Natural Source Zone Depletion of crude oil, gasoline, diesel and aviation gasoline petroleum in the same soil/aquifer system - An intensive intercomparison field study and simulation.

Petroleum biodegrades and naturally depletes. Natural Source Zone Depletion (NSZD) quantifies this at petroleum affected sites in support of management decisions for cessation of active remediation efforts. Whilst a range of NSZD estimates and methods are available, side by side comparison of NSZD rates across petroleum types in the same soil/groundwater system are lacking, especially linked to the weathering status of petroleum. At a former refinery site near Perth Western Australia, locations contaminated by crude oil, gasoline, diesel and aviation gasoline, have been intensively instrumented to enable (i) measurement of vadose zone major gas (O, CO, CH), volatile organic compounds (VOCs) and temperature depth profiles, (ii) online near-continuous temperature, water level, O and VOC concentrations, (iii) depth profiles of groundwater parameters, and (iv) in-well gases, temperatures, water and LNAPL thickness. These measurements were compared to a background location with no history of contamination. Multiple coring events were also undertaken to determine LNAPL mass and its vertical distribution at each location. Additionally, LiCor and Eflux was conducted to measure CO fluxes at ground surface. NSZD rates were estimated from the measurement methods across the four petroleum types and the background site. Despite NSZD estimates that differed across some methods at sites (for example at the gasoline and diesel sites Eflux/LiCor estimates were consistently lower by a factor of 3-4 than those obtained using oxygen/temperature depth profile data) the minimum-maximum range of mean NSZD rates showed a distinct decreasing order across the fuel types: highest rates being aviation gasoline (69,000-91,000 L/ha/y), then gasoline, diesel and crude oil (2700-6200 L/ha/y). Reasons for differences are explored. Analysis of LNAPL in cores and from wells, historical data comparisons and simulations over 50 years, show that composition, age and weathering of the releases are critical to current and long-term NSZD mass losses and rate estimates.