Evaluating Groundwater Quality

This study was designed to identify evidence of fluids from oil and gas development in groundwater and the sources and migration pathways responsible for the presence of those fluids, including those not associated with oil and gas development. The study approach consisted of the following components:

Purpose and Scope

This report describes the quality of groundwater and discusses possible sources and migration pathways for the chemical constituents identified in samples that are potentially associated with hydrocarbon-bearing formations in and near the Montebello Oil Field, including formations used for commercial oil and gas production and naturally occurring oil and gas shows in overlying sediments. Groundwater samples were collected during 2014–18 from 21 preexisting monitoring, supply, and irrigation wells with completion depths ranging from 28 to 424 m below land surface (bls). Water samples were analyzed for a wide range of chemical compounds and isotopes to identify those potentially associated with fluids from hydrocarbon-bearing formations, including inorganic elements; nutrients; volatile and semi-volatile organic compounds; dissolved organic carbon and its optical properties; hydrogen and oxygen stable isotopes of water; isotopes of strontium (⁸⁷Sr/⁸⁶Sr); radium-224, radium-226, and radium-228; groundwater age-dating tracers (tritium [³H], sulfur hexafluoride [SF₆], and carbon-14 in dissolved inorganic carbon [¹⁴C-DIC]); dissolved noble gases; light hydrocarbon gases (methane through hexane); carbon stable isotopes in dissolved inorganic carbon; and hydrogen and carbon stable isotopes of methane.

The primary constituents discussed in this report are those with results most helpful for identifying potential effects of oil and gas development on groundwater in and near the Montebello Oil Field: inorganic elements, volatile organic compounds (VOCs), groundwater age-dating tracers (³H and ¹⁴C-DIC), light hydrocarbon gases (referred to as gases herein), and isotopes of methane. Additionally, oil companies were invited to share relevant information for inclusion in the study and select oil and gas wells, injectate sites, and groundwater wells for sampling. U.S. Geological Survey scientists also consulted with the California Department of Conservation Division of Geologic Energy Management (CalGEM) and Los Angeles Regional Water Quality Control Board staff that were familiar with the oil field. The goal of these compilation efforts and consultations was to obtain the most complete record of information possible.

Contents of this report focus on groundwater chemistry results from the RMP sampling events completed between 2014 and 2018 and available historical data inside or within 4.8 kilometers (km) of the Montebello Oil Field administrative boundary (referred to as the Montebello Oil Field study area hereinafter) and potential factors that may explain or have the potential to influence groundwater quality, including measures of oil development density, intensity, and proximity; water production and injection volumes; vertical fluid pressure differences; occurrence of oil and gas shows in sediments above oil- and gas-producing zones (zones from which commercial quantities of oil and gas are extracted; appendix 1); oil and gas well construction and integrity; other anthropogenic sources of analyzed constituents; and directions of groundwater flow. A 4.8-km buffer of the administrative boundary was selected as the boundary of the study area to include the estimated potential area of influence from oil and gas development in the Montebello Oil Field.

Study results are presented in three sections. The “Factors That Could Affect Groundwater Quality” section includes an analysis of the potential explanatory factors that represent sources and migration pathways of chemical constituents associated with hydrocarbon-bearing formations. The general distribution of groundwater quality across the study area is presented in the “Groundwater Quality near the Montebello Oil Field” section. The “Potential Sources and Migration Pathways of Chemical Constituents” section includes a detailed discussion of the RMP samples that contained chemical constituents associated with hydrocarbon-bearing formations. These analyses are intended to help improve understanding of how or if oil and gas development and naturally occurring sources of petroleum hydrocarbons outside of production zones have influenced groundwater quality in the Montebello Oil Field study area. Limitations for fully characterizing risk factors are presented in the “Study Limitations and Additional Analysis” section. See appendix 1 and figure 1.1 for definitions and diagram of the oil field features and chemical constituent sources used in the report.

Hydrogeologic Setting

Important hydrogeologic features in the area include the Montebello Hills, an east to west ridge in the Montebello Oil Field, and the Whittier Narrows, a narrow, alluvial-filled gap cut between the Montebello and Puente Hills by the San Gabriel River that runs along the eastern side of the field (fig. 1). The Montebello Hills are the surficial expression of anticline structures that trapped the oil and gas reserves in the oil field. Water-bearing sediments underlying the Montebello Hills are shallow (as little as 4 m bls) and minimally productive (Entrix, Inc., 2009). Shallow groundwater in the hills is frequently in discontinuous perched zones that are sometimes affected by confining conditions (Entrix, Inc., 2009). Locally, some of the limited shallow groundwater flows south-southeast toward the thick (greater than 670 m) valley alluvium in the Whittier Narrows adjacent to the Montebello Hills (fig. 1; Entrix, Inc., 2009), where water is used for supply. The study area has more than 180 public-supply wells, and about half of those wells are downgradient from the Montebello Oil Field.

Regional groundwater flow is to the west and southwest (fig. 1; Reichard and others, 2003; Ponti and others, 2014; Water Replenishment District of Southern California, 2016). Whittier Narrows is a 2.4-km gap that provides a natural conduit for groundwater and surface water to move from the upgradient San Gabriel Valley Basin to the Central Basin (fig. 1; U.S. Environmental Protection Agency, 2011). Inflow from the San Gabriel Valley Basin provides most of the natural groundwater replenishment for the Central Basin (California Department of Water Resources, 2004). In addition, artificially enhanced recharge to the alluvial valley aquifer beneath the San Gabriel River for downgradient water supply in Los Angeles groundwater basins has been active since the 1950s (Water Replenishment District of Southern California, 2019a), and two large managed recharge areas are immediately downstream from the Montebello Oil Field (fig. 1). Artificial recharge water has primarily been composed of imported water from the Colorado River (not shown) and northern California via the State Water Project, recycled water that has gone through tertiary treatment, and stormwater (Water Replenishment District of Southern California, 2019a). As of 2019, local stormwater and recycled water became the sole sources of supplemental recharge (Water Replenishment District of Southern California, 2020). Groundwater affected by artificial aquifer recharge in the Montebello Forebay tends to have modern (post-1952) groundwater ages, and detections of anthropogenic VOCs from surface or atmospheric sources are common (Shelton and others, 2001; Reichard and others, 2003).

Oil Field Geology and History

The Montebello Oil Field was discovered in 1917 and quickly became one of the primary oil-producing fields in California (McLaughlin, 1918; Save The Montebello Hills Task Force, 2011). The field consists of eight producing zones in the early Pliocene Repetto and late Miocene Puente Formations (fig. 2; California Department of Conservation, 1992). Zones one through three, the main area production zones (fig. 2), are the shallowest (averaging about 670–1,400 m bls) and produce oil from the Repetto Formation; these zones generally have had the largest volume of oil production since 1977 (fig. 3A). Zones four through eight (about 1,500–2,300 m bls), in the west area, were discovered in the 1930s and have produced oil and gas from the Repetto and Puente Formations. The east area production zones (Upper Nutt, Lower Nutt, Farmer, Cruz, and Baldwin) have produced very little oil or gas (fig. 3A). The oil- and gas-producing zones are overlain by late Pliocene siltstones, early Pleistocene conglomerates and siltstones, and late Pleistocene alluvial fans (Quarles, 1941).

As of 2018, the Montebello Oil Field had 832 (208 active, 20 idle, and 604 abandoned) wells (California Department of Conservation, 2018). The field produced more than 33 million cubic meters (m³) of oil from 1917 to 2018, mostly before 1960 (California Department of Conservation, variously dated). Peak oil production was in 1919. Thermogenic gas, methane and heavier gases produced by the action of high heat and pressure on buried organic matter at depth (appendix 1), was produced from the deeper hydrocarbon-bearing formations of the Montebello Oil Field until formation pressures were reduced. More than 14 million m³ of thermogenic gas was extracted from 1929 to 2018, with almost 11 million m³ of the volume originating as gas that was injected for storage between 1956 and 1996 (California Department of Conservation, variously dated). Starting in 1977, the eighth zone (about 2,300–2,326 m bls) has been used to store thermogenic gas imported from the central United States (fig. 3B; California Department of Conservation, 1992; California Public Utilities Commission, 2001). Imported thermogenic methane gas has been distinguished from locally produced gas in the Montebello Oil Field by the presence of elevated helium concentrations in the imported gas (California Public Utilities Commission, 2001). About 105 million m³ of water was injected into producing zones of the Montebello Oil Field between 1956 and 2017, mostly to enhance oil recovery in the main area (fig. 2; California Department of Conservation, variously dated). Water injection and production volumes (figs. 3C, 3D) are discussed in more detail in the “Potential Subsurface Sources and Migration Pathways” section of this report. Hydraulic fracturing and acidizing practices are rarely used to enhance oil recovery in the Montebello Oil Field. As of July 2014, only 2 percent of the active wells in the Montebello Oil Field had been hydraulically fractured (California Council on Science and Technology, 2015), and hydraulic fracturing was not reported during 2014–21 (California Department of Conservation, 2021).

