for the 5- and 10-day surface water test (39.6 to 46.7 |g/L), it did drop in the groundwater test (11.6 |g/L). Overall, results observed in this study are consistent with results reported by McBeth et al. . A detailed discussion regarding geochemical processes controlling trace elements in semiarid environments is presented later in this section.
In summary, trace element analyses of column studies suggest that sediments of Dead Horse Creek drainage will alter the quality of produced water and precipitation surface- and groundwater runoff. Though produced water discharges can slightly alter the water quality of runoff, the sediment appears to be a stronger contributor to the overall water quality of stream channel runoff. Precipitation events will likely alter water quality of stream runoff by dissolving and leaching chemicals and thus increasing the chemical concentrations of downstream flows. Trace element availability seems to be predominantly affected by sediment interaction, with Mn and Ba decreasing and As and Se increasing from surface water retention to groundwater retention. Iron and F appear to be less affected by this difference in interaction.
Plants did not alter the sediment trace element water chemistry but did increase the rate of infiltration. Release of produced water may increase the pH of stream channel runoff and the mobility of anions, while decreasing the solubility of cations in soils. Overall results suggest that most constituents in produced and distilled water runoff were below the established criteria required for human consumption, except for Mn and Fe. Further research to determine the potential balance of produced water discharge and precipitation events (flushes) on downstream water quality will be useful in determining potential beneficial uses for produced water in the Powder River Basin.
This section reviews geochemical processes controlling trace elements in produced water. McBeth et al.  examined the chemistry of trace elements in produced water at discharge points and in associated holding ponds across the PRB, Wyoming. Produced water samples from discharge points and associated holding ponds were collected from three different watersheds within the PRB (CHR, BFR, and LPR) and were analyzed for pH, Al, As, B, Ba, Cr, Cu, F, Fe, Mn, Mo, Se, and Zn. The chemistry of trace element concentrations was modeled with the MINTEQA2 geochemical equilibrium model.
Results of this study show that pH of produced water for three watersheds increased in holding ponds. For example, the pH of produced water increased from 7.21 to 8.26 for LPR watershed. Among three watersheds, produced water from CHR watershed exhibited relatively less change in trace element concentrations in holding ponds. Concentration of dissolved Al, Fe, As, Se, and F in produced water increased in BFR watershed holding ponds. For example, concentration of dissolved Fe increased from 113 to 135 |g/L. Boron, Cu, and Zn concentrations of produced water did not change in BFR watershed holding ponds.
However, concentration of dissolved Ba, Mn, and Cr in produced water decreased in BFR watershed holding ponds. For instance, Ba and Cr concentrations decreased from 445 to 386 | g/L and from 43.6 to 25.1 |g/L, respectively. In the LPR watershed, Al, Fe, As, Se, and F concentrations of produced water increased substantially in holding ponds. For example, Fe concentration increased from 192 to 312 |g/L. However, concentration of dissolved Ba, Mn, Cr, and Zn decreased in holding ponds.
Geochemical modeling calculations suggested that observed increase of Al and Fe concentrations in holding ponds was due to increase in concentration of Al(OH)4- and Fe(OH)4- species in water, which were responsible for pH increases . Decreases in Ba, Mn, Cr, and Zn concentrations were attributed to the increase in pH, resulting in precipitates of BaSO4 (barite), MnCO3 (rhodo-chrosite), Cr(OH)2 (chromium hydroxide), and ZnCO3 (smithsonite) in pond waters, respectively. Several studies suggested that solubility of As, Se, and F in alkaline environments is controlled by the adsorption and desorption processes and increases as pH increases. Because As, Se, and F are anionic in nature and most of the soil mineral phases contain negative surface charge, these elements become soluble and mobile in alkaline watershed soils. The results found in this study are consistent with those in previous studies [24-27]. For example, concentration of dissolved As, Se, and F concentrations increased substantially in LPR watershed pond waters, which also had the highest increase in the pH.
