This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Natural gas is the fastest growing primary energy source. Its use is projected to double between 1999 and 2020. Natural gas reserves are plentiful around the world, but many of them currently are too small or too remote from sizable population centers to be developed economically. Estimates of remote or stranded gas reserves range from 40 to 60% of the world's proven gas reserves. This massive global gas reserve is largely untapped, and conventional means of development face logistical and economic barriers.
Figure 1.1: World primary energy demand 
Natural gas is of little value unless it can be brought from the wellhead to the customer, who may be located several thousand kilometers from the source. As natural gas is relatively low in energy content per unit volume, it is expensive to transport. The most popular way to move gas from the source to the consumer is via pipelines. For onshore and near-shore gas, pipeline is an appropriate option for transporting natural gas to market. However, as transportation distances increase, pipelines become uneconomical; Liquefied Natural Gas (LNG), Gas to Liquids (GTL) and chemicals are more viable options.
With the world's second largest proved gas reserves, Iran has great potential to export gas to markets in Europe, Asia, and India by pipeline and as LNG. The government is considering at least four projects, each of 390 to 490 Bcf (8 to 10 million tons) per year, to process reserves in the South Pars-North field in partnership with companies in Europe and Asia.
Total has concluded an agreement with the National Iranian Oil Company (NIOC) setting the framework in place for the future Pars Liquefied Natural Gas (LNG) project and its main commercial terms. The agreement organizes the relationship between the Pars LNG, in charge of liquefaction activities, and block 11 of South Pars (SP11) attributed to Total (60%) and Petronas (40%) to supply the LNG plant.
The Pars LNG joint venture is a partnership between NIOC (50%), Total (30%) and Petronas (20%). The project is designed for an initial capacity of two trains of 5 million tons of LNG per year each. This agreement allows the start of engineering studies for both the LNG plant and the SP11 development. These studies will be performed during 2005 and should lead to the decision to launch the project end of 2005 early 2006. 
1.2 What is LNG?
Liquefied natural gas (LNG) is primarily methane, which has been cooled and liquefied. LNG is colorless, odorless, non-corrosive, less dense than water, and non-toxic. LNG vapor typically appears as a visible white cloud since its cold temperature causes moisture in the air to condense. LNG (the liquid itself) is not flammable or explosive. Liquefying natural gas reduces its volume by more than 600 times, making it more cost effective to store and transport from remote locations.
Natural gas is composed primarily of methane, but may also contain ethane, propane and heavier hydrocarbons. Small quantities of nitrogen, oxygen, carbon dioxide, sulfur compounds, and water may also be found in natural gas. The figure below provides a typical natural gas composition. The liquefaction process requires the removal of some of the non-methane components such as water and carbon dioxide from the produced natural gas to prevent them from forming solids when the gas is cooled to about LNG temperature (-256Â°F). As a result, LNG is typically made up mostly of methane as shown in the figure below. 
Figure 1.2: Typical natural gas composition
Figure 1.3: Typical liquefied natural gas composition
1.3 Future Energy Demands for LNG
Today's LNG developments are motivated by a sustained worldwide rise in natural gas consumption, exceeding the growth of other fuels. Its current share of total primary energy consumption is 23.8%, and has risen in all regions over the last ten years. This not only reflects the ability of natural gas to compete directly on a cost basis, but also the growing recognition of its environmentally friendly characteristics that has replaced other fossil fuels in urban areas.
The LNG market currently covers five continents (Africa, Asia, Australia, Europe, and North America), but consists of only a small group of players, nine importers and nine exporters. The double-digit growth figures for imported LNG for a selection of countries are presented in figure (1.4). Similarly high growth figures have been reported for LNG exporters, figure (1.5).
Worldwide, there was a 10.3 % growth in LNG traded from 1996 to 1998 compared to only 3.5% for pipeline movements of natural gas in the same period. Greater flexibility as a fuel, combined with increased spot trading and new technology places greater demands on operators to seek cost effective recovery methods and reliable LNG supplies. Furthermore, the market growth presents obvious opportunities to maximize LNG production profitability. In recognition of the changing market trends and the opportunity to maximize profitability, operators are exploring alternative solutions for the economical production of LNG. 
