3. Mid-Term Natural Gas Supply: Analysis of LNG Imports

Introduction

During the winter and spring of 2001, when U.S. natural gas prices reached record highs and strong growth in natural gas consumption was being forecast both worldwide and in the United States, many analysts and investors expressed the opinion that liquefied natural gas (LNG) might in the future provide a much larger share of U.S. natural gas supply. Given the natural gas consumption forecast of 33.8 trillion cubic feet in 2020 in the Energy Information Administration’s (EIA’s) Annual Energy Outlook 2002 (AEO2002),⁶⁴ existing design capacity of just over 1 trillion cubic feet per year at the four U.S. terminals⁶⁵ and proposed expansion of about 0.4 trillion cubic feet per year would be able to meet about 3.3 percent of projected total demand in 2020.

Costs throughout the LNG supply chain have fallen, making it a much more attractive economic investment, especially for suppliers who believe that prices will reach and be sustained at a level high enough to make LNG competitive. Although natural gas prices have fallen considerably in the past few months,⁶⁶ many suppliers are still confident that prices will increase and remain above the level at which they feel LNG is competitive.

International supplies are plentiful, but the capacity of U.S. import facilities limits the amount of gas that can be received and regasified. There are currently three facilities in operation, at Everett, Massachusetts; Lake Charles, Louisiana; and Elba Island, Georgia. A fourth facility at Cove Point, Maryland, was scheduled to reopen within the next year, but the opening may be delayed as a result of rehearing requests received by the Federal Energy Regulatory Commission (FERC).⁶⁷ The FERC will make a decision on December 13, 2001, on whether to allow a rehearing.

Numerous additional facilities are under consideration, but siting an LNG receiving terminal can be a formidable task. Aside from the facility site requirements, local opposition (often referred to as the NIMBY⁶⁸ factor) can be close to insurmountable and is likely to be the most important factor in whether a facility is built in a particular location. A method for circumventing this obstacle is to site facilities to serve U.S. markets outside U.S. borders, such as in Baja California (Mexico) or in the Bahamas. While this may reduce the NIMBY opposition, it will not eliminate it entirely; local Baja residents have voiced resistance to the siting of an LNG facility in the Baja region.

While the international supply needed to satisfy the potential U.S. demand for LNG is available, worldwide LNG demand is also seeing strong growth, which could lead to competition for available supplies and higher prices. Another potential limiting factor for U.S. LNG import growth is tanker availability. Although tanker construction costs have fallen, shipping cost is still a major component of LNG prices. Because most LNG tankers are built within the context of long-term purchase commitments, there are few tankers available to handle short-term purchases, even though short-term sales are becoming a larger proportion of the LNG market.

Recent Worldwide LNG Developments

The growth in demand for natural gas worldwide is outpacing the demand for any other hydrocarbon fuel. This is due to a number of factors, including price, environmental concerns, fuel diversification and/or energy security issues, deregulation of both natural gas and electricity markets, and overall economic growth. In EIA’s International Energy Outlook 2001 (IEO2001), worldwide natural gas use is projected to almost double between 1999 and 2020, growing from 84 trillion cubic feet to 162 trillion cubic feet. The largest increases in natural gas use are expected in Central and South America and in developing Asia. The largest increases in industrialized countries are expected in North America (primarily the United States) and Western Europe.

Given the anticipated growth in world demand for natural gas, it will be necessary to develop new natural gas fields and infrastructure to assure adequate supplies. As of January 1, 2001, 10 countries held 77 percent of the world’s natural gas reserves, with Russia, Iran, and Qatar accounting for more than 55 percent (Table 7).

Natural gas reserves that would be extremely expensive to transport through pipelines to potential markets are commonly referred to as “stranded reserves.” Stranded reserves are expected to be a major source of natural gas for world LNG trade. It has been estimated that stranded reserves make up about 50 percent of the natural gas reserves held by the top 10 countries shown in Table 7 and between 2,755 and 3,350 trillion cubic feet worldwide.⁶⁹

