CI OGI Inspector Plus, Sensys MagDrone R4, and AI/ML for Detection of Abandoned Oil Wells and Methane 

open_in_new https://www.aerofrohne.com

Leveraging Drones With ICI OGI Inspector Plus, Sensys MagDrone R4, and AI/ML for Detection of Abandoned Oil Wells and Methane

Executive Summary

The rapid growth in drone and AI technology has opened up new possibilities for addressing long-standing challenges in the energy sector. One critical issue is the identification and monitoring of abandoned oil wells, which pose significant environmental and safety risks due to the potential for leaks and contamination. Traditional methods for locating these wells are time-consuming, costly, and often ineffective. By deploying drone technology, coupled with AI-driven image processing and machine learning algorithms, we can efficiently detect and monitor abandoned wells, providing a scalable and accurate solution that reduces costs and environmental impact.

1. Introduction

Oil well abandonment is a growing environmental challenge. Hundreds of thousands of wells across various oil fields globally remain unmonitored, posing threats to ecosystems and communities. Current methods for locating these wells rely on historical records and time-intensive surveys, which often fail to capture all abandoned wells accurately.

This white paper explores how combining drone imagery with AI and machine learning can automate the detection of abandoned wells, increase accuracy, and reduce costs and response times in environmental monitoring.

2. Problem Statement: Challenges in Detecting Abandoned Wells

• Environmental Risks: Abandoned wells can leak harmful chemicals, contaminating groundwater and releasing methane, a potent greenhouse gas.
• Data Gaps: Records of well locations are often incomplete or inaccurate, leading to missed or unaddressed abandoned sites.
• Costly and Inefficient Detection: Manual surveys are labor-intensive, costly, and time-consuming, limiting scalability and responsiveness.

Traditional detection methods are insufficient for the scale and urgency required to address these environmental risks.

3. Solution Overview: Drones, AI, and Machine Learning

Using drones equipped with high-resolution imaging sensors, combined with AI and machine learning algorithms, offers an effective solution for identifying and monitoring abandoned wells. This approach can cover vast areas quickly, generating real-time data and identifying well locations with precision.

Key Components

• Drone Imaging: High-resolution RGB and thermal imaging drones can capture detailed data from large areas, detecting visual and thermal anomalies associated with abandoned wells.
• AI-Based Image Analysis: AI algorithms process imagery, distinguishing between active and inactive sites and identifying wellheads and infrastructure associated with abandonment.
• Machine Learning Models: Machine learning models are trained to recognize patterns specific to abandoned wells, enabling the system to detect wells more accurately over time.
• ICI OGI Payload: The ICI OGI Inspector Plus could help identify abandoned wells in oilfields by detecting gas leaks indicative of residual hydrocarbons, a common issue at such sites. With its high-sensitivity optical gas imaging capabilities, it can locate methane and other gases emitted from wellheads, pipelines, or other oilfield infrastructure.

4. Technology Stack and Methodology

4.1 Drone and Sensor Technology (with Drone Magnetometers)

Drone and Sensor Technology now includes magnetometers, which are essential for detecting underground metallic structures associated with abandoned wells. These sensors work in tandem with high-resolution RGB cameras, thermal imaging, and optical gas imaging to provide a comprehensive solution.

• High-Resolution RGB Cameras: Capture detailed ground imagery, helpful for identifying surface markers, terrain disturbances, or residual equipment.
• Thermal Imaging: Detects heat signatures that might indicate methane leaks or disturbed soil, common signs of abandoned wells.
• ICI OGI Inspector Plus: Combines high-sensitivity infrared (IR) imaging with advanced optical gas imaging to detect volatile organic compounds (VOCs) like methane, allowing real-time monitoring of emissions.

• Magnetometers: By measuring magnetic anomalies, magnetometers can detect subsurface ferromagnetic objects, such as metal well casings, even in challenging or obscured locations. This enhances detection accuracy, particularly where surface indicators are absent or compromised by vegetation or topography.

Updated Workflow with Magnetometer Integration

1. Data Acquisition: Drones equipped with magnetometers, RGB, thermal, and OGI sensors autonomously capture data across targeted areas.
2. Geo-Tagging and Preprocessing: Data from multiple sensors is synchronized, enhanced, and tagged for precise mapping.
3. AI-Powered Object Detection: AI algorithms process imagery and magnetic readings to identify surface and subsurface markers indicative of abandoned wells.
4. Classification and Filtering: Machine learning models classify detected structures as abandoned wells, leveraging magnetic signatures to confirm wellhead locations.

