This disclosure relates to a process for producing Liquid Natural Gas (LNG) from natural gas feedstock delivered at customary pipeline conditions to LNG process sites.
The LNG industry has its roots in 19th century gas liquefaction science for gasses of all descriptions including air and has matured to its present forms since the first commercial shipping began in the late 1950s/early 1960s. The one dozen or so process methods in commercial use today can generally be divided into two broad categories: the cascade chilling process typical of developments by Conoco Philips, and the mixed refrigerant process primarily commercialized by Linde AG and Air Liquide. Both these methods advocate heat extraction from the process gas by exchange with external chilling loops that involve the expansion chilling principle of Joule Thomson applied to refrigerants entirety.
A third method economically used on smaller installations is derived from the Claude process where fractional side streams of the feedstock expand through pressure restrictors to provide a chilled heat exchanger sink. Passing the remaining main gas stream through the heat exchanger prechills it prior to it then expanding further reduce its temperature. Repeated cycles reduce rate of flow of the main gas stream while providing stepped reduction in temperatures of main gas stream to provide liquefaction of the gas.
The cascade chilling process is the more widely used method amongst large industrial installations, and at one time was the preferred configuration in large scale commercial service. In this process a series of chilling stages use heat exchange from the process gas passing through “cold boxes”. Each of the series of cold boxes is serviced by an externally refrigerated loop employing a refrigerant specifically suited to the particular stage of temperature reduction. The refrigerants suited to progressively colder temperature bands are most often single component propane, ethylene and methane products.
In mixed refrigerant processing selective component refrigerants provide heat transfer refinements aimed at improving overall efficiency in liquefaction. In this process the chilling is again made by heat transfer to external refrigeration loops and tracks closer to the trace line defining the gas to liquid interface conditions than does the cascade chilling process. Chilling conditions are dictated primarily by the methane constituent of the produced LNG. Use of tight coil piping in heat exchangers has in recent times been displaced with the introduction of less expensive plate and fin “printed circuit” exchangers operating with benefits of higher heat flow coefficients associated with high pressure flow streams to further enhance process efficiency.
Mixed refrigerant process plants presently form the majority of new efficient large scale plant proposals worldwide, having abandoned coil piping for heat exchange. The cost benefits to cascade processing having abandoned traditional cold boxes and recently adopted these new configurations of printed circuit heat exchange promises to put them commercially on par with the aforementioned mixed refrigerant proposals.
Modifications of the Claude process forming the third category of installations uses the J-T effect of chilling through pressure drop of the actual feed gas flow as well as in an extracted split flow stream where the gas acts as a refrigerant to external heat exchanger by passing it through a series of pressure reducing orifices/flow restrictors that are variants of system originally proposed by Claude for liquefaction of a multitude of gasses including air.
In such natural gas processing scenarios, expanded split side streams acting as refrigerants in heat exchanger loops, prechill the remaining mainstream stream gas prior to direct cryogenic expansion chilling. Repeated energy intensive recompression and orifice expansion further chills the flowstream towards liquefaction threshold conditions. A point is reached where fall-out in the partial flow mainstream in the liquid phase can be extracted. Left over outflow of fractional gas streams from these supplementary external refrigerant loops and mainstream partial liquefaction stages is commonly used to provide fuel for prime movers of process and pipeline compression equipment. Variations of chilling are evolving to more efficiently produce small scale LNG in an increasingly competitive field.
Improvement in LNG Plant Efficiency
For all types of LNG processing higher front-end extraction of undesirable constituents in feedstock, better plant metallurgy, improved heat exchangers, and development of customised refrigerants, have all played their part in a continuous quest for improvements in efficiency of LNG production of all traditional processing.
Present day pre-conditioning processing can extract most of the Natural Gas Liquid (NGL) constituents from natural gas deliveries to LNG process trains. This yields near 100% methane inflow mixture to the LNG chilling process.
In the early days of commercialization, fuel gas was measured as a gas equivalent of over 15% of the inflow of the gas delivered to the plant by the pipeline. As a result of evolving selection of refrigerants and/or continuous process improvements, this fuel gas requirement has broken through the barrier to values typically less than 10% in today's plant configurations. Economies of scale and a migration to floating/offshore LNG plant configurations have all yielded positive changes in recent times to LNG processing making more of the inlet process stream available for conversion to LNG.
Yet, in spite of improvements, the industry is still confronted with emerging concerns over plant complexity, delays in start of operations and cost overruns amidst ever tighter permitting conditions.
Enthusiastic acceptance by the industry of small gains in operational efficiency that are measured as single percentage points is indicative of the plateau effect of “traditional” processing methods. Improvements now have to be justified by cost savings that are increasingly difficult to come by in an industry that can be considered as mature in its present state.
Increasingly Capex, operating costs and the output of CO2 emissions is being questioned in process selection, plant permit applications and financing approvals. The push for less fuel gas consumption equates to lowering these CO2 emissions, and also enables more produced LNG to be made available for marketing.
