Introduction
To control air pollution, optimize energy structure, and achieve sustainable development, China has promulgated serious policies [1–3]. As the cleanest fossil energy, natural gas (NG) proportion in the primary energy is expected to increase substantially in China. In July 2017, “Opinions on Accelerating the Use of Natural Gas” was proposed by the National Development and Reform Commission of the People’s Republic of China [4], and cultivating NG as one of the main energy sources was emphasized.
NG proportion in the primary energy of China is gradually increasing, which was 7.8% in 2018, and will exceed 10% by 2020 and 15% by 2030, according to the government’s work report [5]. However, it is still far below the average NG ratio of the world, which was more than 23.8% in 2018 [6,7]. In the near future, with the promotion of social development and welfare, a huge amount of NG will be demanded. However, NG production in China is insufficient, and thus the vast NG needed has to be imported. In other terms, China has become dependent on foreign NG in recent years. According to the National Bureau of Statistics, the proportion of imported NG rose to 42.9% in 2018, making China the largest importer of NG in the world [5].
China is relatively rich in NG resources. However, the distribution of these resources is very uneven. According to the National Bureau of Statistics, the gas reserves in China exceed 1 × 1014 Nm3 [5]. However, more than 65% of these reserves is unconventional gas, i.e., coal bed methane, shale gas, and oil field associated gas. These gas fields are mostly small and medium-sized. The majority of them are located in remote areas (far away from the users), and the geological structure of most of the gas fields is relatively complex. Thus, from the technical and economical point of view, the traditional pipeline network and the centralized liquefaction transport mode are not suitable for these isolated and scattered fields. As a result, more than 3 × 1010 Nm3 of coal bed methane are not utilized [8], which is enough to meet the NG demand in Beijing for about 1.5 a. This leads to a great waste of resources as well as environmental contamination. Therefore, extraction of the unconventional NG is especially important in China for energy security and independence, besides environmental protection reasons. Under this background, there is an urgent need for the development of the mobile and flexible NG liquefaction technology.
In this paper, the current mainstream NG liquefaction technologies are reviewed, aiming to search for suitable technologies for isolated and scattered gas fields. Moreover, a mobile NG liquefier driven by oil-lubricated single-stage screw compressor designed by Technical Institute of Physics and Chemistry (TIPC) is introduced. Furthermore, the challenges as well as the development of this kind of mobile miniature NG liquefiers are presented.
Review of the current liquefaction technologies
In a liquefied natural gas (LNG) production plant, the NG is generally cooled from ambient temperature to approximately 120 K, and then stored in the liquefied form in a near-boiling state. Thus, NG liquefaction is a typical temperature-distributed heat load-cooling process. Figure 1 shows the temperature-enthalpy (T-h) diagram (at constant pressure varying from 1 to 10 MPa) for a specific NG [9] having the composition reported in Table 1. As the isobar of 5 MPa presented by the dashed line indicates, the liquefaction flow of NG can be divided into pre-cooling, condensation, and sub-cooling. According to the Second Law of Thermodynamics, as the temperature difference in heat transfer decreases, the entropy generation and the exergy loss decrease, too. Therefore, the decrement in temperature difference between NG and the coolant is crucial, especially in the low-temperature range. Based on this consideration, the liquefaction process is performed around the dew point line of NG, as shown in Fig. 1.
Tab.1 Composition (on a molar basis) of a specific NG |
| Composition | Percent/%% |
|---|---|
| Methane | 94.59 |
| Ethane | 3.43 |
| Propane | 0.59 |
| n-Butane | 0.12 |
| Iso-Butane | 0.10 |
| n-Pentane | 0.03 |
| Iso-Pentane | 0.03 |
| n-Hexane | 0.05 |
| Nitrogen | 1.06 |
To achieve the liquefaction of NG, both sensible and latent heat have to be removed over a wide temperature range (approximately 300 K to approximately 110 K). Therefore, to achieve the smallest temperature difference between NG and the coolant, one or more refrigerants (either single or cascaded) are employed. After years of development, the cascade refrigeration cycle (CRC), mixed refrigerant cycle (MRC), and reverse Brayton cycle (RBC) have become the main NG liquefaction technologies [10].
