In This Issue
Summer Issue of The Bridge on Shale Gas: Promises and Challenges
June 15, 2014 Volume 44 Issue 2

Produced Water from Hydrofracturing Challenges and Opportunities for Reuse and Recovery

Thursday, June 26, 2014

Author: James M. Silva, Rachel M. Gettings, William L. Kostedt, and Vicki H. Watkins

Technological advances in exploration and production have enabled the economical production of natural gas from shale, which has led to dramatic increases in both shale gas production and reserves. A key challenge associated with this production is the management of the resulting produced water, which is considered a waste product. In Pennsylvania, site of the Marcellus shale, the near absence of salt water disposal facilities has led to the recycling and reuse of about 90 percent of the state’s produced water as blendstock for hydrofracturing subsequent wells. However, as well fields mature, the supply of produced water is expected to exceed demand.

Thermal water and salt recovery present opportunities to extract value from this produced water. For a “design case” produced water (based on a Marcellus produced water survey), about 55 percent water recovery is possible with evaporation alone, which requires only a simple pretreatment process (the design case is detailed in Silva et al. 2012). High (e.g., 90+ percent) water recovery methods generate a solid NaCl product. Marcellus produced water has relatively high barium and radium content and thus requires further pretreatment to remove these species before NaCl crystallization.

Based on a Marcellus produced water composition survey and a simple material balance, some Marcellus pretreatment waste sludges require disposal as naturally occurring radioactive material (NORM) waste. Economical and environmentally sound options are available, however, for high water and salt recovery from many of these challenging waters.

Introduction: Two Shale Gas Plays

We consider two shale gas plays: the Marcellus (Pennsylvania portion) and the Barnett (Texas). The full Marcellus, underlying five states and covering about 95,000 square miles, has the potential to become the world’s second largest gas field (Considine et al. 2010). The Barnett, which underlies 14 counties in the Dallas-Fort Worth area, covers about 5,000 square miles. Both plays are 5,000–8,000 feet below the earth’s surface, with a shale layer thickness of 100–500 feet.1

Table 1

As shown in Table 1, there are many similarities in water management between the Pennsylvania (PA) Marcellus and the Barnett. For example, both plays use comparable amounts of water for drilling and hydraulic fracturing. Although the Marcellus saw more drilling activity in 2012, both plays had roughly the same number of well completions during this period. (Since the Barnett started development almost 20 years before the PA Marcellus, it has a higher number of active horizontal wells.) Both plays have comparable rates of wastewater generation, and in 2012 only about 1 percent of wastewater was recovered (as distilled water) in each.

There are also sharp contrasts between the two plays. Fresh water is abundant in the Marcellus but scarce in the Barnett. The PA Marcellus has only eight Class II injection wells (also referred to as underground injection control, or UIC, wells) for salt water disposal, with a total disposal capacity of about 0.3 million gallons per day (MGD). In contrast, the Barnett has a large number of high-capacity wells (1 MGD typical), owing to the fact that it overlies the Ellenburger formation, which has porous rock containing naturally occurring salt water. In 2012, 87 percent of the PA Marcellus wastewater was reused in subsequent well completions; 12 percent was deep-well injected. The opposite was true for the Barnett—94 percent was deep-well injected and 5 percent reused.

Water recycling and reuse has been adopted as a best practice by producers in the PA Marcellus because it is considerably less expensive and much more environmentally sound than hauling large amounts of produced water from eastern Pennsylvania to Ohio, where the nearest site for deep-well injection is located. Further, field reports show that using a blend of produced and fresh water does not negatively impact either the formation or gas production (Minnich 2011). In fact, one report shows that wells completed with a blend of fresh and recycled produced water are among the company’s top gas producers (Gaudlip 2010).

The focus of this paper is produced water recovery in the PA Marcellus, because of both the challenging composition of the water and the near absence of in-state UIC disposal capacity.

Management of Produced Water

Completion of a well typically requires approximately 5 million gallons of water, whether from fresh (e.g., municipal or surface water) or brackish water sources. Depending on the shale play, the source water may also include treated flowback and produced water from earlier well completions.

Chemicals—primarily friction reducer, biocide, and acid—are added to the water, which is pumped under pressure into the well. Friction reducer lowers pressure losses in the tubing that carries the water to the shale layer, biocide kills bacteria that can corrode the well tubing, and acid clears cement debris and carbonate minerals from the wellbore. When the shale rock starts to break, graded sand (proppant) is added to the water being injected to keep the cracks and fissures open so that the gas trapped in the rock can flow into the well.