Carbon Dioxide Emissions Scaling in the LNG Industry
A common way to measure LNG plant performance is to compare the tonnage of CO2 emissions from hydrocarbon fuel burned in providing plant power needs to the tonnage of LNG produced by the plant. This ratio is known as the “Carbon Intensity” factor. Early plants offered a ratio of 0.30 or higher. Modern day plants aim at a level of 0.22 to achieve a “Good” rating. A present-day watermark of a Carbon Intensity of 0.18 has been achieved by the offshore Norwegian Snovit LNG Installation by using efficient aero-derivative gas turbine power generation and electric motors for process equipment drivers.
For future advancements, there is a need to look beyond the status quo approach of plant design through a reconsideration of the basic process. The most used practices of multi stepped chilling systems have improved over the years with specific refrigerant mixes and heat exchange equipment being selected for best efficiency. This provision of chilling however remains for the most part an external process, where these plant design constraints and repeated compression within complex plants persist, along with high capital costs.
A process for producing liquefied natural gas from a natural gas feedstock of a pipeline comprising:
In embodiments the turbo expansion device is a stepped series of turbo expansion devices.
In embodiments the final temperature of the expanded gas leaving the turbo expansion device is above about −262° F., the expanded gas is further subjected to extraction of sensible heat to render the liquid state at a temperature of about −262° F.
In embodiments the conditioned feedstock at the turbo expansion device inlet is 100% methane, or it is a mixture of methane or methane and NGLs that has been preconditioned to remove undesirable quantities of water vapor, acid gas, excess NGLs and heavier hydrocarbon liquids, CO2, N2, and mercury.
In embodiments the feedstock comprises up to:
In embodiments the molecular weight of the feedstock does not exceed about 23.2, HHV of the feedstock does not exceed about 1395BTU/ft3, and modified Wobbe Index of the feedstock as calculated for 60° F. does not exceed about 62.20.
In embodiments the elevated pressure of the conditioned feedstock is between about 3400 psig and about 600 psig, the elevated temperature of the conditioned feedstock is between about −20° F. and about 210° F., and the elevated temperature and the elevated pressure of the conditioned feedstock intersect on the offset turbo expansion chilling curve profile to provide an expanded gas having a temperature of between about −145° F. and about −262° F. and a terminal pressure of between about 5 and about 15 psig.
In embodiments the offset turbo expansion chilling curve profile is the 170° F. curve or 190° F. curve of
In embodiments the turbo expansion device is coupled to a shaft, to recover energy released by the expansion.
In embodiments the shaft is either a single shaft or a multi shaft configuration wherein the shafts operate at different or the same speeds.
In embodiments the process further comprises the step of interrupting the expansion of the conditioned feedstock step to remove excess liquid fractions formed during the process. In embodiments the step of interrupting the expansion includes the introduction of guide vanes within liquid bleed off chambers.
In embodiments the expanded gas is further cooled using externally refrigerated heat exchange equipment situated downstream of the outlet of the turbo expansion device to extract sensible heat for final liquefaction of the expanded gas.
In embodiments a condensation loop of refrigerant in the heat exchange equipment is integrated with the turbo expansion device or with the expanded gas emerging flowstream.
In embodiments the process further comprises chilling a stream of NGLs with the heat exchange equipment, to enhance the HHV heat content of the produced LNG by intermixing of the streams.
Described is a process for production of Liquefied Natural Gas (LNG) that optimises an offset chilling curve attributed to the refrigerant properties inherent to the natural gas feedstock for the majority of chilling, followed by a smaller externally sourced heat exchange to make the final phase transition to LNG. “Offset” chilling curve, as used herein, refers to a deeper cut in the turbo expansion chilling curve profile that terminates at the lower pressure levels of the gaseous region of the feedstock natural gas phase envelope, thus reducing the amount of heat transfer required for the transition of the feedstock to the LNG state. The offset chilling curve profile is created by increasing the temperature and pressure of the feedstock.
Abnormally hot inlet conditions, relative to those commonly associated with LNG processing, enables an offset single step chilling path via Joule-Thomson (J-T) expansion of the process gas, to conditions of temperature and pressure near to the LNG state. This avoids a region of energy-sapping recompression of refrigerants in external heat exchange loops commonly employed in the mid range cryogenic chilling steps of traditional LNG systems. The chilling process presented here may achieve a carbon intensity factor for LNG processing in the region of 0.03.
Recovery of compressed feedstock gas energy is available via the substantial outlet shaft power of the turbo expansion step. This recovered power can be applied to drive the final refrigeration loop for heat transfer in the liquefaction step, and to preprocessing of the gas stream upstream of the process. This preprocessing assures controlled steady state operation demanded of the turbo expander and the subsequent extraction of sensible heat for the LNG product delivered to storage at just above atmospheric pressures.
