Noble Gas Analyses to Distinguish Between Surface and Subsurface Brine Releases at a Legacy Oil Site

Attributing the sources of legacy contamination, including brines, is important to determine remediation options and to allocate responsibility. To make sound remediation decisions, it is necessary to distinguish subsurface sources, such as leaking oil and gas (“O&G”) wells or natural upward fluid migrations, from surface releases. While chemical signatures of surface and subsurface releases may be similar, they are expected to imprint specific dissolved noble gas signatures, caused by the accumulation of terrigenic noble gases in subsurface leaks or re‐equilibration of noble gases following surface releases. We demonstrate that only a historic surface release influenced the dissolved noble gas signature of groundwater in monitoring wells contaminated with brine near an abandoned O&G well, rather than subsurface leakage from the well. Elevated brine concentrations were associated with lower terrigenic helium concentrations, indicating re‐equilibration with atmospheric helium at the surface during the release. Geophysical surveying indicating elevated salinity in surficial soils upgradient of the wells further supported the interpretation of the noble gas data. Eliminating the possibility that subsurface leakage was the source of the plume was critical to selecting the proper remedial action at the site, which otherwise may have included an unnecessary and costly well re‐abandonment. This study demonstrates the use of noble gas analysis to compare potential sources of brine contamination in groundwater and to exclude subsurface leakage as a potential source in an oilfield.


Introduction
The presence of chemical contaminants in groundwater overlying oil and gas (O&G) fields can have a such as agriculture, manufacturing, or other industrial sites (e.g., Stanton et al. 2023).
One of the most common groundwater contaminants found within O&G fields is brine, predominantly sodium chloride.Brine contamination may originate from surface or subsurface releases of oilfield-produced water, which is the water produced along with oil that is typically naturally enriched with salts.Brine concentrations are usually expressed as chloride or Total Dissolved Solids (TDS).The U.S. Environmental Protection Agency (EPA) lists both chloride and TDS on the National Secondary Drinking Water Regulations (NSDWRs) that set non-mandatory water quality standards at 250 mg/L for chloride and 500 mg/L for TDS (U.S. EPA 2024).Since chloride behaves as a conservative tracer in groundwater and has high solubility, chloride is commonly used as an indicator of produced water impacts in groundwater near O&G sites (e.g., Warner et al. 2012;Molofsky et al. 2013;Darrah et al. 2014;Darrah et al. 2015;Molofsky et al. 2016;Kreuzer et al. 2018).Since chloride has other common sources besides oilfield-produced water (e.g., road salt, water softeners, sewage, fertilizers, sea water intrusion), chloride is often analyzed in conjunction with other ions, including bromide.Bromide is an ideal tracer of formation brines since bromide stays in solution during sea water evaporation and is not modified by diagenetic processes.Therefore the ratio of Cl/Br is a commonly used fingerprint for distinguishing brine contamination from oilfield-produced water from other sources of brine contamination (Freeman 2007).
Once it has been determined that the chloride in groundwater is from oilfield-produced water (Chesley-Preston et al. 2014;Lauer et al. 2016) using the Cl/Br ratio or other geochemical measurements (Whittemore 2007), it still leaves several possible mechanisms for the produced water to have influenced the groundwater, such as surface releases, wastewater injection, or subsurface well leakage.These different transport pathways of brine releases with potentially identical geochemical compositions may or may not allow for the equilibration of noble gases with the atmosphere.Consequently, they are expected to result in unique dissolved noble gas compositions that allow for the differentiation of release pathways.Surface releases and potential wastewater injection is more likely to allow for noble gas equilibration through exposure to atmospheric air, while subsurface leakage below the groundwater table likely perseveres the noble gas composition of the oil reservoir because the fluids are not exposed to the atmosphere.
Noble gases have distinct source signatures (Hunt et al. 2012;Barry et al. 2018), are chemically inert, and fractionate via physical processes in predictable ways (Stute et al. 1995;Aeschbach-Hertig et al. 1999).Dissolved noble gas analysis is routinely used in the field of hydrogeology to determine aquifer characteristics such as groundwater age (Schlosser et al. 1988;Solomon et al. 1992;Visser et al. 2014Visser et al. , 2016;;Sprenger et al. 2019) and recharge temperatures (Stute et al. 1995;Segal et al. 2014;Visser et al. 2018).Noble gases have also been explored to better understand oil reservoir properties, such as fluid migration (Barry et al. 2017;Byrne et al. 2020).Noble gas analysis has been used in limited cases to determine whether methane impacts were caused by stray gas from hydraulically fractured wells (Hunt et al. 2012;Darrah et al. 2014;Kreuzer et al. 2018) or from infiltration through unlined disposal ponds (Mcmahon et al. 2019;Karolytė et al. 2021).
Based on their unique properties, the analysis of dissolved noble gas concentrations is proposed to differentiate subsurface from surface sources of groundwater contaminants at O&G sites.Subsurface sources, such as leaking O&G wells and natural upward fluid migrations, are enriched in terrigenic noble gases, whereas the fluids of surface-derived sources, such as surface releases and injected wastewater, will have equilibrated with atmospheric noble gases.Dissolved noble gases measured in groundwater are composed of atmospheric noble gases that equilibrate at the groundwater table before groundwater recharge, as well as terrigenic noble gases sourced from the Earth's crust and/or mantle.Terrigenic helium can be particularly useful for this analysis since it is the most abundant terrigenic noble gas.Terrigenic helium can be produced by radioactive decay of naturally occurring uranium and thorium in the aquifer (Marine 1979;Torgersen 1980;Stute et al. 1992) or accumulate by upward diffusion of crustal or mantle helium from deeper formations (Kulongoski et al. 2005).Terrigenic helium dissolves and accumulates in groundwater over time.Fluid from deep O&G-producing formations contains higher concentrations of terrigenic helium than younger shallow groundwater because of the reservoir age (Darrah et al. 2015;Liu et al. 2023).For many U.S. reservoirs, helium concentrations in O&G are known (Blondes et al. 2018).Subsurface leaks below the groundwater table-either natural or via wellbore-are expected to maintain their high helium concentrations, providing a sharp contrast with the overlying aquifer (Darrah et al. 2014;Barry et al. 2018).On the other hand, a surface release reveals the formation of fluids exposed to the atmosphere which causes a loss of dissolved helium to fluid equilibration with atmospheric air.Produced water that is processed on the surface and then reinjected underground as wastewater may have a varied noble gas composition, but as these waste handling processes are not airtight, it is expected that helium concentrations would be lower than produced fluids.The concentration of dissolved terrigenic helium can therefore be an excellent indicator of the primary source of the contamination.
However, elevated terrigenic helium concentrations have been observed in shallow aquifers which may confound the attribution of terrigenic helium to subsurface O&G sources (Ma et al. 2005;Darrah et al. 2015;Kreuzer et al. 2018).Elevated terrigenic helium concentrations have been attributed to (1) diffusion from aquifer sediments derived from old protoliths at rates well above the U/Th production rate (Solomon et al. 1996), (2) upward diffusion and/or advection from mantle through fault zones (Kulongoski et al. 2013), and (3) upwelling or diffusion and/or advection along fault zones of formation brines containing high levels of helium and salinity (Ma et al. 2005;Ma et al. 2009).Even under these conditions, with natural background aquifer helium concentrations elevated above atmospheric equilibrium, surface releases are expected to result in lower helium concentrations than subsurface leaks.Ideally, the noble gas concentrations of both the background groundwater and oil reservoir can be measured to increase confidence in source attributions.
In addition to terrigenic helium, the concentrations of other atmospheric noble gases, including neon, argon, krypton, and xenon, can provide additional lines of evidence to understand the origin of subsurface leaks from surface releases.The solubilities of noble gases are strongly dependent on the salinity at equilibration (Weiss 1971).Small increases in salinity lead to a sharp decrease in noble gas solubility (Jenkins et al. 2019).On the other hand, binary mixing between atmosphereequilibrated groundwater and reservoir fluids results in a sharp increase in salinity and a modest decrease in noble gas concentrations.This contrasting behavior provides a second line of evidence to distinguish between mixing and re-equilibration at elevated salinity.
Table 1 summarizes the potential sources for oilfield brines or hydrocarbons to enter groundwater and whether their chloride concentrations and expected terrigenic helium concentrations would be higher or lower than other sources.While most sources yield the same liquid geochemical composition, the noble gas composition will be altered depending on whether the fluids were in contact and equilibrated with atmospheric air.
The objective of our study was to identify the primary source of two brine plumes within 1000 feet of each other at a legacy O&G production site in the American Midwest.We used a combination of dissolved noble gas analysis from multilevel monitoring, supplemented by geochemical and geophysical investigations.

