Understanding and Validating Probability of Detection in Qube's Continuous Emissions Monitoring Technology

Introduction

Continuous Monitoring (CM) has the capability to detect, locate, and quantify emissions, 365 days a year. As a leading provider of continuous monitoring technology, Qube is often asked to provide performance metrics of the technology to regulators, certifiers and customers.

What is Probability of Detection?

Probability of detection (POD) curves are an industry accepted metric for understanding the detection capabilities of various methane detection technologies including OGI cameras, aerial surveys, and continuous monitoring systems. They plot the likelihood of detecting an emission event with a given technology across various emissions sizes that are often represented in grams or kilograms per hour.

For continuous monitoring technologies, probability of detection is a complicated subject. POD can be impacted by many variables including number of sensors deployed, wind conditions, duration of the test, and distance from the source. Therefore, to accurately estimate POD, numerous tests must be performed with different device configurations.

Third Party Blinded Studies

Qube works with industry and academia to understand and continually improve the performance our technology. One way we do this is through blind controlled release test studies conducted by 3rd parties. These studies are built to test and benchmark the performance of solutions in a controlled environment. Over the past 6 months, Qube ran two such studies, one with consulting firm Highwood Emissions, and the second through the METEC Advancing Development of Emissions Detection (ADED) Program at Colorado State University.

Qube Probability of Detection

For the METEC ADED program, controlled releases were performed over a fifteen week period from February to May 2022. Based on the test results, as indicated in Figure 1, Qube could detect 90% of emissions greater than 1.5 kg/hr. For very small emissions, around 0.1 kg/hr., Qube’s technology could detect emissions around 60% of the time.

Figure 1 - METEC ADED program results

These results compare favorably to the blind study performed by Highwood Emissions, in Figure 2, where Qube could detect 90% of emissions greater than 1.6 kg/hr. The results from both studies are well correlated and the differences can be explained by device distance from source, device configurations at each test site, and differing wind regimes.

Figure 2 - Highwood Emissions Probability of Detection results for Qube Technologies

Detecting the Majority of Emissions

"Qube's solution can consistently detect the leaks responsible for the majority of total emissions."

According to Highwood Emissions, the probability of detection results derived from the blinded study indicate that “Qube's solution can consistently detect the leaks responsible for the majority of total emissions."

The term "majority of emissions" is in reference to our current understanding of actual emissions from major oil basins in the Unites States. According to Highwood, emissions rates of 1.5kg/hr. or more make up 65% of all emissions in the Marcellus, 87% of emissions in the Barnett, and more than 99% of all emissions from the Permian basin. Industry's ability to quickly detect and remediate leaks of this size will play a significant role on total industry emissions reduction.

Upcoming blinded studies

This fall, Qube will publish results from blinded studies on our localization and quantification capabilities. Until then, read the Highwood Emissions Management whitepaper on Qube’s Probability of Detection, and contact us for more information.

Download the Study

The complete Qube Technologies Continuous Monitoring Probability of Detection whitepaper was published in the August 2022 Highwood Emissions monthly bulletin, and is available for download here: