3

Regional Similarities and Differences: Environmental and Regulatory Context for Potential Use of Flowback and Produced Water

The first panel of the workshop provided an overview of the environmental and regulatory context for potential use of flowback and produced water in a regional context. Three 10-minute presentations by three panelists (Danny Reible, Texas Tech University; Radisav Vidic, University of Pittsburgh; and James Silva, GE Global Research [retired]) were followed by a panel discussion moderated by William Stringfellow (University of the Pacific and Lawrence Berkeley National Laboratory).

In his opening remarks to begin the panel, Stringfellow shared some information from California, where hydraulic fracturing is applied in already developed oil and gas areas and, as a result, “flowback” water does not exist. California thus approaches produced water as a blended material, representing some naturally occurring and anthropogenic materials. Because of the existing infrastructure, water from oil and gas wells is typically put directly into the produced water system after hydraulic fracturing. He commented that relative to the 10:1 water-to-oil ratio previously mentioned, the ratio in California is about 15:1 (barrels of water to oil) and this water needs managing. And a great deal of interest exists in using produced water because of recent droughts in the state. At a state scale, the amount of produced water may not be a lot in a relative sense of overall use, but volumes may be locally important. Another reason for interest in produced water in California is the water’s salt content. Produced water in California currently does not have a lot of salt in it; desalinization is not a big issue. If the oil and other materials are removed, produced water may be (and is already being) used for agriculture. Other uses include flooding and steaming, pressure mitigation, subsidence mitigation, cooling, and steam generation. However, he said, the rules of operation are changing in California and the current approach is probably not going to continue. For example, disposal practices such as shallow reinjection and percolation ponds to infiltrate water back to the groundwater are being examined as an opportunity to use produced water.

PANEL PRESENTATIONS

Use of Flowback and Produced Waters in Permian Basin, Eagle Ford, and Barnett Shale

Danny Reible, Texas Tech University

Reible emphasized the importance of regional differences in considering water quality, quantity, and potential use of produced waters. For example, in Texas, produced water does not have similar quality advantages that exist in California. Thus, recognizing differences and recognizing where opportunities exist is extremely important. He suggested an approach that considers water fit for a specific use that can thereby target water treatment; such an approach requires us to adapt to the water, rather than trying to get the water to adapt to us. Reible’s remarks then reviewed the characteristics of flowback and produced water in Texas; the regulation of produced water; the importance of regional context to identify opportunity, logistical, and economic considerations; and the nature of fracturing fluids.

Reible noted a large volume of produced water generated relative to oil production in Texas plays; however, the quality of the produced water is typically bad. He mentioned total dissolved solids (TDS) concentrations of 100,000 milligrams per liter. This high level of salinity excludes a lot of uses and raises some barriers for using produced water. He suggested that the primary viable option for that high-salinity water is direct use in hydraulic fracturing or in the development of oil and gas fields to enhance recovery. He also emphasized that the volume of water produced is actually relatively small, adding a further barrier to a range of potential uses because of a set of alternative, low-cost water resources that may also be available. For example, quite a lot of brackish water (from perhaps 1,000 to 10,000 milligrams TDS per liter) is available in Texas near the surface; however, the amount that can be withdrawn compared to the total volume is small, at most a few percent. In the Southwest, approximately 10 Great Lake volumes of brackish water have been estimated as being near the surface and potentially available for use, he said. These volumes could serve as alternative source waters for small communities following desalination, as makeup water for agriculture, and as waters used directly for hydraulic fracturing.

In Texas one of the biggest differences, especially compared to the Marcellus shale play in the northeastern United States, is that there are 12,000 saltwater disposal wells providing a relatively low-cost water disposal option. If the disposal wells are owned by the operator, the disposal costs can be as little as $0.10 per barrel of water. This cost may increase if the water has to be transported by truck to dispose of it in a commercial well.

Until recently in Texas, Reible noted, it was quite difficult to transfer produced water to other users due to water rights and ownership issues. Freshwater is the property of the surface land owner and selling water is a potential source of income to the land owner. As a result, any transfer of that produced water to another nearby operator had been quite difficult although recent efforts have worked to try to alleviate this issue.

Reible also noted incentives to use produced water given water’s regional importance in central and west Texas. The worst single-year drought in Texas history was 2011. The drought occurred during a period of growth in the oil and gas industry and hydraulic fracturing. The image of tens of thousands of trucks transporting freshwater per year and pumping it down a hole was not a good image for the oil and gas industry when farmers did not have enough water and cities were within a few months of running out of drinking water, he said. A good example of using produced water was the Apache Corporation experience (previously discussed; see Chapter 2, Moderated Discussion), where produced water was used in a large interconnected wellfield. Thus, the need for water for hydraulic fracturing was balanced by the adjacent wells where produced water was coming to the surface. Reible also said that transportation costs (putting water into trucks or pipelines and trans-

porting water off site) dramatically increase the cost of managing produced water. In the Apache Corporation example, the use of produced water onsite overcame many of these types of logistical and economic constraints.

Reible remarked further that the management and use of produced water is not governed by technology but by logistical and economic concerns. Large interconnected wellfields present an opportunity for using produced water. The use of water onsite avoids transportation costs and associated issues, such as impacts on the local community road infrastructure. As an example, the lifetime of many rural roads in west Texas went from a design life of 50 years to less than 10 years when truck activity (associated with oil and gas) to transport produced water increased.

There is a trend in industry throughout the United States and Texas to make fracturing fluid compositions simpler. For example, a recent increase in the amount of slickwater¹ fracturing makes it easier for operators to use very high TDS waters in hydraulic fracturing activities. Reible emphasized that although saltwater disposal wells in Texas will continue to be used for disposal, a feasible goal is to maximize direct use of produced water and that many opportunities exist to do that. For every barrel of produced water that is applied to subsequent uses within an oil field, one less barrel of water has to be transported for disposal, he said.

Radisav Vidic

University of Pittsburgh

Vidic noted that the potential use of flowback and produced water is a difficult topic to address because of many regulatory and environmental concerns that are constantly changing. He indicated that the American Petroleum Institute (API) recommends operators consider options for recycling fracture treatment flowback fluid. The API also discusses management and disposal options for such waters which include land spreading, road spreading, onsite burial, onsite pits, annular injection, underground injection wells, regulated and permitted discharge of this fluid, and offsite commercial facilities. Underground injection wells for water from oil and gas activities are Class II-D² wells. Such disposal wells are by far the cheapest option to dispose of produced water and are also the least troublesome; the combination of logistical and economic practicalities are drivers for a lot of decisions being made with respect to produced water, Vidic said. He began his discussion by describing some of the regulatory framework that governs disposal of produced water in various states and followed that with the example of produced water management in Pennsylvania.

All states, he said, except for North Carolina, permit injection and Class II-D disposal wells. States like Montana and Ohio actually require that any water above 15,000 milligrams TDS per liter has to be injected into a disposal well unless a regulatory hearing determines otherwise. A lot of states are cognizant of produced water disposal wells and the associated potential to generate earthquakes. In some instances, these concerns have led to limiting the use of disposal wells that lie within a certain distance from known faults. Induced seismicity³ is becoming much more prominent in the public context with the result that greater pressure exists to reduce the use of disposal wells and provide alternative solutions to managing produced water.

Pennsylvania has a unique regulatory environment, Vidic said, as there is no ability for the state to permit Class II wells. This regulatory capacity lies with the federal government, and the Envi-

___________________

¹ Slickwater is a type of fracturing fluid that contains a friction reducer (HexionFracline, 2012). Available at http://www.hexionfracline.com/fracturing-fluids-101 (accessed August 30, 2016).

² Class II wells are used to inject fluids associated with oil and natural gas production. These fluids are primarily brines that are brought to the surface while producing oil and gas. SOURCE: EPA, 2016a. Underground Injection Control (UIC). Available at https://www.epa.gov/uic/class-ii-oil-and-gas-related-injection-wells (accessed August 17, 2016).

