Improving the Safety of Marine Pipelines (1994)

A reliable, safe, environmentally responsible pipeline industry is vital to the nation. First, and most immediately, a significant fraction of the nation's energy supply depends on uninterrupted operation. In the longer run, the offshore oil and gas industry in general could be dramatically affected by major pollution incidents or fatal accidents resulting from pipeline failures. The safety record of marine pipelines is good, but it should be improved, since society's willingness to accept risk seems to be decreasing. Transmission and production pipelines account for about 98 percent of all the oil spilled by outer continental shelf (OCS) oil and gas operations (not including transportation by tanker or barge).

This committee's assessment leads to the following conclusions and recommendations.

Pipeline failures and spills are reported to several different agencies, with different reporting formats and information requirements. The available data on failures of off-shore pipelines are correspondingly incomplete. Most important, data for state waters is unavailable; even on the OCS, the lack of consistent reporting standards prevents detailed statistical analysis. It is important to improve the process of information gathering, to put risk management priorities on a sound and cost-effective basis.

The various regulatory agencies involved should develop a common safety database, covering both state and federal waters, and should periodically review theirdata requirements. The focus should be on collecting, archiving, analyzing, andreporting safety data with the intent of improving design and operating regulations. The extended data base should include the information needed for risk andcost-benefit analyses. MMS, which has the greatest experience and resources indata gathering, should coordinate this effort.

Despite the incompleteness of the safety-related data, several important patterns can be discerned in the data base for the OCS:

• Corrosion, although it was the reported cause of nearly half of the 1,047 OCS pipeline failures recorded between 1967 and 1990, produced only about 2 percent of the pollution from pipelines.

• Damage from vessels (and especially from anchors and groundings) is dramatically more significant than corrosion as a source of pollution and other consequences, including deaths and injuries. Anchor damage alone accounted for 90 percent of the pipeline-related pollution on the Gulf OCS.

• A very few incidents have resulted in the majority of consequences. The 4 largest pipeline spills, all caused by anchor damage, accounted for 85 percent of the pollution from pipelines for both production and transmission, on the Gulf OCS between 1967 and 1990; the largest 11, all but one caused by vessels, produced 98 percent.

• Deaths and injuries are rare. Six incidents (two vessel groundings and four repair accidents) resulted in all of the deaths and injuries associated with pipeline failures.

Even in the absence of better safety data, it is possible to improve safety planning. Modern risk analysis methods, using incomplete data supported by inferences and expert opinion about the nature and distribution of risks, can clarify priorities for risk management. For example, the risks to human safety and to the environment due to failures of marine pipelines are not uniform across the Gulf of Mexico. Resources being limited, a risk analysis model that compares various risks by geographic zone would allow cost-effective risk management decisions that address safety everywhere and provide the basis for strengthening regulations in high-risk areas. The goal is a consistent risk management strategy that unites all of the regulatory agencies and pipeline operators in developing criteria for the reduction of human and environmental risks.

The committee recommends that safety regulations be based on sound risk analyses and cost-benefit analyses. Specifically, regulatory agencies should agree on aconsistent risk management strategy to set priorities about human safety criteria,and about the use of cost-benefit analysis for the reduction of property and environmental damage. A zone-based risk analysis model, based on the zonation approach outlined in Chapter 3 of this report, should be developed on the basis ofcurrently available information and regularly updated, to help determine whetherregulations should be revised, strengthened, or relaxed, and assist in establishingpriorities for the operational use of resources by both government and industryfor enhancing pipeline safety (such as inspection coverage and frequency, use ofinternal inspection devices, and establishment of burial depths for areas havinghigh erosion rates).

Enforcement of safety regulations, like information collection, reflects a lack of consistency among state and federal agencies. The enforcement programs of the OPS and MMS are radically different in approach and in scope. OPS inspection efforts are conducted primarily through periodic audits of company records. MMS inspection efforts consist of the periodic inspection of pipeline maintenance and safety systems (during annual inspection of offshore facilities) and spot inspections of construction and repair activities. There is also a dramatic difference in enforcement personnel; MMS assigns 70 inspectors to the Gulf of Mexico OCS region, while OPS assigns 2 inspectors. To some extent, these differences reflect differences in the safety issues faced by the two agencies, but it appears likely that OPS enforcement personnel are too few to adequately cover the

nearly 13,000 miles of pipeline and more than 160 operating companies in the region that are under OPS jurisdiction.