The process uses compression of transmission gas to induce uncommon high pressure hot entry conditions into an expansion chilling device, such as a J-T device or turbo-expansion device, and by subsequent expansion provides the primary method of chilling by using the inherent Joule-Thomson (J-T) properties of this feedstock. Best performance is achieved when the LNG feedstock is close to pure methane, and methane is the preferred feedstock for the process described herein.
Most LNG processing, receiving pipeline delivery of gas starting at 1100 psig and below and rarely above 100° F. in temperature, first uses pre-cooling to set the scene for subsequent low pressure low temperature efficient expansion of the refrigerant and/or feedgas. The method disclosed herein runs contrary to common practice and starts by direct turbo expansion chilling demanding higher pressure and temperature start conditions than are commonly accepted pipeline delivery conditions.
A single step chilling (turbo expansion) stage takes the feedstock from selected high temperature and high pressure inlet conditions to near atmospheric low pressure and cryogenic temperature outlet conditions. Subject to the efficiency of turbo expansion, a small capacity external refrigeration loop may then be the only additional requirement to complete the phase change to LNG. The feedstock is not further compressed during the single step chilling stage, nor is any other means of chilling used during the single step expansion chilling stage.
When normal efficiency turbo expansion is used (about 80-85% efficiency with light gas mixes for machines presently in common use), external refrigeration may be required for transition to the liquid phase. When next generation high efficiency turbo expansion is used (about 95% or higher efficiency on light gas mixes), external refrigeration capacity may be subject to reconsideration and reduced. The process disclosed herein may use, in the single step chilling stage, multi-step back-to-back present generation turbo expansion machines, or single-step standalone next generation turbo expansion machines.
The process initiates the expansion of refrigerant/feedstock gas at higher temperatures than contemporary practices, to offset the trajectory of the conventional chilling curve away from the dewpoint of the feedstock. This path is taken to maintain the J-T chilling, only occurring in the gas phase, as long as possible before the expansion process crosses the dew point curve. By using this method to achieve a deep temperature drop (ΔT), the process eliminates the need for high heat exchanger loads commonly experienced to subsequently achieve the liquid state in most LNG processing. Further, by recovery of greater compressive energy through the turbo expander the process exhibits reductions in fuel, CO2 emissions levels and process energy footprint through plant simplicity.
In practice, the method described herein may use off-the-shelf components. The higher temperatures used for this process can be obtained for example, in a preferred embodiment, purely by heat of compression provided by pipeline station spec compressors and modulated with air coolers. The specifications of the compressor and related size specification of air coolers can be selected by a person of skill in the art depending on the capacity required, In another embodiment, if a high pressure gas feedstock is provided directly from the pipeline, a source heat may be provided to raise the temperature as required. This option is attractive if waste heat is available,
The process results in plant simplification through having the bulk of chilling coming from the process fluid itself, akin to that advocated by Clause, but without the drawbacks and the inefficiencies of providing recompression part way through the expansion chilling. An objective of the proposed method is to eliminate this step of recompression of a chilled refrigerant medium for external or internal heat exchange as used in conventional LNG processes. This recompression traditionally takes place in a cryogenic region around the critical point of the gas where heat coefficient values Cp are high, and excessive heat of compression also has to be handled. Process energy is more effectively used by starting with offset high temperatures at high pressures upstream of the turbo expansion step than by starting with more moderate status quo pre-chilled temperatures at moderate pressures associated with contemporary LNG processing.
The expansion chilling profiles of natural gas used in the proposed method generally can be considered to begin at a pressure of 3400 psig where temperatures may range between 95° F. and 250° F. Lower temperatures are preferable for pure methane mixtures and higher temperatures are preferable for mixtures of higher NGL content. Chilling performance along these curves can be subject to interception at any lower pressure and temperature provided by inlet conditioning.
The chilling profile provided by external heat exchange or self-chilling in contemporary LNG processing starts at pipeline delivery pressures of 1100 psig or less, preferably at between 800 and 600 psig. Process entry temperatures in general are those provided by the feedstock supply pipeline and are at 105° F. or less depending on the temperature surroundings of the pipeline, and preferably between 68° F. and 75° F. Pre-cooling by heat exchanger or small ΔP reduction through turbo expansion prior to first cold box entry or self-chilling would typically be to the 45° F. region with air coolers, and to −31° F. region using external propane chilling loops.
In Minta (U.S. Pat. No. 9,140,490), which advocates the benefits of high-pressure heat transfer properties of natural gas in a process using external heat exchanger loops, pressures between 1000 psia and 5000 psia were investigated. Here benefits of thermal heat transfer for plate exchangers were found to be optimal where refrigerant and process gas pressures were between 1500 psia and 3000 psia. Minta's final high pressure to atmospheric pressure reduction in a liquid state for the delivered LNG takes place from cryogenic temperatures. This offers minimal J-T temperature reduction for feedstock in the gaseous state and no J-T temperature reduction for feedstock in the PLNG (pressurized liquefied natural gas) state. The present disclosure exhibits high values of J-T coefficient in a gaseous state at cryogenic temperatures at pressures less than 1000 psig.
