Technical Options for Processing Additional Light Tight Oil Volumes Within the United States
Release date: April 6, 2015
Preface
U.S. oil production has grown rapidly in recent years. U.S. Energy Information Administration (EIA) data, which reflect combined production of crude oil and lease condensate, show a rise from 5.6 million barrels per day (bbl/d) in 2011 to 8.7 million bbl/d in 2014. Increasing production of light crude oil from low-permeability, or tight, resource formations in regions like the Bakken, Permian Basin, and Eagle Ford, often referred to as light tight oil (LTO), accounts for nearly all the net growth in U.S. crude oil production.
EIA's March 2015 Short-Term Energy Outlook (STEO) forecasts U.S. crude oil production averaging 9.3 million bbl/d in 2015 and 9.5 million bbl/d in 2016, well above the 2014 average level but only moderately above production during December 2014. EIA's Annual Energy Outlook (AEO) projects further production growth, but its pace and duration remain uncertain, as shown by the significant differences in both the timing and level of the highest volume of U.S. crude oil production between the Reference case and the High Oil and Gas Resource case.
Recent and forecast increases in domestic crude oil production have sparked discussion on the topic of how rising crude volumes might be absorbed. As EIA noted nearly two years ago, relaxation of restrictions on U.S. exports of crude oil is only one among several ways to accommodate growing near-term flows of domestic production (EIA, This Week in Petroleum, “Absorbing increases in U.S. crude oil production,” May 1, 2013). Recognizing that some options, such as like-for-like replacement of import streams, are inherently limited, the question of how a relaxation in current limitations on crude exports might affect domestic and international markets for both crude oil and products continues to hold great interest for policymakers, industry, and the public. In response to multiple requests, EIA is developing analyses that shed light on this question, including earlier reports on U.S. crude oil production by type (EIA, U.S. crude oil production forecast — analysis of crude types, May 29, 2014), gasoline price determinants (EIA, What drives gasoline prices?, October 30, 2014) and changes in U.S. crude oil imports to accommodate increased domestic production (This Week in Petroleum, “Crude oil imports continue to decline,” January 23, 2014).
This report examines technical options for processing additional LTO volumes within the United States. Domestic processing of additional LTO would enable an increase in petroleum product exports from the United States, already the world’s largest net exporter of petroleum products. Unlike crude oil, products are not subject to export limitations or licensing requirements. While this is one possible approach to absorbing higher domestic LTO production in the absence of a relaxation of current limitations on crude exports, domestic LTO would have to be priced at a level required to encourage additional LTO runs at existing refinery units, debottlenecking, or possible additions of processing capacity. The cost of such adjustments or capacity additions, together with the perception of market and policy risks surrounding potential investments, will determine the extent to which LTO might need to be discounted to spur those investments.
The analysis of technical options for additional domestic LTO processing discussed in this report, together with the previous analyses cited above, provide a foundation for further analyses of the market outlook and the effects of a possible relaxation of existing restrictions on U.S. crude oil exports.
Executive summary
With the growth in U.S. production of light tight oil (LTO) in recent years, domestic refiners have been processing greater LTO volumes. To date, increased runs of domestic LTO have mainly been facilitated by a reduction of light crude oil imports, particularly to refineries on the U.S. Gulf Coast (USGC) and the East Coast. In addition, refinery utilization rates have increased, and some imports of heavier crude types have also been displaced in some U.S. regions.
The sharp decline in oil prices in recent months may slow domestic LTO production growth in the immediate future. However, there is significant potential for further growth in domestic LTO production. This is particularly the case in scenarios with favorable resource availability, technology development, and oil prices that rise above their level in early 2015, even if they remain below the range sustained from 2011 through mid-2014. For this reason, there is considerable interest in how additional volumes of domestically produced LTO might be accommodated.
This paper focuses on technical options for processing more LTO within the United States. With the rise in domestic production of petroleum products and a general decline in U.S. petroleum product use since 2005, the United States, until recently the world’s largest net importer of petroleum products, is now the world’s largest net exporter of these products. There are no limitations on U.S. trade in petroleum products. Trends in U.S. consumption of petroleum products such as gasoline and diesel fuel reflect prices determined on global product markets, fuel economy and alternative fuel policies, and demographic and economic drivers. Because petroleum product use is not significantly affected by the level of U.S. refining activity, additional processing of crude oil in the United States would likely increase net exports of petroleum products.