³ “Earthquakes attributable to human activities are called induced seismic events or induced earthquakes” (NRC, 2013, p. 1).

ronmental Protection Agency is the ultimate permitting agency for disposal wells in Pennsylvania. The state has 200,000 oil and gas wells, many of which are legacy conventional oil and gas wells, with unknown locations, he noted. This number of legacy wells creates complications for disposal wells. These considerations compromise the use of Class II disposal wells as a primary management solution for Pennsylvania, like they are for Texas, and for this reason, Pennsylvania developed an approach to use produced water. Vidic noted the efforts of the industry sector in Pennsylvania in trying to identify alternative approaches to disposal of produced water in deep injection wells. He emphasized that use of produced water in hydraulic fracturing jobs actually works better than use of tap water or river water where potassium chloride must be added to the water to prevent subsurface swelling in the rock formation. He mentioned in Pennsylvania that about 90 percent of the produced water is used again for subsequent hydraulic fracturing activity. This proportion makes up a small fraction of the amount of water needed for the next hydraulic fracturing job. Vidic did provide a caveat by indicating that using produced water in this way only works if another well is available for hydraulic fracturing. Once hydraulic fracturing of wells ceases, the water cannot be used further for that purpose. He noted that this potential situation is not an immediate concern in Pennsylvania because approximately 10,000 wells have been hydraulically fractured to date with the future total estimated to be much higher if production of oil and gas from shale and other tight formations continues.

Pennsylvania regulates produced water from oil and gas operations in a two-part system. Conventional wells and the conventional oil and gas industry have been grandfathered and permitted National Pollutant Discharge Elimination System (NPDES) facilities exist to dispose of produced water from these activities into streams. This kind of practice has been done for decades and was not a major issue when small amounts of produced water were being generated in Pennsylvania from conventional fields. Such produced waters are permitted for discharge at 500 milligrams TDS per liter. Water with that quality has value, he said, and for that reason, produced water from conventional oil and gas wells is permitted differently than produced water from unconventional wells. Although their salinities are similar, some differences exist, for example, in ratios of strontium isotopes, Vidic said.

Various potential uses for treated produced water in Pennsylvania each offer different kinds of challenges, Vidic said. Use of treated produced water for agriculture is not an option in Pennsylvania because the state does not have irrigated agriculture. Other uses for produced water are limited to industrial and livestock use. Produced water use for hydraulic fracturing is also not straightforward because of the complications in terms of fracture fluid design. For example in Texas, gel-based fracture fluids are used, while in Pennsylvania slickwater fracturing is generally used. Texas is moving toward slickwater fracturing.

Recovery of valuable products from produced water is an attractive option, Vidic indicated, but that process is incredibly complex because one is trying to extract milligrams or micrograms per liter of compounds from produced water that has hundreds of thousands of milligrams per liter of sodium chloride. From an engineering standpoint, this kind of extraction is difficult to do. Even if sodium chloride is extracted as a byproduct, Pennsylvania could be making 8 to 10 million tons of sodium chloride. For context, the entire United States uses 15 million tons of sodium chloride for deicing roads to make roads safe during cold periods; thus, a market for excess sodium chloride extracted through treatment of produced water may not exist.

Another problem in Pennsylvania is the naturally occurring radioactive material (NORM) already in the subsurface rock formations; because of the existence of NORM, the state has produced water containing radiogenic nuclides produced from radioactive decay of these materials. Thus, any effort to recover, treat, or use produced water also has to address NORM. At the moment, Pennsylvania allows NORM to go into landfills. Once Pennsylvania reaches about 30,000 unconventional wells, however, the quantity of NORM will exceed what Pennsylvania landfills can handle.

James Silva

GE Global Research (retired)

Silva noted that one-quarter of the nation’s gas is from unconventional sources. Unconventional sources were not even considered 10 or even 20 years ago, he said. Some of the big differences among the different shale plays revolve around water. Source water for hydraulic fracturing to develop shale plays like the Marcellus in the northeastern United States is not an immediate problem because of the relative abundance of natural surface water and precipitation; the bigger question in states like Pennsylvania is rather what to do with the produced water. In contrast, the Barnett play in Texas has a source water problem for hydraulic fracturing due to the relative scarcity of surface water and competing needs for groundwater resources. However, disposal of produced water is relatively straightforward due to the existence of many underground injection wells. In the remainder of his remarks, Silva focused on experiences he had gained through his work with GE in examining the issue of produced water in Pennsylvania.

