Application of QMU to the design of a nuclear waste storage tank

Defining the reliability of complex physical systems is crucial for system certification and guaranteeing safety, but is often complicated by the presence of many uncertainties and system sub-components. One such example is the certification of the Hanford Waste Treatment Plant, where radioactive and toxic chemical waste that spontaneously generates flammable gases is stored in large tanks. We introduce the QMU methodology to quantify the system safety of waste storage tanks in the event of a loss of tank ventilation, which poses a safety risk to workers. We also show how QMU can be used as a design tool to find operating regimes that maintain a desired confidence level. A simplified physical model is used to analyze the build-up of hydrogen gas in storage tanks and conduct an uncertainty propagation study to quantify the operating margin with respect to the flammability conditions. A demonstration problem is shown in which the radiation loading in a waste tank is iteratively modified until a desired confidence ratio is achieved. The results show that the resulting designs are sensitive to the use of interval or full QMU to achieve a confidence ratio, and it is not always possible to realize a design at a particular confidence ratio.