The first session of the workshop provided an overview of the geology and unconventional hydrocarbon resources in the Appalachian Basin, the technical and engineering processes used to produce unconventional resources, and the potential of these production activities to induce earthquakes. Following three plenary talks, workshop participants broke into working groups to discuss unconventional hydrocarbon resources, engineering practice, and induced seismicity, and to identify priorities for future research, as summarized below.
Geology, Resources, and Potential Activity Levels
Ray Boswell, Department of Energy National Energy Technology Laboratory
Boswell noted that the origins of unconventional hydrocarbon resources in the Appalachian Basin date back to the early Paleozoic, hundreds of millions of years ago. At that time, the basin was covered by ocean and flanked by mountains to the east. Periodic mountain building cut off the basin from the global ocean, creating anoxic conditions that allowed organic matter to be buried in marine muds and eventually converted to hydrocarbons. Some of the hydrocarbons stayed in place as the mud was converted to shale through heat and pressure, and some migrated into overlying sand reservoirs.
These geological events created five main unconventional hydrocarbon plays in the north-central Appalachian Basin:
1. Utica–Point Pleasant shales, which produce dry gas in the east (western Pennsylvania and central New York), wet gas near the Pennsylvania—Ohio border, and oil in the west (eastern Ohio), reflecting an east-west temperature gradient reached during burial. The shales are amenable to fracture stimulation because they are low in clay and high in calcite or quartz. Production is in the early stages and is focused primarily in eastern Ohio.
2. Clinton–Medina basin-centered sands, which has produced naturally flowing or technologically stimulated gas for more than a century. More than 70,000 wells have been drilled, primarily in Ohio. Development of the play in New York, Pennsylvania, and West Virginia has been discouraged by the economics of production.
3. Marcellus Shale, which is a low-clay, high-quartz shale that produces dry gas in northeastern Pennsylvania and more liquid-prone gas in southwestern Pennsylvania and northern West Virginia (Figure 2.1). Approximately 8,000 wells have been drilled thus far, perhaps less than 10 percent of the total potential.
4. Ohio and other Upper Devonian shales, which have produced gas for more than a century in West Virginia and eastern Kentucky. These clay-rich, naturally fractured shales supplied one of the biggest gas fields in the United States in the 1930s.
5. Upper Devonian–Lower Mississippian basin-centered sands and silts, which produce gas from conventional reservoirs in favorable settings. Hundreds of thousands of wells have also been drilled in lower quality tight sands and silts. These formations began producing in 1859.
Boswell next summarized the hydrocarbon resources (see Box 2.1 for definitions) contained in these five plays and the methods used to assess them. He described two basic methods: (1) assessments of gas in place, which estimate how much gas exists in a play using geologic parameters (e.g., porosity, hydrocarbon saturation) and an engineering factor; and (2) assessments of techni-
FIGURE 2.1 Thickness and extent of the Marcellus Shale. The wet gas–dry gas transition is marked by the dashed black line. SOURCE: Modified from Erenpreiss et al. (2011).
Definitions of Resources
In his presentation, Boswell distinguished four types of resources:
1. Gas in place—all gas that exists in a play, which is a function of the geology.
2. Technically recoverable resources—all gas that could be expected to be produced, which is a function of the geology as well as the technologies, policies, and regulations.
3. Economically recoverable resources—the subset of technically recoverable resources that can be produced at a profit, which is a function of market conditions.
4. Reserves—the economically recoverable resources that have been proven to exist by drilling and are available for economic production.
Volumes of different types of resources over time. Technological breakthroughs periodically increase the technically recoverable resources. Estimates of gas in place remain relatively constant, but uncertainties (gray shaded area) decrease with time. SOURCE: Ray Boswell, National Energy Technology Laboratory, from Boswell and Collett (2011). Reproduced by permission of The Royal Society of Chemistry.
cally recoverable resources, which estimate how much gas could be ultimately recovered from the play from historical production data for each well, the number of remaining well locations, and assumptions about the geology, market conditions, technology, and regulations. Assessments of technically recoverable resources are made periodically by federal government agencies such as the U.S. Geological Survey (USGS) and the Energy Information Administration (EIA). The assessments can change significantly over time because there is little production data in newly developing areas, and technological, economic, and regulatory factors evolve. For example, before the first Marcellus well was drilled, the USGS assessed a mean of approximately 2 trillion cubic feet (TCF) of gas for the Marcellus Shale (Milici et al., 2002). Less than a decade later, the Marcellus assessment had increased to a mean of 84 TCF of gas (Coleman et al., 2011).
Boswell concluded that, despite uncertainties, the Marcellus play has the potential to sustain production in the Appalachian Basin over several decades. In addition, a large number of wells could potentially be drilled, on the order of 2,000 per year for both the Marcellus and Utica shales. Detailed production and development data could be used to improve understanding of the resource and help yield more accurate assessments.