Field Site Background
Groundwater at a historic oilfield in the Midwest was found to contain concentrations of chloride higher than the surrounding aquifer.A map of the site is included as Figure 1.Nested monitoring wells were installed, which delineated the extent of the brine plume.Interpretation of the monitoring well data suggested there were two separate groundwater plumes surrounding two adjacent abandoned oil wells (OW-A and OW-B).A schematic of the monitoring wells and screened intervals sampled in this study is also presented in Figure 1.
Groundwater salinity around OW-A was found to be increasing with depth, from 551 mg/L chloride at 56 feet (17 m) depth to 6470 mg/L at 69 feet (21 m), and the surface soil total chlorides (measured by EPA Method 325.2) was within the local background range.It was suspected that the source of the groundwater salinity was a subsurface leak from OW-A.OW-A had previously been abandoned so it would be logistically challenging to reenter the well and assess the abandonment integrity.
Groundwater salinity around OW-B was found to be decreasing with depth, from 1820 mg/L chloride at 53 feet (16 m) depth to 722 mg/L at 115 feet (35 m), and the surface soil total chlorides was found to be elevated relative to the local background.It was suspected that the source of this groundwater brine plume was from a historical surface release near the OW-B wellhead.
Chloride concentrations in groundwater around both OW-A and OW-B were observed to be stable over the past 10 years of quarterly monitoring and there did not appear to be any seasonal variations or fluctuations due to groundwater elevation changes.No volatile or semivolatile organic compounds have been detected.Methane concentrations are low (<1 mg/L) for all locations except background well MW-C1 at 17 mg/L.
Major and minor ions were previously analyzed to determine the source of the brine.Chloride/bromide ratios from the brine-impacted monitoring wells (ranging from 110 to 150) were consistent with published chloride/bromide ratios for the oil-producing formation (Wilson and Long 1993).For comparison, the median Cl/Br ratio from atmospheric deposition in the Midwestern USA was found to be 47 (Panno et al. 2006), 1730 for septic systems in the Midwestern USA (Panno et al. 2006), and 1000 to 10,000 for road salt (Davis et al. 1998).This analysis concluded that the brine likely originated from the oil-producing formation but could not identify the primary release source for the brine to enter groundwater.
A generalized description of the soils at the site consists primarily of reworked glacial sediments.The  depositional environments are beach and back-beach deposits, near-shore lacustrine deposits, as well as stream and over-bank deposits.Soils consist of poorly sorted lacustrine clay, silt, sand, and gravel deposits that are characterized by interbedded layers of dense clay, silt, and intervals of poorly sorted fine to coarse grain sand with fine gravel.The primary drinking water aquifer in the area is located between 33 and 82 feet (10 and 25 m) below ground surface (bgs) and typically ranges between 10 feet (3 m), to greater than 66 feet (20 m), in thickness.
The site is marshy with a gentle slope to the northeast.The hydraulic gradient is correspondingly shallow with a northeast flow direction.It is difficult to correlate the more permeable strata from available soil boring logs, but some paleochannels are possible mirroring topography.There is an abrupt lithologic change to the northeast of OW-B where the soil becomes mostly low permeability clays and silts which could explain the absence of chloride impact further downgradient.Perched water was not encountered.
The two oil wells were previously produced from the Dundee formation, which is a highly saline Devonian Limestone reservoir.The total dissolved solids (TDS) and chloride concentrations measured in this reservoir were 385 and 242 g/L, respectively (Blondes et al. 2018).Chloride concentrations are high and approaching solubility limits.Dissolved helium in formation fluids was measured at 1.76 × 10 −4 cm 3 -STP/g (176 μL-STP/kg) (Ma et al. 2009) which is approximately 3700 times higher than the atmospheric equilibrium He concentration of 4.79 × 10 −8 cm 3 -STP/g (0.0479 μL-STP/kg), suggesting that the formation helium is predominantly terrigenic in its origins.
Site access to install additional monitoring wells is limited, so alternative less invasive approaches to understand the source of the chloride were desired.

