Experimental study of flow distribution in plate–fin heat exchanger and its influence on natural gas liquefaction performance

Highlights

• Effects of gas–liquid ratio and tilt angle on gas–liquid distribution under tilt condition analyzed.

• Tilt condition aggravates mal-distribution of liquid phase.

• Sloshing can reduce the mal-distribution of fluid compared with tilting condition.

• Effects on heat transfer of mal-distribution of feed gas, nitrogen refrigerant, and mixed refrigerant simulated.

Abstract

The mal-distribution of gas–liquid mixtures has serious influences on the heat-transfer performance of plate–fin heat exchangers (PFHEs), which would result in an adverse effect on the stable and efficient operation of a natural gas liquefaction plant. Therefore, it is necessary to investigate the phase distribution performance of PFHEs and its influence on the natural gas liquefaction performance. In this study, a two-phase flow distribution experimental system was built to investigate the flow distribution characteristics of a PFHE under different working conditions. The effects of gas–liquid ratio, tilt angle, and sloshing on the gas–liquid distribution were studied. In addition, the influences of mal-distribution of the feed gas, nitrogen refrigerant, and mixed refrigerant (MR) on the heat transfer and natural gas liquefaction performance were analyzed through the Aspen Muse software. The results showed that the greater the gas–liquid ratio and tilt angle, the more uneven the liquid flow distribution. The efficiency of the MR process in the tilt condition was reduced by 5.2% to 18.5% compared to that in the horizontal condition. Compared with the tilt working condition, the sloshing could reduce the unevenness of the flow distribution. To satisfy the natural gas liquefaction rate of more than 90%, the critical standard deviation values of flow distribution unevenness of the feed gas, nitrogen refrigerant, MR, and MR under the tilt condition were 8.33, 2.95, 4.42 (liquid phase), and 1.42 (liquid phase), respectively. In the process design, to ensure the output and liquefaction rate, the amount of the refrigerant circulation and power consumption should be kept at the margin of approximately 6% and 4% for the nitrogen expander and MR liquefaction processes, respectively.

Introduction

Climate change is one of the main concerns of the 21st century [1] and one of the most pressing challenges is to eliminate the greenhouse gas emissions from industrial and transportation processes [2]. The combustion of natural gas, compared with that of oil and coal, can greatly reduce the emission of carbon dioxide while generating the same amount of energy [3]. According to the Outlook for Energy [4], the demand for natural gas is forecasted to increase by 45% from 2015 to 2040 [5]. The growing demand for natural gas is encouraging more companies in many countries to transfer its exploitation from land to the sea [6]. The floating liquefied natural gas (FLNG) platform can be a candidate for the exploration of offshore natural gas [7]. In previous studies (He et al. [8], Qyyum et al. [9], Khan et al. [10]), the mixed refrigerant (MR) liquefaction process and N₂ expander-based process were considered to be the most attractive and suitable candidates for offshore LNG production. For the offshore MR liquefaction process, Barclay and Shukri [11] introduced an enhanced single mixed refrigerant (SMR) liquefaction process, which used seawater as the cooling medium for offshore natural gas liquefaction. Li and Ju [12] designed a MR liquefaction process for LNG-floating production storage and offloading (LNG-FPSO), which takes the performance parameters, economic performance, and layout, among other aspects, into consideration. Lee et al. [13] also designed three different liquefaction processes for FLNG based on SMR liquefaction process. They claimed that the proposed process was suitable for FLNG because it could reduce the power consumption and number of columns. Hwang et al. [14] presented a dual MR liquefaction process for LNG-FPSO and optimized the operation parameters based on a genetic algorithm and a sequence quadratic program. For the expander-based liquefaction process, Remeljej and Hoadley [15] recommended the nitrogen expansion process for offshore LNG plants based on the exergy analysis of four different liquefaction processes. Li and Ju [12] compared the nitrogen expansion liquefaction process with MR liquefaction for LNG-FPSO and found that the nitrogen expansion liquefaction process was more suitable for LNG-FPSO. Khan et al. [6] presented a N₂-CO₂ expansion process for offshore applications and optimized the operation parameters to minimize the energy consumption. The PFLNG Satu, which has already been producing LNG worldwide, adopted a modified nitrogen expansion process (AP-N™) [16]. Qyyum et al. [17] proposed an innovative propane–nitrogen two-phase expander refrigeration cycle for energy efficiency, and the results showed that its specific energy consumption can be reduced up to 46.4% and the overall energy can be reduced from 79.2% to 29.5% as compared to those of the previously reported N₂ single-expander LNG processes. Later, they proposed a hybrid vortex-tube turbo-expander refrigeration cycle for performance enhancement of the nitrogen-based natural gas liquefaction process and found that the proposed hybrid configuration saved up to 68.5% in terms of the overall specific energy requirement in comparison with that presented in previous studies [18]. Recently, Qyyum et al. [19] proposed an innovative self-recuperative expander-based integrated process to produce LNG–LPG–pentane plus (condensate) at an offshore site, which uses N₂ self-recuperation rather than external precooling with 80% less energy consumption than that required by existing single N₂ expander processes.

