Quantifying timeframes for transitioning the management of a fresh release of LNAPL petroleum from active recovery to natural source zone depletion: Evidence from intensive field investigation of a new diesel release.

Often rates of biodegradation driving natural source zone depletion (NSZD) are estimated for weathered and aged petroleum releases, and often after active recovery is less feasible. But an open question is "When do natural biodegradation rates of petroleum hydrocarbon mass removal exceed what's removable by active engineered approaches for a fresh release of petroleum fuel?". Here we report on intensive site investigations and measurements over 15 months, that quantified active LNAPL mass recovery and NSZD rates implemented shortly after a new release of light non-aqueous phase liquid (LNAPL) diesel into a sandy vadose zone and groundwater. Masses of LNAPL actively recovered via direct pumping, multiphase extraction and primarily skimming were compared to NSZD rates estimated from LiCOR surface chamber measurements of fluxes of carbon dioxide (27 locations) and from soil gas depth profile sampling from nine short-screened recovery wells (RW) and from six multilevel depth samplers (VZ). To maximise recovery of LNAPL up to 13 recovery wells were operating at any one time. On average, LiCOR measured NSZD rates were higher than those estimated by oxygen measured in recovery wells, but those measured by sampling for oxygen in soil gas from VZ samplers were much higher again. Although active recovery was highest initially compared to NSZD rates, at individual paired well locations, the time-averaged LiCOR NSZD rates (5-25 L/m2/y) were higher than averaged skimming rates by up to c. 8 times. Spatial integration across the impacted area showed LNAPL recovery rates (13 locations, up to 82 readings) were substantially higher (5-35 L/m2/y) than NSZD (25 locations and up to 12 readings) rates (2.5-5 L/m2/y) for the first 7-8 months since the release. After that elapsed time, NSZD rates increased and remained at between 2 and 5 times higher (∼ 10 L/m2/y) than skimming based on conservative LiCOR measurements. This was found to be the case whether averaging recovery based on the radius of capture of recovery wells or based on the areal extent of NSZD across the release footprint. This provides an intensively investigated example case of timeframes whereby management of new petroleum releases may be acceptable based on NSZD, and perhaps within an initial 12-18 month period post release. It also provides a dataset that can be used to verify models for assessing the cross over point relevant to other fuel types and site conditions.