Figure 1.4: LNG importing nations 
Figure 1.5: LNG exporting nations 
1.4 History of LNG Production Plants
Gas liquefaction dates back over a century. The British scientist Michael Faraday experimented with liquefying different gases, including natural gas in the 19th century. Thereafter, the German scientist, Carl von Linde, built the first compressor refrigerator in 1873. He liquefied oxygen in 1895 using a spiral-wound heat exchanger. The first LNG liquefaction plant was built in Virginia in 1912. Godfrey Cabot submitted a US patent for the barge transport of LNG in 1914. In 1917, gas was liquefied in Texas to extract helium. During the 1920 and 1930, LNG was produced in the USA as a means of storing natural gas in the USA. The first commercial LNG plant (peak shaving) was constructed in Cleveland, Ohio in 1941. Floating LNG production first occurred in 1951 from a barge-mounted plant in Louisiana. The first LNG export was from this facility to Canvey Island in the UK in 1959.
Shell purchased a 40 percent interest in Constock which became Conch International Methane ("Conch"). Conch, in turn, took a 40 percent interest in Compagnie Algerienne de Methane Liquide ("CAMEL"). The CAMEL plant started up at Arzew, Algeria in 1964. It utilized the Technip-Air Liquide ("TEAL") cascade process with steam turbine-driven compressors and seawater-cooled condensers. The three train plant had a capacity of 1.1 million metric tones per annum ("MMtpa"). It provided the first commercial deliveries to the UK and France. The plant was expanded in 1981. In 1969, the 1.4 MMtpa Kenai plant in Alaska started up. This was the first application of the Phillips Petroleum Company ("Phillips") Optimized Cascade Cycle. This single train process was the first to utilize combustion gas turbines ("CGT"). Kenai also used Phillips plate fin heat exchangers. This technology was not used again for thirty years. Since the mid-1990's, the Phillips Optimized Cascade Process has been used for the first four trains of the Atlantic LNG plant, the first two trains of Egyptian LNG and Bayu Undan LNG. Atlantic Train 4 is designed for a nominal five million metric tones per annum production capacity under tropical conditions.
Development of the Marsa el-Brega LNG plant in Libya was led by Exxon. The plant started operation in 1970. It was the first plant to utilize the single mixed refrigerant ("SMR") process from APCI. The process is less complex than the cascade in that it reduces the number of compressors and heat exchangers, but it is also less efficient. The Marsa el-Brega LNG plant employed four 750,000 metric tone per annum liquefaction trains, each designed around a single spool-wound MCHE and compressor string. The mixed refrigerant comprises methane, ethane, propane, etc.
Shell's Brunei LNG plant entered operation in October 1972. It had five 1.1 MMtpa trains. Brunei was the first application of the PMR process by APCI. Shell represents that its philosophy has been to evolve technology further with each new plant. It represents that the thermal efficiency of the PMR process approached 90 percent. Brunei has recently completed a major overhaul that included the replacement of APCI main cryogenic heat exchangers with the competing Linde units.
In November 1972, the Skikda plant in Algeria started up. This used a TEAL SMR process. Steam turbines were used to drive the compressor strings of three 1.0 MM trains. The plant was expanded in 1981. This used a Pritchard designed SMR process. Skikda was badly damaged by an explosion in January 2004. In 1977, the Botang LNG Plant in Indonesia was started up by a joint venture of Pertamina and Roy M Huffington Inc. This plant utilized the PMR process in two 2.0 MMtpa trains. Once again, steam driven turbines were used to drive the compressors. This used a large MCHE. The plant was expanded in 1983, during which year an explosion occurred in a MCHE. In 1978, the Mobil-led Arun LNG plant started up in Indonesia. This plant used General Electric Frame 5 CGT with a dual shaft and a modified PMR process. Plant expansions started up in 1983 and 1986.