Research Methods
To understand the brine release source that caused the chloride plumes around the two oil wells, an additional assessment was conducted which included dissolved noble gas analyses and geophysical surveying.

Field Sampling
Groundwater samples for dissolved noble gas analyses were collected from nine groundwater monitoring wells (Figure 1).Four samples were collected from a deeper and a shallower zone in two nested wells (MW-A1-A, MW-A1-B, MW-A2-A, MW-A2-B) adjacent to OW-A.Three samples were collected from a shallow, middle, and deep zone in a nested well (MW-B-A, MW-B-B, MW-B-C) adjacent to OW-B.Two background samples were selected from wells (MW-C-1, MW-C-2).These background wells were the two monitoring wells closest to OW-A and OW-B which had chloride concentrations within the natural background range found in this region.Noble gas samples were collected in 0.95 cm diameter (3/8 ) copper tubes, pinch-clamped on either side to prevent atmospheric contamination.Samples were collected using a submersible pump (Grundfos MP-1) maintaining a positive pressure in the sampling line to avoid contact with the atmosphere during sampling.The transparent sampling line was vigorously tapped with a wrench and inspected for trapped bubbles, before the copper tubes were clamped on either end, sealing 10 g of water between the clamps.

Laboratory Analysis
Noble gas samples were analyzed on a dedicated multi-port automated noble gas mass spectrometry system (Rademacher et al. 2001;Visser et al. 2016).Reactive gases were removed with multiple reactive metal getters.Helium isotopes were measured after quantitative purification on a 10 K cryo-trap using a VG5400 sector field mass spectrometer (Surano et al. 1992).Neon, krypton, and xenon were cryogenically separated and measured by quadrupole mass spectrometer via isotope dilution with enriched 22 Ne, 86 Kr, and 136 Xe.The argon concentration was determined by pressure measurement using a high-sensitivity capacitance manometer.The procedure was calibrated using water samples equilibrated with the atmosphere at a known temperature and air standards spiked with known quantities of the noble gases.Typical analytical uncertainties are ±2% for the 3 He/ 4 He isotope ratio and the dissolved concentrations of He, Ne, and Ar, and ±3% for Kr and Xe (Visser et al. 2013).

Data Interpretation
Measuring the concentrations of all atmospheric noble gases enabled us to calculate "derived parameters" related to the conditions of the last equilibration with the atmosphere.The noble gas-derived parameters include the equilibration temperature (referred to as the noble gas recharge temperature and NGRT), and "excess air" component related to the dissolution of entrapped bubbles in the unsaturated zone (EA).The derived parameters were calculated using the unfractionated excess air (UA) model (Heaton and Vogel 1981;Visser et al. 2016).The two free parameters of the UA model (NGRT and EA) were estimated by minimizing the uncertainty-weighted squared differences between modeled and observed concentrations (Ballentine and Hall 1999;Visser et al. 2014).
The equilibration elevation (controlling atmospheric pressure) was set at 200 m, reflecting the ground surface elevation at the site (0.98 atm).Helium was excluded from the fitting because of the potential for a terrigenic source of helium.To fit the UA model to the atmospheric noble gases, the salinity at equilibration was initially assumed to be the salinity of the sample.Derived parameters are only reported if the chi-squared probability of the obtained model-data fit is greater than 1% (Visser et al. 2016).The propagated uncertainty of the estimated parameters (NGRT and EA) was calculated according to Ballentine and Hall (1999).
After calculating the derived parameters and atmospheric helium component, the terrigenic helium-4 concentration and terrigenic helium isotope ratio (R ter = 3 He ter / 4 He ter ) were calculated as follows.The concentration of terrigenic helium is calculated as the measured concentration ( 4 He sample ) minus the atmospheric component ( 4 He atm ), which consists of the equilibrium ( 4 He eq ) and excess air ( 4 He ea ) components derived from the UA excess air model: The isotopic ratio of the terrigenic helium component (R ter ) is then calculated as the ratio between the terrigenic helium-3 and the terrigenic helium-4 component, by subtracting the atmospheric component from the measured helium-3 concentration, using the atmospheric isotope ratio (Ra) and the solubility fractionation coefficient (α): R ter = 3 He ter 4 He ter = 3 He sample − 3 He atm 4 He ter = 3 He sample − R a × (α × 3 He eq + 3 He ea ) 4 He ter (2)