Other oil fields near the Montebello Oil Field are the Bandini, East Los Angeles, Whittier, and Lapworth Oil Fields (fig. 1). These fields had fewer wells installed and produced less oil and gas during 1977–2018 compared to the Montebello Oil Field (California Department of Conservation, 2020c). The main oil- and gas-producing zones of those fields also are in the Repetto and Puente Formations, and average depths to the oil-producing zones range from 275 to 2,600 m bls (California Department of Conservation, 1992). An additional 334 oil and gas wells (41 active, 20 idle, and 276 plugged or buried) are in other oil fields in the Montebello Oil Field study area. As of 2018, most of those active and idle wells were in the Whittier Oil Field southeast of the Montebello Oil Field (California Department of Conservation, 2018).

Sample Collection and Analysis

Groundwater samples were collected in 2014, 2017, and 2018 from 21 preexisting monitoring, public-supply, and irrigation wells (table 1). Groundwater samples from four monitoring wells were collected downgradient from the Montebello Oil Field in 2014 as part of the exploratory phase of the RMP; those sites were sampled to assess the effectiveness of chemical, isotopic, and groundwater-age tracers for identifying potential sources of salinity, methane, petroleum hydrocarbons, and other constituents in groundwater in selected oil-producing areas of California (McMahon and others, 2017). Groundwater samples were collected from eight monitoring wells, eight public-supply wells, and one inactive irrigation well in 2017 and 2018. For those samples, potential well locations in the Montebello Oil Field study area were selected based on specific criteria: (1) areal and vertical position along groundwater-flow paths, with priority given to locations with shallow and deep well pairs; (2) proximity of wells to current (2018) and historical oil and gas development such as active and abandoned oil and gas wells, produced water injection, and disposal ponds and sumps; and (3) availability of historical water-quality data.

Maximum well-perforation depths for the 21 sampled wells ranged from 28 to 424 m bls (table 1). Most wells were less than 200 m deep and had well-perforation intervals that corresponded with aquifers used for public supply. Wells were categorized into shallow (less than or equal to 122 m bls), intermediate (122–305 m bls), and deep groundwater zones (greater than 305 m bls). The depth of 122 m bls was chosen as a break point between shallow and intermediate groundwater that could be used for supply; the depth of 305 m bls was chosen as a break point between groundwater that could be used for supply and deeper groundwater not likely used for supply. Samples from public-supply wells were collected before treatment and do not represent water delivered for consumption.

Groundwater samples were processed onsite in a mobile laboratory using procedures designed to minimize changes to the water-sample chemistry or introduction of target analytes during sample collection (U.S. Geological Survey, variously dated). Water was pumped from monitoring wells with a portable stainless-steel submersible pump. Water collected from all sampled wells was delivered from the well to the mobile laboratory through Teflon tubing with stainless-steel connections. Before sample collection, water was flushed from the wells by purging at least three casing volumes. Precleaned bottles were then filled in an enclosed chamber to prevent sample contamination and preserved by chilling, filtration, or chemical treatment as needed to prevent sample degradation. Specific conductance, pH, water temperature, turbidity, and dissolved oxygen were monitored throughout the duration of pumping to assess the stability of the pumped water (U.S. Geological Survey, variously dated). Blank, replicate, and matrix-spike samples were collected to evaluate the reliability of sample processing and analytical methods (appendix 2).

Groundwater samples were analyzed for a wide range of dissolved chemical, gas, and isotopic constituents to identify potential indicators of fluids from hydrocarbon-bearing formations. Where these potential indicators were detected in groundwater, an attempt was made to evaluate available information to ascertain potential sources and migration pathways of those constituents. Chemical constituents analyzed included inorganic elements; nutrients; volatile and semi-volatile organic compounds; dissolved organic carbon and its optical properties; hydrogen and oxygen stable isotopes of water; isotopes of strontium (⁸⁷Sr/⁸⁶Sr); radium-224, radium-226, and radium-228; groundwater age-dating tracers (³H, SF₆, and ¹⁴C-DIC); dissolved noble gases; light hydrocarbon gases (methane through hexane); carbon stable isotopes in dissolved inorganic carbon; and hydrogen and carbon stable isotopes of methane. The primary constituents discussed in this report are those with results most helpful for identifying potential effects of oil and gas development on groundwater in the Montebello Oil Field study area (inorganic elements, VOCs, groundwater age-dating tracers [³H and ¹⁴C-DIC], light hydrocarbon gases [referred to as gases herein], and isotopes of methane). Results and analytical methods for these analyses are available in an associated data release (Rodriguez and others, 2022) and in Dillon and others (2016).

For purposes of this report, the term “laboratory detection level” (LDL) is generally defined as the smallest concentration reported by the laboratory as a numeric value. Concentrations smaller than the LDL are typically coded as “less than (<)” the LDL. However, VOCs often are reported as numeric values that are less than the LDL because they are analyzed by an information-rich technique (Rodriguez and others, 2022). A detailed description of LDLs is provided in U.S. Geological Survey (2015).

The groundwater-age tracers ³H and ¹⁴C-DIC were used to classify samples as modern (recharged during or after 1953), pre-modern (pre-1953), or a mixture of both. Tritium (³H) is a naturally occurring, radioactive isotope of hydrogen that decays to helium-3 (³He) by beta-decay with a half-life of 12.32 years (Lucas and Unterweger, 2000). This characteristic makes tritium a conservative tracer for defining modern-age waters with detectable ³H. The ³H atom is incorporated in the water molecule in the atmosphere and is introduced and isolated to the groundwater-flow system during recharge. With time, the ³H decays to ³He, which resides as a helium component of the dissolved gas composition of the groundwater. Atmospherically derived carbon-14 (¹⁴C) is incorporated into the groundwater’s dissolved inorganic carbon (DIC) component and can be used qualitatively or quantitively to evaluate time since isolation from the atmosphere (Plummer and Sprinkle, 2001; Jurgens and others, 2010; de Jong and others, 2020). Carbon-14 in DIC has a half-life of 5,730 years and is used to date groundwater on the scale of 1,000 to tens of thousands of years before present. Geochemical processes other than radioactive decay act to dilute the atmospheric contribution of ¹⁴C required for age dating. Carbon mass and isotope exchange between DIC, soil carbon dioxide (CO₂), and aquifer carbonate rock and oxidation of organic carbon are expected to be the primary processes affecting DIC from the sample.

For this study, pre-modern groundwater is defined as having a ³H concentration less than 0.2 tritium units (TU) and a ¹⁴C-DIC concentration less than 75 percent modern carbon (pmc; Plummer and Sprinkle, 2001; McMahon and others, 2017). Modern groundwater is defined as having a ³H concentration greater than 1 TU and ¹⁴C-DIC greater than 90 pmc (Jurgens and others, 2010; Kulongoski and others, 2010; McMahon and others, 2017). Any concentrations outside of those ranges are defined as a mixture of modern and pre-modern groundwater. These assumed thresholds have been used in other California groundwater studies and represent approximate values for the purposes of classifying groundwater ages and qualitatively assessing relations of groundwater age to concentrations of selected chemical constituents.

Historical Groundwater Chemistry Data

Historical groundwater chemistry data available from other sources were used to supplement data collected as part of the RMP. More than 20,000 groundwater samples have been collected from 849 sites in the Montebello Oil Field study area since 1940 (fig. 4). About 80 percent of those wells have well perforation or total well depth information that can be used to associate chemical properties with depth intervals. Well depths ranged from 2 to 458 m bls, with a median of 101 m bls. Historical data were obtained from five sources: the USGS National Water Information System (U.S. Geological Survey, 2021), the Water Replenishment District of Southern California (variously dated, 2019b), the California State Water Resources Control Board Division of Drinking Water (2018), the California State Water Resources Control Board (2017a, 2017b) GeoTracker, and the California Department of Water Resources (2018).

Historical oil-field water-chemistry data were available from Metzger and others (2020) for 10 oil and gas wells (fig. 4) and 2 injectate samples in the Montebello Oil Field study area. The data were compiled from underground injection control project files of the CalGEM. Depths to the top of sampled oil and gas well perforations ranged from 592 to 2,522 m bls, with a median of 931 m bls. One injectate sample represents produced water obtained from neighboring oil wells and injected into the first zone. The other injectate sample represents blended and treated water obtained from multiple oil and gas wells across the Montebello Oil Field and injected into the eighth zone (fig. 2).

