Monitoring Hydrocarbon Pollution Plumes in Groundwater
Sensor Data Fusion using Machine Learning
C.L. Wu (TU Delft - Civil Engineering & Geosciences)
L.C. Rietveld – Promotor (TU Delft - Civil Engineering & Geosciences)
B.M. van Breukelen – Promotor (TU Delft - Civil Engineering & Geosciences)
R.M. Wagterveld – Promotor (Wetsus, European Centre of Excellence for Sustainable Water Technology)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Oil remains the largest contributor to global energy consumption, powering vehicles, heating homes, and fueling the economy. This dependence on oil has led to widespread groundwater contamination, primarily through the release of light non-aqueous phase liquid (LNAPL). These contaminants typically enter the subsurface from leaking underground storage tanks, pipelines, and refineries. Once in the subsurface, LNAPLs migrate through soil and form an immiscible layer that floats on the water table. The formed layer acts as a persistent source of dissolved petroleum hydrocarbons (PHCs) and creates contaminant plumes that last for decades.
PHC compounds are highly toxic and carcinogenic, posing risks to human health and the environment upon exposure. Although groundwater serves as a major drinking water source, drinking water wells are monitored only quarterly or yearly for PHC contamination. Conventional monitoring relies on time-consuming manual sampling and costly laboratory analysis. Sensors exist for directly detecting and measuring PHC levels in the field, but most of these sensors remain experimental and/or expensive.
Microbial communities in aquifers can degrade and thereby attenuate PHCs, making monitored natural attenuation a preferred long-term management strategy for PHC contaminated sites. This is especially the case when active remediation (e.g., pump-and-treat or excavation) is technically challenging or costly. However, there is still a gap for cost-effective tools for real-time, continuous groundwater monitoring for PHC contamination...