The Malaysian LNG Satu plant came on stream in 1983. Sponsors of the project comprise Petronas, Shell, Mitsubishi and Sarawak. It was the last Greenfield plant to use steam turbine drives for the compressors.
In June 1989, the North West Shelf LNG plant in Western Australia came on stream. The North West Shelf Joint Venture partners comprise Woodside (the operator), Shell, BHP, BP, Chevron Texaco and Mitsui/Mitsubishi. This PMR facility used four Frame 5 CGT, each siterated at 28 MW. Environmental constraints on water cooling led to the use of air cooling for the condensers; air-cooling of condensers has become commonplace in tropical environments. Shell was an advisor to this project. Kellogg/JGC and Kaiser were the joint venture EPC Contractor. The MLNG Dua Expansion started up in 1995. Sponsors of the project comprise Petronas, Shell, Mitsubishi and Sarawak. This project was the first project to use the 80 MW GE Frame 7E CGT. The Project employs PMR technology. It used a Frame 7E single shaft turbine to drive two compressors on the mixed refrigerant service. A Frame 6 drives the propane pre-cooling cycle. Compared to Frame 5s, this delivered a 15 to 25 percent reduction in capital cost and a 10 to 15 percent reduction in fuel consumption. The EPC contractor was a Kellogg/JGC joint venture. The Qatargas LNG Project, the first LNG project to be built in Qatar entered operation in 1996; the second train was commissioned in 1998. Sponsors comprise QGPC, Mobil, Total, Marubeni and Mitsui. Chiyoda was the EPC Contractor for the LNG plant. The facility uses PMR technology. The Ras Laffan ("RasGas") LNG Project entered service in 1999. This facility is located adjacent to Qatargas with Ras Laffan Industrial City. A second train came on stream in 2000. Sponsors comprise QGPC, Mobil, Itochu, Nisso Iwai and Korea Gas. The EPC Contractor was a Kellogg Brown & Root/JGC joint venture.
The Nigerian LNG Project entered service in 1999. Sponsors of this project comprise NNPC, Shell, Elf and AGIP. It uses PMR technology. The two trains have a combined capacity of 5.8 million metric tones per annum. A joint venture ("TSKJ") was the EPC Contractor.
The Atlantic LNG Project entered service in 1999. This was a single train plant that utilized the Phillips Optimized Cascade Process. Project sponsors comprise BP, BG, Repsol, Cabot and NGC. Trains 2 and 3 came on-stream in 2002 and 2003 respectively. At present, a fourth train is under construction. Bechtel has been the contractor for all four trains.
The Oman LNG Project entered service in 2000. This has two 3.3 MMtpa trains. The Project employs PMR technology. The EPC Contractor was Chiyoda. The MLNG Tiga Project, a 3.9 MMtpa train, utilizing two Frame 7 EA CGTs, was scheduled to enter service in 2003. NWS Train 4, a 4.2 MMtpa train utilizes two Frame 7 EA CGT drivers. As noted above, it is the first application of the Linde aluminum spool-wound MCHE in LNG service. Projects currently under construction are shown in Table (1.1) while projects at the planning stage are shown in Table (1.2). [3, 19]
Table 1.1: Current LNG projects 
LNG plant current project
Selected process technology
Largest train (MMtpa)
Planned start up
ALNG Train 4
Phillips Optimized Cascade
NWS Expansion Train 4
RasGas Expansion Trains 3 and 4
Egypt Damietta LNG
Nigeria Plus Train 4 and 5
Statoil-Linde MFC Process
Egyptian LNG Trains 1/2
Phillips Optimized Cascade
Damietta LNG, Egypt
Phillips Optimized Cascade
Table 1.2: Proposed LNG projects 
LNG plant proposed project
selected process technology
Largest train (MM mtpa)
Planned start up
Mariscal Surce, Venezuela
1.5 LNG versus Pipelines
1.5.1 Indicative Pipelines and LNG Costs at Full Utilization
The changes in costs also affect the relative attractiveness of the pipeline and LNG options. In determining the most economic transportation method for a given supply route, distance and the volumes transported are the key factors. For short distances, pipelines - where feasible - are usually more economic. LNG is more competitive for long distance routes, since overall costs are less affected by distance. The normal breakeven distance for a single-train LNG project against a 42Ë onshore pipeline (not allowing for transit costs) is around 4,500 km at a cost of around 1.60$/millionBtu. The breakeven point has tended to fall over the last decade, as LNG costs have fallen faster than pipeline costs. But technology advances have made possible short-distance offshore pipelines where previously LNG had been the only viable option.