Geophysical Surveys
Geophysical surveying was completed using a Geonics EM31-MK2 Ground Conductivity Meter to assess electrical conductivity in the shallow subsurface (∼20 feet bgs, 7 m bgs).This is a common method used to infer salinity levels in soil, which correlate with areas of higher electrical conductivity (Greenhouse and Slaine 1983;Monier-Williams et al. 1990;Bahr 1999).Readings were taken approximately every 10 feet (3 m) in a grid pattern over the investigation area.Data was interpolated via inverse distance weighting using the Spatial Analyst Toolset within ArcGIS.Concentrations of terrigenic helium found at the site were relatively high (0.4 to 2.6 μL-STP/kg) compared with many groundwater aquifers throughout the world (Solomon et al. 1996) but within the range of values observed in the local aquifer (Ma et al. 2005).These concentrations are two orders of magnitude lower than observed in formation fluids (176 μL-STP/kg) (Blondes et al. 2018) suggesting dilution with local groundwater has since occurred.Concentrations of terrigenic helium in local aquifer groundwater are higher than expected from in-situ radioactive decay of uranium and thorium and are attributed to the upwelling of deeper formation brines into shallower groundwater (Ma et al. 2005).Evidence for this included (1) strong correlations between terrigenic helium and bromide and (2) terrigenic helium increasing with groundwater age and with proximity to discharge areas.

