Health Impact Assessment of Shale Gas Extraction: Workshop Summary (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

3 Geographic Footprint of Shale Gas Extraction This chapter provides a summary of presentations on the process of shale gas extraction and its potential geographic impact. The first presentation provides an overview of the characteristics of a shale gas production site, the method of hydraulic fracturing, and the safety procedures used by one company. The second presentation describes federal efforts to assess the effects of shale gas production on U.S. geography, or its “geographic footprint.” Both presentations feature examples using the Marcellus Shale, a source of natural gas reserves located in the Appalachian Basin. The presentations are followed by a summary of the discussion that ensued.1 FRACTURING: ACCESSING SHALE AND TIGHT GAS David Cole, M.S. Regional Discipline Leader—Production Technology/Chemistry Shell Upstream Americas David Cole began by noting that hydraulic fracturing is, at its most basic, a process of pumping fluid into a rock faster than that rock can absorb the fluid. This results in cracks that can be held open by the injection of a solid material in order to extract the gas and oil resources in the rock. Fractures such as these can occur naturally, although some rock layers, such as tight shale layers, are naturally impermeable to fluids and gas. Since the 1940s, hydraulic fracturing has been used to extract oil or gas from the tight shale layers of rock, and this practice has been used in more than a million wells in the United States alone. Primarily, this 1 Dr. Charles G. Groat gave a presentation on “Assessing the Perceived and Real Environmental Consequences of Shale Gas Development: Report on an Initiative of the Energy Institute, The University of Texas at Austin.” A summary of that presentation is not included here because of questions that have arisen regarding the conduct of that study. 17 

18 HEALTH IMPACT ASSESSMENT OF SHALE GAS EXTRACTION process is used for extracting shale gas, but there has been movement toward extraction from oil-bearing shales as market demand increases. Hydraulic fracturing relies on the sophisticated use of pumps to create a pathway into the rock. Sand or an engineered ceramic material is then placed into the pathway to keep the cracks open after the hydraulic pressure is removed. Mr. Cole stated that the typical dimension of a hydraulic fracture is microns to 0.25-inches wide, 500- to 2,000-feet long, and 20- to 400-feet high, depending on the geography. Thirty years ago, there was typically 1 well for every 40 acres, which translated to approximately 16 well locations in a square mile, connected by gravel roads. With advances in technologies, there is often now one location per square mile and all the wells are drilled in the direction of the least principal horizontal stress. For example, wells near mountainous regions will run parallel to the mountain. Newer technologies allow engineers to use a steel drill pipe to turn and bend in order to orient the well in any direction. Mr. Cole noted that clustering wells to one surface location has a number of advantages, including reducing the number of trees cut down, reducing traffic, and reducing emissions. Nonetheless, each drilling site will affect the surface geography with wells, roads, and supporting facilities. Mr. Cole explained that careful well planning is crucial to isolate the fluids in the well and avoid contamination of drinking water. Different companies have varying strategies for water management; for example, Shell captures the water used in the fracturing process and places it into tanks with secondary containment for recycling. Other companies may use lined pits to capture this fluid, and the location is chosen based on knowledge of the depth of the groundwater and other local receptors. When planning a new well location, Shell measures the resistance to an electrical current in order to determine where fresh water is located, so that efforts can be taken to protect this water when building the well. The first step in drilling is to put in a conductor pipe, which is a structure designed to carry the load, akin to the foundation of a house (see Figure 3-1). Additional steel casing strings, blowout preventers, and other equipment are installed through the surface using a drilling mechanism and are cemented into place. The surface casing string, which consists of steel pipes coupled together with screws, is lowered into a drilled hole that runs the depth of the freshwater layer, to protect the groundwater. Cement is pumped down into this casing to seal it into place. Check valves on the bottom of the casing will help to prevent contamination and flowback and preserve isolation of the groundwater. The casing and the cement that make up the wellbore are tested to meet strict specifications of integrity before the drilling begins. Intermediate casing may be necessary, depending on where the drilling takes place. 

