Plugging the Gaps in Inactive Well Policy Jacqueline Ho, Alan Krupnick, Katrina McLaughlin, Clayton Munnings, and Jhih-Shyang Shih With assistance from RFF Visiting Fellows Nathan Richardson (University of South Carolina) and Lucija Muehlenbachs (University of Calgary) MAY 2016 Resources for the Future Ho et al. Plugging the Gaps in Inactive Well Policy Jacqueline Ho, Alan Krupnick, Katrina McLaughlin, Clayton Munnings, and Jhih-Shyang Shih With assistance from RFF Visiting Fellows Nathan Richardson (University of South Carolina) and Lucija Muehlenbachs (University of Calgary) Contents 1. Introduction ......................................................................................................................... 3 2. Classifying Wells by Production, Abandonment, and Ownership Status ..................... 4 3. The Scale of the Inactive Well Problem ............................................................................ 5 Key Findings and Recommendations ................................................................................. 5 The Literature on the Environmental Risks of Inactive Wells ........................................... 6 Pollutants and Impacts ........................................................................................................ 6 Risk Pathways and the Role of Well Construction in Minimizing Leakage Risk .............. 6 Factors Affecting the Magnitude of Risk............................................................................ 9 Empirical Estimates of Magnitude of Pollution Potential ................................................ 11 Number of Inactive Wells in 13 States ............................................................................. 13 4. Regulations on Inactive Oil and Gas Wells ................................................................... 16 Key Findings and Recommendations ............................................................................... 16 Methodology ..................................................................................................................... 16 Review and Comparison of State and Federal Regulations .............................................. 17 Discussion of Regulatory Elements .................................................................................. 21 Policy Recommendations and Conclusions ...................................................................... 44 Ho, research assistant, Resources for the Future (RFF); Krupnick, RFF senior fellow and co-director of RFF’s Center for Energy and Climate Economics; McLaughlin, research assistant, RFF; Munnings, research associate, RFF; Shih, fellow, RFF. Resources for the Future (RFF) is an independent, nonpartisan organization that conducts rigorous economic research and analysis to help leaders make better decisions and craft smarter policies about natural resources and the environment. © 2016, Resources for the Future www.rff.org 1 Resources for the Future Ho et al. 5. Conclusions, Recommendations and Future Research ................................................. 46 Bond Amounts .................................................................................................................. 47 Well Management and Monitoring ................................................................................... 47 Inactive and Orphaned Well Programs ............................................................................. 48 Environmental Risks of Inactive Wells ............................................................................ 48 Inactive Well Regulations and Programs .......................................................................... 49 Appendix A. Average Well Depth and Associated Average Bond Amounts by State .... 51 Appendix B. State Oil and Gas Regulations ....................................................................... 55 Acknowledgements ............................................................................................................... 78 References .............................................................................................................................. 78 www.rff.org 2 Resources for the Future Ho et al. 1. Introduction The environmental and financial consequences of a large and probably growing number of inactive wells remain largely unexplored. Based on some new work on methane leaks from such wells and reports of state liabilities for plugging wells and restoring production sites, a closer look at these issues is warranted. Regulatory, environmentalist, academic and industry attention has focused much more on the environmental consequences of oil and gas development from active wells than on those from inactive wells, or wells that have ceased production. This focus is understandable given concerns about drilling, fracking, waste handling and the like; but there are many more inactive wells than active wells—one estimate suggests that at least 3.5 million oil and gas wells have been drilled in North America (Brandt et al. 2014), of which 825,000 are currently in production.1 The remaining wells are presumably inactive. Left unplugged or not properly plugged, inactive wells threaten human and environmental health. Recent research suggests that these wells can leak methane (a powerful greenhouse gas) into the atmosphere (Kang et al. 2014). They could also provide a pathway for surface runoff, brine, or hydrocarbon fluids to contaminate surface water and groundwater (Kell 2011; King and King 2013; King and Valencia 2014). Well sites that are not properly reclaimed can contribute to habitat fragmentation (Drohan et al. 2012) and soil erosion, and equipment left on-site can interfere with agricultural land use and threaten wildlife habitat (DOI 2015). Whether even properly plugged wells can leak is still an open question. Even if wells have a responsible operator on record, they may still represent a potential envi- 1 This total may be an underestimate—many historic wells were drilled before well-permitting regulations were introduced and thus may not be recorded. ronmental risk and financial liability to states. Due to a lack of monitoring capacity, a well that has been inactive for an extended period of time and is noncompliant with environmental standards may be allowed to remain in temporary abandonment or inactive status, such that they can be reactivated when market or technology conditions improve, instead of being permanently plugged and abandoned. Eventually these wells may become orphaned. For instance, a 2014 performance audit of the inactive well program managed by Louisiana’s Office of Conservation found that 46.5 percent of 11,269 wells identified as having future utility had held that status for more than 10 years; 22.8 percent of the 8,682 wells that were ultimately orphaned had been in future utility status prior to becoming orphaned (LLA 2014). Any growth in the number of orphaned wells adds to the already-large population of legacy orphaned wells from an earlier era. A further risk is posed by wells that will become inactive in the future. It is possible that future wells will be less problematic than historic wells because of better regulations for plugging and abandonment, improved technologies for well construction (such that the original bore hole and casings are in better shape for plugging), and growing public pressure on regulators and industry to protect against environmental risk. However, even if less risky, each additional well produced will eventually add to the growing stock of inactive wells. In addition, even wells that have been properly plugged with modern technologies may leak as cement is subject to shrinkage, cracking, and other types of failure. This report discusses the environmental and regulatory challenges of inactive wells, with an eye towards reforming their regulation. Section 2 briefly reviews definitions and classifications. Section 3 assesses the magnitude of the concerns related to inactive wells that are left unplugged by identifying the specific environmental threats posed by leaking wells and by estimating the number of inactive wells in the United States. Stringent regula- www.rff.org 3 Resources for the Future Ho et al. tions are essential for mitigating such environmental and financial risks; thus, Section 4 reports on a survey of inactive well regulations in 22 oil and gas states and on BLM lands. The section identifies the regulations that are the most crucial and discusses the heterogeneity in regulatory approaches across the governments. Policy recommendations are aggregated and presented in Section 5. A forthcoming paper (Shih et al.) estimates the costs of plugging inactive wells in order to reduce these risks, including a discussion of the financial liability to governments that they represent and the extent to which these costs are internalized by private operators. At times we refer in this report to findings in that paper. 2. Classifying Wells by Production, Abandonment, and Ownership Status Some states use different definitions to describe similar well statuses. We therefore introduce generic terms meant to coherently capture categories of inactive wells, while acknowledging that these terms differ from definitions used by a significant number of states. Indeed, many states do not use the term “inactive wells” as we do throughout this report. We identify seven terms that classify wells into different status and ownership categories, as displayed in Figure 1 below. A well’s status switches from active to inactive (or idle) after it stops producing oil and gas after a certain period of time, which ranges from one month to one year for most states. If an operator maintains ownership of that well, it either undergoes decommissioning (which we define as plugging the well bore, removing equipment, and restoring land surrounding the site) at the expense of the owner, or it becomes temporarily abandoned. Temporary abandonment is technically a transitory state, where the well might return to production or be decommissioned in the future; in practice, however, wells can remain temporarily abandoned indefinitely in certain states and circumstances. If a well does not have an owner, it is deemed an orphaned well and either undergoes decommissioning at the expense of the government or becomes abandoned. A well may become orphaned as it becomes inactive (resulting FIGURE 1. STATUS AND OWNERSHIP OF OIL AND GAS WELLS Active (i.e., production > 0) Idle (i.e., production = 0) Temporarily Abandoned Orphaned and Idle Orphaned and Temporarily Abandoned Decommissioned by Owner Decommissioned by Government www.rff.org 4 Resources for the Future Ho et al. in an orphaned inactive well) or after it is temporarily abandoned, which results in an orphaned temporarily abandoned well. Well operators becoming financially insolvent, or simply not found at the time a well requires decommissioning, is a primary cause of wells becoming orphaned. It is currently unclear what number of the approximately 3.0 million inactive wells in the United States belong to each of these categories. However, given the advent of unconventional wells and the growing importance of natural gas domestically and in export, the United States will likely face a rise in the number of inactive wells in the coming year. Most of the news, popular press, and academic literature on the topic of inactive wells focuses on orphaned or temporarily abandoned wells (Mitchell and Casman 2011; Frosch and Gold 2015). Our study considers all six categories of inactive wells, because all of them (regardless of ownership or operational status) can pose environmental risks. 3. The Scale of the Inactive Well Problem Key Findings and Recommendations  Inactive wells can leak pollutants, including methane and brine, as well as heavy metals and naturally occurring radioactive substances; these pollutants may contaminate groundwater, surface water, or, in the case of methane, be released into the atmosphere.  The pathways through which leakage may occur are well documented in the literature. These pathways include mechanical integrity failure, failed well casings, and cement failure. Well construction and well plugging regulations should protect against these failures.  The likelihood of leakage from an inactive well depends on a number of factors, most importantly, the quality of well construction at the time it was drilled and the abandonment measures that have been taken.  The empirical literature provides anecdotal evidence of leakage from wells left unplugged but does not characterize the rate at which these wells leak. We are aware of only one piece of research that provides measurements of methane leakage rates from inactive wells.  The empirical literature does not distinguish between the environmental damage caused by different types of inactive wells (e.g., temporarily abandoned vs. plugged and abandoned wells; historic wells vs. wells drilled more recently). Although wells that have been plugged might still leak due to cement shrinkage, opinions on the extent to which this happens are divided. This is an area in need of further research.  Data from 13 states with significant oil and gas production shows that about 12 percent of the inactive wells in these states have not been decommissioned. The percentage in each state varies significantly from one percent to 56 percent. How much of an environmental threat are inactive wells in the United States? To answer this question comprehensively and empirically, four key pieces of information are needed: the type, quantity, and toxicity of pollutants that may leak from each well; the abandonment status (e.g., whether they are plugged) and characteristics of inactive wells (e.g., the quality of their construction) and how these affect how much of a risk they pose; the number of inactive wells; and the proximity of human and ecological populations to hazardous wells. Because currently available data and literature on www.rff.org 5 Resources for the Future Ho et al. these four components are limited, answering Pollutants and Impacts this question is challenging. This section adMethane is the primary pollutant of concern dresses the first three of these four components, in natural gas. Methane from leaking wells enfirst by reviewing the literature on the enviters the atmosphere directly, contributing to ronmental risks posed by inactive wells. In regreenhouse gas emissions concentrations viewing this literature, we highlight the major (Dusseault et al. 2000; Kang et al. 2014). Mepollutants of concern, then we identify the thane can also pose human health risks when pathways through which inactive wells can entering shallow groundwater or surface water cause environmental damage and describe how and contaminating household drinking water. inactive wells of certain types and characterisMethane poses an explosion and an asphyxiatics are more risky than others. Understanding tion hazard,2 either during well water extraction this then allows us to identify specific regulaor by accumulating in basements and well pits tions that are important for managing the risk (Jackson et al. 2013).3 Other pollutants of conin inactive wells, which we address in section cern in natural gas include nitrogen oxides, sul3. Next we estimate the number of inactive fur dioxide, and hazardous air toxics like benwells in the United States using data from and zene, toluene, ethylbenzene, and xylene (Latindividual states. tanzio 2013). The Literature on the Environmental Brine is another key pollutant that can miRisks of Inactive Wells grate from hydrocarbon formations to surface The pathways through which inactive oil and gas wells can cause environmental harm, if they are not properly plugged, are well documented in the engineering literature on well integrity and procedures for proper plugging and abandonment (see, e.g., King and Valencia 2014). Additionally, the literature has also commented on the conditions under which environmental risk may be exacerbated, such as subsurface geologic conditions and the proximity of ongoing production activities, as well as the effect of well construction and well plugging regulations on the degree of risk posed by inactive wells. Less well understood is the actual, quantified risk posed by the population of inactive wells in the United States, both plugged and unplugged, as there have been few empirical studies done on the topic (see, e.g., Kang et al. 2014). Although the literature treats oil and gas wells as a collective group, we focus on the risks from gas wells. Nonetheless, most risks from oil wells would be of the same type, with the exception oil leaks. water or freshwater aquifers, rendering the water non-potable, particularly if the brine has elevated total dissolved solids or contains naturally occurring heavy metals, such as barium, and radioactive materials (Jackson et al. 2013). Pollutants in surface runoff may also flow into an unplugged wellbore and contaminate groundwater (API 1993). Risk Pathways and the Role of Well Construction in Minimizing Leakage Risk Oil and gas wellbores penetrate shallower strata before reaching the target hydrocarbon formations, and these strata may contain groundwater for drinking or other surface uses This asphyxiation hazard arises as a result of methane’s ability to displace the oxygen in an enclosed space. 2 It should be noted, however, that leaking wells are not the only source of methane. The presence of natural seepage pathways allows methane to migrate slowly from hydrocarbon zones to the surface (King and King 2013). 3 www.rff.org 6 Resources for the Future Ho et al. (Davies et al. 2014). Nonproducing wells left unplugged or that have been improperly plugged may facilitate the migration of pollutants between these zones and/or the surface or atmosphere (Calvert and Smith 1994; Kell 2011; King and Valencia 2014). Leakage pathways include the migration of methane from producing or nonproducing hydrocarbon formations, or sometimes from aquifers, to the atmosphere; of brine from saltwater zones to freshwater aquifers, surface water, or surface soils; of oil and gas from hydrocarbon formations to freshwater aquifers, surface water, or surface soils; or of pollutants in surface runoff into freshwater aquifers (API 1993). Two major types of leakage pathways are surface casing vent flow (leakage between the production and surface casings) and gas migration (leakage outside the outer casing; Erno and Schmitz 1996). For a well to leak, there must be (1) a source of fluid (gas or liquid), (2) a breakdown of one or more well barriers—that is, a pathway for the fluid to migrate either within the cement medium or adjacent to it, and (3) a driving force for the migration of fluid, such as a pressure differential in the wellbore due to a higher pressure in the hydrocarbon formation than in the wellbore annulus (the space between the wellbore and the casing; Davies et al. 2014; Bonett and Pafitis 1996). Proper well construction and P&A procedures should likely prevent such conditions and therefore protect against fluid migration, at least in the early life of the decommissioned well. During well construction, it has been common practice since well integrity regulations were introduced to protect the various zones— groundwater aquifers, hydrocarbon formations, and the surface—using barriers such as well casing and cement, to perform what is known as zonal isolation (King and King 2013; King and Valencia 2014). Well construction elements that protect against fluid migration to the subsurface and gas emissions to the atmosphere fall into a few categories: layers of well casing, cement used to fill the annular space between casings or between the outermost casing and the wellbore, and the wellhead or Christmastree assembly (API 1993). Depending on the unique geologic conditions and depth of the well, there may be one to three barriers in a low-risk area and two to five barriers in a high-risk area, where casing and cement are each considered individual barriers (King and King 2013). The most effective practices for zonal isolation include placing surface casing below a freshwater aquifer and cementing it to the surface, as well as setting production casing from the surface to the production zone and cementing it (at least for a substantial distance, if not all the way to the surface) to prevent the vertical migration of fluids behind the pipe (API 1993). There may also be multiple layers of intermediate casing between the surface and production casings depending on the depth of the well (Dusseault and Jackson 2014). To ensure the integrity of the barriers, a number of other well construction practices are important, including ensuring that the density of the cement slurry is properly designed and that mud is removed from annular spaces in the wellbore (Bonett and Pafitis 1996). Figure 2 is a diagram of a properly abandoned well showing the different zones that need to be plugged in order to ensure zonal isolation. Thus, proper well construction is the first step towards ensuring zonal isolation over the entire lifetime of the well, including during production, after the well becomes inactive, and after P&A. P&A then builds on the completion design, further isolating parts of the wellbore. Effective P&A designs depend on robust evaluations of potential leakage pathways unique to the well (King and Valencia 2014). Depending on the quality of the well construction and P&A, leakage pathways may form in modern well construction through one or more mechanisms (leakage pathways associated with pre-regulatory wells are discussed in a later section): www.rff.org 7 Resources for the Future Ho et al.  Mechanical integrity failure. The wellhead or Christmas-tree assembly may be inadequate to contain fluids, creating a pathway for methane to leak to the atmosphere (API 1993).  Casing failure. Casing may fail due to failed casing joints, casing collapse from sustained casing pressure, and/or corrosion over time due to the presence of brine or of hydrogen sulfide, which forms sulfuric acid upon contact with water (Davies et al. 2014; Watson and Bachu 2009; King and King 2013).  Cement failure. Multiple issues can contribute to cement failure. For instance, cement may shrink over time. This is particularly likely if the water content in the cement is too high, which causes the cement to lose water while setting (Dusseault et al. 2000). This causes a microannulus to develop between the cement and the rock wall and/or casing. Figure 3 is a visual representation of how cement shrinkage can create a fluid migration pathway. There is a possibility that all wells plugged with cement will eventually leak, given enough time, due to this issue of cement shrinkage (Kunz 2015), although this has not been supported by empirical research. These basic pathways can cause leaks regardless of whether the well has been permanently abandoned, temporarily abandoned, or merely shut in. The risk that any of these leakage pathways may develop may be greater or lesser, depending on a variety of factors discussed in the next section. FIGURE 2. SCHEMATIC OF A PROPERLY ABANDONED WELL Source: King and Valencia (2014). FIGURE 3. CEMENT SHRINKAGE CREATING MICROANNULI Source: Bonett and Pafitis (1996). www.rff.org 8 Resources for the Future Ho et al.  