ATTACHMENT 2 United States general Accounting Office Fact Sheet for the Chairman, Committee on Governmental Affairs, U.S. Senate NUCLEAR HEALTH AND SAFETY Examples of Post World War II Radiation Releases at U.S. Nuclear Sites OSD SASE = 9350-A Section 1 The Green Run Test and Its Safety and Health Implications Details of the Green Run test and its historical context indicate that it was an atomic energy intelligence collection experiment. The test occurred during a period of heightened interest in Soviet nuclear capabilities, shortly after the first Soviet nuclear bomb detonation. The test was not considered unsafe at the time, when radiation protection standards were generally less stringent than they are today. However, at some locations, the release exceeded then-existing local Hanford, Washington, tolerances for deposition in vegetation and animal tissue, and it may not have been permissible under today's nuclear safety standards. Presently, potential health effects from the test and other iodine releases at Hanford during the 1950s are being addressed in an ongoing dose reconstruction study. A classified report of the test was issued in May 1950 by the former Atomic energy Commission (AEC), but the report remained classified in its entirety -- and the test remained undisclosed - - for almost four decades. Details of the test and concerns about its potential health and safety effects first surfaced in the latter part of the 1980s. When references to the test appeared in other AEC documents that were declassified over the years, several Green Run-related Freedom of Information Act requests and appeals were filed. As a result, the test report was largely declassified in 1989. (Several passages in the report remain classified by determination of the Air Force, on the basis that further declassification of the report could compromise Air Force missions and thereby damage the national security). Test Purpose and Historical Context The Green Run test was a special test of detectibility as well as a research experiment into the atmospheric diffusion of radioactive gases. As such, it was related to postwar classified AEC/military research into the nature and effects of radioactive fallout and bomb debris. Test Purpose The Green Run test was conducted at Hanford, Washington, on December 203, 1949, by the AEC and the Air Force. The test took place in a postwar climate of U.S. concern about Soviet nuclear capabilities following the first detected explosion of a Soviet nuclear weapon in August 1949. According to a test participant, a premise of the test was that aerial monitoring and sampling of a radioactive cloud, even long distances from the source, could give evidence of nuclear materials. The diffusion of the released gases was to be monitored in order to develop air, ground, and aquatic methods of collecting data on nuclear operations and weapons Page 6 GAO RCED 94-51 Radiation Releases Section 1 The Green Run Test and Its Safety and Health Implications tests. The radioactive cloud was generated by a special spent fuel reprocessing operation. For the test, the plant's radiation emission control procedures were intentionally relaxed. The spent fuel used in the test was aged about 16 days instead of the usual longer period of up to 90 or more days, which accounts for the term 'green' run (i.e. the test involved the reprocessing of "green" full). In addition, the plant's off-gas water scrubbers -- used to minimize the release of radioactive off-gases from the stack -- were not operated. According to the test report issued in May 1950, as a result of these steps, the test released about 27,800 curies of radioactive production off-gases, including about 7,800 curies of iodine and about 20,000 curies of less hazardous xenon, into the atmosphere in southeast Washington and Oregon. The total recorded iodine release was about twice the almost 4,000 curies predicted in pretest calculations. During the test, despite unexpected adverse weather patterns that developed and limited the range of diffusion, the radioactive cloud was detected by an aircraft over 100 miles northeast of the site. After the test, radioactive iodine was found on vegetation over large areas of southeast Washington and Oregon, as shown in figure 1.1. Page 7 GAO RCED 94-51 Radiation Releases Section 1 The Green Run Test and Its Safety and Health Implications Figure 1.1: Areas Where Radioactive Iodine Was Found on Vegetation Following the Green Run Test. FOR REFERENCE SEE (8bb27.gif) Historical Context As a research experiment into atmospheric diffusion, the test was related to postwar classified AEC/military research into the nature and effects of radioactive fallout and bomb debris. Such research began as early as the Operation Crossroads test series in the Pacific Ocean in 1946 -- during which, fallout was monitored aerially by the Air Force and on the surface by naval vessels -- and continued throughout succeeding Page 8 GAO RCED 94-51 Radiation Releases Section 1 The Green Run Test and Its Safety and Health Implications atmospheric-testing sites. Effective instrumentation was an important aspect of research into radioactive effects, and at the time of the Green Run event