Qatar began exporting LNG in 1997 and currently has two export terminals and an additional one in the planning stage. Iran has two terminals in planning stages, currently scheduled to be operational in 2005 or 2006. Nigeria began exporting LNG in 1999, and Venezuela has plans to begin in 2003. Indonesia, Algeria, Malaysia, and the United Arab Emirates, four of the five largest exporters of LNG in 1999 (Qatar is the fifth), have been exporting LNG for close to 20 years, and Australia has been exporting LNG since 1989. Considerable expansion is planned at existing liquefaction facilities, and at least 15 new projects are under consideration.⁷⁰ Although it is the newest project in the industry, future LNG production from Trinidad’s three trains⁷¹ will make Atlantic LNG the fifth largest exporter of LNG in the world.⁷²

Given this distribution of resources, an increase in the amount of natural gas traded across international borders will be inevitable. With many natural gas resources located far from demand centers, LNG will become progressively more attractive as a method of transport. Although in 1999 barely 20 percent of the natural gas consumed worldwide was traded across international borders, 22 percent of that was in the form of LNG. LNG will both satisfy some of the increasing demand and provide source countries a means of monetizing these otherwise stranded natural gas reserves.

One factor contributing to the world growth in the LNG trade is the declining cost structure of all phases of the supply chain, which has allowed the cost at which LNG becomes economic to fall within the year 2001 range of natural gas prices. With new suppliers coming on board, competition has forced cost-cutting measures and price reductions. Liquefaction costs between 1996 and 2000 averaged $230 per ton, compared with $560 per ton between 1986 and 1990. Between 1996 and 2000 the cost of a new tanker dropped by approximately 30 percent.⁷³ The construction costs for regasification terminals have seen similar decreases. In addition to the numerous planned expansions and new facilities for liquefaction, the tanker fleet is expanding. At the end of 2000, the fleet stood at 127 ships, with 22 on order and 7 under option. More than 20 new LNG receiving terminals are either planned or proposed, and more than 10 are under either renovation or construction.

LNG Technology and Economics

Although worldwide natural gas supplies for LNG facilities are abundant and can be produced inexpensively, the processing and transportation equipment is capital intensive and highly specialized, requiring hundreds of millions of dollars of investment for each new facility. For each cubic foot of natural gas delivered to end users, less than 30 percent of the cost is for the commodity itself. The balance reflects the costs of processing and transportation. LNG project costs can vary significantly because of site-specific construction costs. LNG projects comprise several distinct elements, each of which is necessary to implement a successful project:

• Abundant low-cost natural gas reserves. A successful LNG project must have enough proved reserves of natural gas available to support liquefaction capacity for the life of the plant (20+ years). In addition, production costs (including applicable production taxes levied by the host government) need to be low (typically, less than $1 per thousand cubic feet, and preferably on the order of $0.50 per thousand cubic feet).

• A liquefaction facility, including a jetty and loading facilities for LNG tankers. The liquefaction plant is typically the most expensive element of an LNG project. The cost depends on a host of site-specific factors, including the project’s scale, with larger projects having lower unit costs. Operating costs are relatively minor. Liquefaction is a very energy-intensive process, with typically about 8 to 9 percent of the plant’s input used as plant fuel.

• LNG tankers. Each project requires several dedicated LNG tankers. These are among the most complex and expensive merchant ships ever built because of their double hulls and special cryogenic lining. Each new 135,000 cubic meter (3 billion cubic foot) capacity tanker costs approximately $260 million. The tanker’s LNG cargo is kept cool by evaporating a fraction of the cargo (“boiloff”) and burning it as boiler fuel. Typically, 0.15 to 0.25 percent of the cargo is consumed per day, during which the tanker will travel about 480 nautical miles.

• Regasification plant. LNG can be unloaded only in specialized terminals, which typically include a jetty and unloading facilities, LNG storage equal to at least a single tanker cargo, regasification facilities, and connections to pipelines. The cost of the regasification terminal varies with capacity, local construction costs, and the amount and type of site preparation costs. Regasification plant costs are typically considerably lower than liquefaction plant costs. Regasification energy requirements consume a further 1.5 percent of the delivered LNG. The marginal cost of either utilizing excess capacity at an existing regasification plant with excess capacity or expanding the capacity of an existing plant would be far lower than the cost of building a new greenfield facility.

The large capital costs of each link in an LNG project imply that projects can be undertaken in general only by organizations with sufficient financial capacity. Under the traditional LNG project structure, successful LNG projects required the cooperation of the host government (where the natural gas resources are located), the entity that owns the natural gas rights (private or state), the government of the consuming country, consuming organizations (national or private electric utilities, gas companies, etc.), and a host of specialized organizations, including shipyards, financiers, tanker operators, construction companies, and process technology licensors. In the past, protracted negotiations were often needed to reach agreement regarding the distribution of the costs, the benefits, and the considerable risks associated with the project. This project structure may be evolving, however, as a result of the proliferation of spot market trading of LNG in recent years.