Embodiments of the present method contemplate fluctuations in the constituents of the feedstock and make allowance for a final external chilling loop for the last few degrees of temperature drop and removal of sensible heat from the flowstream to convert it to LNG. This loop ideally uses nitrogen as the refrigerant best suited to offshore needs of non-explosive specifications. Methane or mixed refrigerant also remains as a convenient and slightly more efficient refrigerant option under selective service conditions.
Consideration of Lean and Rich Feedstock Mixtures
Minor adjustments in the process allow for inflow of all common transmission quality mixtures (including those with higher NGL content and lower Joule-Thomson coefficients). The allowance for extraction of fallout NGLs through intercept chambers within the turbo expansion step brings the process thermodynamics closer to that of pure methane feedstock.
The use of a pure or near-pure methane feedstock as the primary self-chilling refrigerant is a preferred consideration that, as stated earlier, can be incorporated in today's improved pre-treatment processes.
Exemplary Options for Equipment Layout
Option 1: A series arrangement of individual high efficiency turbo compressors that can yield staged reduction of pressure to the point of liquefaction—for example, a high pressure machine configuration for the first unit followed by a low pressure machine configuration for second unit to make the final transition of conditions to the LNG zone.
Option 2: A staged wheel configuration within a single casing is a consideration, and feasible to the point of liquefaction of the process stream.
This process supports improvements in wheel efficiencies beyond 90% as is apparent when employing a radial flow wheel as the first stage and radial to axial flow blade profile on the second stage wheel, a layout that could improve upon the staged multi-wheel configuration described in Option 1.
A person skilled in the art will recognize that, subject to the purity of the methane constituent, temperature and pressure reductions can reach the low vapor fraction levels of LNG. This provides the opportunity of eliminating or reducing the external exchange of sensible heat to yield the vapor fraction of 100% using specifically configured expander machinery within a single casing.
“Sensible” heat extraction as used herein refers to the removal of latent heat causing the phase change from gas to liquid without a change in temperature.
The process described addresses the problem of the high initial fuel consumption of a typical LNG process plant, which is an objective of process improvements. Given that most plants are powered by fuel streams extracted from the inflowing natural gas feedstock, the reduction in the fuel gas usage of the instant process yields an increase in the available sales gas volumes in the form of LNG leaving the plant.
For peak J-T chilling, the removal/reduction of NGLs (e.g., ethane, propane, normal butane, isobutane and pentanes) is desirable for the process described herein, generally leaving only trace amounts of the lighter NGL constituent of ethane in the feedstock.
Modern day processing for separation of liquids by front end plants upstream of the LNG process can make available an LNG feedstock of almost 100% methane devoid of acid gas, water vapor, CO2, N2 and mercury. For conversion into LNG, these extracted constituents may be removed in the chilling process feedstock because they freeze and clog the process, reduce heat value or corrode equipment.
The process described herein results in temperature reductions to below the critical temperature well into the cryogenic region utilizing the J-T effect through a larger ΔT range to approach the liquid state for the feed gas.
The process considers several factors in setting start conditions for expansion chilling, to maintain continuity of J-T chilling over a single stage and a large ΔT drop for both NGL rich and light gas mixtures:
Process Evolution
First, one must consider the conditions required for J-T chilling through a large ΔT. Even starting at an ambient temperature of 75° F., some 337 F degrees of temperature reduction is required to reach the targeted −262° F. liquefaction state of LNG at atmospheric conditions. The larger ΔT requirements of this method require a considerable pressure drop using the feedstock as the primary refrigerant. The process takes place in a region where overall temperature reduction outpaces deficiencies of low J-T coefficients associated with a high starting pressure.
High Starting Pressure
Pre Chilling of Gas Mixtures
This was found to be both energy intensive and to result in the chilling trajectory of single step expansion coming to an abrupt halt when making a premature interception with the dew point curve of the process gas mixture well short of the liquefaction temperature zone.
Air Cooling Downstream of Compression
Compressing gas to high pressures results in an increase in temperature, and heat extraction is economically handled by air chilling the compressor discharge to approach the temperature of ambient air. To prechill below this limit, uninterrupted J-T chilling of feedstock expansion is desired to minimize further expenditure of chilling energy that would otherwise have to be considered for an external heat exchange loop.
