Applied Geochemistry (v.22, #10)

by Yousif K. Kharaka; James K. Otton (2095-2098).

Past disposal of oil-field brine at the surface has caused substantial contamination of water resources in Kansas. Natural saline water occurs in and discharges from Permian bedrock in parts of the state, and other anthropogenic sources of saline water exist, requiring clear identification of different sources. Time-series analysis of Cl concentration and streamflow relative to pre-contamination contents, and end-member mixing plots, especially for Br and Cl, are practical methods for source differentiation and quantification. Although regulations preventing escape of saltwater from oil wells were first passed in Kansas in 1935, much oil and gas brine was disposed on the surface through the 1940s. Hydrogeologic characteristics of the areas with past surface disposal of oil brine differ appreciably and result in large differences in the ratio of saltwater transported in streams or ground water. Much of the brine disposed during the 1910s to 1940s in an area of silty clay soils overlying shale and limestone bedrock in south-central Kansas soon ran off or was flushed from the surface by rain into streams. Chloride concentration in the rivers draining this area often exceeded 1000 mg/L after the start of oil production up to the 1950s. Chloride content in the rivers then generally declined to about 100 mg/L or less in recent low flows. Oil brine was also disposed in surface ponds overlying the unconsolidated High Plains aquifer in south-central Kansas from the latter 1920s into the 1940s. Most of the surface-disposed brine infiltrated to the underlying aquifer. Where the High Plains aquifer is thin, saltwater has migrated along the top of clay layers or the underlying shaly bedrock and either discharged into small streams or flowed into thicker parts of the aquifer. Where the aquifer is thick, surface-disposed oil brine moved downward until reaching clay lenses, migrated latterly to the edge of the clay, and again moved downward if still dense enough. Water-level declines from pumping have increased the lateral migration rate of the saltwater contamination in the aquifer towards water-supply wells. The period of flushing most of the surface-disposed saltwater from the area of shale and limestone bedrock is on the order of many decades but is at least many centuries for the deeper parts of the High Plains aquifer.

The first geological materials impacted by oil field wastes released into near-surface environments in southern Louisiana, USA, are typically clays and silts. Clay minerals within these siliciclastic sediments have the potential for altering the composition of produced water wastes through cation exchange. The general relations between the composition of adsorbed cations and interstitial water salinity in brine-contaminated samples from a site in southeastern Louisiana are consistent with previous studies of multicomponent exchange in groundwater systems of varying salinity. The divalent cations Ca and Mg dominate as adsorbed cations at low salinities (<1200 mg/L), but Na is dominant at moderate to high salinities (up to 53,000 mg/L). The change in the proportions of adsorbed cations is a non-linear function of salinity, and the transition from Ca-dominated adsorption to Na-dominated adsorption occurs over a narrow range of salinities.Calculated interstitial water compositions, assuming exchange equilibrium, are consistent with the source of contamination being produced waters having Na as the dominant dissolved cation, followed by Ca, rather than some other type of saline waste. The calculated partitioning for Ba indicates that in low to moderate salinity pore waters, Ba, and by extension Ra, are nearly quantitatively adsorbed on the clays and would be of low mobility in a physically active groundwater system. However, at the elevated salinities typical of many produced waters, Ba and Ra are not preferentially adsorbed.