In the CRC system [11–13], generally, three pure refrigerant cycles (propane, ethylene, methane) each with a 3-stage compression are employed, as demonstrated in Fig. 2. The CRC achieves a high liquefaction efficiency due to the small temperature difference between NG and the coolant since there are 9-stage evaporation processes to match NG heat load. Meanwhile, the refrigerants in CRC are easily charged, owing to only one refrigerant for one cycle, which gives CRC good robustness. However, the system is complex and hard to maintain because it consists of nine compressors and nine heat exchangers.
In the RBC cycle [14–16] which is dominated by the N2-based cycle (Fig. 3), a multi-stage expansion process is used. The cold box and equipment in RBC are simple. The liquefaction efficiency is relatively low but could be improved with the pre-cooling stage [17]. Besides, in some cases, NG can be used as the refrigerant medium, which requires no external refrigerant inventory and associated storage, thereby reducing the size and complexity of the liquefaction process [18,19].
In a single-stage MRC (SMRC) system [20,21] (Fig. 4), normally a single centrifugal compressor is used in a large-scale system. The mixed refrigerant consists of four to eight components to match NG heat load. The refrigerant concentration and the heat exchanger are crucial in achieving a high liquefaction efficiency. The suction pressure of the compressor varies from 0.1 to 0.3 MPa, while the discharge pressure varies from 3.5 to 5.0 MPa [21], which means the compression ratio will reach up to 20–30.
In both CRC and MRC, phase transition processes occur in the refrigerants, while the phase of the refrigerant does not change in expander-based technologies, which makes the configuration of RBC relatively simple. A notable disadvantage of RBC is the higher specific power requirement (compression power required for liquefying one unit of NG) compared to CRC and MRC [9,10,22–25]. CRC is still used in some large-scale factories (mainly constructed in the early days); however, it is not suitable in small-scale applications, especially for skid-mounted liquefaction systems due to its complexity and low efficiency. Therefore, most of the liquefaction systems for small-scale applications are developed based upon RBC and MRC or their modified cycles.
In the pre-cooling MRC system [26,27] (Fig. 5), the robustness of CRC and the high efficiency of SMRC can be integrated by a 3-stage propane cycle used for pre-cooling and an MRC cycle used for condensation and sub-cooling of NG. The liquefaction efficiency benefits from the heat load match in low-temperature areas. Since the MRC can be optimized and operated nearly independently of the pre-cooling propane cycle, the high boiling point refrigerant (e.g.,iso-butane and iso-pentane) is no longer needed. It was the most widely preferred cycle in the past decades. Recently, many works aiming at the selection of pre-cooling stage refrigerants have been reported [28–30].
As the name suggests, the dual-MRC (DMR) system [31,32] (Fig. 6) consists of two MRC systems: the pre-cooling MRC with high boiling points components, and the low-stage MRC with middle-low boiling point components. One major reason why attention has been focused on DMR is that it can reduce propane inventory, compared to C3-MRC. High efficiency was achieved in DMR with a better match of heat load compared to SMR. In theory, it has the highest efficiency among the NG liquefaction cycles [33,34]. In real applications, the C3MR and DMR processes have similar efficiencies [35].
The cycle combining C3-MRC and N2-RBC was proposed because it is difficult to achieve the uniform distribution of multi-component two-phase flow in the main heat exchanger in super large-scale applications [36]. This process is also called AP-X (Fig. 7) by the Air Products (AP) Company. The N2-RBC loop in low stage benefits from the heat load match and the uniform distribution in the heat exchanger. With the aid of N2-RBC at the low-temperature stage, this process may possess the largest liquefaction capacity with a reasonable production cost, although its thermodynamic efficiency is smaller than that of C3-MRC.
