Seasonal changes in petroleum NSZD biodegradation rates associated with water table fluctuations within the zone of...

Natural Source Zone Depletion (NSZD) is widely used to manage petroleum-contaminated sites, but estimating NSZD rates is challenging because most methods rely on short period, averaged or point-in-time measurements. These can be misleading due to seasonal variability in biological activity, moisture and water table fluctuations influencing NSZD rates. Here we utilise continuous inground sensor measurements and light non-aqueous phase liquid (LNAPL) distributions from cores to understand seasonal processes affecting NSZD rates. Linked to multiphase modelling these data are used to test the hypothesis that seasonal NSZD variations result from LNAPL distributions and water table fluctuations. Using ~3.5 years of continuous oxygen and temperature data across four fuel types, NSZD rates were shown to vary by up to an order of magnitude seasonally. Rates ranged from 0.1 to 0.8 kg/m²/y for diesel to ~5 to 13 kg/m²/y for aviation gasoline, with broadly comparable trends between oxygen- and temperature-based methods. Rising water tables generally reduced NSZD rates at the diesel, gasoline, and crude oil sites by inundating the LNAPL depth distribution and disconnecting it from oxygen in the vadose zone. In contrast, at the aviation gasoline site, where LNAPL persisted above the maximum water table, NSZD rates remained high or increased. Modelling reproduced observed trends at the diesel site and showed that NSZD rates are influenced by the exposure of LNAPL to the vadose zone air phase (i.e., oxygen), with inundation suppressing rates, while exposure enhanced it. This is further supported by soil core TPH measurements that show loss of shorter chain TPH ranges at shallower depths above and within the zone of water table fluctuation relative to deeper in the formation. Continuous monitoring and modelling have provided insights into the dynamic NSZD rate estimates, indicating that care is needed to align field measurements with conceptualisation of LNAPL distributions and water table dynamics.