Efficient Hydrogen Liquefaction Processes

Abstract

The problem is that today every H2 liquefaction plant has low exergy efficiency of just between 20–30%. It is based on the pre-cooled Claude system, which is still the same as 50 years ago with littleimprovement. Method to resolve the challenges of the future plants is finding a completely newconfiguration with more efficient system. For this dissertation, a multi-component refrigerant (MR)refrigeration cycle is proposed to solve the problem. The work is divided into four parts: a literaturereview, a design and simulation of a small-scale laboratory plant, an experiment with the smallplant, and a design and simulation of a proposed large-scale plant. First, this study investigated thesimulation of a newly proposed small-scale laboratory liquid hydrogen plant with the new,innovative MR refrigeration system. The simulated test rig was capable of liquefying a feed of 2kg/h of normal hydrogen gas at 21 bar and 25 oC to normal liquid hydrogen at 2 bar and −250 oC.The simulated power consumption for pre-cooling the hydrogen from 25 oC to −198 oC with thisnew MR compressor was 2.07 kWh/kgGH2. This was the lowest power consumption available whencompared to today’s conventional hydrogen liquefaction cycles, which are approximately 4.00kWh/kgGH2. Exergy analysis of the test rig’s cycle, which is required to find the losses and optimize the proposed MR system, was evaluated for each component using the simulation data. It was foundthat the majority of the losses were from the compressors, heat exchangers, and expansion valves.Then, a small-scale laboratory hydrogen liquefaction plant that contains the new innovative MRrefrigeration system was constructed to verify the simulation of this system. Initial experimentsindicated that the rig was able to adequately cool normal hydrogen gas from 25 oC to −158 oC at aflow rate of 0.6 kg/h using a simplified 5-component MR composition refrigeration system. Thepower consumption of pre-cooling from the MR compressor was 1.76 kWh per kilogram of feedhydrogen gas. After two weeks, the lowest attained temperature was about −180 oC when a fewadditional grams of nitrogen gas were charged into the rig. There were some differences, but mostof all, the simulation and experimental data were in good agreement. The primary conclusion wasthat pre-cooling hydrogen gas with the MR refrigeration system resulted in a lower energyconsumption per kilogram of feed hydrogen gas compared to conventional refrigeration systems.Finally, a liquid hydrogen plant based on the MR refrigeration system is proposed. A cycle that iscapable of producing 100 tons of liquid hydrogen per day is simulated. The MR system can be usedto cool feed normal hydrogen gas from 25 oC to the equilibrium temperature of −193 oC with a highefficiency. In addition, for the transition from the equilibrium temperature of the hydrogen gas from−193 oC to −253 oC, a new proposed four H2 Joule-Brayton cycle refrigeration system withoptimization is recommended. The overall power consumption of the proposed plant for the basedcase is 6.35 kWh/kgLH2. The current plant in Ingolstadt is used as a reference, which has an energyconsumption of 13.58 kWh/kgLH2 and an efficiency of 21.28%. The efficiency of the proposedsystem is around 45% or more, where this depends on the assumed efficiency values for thecompressors and expanders, together with effectiveness of heat exchangers. Importantly, thevariables and constraints are preliminary studied together with how to adjust these to achieveoptimal steady-state operation. The optimization problem has 23 variables and 26 constraints. Asimplified 5-component composition of refrigerant suggested for the plant is found. The plantoptimization was also conducted with two more pinch temperatures (1 and 3 oC). Power saving isincreased with a pinch temperature of 1 oC as compared to 3 oC. This figure can have a significantimpact on plants selection. In addition, pressure drops in heat exchangers are also employed in thesimulation for the study, however it is shown that they don’t have much significant impact on theoverall plant total power consumption. The proposed system has smaller compressor motors andsmaller crankcase compressors; thus, it could represent a plant with the lowest construction costwith respect to the amount of liquid hydrogen produced in comparison to today’s plants, e.g., in Ingolstadt and Leuna. Therefore, the proposed system has many improvements that serves as anexample for future hydrogen liquefaction plants.