THE USE OF STOCHASTIC SAFETY ANALYSIS METHODS TO ESTABLISH CONTROL STRATEGIES AND ESTIMATE FUTURE RISK

摘要

Stochastic safety analysis methods have been used at the Hanford Site in Washington State to establish control strategies and to estimate the future risk of waste transfer leak accidents. The Hanford Site tank farms currently store approximately 208 ML (55 Mgal) of radioactive/hazardous waste in 177 underground storage tanks. To move this waste a complex network of transfer piping and equipment is used. Each waste transfer carries a potential risk that a waste leak may occur. Waste transfer leaks into waste transfer structures, the soil (above ground and below ground), and into actively ventilated facilities due to a variety of causes were analyzed. A number of parameters (e.g., leak size, radiological composition of the leaked waste, wind speed and atmospheric dispersion) were treated as probability density functions (pdfs) in the safety analysis to determine the full range of potential consequences of waste transfer leaks. Each pdf was sampled in a Monte Carlo simulation model to determine the distribution of potential consequences from the various leak events. The resulting consequence distributions were evaluated with respect to risk guidelines to determine the appropriate control sets for daily operations to ensure that the risk from waste leak events could be minimized. The results were also used in the development of a probabilistic risk assessment to estimate accident risks to the public and co-located workers for both the tank farms and the future Hanford Site Waste Treatment Plant.