The historical chemistry data were primarily used for mapping the spatial distribution of major ions, selected trace elements, and VOCs as a supplement to the samples collected by the RMP, which had more analytes and lower LDLs but were only available for a limited number of sites. When using historical major ion and trace element data for this study, the most recent sample with complete data was generally used because it was closest in time to the samples collected for this study and substantial changes to those chemical constituents through time were unexpected. For VOCs and gas data, all historical samples at each well were used to evaluate if there had been a detectable concentration in any sample in time. This approach was used to best determine the locations where groundwater has been affected by VOCs and gases regardless of subsequent treatment or attenuation. The LDLs for VOCs obtained from historical data sources were often substantially higher than for RMP samples, typically by about an order of magnitude. To help minimize the effect of variable LDLs from different data sources and laboratory methods, a single benzene, toluene, ethylbenzene, and xylenes (BTEX) compound (benzene) was evaluated using a common concentration level of 0.5 microgram per liter (µg/L; the most commonly used LDL from historical groundwater chemistry data). Benzene was chosen for this assessment because it was the most commonly detected BTEX compound in RMP samples.

Potential Explanatory Factors

Newly collected RMP and historical water-sample data were compared to explanatory factors that have the potential to affect water quality. Data for these factors were compiled from many sources and included measures of oil and gas well density and age; occurrence of oil and gas in sediments above commercial oil- and gas-producing zones; water production and injection volumes (figs. 3C, 3D) and vertical fluid pressure differences; oil and gas well construction and integrity issues; other anthropogenic sources of analyzed constituents; depositional environments of aquifer sediments; and directions of groundwater flow. Aquifer geochemical conditions and groundwater ages determined from historical and RMP samples also were factors for understanding sources and migration pathways of chemical constituents in groundwater.

Records for 160 oil and gas wells from the California Department of Conservation (2018, 2020c) within 500 m of RMP sampled wells were analyzed to assess density, ages, integrity issues (table 2; appendix 3, table 3.1), and height of cement within the well annulus (space between the casing and the formation). A 500-m radius might not represent the distance in which oil and gas wells could affect groundwater quality but provides a consistent approach for evaluating potential relations of oil and gas development to groundwater quality in sampled wells and is a radial distance for assessing relations to surrounding land use that has been used by other groundwater studies (Johnson and Belitz, 2009). The height of cement within the well annulus was estimated based on casing diameter, borehole diameter, casing bottom, and the volume of cement and cement additives used (appendix 4). These data were available for 90 oil and gas wells (fig. 5), representing construction dates between 1922 and 1984. Calculated annular cement heights were compared with operator-reported heights that were available for a subset of the wells to assess the accuracy of the calculated results.

Mud logs obtained during the well-drilling process (available from the California Department of Conservation [2020a]) were used to identify the occurrence of oil and gas in drilling mud and cuttings at intervals shallower than (or areas outside) economically productive oil and gas zones (fig. 5). Gas chromatograph instruments determined individual gas components (methane, ethane, propane, butane, and pentane; Crain, 2010) and hydrocarbon fluorescence measured oil as a percentage of mud-log cuttings (Swanson, 1981). Some mud logs also identify traces of hydrocarbons, such as tar, from visual inspection of cuttings. In addition to mud logs, some oil and gas well drilling records were reviewed to determine if petroleum hydrocarbons were present in formations shallower than oil- and gas-producing zones (California Department of Conservation, 2020c).

Spatial patterns of the fluid volume balance (the difference in volume of produced oil and water and injected water and steam) from January 1977 to February 2018 (California Department of Conservation, 2020c) was assessed for the Montebello Oil Field to estimate areas where positive fluid balances would cause positive pressure and present a risk factor for fluid migration (Shimabukuro and others, 2019). In this analysis, fluid production and injection volumes associated with individual wells were distributed to a small surrounding area near each well; a two-dimensional Gaussian function, to allow summing of volume contributions from individual wells, was used. The spatially summed difference between the injected and produced volumes was then used to create a spatial distribution of fluid volume balances expressed as net fluid heights.

Idle well-fluid levels (appendix 1) that were measured in 82 oil and gas wells across the study area from 1991 to 2017 were available from the California Department of Conservation (2020c). These data were used in three ways: (1) 13 measurements (from 2015 through 2017) were compared with areas of positive and negative fluid volume balances, (2) 14 measurements (in 2015) were compared with water-level elevations in overlying groundwater zones during the same year (Water Replenishment District of Southern California, 2016), and (3) 11 measurements (from 2003 through 2014) within 500 m of sample BNE-3 (appendix 3, table 3.3) were compared with the groundwater level in that sampled well.

Hydrogeologic factors were compiled to examine their potential effect on groundwater quality. Available data about sources of recharge, direction of groundwater flow (fig. 1), distance along flow paths inferred from groundwater-level elevation contours (Reichard and others, 2003; Entrix, Inc., 2009; Ponti and others, 2014; Water Replenishment District of Southern California, 2016), and groundwater residence times were considered when interpreting detections of petroleum hydrocarbons, salinity, and other chemical constituents. A three-dimensional chronostratigraphic model developed by Ponti and others (2014) was used to determine whether the dominant depositional environment of sediments was marine or nonmarine at well-perforation depths (table 1).

Potential sources of anthropogenic contaminants not associated with oil and gas development included two U.S. Environmental Protection Agency (EPA) Superfund areas (see the “Potential Surface Sources and Migration Pathways” section), locations of soil or groundwater State Water Board Cleanup Sites (California State Water Resources Control Board, 2019b), and Toxics Release Inventory sites with a BTEX release in 2018 (U.S. Environmental Protection Agency, 2020a; fig. 6).

Potential Subsurface Sources and Migration Pathways

Injection well pressures, historical depletion of reservoir pressures and any resulting subsidence, and the distribution of net volumes of injection and production in oil reservoirs in the Montebello Oil Field are factors to consider in assessing the potential for fluid migration out of oil- and gas-production zones. Swanson (2018) described conditions in water-flooded oil fields in the Los Angeles Basin, including Montebello, that were operating near injection capacity. Those oil fields have many wells operating near or at the maximum allowable surface pressure, which is approximately 95 percent of the formation fracture pressure. Because injection well efficiencies change through time as wells become clogged (Swanson, 2018), injection rates must be reduced to keep pressures below maximum allowable values until wells are periodically stimulated, using acid washing to remove clogs. Therefore, injection rates and well efficiencies require careful management. Sustained pressures above the formation fracture pressure can increase the risk of fluid migration away from injection sites. Large amounts of injected water also can flush oil-producing formations and change the chemical character of the fluids in the oil reservoir (McMahon and others, 2018; Tyne and others, 2021). However, injected water in the Montebello Oil Field is recycled produced water, so the chemical differences between formation water and injected water are primarily expected to be limited to changes in concentrations of dissolved and noble gases and in oxidation-reduction conditions because of potential atmospheric exposure.

Volumes of injection and reservoir pressures have changed over time in the Montebello Oil Field. Before 1998, produced water was handled through a combination of injection (California Department of Conservation, variously dated), evaporation in disposal ponds, and conveyance to the ocean through pipes (Piper and Garrett, 1953). As a result, injection volumes were smaller than production volumes (fig. 7), potentially causing decreases in reservoir pressure. Sustained decreases in reservoir pressure have resulted in land subsidence in some California oil fields, which can induce well-integrity failures (Myer and others, 1996; Dale and others, 2000). However, based on available Global Positioning System station measurements and calculated estimates, it is likely that any land subsidence in the Montebello Oil Field is no greater than 2 inches (California Public Utilities Commission, 2001; California Department of Water Resources, 2021). The extent of subsidence-related well-integrity issues has not been analyzed in the Montebello Oil Field. Since 1998, the volume of produced water has increased substantially and has mostly been reused for water flooding (fig. 7) to enhance oil production. Part of that produced water had likely been previously injected. Some produced water from the deeper gas storage zone is treated and discharged to the Los Angeles County Sanitation District industrial sewer system or sold to other field operators (California Public Utilities Commission, 2001).

Production and injection data were available for individual wells from January 1977 to February 2018. Those data, combined with the elevation of the uppermost screened interval for about half of the wells, indicated that produced water was primarily injected into the same horizontal and vertical zone as most of the extracted waters (California Department of Conservation, 2020c; figs. 3C, 3D). About 70 percent of wells with available well-construction data in the Montebello Oil Field study area have an uppermost screened interval shallower than 1,400 m bls; for these wells, the volume of water produced was 48 million m³, and the volume of water injected was 40 million m³. For wells with an uppermost screened interval deeper than 1,400 m bls, the volume of water produced was 5 million m³, and the volume of water injected was 0.2 million m³.