Cement quality: The American Factors Affecting the Magnitude of Risk Petroleum Institute published cement standards for well construction and well plugging in 1953, specifying eight classes of cement designed to resist various subsurface conditions such as high pressure, salinity, and sulfate concentrations (NPC 2011), although King and King (2013) cite the mid-1970s as the time period when cementing standards improved systematically throughout the industry through the introduction of cement design software and the introduction of more robust cements into the market. The magnitude of leakage risk that any given well presents is determined by a number of factors, including the quality of well construction, the plugging and abandonment measures that have been taken, and other factors. Well Construction Proper well construction is the first critical safeguard against fluid migration, not just during a well’s production life but also after it becomes inactive. For instance, an inadequately cemented annulus provides a conduit for gas migration to occur between hydrocarbon formations and freshwater aquifers (Dusseault and Jackson 2014). Well construction elements such as properly cemented production casing and surface casing also enhance the success of plugging operations by improving the effectiveness of cement plugs (API 1993). In addition to the stringency of well construction regulations, market conditions at the time of well completion have also affected the integrity of construction. In their study of how wellbore characteristics affect the leakage potential of wells in Alberta, Watson and Bachu (2009) find that high oil prices are highly correlated with high leakage occurrence between 1973 and 1999. They hypothesize that increased production activity in response to high oil prices can result in limited supplies of equipment and manpower and therefore suboptimal cementing practices.4 The integrity of a well’s construction depends primarily on its vintage, as the quality of construction depends heavily on the well construction regulations in place at the time that the well was drilled. Many historic wells in the first oil and gas states like Pennsylvania, Texas, and Ohio were drilled in the nineteenth century before well construction regulations were introduced (Calvert and Smith 1994; King and Abandonment Measures King 2013; King and Valencia 2014). The earWhether open annular spaces allow for fluliest wells were drilled before operators began id migration in an inactive well also depends to use steel pipe, and those wells were cased with wood (King and Valencia 2014). King and on the abandonment measures that have been King (2013) list the major changes in well con- taken in that particular well. This should not be struction regulations that have been introduced confused with a well’s official abandonment since the 1820s and the estimated pollution po- status, as different jurisdictions have different tential associated with wells constructed at dif- definitions for each abandonment status, and terms such as “shut in,” “temporarily abanferent times. The well construction elements doned,” “suspended,” and “inactive” are often most crucial for reducing the pollution potential from inactive-unplugged wells are:  Zonal isolation: Most wells constructed after the late 1930s were required to have multiple cement and casing barriers to prevent fluid migration into freshwater aquifers (API 1993). An alternative explanation for this correlation is that small, independent operators tend to emerge during times of high oil prices, and the integrity of wells drilled by these operators may be lower. 4 www.rff.org 9 Resources for the Future Ho et al. used interchangeably (API 1993). Rather, the relevant question is what barriers are put in place after the well has stopped producing. Three major categories distinguish between these abandonment measures: 1. No isolation of hydrocarbon zone: An operator may shut off production from a well for short periods of time in response to temporary market conditions. The operator shuts off the wellhead but leaves the casing exposed to the completion interval. 2. Temporary isolation of hydrocarbon zone: In most cases, a well will only be classified as temporarily abandoned if the completion interval has been isolated. However, the interval is only temporarily isolated if the isolation barrier (such as a bridge plug) can be easily drilled through and the hydrocarbon formation re-accessed. This might be the case if the operator wishes to bring the well back into production. 3. Permanent isolation of hydrocarbon zone and freshwater aquifers: In a permanent P&A operation, the completion interval, any intermediate oil and gas-bearing zones, and any freshwater aquifers are isolated, and the rest of the wellbore that is not cemented is filled with mud. 2015) and by a representative of the Alberta Energy Regulator (Taylor, 2016) that we spoke to. However, there is very little anecdotal evidence available to support this, and a few reviewers of the draft of this report said that this concern about properly plugged wells leaking was exaggerated or nonexistent. Among wells that have been permanently plugged and abandoned, there is heterogeneity in leakage potential depending on the specific abandonment methods used. Watson and Bachu (2009) find that wells plugged using bridge plugs are more likely to leak than wells that have been plugged using cement plugs and cement retainers. A detailed description of different plugging methods can be found in NPC (2011). Of wells that are plugged, improperly plugged pre-regulatory wells pose the greatest problem. These wells were drilled before P&A regulations were systematically introduced and were simply plugged with materials such as brush, wood, and rocks (NPC 2011). For instance, the Texas Railroad Commission began to regulate well plugging in 1919, although cementing procedures were not introduced until 1934 and freshwater aquifers were not required to be protected until 1957 (Texas RRC 2000). In general, oil and gas states began to require cement in P&A operations in the 1950s and introduced stricter standards to protect freshwater aquifers in the 1970s, along with the passage of the Safe Drinking Water Act in 1974 (NPC 2011). In general, wells that have been permanently isolated are less likely to leak than are wells that have been only temporarily isolated, or not Other Oil and Gas Activities isolated at all. Kang et al. (2015) find that As mentioned earlier, another crucial factor plugged wells have lower leakage potential influencing leakage potential is the presence of than wells that have not been plugged, although this result is not statistically significant. None- a pressure gradient or fluid buoyancy gradient within the wellbore. If there are unplugged or theless, plugged and abandoned wells could improperly plugged wells in an area, it bestill leak. Alberta’s Abandoned Well Integrity comes especially important to pay attention to Assessment Project finds that of the wells that were plugged in and after 2008, 11.6 percent of the likelihood that the hydrocarbon formation them leak (Boyer 2015). This concern was also that these wells penetrate becomes repressurized. Re-pressurization may occur due corroborated by an industry consultant (Kunz, to nearby gas drilling, completion, and well www.rff.org 10 Resources for the Future Ho et al. stimulation activities (Jackson et al. 2013). For instance, the injection of fluids at high pressure during hydraulic fracturing can pressurize nearby offset wells that have not been shut-in (Dusseault and Jackson 2014). The pressure from the injection of CO2 if a formation is used for CO2 storage also presents a similar risk (Watson and Bachu 2009). Alternatively, the buoyancy of the CO2 may itself cause CO2 leakage to the surface after it has been injected. Subsurface Geology Finally, the subsurface geology of the area around an inactive-plugged or inactiveunplugged well can influence the leakage potential of the well both by increasing the risk that a leakage pathway will develop and by influencing the pressure or fluid buoyancy gradient. Wet areas and hydrogen sulfide-bearing zones can all accelerate corrosion (King and King 2013). Salt zones may increase the risk that cement will be contaminated by salt and set prematurely, thus compromising the longterm integrity of the cement plug (NPC 2011). High-pressure areas may also increase the risk of fluid migration; King and King (2013) estimate that wells in these environments may have a life of a decade or less before permanent plugging and abandonment is required. Other Factors Finally, the ownership status of a well and its location relative to water resources and/or human population centers are correlated with or contribute to environmental risk. A well’s ownership status refers to whether it has a responsible operator on record. On average, orphaned wells are likely to have been drilled earlier than wells with an owner and are thus more likely to have lower-integrity well constructions and/or be in a deteriorated condition. In addition, operators may be willing and able to plug and abandon only the wells that are cheaper to plug, and may choose to leave the wells with higher plugging costs in a temporarily abandoned state or transfer these wells to smaller operators, who are more likely to de- fault on their bonds, resulting in orphaned wells. These wells that are more expensive to plug may also be the wells that are in the worst condition and thus more environmentally risky. The proximity of a well to human populations or groundwater supplies is also a crucial factor in determining the inactive wells that deserve closer attention and monitoring. Oil and gas states with well plugging programs generally have criteria for prioritizing wells to be plugged, including their location. The Kansas Corporation Commission (KCC), for example, prioritizes wells in a poor condition based on whether they are a threat to sensitive surface water or groundwater areas, and whether they are a threat to public safety in urban or suburban settings (KCC 2015). Empirical Estimates of Magnitude of Pollution Potential The basic leakage pathways that cause methane leakage or groundwater contamination from production wells, such as uncemented annuli or casing corrosion, are also responsible for pollution from inactive wells. The failure rate of oil and gas wells in general has been documented in empirical studies. A 1995 study by Westport Technology found that 15 percent of primary cement completions in the United States fail (Dusterhoft et al. 2002). In a more recent study, Ingraffea et al. (2014) use state monitoring records and report that 1.9 percent of the 32,678 oil and gas production wells drilled in Pennsylvania between 2000 and 2012 have some evidence of leakage and have been issued a Notice of Violation (NOV).5 In addition to failure rates of cement and casing, local Based on this, Ingraffea et al. conclude that these 1.9 percent of wells experienced a loss of structural integrity. However, this conclusion has received criticism for conflating being issued an NOV and experiencing a loss of structural integrity, see e.g. Brown (2014) for a discussion from an industry viewpoint. 5 www.rff.org 11 Resources for the Future Ho et al. instances of pollution from both producing and inactive wells have also been documented. Erno and Schmitz (1996) and Van Stempvoort et al. (1995) have measured gas leakage through surface casing vent flow and soil gas migration from oil and gas wells in the Lloydminster area of Alberta, with the latter documenting instances of groundwater contamination. Instances of pollution specifically from inactive, improperly plugged and abandoned wells also have been documented: Lyverse and Unthank (1988) document an incident of excess chloride discharge from abandoned exploration wells into a shallow aquifer near Fort Knox, Kentucky, and Chafin (1994) describes methane discharge into shallow groundwater from abandoned wells drilled in the 1930s in the San Juan basin in New Mexico and Colorado.6 Although these studies are useful for understanding the mechanisms through which methane leakage and groundwater contamination from inactive wells can occur, and for providing anecdotal evidence that such pollution does occur, they do not provide empirical estimates of the rate at which pollution occurs for inactive wells specifically. Of the empirical estimates that have been published, some are disputed. Thus, given the current state of the literature, it is difficult to estimate the scale of the problem of pollution from inactive wells, whether plugged or unplugged. Pennsylvania, Kang et al. estimate that methane emissions from orphaned wells may have been responsible for 4–7 percent of total anthropogenic methane emissions in the state during 2010, although they acknowledge that they cannot guarantee the representativeness of their samples. Furthermore, King and Valencia (2014) argue that this figure is likely to be an overestimate, as the sample wells are not likely to represent all abandoned or orphaned wells. Using data on methane emissions from 42 plugged and unplugged wells, Kang et al. (2015) estimate the effective permeabilities of these wells—that is, the wells’ potential to leak methane. The authors estimate the effect of plugging status (plugged or unplugged), geographical location, and well type (oil, gas, or combined oil and gas) on the permeability of a well. They find that the average effective permeability of unplugged wells is higher than that of plugged wells (although this difference is not statistically significant), that permeability of plugged wells is highly variable, and that the permeability of gas and combined oil and gas wells is higher than that of oil wells. Outside of the United States, we discussed the issue of leakage from temporarily abandoned wells with Michael Taylor, VicePresident for Climate Policy Assurance at the Alberta Energy Regulator. Data on temporarily abandoned wells in Alberta reveals that, in 2015, of 80,000 wells with this status, 5,000 To our knowledge, Kang et al. (2014) prowere reported by owners to be leaking methane vide the only U.S. estimates of methane emis(a rate of about six percent), with an average sions from inactive-unplugged wells. By meas- daily leakage rate of 13 cubic meters. The maxuring methane emissions from a sample of 19 imum observed leak rate was around 500 m3. orphaned wells and scaling the mean methane Such wells can legally remain in this state for flow rate at these wells to the estimated popula- up to ten years, so an average leaking well tion of 300,000–500,000 orphaned wells in could emit over this period 47,000 m3 before it returns to production or is permanently plugged. The age of these citations should be noted. Newer studies examining wells that have been more recently completed may find that these wells have a lower rate or risk of leakage. 6 Kell (2011) examines the groundwater contamination rate using data on reported groundwater contamination from oil and gas wells in Ohio and Texas. Over a period of 25 years www.rff.org 12 Resources for the Future Ho et al. from 1983 to 2007, 41 of 185 groundwater contamination occurrences (of a total of 65,000 wells) in Ohio were due to leakage from orphaned wells, whereas four were caused by reclamation. In Texas, 30 of 211 groundwater contamination occurrences (of a total of 250,000 wells) were caused by orphaned well leakage, and one was caused by reclamation. Of the 30 orphaned well leakage incidents, 28 were caused by the vertical migration of fluids through inadequately sealed boreholes. Most of these wells were characterized as “old” or “historic.” Number of Inactive Wells in 13 States In addition to reviewing the various environmental risks associated with inactive wells that are not properly decommissioned, another important aspect of understanding the aggregate environmental risk posed by this population of wells is estimating the number of such wells in the United States. Here, we estimate the number of wells that have stopped producing that have not yet been decommissioned, and calculate this as a percentage of the total number of inactive wells. In so doing, we develop, for the states sharing their data with us, an upper bound estimate of the number of wells that could potentially create the types of environmental damage described above.7 We contacted officials from various state oil and gas agencies, prioritizing states with significant oil and gas production as well as states with larger numbers of inactive wells. In total, we managed to obtain data from 13 states.8 Table 1 presents these results. Note that, as laid out in the report’s introduction, we use the term “inactive” to refer to all wells that have stopped producing. At present, the literature on the issue of inactive wells focuses on orphaned wells, particularly those that were drilled in an earlier regulatory era. We argue that this focus needs to expand to include all wells that have ceased production. Even wells with modern well constructions can fail; additionally, all the wells being drilled today have the potential to become orphaned in the future. “Inactive wells” as we define them here include shut-in wells,9 which states generally consider to be part of “active” wells, temporarily abandoned wells, and wells that have been decommissioned, which states generally classify as “plugged and abandoned” and not “inactive”. Although shut-in wells and temporarily abandoned wells are technically wells that have been demonstrated to have future use and do Using this state-by-state approach meant that we were not able to develop estimates of the number of inactive wells in all states across the United States. In order to do that, the proprietary database owned by DrillingInfo provides a starting point, as it contains data on all wells that have been drilled in the United States to date. For a few states, DrillingInfo’s data also has the advantage of being more comprehensive than the states’ own electronic databases, as DrillingInfo has digitized analog data on orphan wells. However, there is little consistency in the way states report data on well production and well statuses: production may be reported at the well level or the lease level, data is updated anywhere from twice a month to once a year, and the wells in some states lack specific well statuses (such as plugged and abandoned wells) and are only very coarsely categorized in DrillingInfo as either active or inactive. Thus, developing estimates of inactive well numbers that are both accurate and comprehensive requires working carefully with both DrillingInfo’s database and data provided by state agencies themselves. This was outside the scope of our work. 8 A shut-in well is a well that is temporarily plugged but capable of producing in the future. The well is secured, but easily re-opened. A well may be shut in due to poor market conditions, inadequate well maintenance and repairing, or lack of equipment to complete it, among other reasons. 9 Note that even wells that have been plugged and abandoned might still leak, although the science on this is not settled and this is an area for future research. 7 www.rff.org 13 Resources for the Future Ho et al. not have to be decommissioned at this time, we include them in our count of wells, as each of these wells has the potential to cause environmental damage if it is not eventually plugged, or not properly plugged. There is also reason to believe that some of them may not be consistently monitored by state regulators; that is, some of them may be in poor enough condition to require decommissioning, but are nonetheless allowed to remain in temporary abandonment status. Table 1 shows the proportion of inactive wells in each state that have not been permanently decommissioned, and therefore, the proportion of inactive wells that may create an environmental concern. It should be noted, however, that the current oil and natural gas price environment has resulted in more wells being shut-in; therefore, the number of inactive wells reported here may be higher than they would be under higher oil and gas prices. Across the 13 states, the population of inactive wells is as large as 557,000, 12 percent of which have not been decommissioned. The percentage in each state varies considerably, with Ohio reporting only 1 percent of inactive wells that have not been decommissioned, and Missouri reporting 56 percent. This should not be read as a measure of each state’s ability to decom- TABLE 1. TOTAL NUMBER OF INACTIVE WELLS IN EACH STATE10 State MO KY MT WV NY PA ND NM WY KS CO AR OH Total Total inactive wells 9,098 29,546 12,358 36,941 12,702 52,091 11,210 46,105 45,913 210,868 37,662 24,660 106,188 635,342 Inactive non-P&A* 5,111 12,338 4,652 14,018 1,730 6,895 1,341 4,773 3,981 15,465 1,881 948 1,178 74,311 Inactive P&A 3,987 17,208 7,706 22,923 10,972 45,196 9,869 37,076 41,932 195,403 35,781 23,712 105,010 556,775 Active wells 1,193 41,371 28,947 18,919 11,406 121,011 14,373 52,903 32,841 91,472 50,861 17,680 61,189 544,166 Inactive nonP&A wells as % of total inactive wells 56 42 38 38 14 13 12 10 9 7 5 4 1 12 Note: We use P&A—“plugged and abandoned”—here as a synonym for “decommissioned.” 10 Different states have different ways of categorizing wells. Inactive, non-P&A wells include various types of nonproducing wells that have not been plugged, including orphan wells, wells of various temporarily abandoned statuses, shut-in wells, and wells approved for plugging. In certain states, for instance, in West Virginia and Montana, production data is reported only in twelve-month cycles such that it was not possible to extract wells that have been shut-in for less than twelve months and include these in our count of inactive wells. For these states, therefore, the number of non-P&A wells reported here is an underestimate. Inactive, P&A wells include wells labelled plugged and abandoned, dry and abandoned, and final restoration. Active wells include all currently producing wells, and exclude wells that were never drilled or wells with expired or cancelled permits. The numbers reported in this table are based on data gathered in February and March 2016. www.rff.org 14 Resources for the Future Ho et al. TABLE 2. INACTIVE WELLS ON FEDERAL LAND VS. STATE LAND Inactive non-P&A wells on State federal land Inactive non-P&A wells on non-federal land Total # of inactive non-P&A wells Percentage of inactive non-P&A wells on nonfederal land (%) Percentage of land under federal ownership (%) NM 2513 1960 4473 43.8 44.4 ND 236 1105 1341 82.4 7.4 PA 195 6465 6660 97.1 2.4 KS 62 15403 15465 99.6 1.2 NY 4 1726 1730 99.8 0.9 -mission inactive wells in a timely manner, as the low percentage in states such as Ohio and Kansas also reflects the fact that many of the wells in these states were drilled a very long time ago and have since stopped producing and been decommissioned. The numbers simply illustrate the size of the population of wells presenting an environmental risk in these 13 states. Not all of the responsibility for managing inactive wells falls to the states. To understand how much of the burden of managing inactive wells is borne by individual states versus the federal government, we also examine the number of inactive wells on different types of land, including federal land 11 as one category and state/local government/private land as another (from now on grouped here as called “nonfederal land”). Wells located on non-federal lands are managed by the states. Federal land includes lands administrated by agencies, such as National Park Service, Fish and Wildlife Service, Bureau of Reclamation, Bureau of Land Management, Bureau of Indian Affairs, Forest Service, Department of Interior, Department of Agriculture, Army, Navy, Air Force, Maine Corps, Coast Guard, Corps of Engineers and Department of Defense. 