A geochemical and multi-isotope modeling approach to determine sources and fate of methane in shallow groundwater above unconventional hydrocarbon reservoirs

Highlights

• Geochemical and isotopic modeling of methane migration in groundwater (Alberta)

• Modeling in situ generation of methane in aquifers

• Quantification of the extent of methane oxidation in groundwater

• Modeling can assess the initial isotopic composition of methane prior to oxidation.

• No deep thermogenic methane contamination was identified in Alberta groundwater.

Abstract

Due to increasing concerns over the potential impact of shale gas and coalbed methane (CBM) development on groundwater resources, it has become necessary to develop reliable tools to detect any potential pollution associated with hydrocarbon exploitation from unconventional reservoirs. One of the key concepts for such monitoring approaches is the establishment of a geochemical baseline of the considered groundwater systems. However, the detection of methane is not enough to assess potential impact from CBM and shale gas exploitation since methane in low concentrations has been found to be naturally ubiquitous in many groundwater systems. The objective of this study was to determine the methane sources, the extent of potential methane oxidation, and gas-water-rock-interactions in shallow aquifers by integrating chemical and isotopic monitoring data of dissolved gases and aqueous species into a geochemical PHREEQC model. Using data from a regional groundwater observation network in Alberta (Canada), the model was designed to describe the evolution of the concentrations of methane, sulfate and dissolved inorganic carbon (DIC) as well as their isotopic compositions (δ³⁴S_SO4, δ¹³C_CH4 and δ¹³C_DIC) in groundwater subjected to different scenarios of migration, oxidation and in situ generation of methane. Model results show that methane migration and subsequent methane oxidation in anaerobic environments can strongly affect its concentration and isotopic fingerprint and potentially compromise the accurate identification of the methane source. For example elevated δ¹³C_CH4 values can be the result of oxidation of microbial methane and may be misinterpreted as methane of thermogenic origin. Hence, quantification of the extent of methane oxidation is essential for determining the origin of methane in groundwater. The application of this model to aquifers in Alberta shows that some cases of elevated δ¹³C_CH4 values were due to methane oxidation resulting in pseudo-thermogenic isotopic fingerprints of methane. The model indicated no contamination of shallow aquifers by deep thermogenic methane from conventional and unconventional hydrocarbon reservoirs under baseline conditions. The developed geochemical and multi-isotopic model describing the sources and fate of methane in groundwater is a promising tool for groundwater assessment purposes in areas with shale gas and coalbed methane development.

Introduction

The recent expansion of the natural gas and oil industry into unconventional hydrocarbon reservoirs in North America and other parts of the world (e.g. Argentina, Poland, Russia) is currently transforming the global energy outlook (US EIA, 2015, 2016). Shale gas exploitation has increased markedly in the past decade due to horizontal drilling and hydraulic fracturing which have allowed the exploitation of hydrocarbon resources trapped in very low permeability source rocks. The production of these resources, however, is often associated with environmental concerns regarding freshwater consumption and potential contamination of surface water and groundwater, the appropriate management and treatment of fracturing chemicals and produced fluids, as well as other issues including induced seismicity, noise, traffic, air quality and atmospheric emissions (e.g. Rivard et al., 2014).

One of the main concerns is the potential migration of natural gas (composed mostly of methane, CH₄) toward freshwater resources (e.g. Jasechko and Perrone, 2017; Lefebvre, 2017; Rice et al., 2018; Bachu, 2017; Cahill et al., 2017; DiGiulio and Jackson, 2016; Harkness et al., 2017; Nicot et al., 2017; Wolfe and Wilkin, 2017). According to the terminology of Milkov and Etiope (2018), methane in natural environments can have biotic or abiotic origins. Biotic CH₄ is derived from biologically produced organic matter while the numerous reaction pathways resulting in abiotic methane production from geological sources are gaining increasing attention (Sherwood Lollar et al., 1993; Glasby, 2006; Etiope and Sherwood Lollar, 2013; Etiope and Schoell, 2014). Biotic CH₄ is produced either by microbial or thermogenic degradation of organic matter. Microbial CH₄ (also often called biogenic) is formed in relatively shallow geological formations through the microorganism-mediated decomposition of organic matter via acetate fermentation or reduction of CO₂. Thermogenic gas is generated by degradation reactions of organic matter in deeper geological formations under high pressure and temperature (Whiticar, 1999; Schoell, 1980; Schoell, 1983).