Questions. A workshop participant asked whether the U.S. government is working to fill gaps in knowledge about reserves in other countries, and Boswell answered that the USGS and the EIA have recently issued reports about global shale resources.1 Another participant asked for the current average estimated ultimate recovery values for wells in the Marcellus and the Utica shales. Boswell deferred to Joseph Frantz, who guessed that they are 6–25 billion cubic feet (BCF) per well. Finally, a participant asked about the advantages and disadvantages to private companies for providing the detailed production and development data needed to improve estimates of reserves. Frantz said that each operator typically records a daily production number, which is used to forecast future well performance. Cumulative values are provided to Pennsylvania every 6 months and to many other states every month, as required by state law. These cumulative values mask what is happening on a daily basis, but he thought that operators would not object to providing the data more frequently if asked.
Engineering and Technology for Developing Unconventional Resources:
Current and Prospective Methods for Exploration and Technology
Joseph Frantz, Jr., Range Resources
Frantz commented that early assessments indicate huge shale resources in China, Australia, and Argentina. Some of these countries are sending experts to the United States to learn how shale resources are being developed and to understand key field issues, such as the transport of water to the wells. With gas prices relatively low, U.S. operators are drilling where the gas contains natural gas liquids (e.g., propane, butane), which provide additional revenue.
In 2000, U.S. shales produced about 1 BCF of gas per day; today the Marcellus Shale is producing 11–12 BCF per day. Production increased dramatically in the late 2000s when the industry succeeded in isolating different intervals in a horizontal well and pumping multiple fracture treatments into that well. Other recent improvements include the following:
• Drilling multiple wells from one pad (pad drilling), which results in a smaller surface footprint of production than the cumulative footprint of the same number of vertical wells and associated roads.
• Using focused ion beam and scanning electron microscope techniques to characterize the pore structure of fine-grained shales, which enables areas with the highest porosity, permeability, and gas in place to be targeted.
• Combining three-dimensional seismic and microseismic techniques with treatment data and models to deduce what part of the fractured area is producing (effective fractured area). Increasing the effective fractured area is an ongoing research effort.
• Improving horizontal drilling techniques, such as using longer lateral wellbores and shorter stages, which increases production.
• Improving the efficiency of drilling and completion through technological advances or practice, which increases the number of stages that can be completed each day and the number of wells that can be drilled each year.
Frantz also touched on some safety and environmental measures undertaken by industry, including using multiple layers of casing, cemented back to the surface, through groundwater zones; placing rubber containment under every piece of equipment that could leak; using impoundments and temporary lines to store and transfer water to the wells, thereby reducing truck traffic; and improving facilities to capture fugitive emissions. Future improvements include further reducing the surface footprint of production and using less water.
These recent changes demonstrate the fast-paced nature of the industry, with ongoing development of new technologies and the constant introduction of new regulations that change industry practices every few years. Frantz concluded that natural gas production is a big opportunity for the United States and that industry must be diligent to maintain its social license to operate.
Questions. A workshop participant asked whether the microseismic data indicate which direction the fractures tend to propagate. Frantz said that hydraulic fractures have a propensity to propagate in the same direction. In the Marcellus, microseismic data and logs show conclusively that the fractures are going northeast-southwest. Another participant asked whether existing wells are being re-stimulated, and Frantz answered that the wells are still too productive to consider using that practice. When a well produces less than 500 million cubic feet of gas per day, the industry will likely look to stimulate the zones between the existing perforations.
A participant questioned whether pad drilling significantly reduces the number of vertical wells that would be needed, and Frantz answered that prior to the recent development of shale-gas reservoirs, most of Appalachia was drilled on about 40-acre spacing and less for economic plays. The participant asked if anything constrains where a well pad is placed or how many well pads can be placed on a landscape, and Frantz said that a small-footprint operation with a small rig requires a narrow access road and a small amount of land that is flat or can be leveled.
Finally, one participant wondered why a lining is used under the entire site. Frantz said that there are advantages and disadvantages (e.g., truck traffic) to lining the entire site, and thought it likely that a fair number of companies are lining only parts of the site.
Earthquakes Induced by Hydrocarbon Production:
What Texas Can Tell Us About Appalachia
Cliff Frohlich, University of Texas, Austin
Frohlich noted that earthquakes associated with oil and gas production have been documented in several states, including Texas, Arkansas, Colorado, Ohio, and Oklahoma. His research has focused on earthquakes in Texas, but the lessons learned are applicable to the Appalachian Basin.
Texas has produced huge volumes of oil and gas for more than 80 years. Statistical analysis shows that earthquakes in Texas are sometimes caused by the disposal of waste fluids in injection wells and, in some areas, by the extraction of gas and fluids. (Earthquakes do not appear to be caused by drilling, and only a half-dozen cases of earthquakes induced by hydraulic fracturing have been documented.) A strong case can be made that earthquakes are induced by waste fluid injection when the earthquakes begin more than a year after injection commenced, the earthquakes are within 2–3 km of injection wells, the injection wells handle high volumes of water (>100,000 barrels per month), the earthquakes occur along previously unknown faults, and earthquake depths are at and below the depth of injection.