11 We were able to obtain well status and location data for five states, including Kansas, North Dakota, New Mexico, New York and Pennsylvania. We merge well data from these five states with the Department of Interior’s Surface Management Agency (SMA) Geographic Information System (GIS) dataset to identify the number and proportion of inactive, non-P&A wells on federal and non-federal lands. Table 2 shows that New Mexico has 43.8 percent of non-P&A wells that are on state land, which is lower than the equivalent figure in the other four states, where more than 80 percent of non-P&A wells are on non-federal lands. This is unsurprising given that the percentage of land under federal ownership in New Mexico is the highest amongst the five states, at 44.4 percent. In the other four states, most of the land is owned by state and local governments, and private landowners. It is likely that the federal government has a relatively larger share of well plugging liabilities in Western states, which have a greater proportion of federal lands. Obtaining data and conducting this exercise for more Western states would help to verify this. www.rff.org 15 Resources for the Future Ho et al. 4. Regulations on Inactive Oil and Gas Wells Key Findings and Recommendations  The individual states and the Bureau of Land Management have very different approaches for regulating the management and decommissioning of inactive wells. This heterogeneity in regulations can be described in terms of their comprehensiveness (i.e., the number of regulatory elements they regulate) and their stringency (i.e., how strict their regulatory elements are).  Shih et al. (forthcoming) note that financial assurances are often inadequate to cover the costs of decommissioning an inactive well. We recommended that bonding amounts should vary according to the major factors influencing costs, such as well depth. In this section (Table 5), we report that many states already do this, to varying extents. We therefore recommend that other states consider this approach to bonding and that all states examine our statistical results for insights into the specifics of how various factors affect costs. Given adequate data, our statistical method may even be used by the states to design bonding requirements that vary with cost factors.  States deal with temporary abandonment in a variety of ways, some more protective of the environment than others. For states that are less protective, shortening the time a well can be temporarily abandoned and raising the bar for proving a well should stay in that condition would help reduce the likelihood that inactive wells will create environmental externalities.  Few operators properly mark decommissioned wells with a permanent sign under the current regulatory regime. We therefore recommend that states adopt more stringent regulations for marking decommissioned wells. State oil and gas agencies regulate a range of industry activities related to the management and decommissioning of inactive wells. Regulatory elements include requirements for operators to post financial assurances intended to cover decommissioning costs and potential environmental damages, and administrative and technical procedures for temporarily abandoning or decommissioning a well. The BLM sets regulations that govern wells undergoing decommissioning on federal lands. As with earlier research by Resources for the Future (Richardson et al. 2013), this section compares regulations across states and the BLM and, where appropriate, compares the stringency of a selection of these regulations, which provides an indication of regulatory heterogeneity across these regulatory bodies. Methodology We examine 31 regulatory elements across 22 states (see Map 1 below). We chose this sample of states by considering three criteria: 1. number of orphaned wells that are on a state’s “wait list” for decommissioning, as reported by the Interstate Oil and Gas Compact Commission (IOGCC 2008); 2. a state’s historical crude oil production from 1981 to 2014, as reported by the US Energy Information Administration (EIA 2015a); and, 3. a state’s historical onshore natural gas production from 1992 to 2014 as reported by the US Energy Information Administration (EIA 2015b). If a state contributes to more than 1 percent of the national total in any of these three criteria, we include it in our sample. Appendices A1 and A2 provide a detailed description of our selection process. www.rff.org 16 Resources for the Future Ho et al. MAP 1. STATES INCLUDED IN OUR SURVEY OF INACTIVE WELL REGULATIONS wells, geothermal wells, coalbed methane In the next section, we describe various wells, and ratholes. regulatory elements and their importance for mitigating environmental risk. We focus only Review and Comparison of State and on state regulations, although we recognize that Federal Regulations permits and field adjustments also play a part The 22 states we examine and the 31 reguin regulating some of these activities. However, latory elements we consider are displayed in such variables are difficult, if not impossible, Map 1 and Table 3 respectively. The 17 regulato capture across the states. We also do not tory elements in the first panel of Table 3 are comment on the quality of monitoring and endirectly relevant to mitigating environmental forcement in different jurisdictions. Two states impact and are therefore included in our strinmay have identical regulations for a given elegency calculations. Following is an explanation ment, but one state’s enforcement might be of how these regulations may partially determore stringent than another’s. Thus, we can mine the degree of environmental and financial only observe regulatory stringency and not efrisk that the public may be exposed to: fective stringency. In addition, we do not 1. The more accurately bond amounts comment on what is “optimal” stringency or reflect decommissioning costs and the what is appropriate versus unjustified heteromore likely that states will be able to geneity across the regulatory bodies. Instead, recover costs, then the more likely that we describe regulations and note the ways in operators will decommission their which the stringency may differ across some of wells on schedule and/or states will these elements. Finally, we do not address cerhave the necessary funds to plug tain decommissioning processes for unique orphaned wells (regulatory elements types of wells, including underground injection 1–5). wells, seismic exploration www.rff.org 17 Resources for the Future Ho et al. 2. The easier it is for operators to idle and to apply and re-apply for temporary abandonment status for their wells, the more likely it is that wells will be left in an idle or temporarily abandoned. status indefinitely and therefore avoid proper decommissioning (regulatory elements 6–10). 3. More stringent requirements for temporarily abandoned wells help minimize the environmental harm caused by these wells (regulatory elements 11 and 12). 4. More stringent regulations on the procedures to be taken during plugging and restoration help minimize the environmental harm caused by decommissioned wells and well sites (regulatory elements 13, 15, and 16). 5. More stringent requirements for marking decommissioned wells and for reporting inactive wells help regulators identify wells that may cause environmental harm (regulatory elements 14 and 17). The remaining 14 regulatory elements are not included in stringency analysis, either because they are not as relevant to environmental impact or are not easily comparable across states. Figure 4 compares the number of regulatory elements that each state (and BLM) explicitly regulates. Figure 5 rates the stringency of state (and BLM) regulations based on regulatory elements that are quantitative in nature, and Figure 6 does the same for regulatory elements that are qualitative in nature. Figure 4 displays the number of elements regulated by each state, indicating the comprehensiveness of each state’s regulations. If a state has explicit regulations for a regulatory element it receives a 1 and if it does not it receives a 0. Regulations for BLM also appear on the figure. As displayed in Figure 4, New York regulates the fewest regulatory elements (10 out of 17) whereas Pennsylvania regulates the most (all 17). Note that the stringency of regulations is not reflected in this figure. Figure 5 displays states (and BLM) ranked by the average stringency of the five quantitative regulatory elements we consider. These five elements include (1) minimum individual bond amounts, in dollars; (2) minimum blanket bond amounts, in dollars; (3) well idle time, in months; (4) duration of temporary abandonment, in months; and (5) timing of restoration requirements, in months. In this figure, each regulatory element is normalized such that the least and most stringent regulations receive a score of 0 and 100, respectively. Then scores are averaged with equal weights across the five elements. We find that Alaska ranks at the top according to our five quantitative elements, with Arkansas having the least stringent regulations for these elements, about two-thirds less stringent than Alaska. No state is superior to all other states on all five elements. Figure 6 displays states (and BLM) ranked by their stringency, as calculated using 12 of the 17 regulatory elements we consider that are binary and qualitative in nature. These include (1) type of financial assurances; (2) well characteristics that determine bonding amounts; (3) operator characteristics that determine bonding amounts; (4) permitting extensions for temporary abandonment; (5) whether notification, reporting, and inspection for temporary abandonment is required; (6) whether economic viability plays a role in determining status of temporary abandonment; (7) shut-in requirements for temporary abandonment; (8) whether well integrity testing for temporary abandonment is required; (9) the types of plugs required during decommissioning of a well; (10) whether marking of decommissioned wells is required; (11) whether restoration requirements are stringent; and (12) whether reporting is required for inactive wells. States (and BLM) with a regulation we judge to be stringent get a “1”: otherwise they get a “0.” So the highest score possible is 12. Pennsylvania leads the pack with a score of 11, while Kansas, Louisiwww.rff.org 18 Resources for the Future Ho et al. ana, and New York are in last place. Note that these calculations make no adjustment for regulatory elements unregulated by a state. TABLE 3. REGULATORY ELEMENTS OF INACTIVE OIL AND GAS WELLS Number Regulatory Element Panel A: Regulatory Elements Considered in Stringency Calculations 1 Types of Financial Assurances (Qualitative) 2 Well Characteristics that Determine Bonding Amounts (Qualitative) 3 Operator Characteristics that Determine Bonding Amounts (Qualitative) 4 Minimum Individual Bond Amounts (Quantitative) 5 Minimum Blanket Bond Amounts (Quantitative) 6 Well Idle Time (Quantitative) 7 Duration of Temporary Abandonment (Quantitative) 8 Extensions for Temporary Abandonment (Qualitative) 9 Notification, Approval, and Inspection for Temporary Abandonment (Qualitative) 10 Role of Economic Viability in Determining Status of Temporary Abandonment (Qualitative) 11 Shut-in Requirements for Temporary Abandonment (Qualitative) 12 Well Integrity Testing for Temporary Abandonment (Qualitative) 13 Types of Plugs Required During Decommissioning of Well (Qualitative) 14 Marking of Decommissioned Wells (Qualitative) 15 Stringency of Restoration Requirements (Qualitative) 16 Timing of Restoration Requirements (Quantitative) 17 Reporting Requirements for Inactive Wells (Qualitative) Panel B: Regulatory Elements Not Considered in Stringency Calculations 18 Separate Bond for Site Reclamation 19 Surface Damage Agreements 20 Statute of Limitations 21 Liens on Equipment 22 Well Integrity Testing 23 Treatment of Wells with Different Casings 24 Treatment of Casing Removal 25 Treatment of Different Well Types 26 Cement Specifications for Plugs 27 Conversion to Freshwater Well 28 Notification, Approval, and Inspection for Decommissioned Wells 29 Ability for Regulator to Order Plugging and Replugging 30 Reporting Requirements for Inactive Wells 31 Considerations for Fugitive Methane www.rff.org 19 Resources for the Future Ho et al. FIGURE 4. NUMBER OF ELEMENTS REGULATED BY EACH STATE (AND BLM) 16 14 12 10 8 6 4 2 NY KS LA TX KY CA ND AR NE MT IL UT OK MO WV OH NM WY AK IN CO BLM PA 0 FIGURE 5. STRINGENCY BY STATE (AND BLM) ACCORDING TO QUANTITATIVE REGULATORY ELEMENTS 80 70 60 50 40 30 20 10 AR UT IL PA KY NY KS WY MO MT NM LA OH WV CA CO NE OK TX IN BLM ND AK 0 www.rff.org 20 Resources for the Future Ho et al. FIGURE 6. STRINGENCY BY STATE ACCORDING TO QUALITATIVE REGULATORY ELEMENTS 12 10 8 6 4 2 Discussion of Regulatory Elements Bonding Requirements An operator must post financial assurance for a well at the time it is drilled. States recover this financial assurance to cover the costs of decommissioning the well in the event that the operator is unable to do so. States vary in the types of financial assurance they accept and the amount they require. Types of Financial Assurances States allow operators to use a range of instruments for financial assurance, as displayed in Table 4. All states allow a surety bond, which involves a third party company that essentially issues and prices the bond. Other popular types of financial assurance include letters of credit, certificates of deposit, and cash. A handful of states also allow escrows, trust accounts, financial statements, and liens to serve as financial assurance. Many of these types of financial assurances come with a variety of stipulations (e.g., irrevocable, automatically renewable, whether interest on deposits accrues NY LA KS OK ND TX KY CA AR NE MT UT IN IL AK OH WY MO WV BLM NM CO PA 0 to operator or state) that may provide better fiscal protection for the states, however there appears to be a lack of analysis on this question. States do not typically require operators to choose a particular type of financial assurance and instead allow operators to choose from a range of options. The BLM allows operators to use surety bonds, letters of credit, negotiable Treasury securities, and cash in the forms of certified or cashier’s checks. The form and amount of financial assurance at least partially determines the likelihood that the regulator will receive the appropriate funds for decommissioning in the event that an operator does not do so. Without sufficient funds, a regulator is less likely to have the financial means to decommission the wells that require it. Consequently, a well either will not be decommissioned or will stay in a status that is more likely to cause environmental harm for a longer period of time. One way to distinguish between strong and weak financial assurances is to consider the mechanism through which the regulator would www.rff.org 21 Resources for the Future Ho et al. TABLE 4. TYPES OF FINANCIAL ASSURANCES ACCEPTED BY STATES (AND BLM) State Surety Bond Letter of Credit Certificate of Deposit AK AR CA CO IL IN KS KY LA MO MT NE NM NY ND OH OK PA TX UT WV WY BLM X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Cash X X X Escrow or Trust Account Financial Statement X X X Govt. Bond Annual Fees X X X X X* X X X X X Lien X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X *This lien amount has limits. receive funds. Cash, for example, guarantees that the state has funds upfront and is therefore quite a strong form of financial assurance (leaving aside the issue of whether the amount of cash is adequate). Other strong forms of financial assurance provide some form of guarantee by a third party that the funds will be allocated to the government in the event of a default and include surety bonds, letters of credit and perhaps escrow or trust accounts. A weaker form is liens, which allow the regulator to collect operator property if they do not pay; but such collections require legal operation, so are costly and may not be fully successful. Another weak form of financial assurance is financial statements, which require that operators provide proof of the financial health of their company—typically up to a set amount. Annual fees represent a special case. These effectively deliver cash to regulators on an annual basis, but the state that allows for them under certain circumstances (i.e., Kansas) sets the fee so low that we count the category as a weak form of financial assurance. We use a binary and qualitative assessment when incorporating type of financial assurance into our stringency calculations for states. A state that allows for financial statements, liens, or annual fees receives a 0 and a state that does not allow for these types of weak financial assurances receive a 1. We do not consider the range of financial assurance types in our stringency calculation because most, if not all, states allow operators to choose between allowed financial assurances. Operators are therefore free to choose the type of allowed financial assurance they view as least stringent www.rff.org 22 Resources for the Future Ho et al. Amount of Financial Assurance Operators choose between individual or blanket bonds when posting financial assurance. The former type of financial assurance covers a single well, whereas blanket blonds cover multiple wells. This financial assurance is intended to cover the expected costs of decommissioning a well; yet, in practice, financial assurance amounts are often insufficient for this purpose, as discussed in Shih et al. (forthcoming) and existing literature (GAO 2011; LLA 2014). All else equal, a higher bond amount provides a more certain guarantee that wells will be properly decommissioned or that the state will have adequate financial resources to plug a well. Table 5 shows that some states tailor bonding amounts based on well characteristics (depth, type, and location of wells) and operator characteristics (number of wells, number of inactive wells, and compliance history). As noted in Shih et al. (forthcoming), differentiating bond amounts based on the most important factors affecting decommissioning costs would help ensure that bonds, or other financial assurance requirements, more accurately reflect cost. Of these factors listed, well depth in particular has been known to strongly correlate TABLE 5. FACTORS DETERMINING INDIVIDUAL AND BLANKET BOND AMOUNTS BY STATE State AK AR CA CO IL IN KS KY LA MO MT NE NM NY ND OH OK PA TX UT WV WY BLM Well Characteristics Depth Type of Wells Location of Wells X* X X X X X X X X X Operator Characteristics Number of Number of Wells Inactive Wells Compliance History X X X X X* X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X *Fee versus bond depending on well type (gas/oil) www.rff.org 23 Resources for the Future Ho et al. with cost due to the amount of plugging material and equipment rental time required. We find this effect in our statistical analysis, described in Shih et al. (forthcoming). Further, states have readily available information on well depth, which may help explain why most states use well depth to at least partially determine bond amounts. Calibrating bond amounts by well depth is important, especially as average well depths in the United States have been increasing. Besides well depth, however, there are also several other factors that may influence cost and that could also be taken into consideration when setting bond amounts (Davis 2015). Four states (Arkansas, Kentucky, Pennsylvania, and West Virginia12) assign higher amounts to certain types of wells (e.g., horizontal) although it is not well understood whether and why unconventional wells may cost more to decommission than conventional wells.13 The BLM, Louisiana, and New Mexico assign higher amounts to wells located in certain regions. This could help to capture the effects of spatial variables on cost, such as proximity to groundwater aquifers, the concentration of coal seams in a play, and the variation in prices charged by service providers operating under different market conditions. Three operator characteristics play a role in determining bond amounts. First, in some states, the larger the number of wells owned by an operator, the smaller the amount of the individual well bond. Most states also permit operators that own many wells in the state to post a 12 Ar. Rule B-2.h; KRS 353.590(9); 58 Pa. Code §3225(a)(1, 2); WVC §22-6A-15. 13 On the one hand, unconventional wells are typically deeper than conventional wells, but on the other hand, plugging of multiple wells can occur on the same pad for unconventional wells such that decommissioning costs may be lower due to economies of scale. single blanket bond covering some, or all, of their wells. On the one hand, this makes sense as firms with many wells are larger, tend to have better access to decommissioning technologies, and are less likely to become insolvent. On the other hand, the price per well in tiered blanket bonds tends to go down quite significantly as the number of wells increases, offering a significant price discount to the operator. While this may help firms pool their risk, it also lowers financial coverage for the state and could leave it especially exposed in certain circumstances (e.g., a large concentrated investment by a small number of firms into a play or resource that goes bust, similar to what Wyoming has experienced with coal bed methane). One benefit of offering blanket bonds from the regulator’s perspective is lower administrative costs to monitor well transfers and bond status. Many states also use a regulator’s compliance history and number of inactive wells to inform bond amounts, given that past performance may be associated with future performance. Regulators, at their discretion, may require operators with poor compliance histories to post higher bond amounts than the standard prescribed or may even prevent operators from posting new bonds or adding wells to a bond. Requiring higher bond amounts for operators with poor compliance histories, or those with a large number of inactive wells that may have an increased risk of being orphaned, helps ensure that the public does not eventually have to bear the environmental or financial burden left behind by irresponsible or bankrupt operators. The BLM is allowed to require additional bonding based on well characteristics (i.e., location, depth, age, production capability of the associated field, and unique environmental issues) as well as operator characteristics including number of wells. We use two binary and qualitative assessments when incorporating factors that determine bond amounts into our stringency calcula- www.rff.org 24 Resources for the Future Ho et al. tion for states. Many states account for well and operator characteristics in an effort to match bond amounts to their conception of costs (i.e., decommissioning costs and costs of potential environmental damages). These states will, all else being equal, more accurately estimate costs of wells that become orphaned or create environmental damages before they are decommissioned (and the states therefore will more often have sufficient resources to deal those costs and damages). Our first stringency assessment assigns a 1 to states that account for well characteristics when determining the monetary value of individual or blanket bonds and a 0 to other states. Our second stringency assessment assigns states that use operator characteristics when setting bond amounts a 1, while other states receive a 0. We recognize that using these factors does not directly mean that the bond amounts are higher than in states that do not, although this appears to be the case in practice. Map 2 displays the lowest possible bond amounts that states require operators to post for a single well. These amounts are typically denoted in dollar-per-well terms, and values among these states range from $500 per well in Kentucky to $100,000 per well in Alaska. Some states utilize other approaches: four states (Kansas, Louisiana, Texas, and Wyoming) calculate bond amounts in terms of dollars per foot of well depth, whereas New Mexico combines these approaches.14 States that denote bond amounts in terms of dollars per well may further differentiate based on well depth, number of wells, and type of well (e.g., vertical or horizontal, inland versus coastal). These differentiations lead to multiple potential bond values for a well, and we therefore choose to display the lowest possible bond value in Map 2.15 We use a continuous and quantitative assessment incorporating lowest possible individual bond amounts into our calculation of stringency across states. This assessment sets the lowest bond amount across states equal to 0 and the highest equal to 1, normalizing all values in between. Map 3 displays the lowest possible blanket bond amounts that states require operators to post for multiple wells. The amounts are all denoted in terms of dollars, and values range from $5,000 in Kansas to $200,000 in California for certain types of blanket bonding situations.16 States differentiate blanket bond amounts based on all aforementioned well and operator characteristics.17 Similarly with Map 2, these differentiations lead to multiple potential bond values for a set of wells, and we therefore chose to display the minimum bond value in the map below. The BLM requires operators to post $25,000 to cover wells within a single state and $150,000 to cover all wells across the nation. We use a continuous and 15 For example, Illinois requires $1,500 in financial assurances for wells less than 2,000 feet and $3,000 for wells deeper than 2,000 feet. We therefore choose $1,500 dollars for our analysis, because it is the lowest possible value. 