For large deliveries (around (3010)9 m3/year), the transport of gas by HP pipelines appears very much competitive. For long distances, LNG appears competitive for capacity below (10109) m3/year. For Middle East supply to Europe for instance (between 4,500 and 6,000 miles), the LNG allows a cost saving of up to 30% with respect to HP pipe technology. Therefore, LNG could be preferred for small field's exploitation on along distance transportation (Figure 1.6 & 1.7). 
Figure 1.6: Pipes/LNG competition for (30109) m3/year capacity 
Figure 1.7: Pipes/LNG competition for (10109) m3/year capacity (us$/million Btu) 
1.5.2 Influence of Market and Financing Factors
However, besides the economics - at full utilization - of each solution, there are three main elements that influence the specific costs as shown on the graphs:
â€¢ Utilization/market size
â€¢ Financing conditions
â€¢ Development of level and size of investment
The development of the three elements has favored the economics of LNG versus pipeline, shortening the breakeven distance:
â€¢ The new projects on the markets are much smaller than they were in the 1980s. Huge projects, like Troll in Norway or SoyuzGasexport or Yamal delivering Russian gas to Europe with 20 or more 109 m3/year are increasingly rare today. Projects of (5109) m3/year are more typical. However the levelized cost shown on the graphs is based on a throughput of 10 to 30109 m3/year.
â€¢ At a (5109) m3/year throughput, specific pipeline transportation costs are much higher than indicated by the line representing the cost of a 42Ë onshore pipeline.
â€¢ Financing of LNG is lower risk (e.g. Oman LNG with an A- rating by Standard & Poor's).
In the case of pipelines, each country border and ethnic enclave crossed ads to the risk premium, which increases the levelized costs. In addition, evolution of investment costs is coming down more favorably for LNG than for pipelines, where cost reductions are mainly linked to an increase in capacity that is difficult to absorb by the market.
In practice, however, LNG projects do not often compete directly against pipeline projects for the same supply route. Competition to supply a given market is usually between different supply sources, either by pipeline or LNG. For example, Trinidad LNG competes against Algerian gas supplied through the Maghreb pipeline to Spain.
1.6 The LNG Value Chain
A typical gas monetization value chain option is shown in Figure (1.8). The steps involved in the monetization of gas via LNG include the following: [3, 5]
Gas Production and field processing
Onshore gas treatment
Gas conversion via liquefaction
End use as fuel (power generation, fertilizer Industry, gas distribution, etc.)
Figure 1.8: LNG value chain 
1.6.1 Exploration and Production
Figure 1.9: Exploration and production
The exploration and production of gas is the starting point for all gas utilization options.
The source of natural gas feed to the LNG plant could be either associated gas or non associated gas. Natural gas from gas fields typically is a mixture of hydrocarbons ranging from methane to heavier hydrocarbon molecules. Methane is invariably the dominant component. Ethane and heavier hydrocarbons are categorized as Natural Gas Liquids (NGL). Liquefied Petroleum Gases (LPG) components refer to a mixture of propane and butane. The quantity of NGL in the gas depends on the type of reservoir from which it originates. Gases with low NGL content are referred to as 'lean gas' while gases with high NGL content are referred to as 'rich gases'. The gas production step includes some field processing, depending on the nature of the gas source and the requirements for pipeline transport to the liquefaction site. Typically, field processing is needed to avoid hydrocarbon liquid drop-out, hydrate formation or corrosion in the pipeline to the liquefaction site.