Once the well has been hydraulically fractured, some of the water returns as flowback. For the Marcellus, about 20 percent of the source water returns as flowback within about 30 days, at rates of 300–8,000 barrels (bbl) per day during the first two or three weeks (King 2012). The flow then slows considerably, to about 5–10 bbl/day of produced water, a rate that continues for the lifetime of the well.

This flowback/produced water (also called wastewater or brine) can be disposed of by one of the following methods. It can be deep-well injected into Class II disposal wells, recovered as clean water and solid sodium chloride salt, or treated and reused as blendstock for the next well completion. The latter, which we discuss in the following section, is by far the least expensive alternative.

Treatment and Reuse

The feasibility of produced water reuse hinges on the ability to create a blend of produced and fresh or brackish water that enables effective and economical well completion. Therefore, the composition of the produced water, the fresh/brackish water, and the target blend must be known.

Table 2

Table 2 shows representative compositions for each of these waters. The design case produced water is based on a PA Marcellus survey that indicates the presence of significant hardness, as well as barium (Ba) and radium-226 (226Ra), a naturally occurring radioactive material (Silva et al. 2011). Although NORM is not a radiation hazard at the levels present in produced water, it must be considered in water treatment and recovery processes (discussed below). The produced water also includes suspended solids, organics, and bacteria. Fresh water is represented by a typical river water composition (Snoeyink and Jenkins 1980, p. 6), and the target blend composition is based on a case study reported by Minnich (2011).

Produced and fresh water are both treated for reuse with well-known techniques to remove iron, suspended solids, hardness, and bacteria; this treatment typically involves chemicals (e.g., lime/soda softening, sulfate treatment) and filtration. The treatment of one or both waters is necessary (1) to avoid BaSO4 precipitation, because produced water contains barium and fresh water contains sulfates (see Table 2), and (2) to ensure that the blend composition supports proper functioning of the hydraulic fracturing chemistry (e.g., friction reducers); the blend hardness is typically kept below about 1,000 mg/L, although for one case study it exceeded 2,500 mg/L (Minnich 2011).

Various methods are used to treat produced water for reuse. For example, one water treatment service provider removes iron, suspended solids, strontium, and barium, then returns the treated water to the customer for reuse. Another uses gravity settling with optional filtration to remove suspended solids before blending the produced and fresh water. Because water chemistry varies significantly among sources, treatment for reuse must be specified on a case-by-case basis.

Supply and Demand

The ability to reuse produced water depends on its supply and demand. The supply of shale gas–produced water in a given geographical area increases with both the active well count and the rate of well completions in that area. In the PA Marcellus, each active horizontal (also called unconventional) well yields 5–10 bbl/day of produced water, and each well completion yields about 1 million gallons within a month of the hydraulic fracturing activity. Figure 1 shows the estimated number of active horizontal wells and annual well completions in the PA Marcellus, based on the Pennsylvania Department of Environmental Protection (PADEP) unconventional gas wastewater database.2


Figure 1

Two key factors affect the demand for produced water as blendstock for well completion. First, well completion activity follows the price of natural gas. In 2012, because of low natural gas prices, only about 540 horizontal gas wells were completed in the PA Marcellus, although 1,365 such wells were drilled. (Noncompleted wells are capped, ready to be completed and put into production when the market conditions are favorable.) Second, weather and ground conditions affect the rate of completions.

Disposal of Excess

When the supply of produced water exceeds the demand, the excess requires disposal. In Pennsylvania some of this wastewater was sent to publicly owned treatment works (POTWs) from 2007 through mid-2011, when, on behalf of the governor, the PADEP secretary asked that all exploration and production companies and drillers in the state voluntarily cease this practice. This request yielded 100 percent compliance.

Since 2011 produced water disposal in Pennsylvania has been solely by injection into Class II salt water disposal wells instead of POTWs (Rassenfoss 2011). However, the cost of transportation from Williamsport, PA, an area of intensive shale gas development, to Youngstown, OH—about 190 miles away—is about $12/barrel. In addition, there is controversy over a possible connection between deep-well injection and small earthquakes (Frohlich 2012). In light of these concerns, deep-well injection is viewed as less than ideal for produced water disposal in Pennsylvania. Another option, which has not yet been implemented on a commercial scale, is water and salt recovery.

Opportunities for Water and Salt Recovery

Excess produced water creates opportunities for water and salt recovery. Produced water with a high level of total dissolved solids (TDS),3 such as the design case, requires the use of thermal methods, such as humidification-dehumidification, forward osmosis membrane distillation,4 and mechanical vapor recompression.

One option is to recover only clean water. For the design case produced water composition, about 55 percent of the feed volume can be recovered (limited by the onset of NaCl precipitation); the remaining water leaves as a concentrate stream. Produced water is first pretreated to remove iron, manganese, suspended solids, and some of the magnesium to protect the thermal equipment from scaling. Disposal of the concentrate may be by deep-well injection, sale as well kill fluid, or dispatch to a crystallizer for further recovery.

Figure 2

Alternatively, both clean water and a solid salt product may be recovered from the produced water. For the design case produced water composition, recovery of 90+ percent water and 90+ percent NaCl may be achieved through NaCl crystallization. Figure 2 shows one scheme for water and salt recovery. First, after appropriate pretreatment, mobile evaporators generate distilled water and concentrate. Next, in a central facility the concentrate is further treated to remove barium and radium to meet salt product specifications. Of seven methods considered for Ba and 226Ra removal from produced water, sulfate coprecipitation was found to be the most cost effective (Silva et al. 2012). Finally, the concentrate is evaporated in a crystallizer to recover solid NaCl and distilled water. A small purge stream is disposed of by deep-well injection or sold as well kill fluid.

Disposal of Pretreatment Sludge

For a given produced water composition, a key economic and environmental concern is the activity concentration of 226Ra in the sulfate pretreatment sludge vis-à-vis the permissible level for disposal in nonhazardous (RCRA-D) landfills (regulated under subtitle D of the Resource Conservation and Recovery Act). As there is no federal standard for these landfills, this regulatory limit varies by state (Veil and Smith 1999). In Pennsylvania, based on guidance from PADEP,5 a limited amount of sludge with up to about 140 pCi/gm radium is acceptable for disposal in RCRA-D landfills, as shown in Table 3. There are 50 RCRA-D disposal facilities in Pennsylvania, but no NORM disposal facilities6; other states, such as Texas, have facilities for NORM disposal. Table 3 shows the relative cost of nonhazardous waste disposal and NORM waste disposal.

Figure 3

The 226Ra activity concentration in the sulfate sludge can be estimated by a material balance. Using 1.1 mole sulfate per mole barium, it is assumed that all Ba and 226Ra in the raw produced water feed coprecipitate as (Ba,Ra)SO4(s), that all excess sulfate precipitates strontium as SrSO4, and that the sludge is 30 wt percent solids. The results of this material balance are shown in Figure 3, where raw produced water Ba concentration is plotted on the x-axis and 226Ra activity concentration on the y-axis. The lines that radiate from the origin are specific 226Ra activity concentrations in the wet (Ba,Ra)SO4 sludge resulting from combinations of Ba and 226Ra concentrations in raw produced water. For example, the design case produced water yields sludge with just under 140 pCi 226Ra per gram of wet sludge.

Sixteen PA Marcellus produced water sample compositions are superimposed on the plot in Figure 3. Several of these produced waters, particularly those with low Ba and high 226Ra levels, would generate sulfate sludges that exceed the maximum acceptable 226Ra activity for nonhazardous waste disposal in Pennsylvania. Others yield sulfate sludges that may be safely disposed of as nonhazardous waste.

This calculation shows that produced water NORM and Ba concentrations are key factors in the economic feasibility of water and salt recovery.

Conclusions

About 90 percent of produced water from the Pennsylvania Marcellus was reused in 2012, demonstrating that reuse is a valid option for management of shale gas–produced water. Often, only very simple treatment is needed for produced water reuse. However, as shale gas development in a given geographical region matures, the supply of produced water will exceed the demand for blendstock use in subsequent well completions. Because of the high TDS levels in Marcellus flowback and produced waters, thermal methods are required for water and salt recovery. Although implementation of water and salt recovery is still in its infancy, solutions exist today that are both technically and economically feasible.

Acknowledgments

Funding for this project is provided by the Research Partnership to Secure Energy for America (RPSEA; www.rpsea.org) through the Ultra-Deepwater and Unconventional Natural Gas and Other Petroleum Resources program authorized by the US Energy Policy Act of 2005. RPSEA is a nonprofit corporation whose mission is to provide a stewardship role in ensuring the focused research, development, and deployment of safe and environmentally responsible technology that can effectively deliver hydrocarbons from domestic resources to the citizens of the United States. Operating as a consortium of US energy research universities, industry, and independent research organizations, RPSEA manages the program under a contract with the US Department of Energy’s National Energy Technology Laboratory.

