U.S. DOT/RSPA - Draft Final Regulatory Evaluation

Page 1U.S. Department of Transportation Research and Special Programs Administration Final Regulatory Evaluation Pipeline Integrity Management in High Consequence Areas (Gas Transmission Pipelines) Docket RSPA-00-7666#

Page 2INTRODUCTION The U.S. Department of Transportation Research and Special Programs Office of Pipeline Safety (RSPNOPS) is changing pipeline safety regulations to require operators of certain pipelines to validate the integrity of their pipelines in high consequence areas. The rule applies to operators of natural and other gas transmission pipelines. The objective of the change is to reduce the risk of pipeline incidents in these areas. High consequence areas are redefined in this rule. Under the redefinition, operators must choose one of two options to designate high consequence areas: A11 class 3 & 4 locations (these are areas where there are at least 46 buildings intended for human occupancy or where buildings with four or more stories above ground are prevalent within 660 feet of the pipeline along any continuous mile of its length) plus areas where a potential impact circle of radius greater than 660 feet includes 20 or more buildings intended for human occupancy, or Locations where the potential impact circle, of whatever size, includes 20 or more buildings intended for human occupancy. Under either option, pipeline operators must also include as high consequence areas any potential impact circle that contains: 8 A hospital, school, prison, day care center or other facility having persons who are confined or of limited mobility, or Outdoor locations or open structures where 20 or more persons congregate at least 50 days in any 12-month period, or 0 A building in which 20 or more persons gather 5 days a week for 10 weeks in any 12- month period (the days and weeks need not be consecutive). The potential impact circle used in any of these determinations is a circle, centered on the pipeline, of a radius calculated based on the size and pressure of the pipeline. The potential impact circle approximates the area that could be affected by a rupture and subsequent explosion occurring on the pipeline. To validate the integrity of their pipelines in high consequence areas under the regulatory change, pipeline operators must implement an integrity management program for such pipelines including periodic inspection and testing and integration of information related to pipeline integrity. The purpose of this report is to assess the benefits and costs of the regulatory change. This rule is similar to rules promulgated earlier for hazardous liquid pipeline operators. High consequence areas were defined differently for hazardous liquid pipelines, because the 1 . * .#

Page 3environmental consequences of leaks from hazardous liquid pipelines are different than those from natural gas pipelines. The elements of an integrity management program required by this rule are similar, however, to the elements previously required of hazardous liquid pipeline operators. This report considers the costs and benefits of these requirements in a manner similar to the analysis of costs and benefits prepared for the earlier rulemakings. TARGET PROBLEM Natural and other gas pipeline breaks can result in explosions and fires that can impact on human health and safety. The magnitude of this impact differs. There are some areas in which the impact of a pipe break will be more significant than it would be in others due to concentrations of people near the pipeline and who thus could be affected. Because of the potential for dire consequences of pipeline failures in certain areas, these areas merit a higher level of protection. The OPS is promulgating this regulation to afford the necessary additional protection to these “high consequence areas”. Numerous investigations by RSPNOPS and the National Transportation Safety Board (NTSB) have highlighted the importance of protecting the public from pipeline failures. The NTSB has made several recommendations to ensure the integrity of pipelines near populated areas. These recommendations included requiring periodic testing and inspection to identify corrosion and other damage, establishing criteria to determine appropriate intervals for inspections and tests, and determining hazards to public safety from electric resistance welded pipe. Congress also directed RSPNOPS to undertake additional safety measures in areas that are densely populated. These statutory requirements included having RSPNOPS prescribe standards for identifjmg pipelines in high density population areas and issue standards requiring periodic inspections using internal inspection devices on pipelines in densely-populated areas. This rulemaking addresses the target problem described above, and is a comprehensive response to the NTSB’s recommendations and Congressional mandates, as well as pipeline safety and environmental issues raised over the years. ALTERNATIVES CONSIDERED The OPS considered several alternatives to provide the necessary increased level of protection to high consequence areas. These alternatives were: 1. No action. 2. Publishing a rule that would adopt requirements of the Pipeline Safety Improvement Act of 2002 and establish procedures to seek waiver of reassessment intervals 3. Prescriptive requirements for inspection and repair of pipelines in high consequence areas. 4. Requiring pipeline operators to develop integrity management programs providing for inspection and testing based on risk factors and integration of information related to pipeline risk. 2#

Page 45 . Requiring pipeline operators to develop integrity management programs providing for expedited inspection and testing. INITIAL SCREENING OF ALTERNATIVES 1 . No action. Regulatory analyses typically consider an alternative in which the agency would not take any action, because it would maintain the status quo. No new requirements would be levied. No costs would be incurred to implement new requirements. No new benefits would result. In this case, however, the “no action” alternative does not maintain the status quo. The status quo has been changed by Act of Congress. The Pipeline Safety Improvement Act of 2002 (PSIA-2002), signed into law on December 17, 2002, imposes requirements directly on pipeline operators. To be sure, PSIA-2002 directs the Secretary of Transportation to publish, within one year, standards for integrity management plans that would require periodic assessment of pipelines in high consequence areas. PSIA-2002 goes on to require that operators of gas transmission pipelines, regardless of whether or not the Secretary publishes such standards, must conduct a risk analysis, implement integrity management programs, and begin baseline assessments of their pipeline facilities in high consequence areas within 18 months. Baseline assessments must be completed on all gas transmission pipeline segments in high consequence areas within 10 years. Each ‘gas transmission pipeline segment in a high consequence area must be reassessed at least every 7 years. The assessment requirements in PSIA-2002 apply to “each of the operator’s facilities in areas identified pursuant to subsection (a)( 1) [of 49 U.S.C. 0 601 091 and defined in chapter 192 of title 49, Code of Federal Regulations, including any subsequent modifications” (emphasis added)’. The cited provision of the U.S. Code is the legislative requirement that the DOT establish criteria for identifying pipelines in high-density population areas, or high consequence areas. The reference to the definition in title 49 of the Code of Federal Regulations is thus a reference to the definition of high consequence areas established by DOT. The applicability of the Act’s requirements to any “subsequent modifications” of that definition allows the DOT to revise its criteria and further focus the actions required of pipeline operators. Absent a change to the definition embodied in part 192, however, operators would be required to conduct assessments of all pipeline segments meeting the current definition. RSPNOPS, in fact, has been considering changes to the definition of high consequence areas in part 192 to further refine the criteria and to focus better on areas most at risk and for which special requirements are appropriate. Making those changes would reduce the amount of pipeline on which integrity assessments are required to be performed without significantly reducing the ‘Pipeline Safety Improvement Act of 2002, Section 14, “Risk Analysis and Integrity Management Programs for Gas Pipelines” 3#

Page 5benefits to be realized from those assessments. PSIA-2002 requires that the pipeline assessments specified in the Act must be conducted using in- line inspection, pressure testing, or direct assessment. The Act also allows assessments to be conducted using “an alternative method that the Secretary [of Transportation] determines would provide an equal or greater level of safety”.’ Here, again, RSPNOPS has been considering additional methods for conducting assessments. Confirmatory Direct Assessment (CDA) was discussed in the proposed rule for integrity management in gas transmission pipelines.’ This method allows for assessment of pipeline integrity at less cost to operators and resulting in less potential interruption in pipeline operations. In addition, RSPNOPS has been considering alternative methods of assessing low-pressure pipeline, for which pipe wall stresses are much lower and failure by leakage is much more likely than ruptures, which would also provide assurance of integrity at less cost than the methods specified in PSIA-2002. Taking no action would leave the definition of high consequence areas in part 192 unchanged. It would also mean that no alternative method, other than those specified in PSIA-2002, could be defined for conducting assessments of gas transmission pipelines in high consequence areas. The result would be conduct of assessments on more pipeline than is needed to address the underlying safety issue using more expensive assessment methods. RSPNOPS evaluated the costs operators would incur under this alternative, as described in the appendix to this analysis, and found that they would be significantly higher than those that would result from other alternatives. In addition, taking no action would not be responsive to the requirement in PSIA-2002 that DOT “issue regulations prescribing standards to direct an operator’s conduct of a risk analysis and adoption and implementation of an integrity management program.’* , ’ For these reasons, the “no action” alternative was not considered further. 2. Publishing a rule that would adopt requirements of the Pipeline Safety Improvement Act of 2002 and establish procedures to seek waiver of reassessment intervals PSIA-2002 allows the Secretary of Transportation to waive or modi@ requirements for reassessments for reasons that may include the need to maintain local product supply or the lack of internal inspection devices, provided that such a waiver is not inconsistent with pipeline safety. Another alternative would be to publish a rule that adopts the substantive requirements of PSLA- 2002, without change, but establishes procedures for approving the allowed waivers. ’bid. ’Federal Register, January 28,2003 (68 FR 4278), “49 CFR 192, Pipeline Safety: Pipeline Integrity Management in High Consequence Areas (Gas Transmission Pipelines); Proposed Rule”. ‘hpeline Safety Improvement Act of 2002, Section 14, “Risk Analysis and Integrity Management Programs for Gas Pipelines” 4#

Page 6I The technical requirements, costs, and benefits associated with this altemative would be identical to those for the “no action” altemative, since the substantive technical requirements would be the same, i.e., those imposed by the Act. As described above, RSPNOPS determined that those requirements would result in unnecessary expense for assessments, using the most costly methods to address more pipeline than RSPNOPS has determined is necessary to address the underlying safety need. For this reason, the option of publishing a rule that only establishes procedures for waivers was not evaluated further. 3. Prescriptive requirements for inspection and repair of pipelines in high consequence areas and for incorporating accident mitigative features. Pipeline circumstances differ, even within high consequence areas. These differences would make it difficult, at best, to establish prescriptive requirements that would appropriately address all possible combinations of pipeline size, type, and configuration or to consider other factors that contribute to the risk of failure of a particular pipeline. It is likely that creating detailed prescriptive requirements would result in a need for a large number of waivers to address the issues of importance to specific pipelines and high consequence areas. The result would be a patchwork of specific, but different requirements. It would be an ineficient use of industry and government resources to establish requirements in this fashion. Prescriptive requirements also would tend to stifle technological innovation. They do not allow for different approaches based on advances in the technology. The technology associated with in- line inspection of pipelines (i.e., pigging) is advancing at a rapid pace. Establishhg prescriptive requirements could slow this advancement, or could preclude use of new techniques that may be developed. In the extreme, prescriptive requirements could stop technological innovation in this area completely. Establishing prescriptive requirements would not assure the integration of information, which experience has shown is vital to preventing pipeline accidents. Two major pipeline accidents have occurred in recent years despite the fact that information about the causative factors should have, or could have, been known - at Edison Township, NJ and Carlsbad, NM. In the first case, in-line inspection (pigging) of the pipeline had taken place. The operator either failed to identify, during the pig runs, the areas of damage that eventually caused the rupture or the damage occurred in the years following the inspection. In addition, the operator failed to integrate information about the pipeline, including the presence of significant construction activity in the area, in a continuing assessment of the line’s integrity. In the latter case, the accident resulted from internal corrosion due to collection of moisture in a low spot which could not be inspected by pigging. The operator failed to consider the possibility of such accumulation of moisture and resulting corrosion and thus did not intercede to prevent the pipeline failure. It appears that information was available that, if correlated to current pig results (in the case of Edison Township) or other information about the pipeline, could have highlighted the need for action regarding the problems 5 8#

Page 7, that ultimately resulted in failure of the pipe. An integrity management program is required to assure this integration of available information. Outside review of the integrity management program by regulators (Federal and state), is necessary to assure that it is complete and properly implemented. This outside review cannot be assured without a requirement for such a program that establishes the requirements against which such a review will be conducted. Most importantly, establishing prescriptive requirements would not be consistent with requirements imposed by PSIA-2002. The Act requires that DOT prescribe rules to direct an operator’s conduct of a risk assessment and adoption and implementation of an integrity management program. Prescriptively establishing when and where integrity assessments must be performed would be inconsistent with the requirements of the Act. For these reasons, the option of establishing prescriptive requirements was not evaluated further. 4. Requiring pipeline operators to develop integrity management programs providing for inspection and testing based on risk factors and integration of information related to pipeline risk. Pipeline operators are uniquely qualified to develop integrity management programs and provide for the necessary integration of information. They have the best knowledge of their pipelines and the factors affecting its risk. Integration of information requires that the management systems of the company be aligned and operated to assure that necessary information is shared and that it is evaluated in its proper context by knowledgeable personnel. These are actions that are difficult to require through prescriptive regulation. Requiring that operators develop such programs is the best way to assure that they occur. Such a requirement also provides the regulatory basis for RSPNOPS and states to audit, review, and assess these programs and their implementation. The best integrity management plans, when implemented properly, can reduce the risk of pipeline accidents. They cannot, however, eliminate that risk. Leaks and ruptures could still occur, from unforeseen outside impacts on the pipeline or from unanticipated interactions among factors contributing to pipeline risk. It is therefore important that features and procedures be available to mitigate the effects of accidents that may occur. Here again, circumstances differ between pipelines and between regions and local jurisdictions. The differences make it difficult to establish prescriptive requirements that will provide the best protection for each high consequence area. Requiring that operators explicitly consider the need for mitigative features and provisions and that they implement those found necessary is the most effective means of providing such protection. Such a requirement also provides the regulatory basis for audit and review by RSPNOPS and state regulators. Establishing requirements for operators to develop and adopt integrity management programs is also most consistent with the requirements imposed on DOT by PSIA-2002. This option also 6#

Page 8allows RSPNOPS to make changes in the definition of high consequence areas to improve the ' focus of the new requirements, and to establish altemative methods that are acceptable for, performing integrity assessments. Both of these changes will reduce the costs imposed on industry without significantly reducing the benefits to be realized. For these reasons, this option was selected for further development. 5. Requiring pipeline operators to develop integrity management programs providing for expedited inspection and retesting. RSPNOPS considered the need for requiring integrity management programs that would require inspection and testing of pipelines to recur over short intervals, e.g., a few years. The ability to require frequent testing is limited by the available resources for testing and inspection. The companion rule covering hazardous liquid pipelines requires reassessments at least every five years, with limited exceptions. The current capacity to perform pipeline inspections will be challenged by this required schedule. The OPS concluded that the spur provided by the regulation would be likely to result in an increase in testing capacity over the next five years that will then be able to accommodate testing at accelerated rates. The OPS also concluded that protection from environmental damage that can be caused by a leak or rupture of a hazardous liquid pipeline, and the occurrence in hazardous liquid pipelines of frequent pressure cycles that can cause defects to grow, necessitated such frequent inspection. Adding requirements for similarly frequent inspection of natural gas pipelines would complicate the existing testing capacity issue and likely make it difficult for any of the testing requirements to be met. , ' Existing regulations already provide some additional protection from accidents on gas transmission pipelines that could affect high consequence areas.' The requirements of 49 CFR 192.61 1 specify that pipelines in class 3 or 4 areas must operate at pressures that produce significantly lower hoop stresses in the pipe than is allowed in more rural areas. As a practical matter, operators meet this requirement by reducing operating pressure or using heavier-walled pipe in class 3 and 4 areas. Hazardous liquid pipelines do not afford similar protection in high consequence areas. The additional protection already provided by natural gas transmission pipelines justifies assessment on a more extended interval than for hazardous liquid pipelines. Additionally, the natural gas pipeline network supplies gas for use in real time. This is not the case for hazardous liquid pipelines, which move product in batches and have significant storage capacity. Assessment of natural gas pipelines can therefore result in interruptions of gas supply. This can have a safety impact, in addition to its economic effect, due to the need to restart gas service in a controlled manner so as to avoid explosions at the point of service. The likelihood of service interruptions, with attendant costs and safety concerns, increases as the assessment interval is shortened, since operators have less flexibility to conduct assessments at times when demand is lower. 7#

Page 9Finally, significant environmental damage is not expected to result from failure of a natural gas pipeline, unlike hazardous liquid pipelines, since gas is lighter than air and dissipates in the ’ atmosphere. The PSIA-2002 requires reassessments on no greater than 7-year intervals. This altemative would be similar to the “no rule” option in that it would require assessments using in-line inspection, pressure testing, or direct assessment at no more than seven year intervals. The OPS has established requirements in this rule that would provide for a more focused assessment on this shorter interval, to reduce the likelihood of supply interruptions. As described above and in the appendix, RSPNOPS has evaluated the costs of options that do not allow the revised scope for 7- year inspections. Costs would increase significantly without addition of commensurate benefits. For these reasons, RSPNOPS concluded that assessment of natural gas pipelines need not be required as frequently as for hazardous liquid pipelines. BASELINE REGULATORY ENVIRONMENT In order to assess the costs and benefits of the new regulation, it is necessary first to ascertain the current level of activity in areas addressed by the rule. As described above, legislation recently enacted will impose a change in this level of activity, requiring the implementation of integrity management plans and conduct of integrity assessments that have not previously been performed. Still, it is necessary to determine the level of activity that has been occurring. The costs, and benefits, of implementing either this rule, or the requirements of PSIA-2002 without a rule, must be measured against this historical background. In this instance, it is necessary to determine the’ rate at which pipeline inspections are being performed, and the prevalence and nature of integrity management plans similar to those required by the rule. RSPNOPS has interacted with gas pipeline operators in recent years as part of development of an integrity management standard by the American Society of Mechanical Engineers (ASME). The standard includes many of the elements of this rule, and has been adopted as a consensus standard. As a result of these interactions, RSPNOPS understands that some gas pipeline operators currently have, or are developing, integrity management programs including many aspects required by this regulation. These current integrity management programs include inspection of their pipelines by some operators. The amount of such inspection is relatively low, however. Much of the testing being conducted by these operators is the initial inspection of pipelines. The rate at which subsequent inspections would be performed is now unknown. It is likely that some pipeline would be identified for reinspection routinely (e.g., every ten years). It is equally likely that some pipeline would not be reinspected at all. Integrity management plans are a key element of this rule. To better understand and promote 8#

Page 10more comprehensive and integrated approaches to safety and environmental protection, RSPNOPS created the Risk Management Demonstration Program, and the System Integrity Inspection Pilot Program. These programs encourage and evaluate operator-developed safety and environmental management processes that incorporate operator- and pipeline-specific information and data to identify, assess, and address pipeline risks. These programs are helping RSPA’s Office of Pipeline Safety refine its regulatory oversight processes. These processes help to ensure that pipeline operators have effective processes in place to identify the most important risks to the public and the environment, and to develop and implement cost-effective preventive and mitigative actions to manage these risks. Many of these initiatives have validated the importance of focusing resources and establishing higher levels of protection in areas where a pipeline failure could have significant consequences. Through the Risk Management Demonstration Program and the System Integrity Inspection Pilot Program, RSPNOPS has improved its understanding of pipeline operator integrity management systems and activities. This experience has shown that a number of pipeline operators have formalized management systems to identify and address the most significant integrity threats to their pipeline systems. In the Risk Management Program, participants perform systematic and comprehensive risk assessments to identify the specific nature and location of the most significant risks posed by operation of their pipeline system. An essential feature of these risk assessments is the integration of information from many diverse sources to hlly understand the integrity threats at specific locations on the pipeline. The impact on nearby population is explicitly considered in these risk assessments. Through formal, risk-based decision making processes, these companies use the risk assessment results to identify projects and activities that address potential system integrity threats, thereby preventing leaks and accidents. These investigative risk management programs, and the preventive and mitigative risk control activities that evolve from them, supplement the minimum regulatory requirements established in 49 CFR 192. The System Integrity Inspection Program is focused on developing a more integrity-based approach to RSPNOPS inspections. Instead of using a “checklist” approach, RSPNOPS is focusing the inspection process on an operator’s integrity management processes and activities. Through working with the operator, RSPNOPS is able to understand and influence the methods and approaches used to assess pipeline integrity, and the approaches to integrating integrity assessment data with other pipeline specific information, to identify the most significant integrity threats to the system. Specifically, RSPNOPS has observed how operators examine internal inspection data in conjunction with other surveillance and operating data, expected population growth, land use, construction activity along the pipeline, and other information relevant to assuring the integrity of the pipeline in high population areas and in environmentally sensitive areas. Through this interaction RSPNOPS is acquiring a broader understanding and a greater confidence that effective programs are in place to address the most significant risks. Similar to the Risk Management Program, the SII Program is emphasizing how operators evaluate their system condition and its risks, and use this information to make sound integrity management decisions. 9#

