Time of Release Estimate for Contained Gas in the Subsurface – A Chemical Data Strategy

Abstract

We propose a model for estimating the time of release (TOR) for a hydrocarbon gas mixture (C1–C4: methane, ethane, propane, and butane). The model uses concentration data and performs a chemical kinetics analysis of the ratio of concentrations: methane to ethane (M/E) or iso-butane to n-butane (i/n). Six criteria are specified to choose the data used to develop the TOR estimate without introducing bias. A data set from the Napoleonville Salt Dome (NSD) gas release was available in which collection time and sample concentrations were known. The NSD release produced the Bayou Corne Sinkhole. That sinkhole appeared on 8/3/2012, establishing the TOR for NSD. Accordingly, the TOR model was calibrated, demonstrated, and tested with the NSD data set. The most immediately affected matrix was soil gas, and using the M/E ratio on the collected soil gas concentration data produced an estimated TOR of 22 April 2012 − 28 November 2012 (10 August 2012 ± 110 days). Using the i/n ratio produced an estimated TOR of 3 March 2012 − 20 November 2012 (12 July 2012 ± 131 days). The NSD sinkhole appeared on 3 August 2012, which is within the standard deviation of both estimates and showed no significant bias.