Although the power consumption of the liquefaction process is important for both small- and large-scale applications, each application has its own emphasis. The initial investment and compactness are of greater importance in small-scale LNG plants compared to large-scale ones.
Thus, some technologies with a low-power consumption of liquefaction efficiency are suitable for large-scale applications (e.g., the CRC process); however, they are not suitable for small-scale ones due to high capital cost and complexity. Besides, for onshore and offshore applications, different technologies are preferred. At present, MRC-based technologies still dominate the onshore LNG plant market, while RBC-based technologies are pursued in the offshore LNG market [37,38]. Table 2 presents a rough comparison of state-of-the-art NG liquefaction technologies. At present, for small scale, NG liquefaction applications, the precooling-MRC process, as well as the nitrogen-based RBC systems, receives the most attention. Although there is no clear agreement on the most suitable liquefaction process for small-scale applications, the precooling-MRC process, with relatively high efficiency, is quite promising to achieve high efficiency and reliability, once some problems, such as refrigerant composition control and the misdistribution of the two-phase flow, are resolved.
Tab.2 Comparison of mainstream NG liquefaction technologies |
| Cycle | Efficiency | Complexity | Reliability | Possibility for miniaturization |
|---|---|---|---|---|
| CRC | High with 9 stages processes | High | Low | Too many facilities, low reliability |
| Precooling-MRC | High | Middle | High | Good, with off-the-shelf facilities, easy assembly |
| N2-RBC | Low, improved by precooling | Low, simple system configuration | High | Difficult in miniature of turbine and compressor |
Considering the urgent NG need in China, and the fact that the conventional liquefaction technologies are not suitable for remote small-scale gas reservoirs, the use of low-pressure MRC driven by oil-lubricated single-stage screw compressor have been proposed in this paper, and accordingly high-efficiency, low-cost, and mobile NG liquefiers developed in Technical Institute of Physics and Chemistry (TIPC) in Chinese Academy of Sciences, China.
Development and applications of mobile miniature NG liquefiers
Several key problems aroused in developing the mobile miniature NG liquefier, such as the improvement of liquefaction, the decrease of the size of the key equipment (for example the cold box), and the shortening of the construction period of the liquefier and the lowering of the initial cost, which can be summarized as three challenges.
Challenge 1: liquefaction miniaturization
The conventional MRC system used for the NG liquefaction in large-scale LNG plants is very complex, which poses a challenging engineering problem to the miniaturization of the liquefaction system. The pressure of refrigerant in the cycle reaches up to 3–5 MPa [21], where the compression ratio increases to 20–30, yielding to the high requirements on the compressor. This is an oil-free multi-stage centrifugal compressor, monopolized by several companies, with high cost and is difficult to meet the miniaturization demand of the liquefaction system. These disadvantages make the construction period of a typical MRC system long and the relocation of it hard or even impossible. Besides, this kind of MRC system has poor adaptability and flexibility in dealing with varying duties.
Progress: low-pressure cycle using a screw compressor
To overcome the above-mentioned drawbacks of the high-pressure MRC process, a low-pressure cycle has been built. Compared to the high-pressure cycle employing an oil-free multi-stage centrifugal compressor, the low-pressure cycle can utilize an oil-lubricated screw compressor, which is widely used in HVAC&R (heating, ventilating and air conditioning and refrigerating). The benefits of oil-lubricated screw compressors are noticeable. First of all, approximate isothermal compression can be achieved by oil injection cooling in the screw compressor. Secondly, the semi-hermetic structure can avoid leakage of the mixed refrigerants, which frequently occurs in oil-free compressors. Finally, the off-the-shelf compressors are economic, highly reliable, and well-supplied, which can significantly reduce the initial manufacturing costs and shorten the manufacturing time.
Progress: low-pressure cycle with optimized mixtures
The selection of components for mixed refrigerants is a significant part of the thermal design of MRC. The optimal refrigerant in the low-pressure cycle is not the same as that in the high pressure cycle. At a suitable temperature of normal boiling point (Tnb) and a low triple-point temperature, N2, CH4, C2H4, C2H6, C3H8, and iso-C4H10 have been widely used in recently developed MRC systems [35]. Therefore, two strategies have been proposed in this paper to give a better match for heat load and temperature in the new cycle.