16 14 Note that annual well fees can be paid in lieu of bonds in Illinois and Kansas under certain circumstances. Specifically, bonds in Illinois are only required for certain operators (those in operation after 1991 and with poor compliance histories) (62 IAR I.240.240.15000(a). In Kansas, regulators allow operators to pay three percent of the amount that would be paid under an individual or blanket bond as a nonrefundable fee, in lieu of a bond (KAR 82.3.120.g). California has different blanket bond amounts based on number of wells and whether the wells are also covered by an idle well fee. The relevant amounts are: $200,000 if there is no idle well fee and the operator has 20–50 wells; $400,000 if there is no idle well fee and 50+ wells; and, $2,000,000 with an idle well. 17 For example, West Virginia has two different blanket bond amounts: $50,000 for conventional and $250,000 for horizontal wells. We therefore chose $50,000 because it is the lowest possible value. www.rff.org 25 Resources for the Future Ho et al. quantitative assessment incorporating lowest possible blanket bond amounts into our calculation of stringency across states. This assess- ment sets the lowest blanket bond amount across states equal to 0 and the highest equal to 1, normalizing all values in between. MAP 2. MINIMUM BONDING REQUIREMENTS FOR INDIVIDUAL BONDS MAP 3. MINIMUM BONDING REQUIREMENTS FOR BLANKET BONDS www.rff.org 26 Resources for the Future Ho et al. Treatment of Site Reclamation in Bond Amount States generally expect financial assurances to cover all stages of well decommissioning, including site restoration. Although separate plugging and site restoration bonds may result in higher aggregate levels of financial assurance, the motivation for separating the two is not immediately clear. Four possible explanations are (1) that site restoration costs have greater heterogeneity and/or are less well understood at the outset, so regulators want to contain this variation in a separate instrument, and (2) that the time between plugging and restoration is prone to either long gaps in which risk of orphaning the well site is large, (3) that contracting for site restoration has a different supply curve than plugging and other well service contractors, and (4) that regulators want to establish a distinct bond forfeiture and return process for site restoration in addition to plugging. This latter explanation could allow for a different process or set of parties (e.g., surface owners) to be involved in bond forfeiture, or it could create an incentive structure that creates greater decommissioning compliance but worse site restoration compliance. A final thought is that because multiple wells are often on a single well site, separate site restoration bonds might better reflect the work flow of plugging different wells at different times, and then returning to site restoration at the end. Surface Damage Agreements One possible reason for having separate site restoration bonds listed above—establishing a different process for bond forfeiture or return that involves different parties—was observed in one novel policy arrangement we encountered during our review of state regulations: the use of surface damage agreements. These surface damage agreements are intended to provide some form of accountability to surface owners in cases where the surface and mineral estate may be severed. Seven states use surface damage agreements, three of which require up- front deposit amounts. In addition to the seven, Kentucky requires a surface owner agreement to the operator’s reclamation plan and details a mediation process if the surface owner does not agree, and Ohio requires liability insurance coverage for property damage.18 This is an interesting approach regarding the local and distributional impacts of oil and gas development. In situations where the surface and mineral estates are owned by the same party, surface use and damage can be covered in lease provisions and other contractual arrangements. In split estate cases, however, protection for surface owners is not given special consideration. Outside of the environmental protection covered in regulations, the only recourse available to surface owners may be post-facto litigation. The surface damage agreements we found in our review generally feature some sort of negotiation between the operator and surface owner prior to drilling, and in some cases the agreement is required as part of the permit application. Although the type of damages covered is not expounded on in great detail, frequent mention was made for crop loss or loss of other surface use. Whether the agreements cover non-market values (such as recreational uses and noise) is unknown. The degree to which surface damage agreements increases monitoring, verification, and enforcement by either the regulator or through the surface owner is unknown but is a promising area of further research. Statute of Limitations The transfer of a well from one operator to another serves as a junction point of liability as the new operator submits new financial assurance or assumes responsibility for existing financial insurance. While states generally have stipulations on notice to be provided to the regulator and any involved financial parties at the 18 805 KAR 1:170(2.3, 3, 4); ORC 1509.07(A) www.rff.org 27 Resources for the Future Ho et al. time of this transfer, some stipulate special liability protection, extended periods for which original or previous well owners may continue to be held responsible for plugging, or provisions establishing how far back through ownership history states may go to find financially able previous operators to defray plugging costs. We can think of these extended liability or statute of limitations provisions as covering two basic scenarios—ones in which the new owner is noncompliant with plugging orders or becomes financially insolvent, and ones in which the previous owner had poor operating practices or well problems that only became apparent after a time delay. While we note the appearance of such language in the statutes and rules we reviewed, such liability provisions are likely also covered contractually between parties outside of systematic regulation by the state. However, the regulations we reviewed typically featured strong discussion of bond transfer, forfeiture, and release. Protections to the state generally took the form of ensuring that the regulators are notified prior to transfer and that the new operator has posted a new (or adopted the previous) financial assurance. Some states (such as Indiana) specify in greater detail that states may deny bond transfer if the new operator has a bad credit or operating history or for other reasons. Extended liability provisions may be highly beneficial in some cases, particularly for wells of an older vintage. Older wells, especially those drilled before modern drilling and casing standards were implemented, pose a greater environmental risk and can result in greater plugging costs. Additionally, older wells are more likely to have been transferred to other operators, and so measures that ensure that these operators have some form of financial liability (as in liability provisions of the socalled Superfund law) may significantly defray fiscal costs for the state. The number of states that stipulate extended liability in regulations is small, although as noted, these provisions may exist outside of specific regulations. Liens on Equipment Some states establish in regulations a lien on oil and gas site equipment or resources, and infrequently such a lien may be used as financial assurance. Some states also specify the process by which they bid out state plugging contracts and how salvage value of any equipment (including casing) may factor into such payment. Use of liens or salvage value is assessed as a weak financial assurance in our review. Additionally, some lien policies allow outside parties (particularly nearby surface owners) to enter onto an orphaned or noncompliant well site and plug a well and reclaim any salvage value. Some states require salvage operators to post their own financial assurance, presumably due to environmental risk that may result from casing removal. Temporary Abandonment Well Idle Time An idle well is one that is not currently producing oil or gas. Wells are not generally permitted to remain idle indefinitely. Instead, after a certain period of time (which we refer to as “well idle time”), operators have a choice: they can start producing again, temporarily abandon the well, or decommission it. We hypothesize that the longer a well remains idle but not properly decommissioned, the greater the odds that the well imposes environmental externalities. Of the 22 states in our survey, 19 impose limits on well idle time. Map 4 displays these values, which range from 1 month on BLM lands to up to 24 months in Arkansas and Ohio. The map masks at least two complexities. First, several states differentiate well idle times based on certain categories of wells—especially those that are uncased, dry, or non-commercial. Dry and uncased wells in particular often have a substantially shorter www.rff.org 28 Resources for the Future Ho et al. MAP 4. MAXIMUM WELL IDLE TIME (IN MONTHS) well idle time (e.g., those in Arkansas, Illinois, and Louisiana).19 Second, two states (Louisiana and Alaska) allow for operators to apply for extensions—granting regulators a significant degree of discretion over effective well idling time.20 We use a continuous and quantitative assessment when incorporating maximum well idle time into our stringency calculation for states. This assessment sets the longest maximum well idle time across states equal to 0 and the shortest equal to 1, normalizing all values in between. 19 AR Rule B-7.c; 62 IAC I.240.240.1120; LAC 43:XIX§137.A.3.a 20 20 AAC 25.115. Duration of Temporary Abandonment When a well no longer produces at an economical rate, an operator may choose to stop production but not to immediately decommission the well (Richardson et al. 2013). This well status is called temporary abandonment and most states we survey allow wells to achieve this status, which essentially allows them to remain idle but—in many cases— requires operators to take various measures to reduce the risk of that well imposing environmental externalities (as discussed in later sections). The prospect that a well may again become active (e.g., if oil or gas prices rise) is an important motivation for states to allow for temporary abandonment, as it is more costly for a well to become reactivated after decommissioning. At least one study, however, shows that operators can use temporary abandonment to simply avoid decommissioning costs even if the wells have very low future economic potential (Muehlenbachs 2015). We again hypothesize that the longer a well is not www.rff.org 29 Resources for the Future Ho et al. MAP 5. DURATION OF TEMPORARY ABANDONMENT (IN MONTHS) decommissioned, the greater the odds that it imposes environmental damages. All of the states we surveyed regulate the duration of temporary abandonment. Map 5 displays these durations, which range from 6 months in Colorado and Texas to 300 months in California. Of the 22 states that regulate the duration of temporary abandonment status, all but New Mexico explicitly allow for some form of extension.21 The extensions granted by these 21 states can be categorized either as unlimited or limited. About two-thirds of these states do not explicitly limit the number or duration of extensions that an operator could receive for a well to stay in temporary abandonment; the majority of these states include some kind of regulator discretion in approving extensions. Some states (e.g., Louisiana, Missouri, Montana, Nebraska, New York) allow for ex- tensions seemingly without explicit regulator discretion.22 The other one-third of the states and the BLM include explicit limits on the ability of regulators to authorize extensions, including:23  Arkansas: wells that have been idled for over 10 years are not eligible for extension.  Kansas: wells that have been shut in for over 10 years are not eligible for extension.  Kentucky: operators can apply for one extension that lasts two years.  North Dakota: operators can apply for one extension that lasts a single year. 22 LAC 43:XIX§137.A.2; 10 CSR 50-2.040(5); ARM 36.22.1240, ARM 36.22.1303; NAC Title 267 Chapter 3 040.01; 6 CCR-NY 555.3.b; 6 CCR-NY 555.2.a. 23 21 NMC §19.15.25.12. AR Rule B-5.h; 82 KAR 82.3.11.b; 805 KAR 1:060(1); 43 NDAR 43-02-03-55; ORC 1509.062.E, F. www.rff.org 30 Resources for the Future Ho et al.   Ohio: after three renewals of temporary abandonment status, the regulator may require a surety bond no greater than $10,000 for each of the owner’s wells that has approved temporary abandonment status. BLM: wells can be temporarily abandoned for 12 months; operators are allowed a limited extension that cannot exceed 12 months some form of inspection.24 The BLM and remaining states with a formalized temporary abandonment process require some form of approval from the regulator. In our view, this approval process is more stringent than notification and less stringent than inspection. Requirements for Attaining Temporary Abandonment Status The link between the level of requirement for attaining temporary abandonment and subsequent environmental or fiscal risk is not direct, but is an important moment for regulatory monitoring. Because temporary abandonment periods can extend for significant time during which operator and regulator monitoring of the site may decline, and because the environmental risk posed by a well increases the longer it is idle, ensuring that the well is in good condition prior to temporary abandonment can prevent serious hazards in the future. To incorporate this linkage between the approval process and environmental risk, we use a binary and qualitative assessment into our stringency calculations. States that do not have explicit regulations or only require notification receive a 0, whereas states that require approval or inspection receive a 1. State regulators may impose three categories of requirements that operators must achieve before gaining temporary abandonment status: notification, approval, and/or inspection. We characterize notification (i.e., requiring the operator to notify the regulator of temporary abandonment status) as the least stringent and inspection (i.e., requiring the operator to receive a positive confirmation from a government inspector that the well is eligible for temporary abandonment status) as the most stringent. As displayed in Map 6, four states (Indiana, Louisiana, Montana, and Wyoming) only require notification, whereas three states (Kansas, Oklahoma, and Pennsylvania) require Of the 22 states we survey, 12 contain provisions that require operators to show some future usefulness of wells that are temporarily abandoned before they are granted an extension (as displayed in Map 7). These provisions likely exist, at least in part, to protect against wells remaining in a status of temporary abandonment only for operators to avoid decommissioning costs and without any intention of returning the wells to active status. We view these regulations as important for limiting the chance of environmental impacts occurring, because the regulations help limit the amount of time an operator can delay decommissioning. However, we would need to We use a continuous and quantitative assessment to incorporate the duration of temporary abandonment into our stringency calculation for states. This assessment sets the longest duration equal to 0 and the shortest duration equal to 1, normalizing all values in between. We also use a binary and qualitative assessment to incorporate extensions for temporary abandonment into our stringency calculation for states. This assessment assigns a 0 to states that allow for extensions (either limited or unlimited) and a 1 for states that do not. 24 IN TR Section 6, 312 IAC 16-5-20.b; WY Rule 3.16.a; 82 KAR 82.3.11.b-c OAC 165:10-11-9; 58 Pa. Code §3214(a). www.rff.org 31 Resources for the Future Ho et al. MAP 6. NOTIFICATION, APPROVAL, AND INSPECTION REQUIREMENTS FOR TEMPORARILY ABANDONED WELLS MAP 7. PRESENCE OF A REQUIREMENT FOR OPERATORS TO SHOW FUTURE USEFULNESS OF TEMPORARILY ABANDONED WELLS www.rff.org 32 Resources for the Future Ho et al. review actual operator reports to assess the rigor of statements of future usefulness, but that is outside the scope of this research.25 The strength of these provisions varies widely by state, as displayed in the bullets below. Some states contain a generic determination from the regulator that the well has future usefulness, or they require a plan that may include an estimate of when a well will return to active status. Texas is a stringent outlier. Some examples include:    Alaska: the request for operation shutdown must provide a full justification, including a description of the proposed condition of the wellbore, approximate date when drilling will resume, and a proposed program for securing the well during shutdown.26  Colorado: usefulness must be shown annually during temporary abandonment status and when a request for extension of temporary abandonment status is submitted. 27  Louisiana and West Virginia: a well must be classified as having future utility.28  Missouri, Nebraska, and New York: a well must be determined to have “good cause shown” or “sufficient good cause” 25 Here and elsewhere throughout the report we recommend increased monitoring, reporting, and verification efforts by regulators as a way to detect and respond to wells before large damage is caused. However, we recognize that states may not be sufficiently resourced to provide this extra involvement. Given this, extra regulator involvement at the time of temporary abandonment approval may be especially beneficial by preventing problematic wells from becoming inactive. 26 27 28   for receiving temporarily abandoned status.29 Montana: the operator must provide a report describing the operator’s plan and time frame for returning to active status, plugging, or converting the well to other purposes.30 Ohio: a well must demonstrate future utility, and the operator has to have a viable plan to utilize the well within a reasonable period of time.31 Pennsylvania: an operator must present a plan for using the well within a reasonable period of time.32 Texas: a licensed geoscientist or petroleum engineer must certify that a well has future utility. That certification must include, among other things, a cost calculation for decommissioning the well and a determination that the expert reasonably expects the well to have future economic value in excess of decommission costs.33 We use a binary and qualitative assessment when incorporating whether regulators consider economic viability in granting temporary abandonment into our stringency calculation for states. A state that does not consider economic viability receives a 0, and a state that does so (either via a general clause or through more prescriptive requirements) receive a 1. Well Closure and Shut-in Requirements for Temporary Abandonment Out of the 22 states we survey, only 12 require operators to place a temporary plug or 29 10 CSR 50-2.040(5); NAC Title 267 Chapter 3 040.01; 6 CCR-NY 555.3.b. 30 ARM 36.22.1240; ARM 36.22.1303. 20 AAC 25.110.a.2. 31 ORC 1509.062.B.5 2 CCR-1-319.b.1. 32 25 Pa. Code §78.102(4). LAC 43:XIX§137.A.3.b; WVC §22-6-19. 33 16 TAR 1.3.3.15.j. www.rff.org 33 Resources for the Future Ho et al. MAP 8. SHUT-IN REQUIREMENTS DURING TEMPORARY ABANDONMENT otherwise prescribe well closure in temporary abandonment status (Map 8). These 12 states require that operators cap the surface of the well (which would likely prevent air pollution emissions), place plugs in the well (which would help prevent water pollution) or both. Nearly all states require capping and most require plugging, with at least six states requiring both. Notably, Louisiana requires the same plugging requirements for temporary abandonment as it does for decommissioning a well—with the exception of installing a surface plug, a seemingly stringent regulation.34 The states that do not require capping or plugging we judge to be at a higher risk for temporarily abandoned wells to cause environmental externalities. 