In the past 8 years, Silva said, the approach toward produced water has changed. In 2008 when he first began examining the issues associated with hydraulic fracturing and the use of produced water, the standard approach was to use freshwater for all hydraulic fracturing rather than taking risks (at surface levels in terms of transportation and in the subsurface with equipment) in using high-TDS produced water. Initially, he said, work was thus focused on ways to recover and use clean water out of high-TDS produced water from Marcellus shale. At the same time, articles were just starting to be published on the use of high-TDS water for hydraulic fracturing.

Silva mentioned that every kind of wellfield area has a life cycle of about 50 years. Within that life cycle are a peak and a decline in generation of oil or gas and produced water. Similarly, if management of produced water is focused on the use in hydraulic fracturing of new wells, the number of new wells where such water may be used is not always guaranteed to meet the produced water supply. At some point, large-scale, high-TDS produced water desalination capabilities will be necessary, he suggested. Technologies exist for this type of desalination and they may be more economical than some other options (e.g., trucking to offsite facilities and deep-well injection). Silva indicated that he has always viewed deep-well injection of produced water as a last resort.

For low-TDS produced water, for example from the shales in the western United States, some success stories are worth noting, Silva said. A 1 million gallon per day desalination plant is in operation in Wyoming, for example. The shale gas produced water is desalinated at that facility essentially to drinking water standards. The water is then discharged where it could be potentially repurposed for aquifer recharge.

For high-TDS waters, Silva noted a couple of choices to consider:

• Desalinate the water to a point just prior to sodium chloride crystallization, and
• Crystallize sodium chloride to get a greater percentage of water recovery.

With these considerations in mind, Silva mentioned a possibility to evaporate the water and then collect the concentrated salt. If water begins with 180,000 milligrams TDS per liter, for example, about 40 percent of the water can be recovered which is a sizable reduction in the amount of water that has to be managed. The remaining water could be disposed through deep-well injection although the water is desirable from a well management standpoint, as it could be used onsite. To achieve 90 percent water recovery, sodium chloride has to be crystallized. Although this option offers the great advantage of almost complete water recovery, it introduces other challenges of management of the crystallized sodium chloride, including the economics and regulation of the new product stream.

In terms of regulating recovered sodium chloride, Silva said, Pennsylvania has a co-product status for minerals recovered from produced water. If sodium chloride is crystallized from produced water, the sodium chloride has to be no less harmful to human health than a product that is intentionally mined or manufactured for that same purpose, he indicated. Thus, for road salt, the comparison would be made to mined rock salt. From a Pennsylvania perspective, produced water can be rich in barium and trace amounts of radium which can affect regulation of subsequent byproducts of water treatment. Silva emphasized that all of these considerations have to be taken into account, managed, and economically modeled to consider any potential alternative use of treated produced waters. The comparison from an economic standpoint is made relative to the $12 per barrel transportation cost for water to dispose of it in a deep injection well.

Silva closed his comments by emphasizing the uniqueness of every kind of produced water and every local economy. Thus, approaches for using produced water have to be specifically addressed. What is right for produced water in the Barnett shale in Texas may be absolutely wrong for the Marcellus shale in Pennsylvania, and vice versa.

MODERATED DISCUSSION

Stringfellow opened the moderated discussion with a question to all of the panelists about water quality standards associated with the various uses of produced water. Reible responded by stating that standards would depend on the intended use, for example, whether the water is intended for drinking or for irrigation. Vidic commented that the guidelines or standards are actually set by whoever is going to use that water, with the idea to seek regulatory approval to use water for a specific application such as irrigation, livestock, or industry. In a social context, water that comes from the oil and gas industry has a negative connotation requiring communication and education to overcome if such water is to be used for other purposes, Vidic said.