Spatial patterns of the fluid volume balance (the difference in volume of produced oil and water and injected water and steam) for January 1977 to February 2018 (California Department of Conservation, 2020c) were assessed for the Montebello Oil Field to estimate localized areas where fluid imbalances would cause positive pressure and present a risk factor for fluid migration (Shimabukuro and others, 2019). These measurements serve as a proxy for hydraulic heads, which are often difficult to obtain. For most of the field, there is a net negative fluid balance (net fluid height less than zero) where more fluid has been produced than injected (fig. 8). Smaller areas of net positive fluid balance (net fluid height greater than zero) are on the eastern and southern areas. When compared with idle well fluid levels, areas with net positive fluid balances are expected to correspond with higher pressures or shallower fluid levels in idle wells unless a net negative fluid balance existed before 1977. Limited available idle well fluid levels from 2015 through 2017 (California Department of Conservation, 2020c) indicated that idle well fluid levels in the area of positive fluid balance were, on average, shallower (mean elevation below sea level referenced to the North American Vertical Datum of 1988 [NAVD 88] =246 m, n=6) than the idle fluid levels in areas of negative fluid balances in the field (mean elevation below sea level referenced to NAVD 88 =364 m, n=7). This approach does not include effects of activities before 1977, spatially heterogeneous sediments, or effects of dissolved gas but can indicate areas where fluid migration risk could be more likely to exist. This assessment for the Montebello Oil Field lacked complete oil field pool codes and well-perforation data that would provide a more detailed assessment of water production and injection imbalances in individual pools and with depth.

Higher hydraulic heads in oil- and gas-producing formations, compared to overlying groundwater, indicate the potential for upward migration of water from those formations into groundwater. The idle well-fluid levels near the Montebello Oil Field generally were lower than water-level elevations in overlying groundwater zones, indicating that vertical hydraulic gradients are downward, at least in the modern era, and downward flow would happen if there was a hydraulic connection between the two zones (Water Replenishment District of Southern California, 2016; California Department of Conservation, 2020c). These limited available data indicate that water from oil- and gas-producing formations likely is not moving upward to groundwater during current (2018) conditions. However, literature from the Los Angeles Basin indicates that upward movement of fluids from thermogenic hydrocarbon source rocks to overlying formations could have happened through geologic time (Jung and others, 2015). These processes have produced oil seeps and shallow oil deposits in parts of southern California (Hodgson, 1987; Hornafius and others, 1999); however, the magnitude and occurrence of upward fluxes of fluids from deep formations to sediments overlying oil- and gas-producing formations (through geologic time) in the study area are unknown.

Groundwater used for supply typically is separated from oil- and gas-producing formations by geologic layers that limit fluid movement between the two zones. The median vertical distance between petroleum resource development (top of oil well perforations) and groundwater (bottom of groundwater well perforations) in the Montebello Oil Field is about 650 m (fig. 3A). However, wells and dry holes constructed or abandoned with improper sealing, deteriorated wells, and well-integrity issues may facilitate migration of fluids from oil- and gas-producing formations to shallower zones (Wisen and others, 2020). Movement of fluids between geologic intervals, especially in the upper oil-producing zones, has been considered a possibility in the Montebello Oil Field because of these potential issues (Stolz, 1939; California Public Utilities Commission, 2001). Of the 21 RMP sampled wells, 15 were within 500 m of abandoned oil and gas wells and 2 were near active or idle oil and gas wells (table 2).

A limited well-integrity assessment of oil and gas wells near RMP samples was completed using CalGEM oil and gas well records (California Department of Conservation, 2020c). Well records indicated potential pathways for fluids from oil- and gas-producing zones to groundwater were documented, but data were insufficient to identify any effects on groundwater quality. Well-integrity issues included casing damage or collapse at depth, open holes, and a blowout (appendix 3, table 3.1). Of the 160 oil and gas wells within 500 m of groundwater wells sampled by the RMP, 35 (22 percent) had a well-integrity issue, although not all intersect groundwater resource zones (table 2; appendix 3, table 3.1). Of these oil and gas wells, 20 had a reported integrity issue potentially intersecting groundwater resource zones, defined as being shallower than the base of freshwater approximately estimated by CalGEM (California Department of Conservation, 2020a). Two oil and gas wells had an observed surface expression of a fluid leak. A surface expression is the flow of oil, water, or steam to the land surface at or around the well. Surface expressions indicate an integrity issue in the well, but the potential intersection of these issues with groundwater is unknown. Of the oil and gas wells analyzed, 13 had well-integrity issues reported at depths within oil- and gas-producing zones, below the groundwater resource zone, that may have low risk of affecting groundwater if the wells have an annular cement seal (cement in the annular space between the casing and the formation) between the oil- and gas-producing and groundwater-resource zones.

Well-construction practices have improved through time with respect to protecting groundwater (California Public Utilities Commission, 2001; King and King, 2013), and regulations on the oil and gas industry that specifically targeted protecting groundwater resources were enacted in the 1970s (California Department of Conservation, 2019a). Before regulations were enacted, some oil and gas well operators in the Los Angeles Basin filled wells and dry holes with construction debris and covered them with soil at the land surface. In addition, well casings and cement seals can degrade over time (Dusseault and others, 2000, 2014; California Public Utilities Commission, 2001; King and King, 2013; Ingraffea and others, 2014). Thus, older wells may be more likely to have inadequate protection for migration of fluids from hydrocarbon-bearing formations even when constructed with current (2018) protective measures.

To supplement the limited well-integrity data, about half of the oil and gas well records (452 wells) in the Montebello Oil Field were reviewed to determine well ages (California Department of Conservation, 2020c). Risk of fluid migration may not coincide with a specific well age, but 1976 was used as an approximate year for classifying oil and gas wells with the possibility of higher risk associated with older wells. This approach is supported by the well-integrity assessment results; 33 of the 35 wells with an integrity issue were constructed before 1976 (appendix 3, table 3.1). Of the 452 wells examined for age, 88 percent (400 wells) were constructed before 1976, and of those, 41 percent (165 wells) were abandoned before 1976. Additionally, some wells constructed before 1976 are in areas of the oil field that have a net positive fluid balance (fig. 8), potentially creating risk for oil-field fluid migration if those wells have integrity issues or were improperly constructed or abandoned.

Another approach for assessing potential pathways for upward movement of water from oil-bearing formations or overlying sediments with poor water quality or oil and gas shows is to determine if oil and gas well-casing intervals have gaps in the annular cement at the time of construction. Gaps in the annular cement can allow any escaped fluids from a well-integrity failure to easily move along the annular space to other geological intervals (Dusseault and others, 2000; Watson and Bachu, 2009; Davies and others, 2014). The limited analysis for oil-field wells within 500 m of RMP samples indicates gaps in calculated annular cement between oil- and gas-producing zones and fresh groundwater (typically less than 600 m bls; appendix 3, table 3.1) for many wells, regardless of the well-construction date (fig. 9); however, deeper oil and gas wells completed after 1964 generally have longer cement intervals than older wells. The calculated annular cement heights were similar to the operator-reported values for most of the oil and gas wells with an operator-reported cement interval available for comparison. The oil and gas wells shown in figure 9 were not distributed evenly across the oil field because RMP sample MT-15 was the only sample collected in an area near many oil and gas wells (fig. 5). Thus, these results primarily represent well-construction practices in an area west of most production and injection activities. However, if those practices are consistent with other wells in the Montebello Oil Field, then oil and gas well boreholes represent a potential pathway for fluids to migrate away from hydrocarbon-bearing formations if a well-integrity issue exists.

A limitation of using gaps in annular cement of oil and gas well-casing intervals to assess potential upward movement of water is that the annular cement record represents a construction record and conditions that existed before the well was abandoned. A well with an uncemented annulus may later be filled during abandonment but would not be accounted for in this assessment. The oil- and gas-well borehole conditions before abandonment could have been present for several decades between well installation and abandonment.

Natural migration of fluids from deep hydrocarbon-bearing formations to overlying groundwater could result from preferential pathways, such as fault planes, fracture systems, or geologic unconformities that intersect porous and permeable formations or from diffuse upward flow through zones where low permeability layers that cap hydrocarbon-bearing formations are absent or thin. Natural upward fluxes from hydrocarbon-bearing formations are typically considered to be limited because overburden pressure minimizes the effects of fractures and faults (California Public Utilities Commission, 2001). However, information regarding upward fluxes of fluids due to natural migration over geologic time in the Montebello Oil Field is unknown. Shallower Pleistocene and Holocene deposits are more permeable than deeper formations (California Public Utilities Commission, 2001) and could allow migration of any oil and gas present in those zones due to geologic oil or gas shows in these sediments or upward migration from deeper formations. The Montebello fault (fig. 5) and other potentially active faults (locations unknown) are in the Montebello Oil Field but are not areally extensive (California Public Utilities Commission, 2001); fluxes of fluids from hydrocarbon-bearing formations along faults are unknown.