The vast majority of injection wells do not cause earthquakes, and the vast majority of injection-induced earthquakes are small (lower than magnitude 3 or 4). However, a few of these earthquakes are larger (magnitude 4 and 5), which can be a concern in populated areas. Why some injection wells cause earthquakes while others do not is poorly understood. It is possible that earthquakes occur when there are suitably oriented faults near an injection well. Perhaps fluid injection pushes the rock on each side of the fault apart, reducing friction and allowing slip. However, Frohlich added that more research is needed to understand why injection near a fault is sometimes a problem and sometimes is not.
Frohlich noted that both Texas and Appalachia are geologically diverse and have natural earthquakes. If unconventional hydrocarbon development proceeds in the Appalachian Basin over the next 50 years, then there will almost certainly be some induced earthquakes.
Questions. A workshop participant asked why some earthquakes occur below the injection point, and Frohlich said he was unsure; it is an empirical observation that faults tend to be at that depth. A participant asked if the fluids being injected are high density and whether that might account for the percolation of fluids to a greater depth. Frohlich deferred to Frantz, who said that most oilfield brines are fairly heavy.
Another participant asked whether it is possible to use reverse modeling to determine the capacity of the well and where it was stimulated by the injection. Frohlich said that careful modeling of the subsurface using detailed information about the injection, hydrology, and geology is only beginning to be carried out. However, he does not expect modeling to yield a definitive answer (e.g., earthquakes will not be induced if injection rates stay below 150,000 barrels of water per month) because the geology differs from place to place.
Finally, a participant asked whether regulators can do anything to prevent injection-induced earthquakes, given that the causal mechanism is unclear. Frohlich said that regulators should invest in some fairly stringent regulations for operations in urban areas, but regulations are not needed in areas where an earthquake would have little impact. Installing seismic networks around all injection wells is not affordable because there are too many wells and not enough earthquake seismologists to examine the data.
Each working group was asked to consider the following themes:
• Geology and hydrocarbon resources—Estimated resources of natural gas, oil, and/or natural gas liquids and current and projected production levels in the main hydrocarbon-bearing geologic formations in and around the Appalachian Basin (e.g., Marcellus, Utica, and Devonian shales);
• Technical and engineering processes—Current and prospective technical and engineering processes for exploration and production of hydrocarbons from unconventional resources; and
• Research priorities—Scientific and engineering research needed to narrow or characterize uncertainties.
Issues that were directly related to these themes or that were raised by more than one working group are summarized below. The complete working group reports appear in Appendix D.
Research and Development Related to Geology and Hydrocarbon Resources
Research and development needs identified in working group discussions included the following:
• Characterizing the shale formations, including porosity, permeability, and spatial variations;
• Improving understanding of the controls (e.g., geologic factors, reservoir pressure) on induced seismicity;
• Improving understanding of the geologic controls (depositional environment, structure, thermal history) on the productive, operational, and economic lifetimes of wells in different areas of the shale plays;
• Developing ways to calculate stimulated rock volume and to relate it to gas in place, estimated ultimate recovery, and resources and reserves; and
• Determining the relationship between well completion strategies and estimated ultimate recovery.
Monthly and daily production data would facilitate this research and improve estimates of resources and reserves for a region, field, or individual well. In addition, seismic measurements and cores, especially around high-rate injection wells, are needed to improve understanding of induced seismicity. Some of these data exist, but are proprietary (e.g., monthly industry data) or are not readily accessible (e.g., some data collected by states). Some workshop participants thought that the shortage of public data adversely affects further development of shale resources and erodes the public trust.
Many of the issues associated with the geology and production of unconventional hydrocarbon resources—such as assessing the regional and cumulative economic, environmental, and social impacts—would benefit from a multidisciplinary, multisector approach. However, it is a challenge to balance industry’s immediate needs with the comparatively slow pace of research and the time it takes to establish relationships with individual researchers, universities, or research consortia. Some workshop participants suggested creating a common vision for developing shale gas (e.g., number of wells, expected revenue) or using common terminology to facilitate collaboration across academic, government, and industry sectors. Others suggested that industry and state regulators work together to determine what data should remain proprietary and what data could be made public.
Communication and Education
Better communication with the public was a common theme of the working groups, although it was not always clear what message to send or how to send it. Some participants suggested that citizens, university scientists, and government scientists, managers, and regulators need to understand hydraulic fracturing as well as industry does. Some thought it would be useful to involve public policy makers and educators in the discussion. Others saw a role for universities to play an honest broker by providing credible information to the public on the costs and risks of shale gas development.
This page intentionally left blank.