Changing US crude imports are driving refinery upgrades

Aug. 10, 2009
In the last 15 years, overall crude oil imports to the US have grown at 3%/year.

In the last 15 years, overall crude oil imports to the US have grown at 3%/year. During this time, the sources of these imports and their proportional shares have changed significantly, with a corresponding movement toward crudes that are heavier and more sour in average quality.

The steady growth in crude oil imports to the US from Canada has been accompanied by significant investments in oil sands production in Canada, development of pipeline infrastructure to transport these crudes to the US, and refinery expansions in the US to enable processing of heavy crudes. These modifications to refining infrastructure, which result in increased requirements for hydrogen, steam, and power, are opportunities for gasification and cogeneration projects.

This article discusses historical trends in US crude oil imports with an emphasis on synthetic Canadian heavy crudes and examines their impact on the configuration and profitability of US refineries.

Import trends

Since 1994, imports of foreign crude oil to the US have consistently exceeded its domestic oil production with a rapidly increasing margin (Fig. 1). A study of data compiled by the US Energy Information Administration shows that during this period, imports of foreign crude oil have increased at 3%/year to more than 10 million b/d in 2007 from 7 million b/d in 1994.1

This timeframe has also seen the steady rise of Canada to the largest country of origin for imported crude oil to the US in 2007 (19% of total imports) from the third largest in 1994 (14%).2 EIA data also show that the imports from such other large suppliers as Saudi Arabia and Venezuela have stayed relatively flat during this period, while imports of North Sea crude have decreased steadily.

Since 2000, crude oil imports from Nigeria have been on the rise, while imports from Mexico peaked in 2004 and have begun trending down since. These trends underlie a remarkable growth in the share of total crude imports: The US imported about 66% of its crude feedstock in 2007 in contrast to about 25% in the mid-1980s.3

These trends in the quantity and origin of imported crude oil to the US have been accompanied by a trend of deteriorating quality of the crude oil processed in US refineries (Fig. 2), namely toward heavier crude oil with a greater sulfur content. This trend is expected to continue due to the decreasing availability of light sweet crude and the relative price discount for heavy sour crudes.

According to EIA, the average sulfur content of imported crude to the US 1985-2005 increased to 1.4% sulfur from 0.9%, while the average API gravity has declined to 30.2° API from 32.5° API as heavier crudes have been imported.4 Processing the increasing quantity and decreasing quality of imported crude oil while meeting increasingly stringent specifications on the allowable sulfur content in gasoline and diesel fuel has wrought modifications to US crude oil refining infrastructure over these years.

In 2004, Canada overtook Saudi Arabia to become the largest exporter of crude oil to the US and has consistently strengthened its position since.5 US imports of crude oil from Canada for 2007 totaled 1.9 million b/d and have increased at 5%/year since 2000.6

Supporting this trend are large project investments in Alberta to increase production of oil sands-derived heavy crude (including bitumen blends), development of pipeline infrastructure to transport these crudes to the US, and investments in refinery projects in the US Midwest to process oil sands-derived heavy crude.

Reduced credit availability, however, increasing market uncertainty, and oil price volatility due to the current economic downturn have temporarily slowed development of current and proposed Canadian heavy crude and downstream refining projects.

Infrastructure developments

Imported crude to the US arrives predominantly by crude oil tankers to various locations along the Gulf Coast and the East and West coasts. The US Gulf Coast currently receives about 55% of US crude oil imports, the majority of which is processed in Gulf Coast refineries, while the remainder moves inland to distribution hubs via the Seaway and Capline crude pipelines.

Crude oil imports via tanker to the East and West coasts are generally intended for regional refinery consumption and represent about 12% and 10%, respectively, of total US imports (or about 1 million b/d each). The remaining US crude imports are mainly from Canada, which currently accounts for about 1.9 million b/d delivered to the upper Midwest and north central US by an extensive network of pipelines.

Fig. 3 shows the average crude oil import flows in 2007 to US Petroleum Administration for Defense Districts and highlights the major countries of origin of the imported crude.

A predominant trend in the recent decade with respect to crude oil imports into the US has been the growth in supply of heavy Canadian crude, in particular from the oil sands region in the western province of Alberta. In addition to investment in upstream production and downstream processing of heavy Canadian crude oil and bitumen blends, several projects are under way further to expand existing pipeline infrastructure. The softening of market demand due to the current economic downturn has temporarily moderated the need for additional pipeline capacity beyond the existing projects.