Two samples of produced-water collected from a storage tank at US Geological Survey research site B, near Skiatook Lake in northeastern Oklahoma, have activity concentrations of dissolved 226Ra and 228Ra that are about 1500 disintegrations/min/L (dpm/L). Produced-water also contains minor amounts of small (5–50 μm) suspended grains of Ra-bearing BaSO4 (barite). Precipitation of radioactive barite scale in the storage tank is probably hindered by low concentrations of dissolved SO4 (2.5 mg/L) in the produced-water. Sediments in a storage pit used to temporarily collect releases of produced-water have marginally elevated concentrations of “excess” Ra (several dpm/g), that are 15–65% above natural background values. Tank and pit waters are chemically oversaturated with barite, and some small (2–20 μm) barite grains observed in the pit sediments could be transferred from the tank or formed in place. Measurements of the concentrations of Ba and excess Ra isotopes in the pit sediments show variations with depth that are consistent with relatively uniform deposition and progressive burial of an insoluble Ra-bearing host (barite?). The short-lived 228Ra isotope (half-life = 5.76 a) shows greater reductions with depth than 226Ra (half-life = 1600 a), that are likely explained by radioactive decay. The 228Ra/226Ra activity ratio of excess Ra in uppermost pit sediments (1.13–1.17) is close to the ratio measured in the samples of produced-water (0.97, 1.14). Declines in Ra activity ratio (excess) with sediment depth can be used to estimate an average rate of burial of 4 cm/a for the Ra-bearing contaminant. Local shallow ground waters contaminated with NaCl from produced-water have low dissolved Ra (<20 dpm/L) and also are oversaturated with barite. Barite is a highly insoluble Ra host that probably limits the environmental mobility of Ra at site B.

Highly saline produced water was released from multiple sources during oil field operations from 1913 to 1973 at the USGS research Site A on Skiatook Lake in northeastern Oklahoma. Two pits, designed to hold produced water and oil, were major sources for release of these fluids at the site. Produced water spills from these and other features moved downslope following topography and downdip by percolating through permeable eolian sand and colluvium, underlying permeable sandstone, and, to a lesser extent, through shales and mudstones. Saline water penetrated progressively deeper units as it moved through the gently dipping bedrock to the north and NW. A large eroded salt scar north of the pits coincides with underlying fine-grained rocks that have retained substantial concentrations of salt, causing slow revegetation. Where not eroded, thick eolian sand or permeable sandstone bedrock is near the surface, and vegetation has been little affected or has reestablished itself after the introduced salt was flushed by precipitation. The extent of salt-contaminated bedrock extends well beyond existing surface salt scars. These results indicate that one of the legacies of surface salt spills can be a volume of subsurface salinization larger than the visible surface disturbance.

Spillage and improper disposal of saline produced water from oil wells has caused environmental damage at thousands of sites in the United States. In order to improve understanding of the fate and transport of contaminants at these sites, the U.S. Geological Survey carried out multidisciplinary investigations at two oil production sites near Skiatook Lake, Oklahoma. As a part of this effort, the hydrology and subsurface transport of brine at OSPER site “A”, a tank battery and pit complex that was abandoned in 1973, was investigated. Based on data from 41 new boreholes that were cored and completed with monitoring wells, a large (∼200 m × 200 m × 20 m) plume of saline ground water was mapped. The main dissolved species are Na and Cl, with TDS in the plume ranging as high as 30,000 mg/L. Analysis of the high barometric efficiency of the wells indicated a confined aquifer response. Well-slug tests indicated the hydraulic conductivity is low (0.3–7.0 cm/day). Simplified flow and transport modeling supports the following conceptual model: (1) prior to the produced water releases, recharge was generally low (∼1 cm/a); (2) in ∼60 a of oil production enough saline produced water in pits leaked into the subsurface to create the plume; (3) following abandonment of the site in 1973 and filling of Skiatook Reservoir in the mid-1980s, recharge and lateral flow of water through the plume returned to low values; (4) as a result, spreading of the brine plume caused by mixing with fresh ground water recharge, as well as natural attenuation, are very slow.

