Environmental Health - Toxic Substances
Environmental Impacts Associated with Disposal of Saline Water Produced During Petroleum Production - Osage-Skiatook Petroleum Environmental Research Project [Completed]
Environmental impacts of improper handling or disposal of produced waters are of concern to petroleum producers, landowners, environmental organizations, and state and federal regulators. Large volumes of highly saline, sodium chrolride (NaCl)-rich produced waters are a ubiquitous by-product of oil and gas development in the Oklahoma Region. Releases of produced water can kill surface vegetation, salinize soils, increase soil erosion, and potentially contaminate local surface waters and ground-waters.
Evidence of environmental impacts from produced water typically is extensive near sites of historic oil and gas production. Prior to the 1970’s, operators commonly discharged produced waters into streams and unlined evaporation ponds. Modern operators are required to be more environmentally conscious when handling produced waters but occasional accidental releases, spills, or leaks do occur. Landowners and land managers, in particular, generally lack sufficient information or guidance to assess the nature, extent, and seriousness of contamination. Such information is needed to best prioritize sites for cleanup and to plan remediation.
In response to these information needs, the Central Energy Resource Science Center (CERSC) of the U.S. Geological Survey (USGS) initiated a project in 2001 to study the transport, fate, natural attenuation, and impacts of produced water contaminants at two study sites--sites A (historic, mostly pre-1935) and B (active, 1939 to present)--along the shores of 4,250-hectare Skiatook Lake in northeastern Oklahoma. CERSC geologists and chemists partnered with other hydrologists, chemists, botanists, and geophysicists from the USGS to conduct a multidisciplinary investigation. Funding was provided by the USGS Energy Resources and Toxics Substances Hydrology Programs, with additional monetary support for drilling from the U.S. Department of Energy. Other stakeholders providing critical information, site access, and logistical and in-kind support of science activities included the Army Corps of Engineers (landowner), the Osage Nation (mineral rights owner), the Bureau of Indian Affairs (trust responsibility), and the U.S. Environmental Protection Agency.
The team of scientists that worked on the project produced many publications, including comprehensive assessments of the impacts of produced water releases on vegetation, soils, ground water, and surface water Project publications also describe numerous methods for site characterization and assessment based on:
These techniques vary in complexity and data requirements but are readily transferrable to other sites and to differing scales of investigation. The techniques and provide a “tool kit” for characterization and assessment of environmental impacts of produced water.
The information below provides an overview of the findings of the project. Please refer to the bibliography of project publications for more detailed discussions.
The spatial distribution of soil salinity in relation to areas of visible salt scarring at sites A and B was evaluated by obtaining 15 centimeter (cm)-deep soil cores and analyzing their aqueous extracts for chloride, bromide, sulfate, and specific conductance. More extensive measurements at site B (active) indicated that surface soils located within, adjacent to, or between salt scars were all anomalously saline and chlorine (Cl)-rich. The ratio of chlorinde to bromide (Cl/Br) in Cl-rich extracts closely matched that of modern produced water. Some deeper cores taken within areas of high surface salinity showed contamination throughout the entire thickness of 0.5 to 2 meter (m) of colluvium and alluvium. Based on these observations, remediation of salt-affected soils will require protection from capillary rise of this deeper salt.
The subsurface distribution of salt within underlying bedrock was monitored by complementary observations of well-water chemistry, aqueous extracts of uniformly crushed rock core, and geophysical well logs. At both sites A and B salts have penetrated as much as 3 to 8 m into underlying bedrock. Plumes of NaCl-rich ground water below the sites move down gradient toward Skiatook Lake and expand laterally and to greater depths. Rock properties and hydrologic characteristics at the two sites retard movement of the saline ground water and act to confine salts within a limited volume of rock at depth. Flow and solute transport modeling indicate that the filling of Skiatook Lake in 1987 likely imposed a hydraulic head that contributed to the stagnation of the contaminant plumes.