The Canadian Association of Petroleum Producers annually provides a detailed overview of existing and planned pipeline infrastructure to transport heavy Canadian crude to US refining markets.7 According to the 2009 CAPP market update,8 around 1.7 million b/d of crude oil from Western Canada moved to US refineries in 2008 via three major pipeline systems: the Enbridge system and Kinder Morgan’s Express and Trans Mountain pipelines.

The largest among these pipelines, the Enbridge system, has an estimated crude capacity of 1.9 million b/d (1.2 million b/d heavy crude) and a trunkline that runs from Edmonton to Chicago in the US Midwest.7 Additional segments in the Enbridge system deliver crude to Cushing, Okla., Patoka, Ill., and Ontario. In March 2006, the 66,000 b/d ExxonMobil-Enbridge Pegasus pipeline was reversed, connecting Patoka to Nederland, Tex., thereby providing Canadian heavy oil producers access to the US Gulf Coast market.7

Kinder Morgan’s Trans Mountain pipeline runs west from Edmonton and delivers predominantly light Canadian crude to refineries near Burnaby in Canada and the state of Washington, while its Express pipeline transports Canadian heavy crude from Hardisty, Alta., south to refineries in the US mountain states of Montana, Wyoming, and Utah.

The lines have estimated crude capacities of 280,000 b/d and 300,000 b/d, respectively, for a total crude capacity of nearly 580,000 b/d.

In Casper, Wyo., the Express pipeline connects with the Platte pipeline that runs due east to Wood River, Ill.7 The existing pipeline infrastructure for Canadian crude, including those discussed thus far, appear in Fig. 4 as solid black lines.

Several projects have been proposed and are under way to expand the distribution of heavy Canadian crude further in the US, with an estimated 1 million b/d of pipeline capacity anticipated by 2011.7

Key projects include construction of TransCanada’s 435,000 b/d Keystone line from Hardisty to Wood River and Patoka by 2010, with a subsequent extension to the US Gulf Coast by 2013. A major expansion of Enbridge’s main trunkline, the Alberta Clipper project, is under way from Edmonton to Superior, Wisc., as well as an extension southwards from Superior to Flanagan, Ill., which is labeled the Southern Access Expansion and Extension project.

Extensions of the Enbridge system to the US East Coast and to the US Gulf Coast have also been proposed, the latter delayed by a short-term outlook for oil sands development due to rising costs and regulatory delays. Enbridge has proposed an interim project to reverse its pipeline sections between Portland, Me., and Sarnia, Ont., which would enable shipment of Canadian heavy crude oil down the US Atlantic Coast to the US Gulf Coast.9 Discussions are ongoing about the timing and market viability of the proposed project.

Enbridge also has a project under way to transport diluent from Flanagan to Edmonton in order to address the demand for condensate among Canadian heavy crude producers.7 Ongoing and proposed Canadian crude pipeline projects appear as dashed blue lines in Fig. 4.

In August 2008, Enterprise Products Partners LP, TEPPCO Partners LP, and Oiltanking Holdings Americas Inc. announced a joint venture to develop a massive offshore crude oil receiving terminal labeled the Texas Offshore Port System (TOPS) in order to meet increasing demand for US crude imports, in particular at regional Gulf Coast refineries of which some are slated for major expansions.

In April 2009, Enterprise and TEPPCO announced their dissociation from the project citing disagreements with Oiltanking.10 Despite this development, Oiltanking remains interested in developing the proposed terminal, which has an anticipated capacity of 1.5 million b/d of crude (nearly 15% of current US crude oil imports) and was scheduled to start up in late 2010.

In January 2008, Sempra announced plans to develop a 500,000-b/d marine terminal near Port Arthur, Tex., for crude oil, LPG, and refined products.11 In October 2008, Valero signed a memorandum of understanding with Sempra with the intention of becoming a major recipient of crude oil at the proposed terminal and also to assist with marketing efforts to third parties and develop connecting pipelines from the terminal.12

These announcements, as well as the ongoing expansions of Canadian heavy crude pipelines, underscore the increasing US dependency on foreign crude oil and represent essential infrastructure developments that are needed to keep pace with the anticipated long-term growth in demand for crude oil imports.

Effects on US refining

The steady increase in crude oil prices between 2000 and mid-2008 has been a key demand driver for the increased use of heavy Canadian crude and bitumen blends in US refineries, as refiners sought to take advantage of the substantial price differentials between heavy and light crude oils. With the increasing emphasis in the US on the security of energy supply, the stable political relations with Canada have served to support this trend.

Increasing investment in production in the oil sands region and pipeline infrastructure to transport Canadian heavy crude to the US Midwest has been accompanied by announcements of refinery expansion projects in the US Midwest to process heavy Canadian bitumen blends. These projects include Marathon’s Detroit refinery (80,000 b/d heavy oil processing capacity after expansion),13 BP’s Whiting, Ind., refinery (260,000 b/d heavy oil),14 and ConocoPhillips’s refineries in Borger, Tex., and Wood River (total of 550,000 b/d heavy oil consisting of 275,000 b/d bitumen).15 While these projects are still under way, a few have been scheduled for a later completion in response to current market conditions.

In 2007, Husky Energy, a Canadian energy company and oil sands producer, purchased Valero’s Lima, Ohio, refinery with the intention of ultimately modifying it for heavy Canadian crude.16 In December 2007, Husky Energy also announced an alliance with BP that included a retrofit of BP’s Toledo, Ohio, refinery to run on bitumen.17 Hyperion Resources is going through the permitting process as a part of its plan to develop a $10 billion, 400,000-b/d grassroots refinery in South Dakota based on heavy Canadian crude.18

In 2008, Valero announced that it has secured capacity on the proposed Keystone XL pipeline extension to the US Gulf Coast and would be expanding and modifying its Gulf Coast refineries to process heavy Canadian crude in 2012.19

These developments highlight the growing market for Canadian oil sands-derived crude and bitumen in the US refining industry.

The modifications required to enable a refinery to process heavy Canadian crude will depend on several factors, such as existing refinery configuration, characteristics of the heavy crude to be processed, and desired refined product mix. For simplicity, this article will focus on bitumen blends, particularly dilbit, which continues to be of great interest to the US refining industry due to its relative price discount to conventional light crude. Dilbit is a mixture of bitumen diluted with light naphtha (or condensate).

It is noteworthy that due to condensate shortages in Alberta, the market share of other bitumen blends and synthetic crudes such as synbit (bitumen diluted with synthetic crude) and Western Canadian Select (WCS) has increased slightly in recent years, but the development of condensate-return pipelines from the US to Alberta such as the Enbridge Southern Lights project may serve to reverse the trend.

While it is technically viable and proven to upgrade bitumen in Alberta to synthetic crude oil for subsequent transportation to US for refining, it is worth noting that the economics of transporting the bitumen directly to US refineries (that are near or connected to Canadian crude pipeline systems) for upgrading and processing can be competitive and potentially superior.

A major modification is typically required to enable a refinery to process bitumen blends such as dilbit. The first step to consider is the recovery of diluent from the bitumen blend before processing, which is an economic decision based on the market value for the condensate and the infrastructure available to transport it to the end market, be it for reuse as a diluent for bitumen or as a petrochemical feedstock. Diluent recovery from dilbit is technically straightforward and accomplished with atmospheric distillation.

The upgrading of bitumen to lighter product streams that are amenable to conventional refinery processing is the central step in adapting a refinery to process dilbit. Two predominant upgrading approaches have been pursued commercially: delayed coking and resid hydroprocessing. Both approaches involve the use of a vacuum column on the front end to recover light fractions, resulting in a heavy vacuum residuum that is then upgraded.

US refiners that have announced modification projects to process heavy Canadian crude have favored the delayed coking approach to bitumen upgrading. This can be largely attributed to the fact that delayed coking is commercially proven and widely used in US refineries and has a lower perceived technical risk.

In this approach, the vacuum residuum is processed in a coker unit to recover the lighter fractions that can be fed to conventional refinery units; the by-product is solid petroleum coke (“pet-coke”). This pet-coke is often stockpiled but is a good feedstock for gasification to produce hydrogen, steam, and power for the refinery.

Refinery expansion projects that use the delayed coking route to bitumen processing typically require a subsequent hydrocracking step to upgrade the heavy coker gas oils into distillate-range products. This typically involves construction of a new hydrocracking reactor (unless spare hydrocracking capacity is available), which also gives the refinery more operating flexibility with regard to production of distillate vs. gasoline. Fig. 5 provides a simplified overview of the typical refinery configuration changes that can be expected when the delayed coking route to bitumen upgrading is selected.

In the resid hydroprocessing approach, the heavy vacuum residuum is typically upgraded in two steps: a severe hydrocracking step in an ebullated bed that cracks the residuum into lighter fractions, and a hydrotreating step that reduces the sulfur and nitrogen to acceptable levels.

The resid hydroprocessing bottoms are typically sent to a solvent deasphalting unit to remove insoluble asphaltenes. The asphaltenes stream is typically not conducive to further refinery processing and is either sold as or blended into low-end refined products or ideally gasified to produce hydrogen, steam and power for refinery use. The gasification option is especially attractive due to the significant hydrogen requirements for resid hydroprocessing as well as for desulfurization of the refined product streams and is discussed later in this article.

The resid hydroprocessing route has been chosen for upgrader projects in Alberta, but it is yet to be selected for bitumen upgrading in US refinery projects.

Refinery units that are typically added or expanded as a part of a refinery expansion project in order to enable the processing of heavy Canadian crude are:

  • Diluent recovery unit.
  • Crude and vacuum units.
  • Delayed coker.
  • Hydrocracker.
  • Hydrogen plant(s).
  • Distillate hydrotreater.
  • Naphtha hydrotreater.
  • Sulfur plant.

The addition and expansion of these units will result in an increase in the Nelson’s complexity index for the refinery, which is a measure of its ability to convert heavy crudes into light products and generally translates to increased refinery profitability.20 The upgraded refinery will have increased requirements for utilities such as hydrogen, steam, and power, which can be generated on site or imported over-the-fence, depending on the relative economics.

These additional utility requirements, however, will increase the greenhouse gas footprint of the refinery due to the incremental CO2 emissions from fuel combustion to meet the increased power and steam requirements and from increased process CO2 emissions when steam-methane reformers are used to satisfy the incremental hydrogen demand.

Recent regulatory2122 and legislative23 developments in the US in support of curbing GHG emissions from manufacturing facilities may eventually pose a barrier to future refinery expansions, particularly those based on heavy crude.

The use of cogeneration to increase the efficiency of steam and power generation can help offset a portion of the incremental GHG from refinery expansion projects. Cogeneration units can be integrated with either conventional hydrogen plants or gasification units to realize additional efficiency benefits.

The possibility of the integrated, on site generation of hydrogen, steam, and power is particularly attractive due to the potential for gasification of the residuals from the refinery processing of heavy Canadian crudes and is described in more detail in the following section.

Gasification

The increasing need for hydrogen, both to process heavier crude oil and bitumen with greater sulfur content and to produce cleaner fuels, is a key driver for petroleum refining-based gasification systems. Other factors include the increasing uncertainty in natural gas prices, reducing the generation of waste, and improving efficiency.

As of January 2008 the US Environmental Protection Agency considers gasification to be a production or manufacturing operation rather than a hazardous-waste management activity,24 while in April 2009 EPA proposed a new rule for mandatory greenhouse gas reporting.21

Gasification is the chemical conversion of any carbonaceous fuel into a mixture of carbon monoxide and hydrogen known as synthesis gas through an exothermic reaction of the fuel with oxygen or steam at elevated temperature.

Additional hydrogen can be recovered through a water-gas shift reaction of the carbon monoxide with steam leaving a stream of concentrated CO2. Another option is to use some or all of the synthesis gas and steam produced for electric power generation.

Commercialization of the gasification process began more than 50 years ago with most applications supporting the production of chemicals or liquid fuels.25

Several vendors, such as Shell, General Electric, ConocoPhillips, and Sasol-Lurgi, provide commercial-scale gasification technologies,26 and hundreds of commercial gasifiers currently operate around the world.27 All the commercial gasification technologies recover heat from the gasifier through steam production.

Gasification of refinery by-products of low or negative value allows conversion of these by-products into hydrogen, steam, or electric power. In many cases, a concentrated stream of CO2 is produced that can be captured for potential use, such as in enhanced oil recovery or for sequestration.

Petroleum coke or asphaltenes produced during the upgrading of bitumen as well as coke and heavy residuals from other refining units are well suited for gasification. Due to its ability to process refinery streams as feedstock and generate products that can be consumed in refinery units, a gasification unit naturally lends itself to integration within a refinery.

To date, several US refineries currently operate gasification units, including those in Delaware City, Del.; Baytown, Tex.; and El Dorado, Kan.25 Although a discussion of how petroleum refining fits into the GHG puzzle is provided in more detail elsewhere,28 the potential integration of gasification is a promising avenue to manage the GHG footprint in refineries that process heavy crude oil.

