The goals of this project are to develop constructed wetland systems for treatment and beneficial use of produced water, and to conduct scientific studies to address ecological, environmental, and regulatory concerns that limit options for managing produced water, including surface discharge water.

Clemson University, Clemson, SC 
Chevron, Houston, TX

New technologies and advances in existing technologies are needed to minimize the environmental impacts associated with hydrocarbon production from our nation's conventional and unconventional resources. Unconventional resources currently account for 30% of U.S. gas production, and the importance of unconventional gas is expected to increase in the next 25 years (NPC, 2007). Most U.S. unconventional gas reserves are in gas shales and coal beds, with the remainder in tight gas sands. Important oil resources are present in oil shale, and industry is moving rapidly to develop these reserves. Unconventional oil resources in the United States include two trillion barrels in oil shales, 16 billion barrels in oil sands, and 73 billion barrels remaining heavy oil-in-place. New technologies for the continued development of both oil and gas resources are needed to help meet our nation's energy needs. These new technologies must be environmentally acceptable, since development of both conventional and unconventional resources is becoming increasingly constrained by environmental concerns and regulations. Technologies are needed that reduce the cost of environmental compliance and protect the environment.

New methods, using environmentally acceptable technology, are needed to efficiently handle the large volumes of produced water generated by oil and gas production. Managing large volumes of produced water and the associated costs have limited development of both conventional and unconventional reservoirs. Cost-effective technologies to treat and manage these waters will contribute greatly to the development of our oil and gas resources. The quality and quantity of waters produced from the development of conventional and unconventional resources varies widely from site to site. Constituents that must be treated prior to re-use or discharge include anions, cations (including metals), organics, dissolved solids, and suspended solids. Examples of specific elements of concern in produced waters are arsenic, iron, magnesium, chromium, zinc, boron, barium, selenium, and manganese.

Current technologies for treating produced waters are costly, especially considering the large volumes of water produced and energy requirements, and are often unable to meet new, rigorous water-quality standards. Current treatment options include ion exchange, reverse osmosis (RO), electro dialysis reversal, and mechanical evaporation. A major limitation of most of these technologies is that their operating costs rise dramatically as the price of energy increases. Although targeted constituents in produced waters vary among oil and gas reservoirs, metals, organics, and biocides are among the more difficult to treat and tend to limit the utility of these waters for re-use or other purposes. Mercury, arsenic, and selenium present a difficult challenge for many types of current treatment options, and concentrations of these constituents in some produced waters can be high. Low-cost and effective technology for treatment and potential re-use of waters associated with development of conventional and unconventional resources is needed to sustain energy production in the United States and to support commercial development of previously sub-economic oil and gas resources. The cost of treatment using constructed wetland systems has proven consistently lower than other technologies. Indeed, these costs are typically an order of magnitude (or more) less than alternative technologies.

Development of low-cost methods to handle large volumes of produced water has the potential to increase production in existing areas of development and to open new areas for exploration and production. The use of constructed wetland treatment systems has great potential for contributing to the development of oil and gas resources by providing an effective and low-cost method of treating produced waters for beneficial re-use and discharge. In addition to greatly reducing the environmental risks associated with current practices, produced waters renovated by constructed wetland treatment systems have the potential to be re-used for a variety of purposes, such as irrigation, livestock watering, cooling-tower water, municipal water use, domestic use, discharge to receiving aquatic systems for other use downstream, and support of critical aquatic life and wildlife. This can allow continued operation of existing wells and lead to increased drilling and production.

Constructed wetland systems for treating produced waters can include centralized facilities (pipe, haul, and treat) or decentralized facilities designed for a single well or for a few nearby wells. Portable "package" constructed wetland treatment systems can be pulled to a site by truck, set up, and begin treatment immediately. These "ready-to-go" systems may be very useful during fracture stimulation and during high initial water production from unconventional gas wells. Cost benefits of using constructed wetland treatment systems rather than conventional treatment approaches have been demonstrated in previous studies (Myers et al., 2001; Mooney and Murray-Gulde, 2008). Developing low-cost and environmentally acceptable solutions for handling produced waters will contribute to meeting the increasing oil and natural gas demands of the future.

Clemson has identified chemical, physical, and risk characteristics of produced waters from conventional gas, coalbed methane (CBM), tight gas, and shale gas. The produced water characterization provided data for identifying the constituents, including dissolved salts, metals, metalloids, organics, and post-RO constituents such as ammonia. A produced water inventory compiled from the literature review and listing these data by geologic basin and formation has been created. To obtain additional data on produced waters, researchers contacted various potential data sources, including state geological surveys, state oil and gas boards, and industry personnel. Much of the conventional gas produced water data have been obtained from basins that co-produce both oil and natural gas. The database currently contains records for 625 samples of co-produced water from 28 geologic basins. The majority of the produced water data for unconventional gas were obtained from formations from which coalbed methane is produced. Data from 319 CBM samples from several geologic basins, including the San Juan, Raton, Powder River, and Arkoma, were entered into the database.

Based on analyses by project researchers, constituents of concern identified in produced waters include arsenic and selenium. These "metalloids" are amenable to treatment in specifically designed constructed wetland systems through targeted biogeochemical processes. In addition, dissolved and dispersed organic constituents can be removed by oxidation in constructed wetland treatment systems. Various combinations of these inorganic and organic constituents may be specifically targeted for treatment depending on the composition of and designated post-treatment use for the produced water. Constructed wetland treatment systems can be designed to promote nitrification for the treatment of post-RO produced water.