1.6.2 Onshore Gas Treatment
The gas from the reservoir may also contain components such as nitrogen, carbon dioxide and sulfur compounds. The feed gas has to be treated for removal of impurities before it can be liquefied. Hence, onshore gas treatment is required for the removal of NGL's and impurities to meet the specifications set by the LNG buyers as well as the requirements of the LNG liquefaction process. Typical LNG product specifications are listed in Table (1.3).
Table 1.3: Typical LNG product specifications 
3 to 4 ppm
30 milligrams per normal cubic meter
0.01 micrograms per normal cubic meter
1 mol %
6 - 8 mol %
3 mole %
Butane and heavier
2 mole %
Pentane and heavier
1 mole %
High heating value
~ 1050 Btu/SCF (Europe and USA)
>1100 Btu/SCF (East asia)
The onshore gas treatment typically comprises of gas reception facilities, acid gas removal & disposal section, gas dehydration, mercury removal and particle filtration.
184.108.40.206 Gas Reception Facilities
This section is provided for the removal of liquids produced in the gas gathering system resulting from condensation as well as reduction in pressure of the fluid. The liquid may be either sent to a condensate stabilizer or to the fractionation section of the liquefaction plant. If a condensate stabilizer is provided, then the overhead gas stream from the stabilizer is sent to the feed gas for the liquefaction plant. The pressure of the LNG feed is adjusted in this section to meet the requirements of the liquefaction facility. If the pressure is lower than that required for liquefaction, then a compression system may have to be added. The pressure may have to be let down if it is higher than that required for the liquefaction plant. The selection of the pressure of the liquefaction plant is one of the key optimization parameters for the facility.
220.127.116.11 Acid Gas Removal and Disposal Section
The purpose of this section of the plant is to remove acid gases (CO2 and sulfur containing components) from the feed gas. The extent of removal of these components is influenced by both the final LNG specification and the requirements of the liquefaction process. CO2 can freeze on exchanger surfaces, plugging lines and reducing efficiency. The feed gas is typically treated to achieve H2S content below 4 ppmv and a CO2 content of 50 ppmv in the feed gas. The total sulfur content is usually limited to about 30 milligrams per normal cubic meter. There are various acid gas treatment processes available for the removal of H2S and CO2 from natural gas.
These processes include chemical solvents, physical solvents, adsorption processes, and physical separation. However, only chemical solvents and physical solvents, or a combination of these two, have been used in existing baseload LNG facilities.
The composition of the inlet sour gas to the LNG plant could vary significantly from one location to the other. Optimal selection of the treating process is dependent on the type and level of impurity in the natural gas stream. The processes most widely applied for LNG plants include amine based processes ranging from conventional MEA to proprietary formulated solvents such as Shell Sulfinol; Union Carbide's Ucarsol and activated MDEA from BASF. Off gases from the acid gas treatment plant may be sent to a Claus unit for sulfur recovery provided that the volumes of sulfur that can be recovered are sufficiently large to justify the investment. When the quantity of sulfur is too low the off gas is often just incinerated or needs to pass though an acid gas enrichment process. Figure (1.10) shows the application of various natural gas acid gas removal technologies in existing baseload LNG plants. Overall the Sulfinol and MEA process have been the most widely applied acid gas removal processes with approximately 70% of the installed LNG capacity. Due to its corrosive nature and high regeneration heat requirement, MEA is losing popularity. In contrast MDEA, usually applied with a proprietary activator to facilitate gas pickup, tends to be less corrosive and has lower regeneration heat duties than other amines. Thus the popularity of MDEA based processes, such as UCARSOL or aMDEA, is growing.
Figure 1.10: Choice of acid gas removal technology for LNG plants 
18.104.22.168 Dehydration Step
The dehydration section removes water from the feed gas. Water vapor must be removed from the gas to prevent freezing in the liquefaction section of the plant that operates at cryogenic conditions. The gas is usually first cooled to a point above the hydrate formation temperature using air or water coolers and refrigerant. Free liquid water is separated and the remaining gas is fed to a molecular sieve adsorption unit for removal of water vapor below 1 ppmv. The key optimization options for the dehydration process include number of beds, type of bed regeneration, source of regeneration gas and source of the regeneration heat.
