The role of hyperspectral cameras in methane gas detection.
Methane leaks are harmful to the environment, the energy sector, and our safety. But how do you detect a gas that’s invisible and odorless? Spectral cameras provide the answer.
You can’t see or smell methane (CH₄), but it’s present. It makes up the majority of raw natural gas found underground and is transported across land and sea to supply energy.
Methane is highly flammable and toxic, and as the second most abundant greenhouse gas after carbon dioxide, it is a significant contributor to global warming. When methane leaks, it becomes a serious issue. Methane emissions pose threats to public health and safety, hinder efforts to meet global net-zero targets, and cost the energy industry an estimated $60 billion annually.
Detecting methane is a challenge, especially with conventional cameras and sensors.
The solution? Spectral imaging.
By combining industrial cameras with hyperspectral imaging systems, it’s possible to identify hydrocarbons like methane in the atmosphere by analyzing the wavelengths they absorb. Machine learning helps these systems process images and videos to automatically detect methane leaks and trace them back to their origin.
In this blog, we’ll explore how hyperspectral cameras can be effectively used to detect methane gas leaks.
Hyperspectral vision is at the forefront of industrial imaging right now. With more energy companies realizing the potential of spectral cameras for gas detection, there’s never been a better time to implement them.
Here at Maxlab, we recommend you harness the power of hyperspectral vision and high-performance edge computing with the Tokay PRO Edge AI Camera. Featuring a low-power and lightweight design, the Tokay PRO is easy to integrate with multiple thermal and spectral
The issue with methane leaks
The global natural gas system is supported by thousands of kilometers of pipelines, forming a vast and intricate network. This infrastructure, which also includes compressor stations, valves, and distribution centers, is incredibly complex. Given its scale, methane leaks are almost inevitable.
Methane leaks can happen during production, processing, storage, or distribution, and the amount released into the atmosphere is higher than many realize. It is estimated that nearly 135 million tons of methane were emitted globally in 2022, with satellite data identifying ‘super emitter’ sites in countries such as Russia, Turkmenistan, and the US as major contributors. For example, in 2022, satellites recorded the Raspadskaya coal mine in Russia emitting 95 tons of methane per hour.
Disasters like the Aliso Canyon gas leak in California, which released nearly 100,000 tons of methane in 2015, and the 2022 Nord Stream pipeline explosion, one of the largest leaks on record, which emitted up to 400,000 tons of methane, further underscore the scale of the issue.
Although methane has a relatively short atmospheric lifespan of seven to twelve years, compared to carbon dioxide, which can persist for centuries, it is far more potent. Methane traps 80 times more heat than carbon dioxide and is responsible for 30% of global warming. This means that methane does more environmental damage in one year than carbon dioxide does over a hundred.
Methane leaks are also financially costly. With many regions relying on limited gas pipelines, energy companies must continuously monitor their infrastructure for leaks. Even a small leak can force a pipeline shutdown, and each hour of downtime can result in millions of dollars in losses.
Challenges of methane gas detection
Methane is invisible—it has no color or smell—which makes it extremely difficult to detect. Many methane leaks from urban, industrial, agricultural, and waste management sources are likely underreported because of this challenge. For example, in the US Permian Basin, methane emissions are 14 times higher than officially reported.
Methane is released into the atmosphere as plumes. Being lighter than air, it disperses quickly and has a small pixel-footprint signature, making it hard to distinguish from background images and difficult to detect.
Traditional cameras and sensors often struggle with methane detection. They frequently require experts to manually sift through visual data, which is neither scalable nor entirely reliable. Human error is a risk, and images can be obscured by background noise, while other hydrocarbons or pollutants can trigger false alarms when attempting to identify methane plumes.
For these reasons, the energy industry is increasingly adopting hyperspectral cameras to improve methane detection.
What is a hyperspectral camera?
Hyperspectral imaging is a cutting-edge technology that enhances industrial cameras by enabling them to detect light wavelengths beyond the visible spectrum. This includes ultraviolet, x-ray, and most importantly, infrared wavelengths, which are key for detecting gases.
These high-resolution infrared hyperspectral cameras analyze the light spectrum in each pixel of an image to provide detailed information about the substances present. They work by identifying the specific wavelengths of light absorbed by different chemicals.
Because of this capability, hyperspectral cameras are increasingly used in remote sensing, particularly for detecting hydrocarbons. They are quickly becoming the most reliable tool for automatically identifying and tracking methane plumes around the world.
Using hyperspectral cameras for methane detection
There are three main stages to finding a methane gas leak with spectral cameras:
– Detection
– Analysis
– Visualization
Detection
As methane plumes rise into the atmosphere, they absorb infrared light from the sun, leaving behind a distinct spectral “fingerprint.” Hyperspectral cameras are designed to detect this unique signature by capturing hundreds of images. Each pixel in these images contains a full spectrum of light, and together, these pixels form a spectral band—a range of wavelengths that helps identify the presence of methane.
Analysis
Hyperspectral systems utilize deep-learning models to analyze wavelengths, examining spectral data from the visible light range, starting at around 400 nanometers (nm), up to the 2,100-2,500nm range, where hydrocarbons like methane are found.
By narrowing down the spectral band, these systems can differentiate methane from other hydrocarbons, often using just a single pixel to detect a hidden gas plume. Even if methane is leaking in small amounts or at low concentrations, the high sensitivity of hyperspectral cameras ensures it will be detected.
Visualization
Once the hyperspectral system has processed and analyzed the spectral data, the methane plume can be visualized in real-time 3D maps and displays. In tandem with mapping software and geo-location technology, hyperspectral cameras can help quantify the size of a methane plume and determine where it’s going, as well as where it came from. And hey presto – there’s the source of the leak.
How are hyperspectral cameras deployed?
Hyperspectral cameras are typically mounted on planes, helicopters, and drones, allowing them to safely scan and monitor methane leaks over vast areas of the air. These cameras are highly resistant to adverse weather conditions and can collect data across diverse terrains and landscapes, helping estimate the rate and direction of methane plume movement. With its wide field of view, airborne hyperspectral imaging offers a cost-effective way to gather large amounts of detailed information at once.
Additionally, satellites equipped with spectral cameras are used to scan broad sections of the Earth’s surface. These space-based systems can detect methane leaks from pipelines across multiple countries and terrains, including remote areas like deserts and oceans while providing high-resolution imagery and reducing the need for airplane emissions.
Satellites can also collaborate, combining their perspectives to enhance weak methane signals and pinpoint the exact source of leaks. They can guide planes to specific areas for further investigation and issue alerts when pipeline leaks occur, enabling quick responses.