Results and Discussion
Plotting terrigenic helium vs. chloride from the two site background wells along with the local aquifer background samples and formation brine (Ma et al. 2005;Blondes et al. 2018) shows that terrigenic helium correlates with chloride (Figure 2).Concentrations in background wells MW-C1 and MW-C2 fall along a mixing line (R 2 = 0.9926; Ma et al. 2005) between the formation water signature and atmospheric signature plotted at the origin (atmospheric helium with no chloride).This indicates that the chloride found in the two background wells included in this study includes chloride sourced from the natural upwelling of deep saline basin formation brines.
For both sets of monitoring wells near OW-A and OW-B, terrigenic helium decreases with increased chloride.The decrease in OW-B (from 175 × 10 −8 to 38 × 10 −8 L-STP/kg between 722 and 1820 mg/L chloride) is steeper than the decrease in OW-A (from 160 × 10 −8 to 60 × 10 −8 L-STP/kg between 551 and 6470 mg/L chloride).This is evidence that the chloride found in monitoring wells near both OW-A and OW-B was sourced from a surface release of brine with no terrigenic helium which increased the chloride concentration but diluted the terrigenic helium concentration.The surface release mixing line connects MW-C1 to a hypothetical brine release that has lost all terrigenic helium to atmospheric re-equilibration.In contrast, a subsurface brine release would have resulted in increased terrigenic helium with increased chloride, following a mixing line toward the formation brine end member (Figure 2).
If bromide had been detected for all monitoring wells, analysis of terrigenic helium vs. bromide data would have likely produced even stronger correlations than chloride since bromide is a more diagnostic tracer of formation brines as it has less potential alternative surface sources than chloride (such as roadsalt, septic leach lines) which may introduce some variability in the data (Dumouchelle, 1987).Unfortunately the bromide in the site background wells was below the detection limit (5 mg/L) of the commercial laboratory utilized, so this analysis could not be conducted.A laboratory with a lower detection limit for bromide will be utilized for future studies.However, for this dataset the strong correlation between terrigenic helium and chloride in local aquifer groundwater makes it clear that the largest component of chloride in the background is from upwelling deep brines.
Xenon concentrations were found to decrease with increasing chloride (Figure 3) in wells impacted by brine releases.Xenon is sparingly soluble in high salinity brines such as produced water from the Dundee formation (385,000 mg/L TDS) (Blondes et al. 2018).As a consequence, re-equilibration of noble gases during a surface release is expected to result in very low xenon concentrations, less than 10% of the local aquifer groundwater, based on a calculation of xenon solubility in saline water (Jenkins et al. 2019).By generating regressions from xenon vs. chloride monitoring well data, the initial chloride concentration of the release may be estimated.A regression of xenon vs. chloride provides a more reliable estimates for the brine end-member than other noble gases because of (1) less variability in background xenon concentrations and (2) a stronger dependency of the xenon solubility on temperature and salinity resulting in lower xenon concentrations of the brine at the surface.OW-A and OW-B samples fit two independent regression lines which suggest that they have been caused by different surface releases.Both lines have an intercept of 16.1 × 10 −9 L-STP/kg, within measurement uncertainty of the xenon concentrations in the two background wells, providing confidence that the limited number of samples yields reliable regression line parameters.Following each regression to zero xenon yields preliminary estimates of the chloride concentrations of the surface releases of 67,000 mg/L for OW-A and 22,000 mg/L for OW-B.Non-linearities between xenon solubility, chloride concentration, and temperature may affect these estimates.These estimated concentrations were much lower than the produced water reported from the Dundee formation.
Little is known about how the produced water was handled or its potential for mixing with surface water before surface release.Differing salinity estimates from these two locations, as a result of different slopes of the regression lines, may suggest that two different brine releases at the land surface have caused the observed elevated chloride concentrations in groundwater.
An EM-31 geophysical survey shows two areas of high conductivity (up to 125 mS/m) corresponding to high salinities in the shallow groundwater (up to ∼20 feet [7 m] deep): one to the southwest of OW-A and one to the northeast of OW-B.MW-B is located in an area with relatively high salinity indicated by an electrical conductivity of 65 to 85 mS/m (Figure 4).The high conductivity overlying the elevated salinity at depth supports the idea  that the brine release from the surface release was near MW-B.OW-A, MW-A1, and MW-A2 are located in an area of low surficial electrical conductivity (15 to 25 mS/m).This supports the explanation that fresh native groundwater overlies the brine-impacted groundwaters sampled by MW-A1 and MW-A2 at 55 to 73 feet (17 to 22 m) depth.Because the surface soil geophysical measurements near OW-A display low salinity, it is plausible that the more concentrated brine release (67,000 mg/L) infiltrated at a topographic high southwest of OW-A and advected toward OW-A at depth.For both releases, the geophysical data are consistent with the chloride profiles and dissolved noble gas analysis and support a surface brine release as the source of the elevated chloride concentrations observed in the monitoring wells.Concentrations of terrigenic helium in the two background wells at the site increase with depth.Depth likely corresponds with the mixing of groundwater ages at a local scale.Groundwater closer to the surface has been diluted with infiltrating precipitation which is void of terrigenic helium (Figure 5).At a regional scale, there was no relationship between terrigenic helium and depth (Ma et al. 2005).Regionally, groundwater age is more a function of groundwater flowpath distance and proximity to discharge areas than depth.
The chloride concentrations in monitoring wells near OW-B decrease with depth, consistent with a surface release mixing with native groundwater.This is supported by the increasing terrigenic helium concentrations with depth.While terrigenic helium in monitoring wells near OW-B follows the same depth-trends as background samples, terrigenic helium concentrations are lower because the released brine was void of terrigenic helium.The xenon profile supports the idea that vertical infiltration of a surface-sourced brine with no xenon mixing with deeper native groundwater.
In contrast, for monitoring wells near OW-A, chloride concentrations are higher in the samples collected from ∼70 feet (21 m) depth than in the samples collected from 55 to 60 ft.(17 to 18 m).Consistent with the chloride profile, OW-A monitoring wells have lower terrigenic helium concentrations at deeper depths, reflecting a larger component of the surface-sourced brine release.This inverted profile is consistent with surface-sourced brines migrating horizontally into this area at depth.This hypothesis is supported by the geophysical data showing high salinity soil upgradient.The high initial concentration of the brine release (22,000 mg/L) has a higher potential for density-dependent flow resulting in a plume that migrates below the fresh groundwater near the water table.

