Decisions on Hydraulic Fracturing Must Be Science-Based.

Abstract

In 2013, in a corner of Palo Pinto County in north-central Texas, a water well exploded and sent several members of the same family to the hospital. Was the methane leak related to hydraulic fracturing of nearby oil wells? Resolution of the case is still pending as consultants and lawyers are busy sharpening their arguments. Legal jurisprudence is probably not the best suited environment to pursue scientific knowledge. The lawsuits regarding dissolved methane in drinking-water aquifers over the past few years are symptomatic of a topic that is not well understood. Hydraulic fracturing regulations in some states require operators to sample water wells in the vicinity of proposed oil or gas wells, the so-called baseline sampling. In other states, where baseline monitoring is not mandatory, many operators nevertheless sample nearby water wells. Because of this, an unprecedented amount of data about dissolved methane in the shallow subsurface has been generated. Methane is not toxic but can clearly be dangerous if allowed to degas from water and accumulate in confined spaces with no venting system. There are now many documented cases of dissolved methane in aquifers. The focus here is not methane originating from gas well blowouts or other clearly defined contamination events; such accidents typically impact a small number of water wells. Rather, our focus is on instances where multiple water wells show the presence of dissolved methane over a large area. Hydraulic fracturing is commonly used to develop so-called resource plays, that is, there is no dry well –only subeconomic wells. In resource plays, many, often horizontal, wells are drilled, each draining a small rock volume whose permeability is enhanced by forming a new or by reactivating an old fracture network. This process creates a diffuse pattern of oil or gas wells over large areas unlike development of conventional reservoirs. Conventional reservoirs can also involve many wells but are spatially constrained above reservoir traps and reservoirs. So, regional presence of dissolved methane, which would tend to suggest a natural origin, does not a priori exonerate the oil and gas industry in aquifers overlying a resource play. Given the tens of thousands of wells drilled and stimulated using hydraulic fracturing every year, some are bound to be defective even with an improved industry track record. However, many occurrences of diffuse dissolved methane have been proven to be natural; few, if any, have been convincingly shown to be directly caused by drilling, oil development, and production activity (excluding acute contamination cases); and several others are still being debated in the scientific community. There is also a large group of cases of dissolved methane in drinking-water aquifers that seem natural, but could have been caused indirectly by the industry for a variety of reasons, such as drilling methods or declining groundwater levels related to water withdrawal to perform hydraulic fracturing. When a causal relationship between dissolved methane and oil and gas activities seems apparent, investigators typically assume mobilization and transport of methane, but are short on describing the processes involved. Focusing on circumstantial evidence and statistics may lead from correlation to causation presumptions. But, in actuality, little is known about dissolved methane behavior in the subsurface and, after years of accumulating data, it is time to understand the actual mechanisms of dissolved methane mobilization and lateral migration. The groundwater community needs to go beyond the well-known fingerprinting techniques of carbon isotopes because, in many cases, naturally occurring shallow gas is also thermogenic or altered thermogenic. Some geochemical indicators, such as nitrogen and noble gas isotopes, provide information on the water/gas ratio and useful clues on transport mechanisms. More established tracers, such a strontium isotopes and bromide/chloride ratios, can help determine whether brine invasion played a role in methane transport. Establishing commonsense monitoring programs, requested by many, presupposes that we define program goals and know where and what to monitor. Federal- and state-led regional programs designed to capture diffuse contamination or local programs designed to investigate hot spots would all benefit from an improved understanding of the actual processes involved in methane migration.