The relaxation of current limitations on exports of crude oil, another possible way to accommodate additional LTO production volumes, is not considered in this report. However, the discussion of technical options for additional domestic processing provided here points to several key issues that will be addressed in forthcoming analysis that considers how markets might evolve with or without changes in current limitations on crude oil exports. These questions include:
- How large is the opportunity to further increase the utilization of existing refining assets to process more LTO, and what are the economic costs associated with such increased utilization?
- What is the actual opportunity for LTO to displace non-similar grades of imported crude oil?
- At what rates of return and payback periods would investments in additional processing capacity become attractive, given the risks associated with future changes to policy, prices, and production?
- How might the costs associated with processing more LTO in domestic facilities be reflected in prices paid to LTO producers? How would any price discount affect projected LTO production?
There are several different ways that U.S. refiners could process additional volumes of LTO. Beyond their varying associated costs, the options have different implications for overall U.S. petroleum product output volumes and their composition. Once the growth in throughput volume from no- or low-cost import substitution or debottlenecking options has been fully realized, further increases in LTO processing would require significant investments in new processing facilities. Refiners would likely prefer low-cost investment options to larger and more-expensive facilities given market, timing and policy risks associated with large-scale investments.
Limited- or no-investment-cost options
The cheapest and simplest option for U.S. refiners to process more domestic LTO production is to replace imported volumes of similar light crude oil types. However, opportunities for additional “like-for-like” substitutions are limited given the significant backing-out of light crude imports that has already occurred since 2010. Refiners are likely to increasingly consider other options, such as displacing volumes of non-light crude oil imports, increasing refinery utilization rates, or making limited investments in debottlenecking existing refinery infrastructure.1
Crude import displacement
Imports of light crude oil into the United States have decreased significantly in recent years, with light crude imports to the USGC almost fully eliminated. As these light crude oil imports have been displaced, refiners looking to process additional domestic LTO production with existing capacity have reduced imports of comparatively heavier crude types. This can result in a lighter overall crude oil input slate to U.S. refineries, which can lead to operational inefficiencies, particularly at refineries designed to process heavier crude oil types.
To offset some of the impact of the lighter crude inputs on refinery operations, refiners can purchase additional volumes of heavy crude oil imports for blending, as medium crude oil imports are displaced.2 The extent to which refiners can do this is limited by the amount of medium crude oil available for backing out, as well as the availability of heavy crude types that can be used for blending. Additionally, although total production volumes would be unaffected by this blending process, the slate of petroleum products may be altered.
Another technical option would be to process LTO in refineries that are optimized to process heavy crude slates, accepting the inherent operational inefficiencies. While there would be no large investment cost, the opportunity cost could prove to be very large, even if LTO was available domestically at a significant discount to comparable global crudes. One consideration is that many of the heavy crude types currently processed in complex Gulf Coast refineries have few alternative markets. For this reason, markets may respond to potential competition from discounted LTO through more competitive pricing of such crudes that serves to either prevent or substantially curtail their displacement from complex USGC refineries. Under such circumstances, the addition of new processing capacity for LTO, discussed below, may prove to have a lower opportunity cost than “LTO for heavy” displacement, even though it entails a significantly higher investment cost.
Increased utilization of existing processing capacity
Increased capacity utilization has also provided U.S. refiners with a relatively simple way to process additional domestic LTO production volumes, with no additional investment in new processing capacity. However, this has largely already occurred, with crude oil refinery inputs reaching record levels last year.3 The additional domestic LTO production that refiners can process by increasing capacity utilization rates is limited.
Increased refinery utilization rates have already enabled U.S. petroleum product production and net exports to increase, and further increases in capacity utilization would cause these exports to rise above current levels. Unlike crude oil exports, which are currently restricted and subject to licensing requirements, U.S. trade in finished petroleum products is not restricted.4
Capacity debottlenecking
As U.S. refinery utilization rates have increased, refiners have made relatively low-cost investments in equipment modifications to remove restrictions on throughput, also known as crude unit debottlenecking. So far, debottlenecking investments have largely been to replace the gathering trays and condenser units needed to collect the greater volumes of lighter distillation products at the top of an Atmospheric Distillation Unit (ADU) column resulting from processing additional LTO. Because the opportunities for such investments are limited, so too is their potential impact on the amount of additional LTO that U.S. refiners will be able to process. Like increased capacity utilization, capacity debottlenecking would lead to increased petroleum product output, which would largely translate into higher net petroleum product exports.
