Using a simple geological setting to model the transport of radionuclides in the shallow subsurface

An underground nuclear explosion (UNE) produces noble gases that migrate and evolve through the earth's subsurface. Among these noble gases, radioactive xenon isotopes, such as Xe-133 and Xe-135, are routinely sampled in nuclear monitoring because they may help indicate whether the release of the radioxenons is caused by a UNE or civilian uses such as power plant or medical isotope production. Accurate interpretation of the gas signals can be quite challenging. First, some radioxenons have a short half-life and can decay below the detection limit before reaching the surface. Second, the transport of the gases depends on many factors, such as barometric pressure, precipitation, and geological properties. Many of these factors are field-dependent and difficult to capture through lab experiments. In addition, geological data for the source region of a UNE can be very limited because onsite measurements are often unavailable. The focus of this study is to see if the transport of radioxenon in a shallow field experiment can be modeled with limited site characterization. This work utilizes the data from a field experiment that involved the introduction of a radioactive xenon tracer into the subsurface environment. The experiment consisted of two wells separated 17m apart with three gas sampling zones each. Xe-133 was injected in the deep zone of one well and air was routinely collected from all sampling ports to measure the Xe activity concentration vs time. In this study, we set up numerical models using PNNL's STOMP (Subsurface Transport Over Multiple Phases) code with time-varying surface pressure and temperature based on the field inputs. Our goal is to model the inter-well connectivity and understand how barometric pumping and dynamic gas sampling affect Xe movement in the shallow subsurface. Our results showed that the simulation tends to provide an overestimation of the Xe concentration for the injection well and an underestimation for the survey well. Simple vertical heterogeneities are introduced to make up for limited site characterization and results in a reasonable match with the measurements. Finally, this work helps improve future field experiments by identifying the important site characterization parameters needed to build numerical models and further enhances understanding of noble gas movement from UNEs.