Page 11The RSPNOPS experience in the Risk Management Demonstration Program and the System Integrity Inspection Program indicates that integrity management programs including many of the elements required by this rule have been developed. They are far from universal, however. CONSIDERATION OF PUBLIC COMMENTS A number of comments were received in response to the proposed rule (68 FR 4278) and the regulatory analysis supporting it (referred to herein as the draft regulatory analysis). Comments from industry generally agreed that the earlier analysis had significantly underestimated the costs of complying with the proposed rule. The Interstate Natural Gas Association of America (INGAA) and the American Gas Association (AGA) made specific recommendations for changes to the proposed rule. In each case, the associations provided estimates of the costs, and savings, that would result from their changes. RSPNOPS acknowledges that the earlier regulatory analysis underestimated the costs of complying with the proposed rule. There were two principal reasons for this understatement. First, RSPNOPS used outdated information for estimating the costs of performing modifications on pipelines (e.g., to accommodate in-line inspection devices) and to perform inspections. The RSPNOPS estimates were based on studies that were performed in the early 1990s. Costs had been escalated to 2001 using the producer price index (i.e., to reflect inflation). Changes since 1990 in how this work must be performed, particularly costs associated with complying with improved environmental standards, significantly outpaced inflation during this period. RSPNOPS acknowledges this deficiency. For this analysis, RSPNOPS has adopted the unit costs (i.e., costs per mile or costs per company) used by INGAA and AGA in their analyses, as described herein for individual cost elements. The unit costs used by INGAA and AGA in their cost estimates differ. AGA explains the reason for this difference as the higher costs to do any kind of work, particularly work requiring excavation, in the urban environment in which its members operate. RSPNOPS acknowledges that costs are likely higher for urban work, and agrees that use of different unit costs for companies in such areas is reasonable. RSPNOPS has restructured this analysis to estimate costs separately for companies operating long-distance transmission pipelines in mostly rural areas (similar to INGAA members) and those operating transmission pipelines in urban areas (similar to AGA members). Unit costs derived from INGAA comments have been used for the long-distance estimates, and those from AGA’s comments are used to estimate costs for companies in urban environments. (RSPNOPS has also updated the estimated costs associated with the proposed rule, using the new unit costs. These estimates are provided in the appendix, in order to illustrate the effect on costs of changes made between the proposed rule and the final rule). Second, RSPNOPS used a much lower estimate for the amount of “overtesting” that will occur than used by INGAA in its analysis. Overtesting refers to pipeline not in high consequence areas that must be tested in order to test segments of pipeline that are within those areas. This 10#

Page 12principally affects in-line inspection (i.e., pigging). Inspection tools (i.e., pigs) must be inserted into the pipeline, run over the length of pipe to be inspected, and removed. Insertion and removal must occur at locations where there are launchers and receivers, structures attached to the pipeline which allow for insertion and removal of these large tools. RSPNOPS's preliminary draft regulatory analysis in support of the proposed rule had estimated the amount of overtesting as 25 percent (i.e., for every 10 high consequence area miles tested, 12.5 total miles would be tested). The Technical Pipeline Safety Standards Committee recommended that this assumption be increased when it reviewed the preliminary draft regulatory analysis. That analysis was' changed to assume 200 percent overtesting. , The comments submitted by INGAA and AGA used different estimates for overtesting. AGA assumed 100 percent overtesting for pigging, and no overtesting for pressure testing or direct assessment. This reflects that a significant portion of the transmission pipeline mileage of AGA members is likely to be in high consequence areas and that the distance between pig launchers and receivers in urban areas is relatively short. INGAA did not report a percentage assumption for overtesting. Instead, INGAA surveyed its members to obtain an estimate of the number of miles that they believed would be classified as high consequence areas and the number of miles that would have to be tested in order to pig those segments. The total number of miles that would be inspected was reported to be 7.25 times the estimated length of the high consequence area segments. RSPNOPS finds that this factor, while considerably higher than the 200 percent overtesting assumed in the draft analysis, is reasonable. High consequence areas are likely to be widely dispersed on long-distance transmission pipelines. The distance between pig launchers and receivers on those pipelines is relatively long, on the order of 50 miles. The combination of these factors makes it likely that a significant amount of overtesting will occur on these pipelines. RSPNOPS has adopted, for purposes of this analysis, an assumption of 625 percent overtesting, for pigging, consistent with the results of the INGAA survey. INGAA assumed 25 percent overtesting for pressure testing, which is the same factor RSPNOPS uses in this analysis. There is no overtesting required for assessments performed by direct assessment. INGAA also commented that the draft regulatory analysis underestimated the costs associated with data integration. The draft analysis had assumed that this would cost $100,000 in the first year for each company operating more than 40 miles of transmission pipeline and $50,000 annually thereafter. (The corresponding numbers for operators with fewer transmission miles were $25,000 and $12,500 respectively). INGAA agreed that more costs would be incurred in the first year, when record systems must be realigned and older records, many in paper form, must be converted for later analysis. INGAA estimated that the costs to perform this work would be $1,359 per mile in the first year, and $1 13 per mile annually thereafter. RSPNOPS agrees that estimating these costs on a per-mile basis (as opposed to per-company) is more reasonable, since the number of records that must be considered is proportional to the amount of pipeline an operator has. RSPNOPS also accepts that the costs for retrieval and conversion of older, paper records is likely to be high. RSPNOPS has adopted the INGAA per-mile estimates for estimating the cost of data integration. 11#

Page 13Commenters also addressed the benefits of the proposed rule. Carol Parker noted that the Carlsbad accident had a significant economic impact on the state of California due to reductions in gas delivery following the accident. Ms. Parker suggested that avoiding future economic impacts of this type is a benefit reasonably attributable to this rule. RSPNOPS agrees. RSPNOPS has estimated the magnitude of the economic impact that the Carlsbad accident had on California and describes that in this analysis among the benefits of the final rule. REVIEW BY TECHNICAL PIPELINE SAFETY STANDARDS COMMITTEE The OPS presented a preliminary draft of this regulatory analysis to the Technical Pipeline Safety Standards Committee (TPSSC) at a public meeting (conducted by teleconference) on July 3 1, 2003. TPSSC is a Federal advisory committee charged with responsibility for advising on the technical feasibility, reasonableness, cost-effectiveness, and practicability of gas pipeline safety standards. The TPSSC, by law, serves as the peer reviewer for RSPNOPS regulatory analyses. The TPSSC noted that the provisions of PSIA-2002 impose restrictions that make it difficult to optimize the costs and benefits of the rule. In particular, the Act requires that pipeline segments in high consequence areas be subjected to baseline assessments within 10 years and be reassessed every seven years. This has two effects. 0 It results in an “overlap” period in the eighth, ninth, and tenth years after the law was enacted, in which both baseline and reassessments will be performed. The increased level of testing in these years increases costs and increases the likelihood that pipelines being taken out of service for assessment will result in curtailment of gas supply It requires pipelines to be inspected more frequently than is likely needed, particularly for pipelines operating at low stress levels. The TPSSC acknowledged that RSPNOPS had done as much as possible, within the restrictions of the Act, to minimize the effect of these restrictions. The committee unanimously found that the cost-benefit analysis supported going forward with the final rule. SCOPE AND PARAMETERS OF ANALYSIS This analysis of benefits and costs takes the following approach. First, the mileage impacted by the regulatory change is identified and estimated. Mileage is estimated separately for long- distance transmission pipeline operators (representative of INGAA members) and for local distribution companies (LDC, representative of AGA members). Then the potential benefits of the rule are discussed. In the next section the potential costs of the rule are examined. Costs are estimated separately for long-distance operators and LDCs, due to the differences in unit costs for work done by each class of operator (as discussed above). Finally, a discussion of the costs versus the benefits is presented. It should be noted that, unless otherwise specified, all dollar values in 12 ‘ 1 .#

Page 14this report are given in constant 2001 dollars.' Furthermore, this analysis will arbitrarily consider only the first twenty years after the effective date of the final rule. Including additional years would not be expected to materially affect the conclusions of this analysis. ANALYSIS Impacted Mileage In this section the total pipeline mileage impacted by the regulatory change is estimated. That mileage is located in or nearby high consequence areas, defined by the change as areas in which defined numbers of people or facilities housing people of limited mobility are within specified distances of the pipeline. The distances vary depending on the diameter of the pipe and the pressure at which it operates. Total Pipeline Mileage In total, there is an estimated 285 tilousand miles of regulated onshore natural gas transmission pipelines in the U.S.6 This rule would not apply to all of this mileage. The principal requirements of the final rule apply to pipeline mileage in high consequence areas, which is estimated below. The rule also applies to transmission pipelines for hydrogen, synthetic gas and other products subject to 49 CFR Part 192 that are not included in the natural gas transmission pipeline totals. RSPNOPS does not have data on the total transmission mileage for these other gases. This analysis uses the available natural gas transmission pipeline total mileage, which is considered to be very close to the total pipeline mileage potentially affected by the regulation. Impacted MileaEe in High Consequence Areas The major elements of the regulatory change apply to that transmission pipeline that is in high consequence areas. The rule provides two options for an operator to determine whether a segment of pipeline is in a high consequence area. Under option a, all pipeline in class 3 and 4 areas, as defined in 49 CFR 192.5, would be included, as well as additional pipeline having an identified site within a potential impact circle. Under option b, operators would determine what pipe is in high consequence areas by evaluating potential impact circles along the entire pipeline. 'Dollars are converted from nominal values to real 2001 values using the Producer Price Index (PPI), Intermediate Materials, Supplies, and Components. The source of the PPI index numbers is the U.S. Bureau of Statistics Web page. 6Jurisdictional natural gas transmission pipeline mileage (onshore) for 200 1. This mileage was obtained from annual reports filed by pipeline operators with the Office of Pipeline Safety. Data available on the OPS web page. 13#

Page 15LDC High Consequence Area mileage Pipeline operators are presently required to maintain data on the population near their pipeline in order to determine pipeline that is in class 3 or class 4 areas. That mileage is not estimated here. This option was included in the rule principally for smaller distribution operators with a limited amount of transmission pipeline mileage. Many of these operators have historically designed and operated their entire systems as if they were class 3 or class 4 rather than collecting and analyzing population data near their pipelines to determine the class location. This option would let them take the same approach to determining high consequence areas, Le., treat the entire length of their transmission pipelines as if they were in high consequence areas. RSPNOPS does not expect that many operators will use this option. Much of the transmission pipeline operated by LDCs is low-pressure pipeline. Much is also small diameter pipeline. Operating pressure and pipe diameter are factors used to define the potential impact radius. For small-diameter, low-pressure pipelines, the radius will be small. For example, the potential impact radius for a 6-inch diameter pipeline operating at 300 psi is 72 feet. As a practical matter, it is very unlikely that 20 buildings intended for human occupancy or an identified site will be located within a 72-ft radius circle. RSPNOPS expects that most LDC operators will recognize that using method b, i.e., defining high consequence areas on the basis of evaluating potential impact circles, will result in little or none of their pipeline being determined to be in high consequence areas. This will minimize the costs to those operators of implementing this rule. In this analysis, RSPNOPS assumes that all operators will use method b. . I , One provision of this rule would apply to low-pressure pipeline in class 3 and 4 areas that is outside of high consequence areas. RSPNOPS relied upon comments by AGA and the American Public Gas Association (APGA) to estimate this mileage. AGNAPGA comments indicated that their members operate approximately 16,000 miles of pipeline in class 3 and 4 areas.’ RSPNOPS has assumed that all of this pipeline operates at low stress (i.e., less than 30% S M Y S ) . RSPNOPS has increased this number by 1000 miles to account for customer owned service lines and operators not members of AGA or APGA. A total of 17,000 miles of low-pressure pipeline is estimated to be in class 3 or 4 locations for purposes of this analysis. The amount of transmission pipeline mileage in high consequence areas has also been estimated based on comments from the trade associations. AGA estimated that 2 1,800 miles of member pipeline (AGA stated that its estimates included pipelines of APGA members) meet the “general concept” of high consequence area.8 This was based on the earlier definition of high consequence .area, which included all class 3 and 4 locations. RSPNOPS adjusted this estimate to account for the reduced pipeline mileage that will be identified as in high consequence areas using method b. As a first step in this adjustment, RSPA apportioned the AGA total estimate to pipeline operating in different pressure ranges. AGA estimated that 45 percent of their member pipeline in high 7L0ri Traweek, AGA, letter to docket dated April 8,2003, RSPA-2000-7666-203, page 3. ‘Lori Traweek, AGA, letter to docket dated April 30,2003, RSPA-2000-7666-294, page 1 I 14#

Page 16consequence area operates below 30% SMYS.' AGA also estimated that 19 percent of its members' pipeline operates above 50% SMYS." This implies that 36% operates between 30 and 50% SMYS. Applying these percentages, AGA's 21,800 mile estimate was apportioned to these operating pressure ranges. RSPNOPS then reduced the number of miles in each range to reflect the likelihood that pipeline in class 3 and 4 locations, operating at those pressures, would not be determined to be in a high consequence area using method b. For pipeline operating at >50% SMYS, little reduction was assumed, since potential impact circles for that pipeline will be relatively large. For pipeline operating below 30% S M Y S , a significant reduction was assumed, since potential impact circles there will be quite small, as described above. The resultant estimate of LDC pipeline transmission mileage in each pressure range is presented in Exhibit 1. Exhibit 1. AGNAPGA Transmission Pipeline in High Consequence Area (Miles) Pressure Range Apportioned AGA Estimated HCA Estimate Miles I >50%SMYS I 4,142 I 4,000 I I 30to50%SMYS I 7,848 I 4,000 I * I <30%SMYS 1 9,810 I 2,000 I Total 21,800 10,000 Long-distance Pipeline High Consequence Area mileage , I RSPNOPS also relied upon industry comments to estimate the amount of high consequence area mileage for long-distance pipelines. Method b for identifying high consequence areas was based largely on a proposal submitted by INGAA in response to the proposed rule. INGAA conducted a survey of its members, asking them to apply the new recommended method and determine the amount of pipeline mileage that would be classified as high consequence areas if it were used. INGAA reported, in a set of spreadsheets submitted to the docket," that members operating 166,000 miles of pipeline had estimated that 8,496 miles would be classified as high consequence areas on the basis of the housing density portion of method b. Extrapolating this estimate to the entire 225,000 miles of long-distance pipelines results in an estimated 1 1,5 16 miles of high consequence area. INGAA's survey addressed only the housing density portion of method b. This method would also include areas where an "identified site", consisting of a building housing a population of 2 'hid., page 2 "Lori Traweek, AGA, letter to docket dated April 8,2003, RSPA-2000-7666-274, Attachment B, page 1 of "INGAA comments to docket, April 30,2003, RSPA-2000-7666-297. 15#

Page 17limited mobility or a place where people gather (meeting specific criteria), is within a potential impact circle. Pipeline operators can not now determine how much pipeline will be classified as high consequence areas based on the presence of an identified site, because they have not historically collected information on facilities within the identified site definition. (An exception is places where people gather that are within 300 feet of the pipeline, since that is part of the definition of class 3). RSPNOPS increased the estimated high consequence area mileage by 30 percent to account for miles that will be added due to identified sites. (The 30 percent factor is the same multiplier used for this purpose in the draft regulatory analysis). The estimated total amount of long distance pipeline in high consequence areas is thus estimated to be 14,970 miles. Most long-distance pipeline operates at high pressures. For purposes of this analysis, RSPNOPS assumes that all of this mileage operates at >50% SMYS. Intrastate Gas Transmission Pipeline Mileage in Texas The high consequence area mileage associated with intrastate pipelines in Texas must be subtracted from these totals. This is because Texas pipeline is subject to a state rule that already requires inspection and assessment. No additional testing costs will be incurred as a result of this rule. (Programmatic requirements imposed by this rule will have an impact on Texas intrastate operators, but these costs are estimated mostly on a per-company, as opposed to a per-mile, basis). RSPNOPS estimated for purposes of the draft regulatory analysis that there are 5,432 miles of intrastate gas transmission pipeline in Texas.'' For this analysis, RSPNOPS reduced that estimate to 3,500 miles, proportionate to the reduction in the estimated amount of total high consequence area mileage as a result of changes in the final rule. RSPNOPS further apportioned this estimate between long-distance pipeline (2,500 miles) and LDC-type pipeline (1,000 miles) on the basis of a rough estimate of the proportion of Texas mileage in each category. RSPNOPS further assumed that the 1000 LDC miles are distributed among the operating pressure ranges in the same proportions as used for LDC operators in general. The resultant estimate of pipeline transmission mileage in high consequence areas is thus as shown in Exhibit 2. "Draft Final Regulatory Evaluation, Pipeline Integrity Management in High Consequence Areas, (Gas Transmission Pipelines), RSPA-2000-7666- 166, page 12. 16#

Page 18Exhibit 2. Transmission Pipeline in High Consequence Areas (Estimated, for purposes of this analysis) (Miles) I Pressure Range 1 LDC HCA Miles I Long-distance HCA Miles I I >50%SMYS 1 3,600 I 12,470 I 30 to 50% SMYS 3,600 0 <30% S M Y S 1,800 0 Total 9,000 12,470 BENEFITS The benefits resulting from the final rule are discussed in this section. Those benefits are expected to result from detection of problems that could cause pipeline failures before the failure occurs, thereby averting accidents. The inspection and assessment required by the rule is designed to detect problems related to internal corrosion, external corrosion, stress corrosion cracking and external damage to the pipeline, all of which can result in pipeline ruptures. Natural gas pipeline accidents usually involve explosions and fire and can result in death, serious injury, and property damage. Preventing accidents will result in reduced numbers of deaths and serious injuries and in reduced property damage. These reductions, then, are principal benefits of the rule. The rule will also help avoid potential accidents with consequences larger than any experienced to date, allow for quicker return to full service following required reductions in operating pressure, provide’ improved assurance of pipeline safety, provide a basis for increased public acceptance of the risks from natural gas transmission pipelines, reduce economic impacts that might result from future pipeline accidents, and will provide other, less tangible, benefits. Each of these categories of benefits is discussed beIow. Pipeline operators also have strong incentives to ensure the integrity of their pipelines. In addition to the positive safety and societal benefits, the lost product and unscheduled downtime for repairs following a major incident can significantly impact the company’s financial performance and its ability to satisfy customer commitments. Operators cannot afford to have these critical transportation assets out of service for lengthy periods of time in today’s competitive business environment. In addition, the damage to the company’s public image and reputation, as well as the legal implications of serious incidents, can pose an even broader and longer term negative impact on the company’s business operations. For these and other reasons, many pipeline operators have implemented and are continuing to improve more systematic safety and environmental management processes, many of which already embody many of the principles in this rule. Benefits from Reduced Conseeuences of Future Accidents 17 . .#

