EcoNav project developing user tools to support more ecologically responsible travel behavior and decisions
YASA Motors named UK’s Best Enterprise in the Lloyds TSB/Telegraph Enterprise Awards

Study finds plausibly high volumes of Canadian oil sands crudes in US refineries in 2025 would lead to modest increases in refinery CO2 emissions

An analysis of the US refining sector, based on linear programming (LP) modeling, finds that refining plausibly high volumes of Canadian oil sands crudes in US refineries in 2025 would lead to a modest increase in refinery CO2 emissions (ranging between 5.4% to 9.3%) from a 2010 baseline, depending upon the supply scenario. The estimates account both for increased oil sands crude in the refinery crude slate and the imposition of new fuel sulfur specifications, which require additional refining effort.

That percentage increase would be equivalent to approximately 0.014−0.024 gigatonnes per year, David Hirshfeld and Jeffrey Kolb conclude in a paper published in the ACS journal Environmental Science & Technology, noting that emissions associated with end-use combustion of refined products exceed by an order of magnitude those associated with refining. [The open access Supporting Information (SI) file accompanying the paper includes a useful tutorial on crude oil, refining, and refinery energy use.]

The estimates of US refinery CO2 emissions developed in this analysis are consistent with recent estimates of the life cycle emissions associated with Canadian bitumen crudes and are an order of magnitude less than the recent estimate [Karras 2010, earlier post] that processing heavy oil or bitumen blends could increase CO2 emissions by 1.6−3.7 gigatons/year. The latter estimate appears to reflect, in part, the assumption that the entire refining sector processes nothing but high density, high sulfur crudes to the exclusion of all other crudes, domestic or imported.

—Hirshfeld and Kolb

Refinery LP models
Refinery LP models embody an analytical framework that computes energy use by summing the direct energy inputs to each refining process (fuel, steam, and power); selects the least cost combination of internal and external energy sources available to the refinery; and estimates refinery CO2 emissions by applying standard carbon emission factors to each of the energy sources.
This framework includes electricity and natural gas used in the production of hydrogen.
It does not include energy used in the production and transport of biofuels; energy used in the production and supply of purchased blendstocks; electricity used in non-process or off-site activities (such as oil movements, product blending, lighting, etc.), and incidental energy losses due to flaring, fugitive emissions, etc.

Hirshfeld and Kolb are the principals of MathPro Inc., a small consulting firm specializing in technical and economic analysis of the petroleum refining industry and related industries. They use a proprietary refinery modeling system (ARMS), which comprises a linear programming model of refining operations, a collection of crude oil assays, operating data on refining processes, and a database of refinery configurations. Clients include government agencies, petroleum companies, automobile companies, chemical and biofuel companies, professional service firms and law firms. The study published in ES&T was funded by Chevron Energy Technology Company.

Refinery LP models are detailed, process-oriented representations of refinery operations and economics; the models are the method of choice for assessing the technical and economic aspects of the refining sector’s responses to changes in operational requirements, such as a major change in crude slate, the authors note in their paper.

The refinery energy accounting in the model is based on process-by-process energy use data obtained from the open literature and from refinery process licensors. The refinery CO2 accounting uses standard, published carbon intensity factors for the various refinery fuels.

...some observers assume that the US crude oil slate is set to shift precipitously and completely to heavy, high-sulfur crudes, and they assert that this shift would lead to steep increases in refinery energy use and CO2 emissions.

The latter view is implausible. As shown in a recent analysis, the estimated per-barrel refinery energy use associated with Western Canadian bitumen crudes is comparable to that of some conventional crude oils already in the US crude slate. More importantly, it is not clear where the assumed large volumes of very heavy, high-sulfur crude would come from or how these volumes could drive existing domestic and imported crudes out of the US market. Running an all heavy, high sulfur crude slate in refineries configured to process lighter, lower sulfur crudes would require substantial construction of new refinery processing capability.

Credible forecasts of increases in the density and sulfur content of the US crude slate (coupled with more stringent regulations governing refined product quality) imply modest increases in per-barrel energy consumption and CO2 emissions from US refineries. These per-barrel increases may be offset by decreases in refined product demand and refinery crude runs resulting from mandated increases in the use of alternative fuels, improved fuel economy of the vehicle fleet, and improvements in refinery energy efficiency driven by prospective state and Federal limits on CO2 emissions (e.g., California’s Global Warming Solutions Act).

