Stationary Combustion to Extract Oil and Gas: Emissions Calculations

Greenhouse Gas Reporting Program (GHGRP)

Facilities that emit more than 10 kt ${\mathrm{C}\mathrm{O}}_2\mathrm{e}$ annually are required to report annual emissions of each of the seven greenhouse gases via Canada's Greenhouse Gas Reporting Program (GHGRP), and have been required to do so since 2004. Since 2022, such facilities have furthermore been required to report their emissions in terms of the reason, or source, such as on-site transportation, waste, leakage, or — of particular relevance to the emission estimate developed for this post — stationary fuel combustion. The GHGRP component of this sector's estimator for years 2022 and 2023 was formed by adding the stationary fuel combustion emissions from facilities whose NAICS6 codes started with 2111, which meant any of "Oil and Gas Extraction (except oil sands)", "Conventional Oil and Gas Extraction", "Non-Conventional Oil Extraction", "In-situ oil sands extraction", and "Mined oil sands extraction". These category labels overlap because the nomenclature changed in 2017: the two-way Conventional vs. Non-conventional distinction was replaced with the three-way differentiation of conventional extraction vs. in-situ vs. mined oil sands.

For years prior to 2022, the GHGRP-reported emissions are not listed by emission source. Facility totals are reported back to 2004, but emissions from stationary combustion prior to 2022 are not listed separately from e.g. on-site transportation, waste, leakage, or any of the other reasons. Since this sector is only supposed to be stationary combustion, a backfilling heuristic was implemented to estimate the stationary combustion emissions of these prior years 2004-2021. The heuristic was to look up the facility's stationary combustion emissions as a proportion of total emissions in 2022, and then apply that same proportion to the facility's total emissions from each of the prior years. Some facilities were no longer operating in 2022; for these, the proportion was assumed to be 80%. Future work could use data to estimate this proportion from data on similar facilities (github issue).

Using this method of counting, GHGRP-registered emitters in the business of oil and gas extraction contributed over 90 Mt ${\mathrm{C}\mathrm{O}}_2\mathrm{e}$ in recent years from stationary combustion alone. This 90Mt is about 85% of the emissions from this IPCC sector, and over 10% of the entire national GHG emissions total.

Petrinex

Petrinex is used in both Alberta and Saskatchewan, but is used differently in the two provinces. The databases records have slightly different semantics.

Petrinex: Alberta

Alberta's Petrinex data makes all upstream oil and gas facilities' monthly movements of material (by Product ID) and by activity (Activity ID) available for public download. Facility names and addresses are in the data files, as well as their type and subtype. Alberta used four facility types as of January 2023:

• battery — a single accounting unit corresponding to a well or group of wells, co-located physical tanks, and related equipment;
• gas gathering system — a set of pipelines, pumps, and compressors that aggregate gas from gas batteries;
• gas plant — a facility for e.g. separating and sweetening the gas received from a gas gathering system;
• injection/disposal facility — a site that primarily disposes of product from the network, rather than e.g. re-selling it.

For the "Small facilities (Alberta)" component of this sectoral estimator, I added up the monthly Petrinex records relating to "FUEL" for all of these facility types. Monthly totals were combined into annual totals, and the annual totals were multiplied by the emissions factors from the NIR (annex 6, Tables 1-2 and 1-3), and then the standard GWP multipliers to get a single annual total in ${\mathrm{C}\mathrm{O}}_2\mathrm{e}$.

Alberta's Petrinex data covers some facilities that also appear in the GHGRP data, but each system has its own set of identifiers: GHGRP lists facilities by GHGRP Identifier, and Petrinex lists facilities by Petrinex identifier. The boundaries of a single "facility" aren't necessarily the same for Petrinex and the GHGRP, and the best alignment table I found was the undocumented API hosted by Pollution and Waste Canada that's intended for the National Pollutant Release Inventory data search. Cross-referencing facilities by their pollution registry ID accounted for most GHGRP facilities, but 71 remained un-accounted-for. In order to skip over about 71 facilities I ignored Petrinex facilities in Alberta and Saskatchewan that emitted over 5 kt ${\mathrm{C}\mathrm{O}}_2\mathrm{e}$/mo. The result of this filtering procedure was to filter about 60-70 facilities from each monthly data dump that were not already identified as being GHGRP-registered facilities. Facilities that emit even 1 kt ${\mathrm{C}\mathrm{O}}_2\mathrm{e}$ monthly are required to register with the GHGRP, so these facilities emitting over 5 kt should certainly have been registered; I hope these >5kt facilities are indeed the ones that have registered as GHGRP facilities and and I just don't have their corresponding Petrinex IDs. Details of this heuristic can be found in the source code of est_nir.py and petrinex.py.