However, as a key equipment in FLNG, heat exchangers are susceptible to sloshing, which has a great impact on their inner fluids [20]. PFHEs are widely used in natural gas liquefaction and vaporization processes owing to their high heat transfer efficiency, more compact structures, less land occupation, and high adaptability. However, the heat transfer performance would be easily influenced by the flow mal-distribution, resulting in lower natural gas liquefaction efficiency [21]. Many studies have explored the factors influencing the mal-distribution of PFHEs by experimental and numerical methods. Zhang et al. [22] used a computational fluid dynamics (CFD) simulation to predict the flow distribution in the inlet header of a PFHE. It was found that the flow mal-distribution in PFHEs is definitely serious owing to the improper header configuration and this is detrimental to heat transfer. The flow mal-distribution is heavier at greater Reynolds numbers. Jiao et al. [23] experimentally studied the effects of the inlet pipe diameter (Φ_1(in)), the first header’s diameter of equivalent area (Φ_1(out)), and the second header’s diameter of equivalent area (Φ_2(in) and Φ_2(out)) on the flow mal-distribution in PFHEs. The results indicate that the flow distribution in PFHEs becomes more uniform when Φ_1(out)/Φ_1(in) is equal to Φ_2(out)/Φ_2(in). Wen et al. investigated the flow characteristics of the flow field in the entrance of a PFHE by means of particle image velocimetry (PIV) [24] and a numerical simulation (with CFD) [25]. Both the numerical and experimental results indicate that the fluid mal-distribution in the conventional entrance is a serious issue, whereas an improved header configuration with a punched baffle can effectively enhance the uniformity of flow distribution [26]. The two-phase flow distribution in a PFHE has been experimentally studied under different operation conditions by Wang et al. [27]. They found that the liquid–phase distribution uniformity deteriorates with the decrease in Re_gas and Re_liq, and the gas-phase distribution uniformity deteriorates with Re_liq but improves with Re_gas. Raul et al. [28] investigated the effect of fluid flow mal-distribution in the inlet header configuration of the heat exchanger by numerical simulation, and then they proposed and simulated a modified header configuration with a double-baffle plate having two arrangements. The numerical results indicate that the flow mal-distribution is serious in the conventional header, whereas in the improved configuration less mal-distribution occurs. Yuan et al. [29] studied the distribution characteristics of a special distributor. The major feature of this distributor is that, ahead of the distributor, gas and liquid enter separately in the distributor. The results show that the proposed distributor can appreciably improve the flow distribution of the PFHE. Yang et al. [30] proposed a mathematical model to evaluate quantitatively by the CFD technology the effect of flow mal-distribution on the thermal and hydraulic performances of a parallel PFHE based on the achieved flow distribution. The obtained results suggest that the mathematical model, which intends to provide a design theory of multi-channel heat exchangers to solve the problem of flow mal-distribution, is flexible, accurate, and informative. Zhou et al. [31] investigated the flow distribution of a single-phase flow into a central-type compact parallel heat exchanger by using CFD simulations. The effect of the interactive influence between the geometrical parameters on the flow distribution was considered, and the influences of key geometrical parameters were also studied. In addition, three types of flow distribution were pointed out. In order to improve the two-phase flow distribution in the header of a PFHE, Tu et al. [3] proposed a method by attaching a porous baffle or modifying the inlet nozzle configuration, and the results showed that a vane swirler with a diffuser enhanced the flow distribution of both the liquid and gas phases of the fluid.

The above investigations were carried out under steady horizontal conditions. Few studies have been conducted on flow distribution under sloshing conditions. Recently, Zheng et al. studied the two-phase distribution performance of PFHEs under sloshing conditions by an experimental method. The influence of the sloshing form, sloshing amplitude, and sloshing period on the gas–liquid distribution is discussed, with flow ratio m*_k and standard deviation STD_k as the distribution index. The results in the experimental condition showed the non-uniformity of the two-phase mixture under sloshing conditions. In their other work, the effects of vapor quality (0.2–0.4) and mass flow rate (647–1310 kg/h) under sloshing conditions were also analyzed. The results showed that by increasing vapor quality, the gas distribution performance in the distributor changed more significantly than the liquid one, and the liquid distribution performance in the low vapor quality cases strongly depended on sloshing conditions [32].

However, the gas–liquid distribution in the header of a PFHE under tilt and sloshing conditions need to be further investigated. Especially, to our knowledge, the effects of flow mal-distribution of the feed gas, nitrogen refrigerant, and MR on heat transfer and liquefaction performance are rarely reported. Thus, it is worth obtaining a better understanding of the flow distribution performance under different conditions and the effect of flow mal-distribution of the feed gas, nitrogen refrigerant, and MR on heat transfer and liquefaction performance. This is important for improving the heat exchanger header to improve the efficiency of the heat exchanger and raise the liquefaction efficiency. In this article, an experimental device for the single-phase flow or gas–liquid two-phase flow distribution of the header structure under horizontal, tilt, and sloshing conditions is constructed. The work focuses on (1) the flow distribution performance in the header of a PFHE under different conditions; (3) the effect of the gas–liquid ratio and tilt angle, (2) the effect of sloshing, and (4) the effect of flow mal-distribution of the feed gas, nitrogen refrigerant, and MR on the heat transfer and liquefaction performance.