Formatted Title
Long-Term Performance of a Large-Scale Pump-and-Treat System Including Novel Approach to Estimate Residual DNAPL Mass
Background/Objectives
The Orica Botany Groundwater Project is one of the world’s largest aquifer remediation programs. Site characterization work commenced in 1989 with remediation trials commencing in 1997. In late 2003, the EPA issued a statutory notice mandating the installation of a groundwater extraction and groundwater treatment plant (GTP). In January 2006, the GTP commenced operations treating on average 5 ML/day of groundwater contaminated with a wide range of chlorinated hydrocarbons (CHCs). The fifth triennial Botany Groundwater Strategy Review Workshop held in March 2020 included a panel of international groundwater and DNAPL remediation experts. An outcome of this workshop was for the review of the existing mass estimates. The purpose of these estimates is to provide a clearer picture of remediation progress and the mass of contamination left in the source zones versus plumes.
Approach/Activities
WSP Golder developed 3D geological and chemical plume models using the Leapfrog™ Works 3D subsurface modelling software. The software allows data to be visualised in ‘real’ space, creating a powerful communication and interpretation tool. All input data are modelled in the same workspace, including borehole lithology and chemical data with connection to the existing groundwater database. This connection allows for an integrated data management system, with the model updated as new data are added without duplication of effort. The geological model utilized lithological information from ~350 boreholes and over 100 CPT logs. The latter, due to their high resolution and objective nature, were vital in helping to define changes in lithology and the extent of numerous peat and clay lenses. The contaminant plume models for the most recent concentrations of the six key CHCs were developed which used over 500 individual analytical results. Combined models were developed for each of the six contaminants by calculating the intersections of each concentration range with each lithological volume. This allowed the modelled concentrations of each analyte to be split into lithological units, and contaminant mass (both dissolved and adsorbed) to be calculated separately for each lithology. The major benefit of this process was that individual porosity and fraction of organic carbon (foc) values could be applied to each unit and contaminant mass calculated more accurately within each lithological unit.
Results/Lessons Learned
Over approximately 16 years over 30,000 ML of groundwater has been extracted and treated in the GTP with approximately 1,575 tonnes of CHC mass destroyed. Modelled estimates of the dissolved phase and adsorbed mass were compared to earlier estimates derived using data from 2005 before the commencement of GTP operations. Comparison indicated that both the dissolved phase and adsorbed mass had decreased by approximately 90%. While the modelling helped inform the nature and extent of source zones, it was not able to be used to estimate the residual DNAPL mass. This DNAPL mass was derived using estimates of the Bulk Retention Capacity (Rs). However, the uncertainty in this parameter yielded a very wide-ranging estimate of DNAPL mass (~4,000 to 40,000 tonnes). An innovative approach was adopted to refine this estimate by examining trends in the GTP mass destruction rates. The wellfield essentially represents a large-scale integral pumping test, which can indicate source strength. Regression analysis of the late-time GTP mass destruction rates indicated a logarithmic trendline with an excellent regression coefficient (~98%). The trendline was extrapolated forward and appears to approach an asymptote at around 5,000 tonnes, which provided a bound around the DNAPL mass estimates.