The natural occurrence of petroleum hydrocarbons and other chemical constituents outside of oil- and gas-producing zones is common in California (Hodgson, 1987) and has potential to affect groundwater resources. In some oil fields in California, mud log data show evidence of naturally occurring oil and gas shows in sediments overlying oil- and gas-producing zones (Rosecrans and others, 2021; Anders and others, 2022) that indicate upward migration of oil and gas into shallow sediments through geologic time. These shallow oil and gas shows could contribute to the presence of some chemical constituents in groundwater. Mud logs were available from nine oil and gas wells in the study area (fig. 5; appendix 3, table 3.2). All nine mud logs were outside of oil field administrative boundaries; two mud logs were within 500 m of RMP sampled wells (MT-1 and MT-14). Mud log data and evidence of oil and gas shows in sediments overlying the Montebello Oil Field are limited but indicated that traces of tar and oil shows were detected in sediments in the study area at depths greater than or equal to 457 m bls (California Department of Conservation, 2019b). In addition, California Department of Conservation (2007) records indicate the presence of hydrocarbons in the aquifer at the base of freshwater, historically interpreted to have a total dissolved solids (TDS) concentration of less than approximately 3,000 mg/L. Another potential source of elevated concentrations of inorganic chemical constituents is the presence of paleo-seawater in marine sediments that have not been flushed with meteoric water (Piper and Garrett, 1953).

Gas leaks have been documented at several locations in the western part of the Montebello Oil Field (fig. 5; California Public Utilities Commission, 2001), but the total extent and prevalence are unknown. The California Public Utilities Commission (2001) noted that operating pressure in the gas storage zone (as of 2001) was about 40-percent lower than initial reservoir pressure in 1937, indicating that fluids should not migrate out of the gas storage zone. Yet, gas leaks were detected in residential areas between the 1970s and 1990s (fig. 5; California Public Utilities Commission, 2001). During a routine leakage survey in the 1970s, stored thermogenic gas (identified from the presence of helium) was discovered in soils of a residential neighborhood. The California Public Utilities Commission proposed two abandoned oil and gas wells and shallow fault zones as possible migration pathways for the gas. The soil gas was successfully treated to undetectable levels using extraction wells in the 1980s. In 1983, a well survey reported that several abandoned wells were leaking stored thermogenic gas (California Public Utilities Commission, 2001). The Southern California Gas Company suspected the leaks were caused by a combination of pressure differentials between deep and shallower zones and compromised cement seals of abandoned oil and gas wells. The abandoned wells were resealed with cement and procedures were improved to prevent additional leakage of stored gas. Leaking gas was discovered in two more abandoned wells; one leak was detected from 1984 to 1987 and another leak was detected in 1999 (California Public Utilities Commission, 2001). Gas probes detected unprocessed (not stored) thermogenic gas, with no helium or high concentrations of heavy hydrocarbons, along the border of the nearby Operating Industries, Inc. (OII) landfill in 1999 (fig. 6). Nearby features were described as wells located under the landfill (fig. 6) and the Montebello fault (fig. 5) to the east of the landfill (California Public Utilities Commission, 2001). In 2018, gas was detected in a shallow oil and gas observation well (less than 200 m bls) and triggered more frequent monitoring (California Department of Conservation, 2020b). The gas contained propane and heavier gases but low concentrations of helium, indicating unprocessed thermogenic gas. The cause of elevated gas levels in the observation well is currently being examined.

Potential Surface Sources and Migration Pathways

Surface spills or releases of hydrocarbons associated with oil and gas development and other activities at or near the land surface are common near the Montebello Oil Field. These spills can affect groundwater quality through infiltration. Historical unlined produced water disposal ponds and drilling fluid disposal sumps could have affected groundwater quality via infiltration of these fluids to unconfined aquifers. However, locations of drilling fluid disposal sumps in the Montebello Oil Field study area were only available for the Remedial Action Plan area (fig. 6). There were at least 99 sump locations distributed throughout that area, and they have not been used since 2008 (Entrix, Inc., 2009; Catalyst Environmental Solutions, 2016). It is unknown whether those sumps are lined. Additional oil field infrastructure features at the Montebello Oil Field include storage tanks, chemical storage areas, a produced water-treatment facility, and a gas separation and compression plant. Each of these features could be a potential source of petroleum hydrocarbons to soils and groundwater in the event of a fluid spill. Other potential anthropogenic sources of contaminants not associated with oil and gas development include EPA Superfund areas, locations of soil or groundwater State Water Board Cleanup Sites (California State Water Resources Control Board, 2019b), and Toxics Release Inventory sites with a BTEX release (fig. 6; U.S. Environmental Protection Agency, 2020a).

Two EPA Superfund areas are in the Montebello Oil Field: (1) the San Gabriel Valley and (2) the OII landfill (fig. 6). The San Gabriel Valley Superfund area is on the east side of the Montebello Oil Field. Previous research identified VOCs present at the site, including tetrachloroethylene (PCE), trichloroethylene (TCE), other solvents, petroleum hydrocarbons, and the trihalomethane chloroform (URS Corporation, 2015). A pump-and-treat system was completed on site in 2002 to remediate contaminated groundwater. As of 2011, treatment had minimized migration of contaminants off-site and to supply wells. Although concentrations of PCE were frequently greater than the drinking-water benchmark (U.S. Environmental Protection Agency, 2011), they were consistently declining. The only other contaminants of concern that were regularly detected in 2011 were the solvents TCE and 1,4-dioxane, and concentrations of those compounds generally were below their respective drinking-water benchmarks (U.S. Environmental Protection Agency, 2011). Regional groundwater flows from this superfund area toward the Montebello Oil Field and the Central Basin (fig. 1).

The OII landfill is on the northwest side of the Montebello Oil Field (U.S. Environmental Protection Agency, 2015). Operations as a municipal landfill for the City of Monterey Park (fig. 1), California, began in 1948, accepting residential and commercial refuse, industrial wastes, liquid wastes, and various hazardous wastes. Landfill leachate was detected off-site in 1982, all landfill operations ceased in 1984, and various measures were put in place to control and monitor leachate and landfill gas migration. In 1996, however, the EPA did a risk assessment and concluded that potential human health risks still existed. The constituents of concern included VOCs and semi-volatile organic compounds, pesticides, hydrocarbons, and inorganic constituents in ambient air, groundwater, and off-site soils. These contaminants included methane (microbial) and constituents in common with hydrocarbon-bearing formations, such as BTEX compounds and inorganic compounds (ammonia and barium). As of 2020, concentrations of contaminants of concern generally are decreasing, but 1,4-dioxane, TCE, and cis-1,2-dichloroethene may not reach cleanup goals by the time frame specified in the Record of Decision (U.S. Environmental Protection Agency, 2020b).

During several site assessments in the mid-1990s, soil contamination from petroleum products was discovered near inactive Montebello Oil Field infrastructure (fig. 6; Entrix, Inc., 2009). Soils were remediated where required to achieve acceptable concentration levels determined by the California Regional Water Quality Control Board (Los Angeles Region) and California Department of Fish and Game. A Phase II assessment was completed in 2006 for the areas not included in the scope of the 1990s remediation (Entrix, Inc., 2009). Shallow soil effects from petroleum products were discovered during these assessments, with an estimated soil volume of about 32,000 m³ requiring remediation near well sites, the produced water treatment facility, the gas plant, and other oil-field infrastructure (fig. 6). Deeper, unsaturated sediments containing petroleum products were assumed to be naturally occurring because they were not near infrastructure and, consequently, remediation was not planned. For soil samples collected as part of the Phase II assessment, sample concentrations were compared with California Human Health Screening Levels (Entrix, Inc., 2009, appendix D) and site or regional background concentrations (Chernoff and others, 2008; Entrix, Inc., 2009) to determine possible health risks. Based on this evaluation, total petroleum hydrocarbons, arsenic, and cadmium were considered chemical constituents of concern.

Published information includes several sites with contaminated groundwater in the Montebello Oil Field (Entrix, Inc., 2009). At one location, a plume of contaminated groundwater extending for about 61 m and containing up to 330 mg/L of total petroleum hydrocarbons was associated with soil contamination (fig. 6) discovered during abandonment of two oil and gas wells on the east side of the Montebello Oil Field. The direction of the plume was not described by Entrix, Inc. (2009), but nearby, local groundwater flow is to the southeast in the direction of the Rio Hondo (fig. 1). The affected groundwater was discovered in 1991, and the petroleum hydrocarbons had naturally attenuated by 1997. In another case, contamination near a bioremediation cell necessitated installation of two monitoring wells. Quarterly groundwater sampling of those wells began in 2001; only one recent sample (from May 2008) contained petroleum products (total petroleum hydrocarbon concentration of 0.56 mg/L) as of December 2009 (Entrix, Inc., 2009).