Page 19Benefits from reduced death and serious injuryi3 Accident reports submitted to the OPS during the period 1986 to 2OO2I4 identify that there’were 1,371 incidents on natural gas transmission pipelines, resulting in 59 fatalities and 224 serious injuries. The consequences of future pipeline accidents could differ, and are likely to be more severe, as discussed below. Nevertheless, it is reasonable to use this 17-year record as an estimate of consequences that would be likely to occur without changes in the manner in which pipeline safety is assured. (The accident data covers 17 years and costs are calculated in this analysis for 20 years. This difference does not affect the analysis, because the accident data is used to estimate an average benefit value per year that is then applied over the 20 year analysis period). The rule is expected to reduce the consequences of future accidents, through identification and remediation of the kinds of anomalies that can cause pipeline accidents before those accidents occur. Accidents that will be prevented by the rule should include a high percentage of those that result in death and serious injury, since the rule is focused on pipelines in areas which have the largest concentrations of people in the vicinity of the pipeline. This rule will not eliminate all pipeline accidents. It is not possible to estimate precisely how effective it will be in reducing such accidents. The theoretical maximum benefit that could be achieved would be elimination of accidents causing death and serious injury. Based on this historical record, the maximum value that could be realized from reducing deaths and serious injuries is thus $289 million over 17 years or $17 million per year.I5 Benefits from reduced property damage I , The same accident data base indicates that $328,3 13,111 in property damage occurred as a result of those 1,371 pipeline incidents. A recent study indicates that this total may be low due to under- reporting of accident COS~S.’~”’ I3With respect to deaths and serious injuries, the following assumptions are made: A life is valued at $3 million A serious injury is valued at $500 thousand These valuations are standard assumptions currently used in Office of Pipeline Safety and DOT benefitkost analyses. I4this period was chosen because this data is publicly available on the OPS website ”The accident data used here is national data. Since intrastate transmission pipeline in Texas has been excluded from this analysis, the contribution of any accidents on that pipeline should also be excluded. This has not been done. RSPNOPS did not examine the specific accident reports that lie behind the tabulated accident information on its web site for the period 1986 to 2002, and does not know whether any intrastate Texas pipeline is represented in that data. None of the major accidents in this period occurred on that pipeline. The OPS expects that the effect of excluding accidents on intrastate Texas pipelines from the computation of benefits in this analysis would be small, likely well within the errors associated with the simplifjmg assumptions used to make the analysis possible. I6“Report on the Accuracy of Cost Data from Incident Reports”, General Physics Corporation, December 2001, un ublished. p 7 R S P ~ O P S has not attempted to exclude any property damage occurring during this period on Texas intrastate natural gas transmission pipelines for the same reasons described above for deaths and serious injuries. 18 . .#

Page 20The study compared accident costs reported to RSPNOPS with other information, including press reports and costs reported in operators' post-accident financial filings. The study considered 49 accidents, of which only four were natural gas pipeline accidents. (Two of these accidents had not been reported to RSPNOPS). The study found that actual costs for accidents involving hazardous liquid pipelines were three times the amount reported to RSPNOPS. For the limited set of gas pipeline accidents considered, costs were under-reported by a factor of 1.62. RSPNOPS believes that a larger study of gas pipeline accidents might show more under-reporting of costs, similar to the situation revealed for hazardous liquid pipelines. For purposes of this analysis, RSPNOPS assumes that costs may have been under-reported for natural gas pipeline accidents by up to a factor of 2. Thus, the true value of property damage experienced in natural gas transmission pipeline incidents over the last 17 years is in the range of $328 to approximately $656 million. This range is used in this analysis as representative of the property damages caused by historical natural gas pipeline accidents. As before, the historical record provides a reasonable estimate of the consequences of potential future accidents. Again, the rule is expected to reduce the number of accidents, and thus the amount of property damage that occurs. The extent of such reduction cannot be estimated with certainty. The maximum benefit that could be achieved if the historical damage is at the upper end of this range and property damage consequences were eliminated by implementation of the rule is $656 million over 17 years, or $38.6 million per year. Consequences of Pipeline Accidents are Likely to Increase Urban areas are rapidly expanding in the United States. Housing starts have increased 57% over the last ten-year period. Increasingly, this brings additional population into the proximity of the natural gas transmission pipelines that serve our urban areas. Rural areas that pipelines may have passed through ten years ago are more likely today to be populated, and that likelihood will hrther increase over time. Natural gas pipeline accidents that occur in rural areas have limited consequences, particularly in causing deaths and serious injuries. Accidents in urban areas can be much more severe. The March 23, 1994, accident in Edison Township, New Jersey is a case in point. This area was already urbanized at the time of the accident, but the accident demonstrates the potential for major consequences. Rupture of a 36-inch diameter natural gas transmission line resulted in an explosion and fire that destroyed six apartment buildings. Property damage exceeded $25 million. Approximately 1,500 residents were evacuated from the apartments. Immediate evacuation prevented any deaths, although one resident living approximately one mile from the scene of the accident suffered a fatal heart attack.18 Had circumstances been only a little different, significant loss of life could have occurred. Increased development makes it likely that the actual consequences of natural gas pipeline '8National Transportation Safety Board, Pipeline Accident Report: Texas Eastern Transmission Corporation Natural Gas Pipeline Explosion and Fire Edison, New Jersey March 23, 1994, January 18, 1995, p. v. 19 '#

Page 21accidents over the next 17 years, assuming no changes in the regulatory environment, would be more severe than suggested by the historical record. There is no information available with which to estimate the precise amount by which population near pipelines will increase. RSPNOPS has therefore not increased the amount of benefits expected in the later years of this analysis, and this increase is not explicitly considered in this analysis. Any understatement of benefits is offset, since the analysis assumes a full benefit from avoiding accident consequences in the first year. In actuality, full benefits would not be realized until all covered pipeline has been subjected to a baseline assessment, i.e., after ten years. (A large portion of the total benefit will be’realized in the first few years, because of the rule’s requirement that assessments begin with the pipeline segments posing the highest risk). The “non-conservative” assumption of full benefit in the first few years offsets the “conservative” treatment of not including expected increases in benefits in later years. Total Benefits from Averted Deaths, Serious Injuries, and Property Damage The maximum total benefit represented by the historical record is approximately $55.6 million per year, the sum of the benefits for deaths and serious injuries and for property damage as described above. RSPNOPS believes that the rule will have significant effect in reducing the,occurrence of the kinds of accidents that resulted in these consequences. Given the magnitude of the potential benefit represented by the historical record and the likelihood that consequences of hture accidents would increase, RSPNOPS concludes that the benefit of the rule in eliminating deaths, serious injuries, and property damage is on the order of $40 million per year. Over the 20 years considered in this analysis, the total benefit is thus $800 million. 8 , Consequential Impact of Natural Gas Pipeline Accidents . . The accident impacts described above are direct effects, i.e., they are caused directly by the pipeline rupture and resulting explosion and fire. The consequences of natural gas transmission pipeline accidents often do not stop there. Other impacts include disruption of business activities in the immediate area of the accident and possibly in areas near the accident. Many communities are served by natural gas distribution companies that receive their product via single lateral pipelines from a natural gas transmission pipeline (so-called “sole-source laterals”). If an accident occurs on the transmission pipeline that results in interruption of the flow of natural gas, service to customers in communities served by sole-source laterals may be cut off. The interruption may be temporary, if gas supply can be restored by valving out the damaged section of pipe and re-establishing supply from undamaged portions of the line. Even so, there is both an economic and a safety consequence to such service interruptions. When natural gas service is cut off, pilot valves on gas appliances go out. Service cannot simply be restored, since gas would enter homes and businesses through the open pilot valves, potentially build to explosive concentrations, and result in fires, explosions and additional collateral damage. For this reason, restoration of natural gas service requires that local distribution companies follow labor-intensive procedures. Representatives of the distribution company must enter each business 20#

Page 22or residence to which service was interrupted. They must close valves to pilot lights. Distribution mains and laterals must be purged to eliminate air that may have become entrained. Only then can service be restored. Restoration of service again requires that an employee of the distribution operator must enter the premises, reopen pilot light valves, and re-light the pilot lights. This process can take several days. A recent service outage involved loss of natural gas service to approximately 4500 customers. Service was restored in 48 hours, but only by the efforts of 400 personnel, many supplied by other local distribution companies to assist in the emergency recovery effort. Economic consequences included business interruption for the period of the outage, overtime for local operator personnel, and the need for the local operator to house and feed personnel loaned from other operators to assist. For purposes of illustration, if we estimate the labor costs of those 400 personnel at $75 per hour and they worked 24 hours (2 twelve hour days) the labor cost alone would be $360,000. Additional costs would include transportation, leased emergency equipment, lost revenue from the customers who were not served. It is easy to see how the cost of such an accident on a small line with only 4500 customers could easily approach $1-2 million. There is a potential that the impact of consequential damages from service interruptions could grow. Natural gas is currently being used to power many new electrical generating facilities. As more of the nation’s electricity is generated from natural gas, the supply of electricity will also become more dependent on reliable, continuous availability of natural gas. It is possible that future accidents on major interstate natural gas transmission pipelines in certain areas could result in loss of natural gas supply to multiple electrical generating stations. Electricity generators typically have a supply margin to account for the unexpected loss of a generating facility. If too many generators are lost simultaneously, however, the margin can be overwhelmed and electrical blackouts, with their attendant consequences, could result. Public Confidence The most significant benefit of the final rule is less tangible. It will provide a basis for improved public confidence in pipeline safety. Public confidence has been shaken as a result of several recent accidents with significant consequences. These accidents were widely reported by national media, becoming known well beyond the communities in which they occurred. These included the 1994 pipeline rupture, explosion, and fire at Edison Township, NJ (discussed above), a June 10, 1999, rupture of a hazardous liquid pipeline in Bellingham, WA, with subsequent fire, and an August 19,2000, natural gas pipeline rupture, explosion and fire near Carlsbad, NM. Three persons were killed in the Bellingham accident. Twelve persons were killed in the Carlsbad accident, (Hazardous liquid pipelines, such as the one involved in the Bellingham accident, would not be affected by this rule. They are covered by similar rules for hazardous liquid pipelines, which have already become effective.) Improving public confidence is, in itself, important. It will, however, also result in economic benefits. 21#

Page 23One way in which public concern regarding pipeline safety manifests itself is in increased public opposition to new pipelines. Local governments can impose additional requirements and restrictions that delay construction and result in significant additional costs. A recent example involved the conversion of an existing hazardous liquid pipeline in Texas. Community reaction in the city of Austin resulted in delays and significant additional costs. In response to the community reaction, the operator replaced 12 miles of the existing pipeline with 21 miles that looped to the south of the city, avoiding most populated areas. This significantly increased the cost of the pipeline project. According to data published in the Oil and Gas Journal, the costs of constructing pipeline have increased significantly over the last 15 years. The average cost of constructing on-shore natural gas transmission pipelines between 1995 and 2000 was $1.12 million per mile. A similar re-route for a natural gas transmission pipeline thus would have cost approximately $23.5 million. Increased public opposition can also result in delays in implementing pipeline projects. In some cases, the related costs associated with responding to public concerns, participation in public hearings, and financing of major construction projects during delays can be as significant as, or more than, the cost of installing new pipeline. In the extreme, increasing public concern could make it impossible to site and construct new natural gas transmission pipelines. The United States needs additional natural gas transmission pipeline capacity to meet current and future needs. The Energy Information Administration predicts that total demand for natural gas will increase approximately 1.8 percent per year from 2001 to 2025, rising from 22.7 to 34.9 trillion cubic feet annually." Demand growth in particular areas could greatly exceed this national average. If operators are unable to construct new pipelines, the existing pipeline system would rapidly reach its capacity limit, and the ability to meet this increased demand could be challenged. New applications of natural gas as a fuel would need to be foregone. The ability to use natural gas as an environmentally-preferable fuel for new electric generating capacity would be lost. Curtailment of existing natural gas usage would likely be required. For all of these reasons, it is vitally important that the public have confidence that the national network of natural gas transmission pipelines is safe. The final rule provides a foundation for an improvement in public confidence. It requires operators to implement inspection and assessment programs directed at identifying the causes of the major pipeline accidents summarized above, and other potential causes of pipeline accidents, and correcting them before pipeline accidents can occur. The resulting increase in public confidence of pipeline safety may well result in less opposition to new pipelines, faster permitting, and less delay caused by local government efforts to require additional provisions or re-routing to improve safety. INGAA has suggested that any increase in public confidence is not likely to improve the ability to "Energy Information Administration, Annual Energy Outlook 2003, With Projections to 2025, January 2003, page 4. 22#

Page 24site new pipelines.*’ INGAA’s assessment is based in large part on those areas in which there may be a near-term need for more pipelines. According to INGAA, these areas are the Rocky Mountain region, the Gulf of Mexico, and Alaska. INGAA notes, appropriately, that the requirements of this rule either would not apply (e.g., offshore piping in the Gulf) or would have limited impact in these areas due to the sparse population and reduced likelihood that pipelines would traverse high consequence areas. INGAA also suggests that public opposition to pipelines is oAen based on property issues (e.g., encroachment on a property owner’s land) more than on safety concerns. RSPNOPS considers that INGAA’s evaluation of the benefit that may result from increased public confidence is too limited. It considers areas in which pipelines may be needed in the immediate future, but does not consider fully where pipelines may be needed in later years. As noted above, the Energy Information Administration estimates that U.S. consumption of natural gas will increase by more than 50 percent between now and 2025. RSPNOPS considers it likely that pipelines will need to be sited to meet that need in areas where the requirements of this rule will have an effect on public confidence in pipeline safety. It is difficult to quantify the benefit that will result from increased public acceptance of natural gas pipelines. Local property issues, as noted by INGAA, could be more important than safety concerns in some areas where siting decisions must be made. Nevertheless, if growth in consumption leads to a need for new pipelines, it is very important that the industry be able to site and construct those pipelines. Gas prices are relatively inelastic. Prices can rise significantly with even a minor shortfall in ability to meet demand. The impact on the U.S. economy if natural gas demand cannot be met could be significant. RSPNOPS has not estimated a monetary value for this benefit. Avoiding an Accident with Very High Conseauences Growth has occurred along many gas transmission pipelines that were originally in rural areas. This growth has included houses, in some areas quite densely packed, and facilities like schools, office buildings, churches, theaters and hospitals. These changes have brought larger numbers of people into areas in which they could be affected by pipeline accidents, and have also increased the number of pipeline miles on which an accident could affect large numbers of people. An accident in one of these areas could result in consequences greater than those experienced in any historical accident. Accidents with larger consequences could occur in one of two ways: affecting a dense concentration of homes or affecting a school or other facility in which large numbers of people congregate. The potential impact circle @e., the area in which the effects of a rupture and explosion would be felt immediately) for a large, high-pressure pipeline can be 1,000 feet or more in radius. A circle of 1000 ft. radius encompasses 72 acres of land. If a housing density of 2 dwellings per acres is ~~ 2?erry D. Boss, INGAA, letter to the docket July 30, 2003, RSPA-2000-7666-342, p.11. 23#

Page 25assumed, 144 houses could be within the potential impact radius.2’ If each dwelling is assumed to house an average of 2.5 occupants, this puts 360 people potentially at risk from a rupture and’ explosion on a postulated pipeline in that area. The chances of an explosion in these areas is, to be sure, already small. Pipeline ruptures are infrequent events, and most areas along pipelines have lower population densities. It is informative, however, to consider the possibilities. Data for the last 19 years indicates that pipeline ruptures occur approximately 22 times per year. Over the 20 years considered in this analysis, 440 ruptures would be expected to occur (without the integrity improvements embodied in this rule). The likelihood that a rupture would occur in this period that would affect areas with very high population densities can be estimated by multiplying the total number of ruptures expected by the fraction of transmission pipeline where such high populations exist. If 0.5% of the total U.S. transmission pipeline mileage (representing 1425 miles of pipeline across the U.S.) is assumed to include high population densities, then 2.2 ruptures would be expected to affect these areas during this analysis period. An explosion in a high density population area would be expected to cause significqt casualties. It is not possible to estimate with certainty how many of the 360 residents potentially involved in such an explosion would be killed and how many would be seriously injured. It is likely that few, if any, would escape without injuries. For purposes of this analysis, RSPNOPS assumes that one- half of the residents at risk would be killed and that the other half would be seriously injured. Using the standard DOT values for attributing value to deaths and injuries (described above) this would be equivalent to an economic consequence of $630 million. This value, of course, assumes that one single accident occurs. It must be converted into an “expected” cost by multiplying times the expectation that such accidents will occur. Using the expectation of 2.2 such ruptures during the analysis period leads to an estimated benefit from avoiding such a high consequence accident of $1.386 billion. An accident occurring near a facility housing large numbers of people could also result in significant numbers of deaths and injuries. This might occur, for example, if an accident occurs in an area where a school or office building exists within the potential impact circle. Informal discussions with pipeline operators have identified that such facilities are often located near urban gas transmission pipelines. In fact, it appears that as much as ten percent of LDC high consequence area mileage is likely to involve identified sites at which large numbers of people congregate on a regular basis. ”Potential impact circles for smaller pipelines operating at lower pressure would be smaller. This is offset, here, by the use of 2 dwellings per acre as the assumed housing density. Housing densities in urban areas can be considerably higher than this. Approved zoning densities for townhouse communities can approach 10 dwellings per acre. The estimate of 144 homes in a potential impact circle is reasonable for higher housing densities near smaller pipelines. 24 -#

Page 26For analysis purposes, RSPNOPS has postulated an accident in which a facility housing 500 people is impacted by a pipeline rupture and explosion. As above, it is not possible to estimate with certainty how many people would be killed and how many injured, but it is likely that few, if any, would escape serious injury. Applying, again, the assumption that half of the people involved would be killed and half seriously injured, the economic consequences from even a single accident of this kind would be $875 million. As before, the likelihood of such an accident can be roughly approximated. Ten percent of LDC high consequence area mileage represents 900 miles of pipeline. Of 440 pipeline ruptures that would be expected to occur in the next 20 years along the entire transmission pipeline network (285,000 miles), 1.4 would be expected to occur along the 900 miles that could affect structures housing large numbers of people. The expected value must be reduced, in this case, since schools and office buildings do not house people 24 hours a day. Assuming these facilities are occupied 40 hours a week results in an expected benefit, in terms of averting the consequences of this type of accident, of $290 million. This rule contains provisions that will contribute to avoiding high consequence accidents and realizing these benefits. Many of these provisions go beyond the requirements in PSIA-2002, making the rule more effective in achieving this end. These provisions include: 8 Repair criteria specified in a very precise manner, going beyond any industry consensus standards. Experience has demonstrated that aggressive repair criteria have had greater impact than other safety actions. Many pipelines that had accidents in the 1990s had been tested, but operators, by themselves, were not making best use of the available data. A requirement for a threat-by-threat evaluation of the pipeline to focus on those threats of significance. A requirement to apply lessons leamed from assessing covered pipe segments to similar pipeline not in high consequence areas. These provisions respond to recommendations of the National Transportation Safety Board, and will significantly expand the impact of the rule. A requirement to gather, monitor and report measures of performance that will allow operators, regulators and the public to observe the downward trends in the number of significant defects in gas pipeline. Preventive and mitigative measures that will significantly reduce the likelihood of accidents caused by time-independent events such as those caused by third-party damage. These provisions will have particular benefit in areas where significant growth is expected, with attendant excavations and potential for pipeline damage. This type of accident would not be addressed by implementing the provisions of PSIA-2202 alone. The total benefit associated with avoiding accidents with very high consequences is the sum of the benefit realized by averting accidents affecting dense concentrations of housing and those affecting facilities housing large numbers of people. Over the term of this analysis, that benefit amounts to $1.676 billion. I 25 1#