To make better use of inspection resources and help integrate the enforcement ofMMS and OPS marine pipeline safety regulations, the committee recommends that the enforcement of OPS regulations offshore be performed by the MMS,through an interagency agreement or redefinition of the memorandum of understanding that defines the jurisdictional division between OPS and MMS. Such asystem would continue OPS's role in regulating offshore pipelines, while strengthening the application and enforcement of such regulations by bringing to bear MMS'greater resources.

Another regulatory discrepancy is visible in the MMS and OPS requirements for internal inspection of pipelines. MMS, under a law requiring the use of the “best available and safest technology,” has established a general requirement for the use of in-line inspection devices (generally known as smart pigs) where practicable. OPS is studying the matter, under congressional mandate. The committee finds that the technology of smart pigs is progressing, and that these devices are seeing increasing use onshore. However, the vast majority of marine pipelines cannot physically accommodate smart pigs, and modification of pipelines generally would be uneconomic. In addition, the current devices are relatively inaccurate in locating flaws. Because the costs of verifying suspected flaws are much greater offshore than onshore, this inaccuracy is a greater handicap. The use of smart pigs offshore will not be widely practical until further technical improvements are made, especially in the reliability and accuracy of three-dimensional measurement of flaws and in the compactness and maneuverability of smart pigs themselves.

The committee recommends that marine pipelines already constructed be exemptedfrom federal or state requirements for the use of currently available smart pigsfor external or internal corrosion control. New medium- to large-diameter pipelines running from platform to platform or platform to shore should be designedto accommodate smart pigs whenever reasonably practical.

Pipeline operators and regulators should continue to assess developments in smartpigging technology and seek cost-effective opportunities for its use.

Detecting and limiting leaks quickly is nearly as important as preventing them in the first place. A variety of techniques is available. Periodic aerial surveillance can detect leaks of all sizes, but sometimes with a delay of days to weeks. Setpoint-limit control systems (which monitor changes in pressure or flow rates) can detect large leaks, but are not effective for pipeline systems with routinely varying pressures and flow rates. For liquid pipelines, manual or automated line-balance calculations (comparing volumes in with volumes out) can detect leaks of varied sizes; manual calculations are generally made only once per day, while automated calculations may be made more frequently (with substantial additional costs for the necessary monitoring and communications equipment).

Many leaks are first detected through visual sightings by parties other than the pipeline operators. The detector of a leak generally cannot identify the operator of the pipeline. Nor is there an agency or entity that can establish the responsible party in a timely

fashion. The responsible operator in turn, once made aware of a leak, can have difficulty contacting in a timely manner all connecting pipeline and platform operators who must take action.

Pipeline operators should use a combination of leak detection methods to ensuretimely detection of a broad range of leaks. Setpoint-limit control systems, wherepractical, should be used to provide quick detection of relatively large leaks.Line-balance calculations—either manual or SCADA-based—should be conducted at least daily, where practical, to monitor pipeline systems for small- to medium-sized leaks (which can be detected in this way with a time delay of 1 to 24 hours).Periodic visual surveillance (with a time delay of 1 hour to 2 weeks) should beused to detect very small leaks and those that have gone undetected by othermeans. The method chosen will depend partly on the product transported, thethroughput of the pipeline system, the potential consequences of leaks in particular locations, and the nature of the pipeline system's operations (such as itsrelative stability of operating conditions and its location and accessibility bypersonnel).

MMS should coordinate an effort by appropriate federal and state regulatoryagencies and industry to establish a system through which leaks detected by thirdparties can be reported to a single agency or notification center with continuouscoverage around the clock. This one central location should have a comprehensive data base permitting easy identification of the operator of any marine transmission or production line based on the reported sighting location. All maritimeentities should be encouraged to use this single reporting center. Pipeline operators, in turn, should have 24-hour telephone numbers or a means of immediatelycontacting all other pipeline and platform operators who must take action.

No sensor technology is available to permit moving vessels to detect nearby pipelines at a distance, and thereby avoid them. Location-determining technologies are too inaccurate. However, there are operational measures by which vessels can lessen the risks of inadvertently interfering with pipelines.

An obvious but difficult problem is the control of the mooring of supply and service vessels in areas adjacent to offshore platform installations. A specific risk is that these vessels may drop anchors on nearby pipelines or flowlines, or interfere with pipeline risers. Clear communications between vessels and offshore platform operators would help avoid these risks.