Four pilot-scale constructed wetland systems were built at Clemson and received various synthetic produced waters. Each system was designed to remove certain target contaminants through different biological and chemical processes. Preliminary sampling and analysis have indicated successful removal of metals (copper, nickel, cadmium, zinc, iron, and manganese), oil, grease, and low molecular weight organics. The addition of calcium carbonate to wetland cells significantly increased metal removal. Selenium removal goals were met when organic amendments were added to increase bacterial activity. Oil and grease removal goals were achieved with both subsurface flow and free water surface flow systems. Based on these analyses, wetland system operating parameters, such as retention time, may be adjusted to improve treatment.

A presentation on ammonia removal in wetlands was given at the South Central Geological Society Meeting on March 28, 2011, in New Orleans. Researchers also made several presentations at the Clemson Hydrogeology Symposium on April 7, 2011. Topics included removal of ammonia, selenium, and metals in wetland systems; produced water characterization, and hybrid wetland systems (http://www.clemson.edu/ces/hydro/symposium/index.html).

Vertical profiles of pilot wetland soil metal levels were created in order to determine the dominant biogeochemical processes responsible for metal removal in the systems. In the detritus zone, reducing conditions supported sulfate reduction resulting in removal of copper, nickel, cadmium, and zinc as sulfides. Metals can also bind to organic matter as complexes. Tests showed that there is little mobility of the metals once they are removed and little seasonal variation in sulfate reducing bacterial activity.

Rate coefficients, extents, and efficiencies for pilot-scale treatment were calculated and performance parameters were compared with treatment goals to assess the success of each pilot-scale wetland system. Ammonia was removed from post-RO produced water to levels low enough to meet irrigation and livestock watering criteria. Copper concentrations met or were below all of the beneficial use guidelines and criteria for 70 percent of the sampling periods. Cadmium concentrations met the beneficial use guidelines for two sampling periods, while 80 percent of the nickel concentrations were below the surface water discharge criteria and 30 percent met the irrigation guidelines. The zinc concentration met the livestock watering criteria in all cases, and irrigation guidelines and surface discharge criteria were met for all sampling periods except one. Concentrations of low molecular weight organics (LMWOs) were below all of the beneficial use guidelines and criteria for every sampling period. Adequate selenium removal was achieved only with the addition of the AquaSmartTM amendment. The concentrations of iron, manganese, nickel, and zinc decreased to below irrigation and livestock watering criteria in the wetlands, with greater removal rates in the subsurface flow series than in the free-water surface series.

Data from the pilot-scale wetlands are being used to design a demonstration wetland system.

The location for the demonstration-scale wetland system was changed from California to Alabama. Dr. Castle and two graduate students traveled to the field site and met with the Chevron field superintendent and reviewed plans for setting up the system. Local sources of plants and soil were examined.

Clemson University researchers and project partner Chevron constructed a demonstration wetland treatment system at a coal bed methane field in northern Alabama. Plants will acclimate for several weeks before the collected site water is passed through the system. Target contaminants include metals, metalloids, oil, and grease. Onsite water treatment has been designed for surface discharge so water hauling practices can be reduced.

The demonstration-scale wetland began to receive produced water on September 3, 2012. The residence time of the wetland was four days for each of the four parallel series, which can receive a total of 2600 gallons per week. Plants have been established for three months. Conductivity will be measured in the produced water holding tank and samples will be taken each time it is refilled. Outlet samples will be collected every two weeks with analyses being performed at Clemson. Sampling will continue for a minimum of six months.

Researchers monitored the performance of the constructed wetland treatment system through sampling and analytical methods for calculating removal efficiency, removal rate, and removal extent of the constituents of concern. Data from two sampling periods in September 2012 indicated removal of manganese, iron, and cadmium. The water quality will be analyzed to determine contaminant removal rates over time. Iron and manganese removal was greater in the two wetland treatment series amended with organic mulch than in the two series amended with a nutrient additive.

Inflow and outflow water samples were taken from the constructed wetland during seven sampling periods in September, October, and November 2012. Ammonia was successfully removed from the wetland during all seven sampling periods. Barium removal was over 90 percent of inflow concentrations until sulfate concentrations decreased in the wetland cells. Cadmium and iron were removed from the wetland to desired levels; manganese removal varied over time.

Reserchers presented posters illustrating project success at the 2012 Geological Society of America Meeting and the annual national meeting of the Society of Environmental Toxicologists and Chemists. The first poster depicted tracer tests showing that evapotranspiration significantly enhances transport of contaminants to the hydrosoil where they may be sequestered or treated through biogeochemical pathways and processes. The second poster showed selenium removal in the pilot-scale wetlands. The targeted removal pathway was dissimilatory selenium reduction, which precipitated selenium out of the water and into the sediments. An added energy source for the bacteria boosted selenium removal to below detection limits for target inflow concentrations.

Barium removal increased after the addition of gypsum to the wetland cells in January 2013 as a source of sulfate. From January to March 2013, ammonia removal rates averaged 80–90 percent, iron removal 70–90 percent, and manganese removal >95 percent. Wetland water sampling ended in April 2013.

The Chevron environmental group discussed keeping the demonstration wetland operating, but has not yet made a decision. If Chevron decides to build a wetland on-site, they will need a permit to discharge treated water to the streams. Researchers presented project results at the 2013 Clemson Hydrogeology Symposium.

The project ended on May 15, 2013. A final report is available below under "Additional Information".

$689,532

$341,441

NETL - Sandra McSurdy (sandra.mcsurdy@netl.doe.gov or 412-386-4533)
Clemson University - James Castle (jcastle@clemson.edu or 864-656-5015)

Final Project Report [PDF-5.61MB]

Demonstration Wetland System Installed for Produced Water Treatment [PDF-99KB] - News Release

Technology Status Assessment [PDF-34KB]

Kick-off Presentation [PDF-1.06MB]