34 The BLM rules contain a general clause that operators must isolate perforations in an acceptable manner, but do not offer any explicit requirements. We use a binary and qualitative assessment when incorporating shut-in requirements into our stringency calculations for states. A state that imposes any such requirements (i.e., plug, cap, or both) receives a 1 while those that do not receive a 0. Well Integrity Testing Out of the 22 states we survey, 18 require well integrity tests before or during the period that a well has attained temporary abandonment status, and we could not find evidence that the remaining states in our survey impose LAC 43:XIX§137.H. www.rff.org 34 Resources for the Future Ho et al. information. The lack of easily trackable regusuch requirements, although it may be incorporated into notification, approval, and inspection latory elements on the specific plugging process may suggest that states approach each well requirements. States may require testing either prior to entering temporary abandonment status on a case-by-case basis. Despite these difficulties, we can identify a handful of regulatory or at specified intervals during temporary abandonment. At least four states require annu- elements pertaining to plugging requirements that may decrease the likelihood of a well imal testing, and at least three states require testing every five years. At least three states stipu- posing environmental externalities. late different testing requirements after a period Types of Plugs Required of time—Wyoming and Colorado after a temAn important part of the typical decommisporary abandonment extension, and Texas after sioning process is placing plugs (usually made 25 years—and Colorado also gives regulator of cement) at the well bore’s surface, bottom, discretion at the time of temporary abandonand in regions in between. The purpose of ment extension to require a well be switched these plugs is to prevent contamination befrom a blanket to individual bond.35 Ohio retween oil and gas strata and subsurface regions quires the operator to inspect wells every six (e.g., freshwater zones) or the surface (e.g., months and submit an inspection report within methane or volatile organic compound emistwo weeks, a stringent outlier.36 In addition, sions). Regulators take various approaches to some states require that operators simply subplugging requirements, and they require operamit a report of results, whereas others require a tors to install a variety of types of plugs (if any witness to be present during testing. This eleat all). We expect that, in general and all else ment could confer large benefits on states by being equal, the more plugs that a regulator allowing regulators to catch problem wells berequires the less chance that a decommissioned fore the environmental and fiscal costs become well will create environmental externalities. exceedingly large. Of the 22 states we survey, 18 contain prePlugging Requirements scriptive regulations that describe the different A decommissioned well is one that is types of plugs (i.e., surface, intermediate, and properly plugged and the surrounding site bottom) that operators must install and the reproperly restored. Most state regulations we quired length of those plugs. The remaining review contain a general phrase about the need four states (Montana, Nebraska, New Mexico, to plug a well such that oil, gas, and water reand North Dakota37) contain a general statesources are contained to their original strata ment (i.e., a performance standard) about the and to prevent any subsurface contamination. need to decommission a well such that oil, gas, Beyond this, states show significant heterogeand water resources are contained to their origneity in how prescriptive they are in their regu- inal strata. Regulations in California and Texas lations; for example, some states describe contain such a phrase as well but also give preplugging materials, and the specific method of scriptive requirements, and Colorado uses a placing plugs of a certain length above or becombination of prescriptive and performance low various resources or strata of interest— whereas other states do not contain any such 35 WY Rule 3.16.c, d; 2 CCR-1-326.c; 16 TAR 1.3.15.l. 36 ORC 1509.062.B.6; ORC 1509.062.C. 37 ARM 36.22.1303; NAC Title 267 Chapter 3 028; NMC §19.15.25.10.A; NDAC 43-02-03-04. www.rff.org 35 Resources for the Future Ho et al. standard approaches.38 A handful of states (e.g., Colorado, Ohio, New Mexico, North Dakota, and Texas39) require that an operator’s plugging plan be approved by the regulator. The 18 states with prescriptive plugging regulations differ significantly in whether they contain prescriptive language requiring a surface, intermediate, and/or bottom plug—as displayed in Table 6. The blank rows indicate a state that relies on a general statement only.40 The BLM approves plugging plans for operators and does not require explicit provisions outside of this process.41 We use a binary and qualitative assessment when incorporating the types of plugs required into our stringency calculations for states. States that explicitly impose prescriptive requirements and/or a performance standard receive a 1, while all others receive a 0. Treatment of Wells with Different Casing Types An active well requires different types of casings—steel tubing inside the well that helps the bore maintain its structure and protects the bore from contamination—including surface, intermediate, and production casings. The latter type of casing is perforated when a well is active, thus allowing oil and gas to seep through those perforations. Portions of well bores may not necessarily be cased and those 38 CCR §1723, CPRC §3228; TAR 16-1-3-3.14(d)(2,3, and 8); 2 CCR-1-319.a.1. 39 2 CCR-1-311.a, 2 CCR-1-319.a.6; OAC 1501:9-1102, OAC 1501:9-11-04; NMC §19.15.7.14.A(1e, 2); NDAC 42-02-03-33; TAR 16-1-3-3.15(l-m). 40 While we treat prescriptive requirements as more stringent than general statements, note that a tightly monitored and enforced general statement that provides operators flexibility may be just as or even more protective than a less well enforced prescriptive approach—and potentially less costly. 41 43 CFR 3162.3-4.a. TABLE 6. PRESCRIPTIVE REQUIREMENTS FOR DIFFERENT TYPES OF PLUGS BY STATE (AND BLM) State AK AR CA CO IL IN KS KY LA MO MT NE NM NY ND OH OK PA TX UT WV WY BLM Bottom X X X X X X Intermediate X X X X X X X Surface X X X X X X X X X X X X X X X X X that are cased may have portions that are not cemented. The extent to which a well is cased and the degree to which that casing is cemented therefore play a critical role in determining the potential environmental risk of the operation (Zirogiannis et al. 2016). Of the 22 states we survey, roughly three-quarters of states differentiate plugging requirements by whether certain casing types are present and/or whether casing is cemented. In general, these differentiations define different categories of wells and the prescriptive plugging methods that must be used. The regulatory element of casing type is closely linked by definition to a state’s casing regulations, which was not in the current study scope (but was covered in Richardson et al. 2013). Additionally, casing conditions are part of many of the pre-plugging notices required by states, and thus may be individually tailored at this point through regulator discretion. If so, then regulations which specify different plugwww.rff.org 36 Resources for the Future Ho et al. ging regulations by casing type may simply be revealing a more prescriptive regulatory approach, and not necessarily guaranteeing higher environmental stringency. However, the environmental risks mitigated by stringency in casing are potentially large. Watson and Bachu (2009) identified uncemented casing intervals as the source of the vast majority of contamination and gas leakage and Dusseault et al. (2000) recommend full cementation from the intermediate casing to the surface. Ingraffea et al. (2014) show that for wells drilled in Pennsylvania from 2000 to 2012, 6.2 percent of unconventional wells have casing and cementing issues, almost six times the rate of conventional wells. The implications of these studies on inactive wells is that monitoring, reporting, and verification of casing and cementing should be a high priority prior to entering temporary abandonment status or for final plugging. Treatment of Casing Removal the odds of contamination of the wellbore, specifically regarding pollution of groundwater and surface subsidence. We did not find evidence of any uniform rules on whether casing removal increases or decreases environmental risk, but in cases of collapsed, compromised, or uncemented casing, removal may lead to better long term environmental integrity. Of the 22 states we survey, approximately one-third do not have special instructions for removal, whereas the remaining two-thirds either require that the regulator approve the removal of certain types of casing or set prescriptive restrictions regarding removal of casing. Seven states require regulatory approval or prohibit pulling of casing, particularly surface casing. At least one state (Wyoming) incorporates a performance standard: any production casing left in place must pass a mechanical integrity test and—if it fails—must be cemented.43 Oklahoma requires a license to pull casing, for which the company involved must show experience and financial responsibility.44 When a well is being decommissioned, cas- Treatment of Different Well Types ing may be removed for a variety of reasons, Different kinds of wells come with differincluding to create more favorable conditions ent environmental risk portfolios. Some states for a plugging job or to capture any salvage tailor requirements for decommissioning based value in the casing material itself.42 As menon well characteristics including type (e.g., tioned previously, some states have specific guidelines for how salvaging of well equipment horizontal well) and subsurface geography may contribute to the state-run decommission- (e.g., whether a well bore penetrates a coal seam, has hydrogen sulfide present, or is in a ing of orphaned and noncompliant sites. The permafrost area). Of the 22 states we survey, 9 casing removal guidelines in technical plughave special decommissioning requirements for ging sections likely refer to both original and well bores that pass through coal seams (relatthird-party operators who might remove well casing for salvage value. Some states explicitly ed to environmental externalities, resource protection, and worker safety concerns) and 6 for ban the removal of any or certain types of cashorizontal wells (Table 7). California may reing, and others require approval prior to pullquire special procedures for fractured ing. However, removing casing may increase 42 This section does not include cutting off of surface casing at plow depth, generally three feet below the surface, a common requirement to allow for continued surface use after decommissioning. 43 WY Rule 3.18.b.iii.C. 44 OAC 165:10-11-6.k. www.rff.org 37 Resources for the Future Ho et al. TABLE 7. SPECIAL DECOMMISSIONING REQUIREMENTS State AK AR CA CO IL IN KS KY LA MO MT NE NM NY ND OH OK PA TX UT WV WY BLM Horizontal Well Coal Seam Permafrost Salt or Sand X X X X X X X X X X X X X X X X X X X X shale or schist, but not horizontal wells as a general category.45 Cement Specifications The integrity of the cement used in securing the well bore and for plugs installed during decommissioning is crucial to limiting environmental externalities. In particular, one industry consultant claimed most decommissioned wells leak at some point due to cement shrinkage, and a report by Watson and Bachu (2009) finds that bridge plugs capped with cement plugs (the predominant plugging method used in Canada) result in leakage in 10 percent of decommissioned wells and recommends against using the bridge plug method. 45 Hydrogen Sulfide We find that eight states include quantitative standards for cement requirements, expressed either in pounds per square inch over a certain length of time or pounds per gallon. Five states require or reference API and/or ASTM standards. Three states use the length of cement plug or other related factors as a standard, in addition to many states that stipulate plug length in the technical plugging section. Dusseault et al. (2000) cite previous studies that assert that more ductile, low compressive strength cements are less likely to crack under stress. Our analysis shows that states that used psi measurements as standards had a mode of 500 psi, with exceptions being Alaska (1500 psi or 0.25 psi/ft) and Colorado (800 psi CCR §1723.1.c. www.rff.org 38 Resources for the Future Ho et al. after 72 hours).46 Using a stringency metric of 1 for states with any quantitative cement requirements, and 0 for those with no evidence of quantitative requirements or a general requirement (e.g., Portland cement), 10 states receive a stringency rating of 0. The use of cement additives is also addressed in some state regulations, either prohibiting their use or requiring regulatory approval. As Dusseault et al. (2000) note, additives can have either a positive or negative effect on cement quality but undergo little thirdparty verification. States also may regulate the materials that are used to fill the wellbore between plugs, and which can include performance standards for mud fluid or specifications on what non-mud materials may be used. Because cement plugs and other materials are the primary barrier against wellbore contamination and leakage, they are critical to long-term wellbore integrity. However, the literature recommends that cement slurries and applications be tailored for individual wellbore conditions rather than be addressed in uniform standards, as factors such as temperature, pressure, and surrounding strata can all affect cementing quality. For this reason it is difficult, and perhaps not even recommended, to assess stringency of prescriptive cement standards. Rigorous monitoring and witnessing of cement jobs may be a more appropriate regulatory path to take. Marking of Decommissioned Wells State databases of inactive wells are likely incomplete. For example, nearly all of the wells in Pennsylvania tested by Kang et al. (2014) for methane emissions were not on Pennsylvania’s list of inactive wells. Although a thorough review of the strategies that states take to identify inactive wells is beyond the scope of this 46 20 AAC 25.112.g; 2 CCR-1-319.a.1. paper, we do focus on one relevant regulatory element—whether a state requires decommissioned wells to be permanently marked. Whereas it is possible that the signs often required to be placed at the beginning of well construction are assumed to last past decommissioning, in our view this is not enough to assure proper identification. Unidentified wellbores complicate state identification of inactive wells and could result in incomplete information during surface purchases and other development decisions, and they could lead to environmental pollution or other externalities in the future. Of the 22 states we survey, 11 require operators to mark decommissioned wells in some fashion (Map 9). These 11 states typically require a permanent marker that is visible above ground or detectable below ground, if casing is cut off below plow depth. The BLM requires a permanent marker for decommissioned wells but the regulator can waive this requirement. We use a binary and qualitative assessment when incorporating marking of decommissioned wells into our stringency calculations. A state that explicitly requires marking of decommissioned wells receives a 1, whereas a state that does not have such requirements receives a 0. Restoration Requirements Requirement for restoring the well site (i.e., revegetation of surrounding areas and removal of equipment) is an area of significant heterogeneity among states and—as with regulations for plugging requirements—is difficult to compare across states. Surface disturbance by oil and gas activities—the well pad, the roads, the gathering lines, and the storage ponds—can leave a significant footprint that fragments habitat and exacerbates erosion and stormwater flows. Restoration Requirements—Timing, Stringency, and Relationship to Bonding General restoration requirements fall into three general categories, as displayed in Map 10. First, nearly one-third of states rely on a www.rff.org 39 Resources for the Future Ho et al. general clause that simply states that operators must restore the surface as near to the original state as possible. No further elaboration or specific requirements is detailed. The two remaining categories involve either a low or high amount of prescriptive restoration requirements, respectively. The “low” category typically includes regulations that describe a list of restoration procedures an operator must complete (e.g., remediate contaminated soils MAP 9. MARKING REQUIREMENTS FOR DECOMMISSIONED WELLS MAP 10. STRINGENCY OF RESTORATION REQUIREMENTS www.rff.org 40 Resources for the Future Ho et al. and drain fluid storage ponds), whereas the “high” category can include highly prescriptive requirements such as specific seed composition and seeding schedules for revegetation. We generally favor a performance standard approach on cost grounds, but with regard to limiting damage, the most prescriptive is probably the most stringent. We use a binary and qualitative assessment when incorporating the stringency of restoration requirements into our stringency calculations. States that contain a general clause receive a 0 and those that offer some level of prescriptive requirements receive a 1. States differ in how important restoration requirements are for the release of financial assurances. Of the 22 states we survey, only 10 require the regulator to inspect an operator’s restored site before releasing the financial assurance associated with that well, which is displayed as a “yes” in Map 10. States also vary by when operators must complete restoration requirements (Map 11). Of the 22 states we survey, 14 explicitly require operators to complete restoration within a certain amount of time, whereas 8 do not.47 Those 14 states require operators to complete restoration within a range of one month to one year—with a handful of states (Alaska, Colorado, Ohio, Oklahoma, and Pennsylvania) explicitly allowing for extensions.48 Similar to well idle time, longer time periods reflect greater likelihood that environmental externalities continue occurring for a longer period of time—all else being equal. We use a continuous and quantitative variable when incorporating this element into our calculation of stringency across states. The 47 AR Rule B-9.e; 805 KAR 1:170; LAC 43:I§21013101; 10 CSR 50-2.060(3)(A)(1); ARM 36.22.1307; NAC Chapter 3 012.14-15, NAC Chapter 3 028.08. 48 20 AAC 25.170; 2 CCR-1-1000-1004; OAC 1509.072; OAC 165:10-3-17.n, §17-53.2.F; 58 Pa. Code §3216. highest value receives a 0 and the lowest a 1, with values between normalized. Conversion to Freshwater Well Some states allow an oil and gas well to be converted into a freshwater well, in which a bottom plug and a plug below the freshwater zone are placed. Of the 22 states we survey, 17 allow for the conversion of a well to a freshwater well. The states differ in the degree of regulator discretion and technical requirements associated with such conversion. Most states that allow conversion require a written statement that the landowner assumes all liability for the freshwater well and must receive approval of the plugging plan from the oil and gas regulator. Four states require approval from the relevant water regulatory body or groundwater rights holder. Regulator Involvement The types of regulator involvement appearing in our survey are limited, and in general are difficult to measure in statutes and regulations. Our review covered only statutes and administrative rules and regulations, not permit documents or other sources, and cannot account for how the regulations are implemented nor many aspects of monitoring, review, and enforcement. Additionally, variance clauses and phrases like “regulator discretion”—for example, allowing time extensions, exemptions from prescribed technical requirements, and others—are found throughout all sections of the regulations surveyed. Although language to the effect of “to the satisfaction or approval of the Director” may seem similar to performance standards, such language lacks specificity as to goals. It does, however, insert some amount of flexibility into otherwise prescriptive command-and-control regulations that could account for well-by-well characteristics that characterize plugging costs and environmental risk. www.rff.org 41 Resources for the Future Ho et al. MAP 11 TIME LIMIT FOR RESTORATION (IN MONTHS) Notification, Approval, and Inspection A well’s plugging plan may require notification, approval, and/or inspection. The specificity of the plugging plan varies on a state-bystate basis, but sometimes it includes proposed plugging methods and depth and length of plugs, as well as well casing information and well integrity test results. Given that GWPC (2014) identifies increasing reliance by regulators on plugging reports over in-person witnessing by a regulator as a possibly worrying trend, conditional inspection requirements might be improved by placing a default requirement on witnessing. Some states surveyed make an explicit statement on the right of the regulator to enter onto the well site and inspect at any time, which similarly gives regulators flexibility in choosing wells to prioritize for inspection. Such an option, depending on how it is applied by regulators, may help mitigate environmental and fiscal risk (especially in the case of orphaning). All of the states in our survey require either notification or approval (and some require both) before and after decommissioning a well, but relatively few require inspection, as detailed in Table 8. The time by which postplugging notice (generally in the form of a plugging report) is required is fairly uniform at 30 days, while pre-plugging notice is subject to greater variation; yet the impact of this variation on regulatory oversight and risk mitigation is unclear. One source of heterogeneity is whether parties other than the oil and gas regulator (e.g., surface owners or coal mine owners) must be notified. A process for comment by these parties on the proposed plugging plan is often included in this notice, and it partially addresses externalities relating to choice of plugging plan on other resources. This approach seems advisable, although the size of the transaction cost may be significant for some operators. Many states have a clause allowing emergency plugging with verbal or no prior approval. Notification is most common www.rff.org 42 Resources for the Future Ho et al. TABLE 8. NUMBER OF STATES REQUIRING APPROVAL, INSPECTION, NOTIFICATION, AND REPORTING Approval Inspection Notification Reporting Pre-plugging 17 3 12 1 Post-plugging 2 1 15 6 Pre-restoration 2 1 1 0 Post-restoration 1 4 1 1 for post-plugging reports, whereas approval of an operator’s plugging plan is most common for the pre-plugging stage. Another surprise in Table 8 is that the restoration phase has so few requirements across the states. Ability for Regulator to Order Plugging or Replugging evidence of such requirements in five states (Arkansas, Illinois, Indiana, North Dakota, and Oklahoma, although even these may have reporting requirements outside of regulations). At least nine states require that operators file reports monthly, semiannually, or annually that detail certain types of inactive wells. At least five states require that regulators file the reports instead, either on a monthly or annual basis. Two states (California and Wyoming) require both.49 We use a binary and qualitative assessment when incorporating this element into our stringency calculations for states. A state receives a 1 if it has explicit reporting requirements (for operators, regulators, or both) and a 0 if it does not. Another regulatory power that some states make explicit is the ability of a regulator to order plugging or re-plugging of a well. Of the 22 states we survey, 17 give regulators this ability. In most cases, the order entails a notice and hearing or other appeals process available to the operator. Commonly cited reasons for allowing such an order include leaks and lack of compliance with notification requirements. Such authority relates to a previously discussed Consideration of Inactive Wells in New element: any outstanding liability or “statute of Well Permitting Processes limitation clauses.” Introduction of a quasiOne environmental risk is that new drilling CERCLA program, in which operator liability may unexpectedly cross an existing inactive for a well extends beyond the time of bond rewell bore, whether plugged or not. Thus we turn, could protect the state against environexamined whether states pay attention to the mental damages that do not appear until after location of existing plugged and abandoned well decommissioning and might deter some wells when permitting new wells. For the seven cases of operator orphaning. On the other hand, states that do require some notation on a well such extensive liability coverage might cause a permit application of plugged wells in the area, short term increase in orphaning rates if small- three include it specifically for wells to be hyer firms exit due to the increased regulatory draulically fractured, two for disposal wells, burden. Establishing clear processes by which and one for gas storage reservoirs. Arkansas is regulators may order re-plugging of a well is a an interesting outlier in that it excludes plugged middle-ground option. wells from consideration for fracturing permits.50 The long-term risks of Reporting Requirements of Operator or Regulator on Inactive Wells State regulations often require that regulators and/or operators report the number of inactive wells, as displayed in Map 12. Of the 22 states we survey, we could not find any 49 CPRC §3227.5; CPRC §3227.a.2; WY Rule 3.16.a. 50 AR Rule B-5.g.1.A. www.rff.org 43 Resources for the Future Ho et al. MAP 12. REQUIREMENTS TO REPORT INACTIVE WELLS unconventional wells once decommissioned is still unknown, and the potential risk of fluid migration or other contamination due to unconventional wells in close proximity to decommissioned wells is one possible pathway that could arise. requires plugging operations to be continuously monitored by a methane gas detector; if the methane concentration exceeds 3 percent, plugging must immediately cease.53 Notably, all of these programs assume well ownership is known. Fugitive Methane Policy Recommendations and Conclusions We found only a few references to fugitive methane monitoring requirements. By far the most noteworthy of these are Kentucky and West Virginia, which allow adjacent landowners to enter a noncompliant site or site where gas is leaking and plug the well. 51 Pennsylvania allows operators to vent gas to the atmosphere at inactive oil wells (or confine to the producing formation), but if this flow is over 5,000 cubic feet per day the regulator must be notified and remedial action taken.52 Illinois In general, and echoing a recommendation by Richardson et al. (2013), it was very challenging to pull out regulations from some state codes, and the lack of information on use of field rules and permitting to adjust rules for individual cases is not well documented. Enforcement data are likewise difficult to find and process. Thus, we recommend that states do a better job in reporting these practices. Our policy recommendations and conclusions are as follows. First, we echo recommen- 51 KRS 353.140, 150, 160; WVC §22-6-31(b); WVC §22-6-32. 