Potential anthropogenic contaminants have been identified at soil or groundwater State Water Board Cleanup Sites (fig. 6; California State Water Resources Control Board, 2019b). Of the 45 State Water Board Cleanup Sites that were within 500 m of an RMP sampled well, 38 were leaking underground storage tank cases and 35 specifically identified petroleum hydrocarbons as potential contaminants of concern. All but two of the RMP sampled wells (MT-1 and MT-8) were within 500 m of at least one of these potential contaminant sources.

Volatile Organic Compounds

Analyzed VOCs were categorized as petroleum hydrocarbon or manufactured compounds (table 3; appendix 3, tables 3.4, 3.5). Petroleum hydrocarbons were detected in 29 percent (6 of 21) of RMP groundwater samples (fig. 10) at concentrations substantially below drinking-water benchmarks (fig. 11; appendix 3, tables 3.4, 3.5; U.S. Environmental Protection Agency, 2018; California State Water Resources Control Board, 2019c; California State Water Resources Control Board Division of Drinking Water, 2020). The most prevalent petroleum hydrocarbons detected were benzene (five samples), 1,2,4-trimethylbenzene (two samples), ethylbenzene (two samples), m-Xylene plus p-Xylene (two samples), o-Xylene (two samples), and 4-isopropyltoluene (one sample; fig. 11; appendix 3, table 3.4). Carbon disulfide (not shown), a naturally occurring or manufactured VOC (Agency for Toxic Substances and Disease Registry, 1996), was detected in four samples, three of which contained hydrocarbons (appendix 3, table 3.5). None of the petroleum hydrocarbon concentrations exceeded 0.1 µg/L, and only 5 of the 14 detected concentrations were equal to or larger than the LDL(s) for the respective compound (fig. 11; appendix 3, table 3.4).

Detected petroleum hydrocarbons in deeper groundwater intervals may be at least partially explained by the presence of naturally occurring hydrocarbons in sediments shallower than commercial oil-producing formations. The limited available mud logs in the study area have evidence of oil shows in sediments at depths as shallow as 357 m bls (appendix 3, table 3.2) in a part of the Bandini Oil Field, near BNE-3, where benzene was detected at a depth of 405–424 m bls (appendix 3, table 3.5; see the “Potential Sources and Pathways of Chemical Constituents” section).

Benzene, toluene, ethylbenzene, and xylenes compounds commonly are measured in groundwater because of their adverse human-health effects (Lawrence, 2006). They also are the most frequently detected VOCs derived from petroleum products (including oil and gasoline) because they have relatively higher solubilities than other hydrocarbons (Lawrence, 2006). For these reasons, BTEX compounds are useful for summarizing the distribution of petroleum hydrocarbon detections from multiple sources. Laboratory detection levels (LDLs) of BTEX compounds for historical data sources were substantially higher than for RMP samples, typically by about an order of magnitude. To help minimize the effect of variable LDLs from different data sources and laboratory methods, a single BTEX compound (benzene, the most commonly detected BTEX compound in RMP samples) was evaluated using a common concentration level of 0.5 µg/L (the most commonly used LDL from historical groundwater chemistry data).

Benzene was detected at concentrations greater than 0.5 µg/L in 18 percent (103 of 585) of wells (historical and RMP data) analyzed for benzene in the Montebello Oil Field study area (California State Water Resources Control Board 2017a, 2017b; California State Water Resources Control Board Division of Drinking Water, 2018; Water Replenishment District of Southern California, variously dated, 2019b; U.S. Geological Survey, 2021; Rodriguez and others, 2022). Detections of benzene greater than the common concentration level (0.5 µg/L) were more frequent at shallower depths (fig. 12); benzene was detected in 12 percent of the 197 wells completed in shallow zones (less than or equal to 122 m bls). Benzene was not detected in any of the 111 wells completed in intermediate zones (between 122 and 305 m bls) nor in any of the 5 wells completed in deep zones (greater than 305 m bls). Of the 272 wells without available depths, 79 had a benzene concentration greater than 0.5 µg/L. Benzene was detected at multiple locations in shallow groundwater, in no apparent spatial pattern, indicating the presence of many sources of benzene at or near the land surface in the study area.

Activities other than oil and gas development have had a larger effect on groundwater quality in the study area with respect to VOCs, even assuming all the petroleum hydrocarbon VOC detections are associated with oil and gas development. Manufactured VOCs were detected in RMP samples more often (15 of 21 RMP samples; fig. 10; table 3) and at larger concentrations (0.0127 to 19.6 µg/L) than petroleum hydrocarbons (appendix 3, tables 3.4, 3.5). Tetrachloroethene (PCE), trichloroethene (TCE), and cis-1,2-dichloroethene were all detected at concentrations larger than water-quality benchmarks for drinking water in samples from one monitoring well (U.S. Environmental Protection Agency, 2018; California State Water Resources Control Board, 2019c); TCE alone was detected at a concentration above the water-quality benchmark in one supply well. This aforementioned supply well is already monitored for VOCs as part of regulatory monitoring, and raw water from this well is treated and blended before distribution as potable water. The concentrations of most manufactured VOCs in RMP samples were well below water-quality benchmarks and many were smaller than LDLs. Most of the detected manufactured VOCs were solvents (42 of the 63 manufactured VOC detections were solvents; appendix 3, table 3.4); other categories of manufactured VOCs detected in RMP samples were trihalomethanes, organic synthesis compounds, gasoline oxygenates, fumigants, and refrigerants. The compounds detected most often were PCE, TCE, and trichloromethane (chloroform). The presence of petroleum hydrocarbons in samples that also contain gasoline oxygenates, such as methyl tert-butyl ether (a common additive to gasoline), could indicate that the source of petroleum hydrocarbons in those samples may be from leaking underground storage tanks and not related to oil and gas production or shallow petroleum hydrocarbons outside of producing zones. Of the 21 RMP samples, 16 were within 500 m of at least 1 State Water Board Cleanup Site (fig. 6) that was a leaking underground storage tank, and almost all those Cleanup Sites had petroleum hydrocarbons included as a contaminant of concern (California State Water Resources Control Board, 2019b). However, petroleum hydrocarbons were not present in the four samples that contained gasoline oxygenates (methyl tert-butyl ether; appendix 3, table 3.5).

The frequent detection of manufactured VOCs is likely related to the long-term urban development of the area. Previous studies have identified relations between the occurrence of manufactured VOCs and urban land use (for example, Zogorski and others, 2006). Studies of VOCs in Los Angeles Basin groundwater have identified the Montebello Forebay (fig. 1), where shallow unconfined groundwater can be directly influenced by overlying land-surface activities, to be a regional recharge area for groundwater and a link to manufactured VOCs downgradient in confined aquifers (Shelton and others, 2001; Dawson and others, 2003; Goldrath and others, 2012). Similarly, manufactured VOCs were widely detected in groundwater underlying the urban area of Bakersfield, California (not shown), in the RMP study of the Fruitvale Oil Field (Wright and others, 2019). Manufactured VOCs have not been detected as frequently in RMP samples collected near oil fields in rural areas (Davis and others, 2018b; Rosecrans and others, 2021; Anders and others, 2022).

Gases

Methane is produced either as a byproduct of the biologic decomposition of organic matter (microbial methane) or in deep geologic formations with high pressure and temperature that produce coal, oil, or natural gas deposits (thermogenic methane; appendix 1; Kulongoski and others, 2018). Microbial methane is produced in highly reducing settings, such as swamps, landfills, or aquifers with geochemically reducing conditions (Whiticar and others, 1986). Thermogenic methane sometimes co-occurs with the heavier gases ethane, propane, isobutane (2-methylpropane), n-butane, isopentane (2-methylbutane), n-pentane, and hexane. The presence of propane and heavier gases is diagnostic of thermogenic contributions to formation gases (Schoell, 1980). Because ethane can also result from microbial processes in some cases, the presence of it is not necessarily diagnostic of thermogenic sources (Taylor and others, 2000).

Dissolved methane was detected in 18 RMP samples, at concentrations ranging from 0.0009 to 110 mg/L and with a median of 0.0085 mg/L (appendix 3, table 3.6). The three samples with a methane concentration greater than 5 mg/L (MT-6, MTE-1, and BNE-3; appendix 3, table 3.6) were from the deepest wells sampled (216, 293, and 424 m bls, respectively; table 1).