Avoiding Peaking Values of Recompression at Lower Temperatures
Unlike the Claude process, the process described herein avoids recompression of the feedstock refrigerant to achieve further chilling through the temperature range of −100° F. to −150° F. Offsetting the chilling curve by beginning decompression at higher temperatures steers the process conditions clear of the zone of peaking values of Cp for methane shown in
Key Requirements for Single Step Pressure and Temperature Drop to Cryogenic Region—High Temperature Starting Conditions
A target temperature of about −262° F. for the LNG state at 5 to 15 psi above atmospheric pressure is an objective of the process described herein. This process seeks to both simplify traditional LNG infrastructure and to reduce fuel consumption for LNG production.
In order to achieve the large temperature drops required, the limitations of expansion wheel efficiency for real world turbo expansion are considered. Target temperature and pressure conditions are also be moderated from the ideal quoted above depending on gas constituents. Starting conditions of temperature and pressure, and levels of turbo expander efficiency commensurate with developments in axial flow as well as radial flow machinery make such deep cuts in temperature practical.
The merging of radial flow blade shape for high pressure reductions with axial blade extensions for low pressure reductions within a single wheel is gaining traction amongst manufacturers of turbo machinery. Optimal wheel tip velocity and associated improvements in guide vane profiles is resulting in a new generation of higher efficiency units that allow for greater temperature differentials between inlet and outlet points.
The single step expansion from a high initial pressure of 3400 psig for primary chilling from typical present-day machinery having a readily available efficiency rating of 83%, as shown in
During computer modelling using compressor and turbo expander efficiencies common to present day use, development studies achieved temperatures down to −255° F. when the liquid fraction began to form for single step turbo expansion from selected start conditions of 170° F. and 3400 psig. At this starting temperature, the aforementioned obstacles of pre-chilling, premature interception of the dew point and recompression are avoided. The trend towards more efficient compound blade profiles directed to this end will serve to enhance the production of LNG in this manner.
Turbo Expansion Starting Conditions
Examination of the temperature/pressure curve profile for offset chilling seen in
As an example, a 1500 psig pipeline compressor delivery pressure boosted upwards would intercept curves (3) and (4) in region C, No higher compression of the gas would be needed, In this event the entry temperature to the turbo expander would be increased so as to intercept curves (3) and (4) in region C.
Proximity of Dew Point Phase Change Boundary to Process Chilling Curve
The dew point line of either the phase diagram of pure methane or phase envelopes of light hydrocarbon mixes of transmission specification gas can be intercepted by the cooling trace from the air-cooled starting conditions. As seen in
The process therefore considers hotter entry conditions. At starting conditions such as 190° F. at 3400 psi in
A starting condition of 170° F. at 3400 psig offers the optimal starting position to turbo expand to attain the highest temperature reduction without exhibiting a vapor fraction for a pure methane feedstock depressured to near atmospheric pressure.
Consideration of NGL Vapor/Liquid Fall Out in Processing NGL Rich Feedstock
The presence of NGLs, even in spec pipeline transmission gas will reduce the rate of temperature drop in turbo expansion. Pre-conditioning of the gas for their extraction is preferable, not just to improve the rate of self-chilling of the feedstock, but to avoid slugging in the turbo expander. As understood by a person of skill, since the expansion can be carried out through a multi stepped array of expansion wheels, it is possible to break the continuous flow at a certain stage in the expansion, direct the flow to a separation drum, collect and remove NGL fall out, and then redirect the remaining (mostly methane) flow back to further chilling through the following stage of expansion. The higher J-T coefficient of the increasingly pure methane flow-stream will permit a deeper ΔT temperature cut than will be obtained flowing mixtures having higher NGL constituents, Different forms of blade configuration will permit residual NGL mist to be carried through the final process without damage to machinery to the point of direct liquefaction or through external extraction of sensible heat to render the final LNG product.
Should NGLs be desired in the LNG mix to enhance HHV combustion heat content of the market product they can be chilled externally or by mixing with a super cold stream of LNG and reintroduced to the sales mix.
Each loop has 4 stages of treating the working refrigerant:
The three loops shown use refrigerants best suited to the reduced temperature steps that each stage imposes on the feedstock gas.
The propane loop (C-D-E-F-C) comprises a compressor (102), a condensing cooler (103), throttle (104) and heat exchange passage (101). The ethylene loop (H-I-J-K-H) comprises a compressor (202), condensing cooler (203), throttle (204) and pre chilling passage through (101) prior to heat exchange through (201). The methane loop (L-M-N-O-L) comprises a compressor (302), condensing cooler (303), throttle (304) and pre chilling passage through both (101) and (201) prior to heat exchange through (301).
It is noted that the propane loop cooler (103) uses air as its heat sink while the ethylene loop cooler (203) and methane loop cooler (303) each use the lower temperature preceding cold boxes (101) and (201) as their heat sinks.
The final throttle at the exit point of coldbox (301) reduces the pressure of the feedstock to atmospheric storage of LNG.
For more efficient temperature reduction and recovery of compressive power throttles (104), (204) and (304) are replaced by turbo expanders, particularly on larger capacity plants.