First, tetrafluoromethane (R14), with a normal boiling temperature of 145.2 K, was introduced to the mixed refrigerant. It is a suitable bridge between methane and ethane, as shown in Fig. 8. Although R14 has a high global warming potential value, the use of R14 can improve the efficiency of the MRC system [39]. R14 can be replaced by ethene (R1150) for environmental protection, although the exergy efficiency of the MRC system will slightly decrease. Second, thermodynamic optimization of the mixed refrigerants based on isothermal throttling effect is applied. Isothermal throttling effect (DhT) corresponds to the cooling capacity gained after throttling. The throttling effects of several pure substances with different boiling points from N2 to iso-C5H12 are shown for various temperatures in Fig. 8. According to the analysis of Gong et al. [40], it is significant to get a component series with relaying temperature ranges of maximum throttle effect in the component selection for mixed refrigerants. A large mixture throttling effect could be achieved in the whole temperature region by relaying the maximum throttling effect of each component with each other. The minimum isothermal throttling effect of the mixture in the whole temperature region is larger than that of any pure component.
Progress: optimized mixtures with accurate properties
The cubic equations of state (EoS), such as Peng-Robinson (PR) [41], Redlich-Kwong (RK) [42], and Soave Redlich-Kwong (SRK) [43], are the most preferred models in liquefaction process simulation [44,45]. The cubic EoS cannot properly describe the liquid properties in most cases [41], especially for the mixtures. As a result, large discrepancies between computational results and experimental data may be introduced by these inaccuracies [46,47]. However, the prediction accuracy can be improved by experimental measurement of thermophysical properties, such as vapor-liquid equilibrium (VLE), vapor and liquid densities, and specific heat. Since for PR EoS, only one adjusting parameter (i.e., binary interaction parameter) is required to describe the thermophysical properties of the mixtures, which can be determined by regressing the experimental data [41,48–50]. As shown in Fig. 9, a series of apparatuses are established, including several VLE devices aimed at different temperature zones [48–50], one pvT measurement apparatus [51] based on Archimedes’ principle of buoyancy, and one adiabatic batch calorimeter [52] for constant-volume specific heat. The accurate experimental data were used for the development of correlation and predictive thermodynamic models. For example, for the R14-based mixtures (R14+ R50, R14+ R170, R14+ R290), the estimated VLE (dashed line in Fig. 10) by setting binary interaction parameters kij = 0 by default presents a big deviation from the correlated VLE by experimental data, as shown in Fig. 10. This inaccuracy will definitely affect the simulation of the MRC system, and eventually lead to a large deviation between the design and actual construction. Therefore, the VLE, vapor and liquid densities, and isochoric specific heat data for more than 30 mixtures were measured, and the equation of state for simulation of MRC system was established [48–50]. This makes the simulations more convincing, compared to other simulations/works supported with no experimental data. At last, the mixture consisting of N2, CH4, CF4 (or C2H4), C2H4, C2H6, C3H8, and iC4H10 is selected for the MRC system.
Fig.10 Temperature-composition diagram for R14+ R50, R14+ R170, and R14+ R290 binary mixtures (The PR [41] equation is employed for regressing or estimating the VLE data. The solid lines are correlated by the experimental data (kij≠ 0), while the dashed line is predicted by setting the interaction parameter to 0 (kij = 0)). |
Progress: low-pressure cycle with a dephlegmator
Although the energy efficiency of the refrigeration cycle (Cycle A without separation) is the highest, as shown in Figs. 11 and 12, the lubricant and high-boiling components may freeze at the low-temperature part of the heat exchangers and block the flow channels. Therefore, a single-stage dephlegmation cycle (Cycle E) was developed by Wu et al.[53] and Gong et al. [54], as shown in Fig. 12(b).
