On December 27, 1993, approximately 0.27 millicuries of plutonium was released to the environment from a tank vent system through the F-Canyon 291-F stack at Savannah River while transferring solution between two underground tanks (ORPS Report SR--WSRC-FCAN-1993- 0065). Facility personnel estimated the release equated to an off-site dose of 0.0021 millirem. The DOE limit for dose to the public is 100 millirem in one year. Although the amount of plutonium released did not present a substantial radiological hazard and a Type B investigation was not required by DOE procedures, DOE personnel initiated such an investigation because of the potential for a larger release and because of concerns regarding operational weaknesses leading to the event.
Facility personnel were emptying a tank containing a strong nitric acid solution with low levels of plutonium. Emptying these types of tanks is unusual, in part, because the steam jet system normally used to transfer tank contents is very inefficient at low tank levels, causing the transfer to take longer and use more steam than normal. When the tank contents reached lower levels, this inefficiency resulted in elevated temperatures in both tanks, which increased tank vapor pressures and caused more vapors containing plutonium to be released to the vent system. The system did not contain High Efficiency Particulate Air (HEPA) filters because radioactivity levels in the system were normally very low. Site personnel indicated that a procedure was used for the evolution, but it lacked precautions related to excessive use of steam and increased heating of tank contents when the steam jet system was used at low tank levels.
The transfer was managed from the outside facilities control room without knowledge of the F-Canyon supervisor. As a result, F-Canyon operators were unaware that a transfer was in progress when they received a high airborne activity alarm for the stack. Facility personnel pulled the filter paper from the associated monitor, counted it, and detected contamination. Because there was no known activity in progress, they concluded that the contamination was caused by high radon levels and decided to recount the filter paper in a few hours, expecting the alpha decay associated with radon. However, when the filter paper was recounted five hours later, the expected decay had not occurred. Health Protection personnel then sampled all contributors to the stack monitor and identified the source of the increased activity.
Investigators continued to review the event to identify root causes and determine which tank the release originated from. However, the event underscores the importance of a number of conduct-of-operations issues including: (1) adequacy of procedure precautions and notes addressing potential problems when using the procedure under unusual circumstances (lack of precautions regarding increased use of steam and tank heating when using the steam jet system at low tank levels), (2) adequate shift management ownership of operations affecting plant safety (the F-Canyon supervisor was not aware of the transfer), and (3) appropriate response to alarms and abnormal indications (operators assumed the stack high activity alarm was caused by radon and not by an actual release of activity). ONS will provide additional information on the causes and lessons learned from this event as warranted.