means that fires, radiation fields and other effects of an attack could preclude operation of active safety systems or implementation of damage-control measures (e.g., provision of water makeup or spray to a drained spent-fuel pool), leading to cascading failures
• Example #3: Reliance on Active Safety Systems
At nuclear power plants, safety systems are typically active rather than passive, and rely on AC or DC electric power. Achieving grid disconnect could be easy for attackers, forcing reliance on potentially vulnerable onsite power sources.
• Example #4: Vulnerable ISFSI Modules
The dry-storage modules used at ISFSIs are not designed to resist attack. At all recently-established ISFSIs in the USA, spent fuel is contained in metal canisters with a wall thickness of about 1.6 cm. Each canister is surrounded by a concrete overpack, but this overpack is penetrated by channels that allow cooling of the canister by convective flow of air. Attackers gaining access to an ISFSI could employ readily-available skills and explosives to penetrate a canister in a manner that allows free flow of air to spent fuel, and could use incendiary devices to initiate burning of fuel cladding, leading to a release of radioactive material to the atmosphere.
A determined, sophisticated group planning to attack a nuclear facility could employ a variety of modes and instruments of attack. To illustrate the vulnerability of nuclear facilities to available instruments, consider the potential for penetration of reinforced concrete structures. Such penetration could be sought in some attack scenarios. Reactor containments and spent-fuel pools are relevant structures. At a typical PWR, the reactor vessel and associated components are inside a cylindrical, reinforced-concrete containment with a wall about 1 m thick. The adjacent spent-fuel pool has reinforced concrete walls up to 2 m thick. An informed attacker is likely to consider a shaped explosive charge as an instrument forpenetrating a structure of this kind.It is, therefore, noteworthy that the US government has published design details for a shaped-charge, cruise-missile warhead intended to penetrate rock or concrete. The warhead's purpose is to open a pathway for entry of a second, tandem-mounted charge. This warhead has a diameter of 71 cm, a length of 72cm, and a total mass of 410 kg. When tested in 2002, it created a hole of 25 cm diameter in tuff rock to a depth of 5.9 m.18
One means of carrying such a device would be a general-aviation aircraft operated remotely or by a suicidal pilot. There are many suitable aircraft. For example, a Beechcraft King Air 90 will carry a payload of up to 990 kg at a speed of up to 460 km/hr. A used King Air 90 can be purchased for US$0.4-1.0 million. Note that there are more than 19,000 airports in the USA. Also, during the period 1998-2003, about 70 aircraft were stolen from general-aviation airports in the USA.
7. Consequences of attack
A successful attack on a reactor or a spent-fuel pool could release radioactive material to the atmosphere by exploiting mechanisms that would be powered by energy sources within the facility -- stored heat, radioactive decay heat, and exothermic chemical reactions (e.g., zirconium-air or zirconium-steam). At a spent-fuel pool, the release could include 10-100 percent of the cesium-137 in the pool, together with other radioactive isotopes. Analyses of reactor accidents suggest that a successful attack on a reactor could also achieve a cesium-137 release fraction of 10-100 percent.21 The reactor release would, in addition, include short-lived radioactive isotopes such as iodine-131.
An attack on a dry-storage module of an ISFSI could potentially achieve a cesium-137 atmospheric release fraction of 10-100 percent. However, achieving this outcome would require the use of an incendiary device to initiate a zirconium-air reaction, and the availability of air to feed that reaction.