A dephlegmator, shown in Fig. 13 [49], is used for replacing the flash phase separator in the traditional MRC. In the dephlegmator, the lubricant and high-boiling components are separated by the dephlegmation effect with the cooling power supplied by the backflow low-pressure mixed refrigerant. Since the backflow provides a low condensation trapping surface temperature, the dephlegmator has a better separation performance than the flash separator. The detailed thermodynamic simulation of the dephlegmator was conducted by Li et al. [55]. The high-boiling components in the liquid phase had a high recovery ratio of more than 90%, and the energy consumption was decreased by more than 30% compared to the traditional distillation tower with similar separation effects. In conclusion, when the efficiency of the optimized separation cycle is close to (or just slightly lower than) the cycle without separation, the separation cycle with the dephlegmator achieves a high efficiency and reliability [56].
Progress: practical effect of low-pressure liquefaction process
Finally, the low-pressure liquefaction system was built. The highest pressure of the mixed refrigerant reached up to 1.5–2.0 MPa, with a pressure ratio varying from 3 to 5. Compared to the high-pressure MRC system [21], at the highest pressure of 3.0–5.0 MPa and the pressure ratio of 20 to 30, the low-pressure MRC significantly eased the design and manufacturing of the critical components, such as compressor and heat exchanger. The intrinsic exergy efficiency of the MRC system with new mixed refrigerants (R14) reaches up to 60%, i.e., increased by 70% (Fig. 14) compared to the conventional high-pressure MRC system (with an intrinsic exergy efficiency of 35%).
The intrinsic exergy efficiency is defined as the maximum exergy efficiency of a MRC system by optimizing the molar concentration of the given components at a given high pressure ph, low pressure pl, ambient temperature T0, and refrigeration temperature Tc. The compressing process is assumed to be an adiabatic process without entropy increase and the pressure drop of the mixed refrigerants is zero. In other words, the highly idealized cycle has no extrinsic losses but only intrinsic losses. Therefore, the thermodynamic performance of the ideal cycle only depends on the mixed refrigerant used, the operating pressures, and the cycle configuration. The exergy efficiency is calculated as the ratio of exergy gained by liquefied gas to the inputted exergy. The exergy gained liquefied gas is the exergy difference between inlet and outlet status. The detailed exergy model can be seen in Ref. [40].
where h is the intrinsic exergy efficiency of the cycle; EQc is the exergy of the temperature-distributed heat loads, in which Qc is the heat loads; W is the electrical power consumed, and Pj is the exergy loss of each element j in the cycle. The intrinsic losses depend on the thermodynamic processes of the cycle and mixed refrigerant properties, such as the throttle irreversibility, the heat transfer temperature difference between the cold and hot streams in the heat exchangers.
Challenge 2: heat transfer and cold box
The low-pressure cycle makes it possible to liquefy the NG driven by the oil-lubricated screw compressor. However, this inevitably reduces the volumetric refrigeration capacity (i.e., kg LNG per volume flow rate of refrigerant). As a result, the heat load of the main heat exchanger for NG liquefaction is almost tripled, which negatively affects the volume of heat exchangers. Furthermore, due to the lack of complex two-phase heat transfer for the multi-component mixtures, it is always challenging to achieve the good two-phase flow mass-distribution in heat exchangers. Conventionally, excessive redundant design for the heat exchanger is adopted, leading to an oversized cold box, which further limits mobility.
Progress: basic heat transfer research
Three visualization testbeds (for nucleate pool boiling [59], flow boiling [60] and flow condensation [61] heat transfer measurement) were set in TIPC (Fig. 15) to investigate the basic heat transfer characteristics. In Refs. [53–59], experiments were performed on nucleate pool boiling, flow boiling, and flow condensation heat transfer of some refrigerants and their mixtures, and new correlations were proposed for two-phase heat transfer and fluid flow with a good accuracy [59–67].