References

  1. “Annual US Crude Oil Imports,” Energy Information Administration, July 2008.
  2. “Crude Oil Imports into the US by Country of Origin,” US Census Bureau, December 2008.
  3. Wisner, Robert, “US Crude Oil Production, Use, and Import Trends,” AGMRC Renewable Energy Newsletter, September 2008.
  4. “Changing Trends in the Refining Industry,” Issues in Focus (2006 Annual Energy Outlook), Energy Information Administration, 2006.
  5. Boll, Theodore, “Canadian Oil Sands: A New Force in the World Oil Market,” a study released by the Joint Economic Committee, June 2006.
  6. “Annual US Crude Oil Imports from Canada,” Energy Information Administration data, July 2008.
  7. “Crude Oil Forecast, Markets, and Pipeline Expansions,” Chapter 4: Crude Oil Pipelines, Calgary: Canadian Association of Petroleum Producers, June 2009.
  8. “Crude Oil Forecast, Markets, and Pipeline Expansions,” Section 3.2, p. 11, Calgary: Canadian Association of Petroleum Producers, June 2009.
  9. “Enbridge delays new pipeline project, opts for East Coast Tankers,” Edmonton Journal, July 10, 2008.
  10. Oiltanking GMBH; www.oiltanking.com, May 20, 2009.
  11. “Sempra Energy Proposes Gulf Coast Marine Terminal and Storage Facility,” Sempra press release, Jan. 9, 2008.
  12. “Sempra Energy and Valero Energy announce plans to develop Texas marine terminal project,” Sempra press release, Oct. 14, 2008.
  13. “Marathon begins Construction on Detroit Heavy Oil Upgrade Project,” press release on detroithoup.com, June 20, 2008.
  14. “BP Plans $3 Billion Project to Refine More Canadian Heavy Crude Oil in the US Midwest,” BP press release, Sept. 20, 2006.
  15. “ConocoPhillips and EnCana to Create Integrated North American Heavy Oil Business,” ConocoPhillips press release, Oct. 5, 2006.
  16. “Husky Energy To Acquire Lima Refinery From Valero Energy Corporation,” Husky Energy Inc. press release, May 2, 2007.
  17. “Husky Energy and BP Announce Integrated Oil Sands Joint Development,” Husky Energy Inc. press release, Dec. 5, 2007.
  18. “Hyperion air quality permit hearings July 15-16,” Sioux City Journal.com, July 9, 2009.
  19. “Valero to Receive Canadian Crude Oil for Its Gulf Coast Refining System,” Reuters, July 16, 2008.
  20. Nelson refinery complexity is based on a series of articles by W.L. Nelson first published in Oil & Gas Journal in 1960-61: Mar. 14, p. 189; Sept. 26, p. 216; and June 19, p. 109. He elaborated on the concept in another series in 1976: Sept. 13, p. 81; Sept. 20, p. 202; and Sept. 27, p. 83.
  21. US Environmental Protection Agency Proposed Rule, Part 98—Mandatory Greenhouse Gas Reporting, Federal Register, Vol. 74, No. 68, Apr. 10, 2009.
  22. US Environmental Protection Agency Proposed Rule, 40 CFR Chapter 1—Proposed Endangerment and Cause or Contribute Findings for Greenhouse Gases, Federal Register, Vol. 74, No. 78, Apr. 24, 2009.
  23. American Clean Energy and Security Act of 2009, (also known as the proposed Waxman-Markey Clean Energy Bill), HR 2454, May 15, 2009.
  24. “Gasification,” US Environmental Protection Agency, September 2008.
  25. Marano, J.J., “Refinery Technology Profiles: Gasification and Supporting Technologies,” US Department of Energy National Energy Technology Laboratory, Energy Information Agency, June 2003.
  26. “Gasification, World Database 2007: Current Industry Status, Robust Growth Forecast,” US Department of Energy National Energy Technology Laboratory, October 2007.
  27. “Gasification Database,” US Department of Energy National Energy Technology Laboratory, September 2007.
  28. Gunaseelan, P., Buehler, C., and Chan, W.R., “In Profile: Carbon Dioxide Emissions from US Petroleum Refining,” Paper No. 2009-A-615-AWMA, Air & Waste Management Association 2009 Annual Conference and Exhibition, Detroit, June 16-19, 2009.

The authors

Praveen Gunaseelan([email protected]) is a manager in Exponent’s Engineering Management Consulting practice. Before joining Exponent, he was an energy market analyst in the Tonnage Gases Division at Air Products & Chemicals specializing in the refinery hydrogen market. Gunaseelan holds a PhD in chemical engineering from Purdue University, is a member of the Society of Petroleum Engineers and the Gas Processors Association, and has served on the hydroprocessing committee of the National Petrochemical & Refiners Association.
Christopher [email protected]) is a senior managing engineer in Exponent’s Thermal Sciences practice. Buehler holds a BS in chemical engineering from Villanova University and an MS and PhD in chemical engineering from Purdue University. He is a registered chemical engineer in Texas and a member of the American Institute of Chemical Engineers, American Chemical Society, and National Fire Protection Association.