22.214.171.124 Mercury Removal Section
Mercury, generally present in trace amounts in the feed gas, attacks piping and equipment made from aluminum and aluminum compounds. Aluminum is a common construction material for low temperature exchangers used in LNG plants. Mercury must therefore be removed prior to the feed gas entering the cryogenic sections of the LNG plant to a maximum content of 0.01 micrograms per normal cubic meter (~0.013 ppb by weight).
Plant feed from gas field
Condensate to storage
NGL to storage
LNG to ships
Figure1.11: Components of a LNG liquefaction plant
Almost every LNG plant will have a mercury removal vessel installed. The mercury removal vessel is usually placed downstream of the dehydration unit. A common method for removal of mercury is by reaction with elemental sulfur to form a sulfide. The sulfur is supported on a high surface area solid carbon bed. This is a non-regenerative process and the spent carbon bed must be disposed to a land fill or is returned to the catalyst vendor for reclaiming.
126.96.36.199 Particle Filtration
Filtering of the gas stream following the mercury removal unit is essential to prevent carrying particles into the liquefaction section of the plant and hence prevent equipment plugging.
1.6.3 LNG Liquefaction
Figure1.12: LNG liquefaction
Feed gas to the liquefaction plant comes from the production field. The contaminants found in produced natural gas are removed to avoid freezing up and damaging equipment when the gas is cooled to LNG temperature (-256Â°F) and to meet pipeline specifications at the delivery point. The liquefaction process can be designed to purify the LNG to almost 100 percent methane.
The liquefaction process entails cooling the clean feed gas by using refrigerants. The liquefaction plant may consist of several parallel units ("trains"). The natural gas is liquefied for shipping at a temperature of approximately -256Â°F. By liquefying the gas, its volume is reduced by a factor of 600, which means that LNG at -256Â°F uses 1/600th of the space required for a comparable amount of gas at room temperature and atmospheric pressure.
LNG is a cryogenic liquid. The term "cryogenic" means low temperature, generally below -100oF. LNG is clear liquid, with a density of about 45 percent the density of water.
The LNG is stored in double-walled tanks at atmospheric pressure. The storage tank is really a tank within a tank. The annular space between the two tank walls is filled with insulation. The inner tank, in contact with the LNG, is made of materials suitable for cryogenic service and structural loading of LNG. These materials include 9% nickel steel, aluminum and pre-stressed concrete. The outer tank is generally made of carbon steel or pre-stressed concrete.
1.6.4 LNG Shipping
Figure 1.13: LNG shipping
LNG tankers are double-hulled ships specially designed and insulated to prevent leakage or rupture in an accident. The LNG is stored in a special containment system within the inner hull where it is kept at atmospheric pressure and -256°F. Three types of cargo containment systems have evolved as modern standards. These are:
The spherical (Moss) design
The membrane design
The structural prismatic design
Figure 1.14: Three types of cargo containment systems 
The figure above shows that currently most of the LNG ships use spherical (Moss) tanks, and they are easily identifiable as LNG ships because the top half of the tanks are visible above the deck. The typical LNG carrier can transport about 125,000 - 138,000 cubic meters of LNG, which will provide about 2.6 - 2.8 billion standard cubic feet of natural gas. The typical carrier measures some 900 feet in length, about 140 feet in width and 36 feet in water draft, and cost about 160 million $. This ship size is similar to that of an aircraft carrier but significantly smaller than that of Very Large Crude oil Carrier (VLCC). LNG tankers are generally less polluting than other shipping vessels because they burn natural gas in addition to fuel oil as a fuel source for propulsion.
The LNG shipping market is expanding. According to LNGOneWorld, as of December 2002, there were 136 existing tankers, with 57 on order.
Figure 1.15: Existing LNG tankers 
Twelve new LNG tankers were ordered in 2002 of which eight tankers have been delivered. About 20 percent of the fleet is less than five years old. The LNG tanker fleet size is estimated to continue to grow to 193 tankers by 2007.