52 25 Pa. Code §78.102(3); §78.102(B)(2)(ii)(D). 53 62 IAC I.240.1140(e). www.rff.org 44 Resources for the Future Ho et al. dations by Shih et al. (forthcoming) that states should require an amount of financial assurance that reflect real world plugging costs. Several states have completed reviews of their inactive well programs and have called for reviews of bonding amounts. Where states have not already revised their bonding amounts, we recommend they explore doing so based upon actual plugging costs among wells they have plugged. Where appropriate, differentiating bond amounts based on well characteristics or other driving factors of well plugging costs might provide better fiscal coverage, and provisions or special attention to operators with a history of noncompliance or a high share of inactive wells are featured by some states and seem advisable for all. Many states already make various adjustments of this type. Additionally, states should review the types of financial assurance offered, particularly those such as annual well fees, liens on equipment, and statements of financial health. These assurances, by design, do not require sufficient funds up front (irrespective of whether required financial assurances are set to cover well plugging and site restoration costs). In some cases, such as Oklahoma, these weak financial assurance measures are only available for operators with a good compliance history, whereas in other states they are specifically targeted at operators that may have trouble meeting the full bond amount. States face an important question when setting bond amounts: how can states pool financial risk across operators, while not making financial assurance prohibitively expensive for some individual operators? One area of further research that could illuminate this question is the take-up rate and default rates for blanket versus individual bonds, and whether certain types of operators (e.g., small versus large, concentrated in one basin or spread throughout the state, new versus old) use blanket or individual bonds more. A system where industry contributes to a dedicated fund for plugging high-cost wells, in addition to standard finan- cial assurance requirements, could provide an extra measure of protection for states while still allowing competitive financial assurance amounts. Second, we observe that almost all states offer the option of blanket bonds, and these bonds feature quantity discounts over the number of wells covered. While individual well financial assurance requirements may be too low relative to plugging and site restoration costs for a sizable fraction of wells, as seen in Shih et al. (forthcoming), per-well coverage from blanket bonds is even considerably lower. Although blanket bonds may reduce administrative costs for state regulators, it is unclear whether the net benefit to the state of setting significant price discounts through blanket bonds is greater than the risk. Third, our review of temporary abandonment practices finds that state well idle time and temporary abandonment time periods are generally well defined, although extensions and a few outliers (e.g., California) might allow for wells to be in an inactive state with no intention to return to active status. Requirements for temporary abandonment—including well testing and monitoring, proof of future economic viability, and mandated well closure requirements—are far less common in state regulations. Although these activities may be addressed during temporary abandonment approval proceedings (if approval is required), implementing systematic procedures for temporarily abandoned wells may greatly reduce the probability of future environmental damages or fiscal costs for the state. In particular, wells that have been inactive longer are likely to cause a greater risk. Thus, we recommend that states develop more stringent temporary abandonment requirements. Fourth, well plugging and site restoration requirements vary greatly in the amount of detail set forth in regulations. This reflects both the different regulatory approaches taken by states, and the fact that well plugging and www.rff.org 45 Resources for the Future Ho et al. abandonment requirements may be dealt with on a case by case basis with a large amount of regulator discretion. In this case, ensuring a robust monitoring, review, and validation framework would be the most important regulatory element. We find that whereas most states require pre-approval of a plugging plan, relatively few require inspection after plugging and almost none inspect after site restoration is complete. Given that state resources are limited, the use of interested parties to inspect, such as surface and other resource owners, could provide additional monitoring capacity. In this regard, we note the use of surface damage agreements in a substantial number of states and their relevance for addressing some of the conflicts that arise in split estate cases. We recommend further study into how these agreements work in practice. 5. Conclusions, Recommendations and Future Research This report has detailed the environmental and financial issues associated with inactive oil and gas wells. We considered the risk pathways and the number of such wells in a group of states, including orphaned wells that are the states’ responsibility to clean up, surveyed the costs of decommissioning wells in a large group of states, and conducted a very detailed statistical analysis of the drivers of costs in decommissioning orphaned wells in Kansas. We then examined the regulations governing inactive wells across 22 states. These analyses lead to the following conclusions. While we understand very well how inactive wells can harm the environment there are very few studies providing empirical information to show precisely how much these wells are affecting the environment or how much each factor contributes to environmental risk. However, the quality of well construction at the time it was drilled and the abandonment measures that have been taken on the well are two factors that stand out. Most disappointing, the empirical literature does not distinguish between the environmental damage caused by different types of inactive wells (e.g., temporarily abandoned vs. plugged and abandoned wells; historic wells vs. wells drilled more recently). As for the size of the inactive well burden, data from 13 states with significant oil and gas production show that about 12 percent of all inactive wells have not been decommissioned, but the percentage in each state varies considerably from one percent to 56 percent. Of course, the number of inactive wells reported here does not include wells that are simply missing from state records. We find that decommissioning costs vary significantly, both within and across states, depending on a combination of factors, including the condition of the well, the quality of its original construction, well depth, market conditions in the production sector, and the market structure of the service provider industry. Average and median decommissioning costs exceed average bond amounts in most of the states studied here. Some wells are particularly expensive to decommission. The cost of decommissioning such wells could be covered by a pool of industry funds. Regarding regulations, the individual states and the Bureau of Land Management take vastly different approaches to regulating the decommissioning of oil and gas wells, meaning that operators face wide heterogeneity in the rules they must follow when a well becomes inactive. This heterogeneity can be described in terms of the number of regulatory elements they regulate and their stringency. But many states offer only vague statements in support of some regulatory elements, while others are very specific. Our most important findings about individual regulations are, for one, that state bonding requirements are set too low to cover decommissioning costs. Thus, when wells become orphaned, states and taxpayers will be on the hook for significant clean-up liabilities. We www.rff.org 46 Resources for the Future Ho et al. also found that states generally set bonding requirements that vary by at least one factor that affects costs, such as well depth. But the approaches are too simple to fully protect taxpayers even if the bonding amounts were larger. Based on our findings, we highlight a number of priority areas for policy reform for state oil and gas agencies, BLM, and other relevant agencies to consider. Bond Amounts 1. Industry bonding requirements should be compared against decommissioning costs in each state and revised accordingly to cover more of the decommissioning costs. Note, however, that the optimal amount of cost coverage that should be built into a bond is uncertain: while pegging bond amounts to the cost of the most expensive projects places an unreasonable burden on operators and can discourage drilling activity, too low a bond amount does not provide an adequate incentive for operators to decommission their wells. 2. Bonding regulations should have provisions to ensure that states do not bear the cost of particularly expensive decommissioning projects. For instance, a pool of funds could be provided by industry that could be drawn on if the cost of a project exceeds a certain threshold. 3. Bond amounts should be calibrated to account for a variety of factors influencing cost. At present, some states set bond amounts that vary by well depth, an important characteristic given that average well depths in the United States have been increasing. Some also set different bond amounts for specific districts, as several factors that affect costs vary spatially. Other factors that a few states consider include whether a well is horizontal or vertical, as well as operator characteristics such as compliance history and the number of wells the operator owns. 4. Consider the use of surface damage agreements in addition to traditional plugging or plugging and restoration bonds. We found that a number of states use some form of surface damage agreement or negotiation in cases where the surface and mineral estates are split. Although it is possible that these arrangements are common on an individual lease-by-lease basis, we view their inclusion in state regulatory programs as a potential way to limit externalities (e.g. crop damage, noise) of decommissioning to affected surface owners when secondary purchasers of the well go bankrupt. Relatedly, the very few states that inspect a site for the quality of its restoration before releasing financial assurances can serve as an example for other states. Well Management and Monitoring 5. The conditions under which wells are allowed to be transferred from one operator to another should be stricter. Anecdotal insights have suggested that wells tend to be transferred from larger operators to smaller companies that purchase these wells for enhanced recovery or other operations. At present, well transfers are typically allowed as long as the buyer can cover the cost of the financial assurance attached to the bond. Regulators should ensure that the operator purchasing the well is financially stable and has a good track record of compliance with regulations. Alternatively, or in addition, states could hold the original owners at least partly liable for decomissioning their wells rather than trasnferring all liability to the new www.rff.org 47 Resources for the Future Ho et al. owners, who are sometimes less financially secure 6. Some states should tighten their requirements for maintaining temporary abandonment status. Otherwise, operators may maintain their wells in this status for longer than appropriate to avoid or postpone decommissioning costs, raising environmental risks. Such measures could include well integrity demonstrations and inspections both prior to and at regular intervals during temporary abandonment, and requirements to demonstrate the future economic viability of a well (perhaps especially relevant during bust periods). 7. States should conduct legislative audits to evaluate the stringency of their monitoring efforts and success of their plugging programs. Regulatory capacity is a crucial factor that we do not evaluate in this study: a legislative audit in Louisiana found that the Office of Conservation did not issue compliance orders for 86 percent of the 482 wells that required them from 2008-2013. Audits should examine questions such as what percentage of the wells on their plugging lists states are able to decommission each year, whether inactive wells are consistently monitored, and what percentage of decommissioning costs have been covered by industry. Inactive and Orphaned Well Programs 8. States should develop more sustainable means of funding their orphaned well plugging programs. At present, recovered bonds represent only a fraction of plugging revenues available for decommissioning orphaned wells in most states. States thus rely on a combination of legislative appropriations (public monies) and permit fees from industry, which are unreliable in times of low oil production. At a minimum, recovered bonds should cover a larger share of decommissioning costs. 9. States need to do a better job in reporting information on numbers of inactive wells of various types and statuses, costs, and regulatory adjustments in the field and through permitting. It was very difficult to find credible numbers of inactive wells and detailed data on orphaned well decomissioning costs. It was also very difficult to pull out regulations from some state codes, and there is a lack of information on how state rules are adjusted at the district level or for individual cases. Enforcement data are likewise difficult to find and process. 10. Given the heterogeneity of state regulations, states should consider using the Regulatory Exchange (supported by the Groundwater Protection Council and the Interstate Oil and Gas Compact Commission) or other bodies to share regulatory information and learn from one another. These recommendations obviously don’t apply to states (and the BLM) already adequately addressing them. And such states can serve as a model for others. Finally, our study has revealed a number of areas for further research—both on the environmental risks of inactive wells and on the regulations governing these wells. Environmental Risks of Inactive Wells  What is the magnitude of environmental risk posed by inactive wells with varying characteristics? Our review of the literature and conflicting opinions from experts we spoke to revealed that there www.rff.org 48 Resources for the Future Ho et al. are limited empirical answers to this question and that this is an area in need of further research. For instance, to what extent do decommissioned wells plugged according to present-day regulatory standards still pose a risk of methane leakage or groundwater contamination? And how do such risks vary as the plug ages? How risky is a temporarily abandoned well as compared to a decommissioned well? What plugging technologies are most effective for minimizing leakage risk? What about the differences in risks posed by wells with modern well constructions, as opposed to historic wells? Understanding this heterogeneity in environmental threat has implications for how to cost-effectively target regulatory efforts.   Inactive Well Regulations and Programs  How do blanket bond amounts compare against decommissioning costs for groups of wells owned by a single operator, and to what extent do operators choose to post blanket bonds? Our data only permitted us to compare bonds for individual wells against perwell costs. Operators are allowed the option of posting a blanket bond for all of their wells in a state. The blanket bond amount is lower on a per well basis than individual bonds, implying that the greater the extent that operators post blanket bonds, the more that plugging costs will exceed bond amounts.  What types of bonds do operators most often post? In most states, operators may post either personal bonds, putting up their own assets as financial assurance, or surety bonds, in which case a thirdparty provider pays out the bond in the event that the operator is unable to decommission the well. There are other options as well. Research is needed on   which types of bonds are more protective of the taxpayer and the environment. To what extent do agencies evaluate the financial capacity of a new operator before they allow wells to be transferred from the primary operator? If they do not, there is no guarantee that the new operator will have the financial means to bear the cost of decommissioning, making it more likely that the well will eventually become orphaned. What are the characteristics of the operators that most commonly orphan their wells? Smaller operators that are less financially stable may be more likely to default on their bonds. A systematic study of the data on this issue and other drivers of default and orphaning would suggest ways to better reform regulations. A related question is whether wells under blanket or individual bonds are more likely to be orphaned. How are states financing their plugging programs? A comprehensive review of state plugging revenues and expenditures could be conducted to assess the magnitude of the cost burden that is currently borne by the public. Such a review should also evaluate alternative financing options that would both increase available revenues and ensure that costs are internalized by industry. Which states have been most successful at monitoring inactive wells and plugging orphaned wells? A mixedmethod study to identify “leadership states” and understand the factors behind their success could be very powerful in revealing best practices. Factors that could be examined include political and economic conditions in the state, bonding amounts, regulator capacity, plugging revenue sources, and degree of citizen involvement. www.rff.org 49 Resources for the Future Ho et al.  What are effective ways of identifying orphaned or abandoned wells not currently in state records? Several states already have programs in place to locate and document orphaned wells, including Pennsylvania, Kansas, Colorado, and West Virginia. In some cases, the public can get involved in these efforts, submitting information about wells to the state website. A review can be done of various states’ programs to identify the most effective identification methods. Another idea that can be considered is the creation of a mobile application that the public can use to report orphaned wells that they find, uploading the images and locations of orphaned wells into a national cloud database. This digitized information can be streamlined, ensuring good data quality, and then shared with state agencies. Continued growth in wells drilled will eventually cause an increase in the number of inactive wells, and therefore a growth in the environmental threat and financial burden from these wells. Further research and regulatory reforms can help mitigate risks. Ultimately, developing effective policy recommendations will depend on a deeper understanding of where the environmental and financial risks are greatest, how operators are currently making decisions about temporary abandonment, well transfers, types of bonds, and permanent decommissioning, and how regulations can best be reformed. The challenge of ensuring that future environmental costs are borne by polluters is common to other sectors, such as mining and waste disposal, and lessons learned from this research will therefore have implications that reach beyond the oil and gas industry. www.rff.org 50 Resources for the Future Ho et al. Appendix A. Average Well Depth and Associated Average Bond Amounts by State APPENDIX A1. STATES QUALIFYING FOR OUR REGULATORY ANALYSIS OF STATE REGULATIONS ON INACTIVE WELLS Number State 1 Alaska 2 Arkansas 3 California 4 Colorado 5 Illinois 6 Indiana 7 Kansas 8 Kentucky 9 Louisiana 10 Mississippi 11 Missouri 12 Montana 13 New Mexico 14 New York 15 North Dakota 16 Ohio 17 Oklahoma 18 Pennsylvania 19 Texas 20 Utah 21 West Virginia 22 Wyoming www.rff.org 51 Resources for the Future Ho et al. APPENDIX A2. THRESHOLD CRITERIA FOR INCLUDING STATES IN OUR REGULATORY ANALYSIS OF STATE REGULATIONS ON INACTIVE WELLS PANEL A: NUMBER OF ORPHANED WELLS ON WAIT LIST IN 2006 AS REPORTED BY INTERSTATE OIL AND GAS COMPACT COMMISSION Number State 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Alabama Alaska Arkansas California Colorado Florida Illinois Indiana Kansas Kentucky Louisiana Michigan Mississippi Missouri Montana Nebraska New Mexico New York N Dakota Ohio Oklahoma Pennsylvania Texas Utah Virginia West Virginia Total Number of Orphaned Wells on State's Wait List 0 15 577 430 45 17 3900 598 7271 10600 3183 100 49 2000 90 6 134 4800 4 2089 2089 8700 11220 8 37 1260 59222 Percent of Total 0.0 0.0 1.0 0.7 0.1 0.0 6.6 1.0 12.3 17.9 5.4 0.2 0.1 3.4 0.2 0.0 0.2 8.1 0.0 3.5 3.5 14.7 18.9 0.0 0.1 2.1 100.0 Included in Analysis? N N Y N N N Y Y Y Y Y N N Y N N N Y N Y Y Y Y N N Y www.rff.org 52 Resources for the Future Ho et al. PANEL B: HISTORICAL CRUDE OIL PRODUCTION BY STATE FROM 1981 TO 2014 AS REPORTED BY THE UNITED STATES ENERGY INFORMATION ADMINISTRATION Number State 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Alabama Alaska Arizona Arkansas California Colorado Florida Illinois Indiana Kansas Kentucky Louisiana Michigan Mississippi Missouri Montana Nebraska Nevada New Mexico New York N Dakota Ohio Oklahoma Pennsylvania South Dakota Tennessee Texas Utah Virginia West Virginia Wyoming Total Total Crude Oil Production (thousands of barrels) 476828 15436216 3479 338067 9611042 1054404 234098 551751 97584 1629315 142549 4151638 479038 817398 4710 809190 140859 45851 2422916 14474 2378928 297293 3374291 101357 49448 15931 21102459 829835 605 84458 2741437 69437449 Percent of Total 0.7 22.2 0.0 0.5 13.8 1.5 0.3 0.8 0.1 2.3 0.2 6.0 0.7 1.2 0.0 1.2 0.2 0.1 3.5 0.0 3.4 0.4 4.9 0.1 0.1 0.0 30.4 1.2 0.0 0.1 3.9 100.0 Included in Analysis? N Y N N Y Y N N N Y N Y N Y N Y N N Y N Y N Y N N N Y Y N N Y www.rff.org 53 Resources for the Future Ho et al. PANEL C: HISTORICAL ONSHORE NATURAL GAS PRODUCTION FROM 1992 TO 2014 AS REPORTED BY THE UNITED STATES ENERGY INFORMATION ADMINISTRATION Number State 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Alaska Arkansas California Colorado Kansas Louisiana Montana New Mexico N Dakota Ohio Oklahoma