Different classification schemes and plots (Bernard et al., 1976, 1977; Schoell, 1980; Schoell, 1983;Whiticar and Faber, 1986) recently revised by Milkov and Etiope (2018) have been used to distinguish microbial gases from thermogenic gases even though post-genetic reactions may alter the initial chemical or isotopic signature of methane and higher alkanes. Natural gas of primary microbial origin usually contains at most trace amounts (<0.05%) of higher alkanes (C₂₊) such as ethane resulting in a dryness parameter (CH₄/(∑C₂₊) higher than 1000 (Bernard et al., 1977) although dryness ratios as low as 100 for microbial gases have been reported more recently (Milkov and Etiope, 2018). Additionally, microbial gas has usually carbon isotope ratios of methane expressed as δ¹³C_CH4 (in ‰ vs. the standard V-PDB) ranging from −120‰ to a traditional upper limit of −60‰ (Schoell, 1983) that was recently revised to a higher value of −50‰ (Milkov and Etiope, 2018). Thermogenic natural gas typically contains >2% C₂₊ alkanes resulting in a dryness parameter of <500, and is usually characterized by δ¹³C_CH4 values > −55‰ (Schoell, 1983) although more negative δ¹³C values for immature thermogenic methane have been reported (Rowe and Muehlenbachs, 1999; Tilley and Muehlenbachs, 2011; Milkov and Etiope, 2018). Milkov and Etiope (2018) revised and enlarged the thermogenic gas field compared to the original diagrams of Bernard et al., 1976, Bernard et al., 1977 and Schoell (1983) with overall δ¹³C values for thermogenic methane from −20‰ to −73‰, and the range of −55 to −73‰ characterizing early mature thermogenic gases (also referred to as immature gas). Such δ¹³C values for early mature thermogenic gas could be misinterpreted as pure microbial gases but may also constitute a mixture between microbial and thermogenic gas (Milkov and Etiope, 2018).

These chemical and isotopic properties of microbial and thermogenic gas form the basis for identifying the sources of methane in shallow environments and in particular shallow groundwaters. Many groundwater monitoring programs mandated by various states in the USA and provinces in Canada rely on these fundamental characteristics to differentiate microbial methane that may have formed in shallow environments from thermogenic fugitive methane. The latter may potentially migrate either from the intermediate depth zone below the base of groundwater protection (BGP, Lemay, 2008), or from deep hydrocarbon production zones.

However, chemical and isotopic fractionation occurring during gas migration and methane oxidation may hinder the isotopic and compositional identification of gas sources (e.g. Etiope, 2015; Prinzhofer and Pernaton, 1997). Solubilities of methane, ethane, and higher alkanes vary as a function of pressure and temperature (Culberson and McKetta, 1951; IUPAC-NIST, 2012). During the migration of fluids from hydrocarbon reservoirs at depths between 2 and 4 km, toward the Earth' surface, alkanes will exsolve into a free gas phase at different rates causing changes in the dryness parameter (McAuliffe, 1963), a process often referred to as solubility fractionation (McIntosh et al., 2018; Milkov and Etiope, 2018). Some studies have reported ¹²C enrichment due to diffusive migration effects in coal leading to more negative δ¹³C_CH4 values in diffused CH₄ compared to its source (Prinzhofer and Huc, 1995; Prinzhofer and Pernaton, 1997). In addition, methane may undergo partial oxidation under aerobic (coupled with O₂ reduction) or anaerobic conditions (e.g. coupled with denitrification or bacterial sulfate reduction). During oxidation, the light isotopes ¹²C and ¹H are preferentially metabolized, leaving the remaining methane enriched in ¹³C and ²H resulting in elevated δ¹³C and δ²H values in the remaining methane (Whiticar and Faber, 1986; Barker and Fritz, 1981). In consequence, it is possible to find microbial but partially oxidized methane with δ¹³C values between −50 and −30‰, i.e. in the range of thermogenic methane. Such elevated δ¹³C values of microbial methane that has undergone partial oxidation is referred to as pseudo-thermogenic. These processes may hinder the accurate differentiation of thermogenic and microbial methane leading potentially to false claims of thermogenic gas occurrence in shallow environments, if the assessment relies on the interpretation of gas composition and gas isotope ratios alone (Rice et al., 2018).

It is therefore essential to develop reliable approaches to differentiate truly thermogenic gases from pseudo-thermogenic methane, namely microbial methane that has undergone oxidation, to accurately detect migration of methane from deep geological strata into shallow aquifers. The incorporation of other aqueous species and their isotopic fingerprints such as dissolved inorganic carbon (DIC), sulfate, and nitrate (if present) in such an approach can reveal processes such as denitrification and bacterial sulfate reduction coupled with anaerobic or aerobic oxidation of methane impacting the concentrations and carbon isotope ratios of DIC and methane. Determining concentration and isotope ratios for the relevant dissolved and gaseous compounds and subsequent joint interpretation of the results is therefore a highly promising approach to assess whether methane in groundwater has been affected by geochemical processes modifying its isotopic composition. The often complex and challenging interpretation of such data sets can be facilitated by the use of geochemical modeling programs. One of the most widely used thermo-kinetic codes is PHREEQC, developed by the US Geological Survey (USGS), able to simulate geochemical reactions between water, gas, and mineral phases in aqueous systems (Parkhurst and Appelo, 1999). Hence, PHREEQC could be a powerful tool for the quantitative assessment of the sequence and the extent of biogeochemical reactions potentially affecting the chemical and isotopic fingerprints of methane, provided that the associated isotope fractionations factors are taken into account. In this case, inverse geochemical modeling of aerobic or anaerobic methane oxidation processes would allow backtracking to the original sources and concentrations of methane in shallow groundwater.

The objective of this study was to develop a method for accurately determining the origin and fate of methane in groundwater by integrating chemical and isotopic monitoring data of dissolved gases and aqueous species into a stringent geochemical model. A multi-isotope PHREEQC module was written for this study and tested on a groundwater data set from aquifers in Alberta, Canada. The module describes generic concepts controlling methanogenesis and methane oxidation that can be transferred to other groundwater study sites, associated with coalbed methane and shale gas development.