No LNG project is likely to proceed unless the developers receive some assurance that they will be able to earn an acceptable return on their investments. A successful LNG project requires a price that is low enough to motivate consumers to use large volumes of natural gas, yet still high enough to persuade developers and borrowers to actually build the project. One risk that cannot be ignored is the likely formation of an LNG cartel, given that so few countries control such a large portion of the world’s stranded natural gas reserves, and its power to affect LNG prices. Although spot sales are on the rise, LNG developers will seek (but not always find) long-term contracts for their product at a price that is sufficient to cover their capital costs and service debts even in a lower-than-anticipated energy price environment. It is also common for consumers to be offered or to take an equity stake in LNG projects, so as to encourage a common interest among the buyers and the sellers.

With natural gas consumption growing rapidly throughout the world, there are many potential and expanding LNG markets. Countries that are potential LNG markets are those with significant demand for natural gas (enough to make LNG trade economically viable) that cannot be satisfied by their own indigenous production or by pipeline imports from neighboring countries, because of a lack of reserves or lack of infrastructure to get reserves to the demand centers. Receiving terminals for LNG are being constructed or considered in numerous locations, including China, India, Korea, Japan, Southern Europe, Latin America, and recently the United States.

When locations for new LNG import facilities are proposed, several tangible and intangible factors must be considered. The major tangible factors include water depth (especially the depth of the channel to the jetty and the potential for silting), availability of reasonably priced large industrial tracts near deep water, and availability of a right-of-way for the pipeline (in high-density areas, rights-of-way may already exist). These three criteria must be satisfied before any location can receive further consideration. In addition, the site needs to have both proximity and access to markets and, of course, access to LNG supplies.

The primary elements of the LNG receiving facility itself are berths for unloading the LNG tankers, storage tanks to receive the ship’s cargo, and vaporizers to regasify the LNG for distribution to market centers through natural gas pipelines. Other elements include site improvements and roads, buildings and services, and miscellaneous components including piping, controls, and utilities. The actual construction time averages about 3 years. In the United States, the approval process for a new site, which usually takes from 18 to 24 months, can be extended considerably if there is strong NIMBY opposition.

The U.S. Market for LNG

LNG in the United States has a sketchy past. Because of rising natural gas prices in the 1970s, LNG project sponsors anticipated large profits and constructed the four U.S. LNG receiving terminals in existence today. Dreams of high profits never materialized, however, because natural gas prices began a precipitous decline after their 1983 peak, and all but one of the four were mothballed. The facility at Everett, Massachusetts, remained in operation only because it was located in a heavily concentrated market center where demand was high and the cost of bringing conventional supplies to market by pipeline was high enough to exceed the cost of LNG.⁷⁴ In 1989, the Lake Charles, Louisiana, facility was reactivated,⁷⁵ mainly to receive spot cargos.

For close to 20 years, LNG was not considered to be an economical source of natural gas. As a result of the high 2000-2001 prices and the growing demand for natural gas, interest in LNG has renewed to the point that not only are the other two facilities, at Elba Island, Georgia, and Cove Point, Maryland, reopening (Elba reopened in October 2001), but at least 13 new facilities have been proposed to serve U.S. markets (Table 8). Some of the parties proposing the terminals readily indicate that although prices have fallen since their proposals were first put forth, they expect future prices to be in a range where LNG is economical relative to competing supply sources. Although LNG was in the past used mainly for peaking purposes, the expanding use of natural gas for electricity generation potentially makes it a less seasonal commodity. Thus, if the economics of LNG become more favorable in the United States, higher utilization of LNG facilities can be expected, just as pipeline capacity utilization is increasing.

Existing LNG Receiving Terminals

Everett Marine Terminal. The Everett Marine Terminal has a design capacity of approximately 160 billion cubic feet per year,⁷⁶ and plans have been announced to add another 200 billion cubic feet per year capacity. Everett is located northwest of central Boston, Massachusetts, on the Mystic River. Construction was completed in 1971, and it has been in operation since that time. It has one unloading berth and two aboveground storage tanks. One tank has a 60,000 cubic meter capacity and the other has a 95,000 cubic meter capacity, for a total of 155,000 cubic meters. Assuming the average LNG ship cargo is 130,000 cubic meters (net), the tanks can hold 1.19 ship cargos.⁷⁷ In addition to supplying natural gas to the Algonquin pipeline, the facility has the capability to load 1 million gallons per day or more of LNG into trailers for over-the-road transport to other facilities.