Fate and groundwater impacts of produced water releases at OSPER “B” site, Osage County, Oklahoma by Yousif K. Kharaka; Evangelos Kakouros; James J. Thordsen; Gil Ambats; Marvin M. Abbott (2164-2176).
For the last 5 a, the authors have been investigating the transport, fate, natural attenuation and ecosystem impacts of inorganic and organic compounds in releases of produced water and associated hydrocarbons at the Osage-Skiatook Petroleum Environmental Research (OSPER) “A” and “B” sites, located in NE Oklahoma. Approximately 1.0 ha of land at OSPER “B”, located within the active Branstetter lease, is visibly affected by salt scarring, tree kills, soil salinization, and brine and petroleum contamination. Site “B” includes an active production tank battery and adjacent large brine pit, two injection well sites, one with an adjacent small pit, and an abandoned brine pit and tank battery site. Oil production in this lease started in 1938, and currently there are 10 wells that produce 0.2–0.5 m3/d (1–3 bbl/d) oil, and 8–16 m3/d (50–100 bbl/d) brine. Geochemical data from nearby oil wells show that the produced water source is a Na–Ca–Cl brine (∼150,000 mg/L TDS), with high Mg, but low SO4 and dissolved organic concentrations. Groundwater impacts are being investigated by detailed chemical analyses of water from repeated sampling of 41 boreholes, 1–71 m deep. The most important results at OSPER “B” are: (1) significant amounts of produced water from the two active brine pits percolate into the surficial rocks and flow towards the adjacent Skiatook reservoir, but only minor amounts of liquid petroleum leave the brine pits; (2) produced-water brine and minor dissolved organics have penetrated the thick (3–7 m) shale and siltstone units resulting in the formation of three interconnected plumes of high-salinity water (5000–30,000 mg/L TDS) that extend towards the Skiatook reservoir from the two active and one abandoned brine pits; and (3) groundwater from the deep section of only one well, BR-01 located 330 m upslope and west of the site, appear not to be impacted by petroleum operations.

Shallow ground water at US Geological Survey research site B in northeastern Oklahoma is contaminated with NaCl-rich brine from past and present oil production operations. Contaminated ground water provides a potential source of salts, metals, and hydrocarbons to sediment and water of adjacent Skiatook Lake. A former brine storage pit 10 m in diameter that is now submerged just offshore from site B provides an additional source of contamination. Cores of the upper 16–40 cm of lake sediment were taken at the submerged brine pit, near an offshore saline seep, and at a location containing relatively uncontaminated lake sediment. Pore waters from each 2-cm interval were separated by centrifugation and analyzed for dissolved anions, cations, and trace elements. High concentrations of dissolved Cl in pore waters (200–5000 mg/L) provide the most direct evidence of contamination, and contrast sharply with an average value of only about 37 mg/L in Skiatook Lake. Chloride/Br mass ratios of 220–240 in contaminated pore waters are comparable to values in contaminated well waters collected onshore. Dissolved concentrations of Se, Pb, Cu and Ni in Cl-rich pore waters exceed current US Environmental Protection Agency criteria for probable toxicity to aquatic life. At the submerged brine storage pit, the increase of Cl concentration with depth is consistent with diffusion-dominant transport from deeper contaminated sediments. Near the offshore saline seep, pore water Cl concentrations are consistently high and vary irregularly with depth, indicating probable Cl transport by layer-directed advective flow. Estimated annual contributions of Cl to the lake from the brine storage pit (∼20 kg) and the offshore seep (∼9 kg) can be applied to any number of similar sources. Generous estimates of the number of such sources at site B indicate minimal impact on water quality in the local inlet of Skiatook Lake. Similar methodologies can be applied at other sites of NaCl contamination surrounding Skiatook Lake and elsewhere.

Releases of NaCl-rich (>100 000 mg/L) water that is co-produced from petroleum wells can adversely affect the quality of ground and surface waters. To evaluate produced water impacts on lakes, rivers and streams, an assessment of the contamination potential must be attainable using reliable and cost-effective methods. This study examines the feasibility of using geographic information system (GIS) analysis to assess the contamination potential of Cl to Skiatook Lake in the Hominy Creek drainage basin in northeastern Oklahoma. GIS-based predictions of affects of Cl within individual subdrainages are supported by measurements of Cl concentration and discharge in 19 tributaries to Skiatook Lake. Dissolved Cl concentrations measured in October, 2004 provide a snapshot of conditions assumed to be reasonably representative of typical inputs to the lake. Chloride concentrations ranged from 5.8 to 2300 mg/L and compare to a value of 34 mg/L in the lake. At the time of sampling, Hominy Creek provided 63% of the surface water entering the lake and 80% of the Cl load. The Cl load from the other tributaries is relatively small (<600 kg/day) compared to Hominy Creek (11 900 kg/day) because their discharges are relatively small (<0.44 m3/s) relative to Hominy Creek (3.1 m3/s). Examination of chemical components other than Cl in stream and lake waters indicates that many species, such as SO4, cannot be used to assess contamination potential because they participate in a number of common biogeochemical processes that alter their concentrations.GIS estimates of well density, well proximity to tributaries (wells within 200 m), and subdrainage area identified tributaries with the most potential for Cl contamination. Tributaries with large measured Cl concentrations (>150 mg/L) were generally in subdrainages with greater well density (>15 wells/km2), relatively large numbers of petroleum wells in close proximity (>2 proximity wells/stream km), and relatively small discharge (<0.005 m3/s). GIS calculations of subdrainage areas can be used to estimate the expected discharge of the tributary for each subdrainage. GIS-based assessment of Cl contamination potential at Skiatook Lake and at other lakes surrounded by oil fields can proceed even when direct measurements of Cl or discharge in tributary streams may be limited or absent.

Airborne, ground, and borehole electromagnetic (EM) induction measurements were combined with surface-water chemical analyses to determine the extent of salinization as well as identify salinity sources and migration mechanisms that elevate total dissolved solids (TDS), Cl, and SO 4 2 - concentrations in Petronila Creek, a shallow stream that drains a small (1600 km2) basin on the Texas coastal plain. A multifrequency airborne EM induction survey measured apparent electrical conductivity of the ground to depths of a few tens of meters along the axis of the stream and along parallel flight lines within a 150 km2 area extending from where the stream is fresh to the estuarine mixing zone. Apparent conductivity sections and maps were analyzed to identify stream segments where high conductivity indicates near-surface salinization and where saline baseflow increases the salinity load. Elevated conductivities at shallow depths beneath the stream lie adjacent to extensive conductive areas in oil fields and along ditches that once carried brine produced from the oil fields. These highly conductive areas delineate salinization that dominantly is caused by past discharge of produced brine into ditches and pits, infiltration into sandy, permeable horizons, lateral subsurface migration, and eventual discharge into the stream. Streamflow measurements and chemical analyses show that the TDS load increased along two highly conductive streambed segments upstream from the estuarine mixing zone. Extensive salinization between the ditches and the stream could provide a continuing source of salinity. A third highly conductive segment coincides with the zone of estuarine mixing, where there is geophysical evidence of possible subsurface seawater intrusion.Multifrequency EM induction spectral images (the sectional depiction of apparent conductivity measured at multiple frequencies) along the stream axis, produced with minimal geophysical processing, helped identify stream segments that receive saline baseflow and guided surface-water sampling that allowed quantification of salinity loads. Apparent conductivity maps helped determine the salinization extent and identify salinity sources and migration mechanisms. Both the stream-axis and gridded airborne EM surveys were effective approaches to investigate salinization and interaction between ground water and surface water in this small basin, but multifrequency stream-axis images are a more practical alternative in larger basins where discrete salinity sources are poorly known.

