LNG PROCESS SELECTION CONSIDERATIONS FOR FUTURE DEVELOPMENTS-外文翻译原文.docx

外 文 文 献 翻 译（1）译文题目： LNG PROCESS SELECTION CONSIDERATIONSFOR FUTURE DEVELOPMENTS 学生姓名： 学生学号： U201215847 专 业： 建筑环境与设备工程 指导教师： 2016年 3 月LNG PROCESS SELECTION CONSIDERATIONS FOR FUTURE DEVELOPMENTSJohn B. StoneSenior LNG ConsultantDawn L. RymerSenior Engineering SpecialistEric D. NelsonMachinery and Processing Technology SupervisorRobert D. DentonSenior Process ConsultantExxonMobil Upstream Research CompanyHouston, Texas, USAABSTRACTThe history of the LNG industry has been dominated by the constant search for economies of scale culminating in the current Qatar mega-trains undergoing final construction, commissioning,start-up and operations. While these large trains are appropriate for the large Qatar gas resources, future, smaller resource developments will necessitate different process selection strategies. The actual LNG process is only one of many factors affecting the optimal choice. The choice of equipment, especially cryogenic heat exchangers and refrigerant compressors, can overwhelm small differences in process efficiencies. ExxonMobil has been developing a dual mixed refrigerant (DMR) process that has the potential of offering the scalability and expandability required to meet the needs of new project developments, while also maximizing the number of equipment vendors to allow broader competition and keep costs under control. The process will also have the flexibility to accommodate a wide range of feed compositions, rates, and product sales requirements.BACKGROUNDThe startup of the 7.8 million tonnes per year (MTPA) trains in Qatar mark the most recent pinnacle in the search for economies of scale in the LNG industry. However, theapplication of these very large trains for general LNG applications is very limited. To produce this amount of LNG requires 42 MSCMD (1500 MSCFD) of feed gas. What is often overlooked in the discussion of large LNG trains is that a resource of about 370 GCM (13 TCF) is needed to support the operation of one such train over a 25-year life. This is nearly as large as the Arun field in Indonesia 425 GCM (15 TCF), which was the backbone of the LNG plant development in that region. For new LNG developments that are often built with a minimum of two identical trains, a truly world-class resource class of 750 GCM (26 TCF) would be required. Even for resources capable of supporting such large trains, very large gas treating and preparation trains with a minimum of parallel equipment are also needed to ensure that economies of scale are not lost in the non-LNG facilities. Given the limited supply of gas resources capable of supporting these large trains, future projects will need to find ways to maintain some cost advantages at smaller capacities. One way to do this is to improve the project execution by selecting a process that gives the maximum flexibility for utilizing compressors, heat exchangers, and drivers with multiple competing vendors. Another desirable feature is using refrigerant as a utility to allow for facilitated expansion if there is a possibility that several resources can be staged for expansion trains.PROCESS COMPARISONLNG process selection has often been highly influenced by the specific power consumption, i.e., refrigerant compression power divided by the train capacity. This is certainly an important parameter, since refrigerant compressors are the largest single cost and energy consumption components in an LNG train. Conventional wisdom would be that lower specific power consumption would result in lower refrigerant compression costs and additional LNG production from a fixed feed gas rate. In actuality it is a more complicated picture. Figure 1 plots the specific power consumptions for a variety of liquefaction processes against the number of cycles employed based on consistent conditions.SMR - Single Mixed RefrigerantC3MR - Propane pre-cooled Mixed RefrigerantC3MRN2 - Propane pre-cooled Mixed Refrigerant plus Nitrogen expander cycleCascade - Pure propane, ethylene, and methaneDMR-SWHE - Dual Mixed Refrigerant with single pressure levels and SWHEsDMR-BAHX - Dual Mixed Refrigerant with multiple pressure levels and BAHXsTMR - Triple Mixed RefrigerantFigure 1 - Process Specific Power ComparisonIn general, mixed refrigerant processes are more efficient than pure component processes and additional cycles improve efficiency. However, both of these efficiency improvements come at the expense of increased process complexity.Another factor that complicates the picture above is that it only considers a process comparison and not a refrigerant compressor or driver comparison. Differences in compressor efficiency, the need for a speed-increasing gear, or driver efficiency can overwhelm some of the differences shown. Considerations for the generation and distribution of electric power for motor driven LNG processes can further