GEOGRAPHIC FOOTPRINT 19 FIGURE 3-1 Illustration of wellhead and casings. SOURCE: Shell, 2013. Reprinted with permission from Royal Dutch Shell. After installation of the surface casing and any intermediate casing, the production interval is drilled. This smaller steel pipe is used to transport the fracturing and produced fluids. The hole for the production casing is drilled in a vertical direction until the correct depth is reached; then, the drilling direction will turn horizontally for thousands of feet before the target formation is reached. This type of long, horizontally drilled well is common in the Marcellus Shale, which composes much of the Appalachian Basin region. The casing string is inserted and, similar to the surface casing, cement is pumped into the casing to isolate it. Finally, the well is pressure tested to ensure its integrity. To access the formation, mechanical punches produce one-quarter- to three-eighths-inch holes in the steel production casing. It is through these holes that pressurized jet streams of drilling fluid will create the fractures in the rock formation. Throughout the life of the well, pressure sensors are used to check for a firm seal. The well is then prepped for fracturing by cleaning out the casing with water. During the first release of fluid, considerable water and solids are produced. A temporary production facility is installed to separate the water from the solid waste. (The 

20 HEALTH IMPACT ASSESSMENT OF SHALE GAS EXTRACTION composition of the fracturing fluid is discussed in more detail in Chapter 5.) Mr. Cole stated that fracturing is an engineered process that takes into account the strengths and properties of the target rock formations. Understanding these properties makes the fracturing of rocks predictable and consistent. Microseismic listening techniques are used as fracturing fluids are pumped into a well. The sound of rock breaking is an indication of where activity is happening underground. This technique is used to optimize gas development, because it allows the engineers to know the orientation and length of the fractures. Further, as more wells are drilled, this information helps to plan for future well placement. One of the side benefits is that microseismic events allow for a company to have an indication of dimension. After the gas or oil is extracted, it is transported via pipeline. The primary concerns in the development of these wells are ensuring that groundwater is protected and that gas and oil are not lost during the process, said Mr. Cole. Fracturing activity takes place thousands of feet from freshwater, and securing the casing with cements helps maintain isolation from drinking water sources. Transparency of information about hydraulic fracturing is also an important consideration. Websites such as Fracfocus.org, a chemical disclosure registry operated by the Groundwater Protection Council, provide a voluntary reporting site for each well (GWPC and IOGCC, 2013). Mr. Cole stated that Shell has reported every well since January 2011 and sees these efforts as best practices for the industry. Additionally, Mr. Cole reported that Shell operates under a series of principles to  ensure the safety of workers and well integrity,  conduct operations to protect groundwater and reduce water use as reasonably practicable,  protect air quality and control fugitive emissions,  work to reduce the operational footprint, and  engage with local communities regarding socioeconomic impacts that may arise from Shell operations. Discussion Following Mr. Cole’s presentation, Roundtable and audience members were invited to ask questions. Christopher Portier began by asking if Shell has ever done a health impact assessment for shale gas extraction, to which Mr. Cole replied “yes.” Bernard Goldstein asked if the tremendous increase in the ability to extract natural gas would continue, and if states that began these activities earlier are receiving as much economic benefit as those that waited for improved extraction technologies. Mr. Cole replied that it has always been known that these resources existed, and only recently was it learned that it could be turned 

GEOGRAPHIC FOOTPRINT 21 into reserves and, in turn, produced. Because development is ongoing and companies are continually evaluating and looking for more efficient ways to extract shale gas, there is likely no problem for states that started early. Richard Jackson asked for more details on what the wells are producing (i.e., oil) and how shale gas is extracted and removed from the sites. Cole described the process, which began with natural gas in the Barnett Shale and a movement toward oil-bearing shales, such as in North Dakota. The movement toward oil extraction is driven by price differentials on an energy basis between oil and gas. Natural gas and oil are both pipelined from the sites, though smaller developments can truck the resources out. An audience member questioned if there was a consensus in the industry on what level of transparency they are willing to provide to the public on fracturing fluids. Mr. Cole reiterated that Fracfocus.org is a valuable source for such information, and has more than 70 different companies with 9,000 wells reporting to that database. Mr. Cole also mentioned that the service companies’ trade secrets may not allow complete reporting, but some states have required that this be reported to them in case they need to respond to a related problem. Most of the fracturing fluid chemicals are not secret, but the formulations of these chemicals are proprietary. Linda McCauley asked about the range of the size of the industry, particularly the number of smaller companies. Mr. Cole noted that major oil companies make up a small fraction of the overall oil produced in the United States, and that the industry is dominated by large independents. An audience member via webcast asked how many Marcellus Shale wells have been drilled in multiple directions, to which Mr. Cole replied that almost everything in the Marcellus Shale is horizontally drilled. Another audience member inquired further on the fluids used in fracturing and expressed concern about the composition of the fluids (e.g., if the chemicals are endocrine disruptors) and the protection of water. Mr. Cole restated that he is not an expert on the chemicals used in these fluids and could not speak to that, and that the real issue for safety is in isolating the fluids being pumped in and out and maintaining the well’s integrity. From Mr. Cole’s perspective, these issues of concern are not happening. In response to another question about how the target shale layer is drilled, Cole responded that directional drilling techniques allow the engineers to always know where the well is relative to the target location. In some cases, geosteering tools can be used. The next question from the audience addressed drilling intensity rates. The audience member recalled learning that wells for fracturing tend to produce a lot of gas in the first year, and then production declines, and that refracturing a well does not produce as much the second time around. In addition, it is suggested that there is a time limit as to how quickly a company needs to drill once it is granted a lease, creating a push to drill faster. Mr. Cole defined the drop-off in production as 