Capacity expansion options
A range of technical expansion options might be considered once the limited- to no-cost options described in the previous section either are no longer available or if the expected margins (revenues less input and operating costs) available from processing additional domestic LTO volumes are more than sufficient to justify costs and risks associated with new capacity additions. Investment decisions are also likely to reflect factors such as scale, location, crude type availability, construction timelines, market risk, policy risk, and the expected value of petroleum product output slates.
Possible expansion projects include those that increase only ADU capacity, those that increase only secondary processing capacity, and integrated projects that include both ADU and secondary units. Within each category, there are tradeoffs between cost and location. Projects to build greenfield5 facilities outside of existing refinery locations include costs that are not included in projects to build brownfield6 facilities located at existing refineries. Greenfield projects are assumed to include additional costs for production area setup, auxiliary equipment, and other utilities that would already be available for facilities built at an existing refinery. However, greenfield project costs could be less than assumed in this paper for a project located in an industrial area with some available auxiliary equipment and on a surface that has already been prepared for unit construction. Brownfield construction costs might be understated, since even for a project that is built at an existing refinery, it is unlikely that all necessary utilities and other facilities will be available on-site. In addition to financial costs, greenfield and brownfield projects differ in terms of their construction timelines. The risk of a greenfield project is generally greater because its longer lead time increases the likelihood that market changes affecting a project’s profitability will occur before construction is finished.
There is also a tradeoff between scale and risk. For projects of a given type, larger-scale projects generally have lower unit costs due to well-known economies of scale.7 However, the lower unit costs of larger units must be weighed against the potentially substantial risks associated with their higher absolute cost and longer lead times, which increase exposure to changes in the cost and availability of crude oil inputs and changing conditions in petroleum product markets.
Potential changes in policies are another source of risk for capacity expansion options. One major policy risk is the possibility that current restrictions on U.S. crude oil exports will be relaxed or eliminated, which could significantly lessen the future value of additional domestic processing capacity.
Some have argued that policy and market risks would pose an absolute and insurmountable barrier to significant investment in additional domestic processing capacity. Although such a view seems extreme, investors may require a high rate of return over a short timeframe when considering such investments. Table ES-1 illustrates how the per-barrel processing margin required to motivate investment in various refinery expansion projects changes when perceived risk causes potential investors to seek a higher expected rate of return (18% rather than 12%) over a shorter payback period (10 rather than 25 years). A sufficient difference, or spread, between petroleum product prices that are determined on global markets and the price at which crude oil is available to domestic processors could make investments in additional processing capacity attractive even in the face of considerable policy and market risk.
The size of the spread required to motivate investment in additional processing capacity could in turn determine the extent to which domestic LTO would need to be discounted relative to comparable global crudes. The discount for domestic LTO needed to spur additional investment in processing capacity at U.S. refineries would change if assumptions are modified regarding the current limitations on crude oil exports and expected growth in domestic LTO resource availability. The difference in price between domestic LTO and global crude oil would in turn have implications for the level of LTO production, which is one key question to be addressed in considering effects of a possible relaxation of current limitations on crude oil exports.
Expanding only distillation capacity
Investments in units that provide only additional distillation capacity have costs significantly below those of facilities that include secondary processing units that can be used to turn naphtha and other distillation products into finished products like motor gasoline and jet fuel. The unfinished product streams from distillation-only facilities can be sold directly as petrochemical feedstock in domestic or export markets. They can also undergo further processing, either on-site or at another domestic or foreign refinery, into finished products.
Splitters and stabilizers offer refiners a less-expensive option for expanding crude oil distillation capacity than building new refineries or adding ADU columns. Splitters and stabilizers, several of which are now under construction or in the engineering, permitting, or planning phases, are typically designed to distill the very light streams that are largely produced in the Eagle Ford region in south Texas. They will operate less efficiently when processing relatively heavier crudes, including crudes produced in the Bakken region in North Dakota and eastern Montana. Although splitters and stabilizers can process crudes with API gravity well below API thresholds that are often cited as representing condensates, doing so will result in decreased throughput levels and increased output of heavier, unfinished petroleum products.
Because the availability of very light streams is relatively less affected by current limitations on crude oil exports, the supply of feedstock for splitters and stabilizers is less subject to market and policy risk than that of other options. Splitters and stabilizers would also benefit from less market risk because their smaller size requires less upfront investment than larger distillation units.
A 20,000-barrel-per-stream-day8 (bbl/sd) brownfield stabilizer is the least expensive option for expanding distillation capacity, in both overall and per-barrel terms. The limited separation provided by such units would mean that while investing less, operators would also receive smaller processing margins compared to higher-cost projects offering potentially higher processing margins.