Petrinex: Saskatchewan

Saskatchewan's Petrinex data is similar to Alberta's in structure, but not identical. Saskatchewan uses a taxonomy of six facility types instead of the four used in Alberta:

• battery — a single accounting unit corresponding to a well or group of wells, co-located physical tanks, and related equipment;
• gas gathering system — a set of pipelines, pumps, and compressors that aggregate gas from gas batteries;
• gas plant — a facility for e.g. separating and sweetening the gas received from a gas gathering system;
• custom treating — a mid-stream hub that processes oil-water-sand emulsions on behalf of nearby wells; Alberta's Energy Regulator might classify the whole set of these things as a single battery, perhaps a "large multiwell proration battery"
• tank terminal — a facility for loading crude oil onto rail cars
• fresh formation water source — a facility that extracts ground water for e.g. steam or hydraulic fracturing

The "Small facilities (Saskatchewan)" component of the sectoral estimator was formed by adding all "FUEL" records in Saskatchewan data for all of these facility types, and ignoring facilities with emissions >5kt for the same reason as this was done for Alberta. For most months, the heuristic to avoid double-counting skipped 6-10 facilities that were, I hope, the ones that actually had GHGRP identifiers, but for which I didn't know the corresponding Petrinex IDs.

Summary

Stationary combustion emissions from the oil and gas sector itself are enormous, and rising. In recent years, this has become Canada's largest single reporting sector. Although it is based on different methodology from the NIR (which uses data from Statistics Canada) the estimator developed here is very accurate, at least for recent years. (I hope it's accurate for the right reasons!) This estimator is based on fuel rows from Petrinex facility reporting data, and emission factors from the NIR annex 6.

Critical Success Factors

To reduce stationary combustion emissions from oil and gas extraction, it seems required to do some combination of the following:

• Reduce the rate of oil and gas extraction (reduce demand, reduce price, or cap production)
• Increase the rate of exhaust capture to long-term ${\mathrm{C}\mathrm{O}}_2$ storage.
• Reduce emissions related to process heat for e.g. creating steam, upgrading bitumen to synthetic crude oil, drying gas, and maintaining fluidity in cold weather.
• Reduce emissions related to mechanical energy, powering e.g. compressors and pumps.

Barriers

It is challenging to make progress on the critical success factors above for several reasons:

• Alignment: Oil and gas continue to see enormous, and still-rising, market demand at prices well above the break-even point of oil and gas extraction operations in Canada.
• Geopolitics: Comparable oil and gas operators in the US are scaling up under policies of the Trump administration.
• Geopolitics: Canadian federal ambition is to position Canada as an energy superpower by exporting fossil fuels.
• Lock-in: Extraction facilities, especially for oil sands, are high-capital (billions) that require decades of operation to pay off.
• Cost: Carbon capture technologies exist, but it's costly and unproductive (carbon capture and utilization is less-unproductive, so to speak). The cost for deploying this technology must be borne by some combination of customers, investors, and citizens of various levels of government, without benefit.
• Cost: electric components exist that could replace e.g. gas-powered ones, but they are not cheaper to buy or operate.
• Maturity: The energy efficiency of oil and gas extraction has been improved by decades of engineering. For example, the heat required to melt bitumen to oil already sometimes comes from a co-generation process in which gas turbines generate electricy and their heat exhaust melts the oil. Not only does this power the site's operations, such sites also sell electricity via the provincial grid.
• Scale: Emissions in this sector are over 100Mt/yr, and rising. The sector involves thousands of facilities and connecting pipelines, especially in western provinces. For perspective, the carbon capture and storage projects in Alberta have been operating at just 1-2 Mt/yr since 2016.

Strategies

The most net-zero-aligned strategy to reduce emissions in this sector is to reduce production, because of course the produced oil and gas is mostly destined to be burned as fuel and release greenhouse gases. Production may be reduced either by reducing demand, reducing price, or capping production via regulation. Canada's output-based pricing system (aka industrial carbon tax) reduces the take-home price for the products from facilities whose emissions are higher than others producing the same product; International and domestic electrification may reduce demand eventually; and capping production has been tried, and proved politically infeasible so far.

That said, it should be possible to reduce the emissions associated with oil and gas extraction, without reducing the quantity of extracted oil and gas. Indeed, if less gas is used to drive extraction equipment, perhaps even more gas could be brought to market. Maybe energy sources such as firewood, sunlight, wind, nuclear, or even alternative non-combustion energy extraction methods from fossil fuels, such as the pyrolysis of methane, could power extraction operations. These alternate energy sources are, at least in principle, capable of providing process heat and mechanical energy. Maybe it can even be cost-effective, especially if some operations could be intermittent, following the availability of e.g. wind and solar power. Research into the feasibility and viability of such alternative energy sources for oil and gas extraction is left for future work.

Conclusion

This post has introduced a relatively accurate estimator of emissions from Stationary Combustion Sources related to the extraction of oil and gas, at least for recent years. The IPCC / Stationary Combustion / Oil and Gas Extraction page has been refreshed to feature this estimator, as well as updated critical success factors, barriers to those success factors, and a handful of as-yet-undeveloped strategies to address those factors and barriers. I had hoped this post would be quicker to put together because I had already incorporated the key data sources, but it wasn't because I ended up using a different version of Petrinex, dealing with missing data, and I had a lot to learn about oil and gas extraction. The next post in this series replicating the NIR will look at a more familiar source of emissions to the general public: residential stationary combustion sources, namely heating. I think it will be a fun one!

Until then,

- James Bergstra