In areas where supply and service vessels operate adjacent to fixed platforminstallations associated with high densities of pipelines or flowlines, permanentmooring systems should be considered. In other circumstances, platform operators should be required to provide detailed and timely information to vessel operators on the configurations of local pipelines or flowlines, so that the vesselscan anchor in designated areas. To lessen the risks of damage further in thosecongested areas, new pipelines should be installed whenever practical in well-defined “corridors.”

In shallow waters (generally less than 200 feet deep), the best protection against the

interference of vessels and pipelines is burial of the pipelines, with enough weight coating to keep them in place. In the shifting, often unconsolidated coastal sediments and eroding shorelines of the northern Gulf of Mexico, however, achieving and maintaining adequate burial requires care and vigilance. Pipeline installation must take into account detailed knowledge of soils, currents, and shoreline processes, so the pipeline can be buried and weighted to keep it in place, even if its surrounding soils are fluidized by wave action.

The committee has no information leading it to believe that the initial burial depths required by regulatory agencies are either adequate or inadequate. Anecdotal evidence suggests that initial cover may be adequate, but loss of cover over time, through erosion or fluidization of surrounding soils, exposes pipelines to interference by vessels. Pending further study, the current regulatory standards for initial depth of burial must be considered adequate.

Much of the Gulf shoreline is eroding rapidly. This erosion may expose pipelines buried at installation, and can be accelerated by the trenching used to install pipelines across the shoreline. The directional bore method of installing pipelines under beaches without breaking the surface eliminates this problem, and is also attractive from the standpoint of construction and maintenance costs.

The need for periodic inspections of pipelines, to ensure that they do not lose cover or become exposed, is not addressed in standard industry practice or in regulations.

Geotechnical studies of soil conditions, with sampling at intervals determined bylocal site conditions, should be required as a condition of marine pipeline construction permits. Soil core samples should be analyzed and interpreted for design parameters relative to weight, specific gravity, grain size, shear strength,and potential for liquefaction and fluidization. Permitting and regulatory agencies should work with industry to develop criteria for specific gravities of marinepipelines in varying soil environments.

To provide baseline data for subsequent depth of cover and bottom status surveys,newly installed pipelines should be surveyed at once and their depths of coverrecorded, with reference to Global Positioning System locations. Maintenance ofthis baseline data should be required by the agencies issuing the constructionpermits.

All agencies involved in the permitting of pipelines crossing shorelines shouldrequire the use of the directional bore installation method wherever feasible.

In waters less than 15 feet deep (where interactions between vessels and pipelinesmay, albeit rarely, expose vessels and crews to fire and explosion), periodicdepth-of-cover surveys in the Gulf of Mexico should be scheduled according tothe specific local shoreline and seabed dynamics, and the passage of severestorms, according to the criteria outlined in Chapter 5 (“Periodic Depth-of-CoverInspections”). In brief, a baseline depth of cover measurement should be estab lished for each pipeline, and subsequent inspections should be made —at intervals determined by local shoreline and seabed dynamics and storms—to determine the direction and rate of change of the depth of cover. Later inspection intervals can

be lengthened or shortened according to this rate; this approach might be called“self-adjusting.”

Pipeline operators and regulatory and permitting agencies should conduct studies to determine the appropriate standards for initial depth of burial under various shoreline and seabed conditions, using the results of the recommended periodic depth-of-cover surveys.

Abandonment of marine pipelines will continue to increase as producing fields reach maturity and are shut-in. Most of these abandoned lines are in shallower state waters. A properly abandoned pipeline poses no risk to public safety or to the environment. Abandoned pipelines have not been reported to cause any loss of life or significant property or environmental damage. The current practice of remediating abandoned pipelines once they come to the attention of the operator is adequate. A more aggressive periodic inspection program is not warranted until, and unless, public safety or the environment is shown to be adversely affected.

Pipeline abandonment standards should include a requirement for a one-timeinspection at the time of abandonment to verify that abandonment requirementswere met. Removal, continuing surveillance, or periodic inspection of abandonedpipelines should be required only where unique public safety or environmental conditions exist, such as rapid coastal erosion in areas of high vessel traffic.Pipeline operators should take timely corrective action when they are made awareof problems caused by their abandoned pipelines. Remediation should be theresponsibility of the owner or successors until or unless the abandoned pipelineis removed.