Probabilistic Risk Assessment: Perspectives on Groundwater Contamination from Hydraulic Fracturing Activities

Concerns regarding the impact of unconventional oil and gas development have motivated a number of scientific studies investigating potential groundwater quality risks posed by directional hydraulic fracturing activities. These studies tend to adopt one of two perspectives: site-specific risks such as torn liners and well casing failures, or regional groundwater sampling campaigns investigating whether groundwater quality changes are associated with local hydraulic fracturing activities. Here we explore the use of fault trees to determine the overall probability of groundwater contamination due to hydraulic fracturing activities referred to here as the probability of failure. The potential utility of fault tree analysis for the quantification and communication of contamination risks is approached and discussed from two perspectives: site-specific failure and cumulative community-based failure. Available scientific data is sub-divided by stage within the hydraulic fracturing process, and the unique basic events required for failure. For example, to quantify the risk of an on-site spill we must consider the likelihood, magnitude, and composition of the spill. However, when considering the impact on a community-based groundwater supply, the subsurface fate and transport of the contaminant, as well as the proximity of multiple hydraulic fracturing sites to the community become relevant as the probability of failure is now a combination of all sites hydraulically connected to the groundwater source. In addition, the definition of failure in these two scenarios are different: risks on site focus on the volume of spill while a community perspective targets contaminant MCLs or the cumulative health risk posed by contamination events. Ultimately, this analysis highlights the differences between site-specific risks and collective regional risks as regulation will likely continue to focus on controllable site-specific risks but needs to consider the cumulative impact of these risks.