Progress: heat exchangers
A new heat exchanger configuration (Fig. 16) was proposed [68], where the invented plate-fin exchangers with varying cross-sections achieved the two-phase flow velocity regulation and the even-distribution of two-phase flow [69–73]. Simulations and experimental tests were performed to further optimize the structure of the new exchangers [74].
Progress: compact cold box
Based on the research on the heat transfer of mixed refrigerants and the recently designed heat exchangers, a compact cold box, a combined plate-fin, and spiral wound heat exchangers were designed and manufactured, which resulted in a high efficiency and a low flow resistance. As a result, the compact cold box was shortened from 10 to 15 m to 3 m. As expected, it has quite a good mobility, allowing the compact box to be mounted on a truck. Figure 17 shows the compact liquefier equipped with a compressor and a cold box.
Challenge 3: fast construction technology
The traditional stationary LNG factories were not designed in the skid-mounted process module; thus, they require a long construction period (normally 8 to 18 months). All facilities are fixed, i.e., they have a poor mobility. Moreover, the long test (debugging) period is ineluctable. Therefore, they are not very useful for scattered gas resources/fields.
Progress: customized normalization technology
Standardized equipment manufacturing and integrated process technology (Fig. 18) have been available after years of effort. Three main technologies, namely production and testing, mixed refrigerants charging, and fast construction have been normalized. Specifically, the liquefaction plant is composed of/divided into several modules (compressor and oil separator module, cold box module and air cooler and fan module), and each module is manufactured on a skid. Combining with modularization interface design, the equipment can be easily and quickly assembled on a trailer skid. Besides, a series of processes and equipment are designed for mixed refrigerants charging.
Progress: flexible liquefaction center
High quality and low-cost are achieved by the standardized equipment manufacturing and device debugging technologies. The construction period is shortened from 8 months (for the traditional construction model) to only 2–6 weeks. In simple terms, “plug and liquefy” has been achieved. In addition, the development patterns of “rapid cluster application and flexible liquefaction center” has been eventually proposed. The flexible liquefaction center (Fig. 19) requires a much lower initial investment/cost than the traditional fixed plants. The reason for this is that the model proposed by the authors of the present paper can be built step by step. The profit of running liquefiers can support the installation of additional liquefiers. Another benefit is the flexibility in dealing with the unsteady gas resources. For example, the skid-mounted liquefier can be easily changed into a new gas well when the original one is exhausted.
Progress: performance comparison
The liquefier was built based on the aforementioned low-pressure MRC process, with an NG engine driving the screw compressor. All the liquefier components were integrated into a skid. The experimental results show that the NG engine consumes about 11% of the feed gas to liquefy the remaining 89% [29]. By using the off-the-shelf equipment and normalization technologies, the initial cost has been notably reduced. Now, it is about USD 2 million for a 10000 Nm3/d liquefier and will be even lower for large-scale plants, which is comparative to large-scale liquefier.
Table 3 compares the performance of representative LNG devices. As reviewed by Song et al. [75], hundreds of simulations have been performed, but only few reported the experimental results of the NG liquefaction processes. Thus, most of the data presented in Table 3 are estimated from the website of the companies. The liquefaction devices of the TIPC are notably enhanceed compared to the ones of the same scale. Currently, more than 20 sets of LNG plants based on technology proposed by the authors of the present paper have been applied in Inner Mongolia Autonomous Region, Shanxi, Shaanxi, and Yunnan provinces, just to name a few. The development process of the LNG plants of TIPC can be seen from Fig. 20. At present, all liquefiers are skid-mounted with a good mobility.
Tab.3 Comparison of the representative small-scale LNG devices |
| No. | Companies or institutes | Liquefaction technology | Scale/(104 Nm3∙d–1) | Power consumption /(kW∙h∙N–1m–3) | Percentage of consumed gasa |
|---|---|---|---|---|---|
| 1 | GTI (U.S.) [76] | MRC | 0.2 | 0.947 | 29.6 |
| 2 | Wärtsilä (Norway)1 | MRC | 6.7 | 0.50 | 15.6 |
| 3 | SINTEF (Norway) [77] | MRC | 2 | 0.6 | 18.8 |
| 4 | Hamworthy (UK)2 | N2-RBC | 10 | 0.57 | 17.8 |
| 5 | Galileo (Argentina)3 | N/A | 2 | 0.65 | 20.3 |
| 6 | KL energy [78] | MRC | 100 | 0.48 | 15.0 |
| 7 | ZY green energy [79] | CRC | 92 | 0.48 | 15.0 |
| 8 | ZJU [25, 80] | MRC | 0.0025 | 0.5 | 15.6 |
| 9 | HEB SL4 | MRC | 7 | 0.47 | 14.7 |
| 10 | KR petroleum5 | MRC | 5 | 1.87 | 58.4 |
| 11 | Yinchuan TianJIA6 | N2-RBC | 10 | 0.65 | 20.3 |
| 12 | Tianjin Zhenjin7 | MRC | 5 | 0.37 | 11.6 |
| 13 | Sichuan KF8 | MRC | 100 | 0.39 | 12.2 |
| 14 | SSCS9 | dual-MRC | 250 | 0.9 | 28.1 |
| 15 | Xinxing New Energy10 | MRC | 100 | 0.40 | 12.5 |
| 16 | Liu [81] | N2-RBC | 15 | 0.47 | 14.7 |
| 17 | TIPC | MRC | 3 | 0.35 | 10.9 |
Notes: a— Percentage of the feed gas consumed to liquefy the remaining NG. Here, the efficiency of the natural-gas generator is estimated as 3.2 kW∙h/Nm3. 1. Biogas liquefaction plant supplied by Wärtsilä to produce biofuel for buses in Norway. 2014–02–12, available at the website of wartsila.com; 2. Small scale & mini LNG liquefaction system. 2020–6–13, available at the website of hamworthy.com; 3. With Cryobox, Galileo achieves mid-scale LNG production. 2020–06–13, avaiable at the website of Galileo Technologies Company; 4. A 70000 Nm3/d LNG liquefaction device. 2015–10–15, available at the website of Liaoning CIMC Hashenleng gas; 5. Small-scale LNG Equipment. 2015–04–01, available at the website of Shandong Kerui Petroleum Equipment CO., LTD; 6. Mobile natural gas liquefaction plant. 2016–04–09, available at website of Yinchuan Tianjia Energy Technology Co., Ltd; 7. 50000 Nm3/d skid-mounted natural gas plant in Inner Mongolia. 2015–05–08, available at the website of Tianjin Zhenjin Oil and Gas Co., Ltd; 8. One million scale of natural gas liquefaction plant in Zhangjiakou. 2014–05–27, available at the website of Sichuan Air Separation Equipment; 9. China Huanqiu Contracting & Engineering Corp. The1st China LNG Expo. 2015; 10. LNG introduction. 2009–1–17, available at the website of Xingxing Energy |
Conclusions
This paper reviewed the development of mobile miniature NG liquefiers in TIPC. To efficiently liquefy the scattered NG and overcome the drawbacks of existing technologies, three main improvements have been made.
First, a low-pressure MRC process was built, including three key components: a low-pressure cycle for NG liquefaction, a mixed refrigerant with good intrinsic efficiency, and a cycle driven by an oil-lubricated screw compressor. The use of off-the-shelf components in constructing the liquefaction facility has greatly shortened the manufacturing time and reduced the costs related.
Next, a compact cold box with high efficiency was designed, including two main improvements: new heat exchangers and a compact cold box with a height below 3 m.
Lastly, the standardized equipment manufacturing and the integration processes were relized, including three main normalized technologies: mixed refrigerants charging normalization, production, test normalization, and fast construction normalization. The development patterns of “rapid cluster application and flexible liquefaction center” were eventually proposed.
The current small-scale NG liquefier achieved a minimum liquefaction power consumption of about 0.35 kW·h/Nm3. This made it suitable for small remote gas reserves, which could also be used in boil-off gas reliquefaction and the distributed peak-shaving of pipe networks.