A project to build a 50,000-bbl/sd greenfield splitter facility would require more investment than a stabilizer, but less than most other distillation expansion options. Although building the facility as a greenfield project means that it would include production area setup, auxiliary equipment, and utility costs, it also means that it would not have to be located at an existing facility. This would provide flexibility to locate the project in an area that may have better access to markets and lower crude oil transportation costs than the location of existing refineries.
By comparison, a 50,000-bbl/sd brownfield splitter would require significantly less upfront capital to build, as it would not require the construction of any other additional equipment already available at an existing refinery site. However, the feasibility of building a brownfield splitter at an existing refinery would be dependent on the proximity of this refinery to supply sources and the market.
Because of their smaller size and cost, splitter and stabilizer projects can be built with less of a commitment to process large volumes of crude oil over the life of the investment. This is important in light of the market and policy risks previously discussed. A refiner would be significantly more exposed to this risk if they added a 250,000-bbl/sd ADU column to an existing refinery. However, they would also benefit from significantly greater economies of scale.
Secondary processing capacity investment options
Secondary processing units receive their feedstock from the ADU for upgrading to higher-value products. These units are not themselves a technical constraint for refineries to process additional LTO, but could be useful in processing additional volumes of naphtha into motor gasoline. Refiners can build new secondary processing units to accompany many of the distillation capacity expansions described in the previous section.
This paper considers two options for secondary processing capacity, continuous catalytic reformer (CCR) units or isomerization units. Although CCR units have a significantly lower per-barrel cost than isomerization units, they are designed to process heavier intermediate naphtha that is distilled at temperatures that are equal to or greater than 180 degrees Fahrenheit (°F). Despite its higher cost, an isomerization unit could still be an attractive investment for a refinery that receives a relatively high volume of lighter intermediate naphtha that is distilled at temperatures below 180°F.9
Expanding both distillation and secondary processing capacity
Expanding ADU capacity by building new, fully equipped facilities that include secondary processing units would enable U.S. refiners to process additional crude oil volumes into finished petroleum products. These finished petroleum products could be either sold to domestic consumers or exported for more revenue than the unfinished naphtha produced from a distillation-only capacity expansion. Refiners would weigh the benefit of this additional revenue against the relatively higher cost of these projects compared to other options described in this paper.
A hydroskimmer refinery consists of an ADU to distill light crude oil (sometimes referred to as a topping unit) and a modest set of secondary processing units. The basic design of a hydroskimmer refinery’s topping and secondary units makes it less expensive and better-suited for processing domestic LTO production than other fully equipped facilities. A hydroskimmer’s topping unit is smaller than the standard ADU at a full refinery. Its secondary units would only include hydrotreaters and a reformer, not heavier processing units like crackers and cokers. The hydroskimmer’s limited capacity would limit both cost and exposure to market and policy risk.
A 100,000-bbl/sd brownfield hydroskimmer would be the least expensive option for expanding both ADU and secondary unit capacity. It would not include the costs for production area setup, auxiliary equipment, or utilities that would be included if the hydroskimmer were built as a greenfield project. An industrial area that has already received some site preparation, with costs somewhere between the brownfield and greenfield options, is another siting possibility for a new hydroskimmer refinery. Therefore, the cost differences in Table ES-1 for a greenfield and brownfield unit represent a range of costs.
Building an entirely new 250,000-bbl/sd greenfield refinery is the most expensive option for expanding both distillation and secondary processing capacity. It has the highest equipment costs, and because it is a greenfield project, would also have higher costs for production area setup and auxiliary equipment. However, it would yield the most additional revenue for a refiner, because it would have the greatest capacity to produce finished petroleum products. As with all other projects listed, its cost in terms of upfront investment would have to be weighed against this greater additional revenue.