Heavier gases in the propane through pentane (C3–C5) range were detected in four of the samples (MT-10, MT-15, MT-16, and MT-17; appendix 3, table 3.6) that contained methane, indicating at least a small fraction of the gases at these sites are derived from thermogenic sources. Concentrations of these gases were small compared to the total volume of dissolved gas (0.0034 mole percent or less). Sample MT-10 was collected from a well to the west of the Montebello Oil Field, in the Bandini Oil Field (fig. 1). Sample MT-10 contained chemical characteristics consistent with a mixture of modern and pre-modern age water (see the “Age-Dating Tracers” section; table 3), manufactured VOCs, and no petroleum hydrocarbons. Sample MT-15 was collected from a deeper well (163 m deep) near oil-field infrastructure (table 2) and contained pre-modern water and petroleum hydrocarbon VOCs, but this sample did not contain manufactured VOCs (appendix 3, table 3.5). Four heavier gases were detected in sample MT-16 and three heavier gases were detected in sample MT-17 (appendix 3, table 3.6). Both samples (MT-16 and MT-17) were collected from relatively shallow wells (less than 60 m deep; table 1) and had a mixture of modern and pre-modern age water (table 3). Sample MT-16 was collected from a well on the north edge of the Montebello Oil Field near the OII landfill (fig. 6). The OII landfill is a possible source of methane and many other constituents, and sample MT-16 included the largest number of manufactured and petroleum hydrocarbon VOC detections of any sample (12 and 5 detections, respectively; table 3; appendix 3, table 3.5). Sample MT-17 was collected from a well near a golf course north of the oil field (fig. 6) and contained manufactured and petroleum hydrocarbon VOCs (table 3; appendix 3, table 3.5). Potential sources of thermogenic gases in samples are discussed in the “Potential Sources and Migration Pathways of Chemical Constituents” section.

The source of methane also can be determined based on the relative abundances of gases and the stable isotopes of hydrogen and carbon in methane (δ²H-CH₄, δ¹³C-CH₄; figs. 13A, 13B; appendix 3, table 3.6; Whiticar and others, 1986; McMahon and others, 2017; Kulongoski and others, 2018). Five samples had a methane concentration large enough (at least 0.13 mg/L) to analyze for isotopes, and isotope results indicated that the methane largely originated from microbial sources, most likely through carbon dioxide reduction processes for samples MT-6, MTE-1, BNE-3, and MTE-2 and fermentation processes for sample MT-17 (fig. 13B), and not from thermogenic sources associated with hydrocarbon-bearing formations in the Los Angeles Basin (Jeffrey and others, 1991). These results are mostly consistent with those from reconnaissance sampling of monitoring wells near oil fields that reported high concentrations of methane in some Los Angeles Basin groundwater are primarily derived from natural microbial processes in aquifers rather than thermogenic sources (Kulongoski and others, 2018). However, unlike results from reconnaissance sampling, MT-17 appears to be from fermentation processes. In addition, background methane was present throughout the intervals recorded by mud logs (fig. 5; appendix 3, table 3.2), providing evidence that its presence is common outside of oil- and gas-production zones. Although this approach indicated that methane in samples MT-6, MTE-1, BNE-3, MTE-2, and MT-17 largely originated from microbial sources, the presence of heavier gases in sample MT-17 is consistent with thermogenic sources. These results demonstrate a mixture of multiple sources of constituents in that sample. However, the relative abundances of gases and δ¹³C-CH₄ data in MT-17 show a strong microbial signal, so the thermogenic component appears to be small. In addition, landfill gas and some other shallow gas sources are associated with fermentation (Coleman and others, 1995).

Methane concentrations were available from 241 historical groundwater samples (collected between 2001 and 2014) from 43 monitoring wells clustered at 10 locations (fig. 4; California State Water Resources Control Board, 2017a). All but one of those well clusters were outside the Montebello Oil Field, and none of the wells had perforation or depth information available. Methane was detected in 21 of the 43 wells. The largest methane concentration from the historical groundwater samples was 4 mg/L, collected from a well at least 0.64 km from the closest oil and gas well. The isotopic composition of methane for historical samples was not available for determining the source of methane. As discussed in the “Potential Subsurface Sources and Migration Pathways” section, thermogenic gas was detected in soils and wells in the western part of the Montebello Oil Field between the 1970s and 1990s (California Public Utilities Commission, 2001) and in oil and gas observation wells in 2018 (fig. 5; California Department of Conservation, 2020b).

Age-Dating Tracers

The age-dating tracers tritium (³H) and carbon-14 in dissolved inorganic carbon (¹⁴C-DIC) were used to determine whether groundwater samples were composed of modern-age groundwater (recharged during or after 1953), pre-modern groundwater (pre-1953), or a mixture of both. The water was classified so potential sources and pathways of VOCs could be explored: if the aquifer contained modern-age groundwater, that would indicate chemical compounds in groundwater could have originated at or near the land surface, whereas if the aquifer contained pre-modern groundwater, it would indicate that chemical compounds in groundwater originated either at the land surface before 1953 or from subsurface sources.

There were 11 samples classified as modern, 5 samples classified as pre-modern, and 5 samples that had evidence of both modern and pre-modern age water (mixed; table 3; fig. 14). Two of the samples containing only modern water (MTE-2 and BNE-2) were collected from wells that were completed with perforations that were at least 150 m bls (table 1), indicating that modern recharge can reach deep intervals in some parts of the study area within a 66-year time frame. Conversely, one of the samples containing only pre-modern water (MT-14) was collected from a relatively shallow well near the OII landfill with perforations that were less than 91 m bls (table 1). The age-tracer data from sample MT-16 are inconclusive because the sample does not contain detectable ³H (or tritiogenic helium-3 [³He]) but does contain a modern concentration of ¹⁴C-DIC (greater than 100 pmc) and manufactured VOCs (table 3; appendix 3, table 3.5). The age indicators contradict each other in that if the sample is ³H “dead,” it should be older than 66 years in age, whereas the ¹⁴C-DIC points out a modern age, which should contain ³H. This result could be explained if the water was recharged shortly before 1953. That type of water would not have detectable ³H or ¹⁴C-DIC, but ¹⁴C-DIC in recharge would still be close to 100 pmc. The contradictory result also could be caused by an analytical issue either with the ¹⁴C-DIC or ³H measurement.

Almost all samples (10 of 11) interpreted as containing only modern-age groundwater had detections of manufactured VOCs; petroleum hydrocarbons co-occurred in one of these samples (table 3; fig. 14). None of the five samples containing pre-modern groundwater had manufactured VOCs, but two contained petroleum hydrocarbons (table 3). All five samples with evidence of mixed-age groundwater contained manufactured VOCs, and three of those five samples contained petroleum hydrocarbons (table 3). One possible interpretation of these results is that petroleum hydrocarbons are more commonly associated with pre-modern groundwater that has not been in contact with the land surface since the 1950s, and the petroleum hydrocarbons present in samples with mixed-age groundwater are more likely to be from the pre-modern groundwater portion than from the modern portion of the sample. The four samples with co-occurring manufactured VOCs and petroleum hydrocarbons (MT-5, MT-9, MT-16, and MT-17) were collected from shallow wells (less than 70 m deep; table 1). Sample MT-16 contained many more manufactured VOCs than petroleum hydrocarbons (fig. 10), and concentrations of manufactured VOCs were larger (appendix 3, table 3.5). In the other three samples, concentrations of manufactured VOCs and petroleum hydrocarbons were generally more similar. Four samples did not have detectable petroleum hydrocarbons or manufactured VOCs: two deep samples (MTE-1 and MT-6) containing pre-modern water, a shallow sample (MT-14) with pre-modern water, and a shallow sample (MT-8) with modern water (fig. 10; table 3). The shallow sample containing only pre-modern water was collected from a well completed in confined marine sediments of the Pico Formation that are closer to the land surface than in other parts of the study area. Groundwater in those sediments may not have been flushed by modern water.

Age-dating tracer results for historical groundwater samples in the Montebello Oil Field study area indicated that modern-age groundwater has reached relatively deep intervals of at least 405 m bls at some locations and pre-modern groundwater is present at relatively shallow depths (85 m) at other locations. Tritium concentrations were available for 107 historical groundwater samples from 103 wells, including 10 of the wells sampled for this study (California State Water Resources Control Board Division of Drinking Water, 2018; Water Replenishment District of Southern California, 2019b; U.S. Geological Survey, 2021). Sample dates ranged from 1994 to 2018. Well-perforation or well-depth information was available for 102 of the 103 sampled wells. Most of the sampled wells were in the valley alluvium adjacent to the Montebello Oil Field.