Page 27This estimate reflects the value placed on the potential loss of life and serious injury that could result from these accidents. It does not include property damage losses. An accident that would kill or seriously injure occupants in 144 houses would, obviously, damage or destroy those houses. Similarly, an accident that kills or seriously injures persons in a school or office building would be expected to cause significant damage to, or loss of, the building. The values placed on death and serious injury, however, mean that potential property damage losses are a small portion of the expected total impact of an accident. For purposes of this analysis, and recognizing the approximate nature of the assumptions made, RSPNOPS has not included the benefits from avoiding potential property damage in these very severe accidents. Preventive Maintenance vs. Accident Response The final rule requires operators to implement programs that are intended to identify areas on their pipelines needing remediation and repair and to accomplish the necessary remediation and repair efforts. If not found and repaired, some of the anomalies could cause accidents, and thus require repair (and recovery activities). The rule can thus be seen, in part, to be a substitution of “preventive maintenance” (i.e., identify problems early and address them) for reactive response to accidents. The rule includes schedules by which required remediation actions must be taken, which vary depending on the severity of the identified anomalies. Operators would be permitted to take longer than these schedules if they provide additional margin of safety by reducing pressure or if they notify RSPNOPS of the circumstances requiring additional time (which will allow RSPNOPS to review those circumstances and oversee the operator’s actions). Remediation of identified anomalies is thus more in the nature of preventive maintenance: operators can schedule their efforts based on important factors such as availability of repair resources and when demand on the pipeline is relatively reduced. If the line must be taken out of service for the repairs, advanced preparations can avoid the need for service interruptions and their consequences (as described above). Not identifjmg and resolving these anomalies could cause some of them to result in accidents. Then, operators have no flexibility. Repair resources must be made immediately available, regardless of other demands. Overtime and use of “borrowed” crews from other pipeline operators is almost always involved. In addition, the accident may cause additional damage to the pipeline involved or to other pipelines on the same right-of-way. Damaged pipelines may be out of service for extended periods (see below). Informal discussions with natural gas transmission pipeline operators indicate that typical costs to repair defects found by inspection range from about $20,000 to $60,000, depending on whether service must be interrupted to effect the repair. The cost of unplanned recovery from a leak can be up to an order of magnitude higher. The cost of recovering from a major pipe failure can be two or more orders of magnitude higher, i.e., in excess of $5 million. Costs to the operator to repair the damage caused directly by the Carlsbad accident amounted to 26#

Page 28approximately $1 million. Indirect damage caused by the accident resulted in approximately an additional $4 million in costs. Repairs and modifications necessary to return the pipeline to 1 service cost an additional $3 million. Total costs directly attributable to the accident were’thus on the order of $8 million. Edison and Carlsbad were both accidents that resulted in significant direct costs to the operator. Recent experience thus indicates that such accidents occur about once each decade. This rule will not prevent all such accidents. Due to the significant amount of overtesting that is expected, however, it can prevent many of them. (Overtesting is discussed below). RSPNOPS estimates that a total of 71,780 miles of transmission pipeline will be assessed to comply with this rule. This represents 25 percent of the total transmission pipeline mileage. If we assume that accidents causing significant direct costs would otherwise be equally likely at any location on the pipeline, this rule will avoid 25 percent of them. If two such accidents are expected during the 20-year period of this analysis, then this rule will result in avoiding 0.5 such accidents during the period, for a benefit of $4 million. Economic Consequences of Accident-Induced Supply Restrictions The largest impact of an accident such as that which occurred at Carlsbad does not fall on the pipeline operator. It is the result of restrictions in the supply of gas and increases in its price that result from a major pipeline being out of service. The Carlsbad accident provides an informative example. Public comments received in response to the proposed rule pointed to the cost impact of the Carlsbad accident on California as justification for this rule. The Carlsbad accident resulted in curtailment of natural gas supply to California. The cost increases resulting from this curtailment were only part of the broader costs associated with energy trading and other issues affecting the 2000 energy price problems in California. A FERC evaluation concluded, however, that evidence “strongly implies that the Carlsbad rupture contributed significantly to extraordinarily high California spot gas prices.”22 The FERC study infers that the effect of the Carlsbad rupture was an increase in gas prices of approximately $ 5 per thousand cubic feet. This amounts to approximately $1 7.25 million per day for Calif~mia.’~ The pipeline was out of service for 324 days and operated at reduced pressure when first returned to service. RSPNOPS cannot say, however, that the supply/demand imbalance that resulted in the $5 price increase lasted for this entire period. There were other pipelines delivering gas to 22 . Final Report on Price Manipulation in Western Markets: Fact-Finding Investigation of Potential Manipulation of Electric and Natural Gas Prices”, Docket No. PA02-2-000, Prepared by the Staff of the Federal Energy Regulatory Commission, March 2003. 23Based on applying the estimated $5 per thousand cf to average daily deliveries from Arizona to California during 2001, estimated in Energy Information Administration, Natural Gas Annual 2001, Table 12, to total 3,440 million cubic feet per day. 27#

Page 29Califomia during the period, and the market likely adjusted to accommodate the reduced supply fiom the out-of-service pipeline. If the reduction continued at its initial magnitude for two months, however, the total effect would be slightly over $1 billion. As described above, implementing the provisions of this rule are expected to significantly reduce the likelihood of accidents resulting in this kind of supply effect on approximately 25 percent of total transmission pipeline mileage. As also described above, accidents of this type would be expected to occur about once each decade if this rule is not adopted. As a result, RSPNOPS estimates that implementing this rule will result in an economic benefit, fiom avoided cost effects of accident-induced supply restrictions, of approximately $500 million during the 20-year period of this analysis. FERC has stated to RSPNOPS that avoiding unexpected gas supply interruptions of this nature would be a major benefit to the U.S. economy. Quantifjlng the avoided cost of potential accident-induced supply interruptions as a benefit should be accompanied by estimation of the cost of supply interruptions that may occur as a result of implementing the rule. These costs are not included in this analysis, for reasons described below. There is a potential for such treatment to significantly increase the estimated costs. While the exact number is under debate, a recent study by EEA under contract to INGAA has estimated the cost to consumers associated with supply restrictions over the first twenty years following implementation of the proposed rule would be $12.4 billion. Several assumptions have been made in the EEA/LNGAA analysis that could cause the predicted cost to swing significantly up or down, including: 6. 7. 8. 9. That operators have perfect knowledge of each other’s assessment plans, so they can plan an optimum assessment schedule. This assumption could clearly lead to an under- prediction of consumer cost impacts. That the mix of assessment techniques used by operators during the time when the baseline assessments overlap the first reassessment will be the same as for the baseline. This assumption ignores the fact that the rule would allow use of confirmatory direct assessment (CDA) for reassessment or low-pressure reassessment of most lines during that overlap period. These methods result in much less disruption of gas service. This assumption will lead to significant over-prediction of consumer costs. That the cost of all gas will follow the spot market price. In reality, much of the gas supply is delivered under long-term agreements for which costs are more stable. That there would be no regulatory means for operators to avoid supply interruptions. Provisions in PSIA-2002 allow waivers from required reassessment schedules to mitigate supply interactions. Those provisions are incorporated in this rule. This allows operators added flexibility in avoiding supply interruptions. RSPNOPS strongly believes that the waiver authority provided in PSIA-2002, and this rule, represents a significant tool in our ability to help relieve supply impacts. Quicker Return to Service Following Required Pressure Reduction 28 , .#

Page 30Perhaps the most important consequence of the Carlsbad accident was that the pipeline was out of service for a total of 324 days. Even then, the line was retumed to service, pursuant to requirements imposed by RSPNOPS, at reduced pressure. Pressure was increased in steps as additional inspections were performed and confidence was gained, but operation at full pre- accident pressure did not resume for approximately an additional year. The reason for this extensive period and cautious approach to return to service was that additional work needed to be performed to ascertain the condition of portions of the pipeline not involved in the accident. Since the accident occurred, the integrity of the entire line was put in question. Additional information needed to be gathered to permit an evaluation of its integrity. Information about the integrity of much of the pipeline will be more readily available in the future as a result of the requirements of this rule. Accidents are not the only reason that pipelines are shut down or operate at reduced pressures. RSPNOPS imposes orders, usually requiring a reduced operating pressure, when incidents or compliance issues necessitate such action. Over the last ten years, fifteen such orders have been issued. Pipeline operators must reduce pressure when such an order is issued, and may not retum to full pressure until RSPNOPS approves of that action. The time required to return to full pressure is dictated by the need to gather additional information about the pipeline in order to assure that operation at full pressure is safe. As noted above, 25 percent of transmission pipeline mileage is expected to be assessed periodically to comply with this rule. This will provide current information about that portion of the pipeline mileage, which will be available to be used following any order requiring reduced- pressure operation. The requirements that operators identify and remediate anomalies found during assessments will also improve the confidence that RSPNOPS will have in that portion of the pipeline. Finally, the data integration requirements in this rule will make other information about the entire pipeline more readily available, and the availability of this information will also shorten the time necessary to retum a pipeline to full-pressure service. As described above, the economic consequences of the Carlsbad accident included a $1 7.25 million per day impact on the Califomia economy. This impact can be used to estimate the benefit that will be realized by allowing future pressure-reduction orders to be lifted sooner. The impact must be adjusted to reflect the conditions normally imposed by corrective action orders. The initial impact of the Carlsbad accident resulted from the simultaneous complete shutdown of three parallel pipelines. Most orders result in reductions that are not as extreme. Most corrective action orders impose a required pressure reduction of approximately 20 percent. The reduction in gas flow that results from such a pressure reduction differs based on the initial operating pressure and other variables. For most pipelines, the reduction is approximately ten percent of the pre-reduction flow rate. Further, most orders do not affect three pipelines simultaneously, and the affected pipelines are often smaller than the 30-inch diameter lines involved in the Carlsbad accident. The average diameter of pipelines for which correcti\)e action orders were issued in the last 10 years is approximately 25 inches. The economic impact of the 29 ' 0 .#

Page 31’ Carlsbad accident must be reduced by approximately 77 percent to reflect the smaller capacity of one 25-inch pipeline compared to three 30-inch lines. The impact must further be reduced by 90 percent to reflect that most corrective action orders result in approximately a 10 percent reduction in flow rather than complete shutdown. Nevertheless, the potential benefit of shortening future shutdowns is significant. An average of 16 months expired before RSPNOPS authorized return to full pressure for the corrective action orders issued in the last ten years for which such authorization has been granted. RSPNOPS estimates that this time could have been reduced by one half if better information had been available regarding the pipelines, similar to the information that this rule wiil require operators to compile. The historical experience thus implies that future imposed pressure reductions will be 8 months shorter. Returning a single pipeline to full pressure (from 80 percent pressure) this much sooner would result in an economic benefit of approximately $96 million, based on the estimated impact of the Carlsbad accident as adjusted for the different circumstances. During the past ten years, 15 orders have been issued requiring pressure reduction. This implies that 30 such orders would be likely to be issued during the 20-year period covered by this analysis. Return to full-pressure service would not necessarily be quicker in all of these cases., As noted earlier, 25 percent of the total gas transmission pipeline mileage will be inspected as a result of this rule. Information will thus be available for 1 out of every 4 miles of pipeline that could allow quicker return to full pressure. This means that 7.5 orders are likely to be closed quicker over this period. The total benefit that will be realized from returning these pipelines to service sooner is thus estimated to be approximately $719 million. This benefit will be higher, of course, ifany of the pipelines for which pressure reduction is ordered suffer an accident or for any other reason must be completely removed from service for any period. . . Consideration of Increases in Operating Pressure Pipeline safety regulations presently limit pipeline operating pressures in order to limit stresses in the pipe, thus providing a safety margin. The allowable pressure is based on the pressure at which the pipe has been tested, which is, itself, determined by the ultimate strength of the pipe. In rural areas, pipelines are allowed to operate at pressures that induce stresses in the pipe wall equal to 80 percent of those that have been demonstrated acceptable by pressure test. In class 3 and 4 areas, which would be included among high consequence areas under the rule, the corresponding limits are 66.7 and 55.5 percent respectively. The corresponding hoop stress may not exceed 72 percent of S M Y S in class 2 locations, 60 percent of S M Y S in class 3 locations, and 50 percent of S M Y S in class 4 10cations~~. The reason for these lower limits has been to provide additional margin against accidents in areas where the population near the pipeline is higher. The margin is, in part, to account for unknown problems and pipe degradation that could result in accidents. 2449 CFR 192.61 1, “Change in Class Location: Confirmation or Revision of Maximum Allowable Operating Pressure”, paragraph (a)( 1) 30#

Page 32This rule will require operators to inspect natural gas transmission piping in areas where the largest concentrations of people exist near the pipeline. Anomalies that could threaten pipe 1 integrity will be identified and appropriate remedial actions will be taken. This will improve knowledge of the condition of the pipe and reduce the need for additional margin, in the form of lower stresses in the pipe wall, to account for uncertainties regarding pipe condition. Accordingly, this rule could provide a basis under which RSPNOPS could approve operation of some natural gas transmission pipelines at higher pressures than are presently allowed. (The particular circumstances of each area would need to be taken into account in deciding whether operation at increased pressures is acceptable). RSPNOPS has some experience in granting such approvals. Operators who participated in the Risk Management Demonstration Program (RMDP) implemented various activities to improve their knowledge of the factors causing risk on their pipelines and to address those factors. Many of the activities impIemented were similar to elements required in integrity management programs mandated by this rule. Some natural gas pipeline operators participating in RMDP were granted authority to implement risk management alternatives in lieu of reducing pressure or replacing pipe in areas where population growth had resulted in a change in class location. The effect was operation of the pipe at a higher pressure than otherwise would be required by 49 CFR 192.61 1, based on the improved knowledge of, and control over, risk in those areas. Mature integrity management programs, addressing risks in high consequence areas could provide a basis for authorizing operation at higher pressures in the same manner as did RMDP alternatives. For a natural gas transmission pipeline, increased operating pressure results in additional throughput, i.e., delivery of larger quantities of gas without need to replace the pipe with one of larger diameter. An increase in operating hoop stress from 72 to 80 percent SMYS, for example, would result in at least 11 percent more gas throughput. The possibility of operating pipelines at higher pressures thus affords operators the opportunity to increase natural gas deliveries from the existing pipeline infrastructure. On a long-term basis, pressure increases could obviate or delay the need for some new pipelines. It would also increase the availability of natural gas to meet all of the needs described earlier. Informal discussions with pipeline operators have indicated that the increased gas deliveries that would result from an increase in pressure associated with a ten percent increase in allowable pipe wall stresses would more than offset the costs of complying with the requirements of this rule, including the costs required to make a line piggable. In emergency situations, like that experienced in California in 2000, temporary increases in operating pressure could make additional gas available in rapid order to alleviate the emergency. RSPNOPS has the authority to allow increases above the stress levels allowed by current regulations via waivers. RSPNOPS would only exercise that authority if it had confidence that the pipeline would operate safely at the higher pressures. The elements of integrity management programs required by this rule would go a long way towards engendering that confidence. Integrity management programs that operators would implement to comply with PSIA-2002, i.e., 31#

Page 33if there were no rule, would likely not produce the necessary confidence. Provisions in this rule, not in the Act, that would contribute to the necessary confidence include: 1. 2. 3. 4. The requirement that operators follow national consensus standards. Operators would likely use these standards for guidance if they implemented programs without a rule, but the degree of adherence to standard provisions would likely be less. The requirement that lessons learned be applied, as appropriate, to the entire pipeline. This will help assure that integrity problems do not occur in areas outside high consequence areas. No such assurance would be likely in programs implemented without this rule. The requirement that performance measures be implemented, and that some of these measures be reported to RSPNOPS. An increasing trend of pipeline performance will be a significant element in assuring confidence in the pipeline. That consistency and quality of operator integrity management plans, and their implementation, can be known. This can only be assured through regulatory oversight, and the rule provides the basis for such oversight. RSPNOPS has not quantified the benefit afforded by the ability to approve operation at increased pressures. That benefit depends on the nature of future energy emergencies and on the location and effect of potential shortfalls in gas supply. These are unknown at this time. When emergencies occur, as seen in California in 2000, any ability to increase the supply of gas to an affected area is a significant benefit. Cohsideration of Alternatives to Pipe Replacement for Class Location Changes As described above, existing regulations require that pipelines operate at lower pressures in areas where the population along the pipeline is greater. As population near the pipeline grows, the class location of a segment of pipe may change. Operators are required to perform a study whenever population increase indicates such a change may have occurred.2s If the study confirms a change in class location, operators must reduce the pressure in the pipeline within 18 months.26 A reduction in pressure would, however, mean lower throughput. Areas in which the population is growing, Le., where class location changes occur, are likely to need more, not less, natural gas. As a result, operators almost always replace pipe, using pipe with thicker walls, so that operation can continue at the same pressures while reducing the stress in the pipe wall to comply with the regulations. INGAA conducted a survey in 1996 in which it asked members to report the costs they incur in meeting class location change requirements. Operators representing 166,000 miles of transmission pipeline responded. INGAA reports that the annual costs reported by these operators, if multiplied by 20 years, would total $1.3 billion for pipe replacement and $149 2549 CFR 192.609, “Change in Class Location: Required Study” 2649 CFR 192.61 1, “Change in Class Location: Confirmation or Revision of Maximum Allowable Operating Pressure” 32#

Page 34million for pressure tests to justify higher operating pressures. Extrapolating these costs to all 225,000 miles of long-distance pipeline results in a 20-year estimate of expenditures totaling, $1.96 billion. Correcting for inflation provides an estimated expenditure of $2.02 billion over the 20 years considered in this analysis. LDC operators may also incur expenses related to class location changes, but they are expected to be much lower, largely since many LDCs already design and operate their pipeline systems as though all areas were class 3 or higher. For purposes of this analysis, LDC costs have been conservatively ignored. RSPNOPS would not change the class location requirements to eliminate the need to replace pipe, at least not initially. Rather, RSPNOPS would entertain requests for waivers from these requirements. As described above, RSPNOPS has experience approving such waivers under the Risk Management Demonstration Program, and this rule will provide information similar to that relied upon for those waivers. That information would not be available until after a baseline assessment had been performed on a given segment. The number of pipeline segments potentially eligible for such waivers would therefore be small at first, but would increase over time until, after 10 years, all transmission pipeline segments in high consequence areas would potentially be eligible. RSPNOPS is likely to be cautious in approving initial waivers. As confidence is gained, and additional information is made available from future assessments, the npmber of waivers approved is likely to increase. It is not possible to estimate precisely how much pipeline mileage might benefit from waiver of class location change requirements. It is possible, though, that relief could be granted by waiver for up to half of all pipeline that would otherwise be subject to those requirements. The benefit to industry, over the course of this analysis, thus'could be as high as $1 billion. Experience gained through review and approval of waivers would also establish a record that could be used at a later time to modi@ or eliminate the class location change requirements. If this is done, hture benefits could be even greater. Improvements in Pipeline Testing Technology This rule will provide a spur to development of new and improved methods of pipeline inspection. Internal inspection of natural gas pipelines has heretofore not been required. Some operators have implemented voluntary testing programs. This rule, and its companion rules for hazardous liquid pipelines, will significantly increase the demand for intemal pipeline inspection services. This demand will be long term and reliable, since the rules require periodic re-inspection. (Voluntary programs might not have resulted in re-inspection of pipelines or in re-inspection at much longer intervals). In the short term, growth is expected among the companies providing inspection services for natural gas transmission pipeline operators. In the longer term, increased competition to provide these services can be expected. The relatively improved economic position of inspection services companies will allow them to invest in research to improve their inspection technology. Improved technologies might be able to detect anomalies that can not now be identified by intemal inspection. Research might be able to improve or develop new techniques to evaluate pipe using direct assessment. Improved methods 33#

