Saskatchewan's oil and gas methane: how have underestimated emissions in Canada impacted progress toward 2025 climate goals?

Using federal inventory methods and detailed operator-reported emissions data, we developed a methane inventory estimate for Saskatchewan's upstream oil and gas sector. Similar inventory estimates have been published for Alberta by Johnson and Tyner (2020) and Seymour et al (2022) so a summary of data sources/methods is provided in table 1 and additional details are provided in SI section S2.

As described in section 2, some operators are subject to fleet-level emission limits at their oil sites. Any excess emissions from these operators can be readily calculated from reported emissions in terms of CO₂e but not in terms of corresponding methane since emissions could come from any combination of venting and/or flaring. To estimate the excess methane emissions, we assume that reported venting would be preferentially reduced before flaring for each operator to meet its fleet emission limit.

Finally, we also perform a simple, short-term forecast of inventoried emissions to demonstrate Saskatchewan's expected inventory up to 2025. This is done by scaling production and emission volumes from the most recent inventoried year (2022) using forecasted production of light/medium oil, heavy oil, and gas volumes published by the Canada Energy Regulator (CER 2020). Rescaled emissions are then reduced assuming all qualifying operators fully comply with their annual fleet-level emission limits. For reference, between 2022 and 2025, the CER forecasts Saskatchewan's production of light/medium oil to decrease by 23%, heavy oil to increase by 32%, and primary gas production to remain stable. Other oil and gas methane forecasting efforts have considered changes to site counts (ECCC 2021b) but since Saskatchewan's emission limits depend on produced volumes, not site counts, forecasting of the latter is not necessary.

We updated the methane inventory estimate to account for underreported emissions from CHOPS wells. While Saskatchewan-specific studies (Baillie et al 2019, MacKay et al 2019, Vogt et al 2022) and those quantifying underestimated flare (Tyner and Johnson 2021, Plant et al 2022, Seymour et al 2022) and combustion slip emissions (Vaughn et al 2017, Tyner and Johnson 2021, Lyman et al 2022, Johnson et al 2023b) were considered for inclusion in an updated inventory, it was unclear how to derive region-specific updates for the current inventory.

Recent equipment-level aerial LiDAR surveys (Johnson et al 2021, Tyner and Johnson 2021) by Festa-Bianchet et al (2023) measured methane emissions from over 900 single-well CHOPS sites in Saskatchewan, finding emissions ∼3.9 times higher than reportedly vented. This finding was incorporated into the inventory using production volume- and site count-scaling approaches, two common methods employed elsewhere in the federal inventory (Clearstone Engineering Ltd 2014, ECCC 2022d, 2023). For production-scaling, since Festa-Bianchet et al (2023) found CHOPS sites were emitting 7% more methane than they reportedly produce, CHOPS site emissions can be scaled up 7% from reported methane production. For active count-scaling, the average CHOPS site emission rate (10.9 kg h⁻¹) can be applied to each CHOPS site.

Since the aerial LiDAR survey measured all on-site methane emissions above its detection threshold (Conrad et al 2023), any other reported/estimated methane emissions on-site (flare, fuel use, fugitives) were removed when applying the updated emission estimates. Finally, because CHOPS wells are not explicitly distinguished from other heavy oil wells in the reported data, we follow a similar definition to Festa-Bianchet et al by considering heavy oil wells 'CHOPS' if they reported sand dispositions (MER 2018).

To support previous findings of underestimated CHOPS methane, and to test the validity of our CHOPS-updated inventory, we conducted regional aerial methane emission measurements in Saskatchewan. Partnering with Scientific Aviation (Boulder, Colorado, US), regional methane emissions were measured using an aerial mass balance approach (Conley et al 2017) for a 40 km by 40 km CHOPS region near Lloydminster, Saskatchewan. The survey made two separate measurements of the area, performing six laps each on Sept. 13th and 22nd, 2022, following the designated flight path (see section 4.4) at different altitudes up to the top of the boundary layer.

Instruments were mounted inside the trunk space of a fixed wing single-engine Mooney M-20 manned aircraft. A Picarro G2301 gas analyzer (Picarro, Inc., Santa Clara, CA) measured in-situ methane concentrations at 1 Hz frequency at a flow rate of 0.2 SLPM from a 1/8 in. OD (3.175 mm) stainless steel sampling probe installed beneath the right wing. A standard aircraft pitot tube measured wind speeds. Flux estimates were based on Gauss's theorem (Conley et al 2017). The methane emission flux divergence integral was calculated for each altitude bin and then summed across all bins to obtain regional methane emission rates. Resulting measured emission rates had contributions subtracted for livestock, solid waste, wastewater, coal mines, and wetlands (Bloom et al 2017, Scarpelli et al 2022); see SI section S3 for details.