Lessons learned in remediation and restoration in the Oklahoma prairie: A review by Kerry L. Sublette; J. Bryan Tapp; J. Berton Fisher; Eleanor Jennings; Kathleen Duncan; Greg Thoma; Josh Brokaw; Tim Todd (2225-2239).
For almost a decade the Tallgrass Prairie Preserve in Oklahoma has been used as a field laboratory for the investigation of aspects of the remediation and restoration of oil and brine spills. Objectives of this work have included: (1) simplification of the remediation process and lowering the cost of remediation; (2) the development of methods to accelerate or jump-start the restoration process; and (3) determining appropriate metrics for assessing the status of soil ecosystem recovery. This research has resulted in a number of lessons learned that will be presented here which can be exported to other exploration and production sites, especially sites located in sensitive ecosystems. Key observations have included the role of a fertilizer amendment in linking the remediation and restoration process at an oil-impacted site, the use of nematodes as ecological indicators in the restoration of oil- and brine-impacted sites, and the development of a two-stage process for remediation of brine impacted sites that does not include significant use of gypsum.

Organic compounds in produced waters from coalbed natural gas wells in the Powder River Basin, Wyoming, USA by William H. Orem; Calin A. Tatu; Harry E. Lerch; Cynthia A. Rice; Timothy T. Bartos; Anne L. Bates; Susan Tewalt; Margo D. Corum (2240-2256).
The organic composition of produced water samples from coalbed natural gas (CBNG) wells in the Powder River Basin, WY, sampled in 2001 and 2002 are reported as part of a larger study of the potential health and environmental effects of organic compounds derived from coal. The quality of CBNG produced waters is a potential environmental concern and disposal problem for CBNG producers, and no previous studies of organic compounds in CBNG produced water have been published. Organic compounds identified in the produced water samples included: phenols, biphenyls, N-, O-, and S-containing heterocyclic compounds, polycyclic aromatic hydrocarbons (PAHs), aromatic amines, various non-aromatic compounds, and phthalates. Many of the identified organic compounds (phenols, heterocyclic compounds, PAHs) are probably coal-derived. PAHs represented the group of organic compounds most commonly observed. Concentrations of total PAHs ranged up to 23 μg/L. Concentrations of individual compounds ranged from about 18 to <0.01 μg/L. Temporal variability of organic compound concentrations was documented, as two wells with relatively high organic compound contents in produced water in 2001 had much lower concentrations in 2002.In many areas, including the PRB, coal strata provide aquifers for drinking water wells. Organic compounds observed in produced water are also likely present in drinking water supplied from wells in the coal. Some of the organic compounds identified in the produced water samples are potentially toxic, but at the levels measured in these samples are unlikely to have acute health effects. The human health effects of low-level, chronic exposure to coal-derived organic compounds in drinking water are currently unknown. Continuing studies will evaluate possible toxic effects from low level, chronic exposure to coal-derived organic compounds in drinking water supplies.

The Black Warrior Basin of the southeastern United States hosts one of the world’s most prolific and long-lived coalbed methane plays, and the wealth of experience in this basin provides insight into the relationships among basin hydrology, production performance, and environmental issues. Along the southeast margin of the basin, meteoric recharge of reservoir coal beds exposed in an upturned fold limb exerts a strong control on water chemistry, reservoir pressure, and production performance. Fresh-water plumes containing Na–HCO3 waters with low TDS content extend from the structurally upturned basin margin into the interior of the basin. Northwest of the plumes, coal beds contain Na–Cl waters with moderate to high-TDS content. Carbon isotope data from produced gas and mineral cements suggest that the fresh-water plumes have been the site of significant bacterial activity and that the coalbed methane reservoirs contain a mixture of thermogenic and late-stage biogenic gases.Water produced from the fresh-water plumes may be disposed safely at the surface, whereas underground injection has been used locally to dispose of highly saline water. Wells in areas that had normal hydrostatic reservoir pressure prior to development tend to produce large volumes of water and may take up to 4 a to reach peak gas production. In contrast, wells drilled in naturally underpressured areas distal to the fresh-water plumes typically produce little water and achieve peak gas rates during the first year of production. Environmental debate has focused largely on issues associated with hydrologic communication between deep reservoir coal beds and shallow aquifers. In the coalbed methane fields of the Black Warrior Basin, a broad range of geologic evidence suggests that flow is effectively confined within coal and that the thick intervals of marine shale separating coal zones limit cross-formational flow.