64 « The Containment of Underground Nuclear Explosions Figure 4-4—Yieid v. Distance 1.000 f= _ LL LL 7 percent of the released material was deposited.’© Therefore, it is thought that Pike is actually a more conservative model than Baneberry. Thesensitivity of the Pike model can be judged by looking at the degree to which its predictions are affected by the amount of material released. For example, consider a test in which 10 percent of the radioactive material produced by the explosion is accidentally released into the atmosphere; in other words, 10 percent of the material that would have been released if the explosion had been detonated aboveground. This also roughly corresponds to the amount of material that would be released if the explosion had been detonated underground at the bottom of an open (unstemmed) hole. The 10 percent Total ist year Total tst exposure exposu 500 mR 100 E- ge a of of 2 bf 10 0 50 ACCIDENT NOTIFICATION Any release of radioactive material is publicly announced if the release occurs during, or immediately following, a test. If a late-time seep occurs, the release will be announcedif it is predicted that the 100 Distance (miles) release can therefore be used as a rough approxima- tion for the worst case release from an underground test. To evaluate the adequacy of the Pike model predictions to withstand the full range of uncertainty of an accidental release, the question is: what effect would a release of 10 percent rather than, say ! percent, have on the location of 170-mR and 500-mR exposure lines? As figure 4-4 illustrates, changing the yield of an explosion by an order of magnitude (in other words, increasing the release from say 1 percent to 10 percent) increases the distance of the 170-mR and 500-mR lines by roughly a factor of 2. Therefore, assuming a worst case scenario of a 10 percent prompt massive venting (as opposed to the more probable scenario of around a | percent prompt massive venting), the distance of the exposure levels along the predicted fallout lines would only increase by a multiple of 2. The Pike model therefore provides a prediction that is at least within a factor of about 2 of almost any possible worst-case Scenario. 170 mi Te | a prompt venting. Baneberry vented through a fissure and decaying radioactive material was pumped out over many hours. Baneberry released more curies than Pike: however. due to its slower release, a higher percentage of the Baneberry material was in the form of noble gases. which are not deposited. The data suggest that much less than Constant Pike Parameters Variab Wind speed ~ 15mph Vertical wind shear = 20° Cloud rise = 5,000ft Yield * Yield (in kilotons) v. distance (in miles) tor projected fallout the Pike Model. TYEindicates total first year exposure. Incre the yield by a factor of 10 roughly doubles the downwind dist of the projected fallout pattern. SOURCE: Provided by National Oceanic and Atmospheric Administ Nationa) Weather Service Nuclear Support Office, 1988 radioactive material will be detected outside boundaries of thetest site. If no detection off-si predicted, the release may not be announ Operational releases that are considered rot (such as small releases from drill-back operati are similarly announced only if it is estimated they will be detected off-site. The Environmental Protection Agency is pre at every test and is therefore immediately awa any prompt release. The Environmental Prote: Agency, however, is not present at post-test « back operations. In the case of late-time releas operational releases, the Environmental Prote: Agency depends on notification from the De ment of Energy and on detection ofthe release( ‘OBaneberry, however, had a limited data set of usable radioactive readings.