The US is home to an extensive network of pipelines transporting natural gas, NGLs, LPG, and olefins, which are in critical need of improved leak detection solutions. Traditional methods often fall short due to high costs, intermittent monitoring capabilities, excessive sensor requirements, and the need for specialized expertise. These systems lack the accuracy and responsiveness, resulting in false alarms and delayed detection, which can compromise safety and efficiency.
Our project aims to develop an innovative leak detection system that leverages existing pipeline sensors, reducing the need for additional instrumentation. By leveraging machine learning (ML) and advanced signal processing, we will create a solution that significantly reduces false alarms, improves sensitivity and robustness, provides real-time monitoring, and adheres to the principles of physics and hydraulics.
The US is home to an extensive network of pipelines transporting natural gas, NGLs, LPG, and olefins, which are in critical need of improved leak detection solutions. Traditional methods often fall short due to high costs, intermittent monitoring capabilities, excessive sensor requirements, and the need for specialized expertise. These systems lack the accuracy and responsiveness, resulting in false alarms and delayed detection, which can compromise safety and efficiency.
Our project aims to develop an innovative leak detection system that leverages existing pipeline sensors, reducing the need for additional instrumentation. By leveraging machine learning (ML) and advanced signal processing, we will create a solution that significantly reduces false alarms, improves sensitivity and robustness, provides real-time monitoring, and adheres to the principles of physics and hydraulics.
We will undertake extensive data collection to capture both normal and leak behaviors, followed by a thorough evaluation and refinement of ML and signal processing technologies. By collaborating with industry operators, we will gather real-world data and conduct physical withdrawals to validate our product. This rigorous testing will ensure that our system is effective under various pipeline configurations and real-world operating conditions.
Our final solution will be integrated into Flowstate’s existing commercial Leak Detection System (LDS), making our advancements easily accessible to the community with minimal overhead. This integration will enable operators to monitor pipeline integrity more efficiently and effectively. By providing a reliable method for leak detection, our system will enhance pipeline safety, operational efficiency, environmental protection, and regulatory compliance, addressing the pressing need for better pipeline monitoring technologies.
Create a real-time leak detection system for compressible hydrocarbon products, validated with real-world data. This system will be integrated into a commercial Leak Detection System (LDS) to enhance pipeline safety, environmental protection, regulatory compliance, and operational efficiency. Our objective is to deliver a reliable and effective solution that addresses the critical need for improved leak detection in the pipeline industry.