Page 35may allow inspections to be done more efficiently, at reduced cost to operators and with less interruption to natural gas service. Inspection companies will have an incentive to develop these improvements in order to improve their ability to obtain contracts from pipeline operators to conduct inspections. Society will benefit from improvements in inspection technology through the safer pipe that will result from identifjmg and remediating anomalies that can not now be addressed. , Summary of Benefits The benefits that will result from implementation of the requirements in this final rule are summarized in Exhibit 3. Exhibit 3. Summary of Expected Benefits (in millions of dollars, over 20 years) Benefit 1 Value Reduced death and serious injury and reduced property damage Avoiding a very-high consequence accident Reduced costs from preventive maintenance vs. accident response Avoiding the economic impacts of supply-interrupting accidents Relief from class location requirements to replace pipe ~ ~ ~ ~~ Quicker return to service following an accident Improved basis for public confidence in pipeline safety, e.g. improved ability to site and construct new pipelines Reduced consequential damages from unexpected interruption of natural gas service Facilitate consideration of increases in operating pressure Foster improvements in pipeline testing technology 800 1,676 4 ’ 500 1,000 I 719 Not Estimated Not Estimated Not Estimated Not Estimated Total Estimated Benefits 4,699 COSTS 34#

Page 36The final rule requires that operators, within one year of the effective date: (1) identify pipeline segments that are in high consequence areas, (2) prepare a written plan for initial (or baseline) assessment of all pipeline that could affect a high consequence area, and (3) prepare a framework addressing each element of an integrity management plan for their pipelines. These documents will detail testing methods to be used, risk factors considered in selecting the appropriate testing methods for each particular high consequence area, and the schedule of testing and inspection. Appropriate testing methods include: (1) pressure testing, (2) internal inspection, (3) direct assessment, (4) equivalent alternatives (in terms of knowledge of the pipeline provided), and (5) confirmatory direct assessment (for those segments for which regular assessments are conducted on intervals greater than seven years). Once the plans have been prepared they will be used for baseline integrity testing. That testing must be completed within ten years of the effective date of the final rule. (Half of the testing must be completed within the first half of the required period). RSPNOPS inspections will verify the plans and assure they are implemented thoroughly. The rule requires that pipeline operators retest their pipeline mileage in high consequence areas periodically. Assessments of some type must be conducted no less frequently than once each seven years. Full retests must be conducted at least once every ten, fifteen, or twenty years, depending on the operating pressure of the pipeline. Pipeline operating above 50% S M Y S must be tested every 10 years. Pipeline operating between 30 and 50% S M Y S must be tested every fifteen years. Low- pressure pipeline, Le., that operating below 30% SMYS requires full retest every twenty years. Full retests may be required more frequently, depending on risk factors, and operators must consider the need for more frequent testing as part of their risk analysis and integrity management program. The risk from pipeline operating at low pressures is different. Failures on such pipelines occur as leaks, rather than ruptures. The most important threat is usually outside force from third party damage. In recognition of these differences, this rule establishes different requirements for the seven-year reassessment of low pressure pipelines. Those requirements focus on assuring that protective systems are functioning properly and include leak surveys. Additional protection is also afforded by requiring additional preventive measures, to reduce the chances of third party damage. These protective measures apply to all low pressure pipeline in class 3 and class 4 areas, including pipeline not in high consequence areas. 35#

Page 37Pipeline operators must also evaluate their pipeline segments in high consequence areas to determine whether installation of automatic shutoff valves or remotely controlled valves, or other preventive and mitigative devices, is necessary to reduce risk. Operators would be required to install such valves where they are found necessary. Based on the foregoing requirements, the costs that can be expected to result from the regulatory change will be those associated with the major provisions of this rule, which are: 1. Identifylng pipeline segments that can affect high consequence areas 2. Framework - setting up integrity management program; 3. Baseline assessment - internal inspection, pressure testing, or direct assessment; 4. Periodic assessment (inspection) & evaluation; 5 . Evaluating automatic shutoff and remotely controlled valves; 6 . Data integration 7. Preventive Measures for low-stress pipeline in class 3 and class 4; and 8. Remedial action. Number of Pipeline Operators Affected RSPNOPS estimated, in the draft regulatory analysis supporting the proposed rule, that 668 operating companies would be affected by the proposed rule. This was based on a review of annual reports submitted to OPS for 2000. In excess of 1000 entities submitted reports. RSPNOPS reduced this number by taking into account known ownershiphbsidiary linkages among the reporting entities. Upon hrther review, RSPNOPS has concluded that there are additional linkages and the actual number of companies that will be affected by this rule is much less than 668. RSPNOPS has taken a different approach to estimating the number of affected companies in the final analysis. INGAA’s web site lists 29 member companies. INGAA’s comments indicate that its members operate approximately 180,000 miles of transmission pipeline, and that there are an additional 45,000 miles of such pipeline operated by companies that are not members. RSPNOPS based its estimate of the number of long-distance pipeline companies that will be affected on the INGAA membership total. RSPNOPS increased this total in recognition of the fact that there are long- distance pipeline operators who are not members of INGAA. For purposes of this analysis, RSPNOPS estimates that there are 35 companies in the long-distance category. AGA’s comments state that there are 166 gas distribution utilities that own and operate transmission pipelines. This includes members of AGA and APGA and companies that are not members of either association. RSPNOPS has relied upon this estimate in this analysis. There are other operators of gas transmission pipelines that are not utilities. This includes intrastate transmission companies and customer-owned service lines. The limited amount of pipeline mileage operated by these companies makes their situation more like that of LDCs than of long-distance operators, and they are included in that category in this analysis. RSPNOPS’s 36#

Page 38’ ’ I review of the 2001 annual reports identified approximately 75 companies that were either intrastate transmission companies or operated customer-owned service lines. RSPNOPS expects that some costs will be less for companies operating only a few miles of transmission pipeline. In the draft regulatory analysis, we assumed that operators with less than 40 miles of pipeline would see lower costs. For this analysis, we have revised that threshold to 30 miles. RSPNOPS’s review of 2001 annual reports identified 89 companies operating less than 30 miles of transmission pipeline, and this number is used in this analysis. Per-company costs for “LDC operators” are thus applied to 240 companies, 15 1 operating more than 30 miles of transmission pipeline and 89 operating less than 30 miles. Costs for long-distance operators are applied to 35 companies, all assumed to operate more than 30 miles of pipeline. The Costs of Identifying Pipeline Segments in High Consequence Areas Natural and other gas transmission pipeline operators are currently required to monitor the population along their pipeline. The number of occupied dwellings in a sliding mile within 660 feet of the pipeline has been the basis for determining the “class” of the pipeline for many years. ‘‘Method a” of the definition of high consequence areas for natural gas pipelines builds off this existing knowledge, starting with the portions of the pipeline that are class 3 and class 4. This information should already be known to the operators, and there is therefore no cost in identifjmg class 3 and class 4 areas as a result of this rule. , . . (Some local distribution company - LDC - operators treat all of their piping as though it were class 3 or 4 rather than performing the population monitoring that would be required to determine the specific class location of a particular segment of pipe. This approach will also be acceptable for the final rule, although all requirements of the rule would then apply to all of the company’s transmission piping. OPS presumes that all of the pipeline mileage of LDC operators following this approach was included in the mileage totals submitted by AGA and APGA. Since this analysis bases its estimate of the affected mileage on those comments, the entire transmission mileage operated by LDCs taking this approach is included in this analysis). High consequence areas for natural gas transmission pipelines, even under method a, involve more than just class 3 and class 4, however. The additional factors include the presence of buildings housing people with limited mobility and places where people congregate in proximity to the pipeline. Some of these locations may exist outside current class 314 locations. Operators using method a will need to conduct additional surveys of the areas near their pipeline to determine if there are any such areas that require additional pipe to be classified as being able to affect a high consequence area. “Method b” of the definition of high consequence areas requires the evaluation of potential impact circles, representing the area that could be affected by a postulated explosion on the pipeline, over the entire length of an operator’s pipeline, Calculation of the size of the potential impact circles themselves is a straightforward exercise. The relevant parameters are maximum allowable 37#

Page 39operating pressure and pipe diameter. These parameters change incrementally along the pipeline. Once the size of the circles is known, the operator must evaluate its pipeline to determine whether any circle contains more than 20 buildings intended for human occupancy or an identified site (meeting criteria specified in the rule). This evaluation will be more resource intensive. Some pipeline, e.g., that operating in undeveloped remote areas or that in very densely populated areas, will be able to be dispositioned easily. Other pipeline mileage will require a detailed evaluation to determine whether the circles contain the number of buildings, or an identified site, that would constitute a high consequence area. ' Existing regulations require that operators monitor the population within 660 feet 'of their pipelines (to determine class location). Operators are therefore expected to have data to evaluate the housing density component of method b for all but the largest transmission pipelines. (The potential impact circle for a 30-inch diameter pipeline operating at 1000 psi is 655 feet). The rule allows operators to use this data to evaluate potential impact circles larger than 660 feet, by prorating the housing density criterion, for a period of three years. This will reduce costs for collecting the necessary additional data for larger pipelines, allowing much to be gathered during the normal course of business. AI1 operators will need to gather data concerning identified sites, and this data must be accumulated within one year. RSPNOPS estimates that each affected operator will incur costs of $10,000 to calculate potential impact circles. This element will thus total $2.75 million. The costs to evaluate potential impact circles will almost all be incurred on those sections of the pipeline that can not be disposed of easily (i.e.','areas that are not remote or highly developed). AGA's comments estimate that evaluating potential impact circles will cost $300 per mile.27 RSPNOPS considers that a per-mile estimate is appropriate, since that is how the pipeline will need to be treated, and has adopted the AGA estimate. The question is how much pipeline will require evaluation? INGAA reported that approximately 6.4 percent of its members' pipeline is in class 3 and 4 areas (based on a survey to which members operating approximately 166,000 miles of transmission pipeline responded). This indicates that much of the 225,000 miles of transmission pipeline in the long-distance category is in rural areas. In many of those areas, only a cursory search for identified sites will be needed, at most. For LDC operators, many potential impact circles will be so small that they will not require detailed evaluation. For example, the potential impact circle for a 6-inch pipeline operating at 150 psi is 50 feet in radius. Circles for smaller-diameter ,or lower-pressure piping will obviously be even smaller. Circles of this size will also be amenable to prompt disposition on the basis of inability to contain 20 housing units. 27L0ri Traweek, AGA, letter to docket dated ApriI 30,2003, RSPA-2000-7666-274, page 8. 38#

Page 40RSPNOPS estimates that long-distance operators will need to evaluate potential impact circles along 25,000 miles of pipeline, roughly 60% more than the number of HCA miles estimated by INGAA (extrapolated for non-INGAA members and members not responding to the survey). RSPNOPS estimates that LDC operators will need to evaluate potential impact circles for 16,000 miles, representing all of the pipeline in this group operating above 50% SMYS, a large portion of that operating between 30 and 50% SMYS, and a small portion of that operating below 30% S M Y S . The cost to evaluate potential impact circles is thus estimated to be $7.5 million for long-distance operators and $4.8 million for LDCs. This makes the total cost for identifjmg high consequence areas $15.05 million. The Costs of Plans and Reports The single most important part of this rule is the requirement for the integrity plan and framework. The creation, development and implementation of these documents will provide for the necessary integration of information regarding pipeline condition. Integration is important to assure RSPNOPS, and the public, that pipeline operators are considering fully the unique risks that gas transmission pipelines pose to high consequence areas. Plan development will assure not only that they are considering these risks but that they have developed a plan that requires extra scrutiny and precautions in these areas to safeguard the public. These safeguards include the use of periodic testing. Evaluation is an ongoing process. Operators will be expected to consider the risk factors and their relative priorities in establishing assessment schedules. This allows operators to develop an internal inspection and testing program that is customized to the particular operating characteristics and risks associated with different portions of their system(s). The plans and reports required by this rule are: (1) a written plan for baseline assessment of all pipelines that could affect high consequence areas, (2) a fiamework addressing each element of an integrity management program, and (3) other documents supporting the decisions made, analyses made, and actions taken in the implementation of the integrity management program. Cost of the Written Plan and Framework Pipeline integrity management plans are relatively new. Such plans were required of hazardous liquid pipeline operators through regulations promulgated in 2000 and 2001. The deadlines under those rules for developing the plans have only recently passed. RSPNOPS has no data on the cost to hazardous liquid pipeline operators for development of the required plans. For long-distance operators: 39#

Page 41. ' INGAA estimated that it would cost an operator $500,000 to develop an integrity management plan and $50,000 per year thereafter to maintain the plan.28 The $500,000 estimate is higher than the value used by OPS in the draft regulatory analysis. RSPNOPS interactions with hazardous liquid pipeline operators who are developing integrity management plans in response to earlier rules indicates that the amount of effort involved is considerable. RSPNOPS concludes that the INGAA estimate appears reasonable. RSPNOPS estimated $1 6,000 per company for plan maintenance in the draft regulatory analysis, included there as part of the reports category (redesignated as recordkeeping in this analysis). RSPNOPS recognizes that $50,000 represents a portion of one full-time professional, and considers this an appropriate estimate for long-term oversight and management of the integrity management plan. RSPNOPS has used the INGAA estimates in this analysis. OPS assumed, in the draft regulatory analysis, that 25 percent of affected operators with more than 40 miles of pipeline already have developed integrity management plans that meet the requirements of the rule. RSPNOPS has learned from its inspections of hazardous liquid pipeline operators that few, if any, integrity management plans developed prior to the integrity management rule did not require significant revision to meet the requirements of that rule. In this analysis, RSPNOPS has not assumed that any operators already have plans meeting the rule requirements. For LDC operators: AGA estimated costs of approximately $1 million to develop plans and $57,000 per year for maintenan~e.~~ RSPNOPS believes these estimates, particularly the development estimate, are too high. (Some of these costs may be associated'with data integration. See discussion below). Many LDC operators will have limited amounts of affected pipeline mileage, particularly after application of potential impact circles. The elements of a plan will be the same as for long- distance operators, but their treatment should be easier for smaller lengths of involved piping. Some operators in this group will have only a few miles of pipeline. This would include, for example, customer-owned supply lines of one or two miles in length. (These operators are included in this group, because their characteristics are more like LDCs than like long-distance operators). Plan development costs for these operators are expected to be considerably lower. RSPNOPS uses the same estimates in this analysis as for long-distance operators (i.e., $500,000 for plan development and $50,000 annually for maintenance) for operators in the LDC group that have more than 30 miles of pipeline. Lower costs are estimated for operators with less than 30 miles of pipeline - $125,000 for development and $25,000 annually for maintenance. Applying these values results in a total estimated cost to develop written plans and frameworks of $1 04.125 million, all of which is assumed to be expended in the first year. (RSPNOPS recognizes 281NGAA spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297 z9L0ri Traweek, AGA, letter to docket dated April 30,2003, RSPA-2000-7666-274, attachment A. 40#

Page 42that plan development is likely to continue after preparation of the initial framework in the first year, but extending these costs over several years would not change the estimated total in this I analysis). Annual maintenance costs are estimated at $1 1.525 million. Recordkeeping ' RSPNOPS estimated in the draft regulatory analysis that recordkeeping associated with integrity management plans, principally consisting of reports, would cost $4,000 per company per year. INGAA's comments estimated a $50,000 per company initial setup charge and $5,000 per year thereafter.30 RSPNOPS accepts that more management involvement and data system changes will be needed in the first year, and that costs in that year will be higher. RSPNOPS has used the INGAA estimates for long-distance operators and LDC operators with more than 30 miles of pipeline. RSPNOPS believes that the setup costs will be less for operators with only a few miles of pipeline. This analysis assumes a $20,000 per company setup charge for LDC operators with less than 30 miles of pipeline. An annual charge of $5,000 is assumed for these companies, as for those with more mileage, since it is unlikely that routine generation and review of reports will be any less costly simply because the reports address fewer miles of pipeline. The total cost for recordkeeping is thus estimated to be $1 1.08 million in the first year and $1.375 million each year thereafter. Total Cost of Plans and Reports Based on the foregoing, the total cost for plans and reports will consist of a one-time cost of ' $1 15,205,000 plus an annual cost of $12,900,00c).3' . . Inspection and Testing The rule requires baseline and subsequent testing of the impacted mileage using in-line inspection, pressure testing, direct assessment, or alternative methods. Acceptable in-line inspection includes high resolution, low resolution, and ultrasonic pigging. Acceptable pressure testing consists of hydrostatic testing. Acceptable techniques for direct assessment are as described in an industry consensus standard. Acceptable alternative methods include any other methods that would provide a level of safety equivalent to that provided by the specified methods. In addition, a more focused application of direct assessment, confirmatory direct assessment, would be required for pipeline segments that would be assessed using one of the other methods at intervals greater than seven years. Internal inspection (pigging) requires that an instrument (pig) be inserted into a pipeline, travel through the line, and be removed. The points at which the pig is inserted and removed are referred 301NGAA spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297 "This does not include cost incurred by the Federal government in setting up the review process (including development of review protocols and training) and in the actual review of the plans and programs. 41#

Page 43to as launchers and receivers. These are usually permanent installations, and are often located at compressor stations. An individual pig inspection thus covers an amount of pipeline mileage roughly equal to the spacing between compressor stations, which is typically about 50 miles for long-distance pipelines. Baseline Testing The rule requires that baseline testing be completed within ten years. As noted above, at least some gas transmission pipeline operators currently have integrity management programs that include some testing. It is therefore likely that some of the affected mileage would be tested even if the final rule were not promulgated. The Interstate Natural Gas Association of America (INGAA) has estimated that 25 percent of the natural gas transmission pipeline system has been pigged since inspection devices became generally available in 1980? This testing did not occur at a constant rate over that period. Rather, pigging was conducted sparingly in the early 1980s and the amount of pipeline pigging being performed has increased in recent years. For purposes of this analysis RSPNOPS assumes that the current rate of pigging for natural gas transmission pipelines would inspect 25 percent of the transmission piping in approximately 10 years, or 2.5 percent per year. The definition of high consequence areas makes it unlikely that these areas will occur uniformly throughout the natural gas transmission pipeline network. Rather, they would be expected to be concentrated in areas with higher population, and to occur much less often in rural areas. This means that it is possible that a pig inspection conducted in a rural area could inspect 50 miles of pipeline and not inspect any pipe segment that could affect a high consequence area. It is likely that some pigging being conducted today is in such rural areas. As a result, it would be unreasonable to assume that 2.5 percent of the pipeline that can affect high consequence areas would be inspected each year if the final rule were not promulgated. For purposes of this analysis, RSPNOPS assumes that 1.5 percent of the pipeline in high consequence areas would be pigged each year under current industry practices and that this rate of pigging would continue indefinitely if the rule did not become effective. Subsequent Testing Once baseline testing has been performed on a segment of pipe, the rule requires that subsequent testing be undertaken on that segment. Some type of assessment, at a minimum consisting of application of confirmatory direct assessment or low-pressure reassessment, must be conducted on all affected pipe segments at least every seven years. Testing using one of the other accepted methods must be performed, based on risk factors, at least once every ten, fifteen, or twenty years, depending on the operating stress level. 3Z"Consumer Effects of the Anticipated Integrity Rule for High Consequence Areas," prepared for the INGAA Foundation, Inc. by Energy and Environmental Analysis, Inc., 2002 42#

