【作者】 刘聪敏；

【导师】 周亚平；

【作者基本信息】 天津大学 ， 物理化学， 2010， 博士

【摘要】 煤层气是一种高品质的气体燃料,具有相当大的实用价值。开发利用煤层气对于充分利用能源,改善能源结构,减少煤矿瓦斯灾害,保护大气环境都有极为重要的意义。然而目前煤层气资源未得到充分利用,这主要是由于煤层气中含有高浓度的非可燃性气体。一般通过抽放气法开采得到的煤层气中甲烷浓度在20% 45%,而管道输送一般要求甲烷浓度高于90%。抽放煤层气无法满足管道输送的最小热值要求。氮气作为煤层气中主要的非可燃性气体,与甲烷之间的分离是煤层气利用的关键技术之一。变压吸附技术由于能耗小、操作灵活,投资和运行费用低等优势在煤层气甲烷分离浓缩方面受到广泛的关注。本文采用计算机模拟方法结合实验对CH₄/N₂在活性炭上的吸附机理进行了深入研究,并在此基础上对传统变压吸附过程进行了改进,实现了低浓度煤层气的浓缩分离。（1）采用10-4-3的势能模型计算了不同孔宽下活性炭与甲烷、氮气之间的相互作用。同时,利用Material Studio 4.0软件中的Sorption模块,采用GCMC方法模拟计算了甲烷和氮气混合气在不同孔宽的活性炭孔内的吸附行为,考察了吸附剂孔宽、吸附压力对甲烷与氮气的分离效果的影响。由计算结果可以看出,甲烷与氮气分离因子受孔宽的影响较大,在孔宽为0.75 nm左右时,氮气和甲烷分离因子达到最大值;当孔宽大于1.3 nm时,分离因子基本不再发生变化。因此微孔（0.75～1.3 nm）含量高的活性炭有利于甲烷和氮气的吸附分离,该结论的得出对于变压吸附分离吸附剂的选择以及吸附剂的合成具有极大的指导意义。（2）采用水蒸气和CO2活化相结合的方法制备了八个具有不同孔宽分布的活性炭材料。通过动态法测定甲烷和氮气混合气在八个样品上的穿透曲线,并计算其分离因子。实验结果表明,低压有利于甲烷和氮气的吸附分离,随着压力的升高分离因子逐渐减小。根据GCMC模拟中计算得到的不同孔宽下甲烷和氮气吸附等温线,建立了基于吸附剂孔宽分布的甲烷/氮气在活性炭上的分离因子计算模型。通过对比实验结果和计算结果可以发现,采用该模型可以准确的计算CH₄/N₂在活性炭上的分离因子,该计算模型的建立可以很大程度上简化筛选吸附剂的实验过程,从分离效果的角度评价吸附剂优劣。（3）传统的变压吸附分离技术中,一般以混合气中的轻组份气体为产品气,相对于强吸附组份,轻组份可以作为塔顶产品具有浓度高,回收率高的优势。然而在CH₄/N₂混合气中,甲烷是强组份气体,同时又是产品气,因此变压吸附浓缩CH₄/N₂与典型变压分离过程意义完全不同,将传统的变压吸附分离过程应用于CH₄/N₂混合体系的分离具有较大的困难。本文在传统的变压吸附流程加入CO2置换步骤,成功地将甲烷由塔底产品转变为塔顶产品,在实现CH₄/N₂变压吸附分离的同时,得到了高浓度的甲烷产品气。本流程包括以下四个阶段:充压,高压吸附,CO2置换以及吸附剂再生。本实验采用了实验室自制的活性炭作为吸附剂,配制了一系列甲烷浓度（17.62%,22.30%,32.06%,40.28%和51.33%）的原料气用于置换分离实验。通过实验确定了不同浓度原料气的最佳吸附时间,产品气浓度与回收率的关系,摸索了吸附床再生条件以及置换压力对于产品气浓度回收率的影响。通过实验发现,采用置换方法分离低浓度CH₄/N₂混合气是可行的,可以将甲烷由塔底产品转换为塔顶产品。在维持90%以上回收率的条件下,产品气中甲烷平均浓度均达到88.5%以上。对于不同浓度的原料气来说,随着甲烷浓度增大,其产品气的浓度和回收率都有明显的提高。甲烷浓度为17.62%时,0.4MPa置换后,产品气浓度为84.4%,回收率为91.4%;常压置换后,产品气浓度为98%,回收率为90%;甲烷浓度为51.33%时,0.4MPa置换后,产品气浓度为91.8%,回收率可达96.5%;常压置换后,产品气浓度为98.5%,回收率为90%。更多还原