The original vaporizer configuration consisted of two trains, with six vaporizers per train. The vaporizers are direct fired with hot water exchangers. Everett can also send out between 90 and 100 million cubic feet per day by truck. The facility has been expanded several times with pipeline connections and increased truck loading capability. Although the facility operators are planning to expand the vaporizing capabilities by adding additional submerged combustion vaporizers, the limited availability of land (and corresponding limits of exclusion zones) precludes additional tankage, which creates a cap on facility growth.

Cove Point Import Terminal. The Cove Point facility is has a design capacity of 365 billion cubic feet per year. Cove Point is located on the Chesapeake Bay at Cove Point in Lusby, Maryland, about 50 miles south of Washington, DC. It was constructed in 1978 and operated as an LNG import and storage facility from 1978 to 1980, before being shut down. Since 1995, it has been providing peak shaving services to customers in the mid-Atlantic and Southeastern regions. The new operator, Williams Companies, had been granted permission by the FERC to return it to an LNG import and storage facility. As a result of national security and safety concerns, raised in the wake of the September 11th terrorist attacks, regarding Cove Point’s proximity to the Calvert Cliffs nuclear facility, however, the FERC is considering whether to grant or deny rehearing requests that have been submitted.⁷⁸

Cove Point has two unloading berths capable of handling large LNG ships and four aboveground storage tanks. All four storage tanks have a capacity of 59,630 cubic meters, or 238,520 cubic meters total. This equals 1.83 ship cargos, assuming a net cargo of 130,000 cubic meters per ship. Total receiving capacity is 435 billion cubic feet (19.8 million cubic meters) per year, or about 150 cargos. The facility has 12 vaporizers (10 fired vaporizers and 2 non-fired using waste heat). Williams received plans to add one additional 160,000 cubic meter tank and recommission idle vaporizers. Cove Point is surrounded by open land and considerable future expansion would be technically possible, but expansion is limited by an agreement with the Sierra Club that prohibits expansion beyond current boundaries.⁷⁹

Elba Island Import Terminal. The Elba Island facility has five submerged-type vaporizers with a total vaporization design capacity of approximately 160 billion cubic feet per year. The operators have announced plans to replace the existing vaporizers with five larger submerged combustion vaporizers, which will give the terminal a total design capacity of 292 billion cubic feet per year.⁸⁰ Located on Elba Island, downriver of Savannah, Georgia, on the Savannah River. It was completed in 1978 and operated until 1980, when it was shut down. It was recently recommissioned, and received its first cargo in October 2001. Presently, it consists of one berth and three above ground storage tanks. All three tanks have a 60,000 cubic meter capacity, or 180,000 cubic meters total. This equals 1.38 ship cargos, assuming a 130,000 million cubic meter net cargo per ship. Elba Island also has a trailer unloading capacity of 10 million cubic feet per day.

Lake Charles Import Terminal. The Lake Charles import terminal has seven submerged-type vaporizers with a design capacity of 365 billion cubic feet per year. Plans have been announced to increase the vaporization capacity by adding another 73 billion cubic feet per year.⁸¹ Other expansion plans are believed to include the addition of another berth for unloading LNG ships. The Lake Charles facility is located on the Calcasui River, south of Lake Charles, Louisiana. It was completed in 1982 and operated until 1983. It was reopened in 1989 and has remained in operation. It has one berth and three aboveground tanks. Each of the Lake Charles tanks has a 95,400 cubic meter capacity, for a total of 286,200 cubic meters. Lake Charles has a maximum receiving capacity of 165 billion cubic feet (7.5 million cubic meters). Based on a net cargo of 130,000 cubic meters, the tanks will hold 2.20 cargos, or about 58 cargos per year.

