Experimental and theoretical investigation of a hydrogen liquefaction process

The world approaches a challenging era where new solutions to the arising energy and climate problems must be found. As the need for clean energy carriers increases due to growing energy demand and increased global warming, hydrogen has been introduced as an energy carrier candidate for the future. There are several barriers in order to introduce large scale hydrogen use; like efficient production, storage and distribution. Hydrogen may be stored and transported as pressurized gas or as cryogenic liquid. The preferred method of transportation is dependent on different circumstances such as quantity and distance. Liquid hydrogen has high gravimetric energy density and has therefore the potential for solving the problems of large scale storage and distribution. However, it also has low volumetric energy density, which creates challenges. The low molecular weight of hydrogen and the low temperature of liquid hydrogen makes the compression and liquefaction processes quite energy demanding. In fact, one third of the lower heating value is typically used for liquefying hydrogen in conventional systems nowadays. In this thesis, up to date technology for hydrogen liquefaction and proposals of future high efficient large scale systems has been described. The focus has been on development of technologies which increase the efficiency of liquefaction, thus decreases the barriers for introducing hydrogen use on a large scale basis. The study was made with special emphasis on properties of key components, which affect the efficiency of such systems (Compressors, expanders and heat exchangers). Key figures for these components were also collected from literature and manufacturers to investigate which assumptions that is realistic. This work also resulted in a draft to an article for publication in a journal (attached in APPENDIX F). SINTEF Energy Research has proposed a process for precooling of hydrogen down to 75 K. Conventional processes are relatively small-scale compared to the proposals for future plants, and the typical precooling process is based on usage of liquid nitrogen as a by-product from air separation. The SINTEF process utilizes a multicomponent refrigerant (MCR) in a closed cycle process, similar to those often used in natural gas liquefaction, and has the potential to efficiently precool large amounts of hydrogen. Simulations of this process have been carried out in this thesis work, by the use of both PROII and HYSYS. The in-built properties of hydrogen in PROII had to be modified to match data from literature, and a special property package had to be used in HYSYS. A comparison has been made between the two programs to verify the simulation results. No large differences were found, but the modification of PROII properties made some errors in the H2 flash gas amount, which propagated to the rest of the process. Different methods for simulating continuous ortho para conversion have also been investigated, and different methods were found preferable in HYSYS compared to PROII. Another investigation that has been performed is the effect of splitting up the expansion process of liquefied hydrogen down to storage pressure into a number of stages with ortho para conversion between. Energy can be saved by introducing more stages, pursuant to the analysis, but it would be a question of CAPEX vs. OPEX to decide if it is defensible in a real process. A small-scale laboratory plant has been built by SINTEF to experimentally investigate the multicomponent precooling process. Experimental work on this plant was carried out to investigate the behaviour of the precooling process in real life. Due to delays in the construction of the plant, the experiments described are early phase experiments. The focus has been on testing of the equipment and then lowering of the temperature towards steady state liquefaction. Problems occurred during the experiments, and the reason for them have been analysed. Temperatures down to -140 °C (without hydrogen heat load) were reached in these early phase experiments (required temperature is -200°C with load). The possibility of replacing the Joule-Thomson valves in the MCR circuit with work extracting expanders has been investigated to see the potential for improvement of the efficiency. This would result in a possible recovery of work, but more importantly better cooling effect. The result is promising, but this is also a question of system complexity, reliability and investment costs. The results from this thesis show that there is large potential for improvement of the efficiency of hydrogen liquefaction, which accordingly may help overcome the barriers for introducing hydrogen as a large scale energy carrier.