the AEC and the military services were conducting several field instrument development programs to support their nuclear weapons research efforts. According to a test participant, the test was also generally related to research in to the safety and health effects of nuclear detonations and nuclear production operations. The Green Run test was preceded by other aerial radiation- monitoring tests that involve routine production releases of radioactive materials. The test was a follow-up to a series of aerial-monitoring tests conducted by the Air Force and the AEC during November 1948 to March 1949 at Oak Ridge, Tennessee, and at Hanford. For these tests, no special releases were conducted. The tests involved monitoring off-gases from routine production operations. At Oak Ridge, during 20 overflights by a C-17 aircraft between November 1948 and February 1949, reactor and separations off-gases were tracked up to 17 miles downwind. At Hanford in March, during three similar overflights, routine separations off-gases (with stack scrubbers in operation) were detectable for less than 2 miles -- results considered so disappointing that further Hanford overlifts were discontinued. In a report on the test series, the authors concluded that further use of similar Hanford operations as a source for aerial tracking was not practicable. Logically, the Green Run test -- with Hanford scrubbers not operating -- provided the needed stronger source.1 In addition, according to a former AEC official, monitoring overflights for the purpose of cloud tracking were conducted wherever sources of atmospheric radiation could be found in the United States, and probably at most or all AEC nuclear production sites. Routine close-in monitoring overlifts at AEC sites began in the early 1950s and developed into a regular monitoring program having, among other things, environmental, safety, and security and safeguards purposes. Also, aerial radiation monitoring by Air Force aircraft was practiced in conjunction with the many nuclear bomb tests conducted at the Nevada Test Site and in the Pacific Ocean during the late 1940s and throughout the 1950s. For example, according to one source, during Operation Sandstone in the pacific in April-May 1948, a fallout-tracking test called Operation 1Also in 1949, at an undetermined time before July 29, aerial monitoring tests of routine production effects were conducted at the Hanshaw Uranium Refining Plant in Cleveland, Ohio. Overflights detected particles, likely uranyl flouride, 1,150 years downwind from the source in concentrations of 0.71 micrograms per cubic meter. Also in 1949, on an undetermined date, aerial effluent monitoring of the ILLEGIBLE Uranium Refining Plant in St. Louis, Missouri, detected uranium concentrations of ILLEGIBLE micrograms per cubic meter in the atmosphere 3,000 feet downwind from the plant. Page 9 GAO RCED 94-51 Radiation Releases Section 1 The Green Run Test and Its Safety and Health Implications Fitzwilliam monitored radioactive fallout gases for several thousand miles at levels many times above background levels. Safety and Health Implications Some routine Hanford radiation safety procedures were intentionally relaxed for test purposes. Specifically, in order to calibrate means of detecting Soviet production from Hanford plant operations, the cooling period for Hanford spent fuel was shortened from 90 or more days to only 16 days to simulate presumably less efficient or careful soviet operations, and separations off-gas scrubbers were not operated. Furthermore, while the release was conducted on a weekend, which may have limited the number of workers on-site, the offsite populace was not forewarned of the event or made aware of it for several decades. The test was also conducted despite less-than-optimal weather conditions, which limited the test results and may have exposed greater-than-expected numbers of the population to the radioactive cloud prevailing wind patterns prior to the test had been inopportune, and wind shifts during the test caused the emission of gases close to the ground, including directional shifts over populated areas in southeast Washington and greater- than-expected deposition at the Hanford site. Because of shifting winds, long-distance tracking of the cloud for several hundred miles was not possible. Two AEC contractor officials responsible for conducting the test differ in their recall of who decided that the weather for the test was acceptable. According to one, AEC contract officials judged the weather to be acceptable. According to the other, the AEC did not wish to proceed, but the Air Force made the decision to conduct the test.2 The recorded total release of iodine 131--about 7,800 curies--was about 2 time the predicted quantity. However, the accuracy of the recorded amounts has been questioned, and they have been recalculated.3 According to officials conducting the test, the amount of the release was not considered unsafe at the time. While the release was extremely concentrated, since it occurred over a 12-hour period, regulatory limits on the amount of such emissions did not exist at the time. In fact, the release was a small fraction of the total releases that occurred during wartime and immediate postwar Hanford operations, before radioactive iodine removal 2 The AEC's Hanford contractor, General Electric Company, had a Health Instruments Division with the day-to-day authority to decide when reactor field could be processed. 