Quantitative decision making for a groundwater monitoring and subsurface contamination early warning network

Abstract

With the increased development of oil and gas activities in northern Colorado, public concerns over the environmental impacts associated with well drilling and hydraulic fracturing have continued to rise. Issues such as leakages of “toxic” products from oil and gas operations to the subsurface environment (such as groundwater contamination) have led to community action and state regulations related to the establishment of groundwater quality monitoring sites in oil and gas activity areas, particularly those adjacent to urban development. Colorado Water Watch was a groundwater quality monitoring network comprised of seven monitoring wells in northern Colorado to monitor groundwater quality near oil and gas wells and give early warnings of contamination. Our study is aimed at developing a quantitative methodology to find ideal monitoring locations as well as evaluate them. We utilized hydraulic and geological data to select the most preferred sites to monitor groundwater quality, understand the temporal trends and identify unique anomaly signals in the oil and gas active area (Wattenberg field, northern Colorado). In addition to the site selection methodology, water quality data from Colorado Water Watch over 2 years is used to do evaluate the performance using entropy information and Principal Component Analysis. The analysis indicates that the earliest functional monitoring site (CHILL) is the most informative monitoring well, and the most recently installed monitoring sites (Gilcrest and LaSalle) are the least informative and least important stations due to their low data efficiency.