Updated: Jun 2, 2020
How can you improve pipeline safety using machine learning and AI?
Pipelines are the key to transporting Alberta's oil and gas to the right market. There are more than 400,000KM of pipelines across the province. In 2018, these pipelines transported more than 1M barrels of crude oil and 5T ft3 of natural gas to the market.
Pipelines are a safe and reliable way to transport oil and gas products. However, most of the pipeline infrastructures have been around for many years and running them safely is one of the ongoing challenges that pipeline operators are facing. Regulatory bodies have established strict regulations and standards to ensure pipelines are designed, constructed, operated, maintained, and abandoned safely for the public and environment. Pipeline companies comply with these regulations in all aspects of their operations by establishing programs to identify threats and mitigate risks, taking necessary actions in case of pipeline failures and emergencies, and taking responsibility for the consequences. Every year, pipeline operators run smart pigs in thousands of kilometers of their pipelines, perform hundreds of integrity digs and repairs, and spend billions of dollars on maintenance, monitoring, and integrity activities.
Despite the strict regulations and standards, investment, and planning for integrity activities, pipelines still fail and cause low to high consequence incidents.
Traditional integrity decision making is not sufficient to prevent incidents from happening.
The pipeline industry requires innovative and effective decision making methods that are supported with data and analytics, are proactive, fast, and repeatable. This is especially true considering the tremendous amount of data that is generated every year from inline inspections and all the uncertainties that naturally come with ILI measurements, aging and old pipeline infrastructures, an environment that is changing constantly and adds to the complexity, and more strict rules and regulations, such as the $1Billion liability bill.
Integra leverages its advanced analytics methodologies, subject matter expertise and its powerful data science platform, Digital Hub™, to develop solution accelerators for better decision making in the pipeline industry.
We recognize the importance of safer pipelines for our province, industry demand for innovative solutions, and incidents that still happen. Motivated by that, Integra has identified improvement areas in the pipeline industry and developed AI/ML solutions that leverage existing data at low cost and generate new insight.
Pipeline Failure Simulation: Simulating millions of failure scenarios for each defect detected by ILI across the entire pipeline in a short time and low cost
Pipeline Remaining Life Prediction: Forecast the years to failure for each defect based on growth rate models and recommend digs and repairs
Inline Inspection Tool Performance: Best fit ILI tool for different pipelines based on their defect susceptibility, defect history, pipeline characteristics and ILI tool probability of detection and identification.
Dig Planning Automation: Automate generating dig plans based on ILI data, POE, engineering assessments and compliance criteria.
Long Range Planning: Generate long range plans based on planned integrity activities and reduce manual and repetitive work