Patz et al.  examined chemistry of trace elements in produced water interacting with Burger Draw and Sue Draw stream channels in the PRB, Wyoming. Significant change occurred for Fe, Mn, and As in flow of Burger Draw. As expected, Fe precipitates at once after discharge from wells (887.90 to 738.15 |g/L) as it transforms from a reduced to an oxidized state in Burger Draw (108.00 |g/L) or Sue Draw (3006.15 |g/L to 92.60 |g/L). Further reduction in Fe concentration in downstream flow did not appear to occur. Manganese reacted like Fe and decrease in concentration was low after initial contact with the atmosphere and the channel system. Arsenic and Se appeared to increase with downstream flow, but As was significantly higher after reservoir storage in Sue Draw (2.58 |g/L above and 4.48 |g/L below storage). McBeth et al.  and Hulin  report similar results.
In general, Al and Fe become less soluble and mobile in natural waters of the western alkaline environments. However, produced water may contain higher concentrations of Al and Fe in disposal pond water and these increases may be explained by the presence of Al and Fe hydroxyl species (e.g., Fe(OH)4-) in waters with a high pH . Other studies also suggest that when anionic species of Al and Fe dominate in water, the solubility increases with increasing pH. Even though Al and Fe concentrations remained relatively unchanged, they may decrease with time in alkaline environments due to the likely precipitation of aluminum and iron hydroxides .
The decreases in Mn and Ba concentrations are likely due to the precipitation of oxides, hydroxides, and carbonates of these metals, common in the calcite-rich, alkaline environment of the PRB. Overall, the preceding review suggested that, in alkaline environments, cationic trace elements (e.g., Al, Fe, Mn, and Ba) become less soluble and mobile due to precipitation of oxides and hydroxides. However, anionic trace elements (e.g., As, Se, and F) become more soluble and mobile as a result of mineral dissolution and desorption processes.
8.6 POTENTIAL IMPACTS OF CBNG-PRODUCED WATER TO RANGELAND PLANTS, RIPARIAN PLANTS, SOILS, AND SEDIMENTS
In this final section, potential impacts of produced water to rangeland plants, soils, and sediment are summarized. Different management alternatives are proposed for produced water within the PRB, Wyoming. These options include surface land application (e.g., irrigation), aquaculture, and livestock watering. Nonetheless, any management option for produced water within the semiarid areas requires a careful evaluation of its salinity (e.g., EC, TDS), sodicity (e.g., SAR), and trace elements.
Common and observed issues associated with plant species and community impacts while developing CBNG in the PRB may be categorized as follows:
• Produced water and toxicity to native plants
• Native plant toxicity when produced water is applied to rangeland soils and to stored sediment in stream channel systems
• Change in plant species composition after produced water is released as perennial channel flow
• Vegetation response and ability to resist accelerated channel erosion after produced water is released as perennial channel flow
• Land application of produced water and change in soil chemistry and structure
• Change in land and stream channel characteristics as land application and channel release ends
• Role of plants in accumulation and turnover of trace elements
In the PRB, toxicity to plants was addressed in a preliminary laboratory study using produced water, nutrient solution, and perlite as a plant support system. Western wheatgrass, a dominate upland and riparian zone grass, was fed 100% produced water; 50% produced water and 50% nutrient solution water; and nutrient water alone. No plants died and no differences in total biomass were observed between 100% nutrient and 50% produced water and 50% nutrient solution; however, growth was reduced when no nutrients were supplied to 100% produced water (Figure 8.7). Plant nutrients in the tested produced water appear to limit growth response, and toxicity of the water did not appear to be an important issue .
Application of produced water to stored sediment of stream channel systems and plant toxicity was surveyed by Hulin  and Jackson  in a laboratory study using column techniques. Of three common wetland plants — Nebraska sedge (Carex nebraskensis Dewey); water sedge (Carex aquatilis Wahl.); and Baltic rush (Juncus balticus Wild) — grown in columns, only Baltic rush survived. These data suggest that when a produced water like that used for their experiments is applied to stored channel sediment, plant toxicity is an issue and that some native plant species cannot tolerate produced water and sediment chemistry under saturated conditions (Figure 8.8).