Salt, trace element, and hydrocarbon inputs to Skiatook Lake by surface water runoff or shallow ground water seepage are extremely diluted by lake water. However, some impacts to benthic organisms may occur in lake sediments located near submerged former brine storage pits or near areas where shallow saline ground water discharges to the lake. Pore waters collected from the upper 16 to 40 cm at such locales offshore from site B (active) had (1) anomalous chloride concentrations; (2) Cl/Br ratios compatible with modern produced water; and (3) concentrations of dissolved Se, Pb, Cu, and Ni that could be toxic to aquatic life. The depth-wise distribution of Cl in pore waters was explained by diffusion-dominated transport from deeper brine pit sediments, and by layer-directed advective flow near an offshore saline seep.
Geophysical measurements of subsurface conductivities obtained from ground surveys (traverses or grids) at the two sites and from borehole conductivity logs provided continuous readings that indicated the volume of rocks affected by highly saline pore fluids. Such distributions are more complex and wide ranging than the surface expressions of salt scars. Ground conductivity surveys focused on the uppermost few meters, whereas borehole logs extended measurements to borehole depths. Measurements taken during low lake levels or from boat-mounted traverses indicated extension of saline ground water plumes for several tens of meters offshore of sites A and B. The geophysical data provided a framework for comparison with other indications of subsurface salinity such as ground water/pore water chemistry and Cl content of aqueous extracts of soil and rock.
Important observed geologic controls on the movement of produced water in the shallow subsurface beneath sites A and B included diverse bedrock units (sandstone, shale, siltstone, dolomite) of contrasting porosity and permeability, rapid lateral facies changes within bedrock units, dip of strata, and the number and extent of vertical fractures. Permeability of clay-bearing units may also be affected by temporal changes in the salinity of pore water. Highly saline waters promote clay flocculation and increased permeability, whereas fresh waters promote clay dispersion and decreased permeability.
Dissolved radium concentrations in modern produced water at site B (active) are attenuated by incorporation of radium into highly insoluble barium sulfate that forms as scale deposits in tanks or pipes or as authigenic precipitates resulting from reaction of barium-rich produced water with soluble sulfate in local soils. Depth-wise changes in the ratio of 228Ra/226Ra in sediment from a brine storage pit were used to estimate the rate of burial of radium-bearing barite particles. Ground waters at site B are oversaturated with respect to barite and shallow soils are clayey and poorly drained. Under these conditions any particles of radium-bearing barite present in soils have limited solubility and environmental mobility.
Dissolved chloride loads (concentration (times) discharge) in tributary streams entering Skiatook Lake and (or) geospatial analysis of well density, well proximity to streams, stream length, and subdrainage areas can be used to identify tributaries with the greatest potential for chloride contamination and to assess the relative contributions of chloride to Skiatook Lake from various subdrainages. Mass balance calculations for chloride indicated that at the time of sampling the lake water concentration could be explained as a flow-weighted average of chloride concentrations provided from stream sources. The inference is that the amount of chloride contributed to the lake from other sources such as ground water seepage was small enough to be accommodated within the error of the mass balance calculations.
Leaf composites from 34 blackjack oaks in woodlands surrounding site A (historic) were sampled with respect to proximity to obvious salt sources such a salt scars and former brine storage areas. Leaves were washed, dried, minced, and submitted for analysis of chloride as a possible indication of root zone impacts by NaCl salt. Results indicated anomalous concentrations (>400 parts per million (ppm)) of chloride in leaves from trees located within 45 m of salt scars. Other trees with anomalous chloride concentrations in leaves were located near a former tank battery site and a trench used to transport produced water. Use of such biogeochemical sampling can complement or substitute for more expensive surveys of subsurface salts by drilling, well sampling, or geophysics.
Back to Petroleum Related Contamination Research Projects
Back to Subsurface Point-Source Contamination Research Projects