Enhanced Results and Benefits

The inclusion of drone magnetometers significantly improves the detection of well casings, even when hidden beneath soil or vegetation. Combined with thermal and OGI data, this approach reduces false positives and enhances well-location precision, offering cost-effective and scalable environmental monitoring.

4.2 AI and Image Processing Workflow

1. Data Acquisition: Drones autonomously capture images across targeted areas.
2. Preprocessing: Images are geo-tagged, enhanced, and prepared for analysis.
3. AI-Powered Object Detection: AI algorithms identify patterns and objects that match characteristics of abandoned wells.
4. Classification and Filtering: Machine learning models classify identified structures as abandoned wells based on shape, size, location, and other contextual data.

4.3 Machine Learning Training and Validation

• Supervised Learning: Historical data on abandoned wells is used to train models, enabling the system to improve its accuracy over time.
• Validation with Field Data: On-the-ground verification is conducted for model refinement and to reduce false positives.

5. Results and Benefits

The integration of drones, AI, and machine learning offers several advantages for abandoned well detection:

• High Accuracy: Enhanced detection capabilities significantly reduce false positives and improve the accuracy of abandoned well identification.
• Cost-Effective Scaling: Automated drone inspections reduce labor costs and are scalable across vast regions.
• Environmental Impact Mitigation: By locating and monitoring abandoned wells, this solution helps prevent leaks and contamination, contributing to environmental protection efforts.

Case Study (Hypothetical)

In a recent test deployment, drone-based AI and machine learning detected 92% of abandoned wells in a 50-square-mile test area, compared to only 67% detected by traditional survey methods. Additionally, the project completion time was reduced by over 60%, highlighting the effectiveness and efficiency of this approach.

ICI OGI Inspector Plus could involve an oilfield operator deploying the device to identify leaks at abandoned or low-production wells. By equipping the Inspector Plus with GPS-enabled imaging, the team conducts systematic field scans to detect VOCs such as methane. The device identifies emissions through thermal IR imaging, allowing rapid localization of high-emission sites. Collected data is then uploaded to a cloud platform for analysis, providing the operator with actionable insights to address emissions, improve site safety, and meet regulatory standards.

6. Future Prospects and Challenges

Advancements

• Incorporating Advanced Sensors: Including methane detectors and hyperspectral imaging can further enhance detection capabilities.
• Real-Time Analysis: Developing algorithms for real-time data processing can support immediate responses to high-risk wells.

Challenges

• Data Privacy and Regulatory Compliance: As drones collect extensive data, addressing privacy and regulatory concerns will be critical.
• Model Generalization: Expanding the accuracy of machine learning models across different geographies and environmental conditions remains a technical challenge.

7. Conclusion

The combination of drone imagery, AI, and machine learning presents a compelling solution for the accurate and scalable detection of abandoned wells. This innovative approach enables environmental and energy sectors to tackle the challenges of well monitoring more effectively, safeguarding communities and ecosystems from potential contamination. By continuing to advance and refine these technologies, we can expect a significant improvement in the industry’s ability to manage and remediate abandoned oil wells worldwide.

Annual Proposed Budget

Data Analyst Base Salary: $80,000

Drone Pilot Base Salary: $120,000

Equipment Costs Breakdown:

• Titan X8 Drone Platform: $27,919
• DJI Matrice 350 RTK + UgSC Skyhub: $18,000
• ICI OGI Inspector Plus Payload: $100,000
• Integration of ICI OGI Inspector Plus & Titan X8 Drone: $1000
• Sensys MagDrone R4 Magnetometer : $50,000
• Data Processing Hardware/Software: $30,000

Other Expenses Breakdown:

• Drone Pilot Travel and Lodging: $26,000
• Site Preparation: $8,000
• Optical Gas Imaging Infrared Thermography Certification: $2,300
• Data Storage and Cloud Computing: $7,000
• Safety Equipment: $4,000
• Field Supplies: $3,000
• Insurance: $5,000
• Administrative & Reporting Costs: $10,000

Total Annual Cost: $492,219

Comments