Pennsylvania Texas Utah West Virginia Wyoming Total Total Onshore Natural Gas Production (MMcf) 69609461 9740182 6865283 23774320 10991955 37126176 1792136 35152190 2466404 2851044 41483614 14430793 147884221 8139257 6620767 39868485 458796288 Percent of Total 15.2 2.1 1.5 5.2 2.4 8.1 0.4 7.7 0.5 0.6 9.0 3.1 32.2 1.8 1.4 8.7 100 Included in Analysis? Y Y Y Y Y Y N Y N N Y Y Y Y Y Y www.rff.org 54 Resources for the Future Ho et al. Appendix B. State Oil and Gas Regulations Alaska Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens and/or special liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability requirement TA technical shut-in requirement TA well integrity demonstration Plugging and Restoration Plugging requirements Special treatment of casing removal Special plugging requirements by well type Cement standards Conversion to freshwater wells Marking after permanent abandonment Restoration requirements Regulator Notification, Approval, and Inspection (NAI) NAI post restoration NAI pre plugging NAI post plugging Other Ability of regulator to order plugging Reporting requirements of operator and/or regulator Protection against fugitive methane at inactive sites Consideration of inactive wells during new well permitting 20 AAC 25.025.b 20 AAC 25.025.b 20 AAC 25.025.a No evidence found No evidence found 20 AAC 25.025 20 AAC 25.025.d; 20 AAC 25.026 20 AAC 25.072.a 20 AAC 25.072.d 20 AAC 25.115 20 AAC 25.110.a.2 20 AAC 25.110.c 20 AAC 25.072.e 20 AAC 25.112 No evidence found 20 AAC 25.112.e 20 AAC 25.112.g 20 AAC 25.140 20 AAC 25.120 20 AAC 25.170 20 AAC 25.170.d 20 AAC 25.105.e 20 AAC 25.070.3 20 AAC 25.105.a 20 AAC 25.115.a 20 AAC 25.066 20 AAC 25.283.a.10; 20 AAC 25.283.1.2.C.i Note: AAC: Alaska Administrative Code www.rff.org 55 Resources for the Future Ho et al. Arkansas Financial Assurance Individual Bond GRR B-2.f.1 Blanket Bond GRR B-2.f.4 Types of financial assurances allowed GRR B-2.d Use of surface damage agreements AC §15-72-203 Bond varies by operator characteristics GRR B-2.g Bond varies by well characteristics GRR B-2.h; GRR B-2.f.4 Liens or special liability provisions GRR G-3; GRR G-2 Temporary Abandonment NAI TA GRR B-5.h TA time limit GRR B-5.h Well idle time limit GRR B-7.d; B-7.c TA future economic viability requirement No evidence found TA technical shut-in requirement GRR B-5.h.3 TA well integrity demonstration GRR B-5.h.3.D.iv Plugging and Restoration Plugging requirements GRR B-8; GRR B-9 Special treatment of casing removal No evidence found Special plugging requirements by well type GRR B-9.d Cement standards GRR B-9.a.2 Conversion to freshwater wells GRR B-11 Marking after permanent abandonment No evidence found Restoration requirements GRR B-9.e Regulator Notification, Approval, and Inspection (NAI) NAI post restoration No evidence found NAI pre plugging GRR B-5.e-g NAI post plugging No evidence found Other Ability of regulator to order plugging GRR G-1.c-g; GRR B-5.c.3; GRR B-1.c.2; GRR B-26.l Reporting requirements of operator and/or No evidence found regulator Protection against fugitive methane at inactive sites No evidence found Consideration of inactive wells during new well GRR B-5.g.1.A. permitting Notes: GRR: Arkansas Oil and Gas Commission, General Rules and Regulations; AC: Arkansas Code. www.rff.org 56 Resources for the Future Ho et al. California Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics CPRC §3204 CPRC §3205 CCR §995.710-760; CPRC §3205.5 No evidence found CPRC §3270.4; CCR §1722.8; CPRC §3202; CPRC §3206 CPRC §3204; CPRC §3205 CPRC §3270.4.d; CPRC §3237 Bond varies by well characteristics Liens or special liability provisions Temporary Abandonment NAI TA No evidence found TA time limit CPRC §3008.d-e; CCR §1723.9 Well idle time limit No evidence found TA future economic viability requirement No evidence found TA technical shut-in requirement No evidence found TA well integrity demonstration CCR 1723.9 Plugging and Restoration Plugging requirements CCR §1723; CPRC §3228 Special treatment of casing removal CCR §1723.6 Special plugging requirements by well type CCR §1723.8; CCR §1723.1.c Cement standards CCR §1723.a Conversion to freshwater wells No evidence found Marking after permanent abandonment CCR §1723.5 Restoration requirements CCR §1776 Regulator Notification, Approval, and Inspection (NAI) NAI post restoration No evidence found NAI pre plugging CPRC §3229-3230; CCR §1714 NAI post plugging CPRC §3232; CPRC §3215.a; CCR §1724.1 Other Ability of regulator to order plugging CPRC §3237.a.1; CPRC §3208.1; CPRC §3206.5 Reporting requirements of operator and/or CPRC §3227.5; CPRC §3227.a.2; CPRC regulator §3260 Protection against fugitive methane at inactive CPRC §3240-3241; CPRC §3850-3865 sites Consideration of inactive wells during new well CPRC §3160.d.1.E permitting Notes: CPRC: California Public Resources Code; CCR: California Code of Regulations. www.rff.org 57 Resources for the Future Ho et al. Colorado Financial Assurance Individual bond 2 CCR-1-706.a Blanket bond 2 CCR-1-706-b Types of financial assurances allowed 2 CCR-1-100; 2 CCR-1-702; C.R.S. §34-60-106(13) Use of surface damage agreements 2 CCR-1-703 Bond varies by operator characteristics 2 CCR-1-707; 2 CCR-1-702.a Bond varies by well characteristics 2 CCR-1-706 Liens or special liability provisions 2 CCR-1-708; 2 CCR-1-320 Temporary Abandonment NAI TA 2 CCR-1-319.b TA time limit 2 CCR-1-319.b.1 Well idle time limit No evidence found TA future economic viability requirement 2 CCR-1-319.b.1 TA technical shut-in requirement 2 CCR-1-319.b.1 TA well integrity demonstration 2 CCR-1-326.c Plugging and Restoration Plugging requirements 2 CCR-1-319.a.1 Special treatment of casing removal 2 CCR-1-319.a.4 Special plugging requirements by well type 2 CCR-1-209 Cement standards 2 CCR-1-319.a.1 Conversion to freshwater wells 2 CCR-1-319.a.7 Marking after permanent abandonment 2 CCR-1-319.a.5 Restoration requirements 2 CCR-1-1000-1004 Regulator Notification, Approval, and Inspection (NAI) NAI post restoration 2-CCR-1-1004.c NAI pre plugging 2 CCR-1-311.a; 2 CCR-1-319.a.6 NAI post plugging 2-CCR-1-311.b; 2-CCR-1-319.a.3 Other Ability of regulator to order plugging 2 CCR-1-208 Reporting requirements of operator and/or 2-CCR-1-308.B; 2-CCR-1-309 regulator Protection against fugitive methane at inactive No evidence found sites Consideration of inactive wells during new well 2 CCR-1-303.b.3.D; 2 CCR-1-608.a permitting Notes: CCR: Colorado Code of Regulations; CRS: Colorado Revised Statutes. www.rff.org 58 Resources for the Future Ho et al. Illinois Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or special liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability requirement TA technical shut-in requirement TA well integrity demonstration Plugging and Restoration Plugging requirements Special treatment of casing removal Special plugging requirements by well type Cement standards Conversion to freshwater wells Marking after permanent abandonment Restoration requirements Regulator Notification, Approval, and Inspection (NAI) NAI post restoration NAI pre plugging NAI post plugging Other Ability of regulator to order plugging Reporting requirements of operator and/or regulator Protection against fugitive methane at inactive sites Consideration of inactive wells during new well permitting 62 IAC Chapter I Section 240.1500(a)(2) 62 IAC Chapter I Section 240.1500(a)(2) 62 IAC Chapter I Section 240.1510 No evidence found 62 IAC Chapter I Section 240.1500(a)(1) 62 IAC Chapter I Section 240.1500(a)(2) No evidence found 62 IAC Chapter I Section 240.1130(c) 62 IAC Chapter I Section 240.1130(f) 62 IAC Chapter I Section 240.1120 and Section 240.1130(a-b) No evidence found 62 IAC Chapter I Section 240.1130(c) 62 IAC Chapter I Section 240.1130(c) 62 IAC Chapter I Section 240.1550 62 IAC Chapter I Section 240.1150(d) No evidence found 62 IAC Chapter I Section 240.1280(a-b) 62 IAC Chapter I Section 240.1280 No evidence found 62 IAC Chapter I Section 240.1160 and 240.1770 No evidence found 62 IAC Chapter I Section 240.1140 (a) 62 IAC Chapter I Section 240.1190 62 IAC Chapter I Section 240.1610(a) No evidence found 62 IAC Chapter I Section 240.1140(e) No evidence found Note: IAC: Illinois Administrative Code www.rff.org 59 Resources for the Future Ho et al. Indiana Financial Assurance Requirements Individual Bond 312 IAC 16-4-2 Blanket Bond 312 IAC 16-4-2.a.5 Types of financial assurances allowed 312 IAC 16-4-2; IC 14-37-6-2; IC 14-37-6-4 Use of surface damage agreements No evidence found Bond varies by operator characteristics 312 IAC 16-4-1.a; 312 IAC 16-3-5-2 Bond varies by well characteristics 312 IAC 16-4-2 Liens or special liability provisions IC 14-37-13-2 Temporary Abandonment NAI TA TR Section 6; 312 IAC 16-5-20.b TA time limit 312 IAC 16-5-20.b, e Well idle time limit 312 IAC 16-5-20.b-c TA future economic viability requirement 312 IAC 16-5-20.f.3 TA technical shut-in requirement 312 IAC 16-5-20.b, d TA well integrity demonstration 312 IAC 16-5-20.d Plugging Requirements Plugging requirements TR Section 14, 15, 20, 1.13; 312 IAC 16-5-19 Special treatment of casing removal No evidence found Special plugging requirements by well type TR Section 19 Cement standards TR Section 3; 312 IAC 16-5-19.f Conversion to freshwater wells 312 IAC 16-5-19(r) Marking after permanent abandonment TR Section 20.e Restoration requirements TR Section 23; 312 IAC 16-5-19.p-r Regulator Notification, Approval, and Inspection (NAI) NAI post restoration no evidence found NAI pre plugging TR Section 7, Section 8 NAI post plugging TR Section 22, IC 14-37-8-4.4; 312 IAC 16-519(k-o) Other Ability of regulator to order plugging TR Section 10; IC 14-37-8-7, 12-15 Reporting requirements of operator and/or No evidence found regulator Protection against fugitive methane at inactive sites IC 14-37-8-10 Consideration of inactive wells during new well IC 14-37-4-5 permitting Notes: IC: Indiana Code; IAC: Indiana Administrative Code; TR: Temporary Rule www.rff.org 60 Resources for the Future Ho et al. Kansas Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed to be posted Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or special liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability requirement TA technical shut-in requirement TA well integrity demonstration Plugging and Restoration Plugging requirements Special treatment of casing removal Special plugging requirements by well type Cement standards Conversion to freshwater wells Marking after permanent abandonment Restoration requirements Regulator Notification, Approval, and Inspection (NAI) NAI post restoration NAI pre plugging NAI post plugging Other Ability of regulator to order plugging Reporting requirements of operator and/or regulator Protection against fugitive methane at inactive sites Consideration of inactive wells during new well permitting KSA 55-155 KSA 55-155 KSA 55-155 and KAR-82-3-120(f) No evidence found KSA 55-155(d)(3) KSA 55-155 No evidence found No evidence found KAR 82-3-111(b)(d)(e ) KAR 82-3-111(a) No evidence found No evidence found KAR 82-3-111(b-c) KAR 82-3-114(a)(1-3), KAR 82-3-114(e ) No evidence found KAR 82-3-114(c ), KAR 82-3-114(d) KAR 82-3-114(e) No evidence found No evidence found No evidence found No evidence found KAR 82-3-113(b) KAR 82-33-111(c) KAR 82-3-112 KAR 82-3-117, KAR 82-3-118 KAR 82-3-1305 No evidence found Notes: KSA: Kansas Statute Annotated; KAR: Kansas Administrative Regulations. www.rff.org 61 Resources for the Future Ho et al. Kentucky Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or special liability provisions KRS 353.590(7-9) KRS 353.590(12, 13, 17) KRS 353.590(18-22) 805 KAR 1:170(2.3, 3, 4) KRS 353.590(12, 13) KRS 353.590(7-9, 12, 17) 805 KAR 1:050; 805 KAR 1:170reg(7); KRS 353.590(25) Temporary Abandonment NAI TA 805 KAR 1:060(1) TA time limit 805 KAR 1:060(1) Well idle time limit KRS 353.150(1) TA future economic viability requirement No evidence found TA technical shut-in requirement 805 KAR 1:060(1) TA well integrity demonstration No evidence found Plugging and Restoration Plugging requirements 805 KAR 1:060; 805 KAR 1:070 Special treatment of casing removal No evidence found Special plugging requirements by well type 805 KAR 1:070 Cement standards 805 KAR 1:070(6) Conversion to freshwater wells 805 KAR 1:060(5) Marking after permanent abandonment No evidence found Restoration requirements 805 KAR 1:170 Regulator Notification, Approval, and Inspection (NAI) NAI post restoration 805 KAR 1:170(6) NAI pre plugging 805 KAR 1:060(2); 805 KAR 1:070(2); 805 KAR 1:080(5); 805 KAR 1:130reg NAI post plugging 805 KAR 1:060(4) Other Ability of regulator to order plugging KRS 353.739; 805 KAR 1:070(1.2); 805 KAR 1:060(6); KRS 353.180 Reporting requirements of operator and/or KAR 1:180(i) regulator Protection against fugitive methane at inactive sites KRS 353.140, 150, 160 Consideration of inactive wells during new well KAR 1:080 permitting Notes: KRS: Kentucky Revised Statutes; KAR: Kentucky Administrative Regulations. www.rff.org 62 Resources for the Future Ho et al. Louisiana Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or special liability provisions Temporary Abandonment NAI TA LAC 43:XIX§104.C.1.a; LAC 43:XIX§104.C.2.a LAC 43:XIX§104.C.1.b; LAC 43:XIX§104.C.2.b LAC 43:XIX§104.B LAC 43:I§3901 LAC 43:XIX§703 LAC 43:XIX§104.C LAC 43:I§2701, 2703 LAC 43:XIX§137.A.3.a; LAC 43:XIX§137.A.4 LAC 43:XIX§137.A.2 LAC 43:XIX§137.A.3.a LAC 43:XIX§137.A.3.b LAC 43:XIX§137.H No evidence found TA time limit Well idle time limit TA future economic viability requirement TA technical shut-in requirement TA well integrity demonstration Plugging and Restoration Plugging requirements LAC 43:XIX§137.F.3.a-f Special treatment of casing removal LAC 43:XIX§137.F.3.j Special plugging requirements by well type No evidence found Cement standards No evidence found Conversion to freshwater wells LAC 43:XIX§137.G Marking after permanent abandonment No evidence found Restoration requirements LAC 43:I§2101-3101 Regulator Notification, Approval, and Inspection (NAI) NAI post restoration LAC 43:I§2503 NAI pre plugging LAC 43:XIX§137.F.1, 2, 3k NAI post plugging LAC 43:XIX§137.F.4 Other Ability of regulator to order plugging LAC 43:XIX§137.C; LAC 43:I§2501 Reporting requirements of operator and/or LAC 43:XIX§137.A.3.b,c regulator Protection against fugitive methane at inactive No evidence found sites Consideration of inactive wells during new well No evidence found permitting Notes: LAC: Louisiana Administrative Code. www.rff.org 63 Resources for the Future Ho et al. Missouri Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics 10 CSR 50-2.020(1) 10 CSR 50-2.020(1) 10 CSR 50-2.020(2) No evidence found 10 CSR 50-2.030(9); 10 CSR 50-2.020(1)(A) 10 CSR 50-2.020(1) Liens or special liability provisions No evidence found Temporary Abandonment NAI TA 10 CSR 50-2.040(5) TA time limit 10 CSR 50-2.040(5) Well idle time limit 10 CSR 50-2.040(5) TA future economic viability requirement 10 CSR 50-2.040(5) TA technical shut-in requirements 10 CSR 50-2.040(5, 6) TA well integrity demonstration 10 CSR 50-2.040(6) Plugging and Restoration Plugging requirements 10 CSR 50-2.060(3)(C) Special treatment of casing removal 10 CSR 50-2.060(3)(C)(4) Special plugging requirements by well type No evidence found Cement standards 10 CSR 50-2.060(2, 4); 10 CSR 50-1.030(C)(3) Conversion to freshwater wells 10 CSR 50-2.060(6) Marking after permanent abandonment No evidence found Restoration requirements 10 CSR 50-2.060(3)(A)(1) Regulator Notification, Approval, and Inspection (NAI) NAI post restoration No evidence found NAI pre plugging 10 CSR 50-2.060(1, 2) NAI post plugging 10 CSR 50-2.060(7) Other Ability of regulator to order plugging 10 CSR 50-2.060(3)(A)(2, 4) Reporting requirements of operator and/or 10 CSR 50-2.080(2, 3) regulator Protection against fugitive methane at inactive No evidence found sites Consideration of inactive wells during new well 10 CSR 50-1.030(1)(A)(3, 4); 10 CSR 50permitting 2.030(2) Note: CSR: Code of State Regulations. www.rff.org 64 Resources for the Future Ho et al. Montana Financial Assurance Individual Bond ARM 36.22.1308.1.a Blanket Bond ARM 36.22.1308.1.b,c Types of financial assurances allowed ARM 36.22.1308.6,7 Use of surface damage agreements No evidence found Bond varies by operator characteristics ARM 36.22.1308.3,5; MCA 82.10.402 Bond varies by well characteristics ARM 36.22.1308 Liens or special liability provisions No evidence found Temporary Abandonment NAI TA ARM 36.22.1240 TA time limit ARM 36.22.1240; ARM 36.22.1303 Well idle time limit ARM 36.22.1240 TA future economic viability requirement ARM 36.22.1240; ARM 36.22.1303 TA technical shut-in requirement No evidence found TA well integrity demonstration No evidence found Plugging and Restoration Plugging requirements ARM 36.22.1303 Special treatment of casing removal ARM 36.22.1306 Special plugging requirements by well type No evidence found Cement standards No evidence found Conversion to freshwater wells ARM 36.22.1305 Marking after permanent abandonment ARM 36.22.1304 Restoration requirements ARM 36.22.1307 Regulator Notification, Approval, and Inspection (NAI) NAI post restoration No evidence found NAI pre plugging ARM 36.22.1301; ARM 36.22.1302; MCA 82.10.401 NAI post plugging ARM 36.22.1301; ARM 36.22.1309; ARM 36.22.1241 Other Ability of regulator to order plugging No evidence found Reporting requirements of operator and/or MCA 82.10.402 regulator Protection against fugitive methane at inactive No evidence found sites Consideration of inactive wells during new well No evidence found permitting Notes: ARM: Administrative Rules of Montana; MCA: Montana Code Annotated. www.rff.org 65 Resources for the Future Ho et al. Nebraska Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator (bad, idle) Bond varies by well charateristics Liens or special liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability requirement TA technical shut-in requirement TA well integrity demonstration Plugging and Restoration Plugging requirements Special treatment of casing removal Special plugging requirements by well type Cement standards Conversion to freshwater wells Marking after permanent abandonment Restoration requirements Regulator Notification, Approval, and Inspection NAI post restoration NAI pre plugging NAI post plugging Other Ability of regulator to order plugging Reporting requirements of operator and/or regulator Protection against fugitive methane at inactive sites Consideration of inactive wells during new well permitting NAC Title 267 Chapter 3 004 Form 3A NAC Title 267 Chapter 3 004 Form 3A NAC Title 267 Chapter 3 004 Form 3A No evidence found NAC Title 267 Chapter 3 004 Form 3A; NAC Title 267 Chapter 3 040.03 NAC Title 267 Chapter 3 004 Form 3A NAC Title 267 Chapter 3 030 NAC Title 267 Chapter 3 040.02 NAC Title 267 Chapter 3 040.01 NAC Title 267 Chapter 3 040 NAC Title 267 Chapter 3 040.01 No evidence found NAC Title 267 Chapter 3 040.01 NAC Title 267 Chapter 3 028 NAC Title 267 Chapter 3 028.04 No evidence found NAC Title 267 Chapter 3 012.04 NAC Title 267 Chapter 2 007 No evidence found NAC Title 267 Chapter 3 012.14-15; NAC Title 267 Chapter 3 028.08 NAC Title 267 Chapter 3 028.08 NAC Title 267 Chapter 3 028.06 NAC Title 267 Chapter 3 007 Form 6 No evidence found No evidence found No evidence found No evidence found Note: NAC: Nebraska Administrative Code. www.rff.org 66 Resources for the Future Ho et al. New Mexico Financial Assurance Individual Bond NMC §19.15.8.9D(2-4) Blanket Bond NMC §19.15.8.9D(1) Types of financial assurances allowed NMC §19.15.8.9A; NMC §19.15.10, 11, 15 Use of surface damage agreements NMC §19.15.8.9B Bond varies by operator characteristics NMC §19.15.8.9C, D(5) bond varies by well characteristics NMC §19.15.8.9D Liens or special liability provisions No evidence found Temporary Abandonment NAI TA NMC §19.15.25.13; NMC §19.15.7.14.A(d) TA time limit NMC §19.15.25.12 Well idle time limit NMC §19.15.25.8 TA future economic viability requirement No evidence found TA technical shut-in requirement No evidence found TA well integrity demonstration NMC §19.15.25.14 Plugging and Restoration Plugging requirements NMC §19.15.25.10.A Special treatment of casing removal No evidence found Special plugging requirements by well type No evidence found Cement standards No evidence found Conversion to freshwater wells NMC §19.15.25.15 Marking after permanent abandonment NMC §19.15.25.10.B, C Restoration requirements NMC §19.15.25.10.D Regulator Notification, Approval, and Inspection (NAI) NAI post restoration NMC §19.15.25.10.F NAI pre plugging NMC §19.15.7.14.A(1e, 2) NAI post plugging NMC §19.15.7.14.E, F Other Ability of regulator to order plugging NMC §19.15.5.10.B(4, 7) Reporting requirements of operator and/or NMC §19.15.5.9 regulator Protection against fugitive methane at inactive No evidence found sites Consideration of inactive wells during new well No evidence found permitting Note: NMC: New Mexico Code. www.rff.org 67 Resources for the Future Ho et al. New York Financial Assurance Individual Bond 6 CCR-NY 551.5.a.1.i; 6 CCR-NY 551.5.a.2.i Blanket Bond 6 CCR-NY 551.5.a.1, 2 Types of financial assurances allowed 6 CCR-NY 551.4.b Use of surface damage agreements No evidence found Bond varies by operator characteristics No evidence found Bond varies by well characteristics 6 CCR-NY 551.5; 6 CCR-NY 551.6 Liens or special liability provisions 11 CCR-NY 1101.3.e,f Temporary Abandonment NAI TA 6 CCR-NY 555.3.b; 6 CCR-NY 555.2.a TA time limit 6 CCR-NY 555.3.b; 6 CCR-NY 555.2.a Well idle time limit 6 CCR-NY 555.3.a TA future economic viability requirement 6 CCR-NY 555.3.b TA technical shut-in requirement No evidence found TA well integrity demonstration No evidence found Plugging and Restoration Plugging requirements 6 CRR-NY 555.5 Special treatment of casing removal 6 CRR-NY 555.5 Special plugging requirements by well type No evidence found Cement standards No evidence found Conversion to freshwater wells 6 CRR-NY 555.6 Marking after permanent abandonment No evidence found Restoration requirements No evidence found Regulator Notification, Approval, and Inspection (NAI) NAI post restoration No evidence found NAI pre plugging 6 CRR-NY 555.4 NAI post plugging 6 CRR-NY 555.5.d Other Ability of regulator to order plugging 11 CCR-NY 1101.3.j Reporting requirements of operator and/or No evidence found regulator Protection against fugitive methane at inactive No evidence found sites Consideration of inactive wells during new well No evidence found permitting Note: CCR-NY: New York Codes, Rules and Regulations. www.rff.org 68 Resources for the Future Ho et al. North Dakota Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage assessments Bond varies by operator characteristics Bond varies by well characteristics NDAC 42-02-03-15-2 NDAC 42-02-03-15-2 NDAC 43-02-03-15-1; NDAC 42-02-0315-8 No evidence found NDAC 42-03-03-13-2 No evidence found Liens or special liability provisions No evidence found Temporary Abandonment NAI TA NDAC 43-02-03-55 TA time limit NDAC 43-02-03-55 Well idle time limit NDAC 43-02-03-55 TA future economic viability requirement No evidence found TA technical shut-in requirement NDAC 43-02-03-55 TA well integrity demonstration NDAC 43-02-03-55 Plugging and Restoration Plugging requirements NDAC 43-02-03-04 Special treatment of casing removal NDAC 43-02-0 Special plugging requirements by well type NDAC 43-02-03-24 Cement standards NDAC 43-02-03-34 Conversion to freshwater wells NDAC 43-02-03-35 Marking after permanent abandonment No evidence found Restoration requirements NDAC 43-02-03-19 Regulator Notification, Approval, and Inspection (NAI) NAI post restoration NDAC 42-02-03-13-7 NAI pre plugging NDAC 42-02-03-33 NAI post plugging NDAC 42-02-03-13-7 Other Ability of regulator to order plugging No evidence found Reporting requirements of operator and/or No evidence found regulator Protection against fugitive methane at inactive sites No evidence found Consideration