22 HEALTH IMPACT ASSESSMENT OF SHALE GAS EXTRACTION “hyperbolic decline,” which occurs as new wells are drilled and experience a peak, followed by a decades-long tail of production. For the final question, an audience member asked if Shell conducts testing on water sources before fracturing, to which Mr. Cole replied “yes.” Shell tests water sources before drilling and is committed to proactive testing after a certain amount of follow-up time. GEOGRAPHIC FOOTPRINT Michael Focazio, Ph.D. Assistant Program Coordinator, Toxic Substances Hydrology Program U.S. Geological Survey Michael Focazio began by noting that human activity generally affects the environment in some way, whether it be as benign as a nature hike or as substantial as clearing a field to plant corn. The evidence of those activities is referred to as a geographic footprint. Environmental health faces a challenge in measuring the impact associated with these footprints. The U.S. Geological Survey (USGS) has endeavored to perform scientific assessments of geographic footprints, both spatially and temporally. Dr. Focazio explained that one way the USGS measures geographic footprints is by surveying the aerial extent of the land and observing land changes associated with the activities. This method has been used to describe the impact of shale gas extraction on land surrounding a well site. As detailed previously in Mr. Cole’s presentation, the well site for hydraulic fracturing involves roads, trucks, water storage, and surface drilling, which contribute to the geographic footprint of that activity. U.S. shale gas extraction increased more than fourfold between 2007 and 2011 (GAO, 2012). As the activity increases, the geologic extent of these recoverable resources, or sources of shale gas, becomes included in the definition of the extraction’s geographic footprint and expands beyond measurements of surface land change. Dr. Focazio stated that the geographic footprint of shale gas extractions can be conceptualized on three levels: national, regional, and local. Nationally, an assessment measures the extent of technically recoverable resources: where they occur and how frequently. It is useful to compare it with the superimposed map of North America that shows the geologic extent of the resource, as displayed in Figure 3-2. On a regional level, the geographic footprint factors in roads, pipelines, and other infrastructure developed to extract and transport the gas. Pipelines in particular are a major component of the infrastructure 

GEOGRAPHIC FOOTPRINT 23 FIGURE 3-2 U.S. mean shale gas resources. SOURCE: USGS, 2012. with potentially tremendous impact when crossing through sensitive ecosystems. The local impact includes surface-level activity, such as land clearing and the construction of well pads or water storage facilities, which entails well site operations, ponds, and roads at the extraction site. The design of a shale gas extraction site is similar to what it is being used in oil and natural gas extraction, in terms of vehicles for transportation, the derrick, water lagoons, and water sources. The USGS used an example of activity in the Marcellus Shale to describe a geographic footprint, said Dr. Focazio. Mostly permits for oil and gas have been issued although coal mining and methane extraction are also occurring in this region. The map in Figure 3-3 shows the extent of the number of permits that have been issued. Technically, this is equivalent to the number of sites expected, and they are located all along the Marcellus Shale, ranging across New York, Ohio, Pennsylvania, Virginia, and West Virginia. Most of the permits are not being used, but the potential geographic footprint can be assessed prospectively. The aerial view in Figure 3-4 is the before (2006) and after (2010) development of a well pad site. From this image, it is clear that the land has been cleared, and that roads and well pads have been added. In itself, the clearing of the land may have important ecological impacts. Dr. Focazio noted that in the Marcellus Shale, the area that is covered by these sites averages 7 acres, ranging from 5 to 10 acres total for shale gas 

24 HEALTH IMPACT ASSESSMENT OF SHALE GAS EXTRACTION FIGURE 3-3 Drilling permit locations. SOURCE: USGS, 2012. sites. In Pennsylvania, there are approximately 10,000 permits issued for shale gas extraction. The approximate land area of that geographic footprint is 76,000 acres. This is a very small percentage (0.3) of the land area in Pennsylvania, and not all of this is currently being drilled. After considering the local impact on the land, Dr. Focazio noted that it is important to understand the factors that can expand the activity to a regional level. A network of roads needs to be developed in order to load, unload, and transport the shale gas. Vehicles moving at an increased frequency than before and traveling to locations that had not been accessed before construction of the well pad may have a tremendous impact on the geography. The well pad will also require pipelines and the infrastructure to move fluids. Processing sites will also be created and, in most cases, those are centralized facilities that differ from the well pad and are often larger. Most of the attention around shale gas extraction has focused on understanding the human health consequences, said Dr. Focazio. Assessments of the geographic footprint are primarily concerned with the measurable ecological impacts of the activities. In many cases, these concerns will overlap. The clearing of land will affect native species, the increase in transportation will change the air composition, and the increased use of water will affect hydrological cycles. This can be taken into perspective if we consider that most concerns from the activity have been driven by the proximity to residential areas. 