The first loop (HMR) uses a compressor (101) and condenser/cooler (102) to move and thermally condition refrigerant through the cold box prior to throttling and expansion. It uses staged chilling steps by splitting the refrigerant into 3 flow streams: Stream A-B-C-D-E-A, Stream A-B-C-F-G-H-A, and Stream A-B-C-F-I-J-K-A.
The first stream comprises the compressor (101), condenser/cooler (102), branching at C to throttle (103), followed by a heat exchange loop through the coldbox to point E where the stream re-enters the compressor (101)
The second stream comprises the compressor (101), condenser/cooler (102), branching at F to throttle (203), followed by a heat exchange loop through the coldbox to point H where the stream re-enters the compressor (101)
The third stream comprises the compressor (101), condenser/cooler (102), branching at I to throttle (103), followed by a heat exchange loop through the coldbox to point E where the stream re-enters the compressor (101).
The second refrigeration loop (LMR) running from L-M-N-O-P-Q-L provides the final chilling to liquefaction of the feedstock. A lighter refrigerant mix works at these low temperatures. The compressor (201), and condenser/cooler (202) move and thermally condition refrigerant through the cold box prior to passing through throttle (203). This throttling provides the pressure reduction and expansion of the refrigerant prior to its heat exchange passage back through the cold box on its way to re-enter the compressor (201) at point Q.
The feedstock gas is pretreated upstream of this liquefaction process for reduction or removal of undesirable amounts of constituents such as acid gas, water vapor, CO2, N2 and mercury. Treated gas enters the process at point R, from where it passes into the cold box vessel (100) and is chilled as far as Point S from where it is withdrawn at pressure and temperature conditions conducive to entering separator vessel (300) at point T. Here residual NGL condensates fall out of phase and are withdrawn from the feedstock at point Z to flow to storage via point AA.
The remaining gas flow is extracted from point U and flows via U-V back into the cold box where it rejoins the heat exchange elements at point W. From W onwards the gas flow is further chilled to liquefaction and emerges at point X as LNG from where it flows to storage at point Y into tank (301).
Many systems, particularly large capacity plants, replace the throttle with a turbo expander for deeper cut chilling and energy recovery. Recent developments to improve efficiency have seen coil wound heat exchangers replaced by more efficient new fin and plate configurations.
This layout is intended for near pure methane feedstock. A staged compression (101) ahead of the turbo expansion inlet (D) may be performed. The supply temperature at the inlet of the turbo expander (202-203) is set to a higher than ambient temperature, to offset the cooling curve as illustrated in
This single temperature drop is sufficiently close to the target figure of −262° F. at 5 psi to permit an external methane/nitrogen/mixed refrigeration loop—LR loop—(U-V-W-X-Y-S-T-U) and low ΔT heat exchanger (401) shown here in the style of the Cascade process, to extract sensible heat of liquefaction from the flow-stream. Loop LR comprises a compressor (102), condenser/cooler (402) linked to the cold flow through turbo expander stage (203), a throttle (403) and heat exchanger between S and T in the coldbox (401).
For required phase transition to LNG the flowline K-L delivers the turbo-chilled gas to the coldbox (401). Chilled LNG product is delivered via line M-N-O-P to traditional storage facilities (503).
The path N-O takes the production stream through the separator 502 where under start up or upset conditions the gas phase is split from the liquid phase. The gas is recycled back to the inlet of compressor 101 via Q-R-S-T that contains compressor 103 to bring up the pressure to entry conditions for compressor 101.
The turbo expander start conditions of temperature and pressure will depend on the climatic region in which the plant is located. Minimum air temperatures in different regions will determine the approach temperature of air cooling that the compressed feedstock can be subjected to.
After compression, to intercept and follow the offset optimal cooling curve as shown in
Shaft power recovery (301) from the turbo expander will vary depending on the start conditions set for the process. Under most conditions it will be substantial and able to cover the power needs of the final liquefaction stage, and/or contribute to the process energy for preconditioning the feedstock, should this plant be located adjacent to the LNG process.
The final chilling stage (LR Loop) is also an assurance that sufficient chilling capacity is available to produce LNG should an upstream process fluctuation in gas composition or conditions introduce small quantities of NGL (primarily trace ethane) into the flow-stream. This could result in an earlier onset of the vapor fraction, slowing down the J-T effect of the remaining methane gas phase and raising the temperature of the product produced by the turbo expander chilling.
Also shown in this embodiment is a break in the turbo expander wheels between sections (202) and (203). This break allows for the extraction of the flow-stream via line E-F to the separator (501). When conditions warrant, NGLs can be removed via line G-H and the drier gas directed back into the turbo expander section (203). The break between turbo expander sections also permits the mixed use of radial and axial flow wheels in the turbo expander configuration. Should different shaft speeds be required between (202) and (203), a physical break or speed compensating gearing can be added at this location.
