A Zero-Accident Strategy for Oil Pipelines
The complete paper provides tools useful for decision-makers to prevent adverse pipeline incidents caused by human action, poor facility performance, accidents, emergencies, and external events.
The authors undertook a research program to evaluate the safety performance of product pipelines in Mexico as part of a zero-incident initiative.
Although pipeline systems exist worldwide, in most jurisdictions, only limited public information on pipeline performance is available. In Mexico, information regarding pipeline design, construction, and operational conditions, as well as detailed data on past accidents, is not available to the general public. Therefore, the research team was obliged to develop innovative means of developing a reliable view of pipeline performance. Data were inferred from publicly available sources with the aid of data-management, data-processing, and analytical tools. For instance, pipeline rights of way were mapped from satellite images and then converted to a collection of georeferenced vector points arranged and ordered in a path. The geometry object created is called a line-string vector—that is, a curve with linear interpolation between vector points, each consecutive pair of points defining a line segment. With the aid of spatial-analysis methods and geographic information systems (GIS), it is possible to extract data from diverse open information sources related to each vector point. The distance between two consecutive vector points can be set depending on the nature of the analysis to be performed. Undertaking such a data-computation process at every vector point and intersecting it with the georeferenced data sets results in a large amount of data that can be stored in a spatial-relational database.
This database is organized into four main categorical layers:
- Pipeline infrastructure systems and interconnections
- Transported product fluxes and operational parameters
- Historical accident data
- Contextual aspects related directly or indirectly to pipeline operation and performance
The database structure, when disaggregated at a vector-point level, provides the required flexibility to plot data against vector points arranged along the pipeline trajectory, resulting in a profile graph of the pipeline system.
To communicate large multidimensional information to decision-makers effectively, a visually rich and interactive front-end interface is required. The project established an analytical platform based on high-quality, current, readily available data. The database communicates with risk-evaluation models and pipeline-integrity programs to provide the necessary data to perform analyses.
Risk is the product of the likelihood of the occurrence of an event and the associated consequences. Cross-country oil and product pipelines are exposed to several threats, each of which could result in an accident.
The first part of the risk equation, likelihood, is often based on the historical record of the pipeline and pipelines of similar characteristics. Accident information and analysis is critical to understand how pipelines have failed and why. This knowledge then is applied to the knowledge of the pipeline segment under assessment, and areas of similarity or contrast are noted. Data integrity is key to ensure that all available knowledge is applied to the assessment effectively.
The second part of the equation is consequence. Consequences are much more difficult to predict and usually require the analyst to draw upon information outside the realm of the pipeline information and records systems. Critical locations where any failure could have severe consequences on people, the environment, property, or the economy must be identified. Current data on topography, population, land use, environmental sensitivity, and economic activity, including navigation and transportation, should be compiled. In addition, the potential for associated or secondary effects of an incident should be considered in the evaluation.
Consequence evaluation lends itself to a GIS-based system, such as the one developed for this project, that is regularly updated from a variety of external sources. The source of data may be affected by the threat being considered; for example, crime rates relate to fuel theft, and seismic activity indicates outside forces. Regular updating of the data sets is essential if the risk-evaluation process is to be valid.
Monitoring Systems and Current Industry Standards
Monitoring pipelines with leak-detection methods is performed in several ways, and usually is categorized as either noncontinuous or continuous. Noncontinuous methods include inspection by aircraft, personnel, or trained dogs, and periodic use of so-called smart pigs or in-line electronic-inspection devices. Continuous monitoring includes either external or internal systems. External systems include such approaches as fiber-optic cable, acoustic systems, and video monitoring that examines conditions on the exterior of the pipeline. Internal systems include pressure-point analysis, the mass-balance method, statistical systems, real-time-transient-based systems, and extended real-time-transient-based systems that monitor temperature, pressure, and flow inside the pipeline and identify irregularities that might indicate an event.
Most pipeline companies use supervisory control and data acquisition (SCADA) systems that consist of remote terminal units connected to flowmeters and a communication unit that sends pressure and temperature information to the control center, where operators monitor the pipeline network.
Modern SCADA systems feature new software based on artificial intelligence, statistical methods, and computation, which facilitates prediction, observation of the internal state of the pipelines, and diagnosis of abnormal events. Software and the speed of the media and measurement devices make possible determination of the status of the pipeline network in real time, monitoring failures and damage in a continuous and automated way.
On the basis of literature reviews and direct knowledge of report authors, among the variety of pipeline-detection techniques available, mass-balance and negative-wave technology are the most-used in the industry because of ease of installation and low installation cost.
Fig. 1 shows the activities that follow an accident. Once an accident occurs, it must be reported immediately to the relevant regulators and appropriate public emergency and security authorities. Emergency-response protocols have been developed in different jurisdictions around the world. The complete paper presents the results of the reviews of emergency-response-protocol requirements of countries that transport hydrocarbons in pipelines. The requirements for emergency-response protocols generally are focused on safety and human-health-and-environmental protection.
In all countries examined, a government body oversees both the establishment and implementation of emergency responses. Most of the countries studied have a long history of producing and transporting hydrocarbons. However, most of these countries examined have either initiated or revised the legislation governing emergency-response protocols during the past 2 decades.
Most countries hold companies accountable and, typically, liable for damages resulting from pipelines losses. Operators are required to have processes in place to identify an event and respond immediately. The exception is Nigeria, which relies on voluntary engagement and support of companies.
In all cases, significant coordination is expected among the various levels of authorities and industry in emergency response. A clear recognition exists that effective emergency response requires effective planning and coordination among all levels of government, industry partners, and community resources. In both Norway and the UK, operators are required to develop emergency-response protocols in cooperation with other operators and in coordination with national and regional authorities and local agencies and services.
While the initial response is expected to be from the operator, other agencies and authorities are prepared to participate quickly and draw in external resources. In many cases, this coordination is part of the risk-management process, with external parties participating along with the operator in planning and training and conducting mock exercises. This is typically established within the regulatory framework that sets out the process for developing and implementing emergency-response protocols.
Nontechnical Tools To Address Integrity Challenges
In some countries, the use of the monitoring technologies discussed are not enough to address integrity challenges such as hydrocarbon theft. The tools themselves are distinguished by the fact that they depend on political will and public support. Consequently, the most-important overarching factor is improvement of transparency and administrative cooperation and action to manage this sector of the economy.
Digitizing commercial processes is a common practice in the energy industry, which typically produces both paper- and computer-based records. This process would allow a common computer-data-recording tool known as check-sum to validate and identify areas of hydrocarbon losses.
A second tool known as blockchain record-keeping has been pioneered by digital-currency brokers to provide reliability in that market. A blockchain is a data structure representing a single financial ledger entry or record of a digital transaction. The transaction is signed electronically or assigned a unique identity that cannot be altered, providing integrity of record-keeping and a certain degree of inaccessibility.
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 199537, “A Zero-Accident Strategy for Oil Pipelines: Enhancing HSE Performance,” by Ricardo G. Suarez Suarez, Fabian Carranza Dumon, and Luis Serra Barragan, Tecnologico de Monterrey, et al., prepared for the 2020 SPE International Conference and Exhibition on Health, Safety, Environment, and Sustainability, originally scheduled to be held in Bogota, Colombia, 28–30 July. The paper has not been peer reviewed.