Potential Sites for New LNG Facilities

North Carolina. El Paso Natural Gas has announced a lease on Radio Island in Morehead City, North Carolina, as the potential site for a 100 billion cubic feet per year LNG facility. Morehead City has deep water and Radio Island is near the channel entrance, in protected water, making it a suitable site for docking large LNG vessels. The primary disadvantage in North Carolina is that the major transmission pipelines run through the western half of the State, and a rather long right-of-way must be acquired to connect to the system. Fortunately, North Carolina is largely rural and the right-of-way can avoid heavily populated areas and the accompanying problems. Nevertheless, local opposition to the proposed terminal has been strong.

Florida. Florida has 14 deepwater ports, but only Tampa (and possibly Jacksonville) has adequate depth for a large LNG tanker. Florida is also a rapidly growing market. Florida’s governor has opposed the President’s energy policy regarding the exploration and development of oil resources within 100 miles of the coast, and public opposition to any new energy facility in the State should be expected. Additionally, most of Florida’s coastline is developed, and a large undeveloped site within a mile of deep water would be difficult to find.

El Paso Natural Gas and Enron have independently announced plans to investigate building a facility in the Bahamas, about 30 miles off the Florida coast, either shore-based or offshore. An underwater pipeline would connect the facility to existing pipelines in Florida and the environmental impact in the United States would be minimal. The facility sizes proposed are 200 and 250 billion cubic feet per year, respectively. British Petroleum (BP) is also considering a 200 billion cubic feet per year facility in Tampa.

Gulf of Mexico. Texaco has recently announced plans to investigate the feasibility of developing an offshore LNG facility with a 365 billion cubic feet per year capacity in the Gulf of Mexico and connecting it to one or more of the existing pipelines in the Gulf. The facility (or facilities) would be a deepwater floating terminal, similar in concept to the Louisiana Offshore Oil Port (LOOP). The hurdle to overcome in this case is regulatory. Discussions with United States Coast Guard (USCG) officials revealed that the existing regulations for this type of facility cover only oil ports; no regulations exist for LNG facilities, and the USCG would require the promulgation and oversight of such regulations. This also raises a jurisdictional issue for the FERC, which has jurisdiction up to the 200-mile limit. Although several studies have indicated that the technical problems are manageable, such an offshore floating terminal would be a first. The ship-to-terminal offloading method has created considerable discussion.

Texas and Louisiana. The coast of Texas and Louisiana has several potential sites for an LNG facility. Lake Charles, Louisiana, already has an LNG facility, and Dynegy has announced plans to construct a 275 billion cubic feet per year LNG facility at its LPG facility in Hackberry, Louisiana. Like other LNG import waterways, Lake Charles has regulations in place for LNG ship transit. Texas ports, such as Houston and Corpus Christi, are already heavily industrialized with oil terminals, and an LNG terminal would not be out of place; however, finding a large, available tract of land may be the greatest obstacle. The Houston Ship Channel is already congested with traffic, and LNG tankers would not be welcomed in the ship safe zones utilized elsewhere. Port Arthur also has some possibilities. Texas and Louisiana also have several existing pipelines, making acquisition of a suitable right-of-way less of a problem in this region. Cheniere Energy has announced plans to construct three LNG receiving terminals, each with a 200 billion cubic feet per year initial capacity, along the Texas Gulf Coast.

Southern California and Mexico. Southern California has many logistical considerations in common with Florida. It has several deepwater ports capable of handling LNG ship traffic, but the combination of existing industrial development and a high population density would make siting an LNG facility difficult. The ports that would be most attractive have already been developed, and their utilization would require innovative approaches. Additionally, the citizenry and government are very protective of the environment and resistant to this type of project. California also has the history of an unsuccessful LNG import project in the late 1970s.⁸²

El Paso has announced plans to build a 250 billion cubic feet per year LNG facility in Mexico and has secured property in Rosarito. The plan is to construct the facility in Mexico and connect it to an existing pipeline near the U.S. border to serve the California natural gas market. El Paso will avoid the NIMBY concerns in California but will be subject to Mexican property laws for siting and right-of-way, as well as approval from Mexico’s energy regulatory commission, the Commission Reguladora D’Energie (CRE). There are no ideal ports in northern Baja California; therefore, the siting and design will not be straightforward.

New Regasification Facility Cost Considerations

The costs for an LNG import terminal depend on several variables. Some have minor impacts and others very significant impacts on cost. Those that have a major impact on costs are storage capacity installed, geology of the area (soil stability and seismic activity), labor and construction costs for the area, and the marine environment (proximity to deep water, need for dredging and/or breakwater). Other factors include public opposition and permitting. Other elements that affect total cost are trestle length, sendout, site improvements, roads, buildings, services, and miscellaneous expenses such as piping, controls, and utilities.

