Beyond Site Rate

Author: Brad Roger, Product Manager

This article discusses why we’ve added Emission Rate to the Site Dashboard

Key Highlights: 

  • Site Rate is instantaneous, but not the most accurate metric for quantification 

  • Qube calculates an Emission Rate which is more accurate, but less instantaneous 

  • In the Qube Platform v2.17 release, the Site Rate line chart in the dashboard is being improved to better visualize the more accurate Emission Rate alongside the Site Rate 

  • In a future release, the API endpoints will be improved to clearly differentiate what data is subject to change 

In a previous blog post (PPM is not enough) we explained how important it is to use flux and site rate instead of raw concentration readings. In this post we will explain why we believe that Site Rate is also not enough.

Site Rate is the rate at the most likely emission source at each minute and is subject to change over the duration of an emission event. The most impactful step in calculating a Site Rate is localizing an emission to the correct source, since distance from the device to the source of the emission is how the physics model determines scale of the plume.

What Qube and other stationary sensor solutions provide as Site Rate is simply the device flux scaled based on an instantaneous localization (usually averaged over some period of time). 

Through extensive testing at our Controlled Release Test Facility in Alberta, we have found that the short-term localization used to generate a Site Rate does not truly reflect how an emission actually behaves.

A Site Rate that only looks at 15 minutes of data is not getting the full picture and is too simple to correctly quantify emissions on an operating oil and gas facility with a high enough level of accuracy. 

The Qube Emissions Model has been built to get smarter over the course of an emission. Each Qube Axon device has its own anemometer so we have a complete view of how the wind field changes over the landscape of a facility through time.

As the wind direction changes across the site, and the methane plume forms swirls and eddies, and washes over the different devices, we can use the presence or absence of gas readings on different devices to tell a more complete story about what a methane plume looks like, and how it is moving around in space over time, and most importantly where the emission is originating. 

When methane readings return to baseline and we determine that an emission event has ended, our model goes through a process we call the Emission Roll-Up. The current model assumes that a methane plume registered by our devices likely came from a single source. (Note: we are working on a more advanced model that can detect and model multiple simultaneous emissions.) 

By looking at an entire emission event from start to finish (however long it may be) through the lens of our physics-based plume model, we can select the most probable emission source with higher confidence, and consequently improve the accuracy of our quantification result. 

The implication here is that results published in real time may change throughout the course of an emission. 

In an effort to be more transparent about the rates used in the platform for determining emission volumes and triggering alarms, we are adding the Emission Rate to the Site Dashboard. 

Emission Rate (red line)

The final inferred emission rate, localized and corrected to a single emission source, and averaged across the emission duration. This rate is used to calculate the total volume of a given emission and accounts for periods where there is limited visibility. 

Site Rate Bounds (shaded area)

The upper and lower bounds of ALL possible emission rates across ALL probable emission sources. Rates are computed at a 1-minute interval for each probable emission source, and the resulting bounds show how large (or small) the final calculated emission rate could be if localized to a different emission source. 

Site Rate (grey line)

Is the rate at the most likely emission source at each minute and is subject to change over the duration of an emission event. 

Example: 

Emission Rate line chart

In the above case, we performed a controlled release at our test facility releasing methane from Release Point 2 (NRP2) at 5.0 kg/h for a 4-hour period. As you can see, the observed Site Rate (the dark grey line) fluctuates as the methane plume hits different sensors. There are even points where Site Rate drops to 0 when the plume passes between sensors and our visibility of the plume is interrupted. After the Qube Emissions Model determines the emission has ended, the Emission Roll-Up is performed to determine the highest likelihood source of the emission and produces a corrected Emission Rate which is shown in red. 

Key Takeaways:

  • We’ve added Emission Rate to the Site Dashboard for additional context on how we determine emission volumes and trigger alarms. 

  • Emissions Rate refers to the final inferred emission rate, localized and corrected to a single emission source, and averaged across the emission duration.

  • This new feature tells a more complete story about what a methane plume looks like, how it is moving around in space over time, and where the emission is originating.  

Previous
Previous

White Paper: Reducing methane emissions: Implementing data science informed operation and maintenance work practices using continuous monitoring technology (abstract)

Next
Next

A Better Way to Think About False Positives