Figure 2 shows our estimate of Saskatchewan's upstream oil and gas methane inventory for 2012–2022. Unfortunately, it is difficult to validate this estimate against the published federal inventory since the latter aggregates upstream, downstream, and end-use estimates in Saskatchewan. Nevertheless, our upstream methane estimate represents between 74% and 86% of the federal estimate of Saskatchewan's oil and gas sector (ECCC 2023). Importantly, our inventory estimate indeed suggests that Saskatchewan's upstream sector has surpassed the federal target of a 40%–45% reduction in methane from 2012 levels (47%); similarly, Saskatchewan has seemingly surpassed its associated gas reduction target of 40%–45% from 2015 levels (67%; see figure S2).

Zoom In Zoom Out Reset image size

Inventoried reductions are overwhelmingly from reduced venting. Reported venting in 2022 dropped to a third of pre-regulation (2019) levels, driven by reductions at light/medium oil and CHOPS sites (see figure S6). Methane loss rates can be derived for Saskatchewan by normalizing emitted methane by produced methane, knowing produced gas volumes (Petrinex 2023) and average gas compositions (ECCC 2023), although this excludes emissions not directly tied to produced gas (e.g. SCVF, abandoned wells). Since 2012 levels, the province's overall methane loss rate of 12% rose to ∼16% in 2018 before reducing to 9% in 2022. Despite Saskatchewan's produced energy having decreased by 25% since 2012 (see figure S5), associated gas production—the dominant source of reported emissions—increased 4% over that same period, implying that reductions in total venting are due to improved loss rates.

Inventoried methane reductions were also achieved from LDAR but to a much lesser extent. For 2020–2022, LDAR reduced estimated fugitive methane by only ∼4–5 kt per year (1%–2% of the inventory). In 2022, the 858 LDAR-requiring sites constituted 91% of gas batteries, plants, and gas gathering systems, but less than 4% of gas wells. More broadly, this means that LDAR surveying was required at 4% of all upstream facilities and ∼1% of all wells in Saskatchewan. By comparison, LDAR in Alberta, Canada, was required at 85% of (non-mined oil sands) upstream facilities and 26% of wells (Seymour et al 2022). Importantly, this assumes full compliance, which has been an issue in other jurisdictions (BC Oil and Gas Commission 2021, AER 2022).

The federal inventory experienced a shift in methodology beginning in 2020, removing federal government estimates for pneumatics, storage losses, etc (see table 1). Instead, such emissions would thereafter be reported by operators under existing vent, flare, and fuel use categories (Petrinex 2019, MER 2019a, 2020a). Unfortunately, it is not possible to confirm that these emission sources are reliably reported by operators since the different types (e.g. venting, pneumatics, storage losses) are all reported to a generic 'vent' category—an issue likely to complicate regulatory development (UNEP 2022). Although Seymour et al (2022) found that the same methodology shift in Alberta likely overstated emission reductions by ∼20% since 2012 levels, these emission categories represent only ∼9% of the Saskatchewan inventory (∼70% in Alberta).

While these inventoried emission reductions appear promising, it is important to consider that many of these operator-reported emissions may still rely on simple engineering estimates or generic emission factors (MER 2019a). Even where operators are conducting measurements, this cannot be distinguished from emission factor-based estimates in the reported data.

Here, we evaluate operators' reported compliance with fleet-level emission limits and calculate the penalties levied against exceedances. For operators with fleet-level emission limits, emissions and produced gas reports were grouped by site ownership and production region, and CO₂e emission limits were calculated for each (Government of Saskatchewan 2020b). Penalties were then assessed after recombining regional data; the findings are summarized in table 2.

In 2022, 8 of the 36 (22%) operators exceeded their respective emission limits, totaling 72 kt CO₂e or an estimated 1.6 kt of methane (0.7% of the 2022 inventory) and a penalty of CA$2.2 million. This is seemingly an improvement from 2020, in which 31% of noncompliant operators emitted an excess 166 kt CO₂e (3.9 kt methane; 1.4% of the 2020 inventory).