A participant then raised a question about NORM and asked whether evaporation, precipitation, or flocculation of the produced water could increase the amount of NORM in the water, thereby making disposal of water difficult. The participant asked further what avenues exist to treat produced water that contains NORM.⁴ Silva suggested that because NORM is soluble species, an approach would be to evaporate the produced water so it is free of NORM, and then deal with the remaining concentrate. Vidic noted that in Pennsylvania, NORM can be recovered together with barite and that barite can meet the specifications for drilling mud to be used in drilling subsequent wells.

Another participant asked the panel about the current state of the art for characterizing constituents in produced water and for assessing their toxicity. The participant used ethylene glycol as an example and stated that there are currently no drinking water standards for ethylene glycol, which can be found in produced water. To set a drinking water standard, one needs to know what is in the water, to know the exposure pathways, and to know the toxicities of the various constituents.

Stringfellow responded by stating that the analytical capacity for a lot of anthropogenic chemicals is not robust and constitutes a science gap. Reible agreed that we do not know as much as we could about characterizing the constituents in produced water and that this area offers some really good science opportunities. Additional areas where scientific understanding is weak include the fluid and fluid-rock chemical reactions occurring in the deep subsurface that are under both high pressure and high temperature.

Another participant asked panelists to comment on current management practices (storage, treatment, transport, and use of the produced water). Reible indicated that much opportunity remains to consider direct use of produced water for other oil and gas activities. A lot of produced

___________________

⁴ NORM also includes radon gas, which could accumulate in enclosed produced water storage tanks due to radioactive decay of uranium in drilling waste and radium in produced water.

water goes directly to disposal wells because that is a simpler option, he said. When moving produced water off site for treatment or storage, one of the immediate concerns from the public is some sort of brine spill that impacts the environment. Direct onsite use of produced water can minimize some of these potential impacts because the water is generally not being handled multiple times.

A participant commented that a lot of opportunity exists for technologic solutions in treating water and reducing volume. The person then asked if any studies have analyzed the value of freshwater when a city (e.g., San Antonio) is 2 months away from running out of water or when a town loses its only freshwater well and has to truck freshwater in to supply the residents, or when the Ogallala aquifer is dropping and farmers’ wells are going dry. Reible responded by saying that water is not currently valued as a resource and people are not willing to pay for water. Another participant thought that this lack of valuing water was a fundamental problem to advancing use of produced water for alternative purposes.

A participant asked the panel to assess the capacity, regionally and as a nation, to undertake studies needed to understand more about the management, treatment, and use of produced water. What is needed to move forward? Vidic responded by emphasizing there is a need to share data and also noted the problem of a lack of common standards. Industry sometimes shares data among its own, but broader data sharing among academics, citizens, and industry does not often take place. Furthermore, he said, without produced water standards, meeting the needs of a user will be difficult because produced water from one field may have a different composition relative to another field and the quality needed even for use in one oil or gas operation may differ from that of another. The operators are not necessarily trying to meet the same standards. He illustrated this point further through an example in which the Department of Energy (DOE) was examining mercury emissions from power plants; little progress was made until DOE defined standards for flow gas composition. After the standards were defined, everybody was able to work on the basis of a similar set of expectations. In the case of produced water, a standard of produced water for subsequent uses could be established for different applications such as agriculture.

A participant said that one big difference between states in handling produced water is the design of pits (storage systems to protect groundwater) (e.g., Kuwayama et al., 2015). In terms of protection of groundwater, the practices seem to differ substantially; in some states, pits are triple lined with sensors between the linings, while in other states, linings are not required at all. The person asked the panelists whether states really differ so much in this practice or if the regulations are simply older and have been selectively adapted over time. Reible suggested that these regulations are historical and they change slowly, if at all. Another participant made a point that industry has at times taken a proactive approach to regulatory requirements and exceeded the state or federal regulation; the example provided was with regard to sampling water in wells in a community where oil and gas drilling may take place. In some cases, the federal or state regulation may only require the company to sample water at 2,500 feet prior to drilling the oil or gas well, but some companies will sample groundwater at 4,000 to 5,000 feet to provide a more complete dataset for comparison to groundwater samples after drilling for oil or gas has commenced.