1.6.5 Storage and Regasification
Figure 1.16: Storage and Regasification
To return LNG to a gaseous state, it is fed into a regasification plant. On arrival at the receiving terminal in its liquid state, LNG is pumped first to a double-walled storage tank, similar to those used in the liquefaction plant, at atmospheric pressure, then pumped at high pressure through various terminal components where it is warmed in a controlled environment. The LNG is warmed by passing it through pipes heated by direct-fired heaters, seawater or through pipes that are in heated water. The vaporized gas is then regulated for pressure and enters the pipeline system as natural gas. Finally, residential and commercial consumers receive natural gas for daily use from local gas utilities or in the form of electricity.
1.7 LNG Economics
LNG projects are very much capital intensive. The cost of the entire chain from wellhead to the receiving terminal can be around 4 billion US$. As in the case of pipelines, economies of scale are very significant:
Liquefaction plants typically consist of one or two processing trains. The economic size of each train is now about 3 to 3.5 million tones per year. With this size of project, the capital cost of just the LNG production facility is in the 1-2 billions$ range. Adding a second train once a plant is built can reduce the overall unit cost of liquefaction by 20-30%. A single-train plant normally costs around 1 billion$, although actual costs vary geographically according to land costs, environmental and safety regulations, labor costs and other local market conditions.
Technological progress achieved in the past decades has led to a sharp decrease in investment and operating costs of liquefaction plants.
Figure 1.17: LNG project cost components 
LNG costs vary considerably in practice, largely as a function of capacity, particularly the number of trains in liquefaction plants and shipping distance. 
1.8 Liquefaction Capacity
It is difficult to give precise numbers for the production capacity of liquefaction plants. Figures cited in the press or in company documents can show significant differences depending on whether they refer to the design capacity of the plant or actual output, which itself can vary from year to year depending on operational factors such as the timing of maintenance.
LNG plants are designed to produce the volume of LNG required by the project sponsors; this is often referred to as the 'design' or 'nominal' capacity. Most plants produce significantly more than their design capacity. This was especially true in the early days of the LNG industry, when security of supply was very important and sponsors and contractors wanted to ensure that the plant would not perform below its design capacity. Hence, the facilities were often over-sized and suboptimal operating conditions were assumed. When these plants came on-stream, they then produced considerably more than their design capacity would have indicated. As experience in designing and operating LNG plants has increased, the degree of over-design has reduced, but plants coming on stream in recent years still typically operate at 10 percent above their design capacity. After plants come into operation, bottlenecks that constrain production are often identified. 'Debottlenecking' the facility to remove these constraints can provide a low cost option to increase capacity further. As a result, the actual capacity of many LNG plants is over 20 percent higher that the original design capacity.
Cost reductions in liquefaction are currently focused on increasing economies of scale from larger train sizes. Several planned plants, such as Melkoya Island in Norway, Gorgon in Australia and the Gulf of Paria in Venezuela, have train capacities in excess of (4106) ton/year. The additional train planned for the RasGas plant in Qatar will have a capacity of 4.7 106 ton/year. Capacities on planned new trains for the Qatargas' plant range from (5106) ton/year to (7.5 106) ton/year (ExxonMobil). These capacities which could reduce unit construction costs by 25% compared to (3103) ton/year trains should become feasible within the next few years. Further improvement in fuel efficiency and unit investment costs can be expected from larger gas turbines as train size increases. Optimization of design parameters, improved reliability, closed-loop cooling systems, the exploitation of cold-recovery and new heat-exchanger designs under development could yield further cost reductions, as well as better engineering and bidding processes, especially for extensions of existing liquefaction plants. Cost reductions will also result from the use of new equipment and more efficient processes.
Figure (1.20) shows that up to 2000 the trend was to increase liquefaction train capacity to fully benefit from the economies of scale. However, from 2000 onwards there is more spread in train capacity ranging between 3 and 8 Mtpa per train. High capacity trains are the most economic when reserves are abundant and relatively easy to produce, assuming that the market can readily absorb large volume in one go. If not all of these conditions are fulfilled, projects with lower capacity trains can also be economically attractive. 
Figure 1.18: LNG train capacity over the years