Newsfront: A Surge in U.S. Refining Capacity  

U.S. petroleum companies are investing in new domestic refining capacity on a scale not seen in decades. Over the next five years the industry will add more than one million bbl/d of distillation capacity (see table), according to the U.S. Department of Energy’s Energy Information Administration (EIA, Washington D.C.; www.eia.doe.gov).

"It’s a step change in expansion," says Rick Boyd, business development executive with Celerant Consulting (Lexington, Mass.; www.celerantconsulting.com). The main driver, he says, is fuel prices, which have not only enriched petroleum companies, but have improved the social climate for the industry. "Back in the 1990s it was difficult to get permits to expand and the public didn’t realize there were supply pressures. But now, with gasoline selling for $3 – 4 a gallon, people are more aware of the petroleum industry and regulators are more inclined to allow expansions."

Current U.S. refining capacity is approximately 17.6 million bbl/d, which is insufficient to meet the demand for fuels. For example, imports of gasoline blending components and finished motor gasoline have been averaging about 30 million bbl/mo in recent months.

As it happens, gasoline is not a major issue, says Joanne Shore, a senior analyst with EIA’s Petroleum Division. She notes that gasoline is available at relatively low prices from Europe, which has a gasoline surplus as Europeans continue to switch to diesel-powered vehicles. Also, U.S. demand for gasoline is expected to decline over the next 15 years because of government regulations that require higher fuel efficiency for light-duty vehicles and higher use of biofuels. At the same time, the demand for distillate will continue to grow, says Shore, "and this will be a challenge for refiners."

The biggest change facing refiners for the next few years will be an increasing use of heavy oils, particularly from Alberta oil sands. Consequently, much of the new capacity is being designed to upgrade heavy oil for the production of gasoline, diesel fuel and other fuels (see box on p. 22). Also, the growing demand for distillate is reflected in the amount of hydrocracking capacity that is being added, compared to the relatively modest capacity of fluid catalytic cracking (FCC) capacity for gasoline production (see table).

Other issues are: a new federal regulation that calls for the ethanol content of gasoline to be increased; the integration of biofuels into the refinery feed mix, and a federal rule that calls for the benzene content of gasoline to be reduced to 0.6% by January 2011. Looming in the background is a law that would limit carbon dioxide emissions, still under discussion in Congress.

Heavy Oil Upgrading

Bitumen, the product of oil sands, typically has a specific gravity of less than 10 API and has a high content of sulfur, metals and total acid number (TAN). Coking followed by hydrotreating is the common way to upgrade heavy oil. These processes may be done in the field or at a refinery.

If upgrading is done in the field, the product is a synthetic crude (syncrude) of 25 – 30 API that can be shipped via pipelines. Otherwise, bitumen is treated in one of two ways to make it suitable for pipeline transmission. One is to dilute it with light syncrude to make "synbit" and the other is to mix it with a diluent (light ends) to make "dilbit." The dilbit method requires two parallel pipelines: one to ship the product and the other to recycle the diluent after the customer has separated it from the bitumen.

A new heavy-oil upgrading process that is said to achieve up to 100% conversion of the heaviest feedstock, thus avoiding the need for coking, has been developed by Chevron Corp.’s Technology Center (Richmond, Calif.; www.chevron.com). "We can put in heavy oil of less than 10 API gravity and produce a blend that is mostly gasoline, jet and diesel fuels," says James Murphy, business manager for the technology. In contrast, he says, conventional refining technology achieves less than 80% conversion.

Called Vacuum Resid Slurry Hydrocracking (VRSH), the process has been extensively piloted and will be tested in a pre-commercial plant at Chevron’s refinery in Pascagoula, Miss. Construction of the 3,500-bbl/d plant is scheduled to start later this year.

FIGURE 1. Pipeline expansions will allow Canadian crude to flow to more areas of the U.S.

Heavy oil or vacuum resid is slurried with a proprietary catalyst, mixed with hydrogen and circulated through several reactors at temperatures ranging from 775 to 850°F and pressures of 2,000 – 3,000 psig. A small amount of catalyst is removed continuously through a slipstream and subsequently reactivated and returned to the process. Murphy adds that the cost of VRSH is expected to be similar to that of Lummus Technology’s LC-Fining hydrocracking process for heavy oil and residue, which is licensed by Chevron Lummus Global, a partnership.

A process for upgrading heavy oil that had not been actively marketed in the past because of low crude and product prices has been revived by UOP LLC (Des Plaines, Ill.; www.uop.com), which has bought the rights from Natural Resources Canada (Ottawa). UOP has improved the process and is offering it for license. "At today’s crude and product prices the economics are attractive," says Dan Gillis, business manager for heavy oil.