—Hirshfeld and Kolb

Four aspects of refining operations must be explicitly addressed in delineating the relationship between crude slate properties and refinery energy use and CO2 emissions, the authors stated: complexity, limited flexibility, sensitivity to crude oil and product properties, and distributed energy use.

For their study, Hirshfeld and Kolb established a set of plausible crude oil supply scenarios in 2025 and estimated energy use, CO2 emissions, aggregate crude oil throughput, and process capacity requirements for each. Each scenario includes forecast product demands and compliance with mobile and stationary source air quality standards expected in 2025, including tighter sulfur specifications on gasoline (Tier 3) and marine diesel fuel (MARPOL Annex VI sulfur standards on marine diesel fuel).

In 2010, US crude imports comprised:

  • About 1.8 MMBbl/day from Western Canada (about 33% conventional crudes, 51% bitumen crudes, and 16% light synthetic crude oils (SCO) produced by upgrading bitumen crudes. Bitumen crudes are delivered in the form of synbit and dilbit—mixtures containing bitumen crudes and diluents (light hydrocarbons, conventional crudes, and/ or SCO).

  • About 5.1 MM Bbl/day from Caribbean and Atlantic Basin (CAB) sources: Eastern Canada, Mexico, South America, and West Africa. Of the CAB imports, about 2.6 MM Bbl/day (51%) were light and medium crudes and 2.5 MM Bbl/day (49%) were heavy crudes. On average, the CAB imports were heavier but lower in sulfur than the rest of the world (ROW) imports.

  • About 2.2 MM Bbl/day from more remote sources in the ROW. Of the combined CAB and ROW imports, about 4.7 MM Bbl/day were light and medium crudes, and about 2.6 MM Bbl/ day were heavy crudes.

In building their four scenarios to represent a reasonably aggressive range of heavier and higher sulfur crudes in the US crude slate, Hirshfeld and Kolb incorporated (i) increased domestic crude oil production and (ii) decreased total US crude oil consumption, as forecast for 2025 in the AEO 2011 Reference Case, in the scenarios. Aside from the baseline scenario, the scenarios each incorporate an additional 2.4 MM Bbl/day of Western Canadian crude oils, corresponding to the optimistic production forecast by the Canadian Association of Petroleum Producers (CAPP). The scenarios are:

  • S1. Baseline scenario equivalent to 2010.

  • S2. An additional 2.4 MM Bbl/day of Western Canadian in the US, displacing all crude oil imports from ROW sources—regardless of crude type—and some from CAB sources. Average properties of the remaining CAB imports are unchanged from 2010.

  • S3. The additional 2.4 MM Bbl/day of Western Canadian crudes displaces only light and medium crude oils from CAB and ROW sources. Average properties of the remaining light and medium crudes and of heavy crudes imported from CAB and ROW sources are unchanged from 2010.

  • S4. The additional 2.4 MM Bbl/day of Western Canadian crudes displaces all crude from ROW sources and a small amount from CAB sources, as in S2. At the same time, the volume share of heavy crudes imported from CAB increases to 75% (from 49%), resulting in a significant increase in the density of the imported crude oil slate.

Among their findings:

  • In the baseline scenario (S1), estimated US refinery energy use and CO2 emissions in 2025 are slightly higher than the reported values in 2010, reflecting increased requirements for desulfurization and hydrogen production capacity, as well as the offsetting effects of the forecast reduction in US refinery crude runs and the slight decrease in domestic crude oil density and sulfur content indicated in EIA’s Reference Case for 2025.

  • In all three other scenarios, estimated refinery energy use and CO2 emissions are higher. Estimated increases in refinery energy use range from 3.7% to 6.3%; estimated increases in CO2 emissions range from 5.4% to 9.3%. These estimates reflect the combined effects of changes in crude oil density and sulfur content and finished product sulfur specifications.

  • Tighter sulfur specifications on gasoline and marine diesel fuel account for about 10%−20% of the estimated increases in refinery energy use and about 25%−30% of the estimated increases CO2 emissions. The effects of tighter sulfur specifications increase with increasing crude slate density and sulfur content.