Overnight cost | Amortized capital cost | |||||||
---|---|---|---|---|---|---|---|---|
Option | Advantages | Disadvantages | Processing Capacity (Mbbl/sd) |
Construction time (years) | Total($MM) | Volumetric input ($/bbl/sd) | 12% annual interest and 25-year amortization ($/bbl) | 18% annual interest and 10-year amortization ($/bbl) |
No investment options | ||||||||
Increase refinery capacity utilization | Allows for additional LTO processing at zero cost | Limited by available unused refinery capacity | 0 | 0 | $0 | $0 | $0 | |
Back out medium and heavy imports | Allows refiners to take advantage of available light tight crude volumes at a minimal cost | Operational inefficiencies, reduced crude oil input and production volumes | 0 | 0 | $0 | $0 | $0 | |
Debottlenecking | Allows for additional LTO processing at a minimal cost | Can only provide a limited amount of additional capacity | Minimal | <1 | Minimal | |||
Purchase less domestic light tight oil production | Zero cost, no operational inefficiencies | Shut-in U.S. LTO production volumes, capped growth in petroleum product output | 0 | 0 | $0 | $0 | $0 | |
Distillation-only capacity investment options | ||||||||
Brownfield stabilizer | Additional ADU capacity at an existing facility; lower per-barrel amortized capital cost than any other option | Constrained by crude type; high unprocessed and unseparated naphtha volumes | 20 | 1.5 | $30 | $1,390 | $0.63 | $1.16 |
Brownfield splitter | Additional ADU capacity at an existing facility | Constrained by crude type; high unfinished naphtha volumes; limited locational flexibility | 50 | 1.5 | $100 | $2,060 | $0.94 | $1.73 |
Greenfield splitter | Additional ADU capacity; locational and operational flexibility | New facility; constrained by crude type; high unfinished naphtha volumes | 50 | 1.5 | $140 | $2,830 | $1.30 | $2.39 |
Brownfield atmospheric distillation column | Brownfield atmospheric distillation column | Policy risk; high unprocessed, unfinished product volumes could require additional investment | 250 | 2 | $370 | $1,500 | $0.70 | $1.31 |
Secondary processing-only capacity investment options | ||||||||
Brownfield isomerization unit | Low cost; increased volume of higher-octane motor gasoline | Only ideal for naphtha processed at temperatures below 180°F | 20 | 1.5 | $110 | $5,250 | $2.38 | $4.37 |
Brownfield continuous catalytic reformer unit | Low cost; increased volume of higher-octane motor gasoline | Only ideal for naphtha processed at temperatures equal to or greater than 180°F | 50 | 2 | $150 | $3,000 | $1.40 | $2.61 |
Combined distillation and secondary processing capacity investment options | ||||||||
Brownfield hydroskimmer refinery | Additional capacity at an existing facility; some economies of scale; finished products | Limited volumes, relatively high cost; limited locational flexibility | 100 | 3 | $530 | $5,280 | $2.64 | $5.06 |
Greenfield hydroskimmer refinery | Additional capacity; some economies of scale; finished products | Limited volumes, relatively high cost | 100 | 3 | $720 | $7,170 | $3.59 | $6.87 |
Full greenfield refinery (ultra-light) | Additional capacity; low impact on existing facilities; finished products | High capital costs, market and policy risk | 250 | 3 | $3,390 | $13,540 | $6.78 | $12.98 |
Sources: U.S. Energy Information Administration, Independent Project Analysis, Inc. | ||||||||
Note: Mbbl/sd = thousand barrels per stream day, or the maximum volume that a distillation facility can process in a 24-hour period under optimal conditions with no allowance for downtime; $/bbl = dollars per barrel; $/bbl/sd = dollars per barrel per stream day; MM = million. Overnight cost is the cost of a project with no interest incurred, or the lump sum cost of a project if it were completed overnight. Amortized capital cost is the revenue per barrel processed needed to pay the cost of the project over a 25-year period with a 12% annual interest rate, or over a 10-year period with an 18% annual interest rate, with the facility operating at a utilization rate of 85% of full stream-day capacity. |
Endnotes
2EIA data through December 2014 indicate a 24% decrease in medium crude oil imports (crudes with an API gravity of equal to or above 27 and below 35) into the United States over the last three years, from an average of 3.3 million barrels per day (bbl/d) in 2011 to an average of 2.5 million bbl/d in 2014. The data indicate particularly large decreases in the Gulf Coast (31% or 0.5 million bbl/d) and East Coast (33% or 0.2 million bbl/d) Petroleum Administration for Defense Districts over this period.
3Gross inputs to U.S. refineries reached a record 16.9 million barrels per day (bbl/d) in July 2014.
4Net exports of finished petroleum products averaged 2.1 million bbl/d in 2014, composed of 2.8 million bbl/d of gross finished product exports and 0.6 million bbl/d of gross finished product imports.
5Built on a site with no existing infrastructure in place.
6Built on a site with existing infrastructure in place.
7The cost advantage obtained due to size, output, or scale of operation.
8The maximum volume that a distillation facility can process in a 24-hour period under optimal conditions with no allowance for downtime.
9This refers specifically to C5 and C6 molecules, and not C8 molecules, which can be isomerized after being distilled at higher boiling temperatures.