In the slurry hydrocracking process, a base metal catalyst is mixed with heavy oil or vacuum resid and the slurry is fed into the bottom of an upflow reactor. Hydrogen is added in the reactor and more than 90% of the feed is converted to distillate and naphtha, which exit the top of the vessel. The reactions take place at 1,800 – 2,000 psig and 800 – 880°F.

A hydrotreating process whose applications include mild hydrocracking of FCC feed, treating of ultra-low-sulfur diesel fuel and severe gas oil hydrotreating is offered by DuPont Stratco Clean Fuel Technologies (Leawood, Kan.; www.dupont.com). An advantage of the IsoTherming process is that H₂ is dissolved in the liquid feed before the liquid enters the reactor, thereby avoiding a high volume of H₂ in the vapor phase. IsoTherming uses standard catalysts and operates under standard hydrotreating conditions, but is said to be less expensive than conventional hydroprocessing (for more details, see CE, October 2007, p. 13). Valero plans to install an IsoTherming unit in its Paulsboro, N.J., refinery, where it will further the company’s goal of producing more distillate for ultra-low-sulfur-diesel fuel.

Hydrogen

The rapid growth in heavy-oil processing is providing a bonanza for hydrogen suppliers. As an illustration, Air Products (Lehigh Valley, Pa.; www.airproducts.com) is now starting up its third hydrogen plant in Canada. Located near Edmonton, Alta., the 105-million-scfd plant is interconnected with a 71-million-scfd plant that started up in 2006. The two plants will serve Petro-Canada’s Edmonton refinery, plus several additional customers. Air Products’ third plant, of 80 million scfd, started up in Sarnia, Ont., in 2006 and supplies two nearby refineries. All use natural gas-based steam-methane reforming.

Stephen Losby, general manager for Air Products in Canada, estimates that about 1 billion scfd of hydrogen is now being produced in Alberta for petroleum processing and expects the demand will grow to 2.5 – 3 billion scfd when syncrude production reaches 3 – 4 million bbl/d. At the same time, he estimates that the U.S. will install up to 1.0 billion scfd of additional hydrogen capacity.

The prospect of having to use a lot more hydrogen has prompted a search for a less-expensive source than steam methane reforming (SMR) of natural gas. Since many refineries have or are installing cokers, one possibility is coke gasification.

The capital cost of a coke-gasification plant is about 2.7 times that of an SMR plant, says Dale Simbeck, vice-president of the consulting firm SFA Pacific, Inc. (Mountain View, Calif.; sfapacific.com). However, with coke priced at $1/million Btu versus $6.60/million Btu for natural gas, plants using either process would have the same hydrogen cost, assuming a five-year payback on the investment. At higher natural gas prices, or after the payback of the investment, coke gasification can have a cost advantage, he says.

Simbeck adds that the economics of coke gasification require large scale and spare capacity to assure a steady supply of hydrogen. The economics can also be improved by polygeneration. For example, an installation might consist of three gasifiers, with two producing hydrogen and a standby unit that is used for cogeneration when it is not needed for hydrogen production.

Gasification also offers flexibility in terms of feed and products. Shell Global Solutions (US) Inc. (Houston; www.shell.com), for instance, offers two variations of gasification technology — one for solid feeds such as petroleum coke, coal and biomass, and the other for liquids such as oil distillates and refinery residues. The processes have been installed in several refineries and make a variety of products, including hydrogen, syngas for fuel, ammonia, methanol and oxo-chemicals.

A pressure-swing-adsorption (PSA) process for recovering hydrogen from process streams that is said to provide a mass-transfer rate up to 100 times higher than that of conventional PSA is being marketed jointly by ExxonMobil Research and Engineering Co. (Fairfax, Va.; www.exxonmobil.com) and QuestAir Technologies Inc. (Burnaby, B.C., Canada; www.questairinc.com). Called Rapid Cycle PSA, the process employs structured adsorbent rather than beads and uses two rotary valves to switch gases between adsorbent beds at up to 100 cycles/min. The installed cost is said to be 30 – 50% lower than that of conventional PSA. A demonstration unit that was installed at an ExxonMobil affiliate refinery in France ( CE, December 2007, p. 13) went into commercial operation in March.

Biofuels

A controversial issue in the U.S. refining industry is the Energy Independence and Security Act of 2007, which mandates that 9 billion gal of renewable fuels (mostly ethanol) be blended into the transportation fuel supply this year, increasing to 36 billion gal in 2022. In a statement before the Senate Energy and Natural Resources Committee in February, Charles Drevna, president of the National Petrochemical & Refiners Association (NPRA, Washington, D.C.; www.npra.org) protested that it was doubtful enough ethanol would be available to meet the near-term requirement.

