Only 3 years ago methane measurements from oil and gas producing sites were sparse. But, recent published studies by the StFX FluxLab, by Carleton University and EDF, by industry, and by regulators, have sketched out a new baseline. It's now well accepted that emissions are somewhat higher than government inventory estimates. Thankfully new methane regulations roll out in 2020 that will reduce the methane-GHG footprint of Canadian oil and gas production.
As part of the StFX-published studies, we visited about 6500 well sites (most on 3 occasions) with a mobile lab to make measurements with Picarro and other high precision analyzers. We have already published on well over half of these measurements, with more papers coming.
Happy browsing. Please visit frequently as we're adding new features soon.
Click here to visit the Fluxlab main webpage to see other interesting projects!
With the right top corner box, you can change the base layer and you can remove the points and the Canada boundaries (in orange) by simply unclicking the boxes.
You can view information for individual sites by simply hovering with your mouse over each point.
If you click on individual red (i.e. emitting site) points, a graphic will pop up to show how that particular site compares to other measured emitting sites across the entire dataset. The graphic is downloadable. Note: Only measurements from emitting sites are shown in this graphic.
Note: The longitude and latitude have been obscured. Thus, the current locations on the map are only approximate.
The map below shows the distribution of methane emissions, modelled using measurement-derived emission factors specific to geography and infrastructure.
You can extract information for individual cell by simply clicking with your mouse over each quadrat. After clicking, a new map will pop up where the longitude and latitude have been obscured. Thus, the current locations on the map are only approximate.
We pooled and analyzed methane datasets from several measurement campaigns, covering six oil and gas producing regions in British Columbia, Alberta, and Saskatchewan. In total, we collected methane measurements from over 6300 well pads, which were taken between 2016 and 2018. During each campaign, we collected atmospheric measurements of three or more gases at ppb levels, geolocated, every second while driving. Preplanned survey routes through regions of high infrastructure density were repeated 2-6 times per campaign. This work has been a collective effort among students and lab technicians in the FluxLab at StFX using vehicle-based surveys. Data from roughly half of these campaigns have been published in Atherton et al., 2017 , O'Connell et al., 2019 and Baillie et al., 2019.
An infrastructure sites (e.g., Single oil well, Single gas well, etc) is defined here as a group of wells and/or facilities within 90 m of each other. For an infrastructure site to be considered sampled, our truck had to drive downwind and within 500 m of the site at least two times. For a site to be classified as emitting, we had to detect sequences of methane enriched concentrations downwind and within 500 m on more than 50% of the times it was sampled. If multiple sites are within range of a plume, the closest site is tagged as the emitter. Volumetric emissions rates were estimated via a point-source Gaussian Dispersion Model that uses both measured, and estimated parameters. Since we do not know the exact source of emissions, individual rates are calculated for each piece of infrastructure within an emitting infrastructure site (in other words, we calculate an emission rate for each well or facility within a single site). The median is then used to represent the overall rate for emitting sites. It is these values that are shown throughout this application. For more information refer to O'Connell et al., 2019 and Baillie et al., 2019.
We used our emission rate estimates to calculate emission factors (EFs). Emission factors were calculated by the following equation: sum of all emissions from emitting sites + the sum of all emissions from non-emitting sites (which is zero) divided by the total number of sampled sites (emitting and non-emitting). This calculation was done individually for all unique combinations of infrastructure types and regions. In other words, emission factors were calculated based on measurements from a specific region and infrastructure type. As a result, a type of infrastructure could have multiple EFs if it is a common type across multiple regions. For example, an EF for a single oil well in Lloydminster might be different than an EF for a single oil well in Peace River. We chose to calculate EFs based on infrastructure type and region because we know from previous studies that emissions can vary substantially based on these two factors. Our method lets us account for the variability that exists within the upstream sector, which in turn helps avoid scenarios of over and under-estimations.
The inventory shown in this application was modelled using calculated emission factors (via the above method). The following steps give a brief description of how the model works: 1. Determine the infrastructure types (e.g., Single oil well) for all sites. 2. Calculate the total number of sites for each infrastructure type within each production region using IHS databases (based on the 2019 infrastructure counts); 3. Allocate an emission factor (EF) to each site by infrastructure type, and region (when possible*); 4. We present the sum of all emissions within a ~6 km2 cell (0.1 degree x 0.1 degree).
*Measurements presented in this application covered Lloydminster (AB and SK sides), Grande Prairie, Red Deer, Medicine Hat, Montney BC, and the Weyburn area in Saskatchewan. Thus, emission inventory estimates for High Level, Bonnyville, St. Albert, Midnapore, and British Columbia are approximations based on EFs calculated from measurements in Red Deer, Medicine Hat and Grande Prairie. CHOPS EFs from Lloydminster were not applied to other areas because that development is unique in Canada.
For more information on the methodology, the readers are invited to consult this document.
Many of these values have been previously published in the peer reviewed literature. All unpublished values were produced using the same methodology, and should be regarded as consistent with peer review values. We reserve the right to update this resource on occasion, based on new approaches to data processing or the inventory model, or to reflect the best available data such as updated infrastructure or activity counts.
Last version: February 2020.
This website was built using RShiny.
The following R packages were used in to build this RShiny application: