Developing monitoring plans to detect spills related to natural gas production.

Surface water is at risk from Marcellus Shale operations because of chemical storage on drill pads during hydraulic fracturing operations, and the return of water high in total dissolved solids (up to 345 g/L) from shale gas production. This research evaluated how two commercial, off-the-shelf water quality sensors responded to simulated surface water pollution events associated with Marcellus Shale development. First, peak concentrations of contaminants from typical spill events in monitored watersheds were estimated using regression techniques. Laboratory measurements were then conducted to determine how standard in-stream instrumentation that monitor conductivity, pH, temperature, and dissolved oxygen responded to three potential spill materials: ethylene glycol (corrosion inhibitor), drilling mud, and produced water. Solutions ranging from 0 to 50 ppm of each spill material were assessed. Over this range, the specific conductivity increased on average by 19.9, 27.9, and 70 μS/cm for drilling mud, ethylene glycol, and produced water, respectively. On average, minor changes in pH (0.5-0.8) and dissolved oxygen (0.13-0.23 ppm) were observed. While continuous monitoring may be part of the strategy for detecting spills to surface water, these minor impacts to water quality highlight the difficulty in detecting spill events. When practical, sensors should be placed at the mouths of small watersheds where drilling activities or spill risks are present, as contaminant travel distance strongly affects concentrations in surface water systems.