Drevna also pointed out that most vehicles cannot use fuel blends greater than E-10 (10% ethanol, 90% gasoline) because ethanol is corrosive, but such blends may be needed to meet the mandate by 2010. The only vehicles able to use these blends are E-85 vehicles, he said, but these represent only 6 million out of more than 240 million registered vehicles.

As an alternative to ethanol, which is simply blended with gasoline, some organizations are developing processes that produce fuels compatible with petroleum-derived products. The goal is to integrate bioprocesses into a conventional refinery, says Jennifer Holmgren, UOP’s director of renewable energy and chemicals.

UOP and Eni S.p.A. (Milan, Italy; www.eni.it) have jointly developed a process in which vegetable oils and animal fats are converted to a "green" diesel that can be blended with petroleum-derived diesel fuel. The UOP/Eni Ecofining process will be commercialized in 2009 – 2010 in refineries in Italy and Portugal. In each case the Ecofining plant will contribute 100 million gal/yr to the diesel fuel pool, says Holmgren (For details on the process, see CE, May 2007, p. 18).

Within the next two years UOP expects to commercialize a process that will convert cellulosic waste to gasoline, diesel and jet fuels. The company’s partners in the development are the National Renewable Energy Laboratory (NREL, Golden, Colo.; www.nrel.gov) and the Pacific Northwest National Laboratory (Richland, Wash.; www.pnl.gov). The partners have piloted the process, in which waste is subjected to fast pyrolysis to obtain pyrolysis oil, which is then upgraded to transportation fuels via UOP hydroprocessing technology (Figure 2).

FIGURE 2. In UOP’s proposed pyrolysis oil refining scheme, several pyrolysis units would supply a central bio-refinery

FIGURE 3. The BenZap process from GTC Technology removes essentially all the benzene from reformate

A drawback of bio-oil is that it contains 10 – 40% oxygen, versus essentially none for petroleum, as well as a high percentage of water. Holmgren says that UOP has developed ways to take out both the oxygen and water, but declines to give details except to say the oxygen is removed by hydroprocessing, followed by a second hydroprocessing step to obtain fuel.

Benzene

Refiners are preparing to meet a new regulation from the U.S. Environmental Protection Agency (EPA; Washington D.C.; www.epa.gov) that calls for the benzene content of all gasoline to be reduced to an average of 0.62 vol.% by Jan. 1, 2011. At present the limit is 1% for reformulated gasoline only. In addition, refiners will have to meet a maximum average benzene standard of 1.3 vol.%, effective July 1, 2012.

Reformate from catalytic reforming accounts for about 50 – 75% of the benzene in the gasoline pool, so a popular way to reduce the benzene content is to fractionate the naphtha feed to the reformer and remove benzene precursors (C₆ paraffins, cyclohexane and methylcyclopentane). However, this method is not expected to be sufficient for the new standards.

A new process called BenZap that is said to reduce the benzene content of reformate by more than 99% has been developed by GTC Technology (Houston; www.gtchouston.com). BenZap is situated downstream of the reformer and employs hydrogenation in combination with a platinum catalyst to convert benzene to cyclohexane and other more acceptable compounds (Figure 3).

ExxonMobil Research and Engineering Co. (EMRE) offers Benzout, a reformate alkylation process in which benzene-rich streams are reacted with light olefins such as ethylene or propylene. Benzout converts benzene to high-octane alkylate, with a 2 – 5 gain in octane numbers, thus avoiding the octane loss and H₂ consumption of conventional processes, according to EMRE.

A pervaporation membrane that selectively removes benzene (or other aromatics) from aliphatic compounds has been developed by PolyAn GmbH (Berlin; www.poly-an.de). The membrane has been tested in combination with extractive distillation to obtain pure benzene. PolyAn is working with Borsig Membrane Technology GmbH (Gladbeck, Germany; www.borsig-mt.com) to scale up the process.

Other companies that offer benzene-reduction technology include Axens North America Inc. (Houston; www.axens.net), CDTech (Houston; www.cbi.com/cdtech), and UOP.