Page 44The planned rate of re-testing of pipeline by operators if this rule is not promulgated is unknown. Some operators currently have programs under which pipe is tested periodically. It is also likely that some pipe being tested for the first time might never be reinspected absent this rule. RSPNOPS has estimated that testing would continue indefinitely at the present rate, i.e., 1.5 percent of mileage that can affect high consequence areas would be inspected by pigging annually. RSPNOPS has assumed that no re-testing by hydrostatic tests or direct assessment would be performed without the requirements of the rule. r Operators will need to do some subsequent testing at the same time they are doing baseline testing. This results from the combination of a ten-year baseline period and the requirement that some form of reassessment be performed every seven years. Pipeline segments that are baseline tested in year 1 will require reassessment in year 8, while other baseline assessments may still be underway. This “overlap” will occur in years 8,9, and 10. The rule allows use of confirmatory direct assessment to perform the reassessments required in those years. Confirmatory direct assessment results in the least service interruption of any assessment method. Use of this method should help to minimize gas supply problems that could occur when pipeline segments are taken out of service for assessments. The Costs of Testing For the purposes of this analysis, it is assumed that all required testing will be accomplished by either (1) hydrostatic testing (pressure testing), (2) smart pigging (internal inspection), (3) direct assessment, or (4) confirmatory direct assessment (for interim retesting of those pipeline segments tested using one of the other methods on intervals exceeding seven years). Alternative methods are not yet to the point where their costs for pipeline testing can be reliably estimated. Hydrostatic Testing The total cost of hydrostatic testing has been previously estimated by RSPNOPS to be $4,656 per mile in 1990 d0llars,3~ which equates to $5,274 per mile in 2001 dollars. (This estimate, as well as those for other testing methods, does not include the cost of any necessary repairs). This cost was used in the draft regulatory analysis to estimate costs resulting from hydrotesting. Comments from industry indicated that this value seriously underestimates the costs operators expect to incur to test using this method. INGAA estimated, in its comments, that the cost to hydrostatically test long-distance pipelines is $29,700 per mile.34 AGA reported an average cost of $40,000 per mile for this work in the more- 330f€ice of Pipeline Safety, “49 CFR Part 195 Economic Evaluation, NPRM - Hydrostatic Testing of Certain Hazardous Liquid and Carbon Dioxide Pipelines,” Docket No. PS-121, Notice 1 , May 13, 1991. 341NGAA spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297 43 I .#

Page 451 urban environments in which their members operate.35 RSPNOPS recognizes that changed circumstances (e.g., new environmental requirements) could have increased hydrostatic testing costs since the 1990 study. RSPNOPS has used the INGAA and AGA estimates in this analysis as the costs for pressure testing of long-distance pipeline and LDC pipelines respectively. Smart Pinning The total cost of smart pigging has been previously estimated by RSPNOPS to be $2,839 per mile in 1992 dollars,36 which equates to $3,2 10 per mile in 2001 dollars. This estimate does not include the cost of making a pipeline piggable (Le., adding pig launchers and receivers or m o d i k n g pipeline that cannot pass instrumented pigs). Again, the industry trade association comments indicated that these costs have been underestimated. INGAA estimated pigging costs to be $3,669 per mile.37 AGA estimated costs for this method at $9,600 per mile.38 As with pressure testing, RSPNOPS acknowledges that costs to conduct this work may have increased since 1992 at greater than the level of inflation. INGAA’s estimate is not considerably higher than the inflation-adjusted RSPNOPS estimate. The AGA estimate is higher, but not unreasonably so, recognizing the higher costs to perform work in an urban environment and the normally shorter distance between launchers and receivers in those areas (over which the fixed costs of conducting an inspection must be distributed). RSPNOPS has again used the INGAA and AGA estimates as the costs for performing pigging on long-distance pipeline and LDC pipeline respectively. Modifving Pipelines for Pig& Much natural gas transmission pipeline is not currently piggable. Comments submitted by the trade associations estimate the percentage of member pipelines in various categories describing . their ability to be pigged as: , Exhibit 4. Piggable Status of Natural Gas Transmission Pipeline (percentage of total mileage) 35L0ri Traweek, AGA, letter to docket dated April 30,2003, RSPA-2000-7666-274, page 10. 360ffice of Pipeline Safety, Instrumented Internal Inspection Devices (A Study Mandated by P.L. 100- 561), Research and Special Programs Administration, November 1992, p. 44. 371NGAA spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297 38L0ri Traweek, AGA, letter to docket dated April 30,2003, RSPA-2000-7666-274, page 9. 44#

Page 46Percentageby 1 Piggable I Easy to 1 Hard to I Not I Category Modify Modify Piggable Long- dis tame3' 30 25 43 2 AGA'' 12 10 43 35 APGA 13 0 41 46 RSPNOPS uses the INGAA estimates, in this analysis, as its assumption for the distribution of pipelines operated by long-distance operators. RSPNOPS assumes that the percentages reported by AGA and APGA represent the proportion of all piping operated by their members. OPS further assumes that a higher proportion of the piping operating at greater than 50% SMYS is piggable than these overall averages. It is this higher- pressure LDC pipe where most assessment by pigging will occur. Thus, it was necessary to develop estimates for this higher-pressure LDC piping. For purposes of this analysis, RSPNOPS assumes that 30 percent of the LDC pipeline operating at greater than 50% S M Y S is now piggable (Le., the same percentage as reported for long-distance transmission lines). This assumption, combined with the reduction in the overall percentage of LDC pipeline expected to be pigged (described above) significantly reduces the amount of LDC pipeline that will require modifications. The AGA estimates submitted to the docket assumed modification of some hard-to-pig piping. RSPNOPS assumes that these modifications will no longer be made, since the final rule will allow use of direct assessment as an equivalent primary assessment mechanism without condition. RSPNOPS has assumed that only pipeline that can be easily modified will be changed by LDCs. The costs to perform the necessary modifications will vary between long-distance and LDC pipelines: For long-distance operators: INGAA estimated costs to add launchers and receivers, make other modifications, and for gas that will be lost when blowing down lines to make modifications for two types of changes. The first was for 40-mile sections requiring little change other than the addition of launchers and receivers. The second was for 1 O-mile sections requiring more significant modifications or pipe replacement. The INGAA estimates for these changes were $1 1,75O/mile and $76,50O/mile, re~pectively.~' This 391NGAA spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297. 40L0ri Traweek, AGA, letter to docket dated April 30,2003, RSPA-2000-7666-274, page 6. 4'INGAA spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297; unit costs derived by dividing stated costs for adding launchers, modifying hard-to-pig pipe, and lost gas over stated number of miles. 45#

Page 47analysis uses these estimates as the cost to modify long-distance pipeline that can easily be modified and hard-to-pig pipe respectively. For LDC operators: RSPNOPS assumes that much less pipe will be modified for pigging by LDC operators, since costs to do so are very high and direct assessment is an available assessment option. Nevertheless, AGA estimates that the percentage of pipeline that will be pigged exceeds the percentage of it’s members’ pipeline that is now piggable. The amount of “easily modified” pipeline exceeds the difference, so OPS assumes that all LDC pipeline that will be modified will come from this category. AGA estimates that such pipe can be made piggable for “less than $50,000 per mile.”42 RSPNOPS has used $50,000 per mile as an estimate for these modifications. Direct Assessment Direct assessment is a technique for assessing the integrity of line pipe that is in the early stages of development. Industry consensus standards governing the application of direct assessment were only recently approved. The process consists of several steps. The first step is pre-assessment. This involves collection and evaluation of data regarding potential threats to pipeline integrity and their relative risks. Based on this evaluation, locations on the pipeline are selected for indirect inspection. Inspection requires use of a minimum of two different “tools”. Examples of direct assessment tools include close interval surveys (CIS), direct current voltage gradient (DCVG), and pipeline current mapper (PCM). The results of the indirect inspections are then used to identify locations for direct examination, excavation of the pipeline to permit visual inspection. Each use of direct assessment requires at least one direct exam. The final stage in the direct assessment process is post-assessment in which the composite set of data (i.e., risk factors identified in pre-assessment and assessment results) are considered to identify continuing excavation needs, determine if additional assessment technologies are needed, and to establish a re-assessment interval. There is very limited experience at this time from which to estimate the costs of performing direct assessment on natural gas transmission pipelines. INGAA estimates these costs at $1 5,000 per mile,43 and RSPNOPS has used this estimate for long-distance pipeline. AGA reported that their members estimate costs for direct assessment at $7,000 to $8,000 per mile, without any verification digsM AGA further commented that verification digs can cost from $2,500 to $250,000 with an average of $40,000 for a typical large transmission line. These estimates did not prove particularly useful due to the wide range stated for verification costs and the expectation that much LDC pipeline is. not “typical large transmission line”. Pacific Gas & Electric (PG&E), which is performing direct assessment in rural and urban environments as part of research and development 42L0ri Traweek, AGA, letter to docket dated April 30,2003, RSPA-2000-7666-274, page 6. 431NGAA spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297. 44L0ri Traweek, AGA, letter to docket dated April 30,2003, RSPA-2000-7666-274, page 9. 46 ‘ I :#

Page 48to validate the process, reported an average cost to perform direct assessment in LDC environments of $29,000 including verification digs. PG&E reported that they are using this value in their , budget planning. RSPNOPS has used the same value in this analysis for the cost to perform direct assessment on LDC pipelines. Confirmatory Direct Assessment Confirmatory direct assessment is a more-focused application of the principles and techniques of direct assessment. This method is concentrated on identifjrlng critical segments of suspected corrosion. The effort involved is less than full application of direct assessment. ’ The OPS has structured the requirements for confirmatory direct assessment in a manner intended to allow maximum flexibility for operators. Indirect examinations may be performed using only one, rather than two, tools. Corrosion regions may be larger than for regular direct assessments. The number of excavations required per region is less. These changes will allow operators to plan and conduct confirmatory direct assessments in a manner that is most cost-effective, i.e., identifies areas of concern at lowest cost. There is no data available at present regarding the cost to implement confirmatory direct assessment. In their comments, INGAA estimated that confirmatory direct assessment (CDA) would cost $1000 per mile.4’ AGA estimated $3,360 per mile, but that appears not to include verification digs.46 The rule requires at least one verification dig per CDA Region. RSPNOPS has used an estimate of $8,000 per mile for CDA costs to LDCs, reflecting the required verification. 3 % Additional Mileage Must Be Tested It is usually not possible to hydrostatically test or pig pipe in high consequence areas without also testing some adjacent piping. Hydrostatic testing requires valves that can isolate the pipe segment being tested. Pigs must be run between available pig launchers and receivers, which are seldom located immediately adjacent to the boundaries of high consequence areas. Valves are spaced closer together than launchers and receivers, meaning that the amount of additional mileage that must be subjected to hydrostatic tests is less than that which will be pigged. RSPNOPS, in the draft regulatory analysis, estimated the amount of overtest mileage at 200 percent for pigging and 25 percent for pressure testing. (No additional mileage must be tested for direct assessment). These assumptions had been increased as a result of comments by the Technical Pipeline Safety Standards Committee. Industry comments indicated that the revised estimates still significantly underestimated the amount of overtesting that would occur for long- distance pipeline that will be tested by pigging. For long-distance operators: 4’INGAA spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297. 46L0ri Traweek, AGA, letter to docket dated April 30,2003, RSPA-2000-7666-274, attachment A. 47#

Page 49INGAA estimated the amount of mileage that would need to be pigged in order to inspect all HCA miles by surveying a sample of their members. They reported that 37,976 miles of pipeline would need to be inspected in order to cover 5,244 miles of HCA pipeline expected to be pigged.47 This equates to a factor of 7.25. RSPNOPS has used this factor for overtesting mileage for pigging of long-distance transmission lines. INGAA’s analysis used a multiplier of 1.3 (Le., 30 percent additional mileage) for pressure testing. RSPNOPS adopts the same factor. For LDC operators: The analysis submitted with AGA’s April 30,2003, letter to the docket assumed 100 percent overtesting for pigging and no overtesting for pressure testing.48 RSPNOPS has adopted these estimates. Choice of Assessment Method Informal discussions with pipeline operators suggest that hydrostatic testing is the least preferred method for assessing pipeline integrity. This was confirmed in discussions at the July 18,2002, TPSSC meeting, and in comments submitted in response to the proposed rule. This is at least in part because hydrostatic testing can be destructive, while the, other two methods are not. Hydrostatic testing also usually requires that a line be taken out of service for a longer period than does pigging. (Direct assessment requires limited out-of-service time or curtailment of pipeline capacity). In addition, care must be taken in drying the pipeline subsequent to a hydrostatic test to assure that all moisture is removed. Remaining moisture can cause internal corrosion problems and can also lead to operational problems in freezing weather. Pigging appears to be the preferred method of assessment for pipeline. As described above, the per-mile cost for pigging is less than that for direct assessment or hydrostatic testing. INGAA’s response to the OPS’s June 27,2001, Federal Register notice indicated that the percentage of its members piping in class 3 and 4 areas that has been pigged at least once is approximately equal to the percentage of their mileage that is easily ~ i g g a b l e ~ ~ , despite the fact that there is currently no requirement that pigging be performed. The choice between direct assessment and pigging is not so obvious for pipeline that must be modified to be made piggable. Direct assessment has the advantage of requiring little or no interruption in normal operation of the transmission pipeline. It would likely be the method of choice for assessing those areas where supply interruption can be most problematic (e.g., single- 47 INGAA spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297. 48L0ri Traweek, AGA, letter to docket dated April 30,2003, RSPA-2000-7666-274, attachment A. 49 INGAA, Subject: Pipeline Safety: Pipeline Integrity Management in High Consequence Areas (Gas Transmission Pipelines) {Docket No. RSPA-00-7666; Notice 2}, no date, page 45. 48 1 ’.#

Page 50source laterals supplying small distribution operators). On the other hand, the costs for direct assessment are higher than for pigging. Finally, while the direct assessment process produces a significant amount of information about the pipeline, pigging still provides operators the most information from actual examination of the pipe wall. The percentage of affected pipeline that will be assessed by each method, as assumed in this analysis, is estimated in Exhibit 5: Exhibit 5. Assumed Assessment Method hercen t age) Percentageby I Pig I DA I P r w w e I Method I Long-distance I 62 1 33 I 5 I LDC 20 75 5 These estimates were drawn from the trade association analyses. They have been adjusted to reflect changes of which the associations were not aware when they made these estimates. INGAA estimated that 225 1 miles of affected piping would be assessed using direct assessment for the first 5-year reassessment, based on an INGAA Foundation report. This is 26 percent of the total HCA mileage estimated under INGAA’s bifurcated definition. Conditions on the use of direct assessment as a primary assessment method (principally that direct assessment could not be used unless it was not possible to assess using pigging or pressure testing) have been eliminated since the Foundation’s report. RSPNOPS has adjusted the percentages between DA and Pressure Testing to reflect its understanding of the strong preference of operators to use a method other than pressure testing. AGA’s comments estimated that the relative percentages would be 35,60, and 5 if direct assessment were treated equivalently to the other assessment methods. Again, these estimates were made before the conditions on use of DA as a primary method were eliminated. RSPNOPS has adjusted the AGA estimates to reduce the amount of pigging and increase the amount of pipeline treated with direct assessment, reflecting the high cost estimated by AGA to modify pipe for pigging. Mileage Tested Per Year Estimating costs requires determining how much transmission pipeline mileage will be tested each year. This requires an estimate of the reassessment periodicity for individual pipeline segments that are subject to the rule. The rule requires that baseline assessments be conducted over a ten-year period and that some kind of re-assessment must be conducted at least every 7 years. The rule allows use of confirmatory 49#

Page 51' direct assessment as an interim reassessment for those segments reassessed using one of the other methods at intervals greater than seven years. Reassessments using those methods are required at 1 0-year intervals for pipe operating above 50% S M Y S , 15 years for pipe operating between 30 and 50% S M Y S , and twenty years for pipe operating at less than 30 percent SMYS. Operators of pipe requiring full reassessment every 15 years would be expected to perform an interim assessment using confirmatory direct assessment seven years after their baseline assessment. These operators would then have the option to conduct another confirmatory direct assessment fourteen years after their baseline and full assessment in the following year, or they could perform their full assessment in the fourteenth year. RSPNOPS assumes that the costs associated with repeating assessments within one year, even if the first were of the limited scope applicable to confirmatory direct assessment, will lead most operators in these circumstances to elect to perform their full assessments every 14 years. Piping for which full reassessment is required every 15 years is thus assumed to be subjected to reassessment every 14 years with an interim confirmatory direct assessment mid-way through the assessment interval. Pipeline that operates below 30% S M Y S is assumed to be subject to a baseline assessment and then confirmatory direct assessment at seven and fourteen years following the baseline. Full reassessments would likely occur at 20 years, but that is beyond the period considered in this analysis. The decision is less straight-forward for operators of pipe required to be assessed at ten year intervals. These operators could perform confirmatory direct assessment seven years following a baseline (or full reassessment) and a reassessment using pressure testing, in-line inspe'ction, or direct assessment at ten years. Seven years could then expire before another confirmatory direct assessment must be conducted, followed in three years by a full assessment. This cycle would repeat. These operators could, instead, perform full assessments at seven year intervals and thus avoid the need to apply confirmatory direct assessment. The decision as to whether to shorten their full reassessment interval, or to use confirmatory direct assessment for interim assessments, will be made by each operator in consideration of the costs and benefits associated with each choice. Holding to the longer reassessment intervals allowed by the rule would involve administrative costs to maintain two different assessment programs (the regular assessment method and confirmatory direct assessment). Field work would also be required twice within three years on affected segments of piping. At the same time, confirmatory direct assessment requires testing of no additional piping beyond that which can affect high consequence areas, has a lower per-mile cost than the other inspection methods, and results in less disruption to pipeline operation. Decisions by individual operators will depend upon factors specific to their pipeline systems, including the total amount of piping that requires assessment and the amount of additional piping that requires assessment when pressure testing or in-line inspection are used. For purposes of this analysis, RSPNOPS assumes that half of the pipeline segments that would be required to be assessed at ten year intervals, and thus to conduct interim assessments using confirmatory direct 50#

Page 52assessment, will instead be assessed at seven year intervals. The other half are assumed to test at ten year intervals and to use confirmatory direct assessment for interim assessments. The number of miles to be tested each year using each assessment method will change as different portions of the affected piping enter reassessment. For the first seven years, all assessments will be baseline assessments. In year eight, reassessments will begin for that portion of affected piping that it is assumed will be tested on seven-year intervals. In the eleventh year, all baseline assessments will be completed and reassessments will begin for that pipe to be reassessed on ten- year intervals. Additional piping will begin reassessment in year fifteen, at which time piping operating between 30 and 50 percent S M Y S will begin reassessment. In each year, the amount of piping assumed to be pigged as a result of this rule has been reduced by the assumption that pigging of 1.5 percent of affected piping would have occurred even without this rule. Confirmatory direct assessments or low-pressure reassessments will also begin for some piping in year seven. These assessments will be conducted for pipe that is to be reassessed using one of the other methods at ten-, fourteen-, or twenty-year intervals. The confirmatory direct assessments will stop in year fifteen for pipe that is to be reassessed on fourteen-year intervals, since this analysis assumes that operators will alternate confirmatory direct assessments and full reassessments on this pipe every seven years. Year after Pigging Hydrostatic Direct Confirmatory DA Effective Testing Assessment Date Base Re Base Re Base Re IO-yr 15-yr 20-yr pipe pipe pipe 1-7 5,125 0 126 0 1,087 0 0 0 8 - 10 5,125 4,261 126 90 1,087 390 4,612 257 Note 1 11 - 14 0 6,403 0 153 0 664 4,612 257 Note 1 ---------- 15 - 17 0 6,403 0 153 0 921 4,612 0 Note 1 18 - 20 0 6,403 0 153 0 92 1 0 0 Note 1 51#

