Improving Oil Pipeline Leak Detection with Aerial Data

If you work in the field of pipeline integrity management, you’re likely finding yourself playing an ever-more difficult game. With aging infrastructure and ever-increasing regulatory burdens, you may be in search of a way to better monitor and mitigate issues along your pipeline and right-of-way (ROW).

Platforms for aerial data can provide a new layer of visualization, that gives you a better view of your asset and helps connect the dots on incomplete data in your workflow. These platforms leverage technologies such as Artificial Intelligence (AI) and Unmanned Aerial Vehicles (UAV) to automate visual sourcing and analysis of your pipeline – delivering the actionable reports via GIS integration or easily-accessible portals.

Diagnose Alerts from Your SCADA System

SCADA systems provide an important toolset for detecting leaks, but the portrait they paint is incomplete. While the sensors that form a SCADA network capture key metrics – such as pressure, flow rate, temperature, and more – the layout of these sensors make localization of potential leaks a challenge. Sensors usually reside on gauges between vast segments of pipeline. When an alarm goes off, an investigation into its source can take days.

An aerial data platform can fill this gap, providing the means to quickly verify leaks and visualize your pipeline SCADA data – all within a matter of hours.

When integrated with a SCADA system, the platform can quickly initiate an automated inspection process, beginning with UAV deployment to the affected segment. During flight, the aerial system captures high-resolution images of the area – be it RGB or infrared. Upon the UAV’s return, the files transmit for analysis via AI-driven computer vision software, which looks for potential spill indicators. All of this data, such as time-stamped and geo-referenced images of anomalies, ports seamlessly into your GIS application for evaluation, risk analysis, and response planning.

Catch Subtle Issues Before They Become Big Problems

SCADA systems are not the be-all, end-all of pipeline integrity management – routine visual inspection of pipelines is also of paramount importance. In a 13-year analysis of a refined petroleum pipeline – approximately 528 miles long, carrying 172,900 barrels per day, and over 50 years of age – the results show that less than 50% of spills and other incidents were detected by SCADA systems¹. In fact, the majority of these incident were detected visually, whether by employees/contractors working on-site, inspections/site visits, ROW surveillance and even the nearby public. Getting eyes on your asset and ROW is one of the best ways to prevent incidents. Don’t you want to get your eyes on the problem before the public does?

The trouble is, a visual inspection is extremely hard to make a routine and consistent process with human personnel – it takes lots of money and lots of manpower. Not to mention, human eyes get complacent and fatigued, and can miss small changes from inspection-to-inspection.

What if you could automate this task? Well, you can. Just as a SCADA system continually gathers sensor data from within the pipeline, autonomous UAVs can continually gather visual data on your pipeline and ROW.

Harnessing the Power of AI for Oil Pipeline Leak Detection

While high-resolution imagery and visual models – such as orthomosaics and Digital Elevation Models (DEMs) – give you an eye on your pipeline, pinpointing the critical details can become a huge undertaking in and of itself. AI-driven computer vision can solve this problem, autonomously analyzing thousands of images for indicators of a leak or ROW issues that could cause damage to your pipeline.

With autonomous right-of-way monitoring, you can identify external issues that can pose great threats to your pipeline integrity. According to the Pipeline and Hazardous Materials Safety Administration’s (PHMSA) 2019 Pipeline Incident Report, external forms of damage such as excavation, natural forces, and other outside issues accounted for over $100 million worth of damage over the course of the year². Spotting these issues and mitigating them before damage occurs avoids the loss of money, time, and your hard-earned reputation.

Beyond the ability to detect objects/anomalies and classify them accordingly, this technology enables changes to be tracked over time. From inspection-to-inspection, AI can identify subtle changes that are indicative of a developing issue along your line – such as approaching construction activity, vegetation encroachment or gradual erosion of the earth.

Change detection can also provide value during the remediation stage. Routine monitoring of a spill clean-up area can show whether the health of the environment is improving, stagnating, or even continuing to decline. This is valuable information that helps prove to regulators that your cleanup efforts are effective.

Predicting Oil Leaks Instead of Reacting to Them

The most powerful way to improve your oil pipeline leak detection methods is by leveraging all of this data to make accurate predictions about future issues. With the ability to collect images, identify anomalies, and track changes over time – these critical points of data are used to build predictive analytics around your asset integrity:

• Readings from your SCADA systems and leak-affected segments of pipeline
• AI-detected anomalies on your ROW
• Important external sources such as weather at the time of a natural force damage event

The fusion of these data sources is used to create a model making predictions with a certain sense of probability. These predictive insights help you to understand:

• Hotspots along your pipeline where issues tend to occur
• The conditions that commonly lead to issues such as certain weather conditions
• How potential ROW issues such as erosion, third-party activity and construction or even vegetation encroachment are likely to grow in the future

Traditional leak detection technologies leave you scrambling to react to vague alerts around potential leaks. Aerial data gives you the data you need to get proactive and prevent damage before it occurs.

Visualize Your Pipeline Integrity Management

An aerial data platform improves your ability to verify alerts from your SCADA system and get in front of external threats on your ROW. With the ability to holistically aggregate key pieces of data from the variety of sources important to your organization, you can utilize predictive modeling to mitigate issues before they manifest.

Have questions about how high-quality aerial data can elevate your organization?
Contact our team to discuss your unique challenges and data requirements.

References

1. Oil Pipeline Safety Failures in Canada, Equiterre, Shelly Kath, 2018
2. Pipeline Incident 20 Year Trends, PHMSA, 2019