  • Estimated total US refinery crude runs, energy use, and CO2 emissions all increase incrementally with average density and sulfur content of the crude slate. The more intensive refinery processing entailed by higher crude density and higher sulfur content leads to some reduction in refinery processing yields (and hence increased crude volume) in these scenarios. This effect accounts for the increased requirement for crude distillation capacity.

  • Processing the higher density, higher sulfur crude slates in scenarios S2, S3, and S4 calls for significant additions to refinery processing capacity, most notably distillation, coking, and hydrogen production. The estimated investment requirements for these three processes alone would range from about $11 billion in S2 to about $19 billion in S4 ($2010, U.S. Gulf Coast location).

Crude slate density and sulfur content are not the sole predictors of refinery energy and CO2 emission intensity. As the results of this analysis indicate, refinery energy use and CO2 emissions are also related to the stringency of regulations governing finished product properties, such as sulfur content.

Changes in refinery energy and CO2 intensity do not, in themselves, imply corresponding increases in emissions of criteria air pollutants and hazardous air pollutants. Refinery emissions of these substances are governed by environmental regulations and standards that are independent of the refinery crude slate.

Individual refineries are configured to process a specific type of crude oil slate. Once its configuration has been established, a refinery’s crude purchase options are limited. A refinery configured to process light and/or sweet crudes cannot efficiently or profitably process heavy and/or sour crudes. A refinery configured to process heavy crudes cannot process light crudes without surrendering the economic benefits of its investments in heavy crude processing capability and leaving some capacity idle. Consequently, a refinery’s configuration allows one to infer the general type of crude slate the refinery is designed to process.

Finally, assessments of refinery energy and CO2 intensity should employ established LP modeling tools and methods, and not econometric models or regression analysis. Such methods cannot capture the complexity of refinery operations and economics and therefore do not yield useful information.

—Hirshfeld and Kolb

The authors also noted that two factors not reflected in their analysis could further reduce future refinery energy use and CO2 emissions, regardless of crude slate: continued improvements in overall refinery energy efficiency; and further reductions in the volume of hydrocarbon-based gasoline and diesel fuel, in response to possible future mandates further increasing the biofuels content of US gasoline and diesel fuel.


  • David S. Hirshfeld and Jeffrey A. Kolb (2012) Analysis of Energy Use and CO2 Emissions in the US Refining Sector, With Projections for 2025. Environmental Science & Technology doi: 10.1021/es204411c



Who financed this study? The recent EU study found an increase of 17+%.


That's without adding the multiple extraction pollution.


right , most of the extra CO2 with the Tar Sands is on the extraction side but also in the cracking of nat gas to make H2 and transform these tars in oil, because Tar is not oil, you need to add H2 to make it what they call a Syncrude.


Not to worry - the delay/cancellation of the Keystone oil pipeline will save more CO2 than this creates.


Since the XL pipe would leave the pollution in Houston while the refined products went to the rest of the world, they can now look at the real culprits.

Coal fired power plants are the number one offender. Not only CO2, but sulfur and mercury. If we could make them IGCC, we would clean up the mess and make synthetic fuels as well.

Co locate them near ethanol plants and use the waste heat to ferment and distill the mash, saving more natural gas and coal combustion. Lots of things we COULD do, but none of them get done.

Thomas Pedersen

Syncrude is what arrives at the refineries. The large additional CO2 emission takes place before the refineries.

According to the conference presentation I have seen about oil from tar sands, syncrude is designed to have roughly similar properties to regular crude in terms of viscosity and C/H ratio. This is also necessary to be able to pump it from Alberta to the Gulf coast...

Since tar sands have very little hydrogen in them, they have to be upgraded with massive amounts of hydrogen derived from natural gas, being the primary source of local CO2 emissions at the extraction plant. Steam reforming of natural gas to yield H2 + CO2 also gives loads of waste heat, which presumably is used for heating the tar sands.


Extra CO2 is only one problem with using tarsands "oil." They've discovered that when you pipe the stuff it not only leaks, it leaks more often and is harder to clean up when it does - sometimes costing up to 10X as much to remove.