Page 53Cost of Baseline Testing Baseline testing costs are estimated by applying the per-mile cost estimates described above to the mileage to be tested each year. This includes the cost to modify piping requiring modification. INGAA has suggested that the cost of baseline testing could be reduced by allowing operators to credit assessments performed before 1997 (the limit in the proposed rule for crediting prior assessment^).^' The final rule has been revised to allow operators to credit, without time limit, prior assessments that were sufficient to meet the technical standards of the rule, but RSPNOPS has not reduced the estimated baseline assessment costs in this analysis. RSPNOPS considers the effect of this change principally as shifting the timing when costs are incurred, rather than reducing their ultimate total. While costs during the baseline period may be reduced by crediting prior assessments, costs in later years would increase. This is because assessments must be performed at ten, fifteen, or twenty-year intervals, with more-limited confirmatory direct assessments, as a minimum, performed at no more than seven-year intervals. Crediting a prior assessment as a baseline may mean that the assessment that must be performed on that pipeline segment during the baseline period could be a less-expensive confirmatory assessment, but that would mean the next full assessment would need to be done sooner. Some of those assessments would otherwise occur after 20 years, i.e., beyond the period considered in this analysis, and would be moved back into the period of concern. RSPNOPS considers that the net effect of allowing credit for all prior assessments in reducing total costs would be small or zero. For simplicity, this analysis assumes that all covered pipeline segments will be tested during the baseline period and re-tested at the appropriate intervals. The change in the rule allows operators maximum flexibility to determine when their expenditures must occur. Cost of Subsequent Testing The cost of subsequent testing will be reduced, since the permanent pig launchers and receivers and other modifications installed for baseline testing will be used and no additional pipeline modifications will be needed. Costs to perform testing may increase slightly due to growth of populated areas near the pipeline. (Costs may not grow as much as populated areas grow, since additional piping that may be determined to be in high consequence areas may have been included in the additional miles pigged or hydrotested during the baseline period and would therefore not add to the total miles to be tested). RSPNOPS has not considered this growth in this analysis. Total Cost of Testing The total cost of testing is thus as shown in Exhibit 7. "Terry D. Boss, letter to the docket dated July 30,2003, RSPA-2000-7666-342, p.6. 52#

Page 54, Exhibit 7. Total Cost of Testing (Annual Costs. in Millions. 2001 dollars) Year after Pigging Hydrostatic Direct Confirmatory DA Effective Testing Assessment Date Base Re Base Re Base Re 10-yr 15-yr 20-yr Pipe pipe pipe 1-7 232.1 0 4.2 1 0 25.75 0 0 0 0 ---------- - 8 - 10 232.1 17.12 4.21 3.0 25.75 7.21 6.41 2.06 Note 1 11 - 14 0 26.08 0 5.1 1 0 12.25 6.41 2.06 Note 1 15 - 17 0 26.08 0 5.1 1 0 19.71 6.41 0 Note 1 18 - 20 0 26.08 0 5.1 1 0 19.71 0 0 Note 1 reassessment at 7-year intervals rather than confimtory DA. Costs for the low-stress reassessment are considered along with preventive measures for that pipeline. Costs of Service IntermDtion The Interstate Natural Gas Association of America (INGAA) sponsored an analysis" evaluating the costs to consumers from service interruptions that they presumed would occur if testing requirements such as those included in the proposed rule were imposed. This analysis is available in the docket. It concluded that consumer costs would rise, perhaps significantly, due to incrqased costs of gas transmission if portions of the transmission pipeline network were required to be removed from service for testing. Both the Department of Energy (DOE) and the Office of Pipeline Safety have analyzed the INGAA report. RSPNOPS contracted with the Volpe National Transportation Systems Center to review INGAA's report. Both evaluations were performed considering the estimates in the preliminary draft regulatory evaluation prepared for discussion with TPSSC in July 2002. Both DOE and the Volpe Center concluded that the proposed rule would not have an impact on the domestic price of natural gas. RSPNOPS notes that the conclusions of DOE and the Volpe Center were both predicated, in part, on the ability of operators to conduct inspections during non-peak demand periods, and under the assumptions used in the preliminary draft analysis most testing should have been able to be performed during off-peak times. This is likely no longer the case. Changes that have been made in this analysis reduce the number of high consequence area miles estimated to require testing compared to the situation reviewed by DOE and Volpe. At the same time, the revised estimate of overtesting mileage results in more total miles being tested. This 5 1 " C ~ n ~ ~ e r Effects of the Anticipated Integrity Rule for High Consequence Areas," prepared for the INGAA Foundation, Inc. by Energy and Environmental Analysis, Inc., 2002 53 ., .#

Page 55. 8 greatly curtails, probably eliminating, the ability of operators to perform all testing during off-peak periods. Other changes in the final rule will result in less pipeline being removed from service for testing. The changes in reassessment requirements for low stress piping are particularly important in this regard. These changes are unlikely to offset, completely, the likelihood of supply effects from the considerable amount of high-pressure long-distance piping that will be assessed. As a result, RSPNOPS acknowledges that there likely will be some cost increase to consumers as a result of curtailment of the ability to transport gas during testing. RSPNOPS ddes not know how large this cost will be and has not included it in the cost estimates in this analysis. A summary of the issues raised in review of the I N G M E A analysis, and the positions taken by the various reviewers, is provided in the appendix. Consideration of Remote Control Valves (RCV) and Automatic Shutoff Valves (ASV) The final rule requires operators to conduct a risk analysis of their pipeline to identi@ additional actions to enhance public safety. Such actions include, but are not limited to, installing ASVs or RCVs, computerized monitoring and leak detection systems, extensive inspection and maintenance programs, etc. If an operator determines that an ASV or RCV is needed to protect high consequence areas in the event of a gas release, the operator must install the valve. Natural gas transmission lines are already required to be equipped with sectionalizing block valves.52 The required spacing of these valves varies for different class locations. Valves must be no more than 5 miles apart in class 4 areas, 8 miles apart in class 3, 15 miles apart in class 2, and 20 miles apart in class 1. Some of these valves are presently remotely operable. The requirement of this rule will have the primary effect of requiring operators to conduct risk assessments to determine if conversion of any of the now-manual valves to remote or automatic operation is needed. RSPNOPS completed a feasibility study on remotely controlled valves on interstate natural gas pipelines in 1999.53 In conjunction with that study, RSPNOPS conducted a public meeting and solicited written comments (see 62 Federal Register 51624, October 2, 1997). The study determined that conversion of valves to remote operation was not economically feasible. Most fatalities and injuries resulting from natural gas pipeline ruptures occur very quickly, before the time that would be required to isolate the pipeline using RCVs. Similarly, a significant amount of the property damage experienced in historical accidents occurred immediately after the rupture. The 1999 feasibility study determined that the value of gas lost before the pipeline could be isolated would be the principal benefit, and that the value of that benefit did not offset the costs of converting the valves. 5249 CFR 192.179, ‘“Transmission Line Valves” 53,‘ Remotely Controlled Valves on Interstate Natural Gas Pipelines (Feasibility Determination Mandated by the Accountable Pipeline Safety and Partnership Act of 1996)”, September 1999. 54#

Page 56Operators will need to re-visit this generic conclusion for particular pipeline segments that can affect high consequence areas. RSPNOPS expects that the conclusion that most fatalities and injuries could not be avoided by conversion of valves will not be changed, since historical accident experience shows that injuries and fatalities occur very quickly after any pipeline rupture. Circumstances specific to individual pipeline segments that can affect high consequence areas could, however, change the generic conclusion that significant property damage cannot be avoided through use of RCVs. This could lead to a need to convert some sectional valves to remote operation. The 1999 study reported on the results of a one-year field evaluation of RCVs conducted by Texas Eastern Transmission Corporation (TETCO) pursuant to a settlement agreement in the compliance case involving the 1994 pipeline rupture in Edison Township, NJ. The study reported that TETCO experience indicated that costs to install a RCV ranged from $150,000 to $500,000, depending on the number of valves at the same locations and variations in permitting costs. The study further estimated that the cost for converting an existing valve, on average, was between $125,000 and $150,000, including efficiencies that could be realized by dividing site costs over a number of valves in an individual location. The study concluded that there was no significant impact on direct operating costs, since the maintenance activities for the additional equipment were absorbed in the hnction of the personnel working valve sites for other purposes. RSPNOPS does not have any information on which to base a conclusion about how many natural gas transmission line sectional valves may be converted to RCVs. Because there are so many factors in an operator’s decision to install or not to install these valves, RSPNOPS has decided there would be no valid approach, including a sensitivity analysis, to make such an estimate. RSPNOPS has thus not estimated the industry costs to convert valves. RSPNOPS assumes that operators will not make such conversions unless the benefits (expected reduced property damage and value of lost gas) exceed the approximately $150,000 conversion cost, and thus there should be no net cost from valve conversions. Operators will be required to analyze their systems to determine whether it is cost-beneficial to convert valves. RSPNOPS assumes that these analyses will be conducted by staff engineers. The time required for these analyses is expected to be relatively small, since generic conclusions are already available and the effect of site-specific factors will be the focus of operator evaluations. RSPNOPS estimates that this will require approximately several man-months for pipeline operators with more than 30 miles of pipeline, at a cost of approximately $50,000. Operators with less than 30 miles of pipeline should be able to conduct these evaluations in about half the time at a cost of approximately $25,000. The total cost for considering the need to convert valves to ASV or RCV is thus $1 1,525,000. These costs will be incurred once and are assumed to be incurred in the first year after the rule becomes effective. The Costs of Data Integration 55#

Page 57As described above, integration of all information relevant to the integrity of the pipeline is a key element of the integrity management plans required for high consequence areas. Assuring this integration will require that operators’ internal data management systems be aligned and managed in such a way that relevant information is brought together. It will also require that the importance of this information be assessed by experienced pipeline safety professionals. RSPNOPS estimated these costs on a per-company basis in the draft regulatory analysis. That analysis projected that first year costs for operators with 40 or more miles of pipeline would be $100,000 and that continuing costs would be $50,000 annually thereafter. For operators with less than 40 miles of pipeline, the draft regulatory analysis estimated that first year costs would be $25,000, and that continuing costs would be $12,500. Industry comments suggested that the draft regulatory analysis had seriously underestimated the initial costs. INGAA estimated that operators would incur costs of $1,359 per mile to retrieve old data and set up the integration system. Annual costs thereafter, by INGAA’s estimate, would be $1 13 per mile.54 RSPNOPS agrees that initial costs to retrieve relevant data could be high. Much existing pipeline was installed many years ago. Records for installation and early years of operation are on paper and are not always readily accessible. Operators will need to access those records and to convert many of them to electronic form for future use. RSPNOPS also agrees that it is reasonable to estimate costs for these efforts on a per-mile basis. The records involved relate to the pipe, and it is reasonable to assume that there would be 10 times as many records for 1,000 miles of pipeline as for 100 miles. RSPNOPS has adopted the INGAA estimates for use in this analysis. The work of gathering records and integrating data applies to more than pipeline segments that are in high consequence areas. The rule requires that a risk analysis be performed that considers the entire pipeline. The rule also requires that lessons learned through assessment of pipeline segments that are in high consequence areas be applied, as appropriate, to other areas of the pipeline with similar characteristics and potential for similar problems. Operators will not be able to comply with these requirements unless they retrieve records and organize data for their entire pipelines. The per-mile costs for data integration must thus be applied to the entire length of transmission pipeline mileage, not just to pipeline in high consequence areas. The result of these changes is a significant increase in the estimated costs for data integration over those estimated in the draft regulatory analysis. RSPNOPS estimates that long-distance operators, with 225,000 miles of pipeline, will incur costs of $305,775,000 for setup (assumed incurred in the first year) and $25,425,000 annually thereafter. For LDC operators, with 60,000 miles of pipeline, the corresponding costs are $8 1,540,000 for the first year and $6,780,000 annually thereafter. The Cost of Preventive Measures for Low-Stress Pipeline 5 4 1 N G ~ spreadsheets submitted to the docket April 30,2003, RSPA-2000-7666-297. 56#

Page 58The final rule recognizes that low-stress pipeline (Le., that operating below 30% SMYS) is different from pipeline that sees higher stresses. Low-stress pipeline tends to fail by leakage rather than rupture. Accordingly, the rule includes different methods for interim assessment of low-stress pipelines, which focus on preventing corrosion failures that could result in leaks. These assessments are required for low-stress pipeline that is in high consequence areas. As described above, the amount of such pipeline is expected to be small, since the corresponding potential impact circles are small. Accordingly, the rule also requires preventive measures to be applied to all low-stress pipeline in class 3 and class 4 areas. These measures include required participation in one-call systems, use of qualified personnel for conducting and supervising excavations, improved public communication programs, and increased monitoring and patrolling. These measures are designed to reduce the likelihood that outside force, from third party excavations, will damage the pipeline. Third party damage is the most likely cause of catastrophic failures on low- stress pipelines. RSPNOPS has estimated the cost to perform the low-stress reassessment applicable to pipeline in high consequence areas as $1,000 per mile. These reassessments would be conducted on the 1,800 miles of low-stress pipeline estimated to be in high consequence areas at seven-year intervals. The annual cost to perform these reassessments is thus $257,143. RSPNOPS estimates that the cost to implement the preventive measures required for all low-stress pipeline in class 3 and class 4 areas is $150 per mile per year. Those measures are required on a continual basis, so these costs will be incurred every year. For purposes of this analysis, RSPNOPS has assumed that these costs apply to all 17,000 miles of LDC pipeline estimated to be in class 3 and 4 areas. The annual costs for these preventive measures is thus $2,550,000. The Cost of Remedial Action Inspection and testing and integration of other relevant data will identify anomalies that must be investigated and remediated. Real improvement in pipeline safety will not occur unless problems are addressed. The number of anomalies that will require action or the cost of that action can not be known until the assessments are performed. In addition, other provisions of Part 192 require operators to repair known anomalies that can impact pipeline integrity; the requirements of this rule merely serve to identify the anomalies. Costs associated with remediation are therefore not estimated as part of this analysis. TOTAL COSTS Costs have been estimated for: (1) identifying pipeline segments that can affect HCAs, (2) plans and reports, (3) evaluating valves for possible conversion to automatic closing or remote operation, (4)testing, (5) data integration, and (6) preventive measures for low-stress pipelines. Exhibit 8 presents a summary of these costs, presented by the year that they are incurred after the effective date of the final rule. 57#

Page 59EXHIBIT 8. THE ESTIMATED COST OF THE FINAL RULE (Costs in millions of 200 1 dollars) Costs to Operators will be Reduced by this Rule As described above, this rule does not impose requirements to assure pipeline integrity where otherwise no requirements would exist. The Pipeline Safety Improvement Act of 2002 (PSIA- 2002), enacted into law on December 17,2002, requires that pipeline operators implement integrity management programs and perform integrity assessments on their pipelines. These requirements forever changed the regulatory landscape. PSIA-2002 requires that operators use one of three methods to perform assessments, in-line inspection, pressure testing, or direct assessment. The Act requires that operators begin baseline assessments within 18 months (Le., by June 17,2004), and that they complete a baseline assessment of all of their pipeline segments that can affect high consequence areas within 10 years (i.e., by December 17,2002). The Act further requires that assessments be repeated on these segments at intervals no less frequent than once every seven years. The Act applies to pipelines that have been determined to be in high consequence areas pursuant to criteria defined in 49 CFR Part 192 pursuant to earlier legislative mandate. PSIA-2002 explicitly 5 8#

Page 60allows, however, for subsequent modification of these criteria.55 This final rule makes such modifications. The changes made in this rule refine the criteria to permit a better focus on those portions of the pipeline system where accidents can truly result in high consequences. As a result of these refinements, the number of pipeline miles determined to be in high consequence areas will be significantly less than it would be if this rule were not adopted. RSPNOPS has estimated, above, that the total number of high consequence area miles of transmission pipeline affected by this rule is 2 1,470 miles. RSPNOPS estimated the number of high consequence area miles in the draft regulatory analysis supporting the proposed rule (Le., before the refinements in this final rule were included) as 36,854 miles.56 If this rule is not adopted, pipeline operators would therefore be required to perform assessments on nearly twice as much pipeline mileage as they will need to assess to comply with this rule. PSIA-2002 also grants authority to the Secretary of Transportation to designate alternative assessment methods that would provide an equal or greater level of safety. This rule designates two alternative methods of assessment. The first alternative assessment method, confirmatory direct assessment, was introduced in the proposed rule. The final rule allows it to be used to perform assessments required to meet the statutory seven-year periodicity. The rule allows inspections using the other, more expensive, methods to be performed at intervals of ten, fifteen, or twenty years, depending on the operating stress levels in the pipeline. RSPNOPS has determined that use of assessments using the methods specified in PSIA-2002 at these longer intervals, combined with confirmatory direct assessment at seven-year intervals, provides an equivalent level of safety as the provisions in the statute. The second alternative assessment method allowed by this rule is low-pressure assessment. This is applicable to pipeline operating below 30% SMYS. It recognizes that the failure mode for pipe at these lower pressures is almost always leakage, rather than rupture. For such pipeline, increased leak surveys, more focused attention on the potential for internal corrosion, and electrical surveys to assure protection against external corrosion are a more effective assessment method. The rule allows this method to be used to meet the statutory seven year requirement. Assessments using the more expensive methods are required at 20-year intervals. If this rule is not adopted, operators would not be able to use these more appropriate, and less expensive, assessment methods. Instead, they would be required to implement the more costly methods specified in the statute at seven-year intervals. This would significantly increase the cost for conducting required assessments. RSPNOPS estimated the costs to operators of complying with the statute without the refinements included in this final rule. This estimate is described in the appendix to this regulatory analysis. ”Pipeline Safety Improvement Act of 2002, Section 14, “Risk Analysis and Integrity Management Programs for Gas Pipelines” 56Draft Final Regulatory Evaluation, Pipeline Integrity Management in High Consequence Areas, (Gas Transmission Pipelines), RSPA-2000-7666- 166, page 12. 59#

Page 61The total implementation costs, over 20 years, were then compared to the costs for implementing the final rule as estimated in this analysis. The costs for implementing the statute, without the changes made by this rule, are estimated to be $6.2 billion higher than the costs for implementing this rule. These savings are not, in the strict sense, “benefits” associated with this rule. They are, however, very real. The more costly requirements have already been imposed, by Act of Congress. The Act allowed flexibility for RSPNOPS to refine its scope, which is accomplished in this rule. These refinements to the legislative mandate will result in savings of $6.2 billion, over 20 years. CONCLUSIONS Issuance of this final rule will ensure that all operators will perform at least to a baseline safety level and will contribute to an overall higher level of safety nationwide. It will lead to greater uniformity in how risk is evaluated and addressed and will provide more clarity in discussion by government, industry and the public about safety issues and how they can be resolved. The flexibility of a performance-based approach provides several advantages. It encourages development and use of new technologies. It is an important feature in supporting operators’ development of more formal, structured risk evaluation programs and in RSPNOPS’s evaluation of them. It provides greater ability for operators to customize their long term maintenance programs. It has also stimulated the development of a supplemental industry standard, which is referenced in the rule. A performance-based approach will also encourage the development and maturing of risk-based approaches to integrity management. Our emphasis on the integrity management system encourages a balanced program, addressing the range of prevention and mitigation needs and avoiding reliance on any single tool or overemphasis on any single cause of failure. This orientation will lead to addressing the most significant risks, and is the best opportunity to improve industry performance and assure that the high consequence areas get the protection they need. It also addresses the interrelationships among failure causes and benefits. It promotes the coordination of risk control actions, beyond what a compliance-based approach would achieve. The rule provides for a verification process, which gives the regulator a better opportunity to influence the methods of assessment and the interpretation of results. This is not to say the regulator would overstep the bounds of oversight, but would provide a beneficial challenge to the adequacy of the operators’ decision process. This leads to greater accountability to the public. A particularly significant benefit is the quality of information that will be gathered as a result of this rule to aid operators’ decisions about providing additional protections. Two essential elements of the integrity management program are that an operator continually assess and evaluate the pipeline’s integrity and perform an analysis that integrates all available information about the pipeline’s integrity. The process of planning, assessment and evaluation will provide operators 60#