GEOGRAPHIC FOOTPRINT 25 FIGURE 3-4 Local footprints of Marcellus Shale gas well sites from 2006 (left) and 2010 (right). SOURCES: Slonecker et al., 2013. The way that geographers have been monitoring the issue is by utilizing programming computers to detect land changes in a certain area. The computer is configured to recognizing patterns; for example, ident- ifying how much fragmentation appears in the land or more sensitive details such as mapping well pads. An important aspect of this kind of monitoring methodology is time. Acreage compares the differences before and after the well pad has been constructed, but pattern recog- nition programs have the capability to record changes over time. After the well has been constructed, the program, for example, can identify the density of the forest before and after a shale gas production site is constructed, which may entail reducing forest to build pipelines. Offsite activities and infrastructure also contribute significantly to the geographic footprint of hydraulic fracturing. Use of sand to fill the fractures is a characteristic of shale gas production that makes sand mining an important consideration. It is an operation that requires an extensive infrastructure and that covers a large expanse of land. For the Marcellus Shale sites, sand mining is mostly done in the Midwest, primarily in Minnesota and Wisconsin, and transportation is mostly done by train. Sand mining has grown exponentially in the last decade with the increase in demand of natural gas and, as a consequence, the geographic footprint of shale gas extraction includes an even larger region. At this time, there is no objective conclusion about what constitutes an unacceptable geographic footprint for shale gas extraction, said Dr. Focazio. Over time, geographic scientists will strive to gain a better understanding of the long-term impacts. Current shale gas resources, such as the Marcellus Shale, are finite. If demand remains high, production may focus on deeper sources such as the Utica Shale, which is located underneath the Marcellus Shale and not currently used in Pennsylvania. 

26 HEALTH IMPACT ASSESSMENT OF SHALE GAS EXTRACTION Discussion Dr. Focazio emphasized the importance of a balanced, scientific perspective when evaluating the impact of shale gas extraction in response to Roundtable members’ questions about the promotion of misinformation and the ubiquity of natural gas wells in certain locations. One audience member asked about the extent of radioactive releases during hydraulic fracturing. Dr. Focazio responded that the geology of the shale structure dictates some of the effects of the extraction. For example, radionuclides naturally occur in the rock formation and may be released in the course of fracturing. It is unclear if hydraulic fracturing mobilizes more radioactive releases than would normally be produced; however, analyses are under way to examine the water produced in the extraction process for the presence of radionuclides and other chemicals. Another audience question centered on asbestos-type molecules and radiation, and asked if there has been any measurement of those at hydraulic fracturing sites. Dr. Focazio replied that these data are collected to characterize aquifer water quality generally, including other chemicals, but are not collected specifically for geographic footprinting. Throughout the discussion, Dr. Focazio reiterated that measuring the geographic foot- print of shale gas extraction requires science that considers the compre- hensive activities of hydraulic fracturing and the related health and safety consequences. REFERENCES GAO (U.S. Government Accountability Office). 2012. Oil and gas: Information on shale resources, development, and environmental and public health risks. Washington, DC: GAO. GWPC and IOGCC (Ground Water Protection Council and Interstate Oil and Gas Compact Commission). 2013. FracFocus Chemical Disclosure Registry. http://www.fracfocus.org (accessed May 30, 2013). Shell. 2013. Hydraulic fracturing: Your questions answered. http://s00.static- shell.com/content/dam/shell/static/usa/downloads/onshore/abc002-hyd-frac- insert0623.pdf (accessed May 30, 2013). Slonecker, E. T., L. E. Milheim, C. M. Roig-Silva, and A. R. Malizia. 2013. Landscape consequences of natural gas extraction in Allegheny and Susquehanna counties, Pennsylvania, 2004–2010. Open-file report 2013- 1025. Reston, VA: U.S. Geological Survey. USGS (U.S. Geological Survey). 2012. National oil and gas assessment 2013 assessment updates. http://energy.usgs.gov/OilGas/AssessmentsData/ NationalOilGasAssessment/AssessmentUpdates.aspx (accessed May 30, 2013).