Change in plant species composition after produced water was released as perennial channel flow was observed by Patz . Approximately 3000 containerized plants of each of the three
wetland species used in Hulin and Jackson's study were transplanted in and along three study sites of a single stream discharging produced water. Only Baltic rush survived and was confined to a narrow zone next to the water edge and bank. Rangeland plant species that had occupied the ephemeral channel area before release of produced water were being replaced by the riparian grass or grasslike species: inland salt grass (Distichlis spicata L.); foxtail barley (Hordium jubatum L.); Nutall's alkali grass (Puccinellia nuttalliana (Schult.) Hitchc.); Western wheatgrass (Pascopyron smithii Rydb.); American bullrush (Scirpus americanus Pers.); and maritime bullrush (Scirpus maritumus L. var. paludosus (A. nels.) Kukenth). These replacement species are known to be salt tolerant, whereas the grasses being replaced are not. This preliminary study suggests that upland rangeland grasses may be replaced by riparian zone and more salt-tolerant species when produced water is released as perennial flow, stored in ponds, or detained behind water spreaders or debris dams for extended lengths of time (Figure 8.9).
The change in plant composition observed by Patz  in addition to follow-up photo monitoring confirms that the vegetation response to perennial release of channel flow creates a stable riparian zone capable of resisting accelerated channel erosion. Plant height, cover, and stem density appear equivalent to riparian zones found throughout the PRB . Three extreme flood events caused by summer convective storms have not caused low flow channel conditions to change. Figure 8.10 illustrates stability of riparian vegetation near Patz's 2002 study sites.
Land application of produced water and change in plant and soil characteristics in the PRB are yet to be confirmed within the scientific literature even though potential impacts of salinity and sodicity are well established. Figure 8.11 illustrates limited observations of soil conditions in which produced water is applied through sprinkler irrigation. Application of saline and sodic water for irrigation is not a problem in humid areas compared to arid and semiarid areas because rainfall is a major source of salt-free water. However, use of salty and sodic water for irrigation in arid and semiarid areas, particularly containing clay minerals with poor drainage, accumulates salts, decreases infiltration, and increases runoff and erosion . A recent study reported that water with SAR as low as 5 will have an adverse effect on the soil structure and infiltration rates .
Surface disposal of produced water in ponds and in stream channels is creating riparian zone and wetland habitat in addition to that which is natural and has been developed through irrigation using historic water supplies. The additional produced water supply is limited to the time that it takes to remove the coalbed gas supply. Once this supply is harvested, it must be assumed that
riparian zone and forage plant resources created because of surface disposal of produced water will suffer drought conditions. The question of if and how vegetation will respond to no produced water has not been answered. It can be assumed that sodicity and salinity of soil and sediment resources should have increased through the influence of CBNG surface water supplies. However, fate and transport of trace elements by plants remain unknown. The questions associated with assimilation, accumulation, and transformation of trace elements by plants that can and cannot tolerate produced water should be answered knowing that reclamation of riparian zones, wetlands, and native range soils will be needed in the future.
Release of produced water into established watershed can have an impact on the water quality of the watershed, but this impact may be similar to that of naturally occurring precipitation events. However, precipitation is sporadic in causing runoff, and release of produced water may be perennial for some given period. From the authors' data, continual release of produced water would be expected to alter the chemistry of channel flow to assume eventually the chemical composition of the produced water discharge and naturally occurring precipitation combined. These data would further suggest that, following completion of CBNG discharge, naturally occurring precipitation would likely bring the concentrations of trace elements deposited back towards original levels.
The travel time of stream flow from the head water tributaries of Dead Horse Creek to the Powder River is not known. However, the authors' data suggest that containment behind debris dams, water spreader dikes, and even in ponds of low gradient stream channel reaches may produce local and high contributions of chemistry associated with trace elements to flow moving towards the Powder River. It did not matter if the stored water was CBNG-produced water or distilled water. When the water types were stored as ground water for approximately 2 months, both types of sediment water assumed about the same water chemistry. Although the pond water types were lower in chemistry after 5 and 10 days, they approached the water quality of the ground water supplies. Therefore, the issue of travel time to the Powder River as channel flow vs. detention storage is an important consideration for developing regulatory and management options for disposing of or using this new and potential CBNG-produced water supply.
It is important to note that stream flow of water from precipitation represents a baseline for evaluating potential water quality issues associated with managing CBNG-produced water. If CBNG water is released into Powder River tributaries, then water from precipitation represents a blending of two different supplies. The data suggest that rain water may contribute substantial amounts of chemistry to the Powder River system and this load should be considered as it varies in time and space. Then, blending and water treatment alternatives for adding produced water sources that range in quality and lie within the time and space distribution of precipitation runoff may be feasible for helping achieve acceptable water quality in a downstream direction.
Was this article helpful?