| 摘要: | 減碳已成為國際趨勢，因此碳捕獲、利用與封存 (Carbon Capture, Utilization and Storage, CCUS) 之相關技術至為關鍵，而碳捕獲占了CCUS技術中最大宗之成本。變壓吸附程序(Pressure Swing Adsorption, PSA)為其中一種捕獲二氧化碳之技術，其具有能耗低、操作成本低等優點，故該技術具備商轉化之潛力。本研究採用此技術進行捕獲台中發電廠燃煤機組排放之煙道氣中二氧化碳之實驗，而研究目標為使捕獲之二氧化碳純度達95%。
本研究利用台灣電力公司綜合研究所提供之三塔式變壓吸附設備，以COSMO 13X 沸石作為吸附劑，進行三塔十二步驟真空變壓吸附程序捕獲燃煤電廠煙道氣中二氧化碳之實驗。經預處理之煙道氣中二氧化碳濃度約為10.48%至13.35%。而三塔十二步驟程序由高壓吸附、壓力平衡、同向減壓、逆向減壓與靜置等程序步驟組成。本研究採用實驗設計法規劃並進行實驗，目標為分析出各響應之顯著因子，並找出最適化操作條件。本研究採用兩水準三因子之全因子實驗設計，選擇以步驟3/7/11時間(A因子)、進料壓力(B因子)、同向減壓壓力(C因子)為探討之因子。
分別建立塔底二氧化碳純度、回收率、能耗之迴歸模型。再藉由確認各響應之相對誤差皆不大，推測迴歸模型應具有不錯的預測能力。
接著由變異數分析之結果，分析出各響應之顯著因子或交互作用。對塔底二氧化碳純度而言，僅主效用B(進料壓力)與主效用C(同向減壓壓力)為顯著因子; 對塔底二氧化碳回收率而言，僅主效用A(步驟3/7/11時間)與主效用C(同向減壓壓力)為顯著因子; 對能耗而言，僅主效用B(進料壓力)與交互作用BC(進料壓力同&向減壓壓力)為顯著因子或顯著之交互作用。
以各響應之迴歸模型找出最適化操作條件，得當步驟3/7/11時間為200秒、進料壓力為3.50 bar，以及同向減壓壓力為0.30 bar時，最適化實驗結果為塔底二氧化碳純度達95.26 %、回收率達65.33 %，以及能耗為 1.33 GJ/tonne of CO2。
最後，將吸附劑COSMO 13X 沸石在經以燃煤電廠煙道氣為進料之長時間真空變壓吸附程序操作後，取少量樣品並以高壓氣體吸附分析儀進行二氧化碳等溫吸附測試，由數據趨勢可推測: 吸附劑13X的吸附量，隨著歷經長時間真空變壓吸附程序之次數增加有持續下降之趨勢。
;Carbon reduction has become a global trend, therefore, technologies related to Carbon Capture, Utilization and Storage(CCUS) are very important. Carbon capture accounts for the most costs in the Carbon Capture, Utilization and Storage(CCUS). Pressure Swing Adsorption(PSA) is one of the technologies to capture carbon dioxide and has great potential to be operated industrially due to its lower energy consumption and lower operational costs. The research aims to capture carbon dioxide in flue gas at Taichung coal-fired power plant with CO2 product purity reaching 95% by pressure swing adsorption.
COSMO zeolite 13X was used as adsorbent. Three-bed twelve-step vacuum pressure swing adsorption(VPSA) experiment to capture carbon dioxide in flue gas at coal-fired power plant was conducted by 3-bed pressure swing adsorption equipment provided by Taiwan Power Company. After pretreatment, the CO2 concentration of flue gas was about 10.48% to 13.35%. The whole process comprising of adsorption, pressure equalization, cocurrent depressurization, countercurrent depressurization and idle steps. Experiments were planed by design of experiments(DOE). The goal of DOE is to find the significant factors for each response and find the optimal operating conditions. Two-level three-factor full factorial design was used, and the factors studied were step 3/7/11 time(factor A)、adsorption pressure(factor B)、cocurrent depressurization pressure(factor C).
Regression model for bottom CO2 product purity, recovery and energy consumption were established. The prediction accuracy of regression model were presumed to be good because the relative error of each responses were not big.
Next, find the significant factors or interaction by analysis of variance(ANOVA). For the bottom CO2 product purity, main effect B and main effect C are significant. For bottom CO2 product recovery, main effect A and main effect C are significant. For energy consumption, main effect B and interaction BC are significant.
Then, the optimal operating conditions were predicted from regression models of design of experiments(DOE). The performance at optimal operating conditions is: the bottom CO2 product purity is 95.26 % with recovery 65.33 %, and energy consumption were measured to be 1.33 GJ/tonne of CO2 with step 3/7/11 time equaled to 200 s, adsorption pressure equaled to 3.50 bar and cocurrent depressurization pressure equaled to 0.30 bar.
Finally, the CO2 adsorption isotherms were measured for adsorbent after long-term VPSA operation. The data shows that the adsorption capacity of zeolite 13X seems to decay with the used times of long-term VPSA operation increases. |