The authors acknowledge technical discussions with Joseph Tinto of GE Power and Water.

References

Considine TJ, Watson R, Blumsack S. 2010. The economic impacts of the Pennsylvania Marcellus Shale natural gas play: An update (May 24). Pennsylvania State University, Department of Energy and Mineral Engineering. Available at http://marcelluscoalition.org/wp-content/uploads/2010/ 05/PA-Marcellus-Updated-Economic-Impacts-5.24.10.3.pdf.

Frohlich C. 2012. Two-year survey comparing earthquake activity and injection-well locations in the Barnett Shale, Texas. Proceedings of the National Academy of Sciences 109(35):13934–13938.

Gaudlip T. 2010. Preliminary assessment of Marcellus water reuse. Presented at the Process-Affected Water Management Strategies Conference, March 17, Calgary, Alberta, Canada.

King G. 2012. Hydraulic fracturing 101: What every representative, environmentalist, regulator, reporter, investor, university researcher, neighbor, and engineer should know about estimating frac risk and improving frac performance in unconventional gas and oil wells. Society of Petroleum Engineers Paper SPE 152596.

Mantell M. 2011. Produced water reuse and recycling challenges and opportunities across major shale plays. Presented at EPA Hydraulic Fracturing Study Technical Workshop #4, March 29–30.

McCurdy R. 2011. Underground injection wells for produced water disposal. Proceedings of the Technical Workshops for the Hydraulic Fracturing Study: Water Resources Management, May. Available at www2.epa.gov/sites/production/files/documents/HF_Workshop_ 4_Proceedings_FINAL_508.pdf.

Minnich K. 2011. A water chemistry perspective on flowback reuse with several case studies. In: Proceedings of the Technical Workshops for the Hydraulic Fracturing Study: Water Resources Management. Washington: US Environmental Protection Agency.

Nicot JP, Reedy RC, Costley RA, Huang Y. 2012. Oil and Gas Water Use in Texas: Update to the 2011 Mining Water Use Report, Table 7. Prepared for the Texas Oil and Gas Association, Austin.

Rassenfoss S. 2011. From flowback to fracturing: Water recycling grows in the Marcellus Shale. Journal of Petroleum Technology 63:48–51.

Silva JM, Matis H, Kostedt W, Watkins V. 2011. NORM removal from hydrofracturing water. International Water Conference, paper IWC-11-07, Orlando, FL.

Silva JM, Matis H, Kostedt W, Watkins V. 2012. Shale gas produced water pretreatment for barium and radium removal. International Water Conference, paper IWC-12-56, San Antonio, TX.

Snoeyink V, Jenkins D. 1980. Water Chemistry. Hoboken, NJ: John Wiley & Sons.

Veil J, Smith K. 1999. NORM disposal options, costs vary. Oil & Gas Journal 97:37–43.

FOOTNOTES

1 Specific data are currently available from the company involved in development of these shale plays; for Barnett, www.halliburton.com/public/solutions/contents/Shale/related_ docs/Barnett.pdf; for the Marcellus, www.halliburton.com/en-US/ps/solutions/unconventional- resources/shale-gas-oil/shale-plays/marcellus-shale.page? node-id=hgjyd46z.

2 The database is available at www.paoilandgasreporting.state.pa.us/publicreports/Modules/ DataExports/DataExports.aspx.

3 For low-TDS produced waters (<35,000 mg/L TDS), reverse osmosis can be used for water recovery.

4 Forward osmosis moves water across a membrane into concentrated draw solution, thus diluting the solution, from which water is then thermally recovered to reconcentrate the solution.

5 The guidance from PADEP is given in terms of radiation dose rather than radium activity concentration. The inferred activity concentration limit is based on a conservative estimate that the radiation dose for a typical sludge rolloff container (in µRem/hr) is approximately equal to the average radium activity concentration for the contents of the container (in pCi/gm). A more precise correlation between dose and activity concentration should be applied if site-specific or updated regulatory guidance from PADEP is provided.

6 PADEP is conducting a study on management of technologically enhanced NORM (TENORM) for produced water and treatment residuals (sludge). Information about the study is available at http://files.dep.state.pa.us/OilGas/BOGM/BOGMPortalFiles/ RadiationProtection/TENORM-Study_SoW_04_03_2013_FINAL. pdf.

About the Author:James M. Silva is a senior chemical engineer, Rachel M. Gettings a lead chemical engineer, William L. Kostedt a lead environmental engineer, and Vicki H. Watkins a senior chemist, all at the General Electric Global Research Center in Niskayuna, New York.