“If you can’t measure it, you can’t improve it.” (Peter Drucker) is a maxim that applies to airline accidents, gas and oil pipeline explosions and leaks and underground utility damage resulting from excavation. Whereas reliable statistics have been available for the civil aviation industry since the early 20th Century, for gas and hazardous liquids pipelines reliable statistics have only been available since the beginning of the 21st Century.
Reliable metrics provide a foundation for assessing the social and economic impact of incidents and the effectiveness of new technologies and policies in preventing and reducing the severity of these incidents. The commercial airline industry is a model of what reliable statistics and data for incidents and policies and technologies designed to reduce these incidents can achieve. In the civil aviation industry there have been reliable statistics since the early 20th Century. The Air Commerce Act of 1926 created an Aeronautic Branch of the United States Department of Commerce. In 1929 there were 51 commercial airline incidents, which represents an accident rate of about 1 for every 1,000,000 passenger miles. If that rate is prorated to the current total number of revenue passenger miles this would be equivalent to 7,000 fatal incidents per year. Fatal accidents per million flights in 2018 have decreased 16 fold since 1970, from 6.35 to 0.39, and fatalities per trillion revenue passenger kilometers (RPK) decreased 54 fold from 3,218 to 59 in 2018.
Since 2004 the total miles of pipeline in the US including crude oil lines, petroleum product lines, gas distribution mains, gas transmission pipelines, and gas gathering lines increased by about 12 % to 1,830,672 miles in 2017. The Pipeline and Hazardous Materials Safety Administration (PHMSA) was created in 2004 by federal legislation. Pipeline operators are required to submit performance measure reports for pipeline infrastructure covered by integrity management (IM) programs. This includes gas distribution, gas transmission, and hazardous liquids. Operators are required to report all pipeline incidents including the amount of location, volume of product released, number of fatalities and injuries, costs and cause. Since 2005, pipeline operators have reported excavation damage as the cause of 1052 incidents, resulting in 48 fatalities, 195 injuries requiring hospitalization, and $ 481,736,551 of property damage.
Possible causes include
- CORROSION - caused by galvanic, atmospheric, stray current, microbiological, or other corrosive action.
- EXCAVATION DAMAGE - incidents resulting directly from excavation damage by operator's personnel, by the operator's contractor, or by people or contractors not associated with the operator
- INCORRECT OPERATION - incidents caused by operating, maintenance, repair, or other errors by facility personnel
- MATERIAL/WELD/EQUIP FAILURE - incidents resulting from poor welds, poor construction, and stresses such as vibration, fatigue, and environmental cracking.
- NATURAL FORCE DAMAGE - incidents resulting from earth movement, lightning, floods and washouts, frost heave, and other natural causes.
- OTHER OUTSIDE FORCE DAMAGE - incidents caused by non-excavation-related outside forces including vandalism and terrorism.
- ALL OTHER CAUSES - incidents whose cause is currently unknown
Since pipeline incident data began to be collected the proportion of incidents attributed to excavation has increased. For example in 2018 34.4 % of gas distribution leaks (unintentional escapes of gas from the pipeline not reportable as Incidents) were caused by incorrect operation. In 2010 only 16.4 % were attributed to incorrect operation. Clearly improved training and other initiatives have reduced equipment operation failures but also being able to better identify causes has made it possible to reduce other types.
Serious Incidents include a fatality or injury requiring in-patient hospitalization. PHMSA statistics suggest there may have been some improvement in number of serious incidents and in fatalities, but prorating these statistics to pipeline mileage suggests otherwise.
On June 10, 1999, a gasoline pipeline operated by Olympic Pipeline Company exploded in Bellingham, Washington. Three people died in the accident. On August 19, 2000 a natural gas pipeline owned by the El Paso Corporation exploded near Carlsbad, New Mexico killing 12 people. After these incidents the government moved quickly to improve the quality of geospatial and other data about underground pipeline assets. Operators were required to identify high consequence areas (HCAs) where there were 20 or more structures intended for human occupancy within a radius (potential impact radius or PIR) defined by the diameter and pressure of the pipe.
One of the reported statistics is the number of significant incidents per 10,000 HCA miles. Significant incidents are defined as incidents involving a fatality or injury requiring hospitalization, $50,000 or more in total costs measured in 1984 dollars, substantial liquid releases, or fires and explosions. The data reveals no trend in improvement in this statistic.
By analogy with the civil aviation industry, reliable statistics resulting from government mandated reporting of pipeline incidents provides the foundation for the next phase of implementing safety management plans designed to produce continuous improvement in these statistics. Safety management systems tailored to specific industries and requirements have been developed often involving regulation for civil aviation, international maritime shipping, and the rail industry in Canada. At the GITA Pipeline Forum in Houston, there was a lot of discussion about pipeline safety management systems (PSMS), specifically the RP 1173 framework developed by the American Petroleum Institute (API) in partnership with PHMSA, state pipeline regulators, and other interested stakeholders. RP 1173 provides best practices for pipeline operators including operational controls, risk management, incident investigation, evaluation, and lessons learned, and safety assurance. This is intended for operators of hazardous liquids and gas pipelines under the jurisdiction of the US Department of Transportation. It provides for the comprehensive and systematic management of safety-related activities for achieving its goal of zero incidents per year.
Tracy Thorleifson of G2 Integrated Solutions suggested that at this stage we should be really worrying about asset integrity risk assessment. Pipelines are distinguished from airplanes in being in intimate contact with their environment which means many more variables to consider. Pipeline operators routinely capture and store operational information about their networks including pressure, flow rates, valve status and so on, but they don't capture and store information about the environment in which the pipeline is operating, such as, earth movement including subsidence, soil types, landuse, vegetation, proximity to water bodies, etc. Asset integrity assessment should include real-time dynamic segmentation using the thousands of data layers for the external environment available from NOAA, NASA, ESRI Living Atlas and other sources. This is an area where machine learning could fruitfully be applied to calculate risks of external events affecting the integrity of pipelines. For a pipeline in an area prone to flooding, subsidence, land slide, or river bank collapse, using dynamic segmentation would permit using these factors to adjust the risk assessment algorithms. Furthermore risk models should include data quality, including location accuracy, as a potential threat factor. For example, not knowing accurately the proximity of hazardous underground utilities can seriously compromise risk assessments.
In summary since 2005 the gas and oil pipeline industry and its regulator have achieved reliable statistics for pipeline incidents. The focus is now on regulation, policies and technologies to improve these statistics to reduce the number and severity of pipeline incidents.