Page 62with better data on which to judge a pipeline’s condition and the location of potential problems that must be addressed. Integrating this data with the safety concerns associated with high consequence areas will help prompt operators and the Federal and state governments to focus time and resources on potential risks and consequences that require greater scrutiny and the need for more intensive preventive and mitigation measures. If baseline and periodic assessment data is not evaluated in the proper context, it is of little or no value. It is imperative that the information an operator gathers is assessed in a systematic way as part of the operator’s ongoing examination of all threats to the pipeline integrity. This rule is intended to accomplish that. The cost estimates in this evaluation reflect the estimated costs for operators to establish the necessary integrity plans and data integration processes, to integrate and analyze the data, to consider the need to convert some valves, to perform testing of piping in high consequence areas, and to implement additional protective measures for low-stress pipeline in class 3 and 4 locations. The evaluation reflects the fact that some operators have begun testing programs and would be expected to continue those programs without this rule. The cost for operators to identify pipeline segments that can affect high consequence areas is estimated to be $15.05 million. These costs will be incurred in the first year after the effective date of the rule. The integrated cost to all operators for developing integrity management plans is estimated to be $104.13 million. First-year costs for establishing required reports are estimated to total $1 1.08 million. Annual costs of $12.9 1 *million are projected to review the plans, make changes as needed, and to prepare routine reports. Data integration is a significant cost, but it is also one of the most important elements of the rule. First year costs to retrieve old data, prepare it for use with future integrity information, and to realign data management systems to facilitate integration are estimated to be $387.32 million. Retrieval of old data is a one-time cost. Costs for data integration reduce to $32.21 million per year after the first year. Testing is a key element of the rule. A portion of transmission pipeline in high consequence areas will be inspected each year, using one of three specified methods. (The rule would allow operators to use altemative methods, with adequate justification, but no additional methods are projected in this analysis). Modification of some pipeline will be required initially to permit the use of in-line inspection tools. These modifications add to the cost of performing the initial, or baseline, inspections. Baseline inspection costs across the industry are estimated to be $262.12 million annually for the ten years of the baseline period. Testing costs will increase in the eighth year, because operators will be required to begin re-assessing piping. Testing costs in years 8,9, and 10 will increase by $35.84 million as a result of the additional testing for reassessments. After ten years, all necessary modifications will be completed. Testing costs will then be reduced to 61#

Page 63between $50.9 and $57.3 million per year, depending on which pipeline is due for reassessment during each year. The total of all costs expected to be incurred over the 20 years following the effective date of this rule is $4.7 billion. The benefits expected to be realized as a result of implementing this rule are approximately the same its costs. As described in the evaluation, total benefits are estimated to be $4.699 billion. Of these, $800 million represents the safety benefit from expected reductions in deaths, serious injuries, and property damage from pipeline accidents. The remaining $3.9 billion in quantified benefits is more uncertain. There is little doubt that an accident of much greater consequences than any previously experienced could occur. There is also no doubt that implementing the requirements in this rule makes such an accident less likely. Quantifylng the value of this improvement is difficult, since it requires making assumptions about very unlikely events occurring concurrently. RSPNOPS is confident that the estimate in this analysis of $1.676 billion represents the order of magnitude of this benefit, although the precise value is highly uncertain. Similarly, the estimate that a $500 million benefit accrues due to avoiding the economic impacts of future accident-induced gas supply interruptions also has a high degree of uncertainty. It is clear from the experience in California in 2000, though, that economic impacts of this magnitude can result from gas transmission pipeline accidents, and RSPNOPS believes that this estimate, though uncertain, is reasonable. The estimate that a $71 9 million benefit accrues due to expected quicker return to full-pressure service following future corrective action orders is also somewhat uncertain, since it is based on the estimated impact of the Carlsbad accident. Orders require pipelines to operate at reduced pressure, and therefore reduced gas throughput. It is certain that this results in an economic consequence on the pipeline operator and on consumers served by the pipeline. Historically, many months have been required for operators to gather additional information about their pipelines and for RSPNOPS to gain sufficient confidence to allow a return to full-pressure operation. Shortening this period will thus result in benefits to operators and consumers. RSPNOPS considers $719 million a reasonable estimate of the amount of this benefit that will be realized over the next 20 years. The final benefit of $1 billion in reduced cost due to relief from class location requirements for pipe replacement is more certain. Granting relief from these requirements is within the authority of RSPNOPS. The agency has experience exercising that authority, through the Risk Management Demonstration Program. There is some uncertainty regarding when sufficient information will have been generated by integrity management program to grant future relief. This could reduce the benefit realized during the 20-year period of this analysis somewhat. It is virtually certain, however, that the entire benefit will be realized, and it is very likely that even more benefit will be seen after this 20-year period from further relief or changes to the class location requirements. 62#

Page 64As described in the analysis, there are also a number of qualitative benefits that will be realized as a result of implementing this rule. Foremost among these is providing a basis for improved public confidence in pipeline safety. Economic benefits are expected to accrue from this increased level of confidence, including reduced costs for siting and constructing new pipeline, thereby allowing access to the environmental benefits of increased use of natural gas in lieu of other fuels. Comments by members of the Technical Pipeline Safety Standards Committee (TPSSC), and members of the interested public who spoke at their July 18,2002, meeting, indicate their belief that the benefit of increased public confidence to be realized by implementing the rule, alone, justifies its costs. The estimated costs are certainly large, but need to be considered in context. Many of these costs will be reflected as increases in the cost of natural gas. During 2001, a total of 20.48 trillion cubic feet of natural gas was delivered to U.S. consumers.57 If the total projected first-year costs of $793.77 million are divided over this quantity of natural gas, the result would be an increase in delivered price of 3.87 cents per thousand cubic feet. The average U.S. residential customer used 79 thousand cubic feet of natural gas in 2001 .58 The effect of these costs would therefore be an increase in the monthly bill for this average residential consumer of 25 cents. After the first ten years, when all pipeline modifications are completed, the increase in average residential cost would be 3.2 cents per month. RSPNOPS believes that the process of developing an integrity framework and schedule and integrating data related to pipeline integrity is an important process for the operator, the government and the general public. The creation, and development of the plan will alert operators to the potential risks and consequences unique to high consequence areas. The planning process, including the testing schedule, also provides a level of confidence to Federal and state pipeline inspectors that pipeline operators are considering, examining, testing, and repairing if necessary, natural and other gas transmission pipelines that potentially pose severe consequences to public safety. Finally, standardizing the requirements nationally for transmission pipelines will save operators having to face potentially different testing and inspection requirements from the various state pipeline agencies. The OPS concludes that the benefits justify the costs associated with implementing the requirements in the final rule. Additionally, promulgating this rule will result in savings of approximately $6.2 billion over the expected costs to industry of complying with legislative requirements absent this rule. Publishing this final rule, and requiring that gas transmission pipeline operators comply, is clearly the appropriate course of action. Appendix 57NaturaI Gas Annual 2001, Energy Information Administration, page 44. 58 . Ibid, page 45. Determined by dividing the reported total volume of natural gas delivered to consumers by the reported number of consumers. 63#

Page 65Costs of Implementing Pipeline Safety Improvement Act of 2002 The Pipeline Safety Improvement Act of 2002 (PSIA-2002) requires that RSPNOPS issue a rule establishing requirements for integrity management plans for gas transmission pipelines. PSIA- 2002 also imposes requirements directly on pipeline operators. The Act requires that operators must develop integrity management plans and must conduct assessments even if RSPNOPS fails to publish a rule establishing requirements. If this rule is not published, therefore, operators would need to develop plans and conduct testing to comply with the law. PSLA-2002 establishes no detailed requirements for integrity management plans. RSPNOPS presumes that most operators would use the industry consensus standard, ASME/ANSI B3 l.SS, as the basis for plans they would develop. That standard is referenced in the final rule, and forms the basis for integrity management plans that the rule requires. Costs to operators would thus be similar for developing and managing their integrity management programs under the rule or under PSIA-2002 in the absence of a rule.59 Costs of testing would increase significantly, however. There are several changes made in this rule that have the effect of significantly reducing testing costs. These are: 1. Changes in the definition of high consequence areas. PSIA-2002 requires that assessments be performed on “each of the operator’s facilities in areas identified pursuant to subsection (a)(l) [of 49 U.S.C. 5 601091 and defined in chapter 192 of title 49, Code of Federal Regulations, including any subsequent modifications” (emphasis added).@ The referenced portion of the U.S. Code reflects previous legislative action directing RSPNOPS to define high consequence areas. The reference to 49 CFR 192 is thus to the definition of high consequence areas. This rule makes changes to that definition. The rule requirements therefore apply to the reduced amount of pipeline mileage expected to be determined to be in high consequence areas under the modified definition. Without these modifications, significantly more pipeline would be subject to assessment requirements. 2. Introduction of confirmatory direct assessment (CDA). PSIA-2002 requires that an assessment be performed every seven years. It specifies acceptable assessment methods as in-line inspection, pressure testing, direct assessment and “an alternative method that the Secretary [of Transportation] determines would provide an equal or greater level of ~afety”.~’ This rule establishes CDA as such an alternative method. If this rule were not adopted, all assessments would need to be performed using one of the other three listed 59The detailed requirements in the rule provide a basis for RSPNOPS and state pipeline safety regulators to oversee operator development of integrity management plans and to use their enforcement authority to assure that the plans are complete and are thoroughly implemented. In the absence of such oversight, some operators might not fully implement all provisions of ASME/ANSI B3 1 AS, resulting in lower costs. For purposes of this analysis, any such reduction in costs has been ignored. 60. Pipeline Safety Improvement Act of 2002, Section 14, “Risk Analysis and Integrity Management Programs for Gas Pipelines” 6’Ibid. 64#

Page 66methods. Those methods are all more expensive to use than CDA, and pose greater potential for interruption in gas service. 3. Low-pressure reassessment. Here, again, RSPNOPS is exercising its authority to determine that alternative assessment methods are acceptable. This rule provides that low- stress pipeline (Le., pipeline operating below 30% SMYS) be assessed every seven years using tests to assure the adequacy of corrosion protection. These tests are less expensive, and less intrusive than CDA, and considerably less expensive than the other specified assessment methods. If this rule is not published, all low-stress pipeline would be required to be assessed using the more expensive methods on seven-year intervals. RSPNOPS has estimated the costs that would be incurred if PSIA-2002 were implemented without the changes made in this rule. This analysis uses the high consequence area mileage estimated for the draft regulatory analysis, before the changes in high consequence area definition included in this rule. The analysis uses the updated unit costs (i.e., costs per mile or costs per company for various elements) adopted from industry comments for this final regulatory analysis, so that a direct comparison between costs estimated here and those estimated for implementing the rule would be valid. The costs to implement PSIA-2002, without the changes made in the final rule, are estimated in Exhibit 9. 65#

Page 67Exhibit 9. Pipeline Safety Improvement Act of 2002 Estimated Cos1 L! i to Implement (in million S of dollars) I Year lsegment I Integrity I Valve I Annual Baseline ID Plans Analysis Reports Testing 1 13.77 104.13 11.53 11.08 8 10.23 2 - 11.53 - 1.38 3 - 11.53 - 1.38 4 - 11.53 - 1.38 5 - 11.53 - 1.38 6 - 11.53 - 1.38 7 - 11.53 - 1.38 8 - 11.53 - 1.38 9 - 11.53 - 1.38 10 - 11.53 - 1.38 11 - 11.53 - 1.38 12 - 11.53 - 1.38 13 - 11.53 - 1.38 14 - 11.53 - 1.38 -_ 15 - 11.53 - 1.38 16 - 11.53 - 1.38 17 - 11.53 - 1.38 - 18 - 11.53 - 1.38 - 19 - 11.53 - 1.38 Subsequent Integration Total Testing - 387.32 1.338.04 - 11 1.69 32.21 156.80 1 11.69 32.2 1 156.80 1 32.21 I 156.801 11 1.69 32.2 1 156.80 20-yr total 10,939.12 The total 20-year cost of $10.94 billion compares to the corresponding total for implementing the final rule of $4.7 billion. The difference of $6.24 billion represents a savings to the pipeline industry from establishment of the requirements in the final rule. Updated Cost Estimates for the Proposed Rule As described in the analysis, RSPNOPS revised the unit costs used in this analysis. Industry comments indicated that the costs used in the draft regulatory analysis were significantly below the current costs for activities that would be required by this rule. Industry comments also indicated that a much greater amount of pipeline mileage would be “overtested”, Le., more miles outside of high consequence areas would be pigged in order to assess miles within high consequence areas. The cost estimates included in the draft regulatory analysis, reflecting the estimated costs for implementing the rule as originally proposed, are much lower than they would have been had the new unit costs been used in that analysis. RSPNOPS has re-estimated the costs to comply with the proposed rule, so that the effect of changes made in the final rule can be better understood. The draft regulatory analysis considered 66#

Page 68two scenarios, one in which a significant amount of pipeline would be modified to accommodate pigs (called the pipe modification scenario) and one in which only easy-to-modify piping would be modified (the limited modification scenario). Industry comments, both written and in the public meetings that occurred during the comment period, strongly indicated that hard-to-modify pipe would be changed and that pigging would be the predominant method used for conducting assessments. Based on those comments, RSPNOPS did not consider two scenarios in the final regulatory analysis. The costs were estimated based on the assumption that significant amounts of pipeline will be modified, similar to the pipe modification scenario in the draft regulatory analysis. Cost estimates have been updated for the pipe modification scenario of the draft regulatory analysis, which is most comparable to the pipe modification assumptions made in the final regulatory analysis. The revised cost estimates for the proposed rule (pipe modification scenario) are presented in Exhibit 10. Exhibit 10. Revised Cost Estimates for Rule as Proposed (68 FR 4278) (in millions of 2001 dollars) 20 - - - 1.38 - 89.11 32.2 1 122.69 ~ 20-yr total 10,3 17.40 The total estimated cost for implementing the rule as originally proposed is $10.32 billion, less than 10 percent below the cost estimated to implement PSIA-2002 if no rule is adopted. This is despite the fact that the proposed rule included use of confirmatory direct assessment (CDA) to 67#

Page 69meet the seven-year assessment requirement of PSIA-2002, significantly reducing subsequent assessment costs. The reason that costs for the proposed rule are not considerably less than costs to implement PSIA-2002 is that the proposed rule included shorter intervals for assessments to be conducted using direct assessment (DA). As proposed, baseline DA assessments were to be completed in seven years (vs. ten) and reassessments would have been required every five (vs. ten fifteen, or twenty). The increase in costs associated with more frequent DA assessments nearly offsets the cost reduction afforded by the inclusion of CDA. Changes made in the final rule thus reduced the estimated 20-year costs by $5.62 billion. The principal changes resulting in these savings are revising the assessment intervals for DA to be like those for other assessment methods and refining the definition of high consequence areas. This reduces significantly the amount of pipeline that will be determined to be in high consequence areas, and thus will require assessment. The change better focuses attention on those areas in which significant population could be at risk. The mileage deleted represents areas that meet the definition of class 3 locations but which are relatively more sparsely populated, in which 20 buildings intended for human occupancy are not within the potential impact circle associated with a postulated pipeline accident. The introduction of low-stress reassessment, better focusing reassessment requirements for low- stress pipelines on the failure mode expected for such pipelines, also contributes to the reduction in costs from the proposed rule. RSPNOPS concludes that the changes made in the final rule are appropriate, are responsive to public comments, and have significantly reduced the estimated costs for complying with the rule. Potential Supply Curtailments due to Assessment Requirements INGAA submitted to the docket an analysis performed by Energy and Environmental Analysis, Inc. that predicted the potential for significant supply interruptions resulting in increases in the price of natural gas to the consumer. The analysis was performed prior to publication of the proposed rule, and considered the likely effects of provisions that might have been included in the rule. The model is designed to look at changes in gas supply and the effect on monthly prices at market hubs. The analysis was reviewed for RSPNOPS by DOE and the Volpe National Transportation Systems Center. The major issues and positions of the various agencies and INGAA are summarized in the following table.62 62At a meeting with RSPA/RSPA/OPS on July 10,2003, FERC noted that recent years have seen high summer load factors, relatively fewer periods of low demand, and thus a higher potential for service interruptions. FERC noted that interruptions for planned maintenance, including assessments, should be less costly than unplanned interruptions due to accidents. 68#

Page 70I I Applicability of EEA Model Ability of natural gas transportation system to accommodate inspections Effect of redundant pipeline capacity, loop lines, etc. Improvements in efficiency and scheduling of testing Reviews of EEA Analysis of Potential Supply Interruption Issues and Positions of Reviewers DOE INGAA Volpe RSPAlOPS Model is integrated tool, best for predicting market prices Lack of capacity in some regions will lead to down time and potentially significant spikes in market price. Model accounts for redundant systems. They are not sufficient to accommodate all needs. Model is proprietary, not designed to examine impact of ongoing rule Given reasonable planning time, inspections can be scheduled to minimize down time, like routine maintenance. Existing storage and redundancies will prevent supply interruptions in most cases INGAA did not account for improvements in efficiency of scheduling and conducting inspections over time The effect of unplanned outages, e.g., accidents, could be far greater Inspections can be accommodated as part of routine maintenance Position is similar to DOE Concurs with DOE RSPNOPS cannot validate the model. Output is very dependent on assumptions. Point estimates of effect of rule are difficult. Agrees with DOE. Whether interruptions can be eliminated is unknown. The situation is without precedent. Sufficiency of redundancies is unknown. Single- supply lines could be a problem. RSPNOPS agrees that experience is likely to lead to greater efficiencies, but they are difficult to estimate without past experience to guide us. Concurs with DOE Effect of unplanned outages Impact on wellhead and production Testing in peak season Some testing will need to be done at peak periods. Dual- peak systems make it more difficult to schedule non-peak Testing done in off- peak seasons will have little effect on prices The rule will have no effect on wellhead prices or production capacity Concurs with Volpe Concurs with DOE Agrees off-peak scheduling will mitigate effect. Ability to complete required testing off- peak is unknown 69#

Page 71Significant changes were made to the rule following completion of the EEA analysis. These reduce the amount of pipeline mileage expected to be in high consequence areas. The distributed nature of these areas, however, may mean that the amount of long-distance transmission pipeline that will be pigged to comply with the rule has not changed significantly. In addition, elimination of conditions on the use of direct assessment makes that a more viable assessment option. The actual amount of testing that will be performed, and whether some expected pigging will shift to direct assessment, are currently unknown. Future additions to pipeline capacity will also affect the likelihood of supply interruptions, but the magnitude of that effect cannot now be estimated with any certainty. In conclusion, the situation is unprecedented. There is a possibility that interruptions in the supply of gas may occur. It is unlikely that gas will be cut off to consumers, but increases in gas price may occur as the market responds to curtailments on individual transmission pipelines. Transmission and distribution operators could face additional costs to provide alternative supply for some customers. RSPNOPS believes that the rule has been revised so that the likelihood of such interruptions has been reduced as much as practicable. This likelihood has probably not been eliminated. RSPNOPS is unable to estimate the costs that may result from supply interruptions. 70#