complicate the comparison.The LNG industry is changing in a number of areas that can also impact the selection of the best liquefaction process. While stick-built LNG plants are still the norm, modularization of LNG facilities are more attractive for offshore applications or where labor costs are very high and/or productivity is low. Modular construction is routinely applied for offshore oil processing. However, oil processing is much simpler than LNG production and process selection is generally not an important consideration. All these factors point to the need for more compact, lighter mechanical designs.Another important future consideration is the increasing need to reduce greenhouse gas emissions. Aeroderivative gas turbine drivers are an obvious choice for higher thermal efficiency or modular application but are not available in sizes as large as industrial gas turbines. Consequently, a process suitable for large 95 MW industrial gas turbines may not be well suited for a 35 MW aeroderivative gas turbine. Combined-cycle power generation is another option for achieving increased thermal efficiency and can be adapted to any of these processes, but is not well suited for modular construction or for offshore application due to the additional weight of motors, generators and distribution equipment as well as limited aeroderivative gas turbine choices for very large (>100MW) power generators.The value of thermal efficiency can also become a more important process selection criterion when the feed gas to the LNG plant is relatively expensive or supply is limited. An efficient process can allow for a reduced cost development plan through a lower gas rate, or extend the gas production plateau from the reservoir to make a more profitable project.IMPACT OF EQUIPMENT COSTSOur process research comparing liquefaction processes has demonstrated that the primary difference in the costs for the different liquefaction processes is the choice of equipment utilized. Process licensors tailor their process to make it capital and thermally efficient given the owners' preferences and constraints. However, they do not always have control over the cost (both equipment and installation) in the final analysis.Gas TurbinesGas turbine costs exhibit a reasonably high economy of scale. Large industrial gas turbines are the least expensive, but their cost advantage is lost in a modular or offshore environment due to their large weight and space requirements. Therefore, aeroderivative based designs will be more attractive. However, once the drivers are selected, then a process that is flexible in allowing a shift in refrigerant power loads to maximize the utilization of the available turbine power would be the best process. A multiple mixed refrigerant process, without the fixed atmospheric boiling temperatures characteristic of pure refrigerants, has the flexibility to allow such shifting. An alternative to mechanical-drive gas turbines would be electric motor drives with very large power generators for economy of scale. In this case, gas turbine costs would be lower because of standard designs, multiple manufacturers, and possibly greater economies of scale, but there would be additional costs for motors, spare turbine generators and power distribution which can reduce the overall efficiency in a simple cycle configuration. This efficiency loss can be overcome with combined cycle, but in simple or combined cycle the net result is usually a higher capital cost. The implementation of an all-electric drive configuration is even more difficult at reduced economies of scale where the use of larger lower cost turbines becomes problematic due to difficulties managing the dynamic response to electrical load changes spread across fewer units. In the end though, the choice of an all-electric drive configuration is condensed to a trade off between a higher capital cost and the increased plant availability that electric motors can achieve.CompressorsCompressors exhibit a very high economy of scale. Refrigerant compression costs areprimarily a function of the number of compressor cases needed. Consequently, it is important to minimize the number of compressor cases. Likewise it is important to limit the required rotor diameter of the centrifugal compressor wheels to stay within the capabilities of multiple vendors. This requires limiting the volumetric flow rate feeding these compressors through reduced refrigerant circulation or higher refrigerant suction pressure. Again the dual mixed refrigerant process allows the process designer the flexibility to optimize the compressor inlet suction volumetric rate to maximize throughput within the design capability of at least four suppliers.Heat ExchangersCryogenic heat exchanger costs are primarily related to the surface area supplied. There will always be a tradeoff between exchanger area and compressor power to reach a minimum overall