Concern has been expressed by IMO member states and concerned citizens that the regulations for the safe transport of radioactive material (RAM) as listed in IAEA ST-1 do not provide adequate protection for RAM packages during sea transport. To support this position, the extreme amount of kinetic energy available in ship-to-ship collisions is frequently cited.
This amount of kinetic energy is orders of magnitude larger than the kinetic energy that is associated with the regulatory impact test that packages are required to withstand in order to be certified. The fallacy of this argument is that, in the regulatory impact accident, all of the kinetic energy of the event is transmitted to the package, while in ship-to-ship collisions only a small fraction of the kinetic energy (or as is often the case, none) is transmitted to the package. However, it must still be demonstrated that the probability of maritime transport accidents exceeding the regulatory hypothetical accident conditions is no higher than the probability of land transport accidents exceeding these conditions. The probability of a release of radioactive material due to a collision at sea is the product of four probabilities:
(1) the probability of having a collision,
(2) the probability of the collision being in the correct configuration to cause damage to the RAM package,
(3) the probability the collision is severe enough to either penetrate to the location of the RAM package or compress cargo around the package, and
(4) the probability that the forces acting on the RAM package are high enough to cause the package to fail.
In this section the last three of these probabilities is discussed, the first having been discussed in Section 5.2.1.
6.2. Mechanics of ship-to-ship collisions
Legal operation of commercial ships does not allow cargo to be stowed in the most forward portion of the ship (ahead of the collision bulkhead). This fact ensures that, if a ship carrying radioactive material package(s) strikes another ship, the package(s) on the striking ship will not be involved in the collision. Also, if a ship carrying the radioactive material package(s) is struck by another ship during a raking collision (a collision with a small angle between the paths of the two ships), the penetration into the struck ship will not be sufficiently deep to involve the RAM package(s) in the collision. Therefore, the only collisions that have a possibility of involving the radioactive material package(s) are nearly right-angle collisions in which the ship carrying the package(s) is struck by another ship. When combined, these factors show that only about one collision in ten will have a configuration that might cause damage to a RAM package. Even if the collision is of a configuration to cause damage to the RAM package, it must still be sufficiently severe to either penetrate the struck ship to the location of the package or compress cargo (or intervening ship structure) around the package.
The severity of the collision is a function of the masses of the ships involved and their relative speeds. In this type of accident, the initial kinetic energy of the two ships is dissipated by deformation of ship structure and by hydrodynamic forces acting on the ships. The exact solution of this mixed structural dynamics and fluid dynamics problem is generally not performed. Instead, approximations on the effect of the hydrodynamic forces are made. One frequently used approach is to account for the hydrodynamic resistance by increasing the mass of the struck ship to increase its inertial resistance. This approach allows the collision to be treated in a purely structural mechanics setting.
The amount of damage to a ship struck by another ship is proportional to the kinetic energy of the collision. In the 1950s V.U. Minorsky investigated a large number of severe collisions and developed an empirical linear equation relating the collision energy absorbed by the two ships (calculated by assuming a perfectly plastic collision) to the volume of ship steel damaged [12].
Recently, Reardon and Sprung [13] developed new constants for this linear equation as the result of additional data gained from more recent collisions (beginning in 1959). They also considered the energy absorbed by the crushing of ship cargo. Using this equation it is possible to determine approximately the penetration distance for any collision. Applying this methodology to a wide range of collisions, Reardon and Sprung determined that only about 15 per cent of all collisions occur with a configuration that may result in crush forces being applied to a RAM package for transportation in a small freighter without other cargo, and about 30% for transportation in a larger break-bulk freighter with other cargo stowed in the same hold as the RAM package.
The structure of the Minorsky equation is such that it treats the energy required to penetrate the hull of the struck ship as a constant (the y-intercept of the correlation). For the smaller ships that may be used to transport RAM, and especially small double hull purpose-built ships, treating the energy to penetrate the hull as a constant can lead to substantial errors. The extensive amount of work done to investigate loss of containment for tanker ships has led to a robust method for determining the energy and penetration distance required to cause hull rupture [14]. This method can be used to develop ship specific penetration and energy absorbed at the moment of hull rupture. Further penetration can be treated in a manner similar to that developed by Minorsky. A discussion of this method is included in the technical appendices of this report. No attempt has been made to improve the probabilities for imposing structural loads on RAM packages derived by Reardon and Sprung using this improved technique.
6.3. Magnitude of forces acting on RAM packages
The results of the preceding section only indicate that a RAM package may experience forces as a result of ship-to-ship collision. In this section we will consider the magnitude of those forces. If a collision is severe enough to cause penetration to the location of a RAM package, there are two ways in which the package might be damaged. First, by direct contact of the striking ship bow with the package. However, because the bow velocity and bow rigidity during this impact are much less than those for the regulatory impact, no damage is expected as a result of direct impact of the striking ship bow onto the package. Second, the penetrating bow might subject the package to crush forces either by crushing cargo around the package or by pushing the package up against ship structures, for example, the far hull of the struck hold.
In order for a package to be subjected to crush forces, there must be forces acting on both sides of the package. If it is assumed that the bow of the striking ship can supply infinitely large forces on one side of the package (rigid bow assumption), the limit to the magnitude of the crush force is the force restraining the package from moving ahead of the advancing bow once the package has been struck.
When the package is first impacted by the bow of the striking ship it is held in place by the tiedowns that attach it to the deck. These tiedowns are required by regulation to be designed to fail at forces much lower than the force needed to damage the package, and they are also required to fail in a manner that does not decrease the functionality of the package with respect to containment, criticality control, or shielding effectiveness. For this reason, it is impossible for the tiedown system to provide restraining force of a magnitude large enough to cause package damage.
Immediately after the tiedowns fail, the only force restraining the package is friction along the deck. Even if the collision has caused the deck to fold or buckle, the magnitude of this friction force is very small. Therefore, the package will slide/roll across the deck until it strikes other cargo or ship structure. When cargo is present, because most cargo is softer than a RAM package, contact with other cargo will result in crushing of cargo rather than crushing of the RAM package. A paper by Radloff and Ammerman [15] shows finite element calculations of the maximum crush force imparted to a radioactive material package as a function of cargo strength and stiffness. In that paper, it was shown that the presence of intervening cargo has little effect on the magnitude of the maximum load acting on the package. For the case with no other cargo, the RAM package will be pushed across the ship until it strikes some other ship structure (typically the hull on the side of the ship away from the collision). The strength of this structure will limit the magnitude of the crush force that can be applied to the package.
Ammerman and Ludwigsen [16] used finite element analyses to investigate the maximum crush force that could be applied to the package as it is pushed through the side shell structure of a single-hull ship. This conservative analysis indicated a maximum force that is very similar to the inertial forces experienced during the regulatory impact accident. Therefore, the probability of the forces acting on the RAM package being high enough to cause the package to fail are very low and the expected result of this scenario is the ejection of the package through the far hull of the struck hold into the ocean.
6.4. Conclusions
The work by Radloff and Ammerman, and Ammerman and Ludwigsen indicates that during the maritime transport of radioactive material, the probability of a RAM package experiencing
loads that are greater than the loads experienced during the regulatory certification tests is very low. Therefore, there is no need to require a different set of mechanical certification tests for sea transport than are required for land transport. In actuality, the probability of a package structurally failing as a result of collision during transport at sea is lower than it is for land transport.
7. CONSEQUENCES OF ACCIDENTS TO SHIPS TRANSPORTING