Gerald Parkinson

The Canadian Connection The ongoing expansion of U.S. refining capacity constitutes only part of the North American petroleum picture. The driver for many of the ongoing or planned refinery projects is that an increasing amount of the future refinery feed will be heavy crude from Alberta oil sands as well as from other sources. Western Canada is in the middle of a major expansion that will increase oil production from 2.4 million bbl/d in 2006 to 4.6 – 5.3 million bbl/d by 2020, according to the Canadian Association of Petroleum Producers (CAPP, Calgary; www.capp.ca). At present about 50% of production is heavy oil and 50% light or medium oil, but oil sands production is increasing while conventional crude output has been declining. Much of the future production will go to the U.S. Also, while pipelines from Canada have until now mostly served refineries in the Midwest and Northwest, new pipelines are in the works that will carry Canadian heavy oil to the Gulf and east coasts (see Figure 1). Canada is also adding new refining and oil-upgrading capacity to serve the U.S. market as well as the domestic market. The investments include a 300,000-bbl/d refinery that will be built in Newfoundland’s Placentia Bay area by Newfoundland & Labrador Refining Corp (St. John’s). Scheduled for startup in 2011, it will be the first new refinery built in North America since 1984. What amounts to a major new refinery is being built by Motiva Enterprises LLC (Houston, Tex.; www.motivaenterprises.com), a joint venture of Shell Oil Co. and Saudi Aramco, at its Port Arthur, Texas, refinery. Scheduled to go onstream in 2010, the 325,000-bbl/d expansion will increase total capacity to 600,000 bbl/d, making it the largest refinery in the U.S. In addition to a 325,000-bbl/d crude unit, the expansion will include a 75,000-bbl/d hydrocracker, a 95,000-bbl/d delayed coker and an 85,000-bbl/d naphtha reformer. The new capacity will be closely integrated with the existing refinery, says Thomas Purves, vice president manufacturing for Shell and Motiva. For example, he notes that the existing fluid catalytic cracker has excess capacity because vacuum gas oils that would normally go to the FCC are used in the refinery’s 40,000-bbl/d lube base oil plant (the world’s largest). Consequently, at present the refinery buys FCC feed. After the expansion, the refinery will generally be balanced on gas oil. Valero Energy Corp. (San Antonio, Tex.; www.valero.com), the nation’s largest refiner, is adding a 50,000-bbl/d hydrocracker and a 45,000-bbl/d coker to its Port Arthur, Tex., refinery, thereby boosting the refinery’s capacity to 415,000 bbl/d. An expansion is also under way at the company’s St. Charles, La., refinery. The company’s goal, like that of other refiners, is to meet the projected fast-growing demand for ultra-low-sulfur diesel fuel. (ULSD). "We are trying to increase our runs of heavy, sour crude, which you can buy at a deep discount to benchmark prices, and turn it into ULSD," says Richard Marcogliese, executive vice-president and chief operating officer. Valero already has the nation’s largest coking capacity, with a current total of 455,000 bbl/d of feed and 25,000 ton/d of petroleum coke. Marcogliese notes that the expansion at Port Arthur will increase coke production there from 6,000 ton/d to 10,000 ton/d. Next, he says, the company plans to install a large coke gasifier at Port Arthur to make hydrogen for hydrotreating and to generate power. A number of U.S. refinery projects are integrated or connected with Canadian heavy oil production. At the end of last year, BP acquired a half-share in the Sunrise (Alta.) oil sands field, operated by Husky Energy (Calgary; www.huskyenergy.ca) while Husky acquired a half-share in BP’s Toledo, Ohio, refinery. The refinery’s capacity will be expanded from its present 155,000 bbl/d (including 60,000 bbl/d of heavy oil) to 170,000 bbl/d by 2015, with the entire capacity devoted to heavy oil and bitumen. BP is also expanding and upgrading its Whiting refinery in northwestern Indiana to refine more Canadian heavy crude and to increase fuel output. Meanwhile, last fall Marathon Oil Corp. (Houston, Tex.; www.marathon.com) obtained a 20% interest in the Athabasca Oil Sands Project (Alta.) through the acquisition of Western Oil Sands Inc. (Calgary). The other 80% is owned by Shell (60%) and Chevron (20%). To process the Canadian oil, Marathon will expand its Detroit, Mich., refinery’s crude capacity from 100,000 to 115,000 bbl/d and increase its heavy oil-processing capacity by 80,000 bbl/d. In another U.S.-Canada venture, ConocoPhillips (Houston, Tex.; www.conocophillips.com) and EnCana Corp. (Calgary, Alta.; www.encana.com) have created two 50/50 operating partnerships. One consists of two EnCana oilsands projects in northeastern Alberta and the second involves ConocoPhillips’ refineries in Roxana, Ill., and Borger, Tex. The partners plan to increase heavy-oil processing capacity at the two refineries from the present 60,000 bbl/d to 550,000 bbl/d by 2015 and to increase total capacity from the current 450,000 bbl/d to 600,000 bbl/d.