The performance of all 3 systems is shown against the heat transfer rate required en route to liquefaction of pure methane for a typical production train flow rate of 4.5 mpta. All systems deliver LNG product at −262° F. and 5 psi to storage.
The space between the lines tracking the process (solid lines) and the methane demand (dashed lines) is therefore representative of efficiency losses in the particular system.
The cascade process is represented by two trace lines shown in
The mixed refrigerant process is represented by
The process described herein is represented by
Recompression/expansion of feed gas along the lines of the Claude process for chilling here would prove restrictive in this tight pressure/temperature zone. Hence the classic Claude trace of chilling and recompression to attain lowest terminal temperature and liquefaction is best avoided here. An external chilling process and heat exchange system (along the lines of the final Cascade process stage, but configured for smaller ΔT), is used here to achieve phase change to LNG.
The following series of chilling profiles show why a direct chilling curve was necessary to approach the target region of −262° F. at 5 psi for LNG formation from turbo expander inlet conditions involving pure methane feedstock.
The relationship of heat of compression to the value of k=Cp/Cv indicates that for recompression in this region where large values of Cp occur, excessive heat has to be removed by subsequent Joule-Thomson expansion of the feed gas. The idea that chilling can be efficiently achieved by repeated compression and expansion through the cryogenic zone is challenging, and not uncommon to other Claude processing systems.
The process described herein, using the feed gas as a refrigerant, avoids recompression.
It will be noted that for process pressures above 1000 psi there is a flatter profile of the corresponding Cp curve. Repeated recompression and expansion at these higher pressures as suggested by the Claude process and used in some process methods, would still demand high levels of driver energy to maintain the high pressure trajectory of the chilling curve through this zone.
Subsequently with the offset chilling curve of
Simplification of the LNG process and power reduction of the new process is apparent when all three systems are examined alongside each other.
In the process described herein, the process fuel energy is primarily entered at the pipeline compressor, and the process compressor upstream of the LNG chilling plant. The turbo expander recovers both the pipeline and process compressor energy —sufficient to power both the second step liquefaction chiller and the gas conditioning plant, however it is distributed.
In general, the trend follows an S shaped curve with lower values of the coefficient less than 1 degree of chilling per 100 psi drop in temperature occurring at high pressures. In the intermediate region of 1000 psig values of 5 to 7 degrees F. per 100 psi drop in pressure are noted. For low pressures at the termination of expansion to pressures below 100 psig, the most useable increasing values in the coefficient of 5 to above 8 degrees F. per 100 psi pressure reduction are observed.
Noting the influence of temperature, it is seen that below 2000 psig colder temperatures bring about higher coefficient values. The curves twist on themselves at pressures above 2000 psig and show better performance coefficients for the higher temperature curves.
The gas behaviour trends shown here are indicative of where the process described herein demands real world consideration. Turbo expanding from pressures around or below 2000 psig and from temperatures above 81° F. shown here to near atmospheric pressures is the direction to follow to reach cryogenic temperatures for near pure methane feedstock. Intercepting the chilling curve (3) of
Described herein, and shown in
The increased pressure and increased temperature are selected to be at a level such that the conditioned feedstock has an offset turbo expansion chilling curve profile that will enable chilling of the flow stream through J-T expansion along an offset chilling curve profile that terminates at the lower pressure levels of the gaseous region of its phase envelope, thus easing the amount of heat transfer for the transition of the feed gas to the LNG state. Here the detail in
The “termination” of the offset curve is the point at which the Dew Point of the feedstock gas on a phase diagram is struck by the offset chilling curve, or when the horizontal zero pressure axis is struck by a chilling curve having a higher start temperature. For example, in
An offset chilling curve profile that terminates at the lower pressure levels of the gaseous region of the phase envelope is desired. In this context, “lower pressure” refers to a pressure of less than about 20 psig, less than about 15 psig, less than about 10 psig, less than about 5 psig; between about 0 to about 20 psig, about 0 to about 10 psig, about 0 to about 5 psig, about 1 to about 5 psig, about 1 to about 10 psig, about 5 to about 10 psig, about 10 to about 20 psig, about 15 to about 20 psig. In embodiments, the lower pressure level is the pressure typically used for industrial storage of LNG.
The next step is delivering the conditioned feedstock to the inlet of a turbo expansion device at the selected elevated temperature and selected elevated pressure. As discussed above, the turbo expansion device may be a stepped series of turbo expansion devices.
The next step is expanding the conditioned feedstock in the turbo expansion device.
The next step is discharging the expanded gas from an outlet of the turbo expansion device at a temperature of between about −175° F. and about −262° F. for pure methane or light NGL gas mixtures and between about −145° F. and −175° F. for rich NGL mixes and at a pressure of between about 5 and about 15 psig.