In addition to new facility construction, additional capacity can be obtained through the expansion of existing facilities. Most facilities are constructed with an initial operating capacity and built-in expansion potential that can be obtained by increasing any one of a number of factors that limit throughput, including number of berths, size of the receiving tanks, capacity of the vaporizers, and capacity of the sendout lines.

Since there are so many variables that contribute to the cost of building and operating a receiving terminal, a number of assumptions regarding facility configuration and site characteristics were made in developing the costs of new facilities for EIA’s analysis. From these assumptions, generic capital costs for a basic LNG import terminal, as well as multipliers to account for unique features of each potential location,⁸³ were developed. Assumptions regarding the regasification facilities include:

• Two 140,000 cubic meter containment tanks are used for storage. The basis of this assumption is the rule of thumb to have two times ship volume of storage for the facility.

• The trestle length is 300 feet. Trestle length could vary significantly; costs are approximately $140,000 per foot.

• The site has adequate soil for construction of roads and building foundations. No piling is required.

• The area is not seismically active.

• Sendout is 0.5 billion cubic feet per day.

• No dredging is required.

The regional cost multipliers are based on wage differentials, land costs, and other factors that vary by region. They are used to increase the accuracy of the assumed construction costs when applied to different regions. The regional cost multipliers are shown in Table 9.

The major costs in operating a facility are personnel and power. Personnel, the largest single expenditure, is a fixed cost; and power is variable relative to sendout. The estimates used assume that administrative functions are provided by parent company personnel rather than by personnel directly associated with the LNG import facility.

The main operating costs of the facility can be divided into fixed and variable costs. The fixed costs are payroll, maintenance, insurance, and taxes. Payroll is estimated at $2.8 million per year for approximately 22 employees (at Gulf Coast wages), and maintenance costs account for an additional $2.8 million per year. Taxes and insurance are estimated at $5.7 million. This is a rough estimate because taxes (and potential tax abatements) vary widely by location, and insurance costs are, in part, a function of the operator’s safety record. Variable costs include fuel, electricity, chemicals, and other consumables. Electricity consumption is estimated to be approximately 480 kilowatthours per day.

Means of Facility Expansion

In addition to new facility construction, additional capacity can be obtained through the expansion of existing facilities. Most facilities are constructed with an initial operating capacity and built-in expansion potential that can be obtained by adjusting any one of a number of factors that limit throughput, including the following. The import terminal operator has some control over most of these factors in that additions may be made to the facility or operations may be tailored to allow for mitigating factors.

Number of berths. A typical ship unloading requires about a 24-hour turnaround time, broken down as follows:

• 4 hours for customs, immigration, custody transfer measurements, connecting the unloading arms, and cooldown

• 12 to 14 hours unloading

• 6 to 8 hours for final custody transfer measurements and calculations, disconnecting unloading arms, provisioning, and deberthing.

A reasonable scheduling assumption for one berth is one ship every 3 days. For a 2.84 billion cubic feet cargo, this is essentially a 0.9 billion cubic feet per day terminal capacity limitation resulting from a single berth. There will be times when there will be delays such that the shipping, inventory, and sendout logistics must be flexible to accommodate occasional delays. Alternative mooring availability is also a consideration.

Contractual arrangements for shipping. Establishing a shipping schedule well in advance (i.e., for a 1-year period) will allow the inventory management necessary to assure adequate cargo arrivals and a minimum of ship demurrage⁸⁴ while awaiting receiving tank space. Scheduling becomes more complicated where more than one ship or shipper is utilized. If there is more than one export terminal as the source, then inevitably there will be times when two ships arrive on the same day and other times when there is twice the average time between ships.

Number of LNG sources and spot cargo activity. If more than one export source of LNG supplies an import terminal, the ship arrival schedule will be much more erratic with occasional to frequent situations where two cargos arrive nearly at the same time. This implies that there will be an extended period with fewer (or no) cargos. Spot cargos are becoming more available but require some time to negotiate. Spot cargos have little flexibility in schedule, either from the supply or ship availability standpoint. In order to accommodate spot cargos, the import terminal must have the ability to take an extra ship out of normal sequence.