cost. Spiral Wound Heat Exchangers (SWHEs) are the standard cryogenic heat transfer equipment for the base load LNG industry. SWHEs have an excellent service record in LNG service; however, they are expensive, have long delivery times, and are limited to two manufacturers.Another option is to use brazed aluminum heat exchangers (BAHXs), which have a lower cost per unit area than SWHEs, and can be aggregated easily into blocks of surface area to meet large heat transfer requirements effectively. BAHXs also easily accommodate side-streams which allow refrigerant systems with multiple pressure-levels to be readily incorporated. BAHXs have been demonstrated in LNG service in cascade processes and smaller mixed refrigerant processes. BAHXs are built in small units (cores) typically manifolded together and insulated in a cold box. A typical design would require about 30 cores to provide the exchanger area needed for a 3 MTPA LNG train. These exchangers are available from five manufacturers. Having multiple vendors ensures not only competitive prices, but also flexibility in acquiring the exchangers in time to meet the project schedule.PROCESS SELECTIONWhat would an ideal liquefaction process look like? It would be a DMR process such as shown in Figure 2 below for low specific power consumption and flexibility to optimize compressor design. Including multiple levels of cooling in the warm mixed refrigerant circuit allows more flexibility to meet compressors volumetric limitations. ExxonMobil has synthesized these traits with known liquefaction processes, adding our own proprietary optimizations resulting in this configuration.Figure 2 - ExxonMobil DMR-BAHX Process SchematicIt would utilize BAHX exchangers to provide: Multiple manufacturers for cost and schedule benefits, Economic scale up over a wide range of throughputs, Ease of modularizationThe BAHX exchangers would be protected from operational and design problems associated with multi-phase maldistribution by effecting refrigerant separation at each pressure level of the warm refrigerant and feeding only liquids to the BAHX cores while bypassing the vapor back to the compression system.It would utilize gas-turbine-driven centrifugal compressors large enough to capture the economy of scale available but small enough to ensure that multiple compressor vendors are capability of supplying the sizes needed.The results of our LNG process research applying these principles to a potential LNGdevelopment are shown in Figure 3. By using BAHXs and a dual mixed refrigerant process to match the best fit of compressors and drivers available from multiple vendors, the resulting process will have a lower specific power requirement, and could have a lower capital cost than traditional technologies. The DMR process with brazed aluminum heat exchangers shows a unit cost advantage across a broad range of plant capacities and optimizes the trade-offs of efficiency versus cost for a wide size range (3-6 MTPA) of plants. EFFICIENT EXPANSIONLNG plants have long benefited for profitable expansion trains, typically provided from the same large resource. While the number of discovered large fields available for multi-train development is shrinking, there is still the potential for economical expansion from nearby smaller resources. In many cases these other fields cannot be aggregated into one large project for a variety of reasons: difficulty aligning several commercial interests, waiting on reduced development costs for more difficult resources, or near-field discoveries identified after the LNG project is underway. For all of these reasons it is desirable to have an easily expandable LNGplant.Treating refrigerant as a utility is a way to maximize the expandability and reliability of a multtrain facility. In this configuration all of the refrigerants that serve the same process function are combined into a single header and delivered as required to the LNG liquefaction sections. The refrigerant as a utility concept can be done with any liquefaction process, but is most suited for dual mixed refrigerants where the refrigerant return pressures can be higher resulting in smaller piping for distribution of refrigerant across the LNG plant. Figure 4 shows one such configurationTreating refrigerant as a utility has several benefits: The trains do not necessarily need to be the same size, leading to customizableexpansion to match commercial needs. All the refrigerants can be re-tuned to match changes in feed gas composition tomachinery limits as new gas supplies are brought on-line. Any spare capacity identified by testing after start-up can be designed for and utilized during expansion. A mixture of gas turbine, steam turbine, and motor drivers can be used giving moreflexibility to the driver selection and energy utilization. In the event of driver failures, the liquefaction train may be able to turn-down instead of shu