3 In June 1992, the Journal Health Physics, Maurice Robkin, a participant in the Hanford Dose Reconstruction Project, estimated the amount of iodine released to be about 11,000 curies, well over twice the predicted quantity. He calculated the release of xenon to be about 16,000 curies for a total of about 27,000 curies. Page 10 GAO RCED 94-51 Radiation Releases Section 1 The Green Run Test and Its Safety and Health Implications systems were installed. For example, during 1945, production releases estimated at over 45,000 curie of iodine per month occurred at Hanford. By one estimate, the Green Run test accounted for about 1.1 percent of the total radioactive iodine released during 1944-49. Test participants said the release was considered to be well within the standards of the time for human exposure to radiation.4 In some locations, the release reportedly exceeded then-existing local Hanford limits for radioactive deposition in animal tissue and vegetation. According to the test report, the release resulted in iodine deposition in animal thyroids up to 80 times above the limit of 4 microcuries per kilogram of tissue. The then-existing local Hanford tolerance for continuous deposition on vegetation -- 9 microcuries per kilogram -- was temporarily exceeded in the areas of Yakima, The Dalles, Spokane, and Blue Mountains. Based on post-test documentation available to us, radiation doses that the off-site population might have received as a result of the test were not estimated at the time. In regard to today's more stringent radiation standards, which are not directly comparable to those of the 1940s, it has not been determined whether the test exceeded present limits for off- site radiation doses and emissions.5 The effects of the Green Run release and other postwar Hanford radioactive iodine releases that may have had effects on the off-site population are being addressed in an ongoing dose reconstruction study, directed by the Centers for Disease Control, focussing on Hanford operations and releases from the site's beginning in 1944.6 In regard to deposition standards that exist today, post-test deposition on vegetation in Richland, Walla Walla, and Pendleton reached levels about the threshold of 50 picocuries per gram listed in recent Environmental Protection Agency guidance for the interdiction of foodstuffs, applicable to accidents 4 At about the time of the test, the National Committee on Radiation Protection -- whose recommendations at AEC followed -- recommended (but did not immediately publish) a public external dose limit corresponding to about 1.5 rem (roentgen equivalent man) annually, or 10 percent of its re commended worker limit of about 15 rem annually. We were unable to document a then-existing specific limit of about 15 rem annually. We were unable to document a then-existing specific limit for internal radioactive iodine doses. Rem is a measure of the dose of any ionizing radiation to body tissues in terms of as estimated biological effect relative to a dose of 1 roentgen of X-rays. 5 Per 40 C.F.R. 61.92, applicable to he Department of Energy under departmental order 5400.5, ILLEGIBLE pathway radiation doses to the off-site populace are limited to 0.01 rem annually. 6 Preliminary dose estimates from the study indicate that,a during 1945-47, when routine Hanford iodine releases were conducted that totalled up to several dozen times more than the Green Run release, doses exceeding present limits may have been recurred downward infants through the air-pasture-cow-milk- thyroid pathway. According to DOE at the time scientists had not identified this as a pathway for significant doses of radioactive iodine to individuals. Page 11 GAO RCED 94-51 Radiation Releases Section 1 The Green Run Test and Its Safety and Health Implications or other mishaps at both civilian and Department of Energy (DOE) nuclear plants.7 Furthermore, if proposed today, the test (including procedures that intentionally increased the amount of the release) might not be permissible under the principle of limiting radiation effects from nuclear production operations to levels "as low as reasonably achievable" (10 C.F.R. 20.1, and DOE Order 5400.5). This principle was not operative in 1949, at the time of the Green run test. In addition, if proposed today, such a test would appear to be imprudent from the point of view of operational safety procedures. DOE has categorized the test as one of the 14 most significant safety-related incidents in Hanford's history. Our work did not document that the test was intended to be a radiation warfare experiment or a field test of radiobiological effects on humans. In particular, we examined still-classified passages in the Green Run test report and found that they did not refer to any such intentions or operations. 7EPA Manual For Protective ILLEGIBLE for Nuclear Incidents, No. ILLEGIBLE 75-001A, Jan. 1990 Page 12 GAO RCED 94-51 Radiation Releases