Although the current inventory method suggests that Saskatchewan has achieved its emission reduction commitments, some emissions remain underestimated. Figure 3(a) shows the updated inventory (forecasted to 2025) updating emission estimates for CHOPS well sites; figure 3(b) shows the relative reductions since each estimate's respective 2012 baseline level. The figure shows that—considering underreported emissions from CHOPS only—Saskatchewan's upstream oil and gas methane emissions are generally at least 30% higher than currently estimated. However, figure 3(b) shows that Saskatchewan may nevertheless reach its relative reduction target (although the production-scaled CHOPS update may only achieve a 39% reduction). In terms of Saskatchewan's CO₂e target, the province will likely meet the target (see figure S3).

Zoom In Zoom Out Reset image size

While these conclusions are encouraging, the spread of relative reductions using different scaling methods in figure 3(b) illustrates the inherent uncertainty in updating the inventory using a survey conducted only in 2021. No doubt this uncertainty will increase as additional measurement data are included in the inventory, as is expected in Canada's next national inventory report (ECCC 2023, OAGC 2023). Despite Saskatchewan's possible success in achieving the 40% reduction from 2012 levels, the updated inventory suggests that CHOPS will have emitted an additional 1.7 or 2.3 Mt of methane between 2012 and 2022, cumulatively. This significant underreported methane could have, for example, provided heat for an estimated ∼100 000 households in Canada over the same 11 year period.

Interestingly, the two updated inventories stay closely aligned, only beginning to diverge for inventoried years far away from the aerial LiDAR survey year (2021). Divergence is not uncommon when extrapolating measurement data over a long period as changes to on-site technology and/or production practices (Husky Energy Inc. 2020) cannot be captured by a short-term survey. Clearly, it will be necessary for measurement surveys to be conducted frequently to minimize the need for measurement extrapolations and their associated uncertainty.

The updates to the inventory show the outsized impact CHOPS has on Saskatchewan's upstream methane emissions. Emissions from light/medium oil sites were previously the largest emission source (often ∼50% of the inventory; see SI section S5), but after correcting for underreported CHOPS emissions, CHOPS emissions are the largest emitting site type. The implicit risk is that CHOPS might have been overlooked as a source of significant and cost-effective emission reductions during the development of current regulations (Tyner and Johnson 2018). Although the federal government did consider the impact of higher-than-inventoried emissions (Government of Canada 2018b), it is unclear whether sufficient emphasis was placed on CHOPS. Clearly, robust measurement, reporting, and verification (MRV) are needed to ensure both the emission magnitudes and source distributions are properly assessed, a critical aspect when designing future regulations (UNEP 2022, ECCC 2022c).

We conducted two aerial mass balance measurements over a CHOPS-dominant region near Lloydminster, Saskatchewan, with non-upstream emissions subtracted. These top–down measurements are compared with the inventory estimates for the region in figure 4. The region contained—based on our site type definition (section 3.2)—552 single-well CHOPS sites, 412 other single-well heavy oil sites, five multi-well CHOPS batteries (20 on-site wells), and 34 gas wells. Measured emission rates were ∼2.8 times higher than Saskatchewan's current inventory would suggest, indicating that CHOPS emissions continue to be underreported despite being a known issue since at least 2016 (Johnson et al 2017). However, our CHOPS-updated inventory estimates for the region—also included in figure 4—agree with the regional measurements within uncertainties. This supports the idea that such inventory updates from aerial LiDAR surveys could be used in the current federal inventory at least until more comprehensive inventory methods can be implemented (e.g. Johnson et al 2023a).

Zoom In Zoom Out Reset image size

Underestimated emissions are likely to evade the control/mitigation required under current regulations. If, however, these emissions can be reliably quantified and reported, and operators meet their emission limits, the current regulations are expected to provide much deeper emission reductions than if underreporting is allowed to continue (i.e. figure 3). To compute possible emission levels, we add the estimated CHOPS emissions derived in sect. 4.3 to each operator's total and reduce them as needed to comply with their respective fleet-level limits. This is repeated for each year between 2023 and 2025 using both CHOPS-updated inventory estimates (production- and site count-scaled) to illustrate achievable emission levels under current regulations. See SI section S6 for a sensitivity analysis of these results in scenarios where CHOPS gas production is also assumed to be underestimated.

Figure 5 compares the current inventory trajectory (with underreported CHOPS emissions) with the scenario where CHOPS emissions are accurately measured/reported and subsequently controlled following current regulations; figure S4 shows the same scenarios for Saskatchewan's provincial CO₂e target. Figure 5(a) shows that full compliance and accurate reporting could reduce emissions by at least 50 kt of methane annually (a ∼15% decrease in emissions from the current trajectory), achieving relative reductions of at least 54% from 2012 levels in figure 5(b). To achieve this, however, improved MRV requirements and full operator compliance are needed.

Zoom In Zoom Out Reset image size