ClientEERC / iPIPE
ScheduleNov 2024 - Nov 2025
PartnersTOKU, HESS

Improving Leak Detection in Complex Gathering Systems with Advanced Pressure Sensing and AI

Effective leak detection in complex gathering systems is crucial for ensuring safety, enhancing operational efficiency, and protecting the environment. Recognizing the challenges in this area, we are embarking on a 12-month joint research project with TOKU, a leading pressure sensor company. Funded by the University of North Dakota’s Energy and Environment Research Center (EERC) through the iPIPE program, this project aims to revolutionize leak detection, localization, and verification in gathering systems by leveraging advanced pressure sensing technology and machine learning (ML).

Project Overview

Gathering systems, which collect oil and gas from multiple wells and transport them to processing facilities, are inherently complex. Traditional leak detection methods often struggle with these systems due to their intricate networks and variable operating conditions. Our project addresses these challenges by integrating TOKU’s high-frequency pressure sensors with our existing leak detection solutions, enhanced by machine learning and anomaly detection techniques.

Innovative Pressure Sensing with TOKU

TOKU’s pressure sensors are uniquely designed to operate in locations where power and connectivity are typically constraints. Key features include:

  • High-Frequency Data Collection: Provides granular pressure data for more accurate analysis.
  • Solar and Battery Power: Ensures continuous operation without reliance on external power sources.
  • Cellular Connectivity: Facilitates real-time data transmission even in remote areas.

By deploying these sensors across the gathering system, we can achieve comprehensive sensor coverage that was previously unattainable.

Enhancing Leak Detection Capabilities

The integration of TOKU’s sensors enables us to:

  • Leverage High-Resolution Data: The abundance of detailed pressure data improves the sensitivity of our leak detection algorithms.
  • Apply Machine Learning and Anomaly Detection: ML models can identify patterns and anomalies in the data that may indicate leaks, even in complex network configurations.
  • Improve Detection Accuracy: Enhanced algorithms reduce false positives and improve the reliability of leak detection.

Localization of Leaks

Detecting a leak is only part of the challenge; pinpointing its location within a complex gathering system is equally critical. Our project focuses on:

  • Regional Localization: Developing methods to narrow down leak locations to specific regions or laterals within the system.
  • Practical Implementation: Ensuring that localization techniques are feasible for real-world applications and can be readily adopted by operators.

Verification and Validation Enhancements

One of the difficulties in current leak detection is confirming the exact start time of a leak and verifying its occurrence. With the new pressure sensors:

  • Enhanced Validation Workflows: Improved data quality and frequency allow for more precise correlation between detected anomalies and actual leak events.
  • Reduced Uncertainty: Operators can make informed decisions with greater confidence, leading to quicker response times.

Importance of the Project

This research initiative holds significant promise for the industry:

  • Safety Improvements: Early and accurate leak detection minimizes risks to personnel and infrastructure.
  • Operational Efficiency: Reducing downtime and maintenance costs by addressing leaks promptly and effectively.
  • Environmental Protection: Preventing leaks mitigates environmental impact, aligning with regulatory compliance and sustainability goals.

Conclusion

This collaborative project represents a significant advancement in addressing the challenges of leak detection in complex gathering systems. By combining TOKU’s innovative pressure sensing technology with our expertise in machine learning and pipeline monitoring, we aim to deliver a solution that not only detects leaks more effectively but also localizes and verifies them with greater precision.

We are committed to enhancing safety, boosting operational efficiency, and safeguarding the environment through this initiative. We look forward to sharing our progress and findings with the industry and invite interested parties to engage with us as we work towards these vital goals.