Rachel Maddow did a bit about this, around 7 min into this clip;


One more thing, it is a political argument to day to say that if the Keystone project was build it could drive the price of gas down. It is the opposite in fact, because to day Syncrude is bought at a discount price because north US is the only customer, bringing the Syncrude in the gulf will open to more customer and the discount will go away and the gas at the pump will be higher, this has been well documented.


"if the Keystone project was build . . gas at the pump will be higher".

That's increduble; in the most literal sense of the word.


I know it is counter intuitive but most people don't understand the complexity of the oil market these days


Greens have gone wild on tar sands simply to slow down (they cant actually stop it) energy production.. they will bring up all kind of false stuff to achieve this.. to save the planet.

Originally many Gulf refineries were setup to handle the heavy oil from Venezuela, since Chavez took over it has been an oil production disaster, it has dropped by 30%.. plus Chavez does not want to sell it to the US and is instead courting the Russians and Chinese.. so now the refineries are starving for product and guess where there is a huge supply of heavy oil?.. Canada.

The pipeline infrastructure is stressed in the middle of the country and thus there is a glut of oil building up, depressing oil prices there without the ability to send the oil to the gulf or to the east coast refineries.. many refineries in the east coast are already shutting down for good and hopefully it wont happen to the gulf refineries.. if that happens we will be forced to import foreign gasoline (often refined with little care for pollution) and drive the price up for US drivers.. and this is the real reason for the XL pipeline debacle.

Luckily plans are being made to deliver the tar by railroad, at an additional cost of $5 per barrel or so. Railroad infrastructure is adaptable and can grow quickly.


Right now Alberta and Dakota oil can not get to any other market than refineries in Montana, Wyoming and other local areas. The price of gasoline is cheaper there because the product has to be sold to pay investments.

Once the XL pipeline is in place, they can sell to a world market at a higher price. This is why gasoline prices in Colorado, Idaho, Montana and Wyoming will go up if the pipeline is installed.

They are building a section of pipeline from Oklahoma to Houston to get some of the stored oil to refineries. That will not help the people in New York, where there are NO refineries at all.

The three refineries in Pennsylvania are being shut down because of the high cost of oil transported to those refineries and low capacity utilization in the region.

The U.S. tends to take a company over country view. There are some issues and times that call for national interests over corporate profits. I hope this is one of those times when people can discuss the issues rationally for the long term benefit of ALL citizens.


Yeah... good luck with that. Obama tried to and he got labeled as the anti-christ.


The wrong wing will label him the devil if it helps their greedy self interests.


You know, I use to be a "Right" winger, and then I noticed something;


Of the top 80 of the largest US refineries, 28 are in Texas and Louisiana and only one is in all of Montana, Minnesota, North Dakota and Wisconsin. They don’t need much crude in those states.

Not laying Keystone is like restricting every exported Prius and Leaf to Hawaii, so that the price in Hawaii stays low – idiotic.

Of the top 142 largest US refineries, 64 are in Texas and Louisiana and only13 are in all of Montana, Minnesota, North Dakota and Wisconsin combined.

You say; “They are building a section of pipeline from Oklahoma to Houston to get some of the stored oil to refineries. That will not help the people in New York, where there are NO refineries at all. “

No kidding. You want to pipe the oil to where there are no refineries?

Where do you get these weird ideas? Rachel Maddow?

“The three refineries in Pennsylvania are being shut down because of the high cost of oil transported to those refineries and low capacity utilization in the region.”

OK, so what? You want to build a pipeline to there too?
And did you know; Shell Oil Co. has chosen a site near Pittsburgh for a major new multi-billion $ petrochemical refinery.


For Alberta to get a more decent ($15/bar higher) price for for their oil, access to more buyers are required, in order of priority:

1. a new large diameter Trans-Mountain pipeline to open the very large Asian market. This should be given the highest No. 1 priority. China has offered to finance this project and has already bought three large producers. The offer includes the construction of an adjacent new highway. PM Harper is re-visiting Asia to accelerate and finalize this project.

2. the XL pipeline to Texas to replace crude from Venezuela and Mexico. This is not primarily for the USA market but to better use existent Texas refineries and to export refined products.

3. extension of the Trans-Canada pipeline network to Montreal and Portland Maine. The Portland to Montreal pipeline exist and the flow could be reversed.



I think you like to argue. I would like to get fuel to people that need it, that is the point.

The comments to this entry are closed.