A natural-gas leak-detection tool uses sensors and machine learning to locate leak points at oil and gas fields, promising automatic, affordable sampling across vast natural gas infrastructure. The Autonomous, Low-cost, Fast Leak Detection System (ALFaLDS) was developed to discover accidental releases of methane, a potent greenhouse gas.
ALFaLDS detects, locates, and quantifies a natural-gas leak based on real-time methane and ethane (in natural gas) and atmospheric wind measurements that are analyzed by a machine-learning code trained to locate leaks. The code is trained using high plume dispersion models and the training is finessed onsite by controlled releases.
Tests demonstrated that ALFaLDS locates the engineered methane leaks precisely and quantifies their size. This novel capability for locating leaks with high skill, speed, and accuracy at lower cost promises affordable sampling of fugitive gas leaks at well pads and oil and gas fields. ALFaLDS's success in locating and quantifying fugitive methane leaks at natural-gas facilities could lead to a 90 percent reduction in methane emissions if implemented by the industry.
ALFaLDS uses a small sensor, which makes it ideal for deployment on cars and drones. The team is developing the sensors that were integrated with a mini 3D sonic anemometer and the powerful machine-learning code in the tests. The code is autonomous and can read data from any gas and wind sensors to help find leaks fast and minimize fugitive emissions from the vast network of natural-gas extraction, production, and consumption.