Development of PSA process for separating high purity hydrogen from methane reforming gas

Abstract

In these PSA systems, two groups of single adsorbent beds undergo a separate cycle while the two groups are interconnected at some steps. Because these two groups can have a different cycle, the PSA process can make the best use of the two adsorbents and obtain a number of very high purity products with great efficiency. However, since these PSA processes need very complicated connecting lines, many valves, and adsorption beds, they lead to high capital cost. Therefore, if many high purity products are not needed simultaneously, two groups of adsorption beds can be combined into beds where two different adsorbents are superposed to each other (such beds are called layered beds). In this case, although it is difficult to produce simultaneously more than two high purity products from a multi-component feed, at least one high purity product can be obtained effectively because this type of a bed can remove many disadvantages caused by using only one adsorbent. In the case of the H2 PSA with layered beds, the activated carbon layer at the bottom of the bed does a bulk separation as a separator and the zeolite layer purifies the raffinate stream from the activated carbon separator. Practically, many H2 PSA processes have employed layered beds with several adsorbents and included guard section additionally in order to prevent the detrimental impurities from entering the main bed. The process performance of a PSA is usually defined by the purity and the recovery, and the optimization of PSA process by executing experiments and mathematical modelling plays a key role in maximising profits in the process design and the operation. Typical decision variables for the optimization include step time, cycle pressure history, purge to feed (P/F) ratio, and throughput, etc. Chapter 1 introduces the importance of separation processes for high-purity hydrogen (99% or more) production, the application of PSA (Pressure Swing Adsorption) process as it is determined to be one of the most promising and economic for hydrogen separation from steam reforming off-gas, and the objectives of the research. Chapter 2 states the required mathematically approached model functions for PSA process, for example material balance equations, energy balance models, and absorption theory equations such as Langmuir and Langmuir-Freundlich. Chapter 3 presents the characteristics of absorbent, the experiment method and equipment for absorption isotherm, the equipment installation for absorption dynamics, and the introduction and experiment method of 4-BED PSA process. In Chapter 4, the adsorption experiments for H2, CO2, CO, CH4 and activated carbon absorbent on zeolite 5A were performed by static volumetric method at temperatures of 293.15, 303.15 and 313.15 K under 30 atm. The parameters obtained from single component adsorption isotherm. Multicomponent adsorption equilibrium are predicted and compared with the experimental data using Langmuir isotherm, Langmuir-Freundlich isotherm and Dual-Site Langmuir isotherm. As a result, Langmuir-Freundlich isotherm and Dual-Site Langmuir isotherm offered more precise result than the other models. In chapter 5, through the dynamic characteristics of adsorption using layered bed were studied. With the model, the optimum length of activated carbon layer that maximizes the bed utilization was determined. This optimum height of activated carbon layer depended on the feed composition and feed velocity. The expected value using the model is similar to the data obtained from the experiment. Chapter 6 illustrates the experiment in order to observe the effect of process parameters such as amount of gas intake, absorption pressure and time on PSA process in the quaternary system. The optimized operation condition investigated by the experiment harmonizes with the computational data result. In this experiment, the optimized operation condition to get high recovery rate and high-purity hydrogen is 9.4 LPM of mass flow rate, 9 atm of adsorption pressure, and 50 seconds of absorption time in the quaternary system. Chapter 7 In Bench-scale PSA process the optimized recharge height for activated carbon and zeolite is 100 cm and 70 cm respectively to produce 1 Nm3/hr. According to the result of experiment and computational calculation, residence time is a significant design factor when modeling an activated carbon bed. Therefore, in order to achieve 99.999% pure hydrogen or purer and hydrogen recovery rate of 75% in the PSA process, the operation conditions with retention time more than 38 seconds and within the appropriate sized of tube diameter in activated carbon bed is suggested.