For the past few months, I’ve been digging into the implications of Canada’s recently announced $170 CAD per ton of CO2 carbon price cap, something the country will move toward annually in $10 and $15 increments before reaching it in 2030. One focus has been Alberta’s electrical grid, where the higher carbon price will radically change the economics of natural gas vs renewables. Another is a national look at heat pumps, where it’s clear both that just putting heat pumps in the worst buildings would have a major impact, and that the carbon price makes natural gas and oil heating much less appealing to the people paying the bills.
A Canadian NGO reached out to me due to my string of analyses and publications, and suggested I consider the impact on electric vehicle ownership costs. One of the interesting questions that arose was how to forecast the price of gasoline. Naturally, this is an arcane art at the best of times given the variety of challenges, but obviously the carbon price itself is going to be applied to the retail cost of gasoline. Each liter of gasoline when burned emits CO2, after all.
As an odd side note, the Canadian government has a difference of opinion with itself about what numbers to use for the grams of CO2 per liter of gasoline, with the carbon pricing using 2.21 kg of CO2 per liter burned in internal combustion cars and NRCan — and many others — providing a figure of 2.3 kg. It’s only a 4% difference, but it does appear to be an unnecessary discount on an otherwise solid policy.
That’s not really going to break the bank. In Canada, the average car gets 8.9 liters per 100 km, at least as of 2017, a bit over 26 miles per gallon. That means that the average car will travel a little over 11 kilometers on a liter, so the tax equates to under a cent a kilometer. And, of course, these are Canadian cents. Even in 2030, that’s only about 3.4 cents a kilometer.
The intent, obviously, is to put in place a gently increasing price of driving inefficient internal combustion cars so that when Canadian families decide to replace their cars, they’ll consider more efficient cars, including electric ones.
However, that’s not all the carbon price will apply to.
As this diagram shows, the gasoline that is pumped out of the gas pump is at the end of a long chain of carbon-intensive activities. All of the portions of those activities which occur in Canada will be subject to some variation of carbon pricing as well.
As I wrote recently here, analysis shows that Alberta’s oil sands upstream activities and part of midstream activities by themselves produce as much CO2 as the entire commodity market for the gas globally every year.
If we want to project the potential rise in the price of gas, in other words, we want to consider the carbon price’s impact on upstream, midstream, and refinery emissions.
There’s another aspect to this that we had to consider: not all gasoline in Canada is created from the same sources at the same refineries. There’s a significant difference between eastern and western consumption, and in provincial level treatments of oil and gas emissions.
Western Canadian refineries process Alberta’s oil sands crude with its relatively high CO2 emissions per barrel. Eastern Canadian refineries process a lot more lighter oil and imported oil, which has an impact too.
There are three categories of gasoline emissions prior to the pump to consider. The first two are heavy and light crude, and the last is whether it is imported or domestic. That’s important because, at least right now, Canada doesn’t have a carbon border adjustment where tariffs are applied to products from countries which don’t price carbon. The implication for gasoline prices is that as the carbon price rises, the upstream and midstream emissions aren’t taxed, so imported oil doesn’t increase in price. It’s being discussed, and you can be sure that Canadian oil and gas companies that sell domestically want it in place.
It’s imperfect, but that suggests the provincial mixes for their gasoline supply look something like this:
Now we get into the nitty gritty: how much CO2 should we consider for each liter of oil from upstream, midstream and refining?
Let’s start with heavy and light crude, as they are fairly clearly delineated in the Carnegie Endowment for International Peace’s Oil-Climate Index (OCI). Its results are more credible than the Canadian energy war room’s misleading numbers. The Pembina Institute did some analysis of the numbers and say:
“The Carnegie Endowment’s Oil-Climate Index suggests most oilsands crude is associated with 31 per cent more emissions than the average North-American crude from the point of extraction through its lifecycle to the point of end use”
It’s imperfect, as not all crude is turned into the same products with the same emissions, but adequate for the purposes of our analysis. Combustion of gasoline in an internal combustion engine is always the same, so the variance is from upstream, midstream, and refinery emissions. Pembina puts the North America weighted average at 541 kg of CO2e per barrel of oil, and the oil sands average at 709 kg per barrel of crude.
A 2018 study, Global carbon intensity of crude oil production, provides a useful filter, as OCI’s numbers don’t separate refinement from use.