【Abstract】 Coalbed methane （CBM） is a kind of natural gas buried in seams that have not been mined and it amounted up to 1x1014m3. The methane rich gases captured on methane drainage in operational mines contain different amounts of non-combustible gases, most of which is nitrogen. The concentration of methane varies for a wide range, but mostly in 20-45 %, which do not meet the specification on pipeline gas for the minimum heating value, for which the methane content is typically larger than 90 %. To utilize CBM and prevent from its emission into troposphere, the separation between CH4 and N2 has to be carried out. PSA is supossed to be a good choice for its low investment and energe cost. In this paper, we have focus on two respects of CH₄/N₂ separation. The frist part of the work is to study on the adsorption mechanism between on the CH4 and N2 on activated carbons; the second part of the work is to modify the conventional PSA cycle to achieve separation of CH₄/N₂.（1） The present work considered the separation on the basis of adsorption mechanism. The adsorption potential was calculated relying on the 10-4-3 potential model and the slit pore model of activated carbons. At the same time, the adsorption isotherms of CH₄/N₂ at 298 K are obtained for different pore widths by using a Grand Canonical Monte Carlo simulation （GCMC）. The results of the calculation and simulations showed that the largest selectivity between CH4 and N2 occurred in pore size of 0.75 nm, and there was no difference between them for pores larger than 1.3nm. It has been known that supercritical gases follow the adsorption mechanism of monolayer surface coverage; however, the surface of adsorbent has not been fully covered when the partial pressure of CH4 or N2 is low, and the adsorption potential dominates the separation.（2） To study the effect of adsorption mechanism on the separation, eight carbon samples corresponding to different extent of activation were prepared .The separation coefficient between CH4 and N2 on the eight carbons was evaluated based on breakthrough curves .It was experimentally shown that the separation coefficient between CH4 and N2 increases following the increase of micropore volume, and decreases following the increase of pressure. The corresponding selectivity of CH₄/N₂ on the real activated carbon is predicted with the PSD determined by the DFT method of the adsorbents and simulation results and the simulation results match the experiment perfectly.（3） For the mixture of CH₄/N₂, methane, which is the strongly adsorbed component and obtained as bottom product in the typical PSA or VSA process, is the targeted product. However, methane is expected to be upgraded and obtained as top product, which is totally opposite with the typical PSA process in the role of each step. A new step, CO2 replacement, was implemented into conventional PSA cycle. The new cycle of pressure swing operation composed of four steps: pressurization, adsorption at an elevated pressure, CO2 replacement, and bed regeneration. The pressure swing adsorption experiments were carried out using activated carbon made in our laboratory, as adsorbent. A series of CH₄/N₂ mixtures containing 17.62 %,22.30 %,32.06 %,48.28 % and 51.33 % methane was used to simulate the methane rich gases. Effect of operational variables on the concentration and recovery of methane in the product stream was experimentally studied. It was shown that methane content in the product stream is higher than 88.5 % with methane recovery higher than 90 % for all feeding gases tested. It was shown also that replacement at ambient pressure is more efficient than the replacement at adsorption pressure. Both methane concentration in product and its recovery remarkably increased following the increase of methane content in feeding streams. When the displacement pressure at 0.4MPa , the recovery and concentration of methane reached 96.5% and 91.8%, respectively, for the methane rich gas containing 51.33 % CH4 and when the displacement pressure at 0.1MPa, the recovery and concentration of methane reached 96.5% and 91.8%, respectively. When the displacement pressure at 0.4MPa , the recovery and concentration of methane reached 91.4% and 84.4%, respectively, for the methane rich gas containing 17.62 % CH4 and when the displacement pressure at 0.1MPa, the recovery and concentration of methane reached 90% and 98%, respectively.更多还原