Size of receiving tanks. The receiving tankage must have the capacity to take the ship’s cargo. There must also be additional volume to accommodate schedule and sendout variability. As a rule of thumb, receiving tankage should be at least two times cargo volume, or about 6 billion cubic feet. Additional volume may be useful and/or economic to facilitate erratic ship scheduling, spot cargos, variable sendout rates, and peak demand opportunities. Because ship storage costs about 5 or 6 times the equivalent on-shore storage, the best overall economic result is achieved by buffering logistic variability with additional tankage at the receiving terminal. Additional storage at the receiving terminal also assists in responding to peak demand markets and general logistics management.

Capacity of vaporizers. The sendout pumps and vaporizers must meet the maximum contractual sendout rate. It is common practice to have at least one spare unit for reliability and maintenance functions. Typically, additional vaporizers can be added, although air emission permits can be a problem. Most tanks will have provision for additional or larger pumps. Additional booster pumps for pipeline pressure can typically be added, but installed standby units are advisable and common practice.

Variability in sendout. Generally speaking, as long as the receiving terminal’s storage is large enough, variability in daily sendout rates will affect only the pumps and vaporizers and will not affect the upstream shipping and receiving functions. Short-term sendout variability problems arise when the sendout rate is interrupted such that there will not be receiving tank space for the next ship. Such a situation can occur if the primary customer is a power plant and the power plant is taken offline. A provision in the contract to allow any excess gas to be sold to other customers in the market may alleviate high inventory problems. Long-term variability problems arise when there is a consistent sendout rate either above or below the contractual supply amount. This will result in either shipping delays with demurrage or very low inventories awaiting ship arrival.

Capacity of sendout lines. The sendout pipelines must have the capacity to take away the maximum sendout rate consistent with maintaining the nominal throughput. Pipeline capacity can often be increased by compressor stations and line looping, but these functions may not be within the control of the terminal operator. For example, the Cove Point sendout line currently has a maximum capacity of about 1.2 billion cubic feet per day and is a limiting factor for the current configuration.

Capacity of local and regional system. The local and regional areas served by the terminal need to absorb the throughput. For example, the Distrigas terminal in Boston has a large market in the immediate area, whereas the Elba Island terminal is relatively remote from concentrated usage areas except Savannah.

Schedule discretion in truck deliveries. An import terminal can facilitate inventory management if there are discretionary markets available. If LNG is delivered by truck to offsite peak-shaving plants, the schedule is typically set so that the peak-shaving tank is filled by the beginning of the heating season. When the deliveries are actually made is inconsequential to the peak-shaving plant, so the import terminal can manage inventory by scheduling the trucking to occur during times in the summer when LNG inventories are high. The ability to have certain customers that are willing to take gas at the terminal’s request serves a similar inventory management function. Short-term sales have been made straightforward with the advent of natural gas marketing, and if the import terminal operator is also a natural gas supplier, this may be relatively easy.

Analysis of LNG Imports

Analysis Cases

To analyze the sensitivity of domestic natural gas production and prices to increased LNG import terminal capacity, two analysis cases were used in addition to the AEO2002 reference case and the carbon dioxide emissions limit case (described in Chapter 2, “Analysis of Access Restrictions,” pages 20-24). All the cases used the same assumptions regarding existing and potential future LNG regasification capacity. Because there is considerable uncertainty surrounding the various costs that make up the delivered cost of LNG, the two LNG analysis cases were developed to examine the impact of those costs on the expansion of existing and construction of new LNG receiving terminals. Both cases were based on the carbon dioxide emissions limit case.

The high LNG cost case assumes that LNG production costs are double the costs assumed in the reference case by 2020, that the rate of return on LNG tankers is 20 percent rather than the 15 percent assumed in the reference case, and that the rate of return on liquefaction plants is 12 percent rather than the 10 percent assumed in the reference case. The low LNG cost case assumes that LNG production costs are 50 percent lower than in the reference case by 2020, that the rate of return on LNG tankers is 10 percent rather than the 15 percent assumed in the reference case, and that the rate of return on liquefaction plants is 8 percent rather than the 10 percent assumed in the reference case.