of inactive wells during new well NDAC 43-02-03-15-2 permitting Note: NDAC: North Dakota Administrative Code. www.rff.org 69 Resources for the Future Ho et al. Ohio Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability TA technical shut-in requirement TA well integrity demonstration Plugging and Restoration Plugging requirements Special treatment of casing removal OAC 1501:9-1-03(A) OAC 1501:9-1-03(A) OAC 1501:9-1-03(F) No evidence found No evidence found No evidence found ORC 1509.07(A) ORC 1509.062.C ORC 1509.062.G ORC 1509.062.A ORC 1509.062.B.5 ORC 1509.062.C ORC 1509.062.B.6; ORC 1509.062.C OAC 1501:9-11-08; OAC 1501:9-11-09 OAC 1501:9-11-06; OAC 1501:9-11-10; OAC 1501:911-08.F; OAC 1501:9-11-09.B Special plugging requirements by well type OAC 1501:9-11-08.A.6; ORC 1571.05 Cement standards OAC 1501:9-11-01.M Conversion to freshwater wells OAC 1501:9-11-13 Marking after permanent abandonment OAC 1501:9-11-10 Restoration requirements OAC 1509.072; OAC 1501:9-1-02.B, C Regulator Notification, Approval, and Inspection (NAI) NAI post restoration No evidence found NAI pre plugging OAC 1501:9-11-02; OAC 1501:9-11-04 NAI post plugging OAC 1509.14 Other Ability of regulator to order plugging ORC 1509.12; ORC 1509.04.C Reporting requirements of operator and/or No evidence found regulator Protection against fugitive methane at inactive No evidence found sites Consideration of inactive wells during new well No evidence found permitting Notes: OAC: Ohio Administrative Code; ORC: Ohio Revised Code. www.rff.org 70 Resources for the Future Ho et al. Oklahoma Financial Assurance Individual Bond Blanket Bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or special liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability TA technical shut-in requirement TA well integrity demonstration Plugging and Restoration Plugging requirements Special treatment of casing removal Special plugging requirements by well type Cement standards Conversion to freshwater wells Marking after permanent abandonment Restoration requirements Regulator Notification, Approval, and Inspection NAI post restoration NAI pre plugging NAI post plugging Other Ability of regulator to order plugging Reporting requirements of operator and/or regulator Protection against fugitive methane at inactive sites Consideration of inactive wells during new well permitting OS §17-518(A) OS §17-518(A) OAC 165:10-1-10-(a)(1), (b); OAC 165:10-1-11 OS §17-519, 520 OS §52-318.1(C); OAC 165-10-1-10(d) OS §17-518(A) OAC 165:10-11-3(b) OAC 165:10-11-9 OAC 165:10-11-9.c.2 OS §68-1—1.F.2 No evidence found OAC 165:10-11-9.d OAC 165:10-11-9.c.4 OAC 165:10-11-6; OAC 165:10-11-3 OAC 165:10-11-6.k OS §52-308 No evidence found OAC 165:10-11-6.p OAC 165:10-3-4.f; OAC 165:10-3-17.j OAC 165:10-3-17.k-n; OAC 165:10-7-2.d; OS §17-53.2 No evidence found OAC 165:10-11-1, 2, 4 OAC 165:10-11-7 OAC 165:10-1-10.g; OS §17-53; OAC 165:10-114; OS §52-309 No evidence found OAC 165:10-12 No evidence found Notes: OS: Oklahoma Statues; OAC: Oklahoma Administrative Code. www.rff.org 71 Resources for the Future Ho et al. Pennsylvania Financial Assurance Individual bond Blanket bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or special liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability TA technical shut-in requirement TA well integrity demonstration 58 Pa. Code §3225(a)(1) 58 Pa. Code §3225(a)(2): 25 Pa. Code §78.307-309; 58 Pa. Code §3225(a)(3); 58 Pa. Code §3225(d)(1) No evidence found No evidence found 58 Pa. Code §3225(a)(1, 2) 58 Pa. Code §3225(a)(3); 58 Pa. Code §3220(a) 58 Pa. Code §3214(a) 25 Pa. Code §78.104 No evidence found 25 Pa. Code §78.102(4) 25 Pa. Code §78.102(2)(C.) 25 Pa. Code §78.103; 25 Pa. Code §78.102(2); 25 Pa. Code §78.903(8) Plugging and Restoration Plugging requirements Special treatment of casing removal Special plugging requirements by well type 25 Pa. Code §78.91, 92, 93, 94, 95 25 Pa. Code §78.91(d) 25 Pa. Code §78.92-93; 58 Pa. Code §3220(b); 25 Pa. Code §78.91(g) Cement standards No evidence found Conversion to freshwater wells No evidence found Marking after permanent abandonment 25 Pa. Code §78.96 Restoration requirements 58 Pa. Code §3216 Regulator Notification, Approval, and Inspection (NAI) NAI post restoration 25 Pa. Code §78.65; 25 Pa. Code §78.903(10); 25 Pa. Code §78.903(11) NAI pre plugging 58 Pa. Code §3211(f)(2); 58 Pa. Code §3220(c, d); 25 Pa. Code §78.903(9) NAI post plugging 25 Pa. Code §78.124 Other Ability of regulator to order plugging 58 Pa. Code §3214.e; 58 Pa. Code §3220(e.); 25 Pa. Code §78.13 Reporting requirements of operator and/or 58 Pa. Code §3222 regulator Protection against fugitive methane at inactive 25 Pa. Code §78.102(3); §78.102(B)(2)(ii)(D) sites Consideration of inactive wells during new well No evidence found permitting Note: Pa. Code: Pennsylvania Code. www.rff.org 72 Resources for the Future Ho et al. Texas Financial Assurance Individual bond Blanket bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or special liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability TA technical shut-in requirements TA well integrity demonstration Plugging and Restoration Plugging requirements Special treatment of casing removal Special plugging requirements by well type Cement standards Conversion to freshwater wells Marking after permanent abandonment Restoration requirements Regulator Notification, Approval, and Inspection (NAI) NAI post restoration NAI pre plugging NAI post plugging Other Ability of regulator to order plugging Reporting requirements of operator and/or regulator Protection against fugitive methane at inactive sites Consideration of inactive wells during new well permitting TAR 16-1-3-3.78(g) TAR 16-1-3-3.78(g) TAR 16-1-3-3.78(d-f) No evidence found TAR 16-1-3-3.78(b)(9) TAR 16-1-3-3.78(g); TAR 16-1-3-3.78(b) TAR 16-1-3-3.78(k) No evidence found TAR 16-1-3-3.15(d)(1); TAR 16-1-3-3.15(e ) TAR 16-1-3-3.14(b)(2) TAR 16-1-3-3.15(j) No evidence found TAR 16-1-3-15(l) TAR 16-1-3-3.14(d)(2,3, and 8) No evidence found TAR 16-1-3-3.14(e-k) TAR 16-1-3-3.14(d)(4 and 9) TAR 16-1-3-3.14(a)(4) No evidence found TAR 16-1-3-3-.14(d)(12) No evidence found TAR 16-1-3-3.15(l-m) TAR 16-1-3-3.14(b)(1) TAR 16-1-3-3.15(b)(3) TAR 16-1-3-3.15(i)(5) No evidence No evidence Note: TAR: Texas Administrative Code. www.rff.org 73 Resources for the Future Ho et al. Utah Financial Assurance Individual bond Blanket bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or special liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability requirement TA technical shut-in requirement TA well integrity demonstration Plugging and Restoration Plugging requirements Special treatment of casing removal Special plugging requirements by well type Cement standards Conversion to freshwater wells Marking after permanent abandonment Restoration requirements Regulator Notification, Approval, and Inspection NAI post restoration NAI pre plugging NAI post plugging Other Ability of regulator to order plugging Reporting requirements of operator and/or regulator Protection against fugitive methane at inactive sites Consideration of inactive wells during new well permitting UAC R649-3-1-5.1-5.4 UAC R649-3-1-6.2, 6.3 UAC R649-3-10 UAC R649-3-38-6 UAC R649-3-1-4.3, 4.4 UAC R649-3-1 UAC R649-3-1-14 UAC R649-3-36-2 UAC R649-3-36-3 UAC R649-3-36-1 UAC R649-3-36-1.1 No evidence found UAC R649-3-36-1.3 UAC R649-3-24-3 UAC R649-3-24-5.3, 5.5, 8 UAC R649-3-28; UAC R649-3-31; UAC R649-3-30 No evidence found UAC R649-3-24-6 UAC R649-3-24-7 UAC R649-3-34 UAC R649-3-15; UAC R649-3-34-17 UAC R649-3-24-1 UAC R649-3-24-5 No evidence found UAC R649-3-6 No evidence found No evidence found Note: UAC: Utah Administrative Code. www.rff.org 74 Resources for the Future Ho et al. West Virginia Financial Assurance Individual bond WVC §22-6-26(b) Blanket bond WVC §22-6-26(c) Types of financial assurances allowed WVC S22-6-26(d)(e) Use of surface damage agreements WVC §22-6A-16; WVC §22-7 Bond varies by operator characteristics WVC §22-6-6(h) Bond varies by well characteristics WVC §22-6-26; WVC §22-6A-7(g); WVC §22-6A-15 Liens or special liability provisions WVC §22-10-7, 8, 9 Temporary Abandonment NAI TA WVC §35-5-5.2 TA time limit WVC §35-5-5.4 Well idle time limit WVC §22-6-19; 35-5-2.2 TA future economic viability WVC §22-6-19; 35-5-3, 4 TA technical shut-in requirement No evidence found TA well integrity demonstration WVC §35-5-5.3 Plugging and Restoration Plugging requirements WVC §22-6-24 Special treatment of casing removal WVC §22-6-19 Special plugging requirements by casing type WVC §22-6-24(b)(c)(d)(e.); WVC §22-6A-13 Cement standards WVC §22-6-1(h) Conversion to freshwater wells No evidence found Marking after permanent abandonment No evidence found Restoration requirements WVC §22-6-30; WVC §22-6A-14 Regulator Notification, Approval, and Inspection (NAI) NAI post restoration No evidence found NAI pre plugging WVC §22-6-6(c)(10); WVC §22-6-23 NAI post plugging WVC §22-6-23 Other Ability of regulator to order plugging WVC §22-6-26(e); WVC §22-6-14(d) Reporting requirements of operator and/or WVC §22-6-29(b) regulator Protection against fugitive methane at inactive WVC §22-6-31(b); WVC §22-6-32 sites Consideration of inactive wells during new well WVC §22-6-14(a)(1) permitting Note: WVC: West Virginia Code. www.rff.org 75 Resources for the Future Ho et al. Wyoming Financial Assurance Individual bond Blanket bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics Liens or special liability provisions Temporary Abandonment NAI TA TA time limit Well idle time limit TA future economic viability requirement TA technical shut-in requirement TA well integrity demonstration Plugging and Restoration Plugging requirements Special treatment of casing removal Wyo. Stat. Ann. §30-5-404; Rule 3.4.a.i, ii Rule 3.4.a.iii Rule 3.5, 3.6 Wyo. Stat. Ann. §30-5-402(c, e); Wyo. Stat. Ann. §30-5-403; Wyo. Stat. Ann. §30-5-406(a); Rule 3.4.(ik) Rule 3.4.(c-e) Rule 3.4 Rule 3.4.f; Rule 3.7; Rule 3.14 Rule 3.16.a Rule 3.16.b No evidence found No evidence found Rule 1.2.eee Rule 3.16.c, d Rule 3-18 Rule 3.18.b.iii.C; Rule 3.18.b.iii.E Special plugging requirements by well type Rule 3.18.c; Rule 3.18.b.iv Cement standards Rule 3.18.b.i Conversion to freshwater wells Rule 3.15.b Marking after permanent abandonment Rule 3.19.a.5 Restoration requirements Rule 3.17.b-d; Rule 3.7.a Regulator Notification, Approval, and Inspection (NAI) NAI post restoration Rule 3.17.c NAI pre plugging Rule 3.15 NAI post plugging Rule 3.17.a Other Ability of regulator to order plugging Rule 3.36 Reporting requirements of operator and/or Rule 3.16.a; regulator Protection against fugitive methane at inactive No evidence found sites Consideration of inactive wells during new No evidence found well permitting Notes: Wyo. Stat. Ann.: Wyoming Statutes Annotated; Rule: Wyoming Oil and Gas Conservation Commission Rules. www.rff.org 76 Resources for the Future Ho et al. Bureau of Land Management Financial Assurance Individual bond Blanket bond Types of financial assurances allowed Use of surface damage agreements Bond varies by operator characteristics Bond varies by well characteristics 43 CFR 3104.2 43 CFR 3104.3 43 CFR 3104.1.b,c OOGO III.D.4; OOGO VI 43 CFR 3104.5; OOGO III.D.5.a 43 CFR 3104.5; OOGO III.D.5.a Liens or special liability provisions No evidence found Temporary Abandonment NAI TA 43 CFR 3162.3-4.c TA time limit 43 CFR 3162.3-4.c Well idle time limit 43 CFR 3162.3-4.c TA future economic viability requirement IM 2012-181 TA technical shut-in requirement IM 2012-181 TA well integrity demonstration IM 2012-181 Plugging and Restoration Plugging requirements 43 CFR 3162.3-4.a Special treatment of casing removal No evidence found Special plugging requirements by well type No evidence found Cement standards No evidence found Conversion to freshwater wells 43 CFR 3162.3-4.b; OOGO IX.B Marking after permanent abandonment 43 CFR 3162.6.d Restoration requirements OOGO II.4.D.j; OOGO XII.B Regulator Notification, Approval, and Inspection (NAI) NAI post restoration GAO 2011 NAI pre plugging 43 CFR 3162.3-4.a; OOGO XII.A NAI post plugging OOGO XII.A Other Ability of regulator to order plugging IM 2012-181 Reporting requirements of operator and/or 43 CFR 3162.4-3; IM 2012-181 regulator Protection against fugitive methane at inactive No evidence found sites Consideration of inactive wells during new No evidence found well permitting Notes: CFR: Code of Federal Regulations; IM: Instruction Memorandum; OOGO: Onshore Oil and Gas Order No. 1 (2007). www.rff.org 77 Resources for the Future Ho et al. Acknowledgements The authors would like to thank the Paul G. Allen Family Foundation for funding this project and, in particular, Courtney Blodgett for providing helpful and extensive comments on the draft. We also want to acknowledge the help of RFF’s Jessica Chu, research associate; research assistants Elaine Swiedler and Alexandra Thompson; plus Jan Mares, senior policy advisor. We also thank Pat Payne (Alberta Orphan Well Association), Rod Smith (Schlumberger), Dale Kunz (Winterhawk, Inc.), John Reitsma (Bureau of Land Management National Operation Center), Patrick Shields (Kansas Corporation Commission) and Lane Palmateer (Kansas Corporation Commission). We offer special thanks to DrillingInfo for making its database available to us. Finally, we thank our many expert reviewers and state officials for their very helpful comments and suggestions. References API (American Petroleum Institute). 1993. Environmental Guidance Document: Well Abandonment and Inactive Well Practices for US Exploration and Production Operations. 31 Jan. ———. 2009. Environmental Protection for Onshore Oil and Gas Production Operations and Leases. July. http://www.api.org/~/media/Files/Polic y/Exploration/API_RP_51R.pdf (accessed 23 Sept. 2015). Bishop, Ronald E. 2013. “Historical analysis of oil and gas well plugging in New York: Is the regulatory system working?.” New Solutions: A Journal of Environmental and Occupational Health Policy 23(1): 103–116. Bonett, Art, and Demos Pafitis. 1996. “Getting to the root of gas migration.” Oilfield Review 8(1): 36–49. Boyer, Gerry. 2015. Vent Flow/Gas Migration Data Trends in the Western Provinces. Industry Operations Branch, Alberta Energy Regulator. May 13. Brandt, A.R., G.A. Heath, E.A. Kort, F. O’Sullivan, G. Pétron, S.M. Jordaan, P. Tans, J. Wilcox, A.M. Gopstein, D. Arent, S. Wofsy, N.J. Brown, R. Bradley, G.D. Stucky, D. Eardley, and R. Harriss. 2014. “Methane Leaks from North American Natural Gas Systems.” Science 343: 733–735. Brown, Katie. 2014. “New Ingraffea Report: Same Old Tactics.” Web log post. Energy in Depth, 10 July. http://energyindepth.org/national/newingraffea-report-same-old-tactics/ (accessed 15 Feb. 2015). Calvert, D.G., and Dwight K. Smith.1994. “Issues and techniques of plugging and abandonment of oil and gas wells.” SPE Annual Technical Conference and Exhibition. Society of Petroleum Engineers. Chafin, Daniel T. 1994. Sources and Migration Pathways of Natural Gas in NearSurface Ground Water Beneath the Animas River Valley, Colorado and New Mexico. No. 94-4006. US Geological Survey. Davis, Lucas. 2015. “Bonding requirements for US natural gas producers.” Review of Environmental Economics and Policy 9(1): 128–144. Davies, Richard J., Sam Almond, Robert S. Ward, Robert B. Jackson, Charlotte Adams, Fred Worrall, Liam G. Herringshaw, Jon G. Gluyas, and Mark A. Whitehead. 2014. “Oil and gas wells and their integrity: implications for shale and unconventional resource exploitation.” Marine and Petroleum Geology 56: 239–254. Derksen, Cheyenne. 2015. “Deserted Wells Create Kansas Sinkholes.” McPherson www.rff.org 78 Resources for the Future Ho et al. Sentinel, 6 July. http://www.mcphersonsentinel.com/arti cle/20150706/NEWS/150709731 (accessed 23 Sept. 2015). DOI (US Department of the Interior). 2015. DOI (US Department of the Interior). 2015. US Fish and Wildlife Service’s Management of Oil and Gas Activities on Refuges. Report No. CR-EV-FWS0002-2014. Washington, DC: DOI. http://www.fws.gov/policy/Manageme ntofOilandGasActivitiesHandbook.pdf (accessed 24 Sept. 2015). Drohan, P.J., M. Brittingham., J. Bishop, and K. Yoder. 2012. “Early trends in landcover change and forest fragmentation due to shale-gas development in Pennsylvania: a potential outcome for the Northcentral Appalachians.” Environmental Management 49(5): 1061–1075. Dusseault, Maurice, and Richard Jackson. 2014. “Seepage pathway assessment for natural gas to shallow groundwater during well stimulation, in production, and after abandonment.” Environmental Geosciences 21(3): 107–126. Dusseault, Maurice B., Malcolm N. Gray, and Pawel A. Nawrocki. 2000. “Why oilwells leak: cement behavior and longterm consequences.” International Oil and Gas Conference and Exhibition in China. Society of Petroleum Engineers. http://www.eia.gov/dnav/ng/ng_prod_s um_a_EPG0_FGW_mmcf_m.htm (accessed 15 Aug. 2015). Erno, B., and R. Schmitz. 1996. “Measurements of Soil Gas Migration Around Oil and Gas Wells in the Lloydminster Area.” Journal of Canadian Petroleum Technology 35(7): 37–46. Frosch, Dan, and Russell Gold. 2015. “How ‘Orphan’ Wells Leave States Holding the Cleanup Bag.” Wall Street Journal, 25 Feb. http://www.wsj.com/articles/howorphan-wells-leave-states-holding-thecleanup-bag-1424921403 (accessed 23 Sept. 2015). GAO (US Government Accountability Office). 2010. Oil and Gas Bonds—Bonding Requirements and BLM Expenditures to Reclaim Orphaned Wells. GAO-10245. Washington, DC; GAO. http://www.gao.gov/assets/310/300218 .pdf (accessed 24 Sept. 2015). ———. 2011. “Oil and Gas Bonds: BLM Needs a Comprehensive Strategy to Better Manage Potential Oil and Gas Well Liability.” Feb 25. GAO-11-292. Washington, DC. http://www.gao.gov/new.items/d11292. pdf (accessed 24 Sept. 2015). Dusterhoft, Dale, Greg Wilson, and Ken Newman. 2002. “Field study on the use of cement pulsation to control gas migration.” SPE Gas Technology Symposium. Society of Petroleum Engineers. Gaspar, J., J. Mathieu, Y. Yang, R. Tomson, J.D. Leyris, K.B. Gregory, and P.J.J. Alvarez. 2014. “Microbial Dynamics and Control in Shale Gas Production.” Environmental Science and Technology Letters 1: 465–473. EIA (US Energy Information Administration). 2015a. “Petroleum & Other Liquids: Crude Oil Production.” https://www.eia.gov/dnav/pet/pet_crd_ crpdn_adc_mbblpd_a.htm (accessed 15 Aug. 2015). GWPC (Ground Water Protection Council). 2014. “State oil and natural gas regulations designed to protect water resources.” Prepared for the US Department of Energy, National Energy Technology Laboratory. ———. 2015b. “Natural Gas: Natural Gas Gross Withdrawals and Production.” www.rff.org 79 Resources for the Future Ho et al. Ingraffea, Anthony R., Martin T. Wells, Renee L. Santoro, and Seth B.C. Shonkoff. 2014. “Assessment and risk analysis of casing and cement impairment in oil and gas wells in Pennsylvania, 2000– 2012.” Proceedings of the National Academy of Sciences 111(30): 10955– 10960. IOGCC (Interstate Oil and Gas Compact Commission) 2008. Protecting Our Country's Resources: The States’ Case. http://iogcc.publishpath.com/Websites/ iogcc/pdfs/2008-Protecting-OurCountry's-Resources-The-States'Case.pdf (accessed May 2015). http://iogcc.publishpath.com/Websites/ iogcc/pdfs/2008-Protecting-OurCountry's-Resources-The-States'Case.pdf (accessed 12 May 2015). Jackson, R.E., A.W. Gorody, B. Mayer, J.W. Roy, M.C. Ryan, and D.R. Van Stempvoort. 2013. “Groundwater Protection and Unconventional Gas Extraction: The Critical Need for FieldBased Hydrogeological Research.” Groundwater 51(4): 488–510. Kang, Mary, Cynthia M. Kanno, Matthew C. Reid, Xin Zhang, Denise L. Mauzerall, Michael A. Celia, Yuheng Chen, and Tullis C. Onstott. 2014. “Direct Measurements of Methane Emissions from Abandoned Oil and Gas Wells in Pennsylvania.” Proceedings of the National Academy of Sciences 111(51): 18173– 18177. Kang, Mary, Ejeong Baik, Alana R. Miller, Karl W. Bandilla, and Michael A. Celia. 2015. “Effective Permeabilities of Abandoned Oil and Gas Wells: Analysis of Data from Pennsylvania.” Environmental Science & Technology 49(7): 4757–4764. KCC (Kansas Corporation Commission). 2015. Abandoned Oil and Gas Well Status Report. Topeka, KS: KCC. Kell, Scott. 2011. State Oil and Gas Agency Groundwater Investigations and Their Role in Advancing Regulatory Reform; a Two-state Review, Ohio and Texas. Ground Water Protection Council. King, George E., and Daniel E. King. 2013. “Environmental risk arising from wellconstruction failure—differences between barrier and well failure, and estimates of failure frequency across common well types, locations, and well age.” SPE Production & Operations 28(4): 323–344. King, George Everette, and Randy L. Valencia. 2014. “Environmental risk and well integrity of plugged and abandoned wells.” SPE Annual Technical Conference and Exhibition. Society of Petroleum Engineers. Krupnick, A., N. Richardson, and M. Gottlieb. 2015. “Heterogeneity of state shale gas regulations.” Economics of Energy & Environmental Policy 4(1). Kunz, Dale. 2015. Winterhawk, Inc., Calgary, Alberta, Canada. Personal Communication. 13 November. Lattanzio, Richard K. 2013. Air Quality Issues in Natural Gas Systems. Congressional Research Service. March 4. LLA (Louisiana Legislative Auditor). 2014. Regulation of Oil and Gas Wells and Management of Orphaned Wells. Office of Conservation—Department of Natural Resources. By Daryl G. Purpera, Paul E. Pendas, and Nicole B. Edmonson. N.p., 28 May. http://app.lla.state.la. us/PublicReports.nsf/0/D6A0EBE279B 83B9F86257CE700506EAD/$FILE/00 0010BC.pdf (accessed 24 Sept. 2015). Lyverse, Mark A., and Michael D. Unthank. 1988. Assessment of Ground-Water Contamination in the Alluvial Aquifer Near West Point, Kentucky. No. 884166. US Geological Survey. www.rff.org 80 Resources for the Future Ho et al. Mitchell, Austin L., and Elizabeth A. Casman. 2011. “Economic incentives and regulatory framework for shale gas well site reclamation in Pennsylvania.” Environmental Science & Technology 45(22): 9506–9514. Muehlenbachs, Lucija. 2015. “A Dynamic Model of Cleanup: Estimating Sunk Costs in Oil and Gas Production.” International Economic Review 56(1): 155–185. NPC (National Petroleum Council). 2011. “Plugging and Abandonment of Oil and Gas Wells.” Working Paper. Railroad Commission of Texas. 2000. “Well Plugging Primer.” January. http://www.rrc.state.tx.us/media/6358/ plugprimer1.pdf (accessed Jun. 2015). Richardson, N., M. Gottlieb, A. Krupnick, and H. Wiseman. 2013. The State of State Shale Gas Regulation. Washington, DC. Resources for the Future. ster. Report for Phase 2 (1995).” March. Walls, Margaret, Carolyn Kousky, and Ziyan Chu. 2015. “Is what you see what you get? The value of natural landscape views.” Land Economics 91(1): 1–19. Watson, Theresa Lucy, and Stefan Bachu. 2009. “Evaluation of the potential for gas and CO2 leakage along wellbores.” E&P Environmental and Safety Conference. Society of Petroleum Engineers. Zirogiannis, N., J. Alcorn, J. Rupp, S. Carley, and J.D. Graham. 2016. “State Regulation of Unconventional Gas Development in the US: An Empirical Evaluation.” Energy Research & Social Science 11: 142–154. Shields, Patrick. 2015. Kansas Corporation Commission. Personal Communication. 6 July. Shih, Jhih-Shyang, Jaqueline Ho, Lucija Muehlenbachs, Clayton Munnings, and Alan J. Krupnick. Forthcoming. The Costs and Fiscal Implications of Decommissioning Inactive Wells. RFF Discussion Paper. Washington, DC: Resources for the Future. Smith, Rod. 2015. Schlumberger, Ltd. Personal Communication. 27 August. Taylor, Michael (VP for Climate Policy Assurance, Alberta Energy Regulator, Alberta, Canada) 2016. Personal Communication. May 10. Van Stempvoort, D.R., E.J. Jaworski, and M. Rieser. Canadian Association of Petroleum Producers. 1995. “Migration of methane into ground water from leaking production wells near Lloydmin- www.rff.org 81 IIB RESOURCES FOR THE FUTURE Resources for the Future 1616 St. NW Washington, DC 20036 ?g