“Producing, transporting, and refining crude oil into fuels such as gasoline and diesel accounts for ∼15 to 40% of the “well-to-wheels” life-cycle greenhouse gas (GHG) emissions of transport fuels.”
This aligns with OCI’s numbers and suggests an easy way to deal with the emissions intensity before burning of gasoline, which is to apply the 15% to 40% numbers to the 2.3 kg of emissions from burning the gasoline. While oil sands crude has improved somewhat, I’ll assume it’s in the 35% range given the recency of study, and as the North American weighted average is higher than the bottom, I’ll use a 20% figure for that.
This suggests that for oil sands heavy crude, that there are 0.81 kg of emissions prior to the 2.21 kg of emissions that the Canadian carbon price attributes to the retail liter, and 0.46 kg of emissions for light crude.
We now have almost all the numbers to figure out what the carbon tax will do to the real price of gasoline as it increases. The last wrinkle is Alberta’s Technology Innovation and Emissions Reduction Regulation, or TIER. It’s an alternative carbon market for ‘heavy industrial emitters’ in Alberta, which is to say the oil sands. It allows providers who cut emissions to earn credits and emitters who don’t meet the benchmark to buy credits. It’s an approved variance under the federal carbon tax, but it too will be forced to scale rapidly given the $170 carbon price.
I’d calculated the impact of the federal carbon price on the cost of a barrel of Alberta’s primary product at $9 USD on top of the $40 USD average they were fetching on the market, or almost 25%. Given TIER, it’s less than that. For our calculations, we’ll assume that the end cost to oil sands crude is 75% of the increasing carbon price. That’s not only the relatively low cost of TIER per barrel, but the money that they have to spend to achieve emissions reductions to maintain social license and not be penalized by TIER.
The combination suggests that the discount that oil sands crude gets still puts it slightly above onshore light oil in terms of the impact of the carbon price, but does clearly give us the uplift for both per liter of gasoline that will find its way to consumers.
We’re now in a position for the final assessment, which is the real additional impact of the carbon price per liter of oil, both in the base purchase cost where the carbon debt of creating and delivering it sits, and in the burning of it in the car. The lack of the carbon border adjustment becomes clear here, as provinces which don’t use domestic oil are paying only about 10% of the total carbon debt, for refinement and final distribution alone.
A few things were interesting to me upon finishing this analysis.
The first was that the oil sands were clearly being given a substantial break by Alberta and Canada on their high emissions, something I found unsurprising, but yet still disappointing. I suspect my analysis of the cost implications for them is overstating the actual impacts, and would welcome feedback from people directly familiar with the application of TIER to the oil sands.
The second was that offshore oil was being given a much bigger break at present, as upstream and midstream emissions are not carbon taxed due to the lack of carbon border adjustment. All else being equal, this could lead to shifting sources of petroleum for gasoline in Canada to imported in Ontario and eastward. This benefits neither Canada nor the climate, so obviously isn’t a good outcome. The recent meeting between President Biden and Prime Minister Trudeau included clear statements of alignment on climate policy and protection from unfair competition from foreign suppliers in countries with weaker policies, so this is more likely to be addressed in the next two years than not.
Third, the variance across provinces is interesting and unlikely to be represented as cleanly in retail prices. Cross-border shoppers for cheaper gas cause adjustments, and neighboring towns with substantially different gas prices is a political headache. More importantly, the federal carbon price doesn’t apply directly to gasoline sold in Quebec, Nova Scotia, the Northwest Territories (not represented in the model), and British Columbia, as these provinces and territories have province-level carbon and pollution pricing systems deemed sufficiently stringent by the federal government. However, just as it is reasonable to expect increasing prices on carbon from oil sands emissions, it is reasonable to expect that these provinces will increase their carbon pricing roughly in step with the federal increments.
It will be interesting to see how actual gasoline prices track with this analysis as we approach 2030. Given that 40 to 50 cents CAD is around a third the price of a liter of gasoline today, that is a significant increase and should drive behaviors. But it will also drive consumer awareness of the cost transactionally, while rebates are only quarterly. Politically, this is not necessarily a solved equation.
All of this is to say that someone who owns an electric car in 2030 will be saving hundreds of dollars per year. The economics of electrics keep getting better, especially in jurisdictions that are implementing good climate policies.