A comment from another participant related to the need for standards for evaporation and percolation pits and the areas around wellheads and the potential for air quality issues in evaporation pits. Stringfellow indicated that air quality issues do exist with evaporation and percolation pits but that he was unsure if specific regulations were currently in place to address the issue. Another participant added a question about the state of standards and practices around the wellheads that would preclude effects on shallow aquifers. Silva responded by sharing information about his experience in visiting some of the wellheads where he said a wide range of provisions were put into practice to avoid any leakage of water into the ground.

Stringfellow asked panelists about calculating the water balance on oil and gas activities? What kinds of data are needed and what data are available? Reible added that operators are prob-

ably focused on a water balance for their individual fields and wells. It may be difficult, given regional differences, to make decisions based on larger-scale overall water balance calculations. Silva mentioned that he calculated a water balance on a past project located in Bradford County, Pennsylvania. He found that the use of produced water in subsequent fracturing jobs was increasing the TDS content of produced water in a predictable manner. The water balance calculations were doable when information was accessible. Vidic gave another example of the Susquehanna River Basin Commission, which accounts for every drop of water in the entire watershed because they have a mandate to do so. Every drop of water taken out of the river basin is reported, as is what is done with that water and where it went after it was removed. He also described an example from the Ohio River Basin where the Ohio River Valley Water Sanitation Commission has no regulatory or permitting authority, so anybody can go to the Ohio River and take whatever water they need.

Regarding data availability to estimate water usage, one participant added that a number of companies routinely cite water volumes in their corporate responsibility reports. This information may include gallons of freshwater and produced water used annually in their operations, often reported by geographic area.

A participant mentioned that the U.S. Geological Survey (USGS) has a Produced Waters Database⁵ that contains approximately 165,000 data points and is freely available to the public as a Web-based tool. The current version is 2.2 and they are continually working to update the database with version 2.3 in progress. The database is designed to accept any water quality data, whether they derive from academia or industry or some other entity. USGS conducts studies and populates the database with some of those studies as well. They are also trying to incorporate radionuclide data, as well as trace metals and trace elements that might be of concern. The data are quality assured and quality controlled.

Vidic commented further on the topic of data sharing by outlining a Pennsylvania example. For the past 5 years in Pennsylvania, an effort was made to bring people, industry, regulators, academics, and others who expressed a concern about water quality together via running workshops. In the first year, a few citizens attended. The following year, more regulators came. Last year, all stakeholders—regulators, industry, and citizens—participated in the workshop. A participant involved in the effort said that it had initially been funded by the National Science Foundation from a Research Coordination Network proposal; the effort is currently being supported by Penn State University with some funding from the University of Pittsburgh, as well.

Returning to the discussion about water budgets, one participant noted that the Bureau of Economic Geology in Texas compared produced water numbers to the water that is injected into producing horizons and nonproducing horizons for direct saltwater disposal. In different fields, for example in the Permian Basin, the data showed that produced water was volumetrically 20 percent less than the volumes for saltwater disposal and injection wells. The participant also mentioned that produced water quality data are not good. In Texas, data reporting for oil is done on a lease-by-lease basis. Such data are going to be very important for designing treatment systems, the participant said.

Discussion then centered around a participant’s question about the value of water and water markets. Specifically, the participant asked the panel to comment on how quickly the country is moving toward a combined examination of water content and pricing of water, which would lead toward markets in water trading. Stringfellow indicated that some water trading is already ongoing in California and Reible said that drought in California and Texas started a lot of innovation in terms of getting water to where it is most needed and valued. Vidic mentioned an effort in Pennsylvania to establish a clearinghouse for water needs within the oil and gas industry but that it was compli-

___________________

⁵ Produced Waters—Homepage, USGS. Available at http://energy.usgs.gov/EnvironmentalAspects/EnvironmentalAspectsofEnergyProductionandUse/ProducedWaters.aspx (accessed November 10, 2016).

cated by concerns related to moving water from one company’s operation to another and potential liability issues.