Conclusions
Noble gas analysis is demonstrated as a powerful tool to differentiate surface from subsurface sources of oilfield groundwater contamination.Terrigenic helium vs. chloride concentration data provided strong evidence for historical surface releases as the source of brine impacts around both OW-A and OW-B.Naturally high background terrigenic helium concentrations had become diluted by brines, which had equilibrated with atmospheric air at the surface.Xenon vs. chloride concentration data suggest the brine impacts around OW-A and OW-B may be from two potentially separate surface release events with different initial salinities.Chloride concentration data were used in this case, but bromide concentration data with better detection limits are recommended for future studies.Electrical resistivity survey data also identified suspected locations of two surface releases.Dissolved noble gas analysis and electrical resistivity surveying produced complementary datasets and corroborated each other's conclusions.
The ability to identify the source of oilfield groundwater contamination is critical in selecting the proper remedial action.If the source of contamination is an ongoing subsurface leak from a well, the leak in the oil well needs to be addressed before groundwater remediation can be successful.If the possibility of a subsurface leak can be eliminated and the source of contamination is identified as a surface spill, well work (e.g., re-abandoning) can be avoided and remediation can be successful.Groundwater overlying or adjacent to oilfields commonly contains brines or hydrocarbons that have naturally migrated into the aquifer.Differentiating natural from anthropogenic impacts can help appropriately allocate responsibility or determine that no remedial action is required.
Noble gas analysis is recommended as a method to identify sources of oilfield groundwater contamination, especially in cases where other lines of evidence are inconclusive and where there is limited site access for additional monitoring wells.Currently there are few laboratories that can perform this analysis so availability may prevent wide-spread adoption.However, the cost of this analysis was much less than drilling additional monitoring wells at the site which alternatively may have led to the same conclusions through conventional analysis.

Figure 1 .
Figure 1.Site map including chlorine concentration contours as measured in groundwater monitoring wells located near OW-A and OW-B.All monitoring wells shown were sampled for chloride and chloride concentration contours were developed from data from monitoring wells with screened intervals ranging between 50 feet (15 m) and 119 feet (36 m) deep.Dissolved gases were measured at chloride-impacted locations MW-A1, MW-A2, MW-B (red), and background locations MW-C1 and MW-C2 (green).The insert depicts a schematic of the monitoring wells and screened intervals sampled in this study.

Figure 2 .
Figure 2. Terrigenic helium (presented in units of the atmospheric concentration of He) vs. Chloride concentration (mg/L).Dashed line shows general trend of terrigenic helium decreasing as groundwater becomes impacted with brine at monitoring wells in the vicinity of both OW-A (possible subsurface source) and OW-B (surface source), suggesting that both brine plumes are sourced from surface releases.Local Aquifer data is from (Ma et al. 2009).Formation Brine data is from (Blondes et al. 2018).Uncertainty ranges for terrigenic helium and chloride concentrations are smaller than the symbols in the figure.

Figure 3 .
Figure 3. Xenon vs. Chloride.Initial brine spill chloride concentrations of 67,000 and 22,000 mg/L for monitoring wells near OW-A (possible subsurface source) and OW-B (surface source), respectively, were estimated as the chloride concentration corresponding to a xenon concentration of zero based on the regression between chloride and xenon for each possible source.Uncertainty ranges for chloride concentrations are smaller than the symbols in the figure.

Figure 4 .
Figure 4. EM-31 geophysical survey shows two areas of high conductivity (up to 125 mS/m) corresponding to high salinities in the shallow groundwater (up to ∼20 feet [7 m] deep): one to the southwest of OW-A and one to the northeast of OW-B.Chloride concentration contours were developed from data from monitoring wells with screened intervals ranging between 50 feet (15 m) and 119 feet (36 m) deep.

Table 1 Relative Chloride, Bromide, and Terrigenic Helium Concentrations from Potential Groundwater Sources Groundwater With Liquid Hydrocarbons/Brines Surface Source Subsurface Source Tracer Shallow Meteoric Groundwater Oil Reservoir Oil Well Surface Release
Dissolved noble gas concentrations are summarized in Table2.Noble gas concentrations, Calculated terrigenic helium, additional geochemistry, and well details are summarized in Table3.

Table 3 Depth, Chloride, Bromide, and Terrigenic Helium.
Note: Cl and Br: dissolved chloride and bromide concentrations; ft bgs: feet below ground surface; +/− denotes the 1-sigma uncertainty propagated through the UA model calculations and Equation 1.