In the method, the conditioned feedstock gas is not compressed again (it is not further compressed) after it is delivered to the inlet of the turbo expansion device and before it is discharged from the turbo expansion device. Nor is any other means of chilling the gas used, after the feedstock gas is delivered to the inlet of the turbo expansion device and before it is discharged from the turbo expansion device.
If the final temperature of the expanded gas leaving the turbo expansion device is above about −262° F., the expanded gas may be further subjected to extraction of sensible heat to render the liquid state at a temperature of about −262° F.
The preferable feedstock at the turbo expander inlet is 100% methane, or a mixture of methane or methane and NGLs that has been preconditioned to remove/reduce undesirable quantities of water vapor, acid gas, excess NGLs and heavier hydrocarbon liquids, CO2, N2 and mercury that would otherwise inhibit the chilling process.
In embodiments, the standalone unrelated maximum content of the constituents is:
In embodiments the molecular weight of the feedstock does not exceed 23.2, HHV does not exceed 1395BTU/ft3, and modified Wobbe Index as calculated for 60° F. does not exceed 62.20.
The phase at turbo expander entry conditions of such mixtures is gaseous. It is possible, subject to delivery conditions of temperature from the pipeline, and compression ahead of the turboexpander that the preferred offset chilling curve can be intercepted at pressures lower than the aforementioned 3400 psig.
In embodiments, the starting pressure for turbo expansion is between 3400 psig and 600 psig depending on climatic location of the process plant and the constituent makeup of the gas feedstock. The corresponding start temperatures to these pressures lie between 250° F. and −40° F. Higher temperatures are considered for NGL rich gas mixes to allow for liquid extraction, and a staged final expansion of the separated stream of near methane gas to the liquefaction region at the lower reaches of its dew point curve.
Lower inlet temperatures are noted for special circumstances involving the receipt of near methane mixes from pipelines operating in Arctic regions whereby gas conditioning for pressure and temperature can intercept the lower temperature reaches of the aforementioned offset chilling curves.
In this process the cryogenic temperature achieved on exit from the turbo expander is between −175° F. and −262° F. for feedstock light in NGL constituents. When NGLs are allowed to fall out at a break point in the turbo expansion, the near methane mix emerging to enter the final expander step was found to achieve temperatures between −145° F. and −175° F. depending on start mixture and turbo expander efficiency.
In embodiments turbo expansion can be coupled to recovered shaft energy. This can be either single shaft or multi shaft configuration operating at different or same speeds to suit axial or radial flow blade and wheel selection.
In embodiments the turbo expansion step, or multiple stages of blade wheels, are interrupted for liquids removal. This interruption may include the introduction of guide vanes within liquid bleed off chambers. These chambers may be located at certain process encompassed pressure and temperature points within or in a branch connection outside of the turbo expander that are attributed to specific fallout properties of constituent NGLs. After the withdrawal of liquids from these operational points, the remaining gas stream may be subjected to further turbo expansion to low temperatures. In embodiments externally refrigerated heat exchange equipment is situated downstream of the turbo expansion unit, to extract sensible heat for final liquefaction of the gas mixture. The condensation loop of the refrigerant of such equipment may be integrated with the turbo expander or emerging flow-stream. In embodiments, heat exchange equipment may be used to further chill a stream of NGLs, enabling enhanced HHV heat content of the produced LNG by intermixing of the streams.
In summary, described herein is a process for liquefying preconditioned natural gas feedstock principally by direct Joule-Thomson chilling in a single step from compression to high temperature/high pressure conditions to cryogenic/low pressure conditions using the refrigerant properties of the single stream of feedstock. The harnessing of site receipt of energy in compressed feedstock at an elevated temperature through a high efficiency turbo expander enables offset chilling to deep cut temperature conditions. This makes it possible to displace much or all of the complex and inefficient heat exchange loops associated with traditional LNG processing employing external chilling, and pre cooling of refrigerant and/or feedstock prior to expansion chilling. Sensible heat transfer to bring about phase change to liquid natural gas (LNG) for storage at or above atmospheric pressure is mixture and expander efficiency dependant and can call upon minor secondary chilling if so required.
While the process has been described in conjunction with the disclosed embodiments which are set forth in detail, it should be understood that this is by illustration only and the process is not intended to be limited to these embodiments. On the contrary, this disclosure is intended to cover alternatives, modifications, and equivalents which will become apparent to those skilled in the art in view of this disclosure.
The present application is a U.S. National Phase entry under 35 U.S.C. § 371 of PCT/CA2021/051842 filed on Dec. 17, 2021 entitled “LNG PROCESS USING FEEDSTOCK AS PRIMARY REFRIGERANT,” which claims the benefit under 35 U.S.C 119(e) of U.S. Provisional application No. 63/127,548 filed on Dec. 18, 2020, incorporated herein by reference in its entirety.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/CA2021/051842 | 12/17/2021 | WO |
Number | Date | Country | |
---|---|---|---|
63127548 | Dec 2020 | US |