Abstract

The Fruitvale oil field is directly beneath the city of Bakersfield and was identified as a high-priority study area for the California State Water Resources Control Boards Regional Monitoring Program of Water Quality in Areas of Oil and Gas Production (RMP) because of the fields proximity to protected groundwater resources and volumes of injection. From 2014 to 2017, the United States Geological Survey (USGS) compiled information on the study area, including available groundwater and produced water quality data, water and oil well construction and geophysical records, oil field development history and injection data, and hydrogeologic features including formation depth surfaces, faults, and vertical pressure gradient data.

Both groundwater and oil field reservoirs were sampled for the Fruitvale RMP.A conceptual model of the Fruitvale study area was developed and generalized groundwater flow paths were identified.Based on these flow paths and the depths of available wells, USGS staff selected wells and sought permission from owners to collect samples.Fourteen groundwater wells were sampled for the full suite of RMP constituents. The oil field reservoir sampling was designed to characterize the predominant oil-producing area/formation combinations in the field and produced water being reinjected into the oil reservoir. USGS staff coordinated with operators to identify active sites meeting the selection criteria and to obtain permission for sampling. Six oil wells and two injectate sites were sampled.

Analysis of samples collected for the RMP, and historical water quality data from the Fruitvale oil field area, show little evidence of formation fluids mixing with the overlying groundwater.Two of the 14 groundwater wells sampled had enriched carbon isotopic values consistent with mixtures of <1 to about 13% produced water, and one of these wells had a trace detection of methane with isotopic values consistent with a petroleum source.The source of produced water mixed in these two groundwater samples cannot be resolved based on available data.Potential pathways allowing for oil reservoir fluids to mix with groundwater include: (1) historic disposal of produced water in surface ponds; (2) upward movement of oil formation fluids via faults, compromised well casings or annular seals; (3) via produced- water injection wells with compromised well casings or annular seals. Overall groundwater quality in the aquifer overlying the Fruitvale oil field appears good, reflecting in part that this is an area of rapid flushing of the aquifer system due to the proximity of the Kern River.

Additional observations supporting these conclusions include:

The stable isotopic signature of groundwater samples was lighter than that of produced water samples, consistent with little, if any, mixing; however, uncertainties in end - member compositions limit the resolution of these data to discern small mixtures of produced water in groundwater.The isotopic data, younger apparent recharge age, and larger fractions of post-1950s water in samples closer to the river indicate that the Kern River is a major source of groundwater recharge in the study area.Mean ages of the post- 1950s fractions ranged from about 15 to 49 years within 1 km of the river to 29 to 46 years at distances greater than 4 km from the river.

Total dissolved solids (TDS) concentration in groundwater samples were generally lower (144754 mg/ l) than produced water samples (58524, 930 mg/ l).The generally low TDS concentration in groundwater indicates that a potentially small amount of mixing with oil formation water has not significantly degraded groundwater quality.Groundwater sample concentrations for total dissolved solids, sodium, chloride and pH increased with distance from the Kern River in a manner consistent with the dissolution of aquifer minerals and cation exchange being the dominant processes controlling salinity in groundwater.Generally, ?13C values were substantially lighter in groundwater samples than in produced water samples, except for the two samples where geochemical (NETPATHXL) model simulations indicate mixing of Kern River water with produced water.

Volatile organic compounds (VOCs), primarily solvents and disinfection byproducts typically derived from industrial/urban land uses, were detected in all groundwater samples at concentrations below drinking-water maximum contaminant levels, with the exception of a single detection of the solvent 1,2-Dibromoethane. Of the classes of VOCs detected in groundwater samples, petroleum hydrocarbons were the least frequently detected class and always co-occurred with other classes of VOCs. Three of the five wells with detections of petroleum hydrocarbons were located in close proximity to known groundwater contamination from leaking aboveground storage tanks at a petroleum refining facility.

The water wells available for sampling were located in the upper 800 feet of the aquifer, the portion used for public water supply; monitoring of the relatively shallow aquifer zones does not address processes occurring more deeply n protected waters overlying the oil field. Deeper monitoring would require installing monitoring wells, which would be costly and may not be justified given the good water quality in the groundwater resources currently being used. Focused studies around some groundwater wells with trace detections of constituents that could be derived from formation water, and that are in close proximity to potential risk factors, may be warranted.