There were 21 historical samples analyzed for ³H and ¹⁴C-DIC, in addition to the 21 samples collected for this RMP study area (fig. 14). Of the 42 RMP and historical samples with ³H and ¹⁴C-DIC results, 19 samples were classified as modern, 9 were pre-modern, and 14 were mixed. In general, modern groundwater was shallower than pre-modern water. Modern groundwater was collected from wells completed between 28 and 366 m (median=134 m) bls, pre-modern groundwater was from wells completed between 90 and 424 m (median=293 m) bls, and mixed groundwater was from wells completed between 53 and 424 m (median=206 m) bls. Two of the historical samples contained manufactured VOCs and were classified as mixed-age groundwater (fig. 14); petroleum hydrocarbons were not detected in these samples. There were 19 samples with no detection of either manufactured or petroleum hydrocarbon VOCs; 8 classified as modern groundwater, 4 classified as pre-modern groundwater, and 7 classified as mixed.

Inorganic Tracers

Water from hydrocarbon-bearing formations is often saline, with TDS concentrations greater than 10,000 mg/L (Richter and Kreitler, 1991; Metzger and others, 2020). The TDS concentrations in RMP samples generally increased with depth (appendix 3, table 3.7), but all except the deepest groundwater sample (BNE-3) contained freshwater (TDS less than 3,000 mg/L). The median groundwater TDS concentration was 509 mg/L. Historical groundwater data indicated freshwater at depths less than 305 m bls and a median TDS of 490 mg/L (appendix 3, table 3.7). In contrast, historical oil-field water samples collected from wells in the study area, with depths to shallowest perforation of almost 600 to more than 2,500 m bls, contained TDS concentrations ranging from about 15,000 to 30,000 mg/L (Metzger and others, 2020), with a median TDS concentration of 24,275 mg/L.

In addition to elevated TDS concentrations, water from hydrocarbon-bearing formations commonly has elevated concentrations of specific inorganic constituents, such as ammonium, barium, boron, bromide, chloride, iodide, and lithium (Angino, 1970; Compton and others, 1992; McMahon and others, 2018; Wright and others, 2019). Available historical oil-field water samples in the study area indicate that the median concentrations of these constituents were several orders of magnitude larger than RMP and historical groundwater samples (appendix 3, table 3.7). The only exception was lithium, which was not detected in three of the four oil-field water samples analyzed for lithium. However, this result is most likely related to the large LDLs for those samples. The largest concentrations of inorganic constituents in RMP and historical groundwater samples were generally identified in the deepest wells (appendix 3, table 3.7).

Chloride concentrations relative to boron, bromide, and iodide indicated that three RMP samples (MT-6, MTE-1, and BNE-3) and several historical groundwater samples (1, 2, 3, 7) have chemical characteristics that plot along an apparent mixing line between either the native groundwater and a produced water end member or between native groundwater and a seawater end member (fig. 15). The RMP samples MT-16 and MT-13 plot along an apparent mixing line between the native groundwater and imported Colorado River water for chloride relative to iodide (fig. 15A). The historical groundwater sample used to represent the native groundwater end member (U.S. Geological Survey, 1998) for the mixing line analysis was selected because it represented typical native groundwater chemistry conditions in the Central Basin. However, the large distances between decile markers indicates that the analysis is not sensitive to changes in the plotted location of the native groundwater end member (fig. 15). Samples 1, 3, and 7 (fig. 15) were collected from wells that were also sampled for the RMP (samples BNE-3, MTE-1, and MT-6, respectively), and sample 2 was collected from a well perforated from 228 to 234 m bls. The samples were collected from wells completed at depths greater than 200 m bls and perforated in marine sediments. Age-dating tracer data from the RMP samples indicate that water in those wells is pre-modern (table 3). These results indicate that some deeper wells contain naturally occurring old groundwater that has (1) interacted with marine sediments containing water with a similar composition as produced water, (2) mixed with water from subsurface hydrocarbon-bearing formations (represented by oil-field water samples), or (3) both. Previous groundwater studies in the Los Angeles Basin do not provide evidence of seawater intrusion as far inland as the Montebello Oil Field (Reichard and others, 2003; Paulinski, 2021), although paleo-seawater may be present in some marine sediments that have not been flushed with meteoric water (Piper and Garrett, 1953). Shallower samples did not have inorganic chemical characteristics associated with mixing of water from hydrocarbon-bearing or marine formations. Recycled water also composes part of the water in the study area and is a potential end member. Chloride and boron concentrations for recycled water averaged 111 mg/L and 270 µg/L (Los Angeles County Sanitation Districts, 2021), respectively, plotting in the main cluster of RMP sample results on figure 15. Therefore, recycled water is not shown as an end member on figure 15.

The relative proportions of major inorganic constituents as charge equivalents can be useful for identifying groundwater origins and interactions (Appelo and Postma, 1999). A Piper diagram was used to show the relative contribution of major cations and anions, on a charge-equivalent basis, to the total ionic content of the water (Piper, 1944). Historical oil-field water chemistry samples in the study area generally clustered and had the highest relative proportions of sodium and chloride (fig. 16) compared to groundwater samples. Most groundwater samples collected for the RMP study are geochemically similar to most of the historical groundwater samples. Exceptions to this pattern are samples from the three deepest wells (MT-6, MTE-1, and BNE-3) and a shallow monitoring well perforated from 64 to 67 m bls (MT-9; table 1). Sample BNE-3 was most geochemically similar to the oil-field water samples and seawater and also was the deepest sample collected as part of the RMP (perforated from 405 to 424 m bls). Concentrations of TDS and many inorganic constituents in sample BNE-3 were greater than most other groundwater samples but less than in oil-field water (fig. 15; Metzger and others, 2020; U.S. Geological Survey, 2021). Samples MT-6, MT-9, and MTE-1 have higher proportions of sodium and bicarbonate and smaller proportions of calcium, magnesium, chloride, and sulfate than most groundwater samples (fig. 16). Their composition is consistent with cation exchange of calcium and magnesium for sodium, calcite precipitation, or sulfate reduction (Appelo and Postma, 1999). This pattern is similar to groundwater that has moved deeper into the flow system in the coastal California aquifers (Izbicki and others, 2003). Samples MT-6 and MTE-1 contain pre-modern water and elevated concentrations of TDS, ammonium, and many other inorganic constituents. Sample MT-9 was collected from a shallow well containing a mixture of modern and pre-modern groundwater (table 3) as well as relatively high concentrations of several trace elements (aluminum, arsenic, fluoride, lithium, and molybdenum), high pH, and relatively low concentrations of TDS. This sample is from a monitoring well in the San Gabriel Valley Superfund area (fig. 6) and its unique chemical composition could be affected by upgradient contamination and remediation activities. The RMP samples composed of modern water have a mixed cation and anion signal that is consistent with younger groundwater that has not extensively interacted with aquifer sediments (fig. 16). Those samples were similar in composition to most of the historical groundwater samples.

Geochemical Results in Relation to Water-Quality Benchmarks

The groundwater samples discussed in this report were compared to water-quality benchmarks to provide a frame of reference for the measured concentrations. Note that some groundwater samples were collected from monitoring or irrigation wells not used as a source of drinking water, and raw water collected from groundwater pumped from wells used for drinking-water supply undergoes treatment and blending before being delivered to consumers. Consequently, none of the samples discussed directly characterize drinking-water quality. These results cannot be firmly linked to hydrocarbon-bearing formations but provide a basis for understanding the relative magnitude of chemical constituents in RMP samples.

Chemical results for seven of the RMP samples indicated that at least one constituent did not meet enforceable U.S. Environmental Protection Agency (2018) and California State Water Resources Control Board (2019c) standards and health advisories (appendix 3, table 3.8). Turbidity was the only constituent that did not meet its standard in four of those seven samples. Of the samples, 67 percent (14 of 21) contained at least one chemical constituent at a concentration greater than nonenforceable drinking-water benchmarks (U.S. Environmental Protection Agency, 2018; California State Water Resources Control Board, 2019c) or USGS health-based screening levels (U.S. Geological Survey, 2018), which were created to ensure that drinking water does not have undesirable cosmetic or aesthetic effects (appendix 3, table 3.8). Most of the concentrations that were greater than benchmarks were for inorganic constituents. The only organic constituents present at concentrations greater than benchmark standards were the solvents tetrachloroethene (PCE; in sample MT-16), trichloroethene (TCE; in samples MT-12 and MT-16), and cis-1,2-dichloroethene (in sample MT-16). Sample MT-16 was collected from a monitoring well near the OII landfill (fig. 6), and sample MT-12 was untreated water collected from a public-supply well in the Bandini Oil Field (fig. 1). Large concentrations of PCE and TCE are consistent with previously published results for the Coastal Groundwater Basin (Shelton and others, 2001; Fram and Belitz, 2012) and with available historical data in the study area (California State Water Resources Control Board, 2017a, 2017b; California State Water Resources Control Board Division of Drinking Water, 2018; Water Replenishment District of Southern California, variously dated, 2019b; U.S. Geological Survey, 2021).