Results

The differences in assumptions for the cases are reflected in the projected minimum regional import prices needed to trigger expansion and/or construction of new facilities. The trigger prices represent summations of the five major costs in the LNG chain: production, liquefaction, transportation, regasification, and receiving terminal site-specific costs such as permitting, special land and waterway preparation and/or acquisitions, and regulatory costs. Because LNG is used primarily to serve local markets in the vicinity of the receiving terminal, it competes with regional prices rather than national average wellhead prices. The regional trigger prices assumed in the reference case and the two LNG analysis cases are shown in Table 10. Projections for domestic natural gas production, consumption, and prices and for LNG imports are summarized in Table 11.

The variation of LNG costs, and thus the availability of more LNG, affect the demand for natural gas (Figure 14). The impacts on natural gas consumption in the high and low LNG cost cases are seen towards the end of the forecast, with the spread reaching 0.6 trillion cubic feet by 2020. The corresponding difference in the wellhead price in 2020 is $0.16 per thousand cubic feet, indicating that the availability of more LNG reduces prices (Figure 15). The largest price spread ($0.21 per thousand cubic feet) occurs in 2017 and represents a difference of about 6 percent between the two analysis cases.

The projected prices in the high LNG cost case are sufficiently high to make expansion at existing facilities over what has already been announced and construction of new facilities uneconomical. As a result, the 2020 projection for LNG import capacity in the high LNG cost case is the same as in the reference case, and LNG imports are the same in the two cases throughout the forecast (Figure 16).

Domestic natural gas production shows a maximum variation across the cases of 0.5 trillion cubic feet in 2015, but the gap narrows to 0.2 trillion cubic feet by 2020 as more LNG capacity becomes available (Figure 17). In 2020, the difference in LNG imports between the two LNG analysis cases exceeds the difference in production by 50 percent. Thus, LNG in these cases not only makes up for the difference in production but also displaces some Canadian imports (Figure 18).

Although additional expansion of LNG import capacity beyond the expansion plans already announced is not projected to occur in the reference case or in the high LNG cost case, additional expansion is projected in the carbon dioxide emissions limit case and the low LNG cost case. In the carbon dioxide emissions limit case, prices reach a level that triggers expansion at existing U.S. receiving facilities and new facility construction, and net LNG imports are projected to increase to 1.35 trillion cubic feet, or 3.7 percent of demand, by 2020. Beginning in 2010, when natural gas wellhead prices peak at $3.81 per thousand cubic feet, expansion begins at existing facilities beyond what has already been announced, and some new construction begins in the South Atlantic region. Partly in response to the availability of the new supply, average lower 48 wellhead prices fall back to $3.37 per thousand cubic feet by 2015. With the demand for natural gas continuing to increase, prices change direction and increase to $3.72 per thousand cubic feet in 2020. The total increase in sustainable LNG receiving terminal capacity by 2020 includes an increase in capacity over existing and proposed capacity at the four U.S. terminals of 526 billion cubic feet.

In the low LNG cost case, lower LNG costs trigger expansion at existing U.S. LNG receiving facilities and construction of new facilities. LNG is projected to meet 4.7 percent of total U.S. natural gas demand by 2020 (as compared with 0.7 percent in 2000 and 2.3 percent in 2020 in the high LNG cost case). The projected level of expansion in the low LNG cost case exceeds that in the carbon dioxide emissions limit case by 912 billion cubic feet. Total LNG imports in 2020 are projected to increase to 1.8 trillion cubic feet in the low LNG cost case, double the 0.9 trillion cubic feet projected in both the carbon dioxide emissions limit case and the reference case. The level of LNG imports grows steadily each year, leveling off in 2016 and then increasing again beginning in 2019. Average wellhead prices in the low LNG cost  case reach a high of $3.81 per thousand cubic feet in 2009 and 2010, which leads to expansion and new construction of LNG receiving facilities in both the South Atlantic and West South Central regions. The projected increase in LNG imports has an immediate effect on prices, which fall to $3.32 per thousand cubic feet by 2016 in the low LNG cost case. Prices then begin increasing, reaching $3.59 per thousand cubic feet in 2019, when LNG capacity again begins to increase.

Conclusion

The results of EIA’s analysis suggest that increased imports of LNG could have a positive effect on U.S. natural gas markets, especially in an environment of high demand. LNG can meet demand that otherwise would have to be met by higher cost sources, thus tempering price increases. If only a fraction of the LNG terminal capacity currently proposed is built, LNG could capture a much larger portion of the U.S. import market for natural gas than it holds today. In some regions, LNG could have a proportionately larger impact.

Notes