A data-sharing and relationship-building example was described by a participant. The person noted that in the last legislative session in Texas, the legislature put forward a bill that required the Texas Water Development Board (TWDB) to map brackish water aquifers throughout the state. The Texas Oil and Gas Association met with the TWDB and talked about specific needs. A process was developed where operators provided data to TWDB to help them map brackish aquifers. Initially TWDB was going to look at brackish groundwater up to 10,000 milligrams TDS per liter. The Association proposed examining brackish groundwater up to 30,000 milligrams of TDS per liter because industry could use that water in place of freshwater in hydraulic fracturing operations and the TWDB agreed to expand its mapping to include water with 30,000 milligrams of TDS per liter.

An online participant asked the panel to comment on the seismic effects of drilling more disposal wells near the San Andreas Fault. Stringfellow said his understanding is that a disposal well would not be placed near a fault zone like the San Andreas and that studies are being conducted across the country to examine potential seismic impacts from deep disposal wells. Vidic explained that in Arkansas there are regulations regarding where disposal wells can be placed relative to proximity to known faults. In Ohio, injection wells were closed at least temporarily because of induced seismicity and public concern. A participant mentioned that one has to recognize that there are known faults in the subsurface as well as many faults that are not known, making it difficult to guarantee that a disposal well would not be completed in the vicinity of a fault. A response to this lack of complete knowledge of the presence and orientation of faults in the subsurface would be to monitor pressure in the well during disposal.

Another example of shared experiences and collaborative effort was described by a participant related to the induced seismicity issue. Through the States First Initiative, the Ground Water Protection Council convened a work group to share practices across states that were having concerns about induced seismicity from disposal wells. Through that convening process, the group brought together about 80 subject-matter experts, technical experts, state regulators, industry, academia, and researchers from the federal government to develop a document about practices and approaches for managing risk, considering issues such as local variability, including local geology.⁶ Previously, the National Academies of Sciences, Engineering, and Medicine did a report on induced seismicity in the 2011-2012 time frame that really drove the conversation of broad stakeholder groups to continue a dialogue.⁷ From an industry perspective, that document was tremendously valuable to get a range of industry groups recognizing the issue and proactively engaging on the topic, the participant added. Stringfellow commented on the SB4 study, which was mandated by legislation in California as part of regulation of hydraulic fracturing. This forced industry, regulatory agencies, and scientists into productive discussion, he said. The report is publicly available.⁸

Another participant added that, partly a result of earthquakes to the north of Texas and in Oklahoma, the state of Texas put close to $5 million into a program called TexNet. This program was followed by the development of an industry consortium called the Center for Integrated Seismicity Research to look at an integrated set of studies, including seismology, structural geology, reservoir engineering, civil engineering, impacts, risks, and social aspects. This kind of integrated research program that brings science and engineering together with industry and regulators is potentially very powerful. These examples from the induced seismicity issue may provide a good lesson for those working on water; many voices are needed to join together and work through the questions and issues in a thoughtful, disciplined way.

___________________

⁶ Available at http://www.statesfirstinitiative.org/#!induced-seismicity-work-group/cwed (accessed July 14, 2016).

⁷ Available at http://www.nap.edu/catalog/13355 (accessed July 14, 2016).

⁸ Available at http://ccst.us/projects/hydraulic_fracturing_public/SB4.php (accessed July 14, 2016).

In closing the session, Stringfellow asked the panel to comment on the kinds of temporal variation in produced water quality that may be anticipated from different oil and gas fields in order to help design the right treatment systems. Specifically, he asked if panelists had thoughts regarding daily variations in TDS per liter and whether we have consistent information to be able to address those kinds of issues related to treatment. Vidic indicated that some information is available to understand how water quality evolves in different wells in different regions but that the treatment needs may change after the water has moved to the treatment plant; different water streams from different fields or even from the same field over time may have different chemical components, leaving a single treatment approach invalid or less useful. Resilience and flexibility in the water treatment systems is important, another participant added.