用於萃取蒸餾塔序列節能技術之比較

Abstract

萃取蒸餾為一常見的共沸物分離方法，因其經濟可行性較高且動態控制較為穩定，然而此方法之能耗相當可觀，屬能源密集型單元操作，為此，研究人員發展出數種節能技術以降低其能源消耗，包括熱耦合、液態側流與熱整合等技術。雖然現有文獻不乏大量針對此些技術之研究與應用，但並無快速且通用的標準可幫助程序設計，系統性的評估與篩選方法仍有待發展。

為了解決此不足，本研究藉由十七組二元最小共沸物系統與重夾帶劑配合六種強化蒸餾序列與兩種基本序列之優化結果得出一系列趨勢及現象，且透過觀察與比較這些系統之能耗與年化總成本，本研究提出一個新指標RCB用以評估這些節能技術之合適性，針對不同條件與情況快速篩選出較具成本效益之技術和序列。

依照RCB之數值變化，本研究歸納出下列結論: (1) 當不考慮再沸器與冷凝器之熱整合時，如RCB小於約8.4，修正型液態側流序列 (MSSS)表現最佳，如RCB大於約8.4，雙塔液態側流序列 (DCSSS)表現最佳。 (2) 當共沸物重成分與夾帶劑之相對揮發度過高時，即使再混合效應相當嚴重，液態側流也不具成本效益。 (3) 當考慮再沸器與冷凝器之熱整合時，如RCB小於約2.2，普通熱整合序列(OSS)具有最佳表現，如RCB大於約2.2，雙塔液態側流熱整合序列(SDCSSS)具有最佳表現。 (4) 當RCB較小時，再沸器與冷凝器之熱整合較有效，當RCB較大時，液態側流序列較有效。 (5) 當RCB介於2.6與8.4之間時，再沸器與冷凝器之熱整合搭配液態側流序列效果較佳，如RCB小於2.6，液態側流幫助甚微，不建議使用，如RCB大於8.4，再沸器與冷凝器之熱整合幫助甚微，不建議使用。 (6) 當RCB較小時，隔牆塔缺乏成本效益，當RCB較大時，隔牆塔可降低能耗與成本，但其表現仍不及液態側流序列。

Extractive distillation is a widely-used but energy-intensive method for separating homogeneous azeotropic mixtures into pure components. The most common case is a minimum-boiling azeotrope separated using a heavy entrainer. Numerous design alternatives have been proposed for reducing energy consumption in such processes, including alternatives with side-streams, column stacking and thermally coupled sequences. However, there is very little guidance in the literature about which alternative is likely to work the best in a given situation.

To address this deficiency, in this work, seventeen industrially-relevant chemical systems comprising two species that form a minimum boiling azeotrope and a heavy entrainer are selected for study. Two ordinary and six energy-saving flowsheet alternatives for extractive distillation processes were also selected. Each of the eight flowsheet alternatives was optimized for each of the 17 chemical systems. A novel metric RCB (the ratio of the entrainer flow rate to the heavy key flow rate) is proposed to facilitate interpretation of the results. The following general trends are observed when comparing flowsheet alternatives for chemical systems with different values of RCB: (1) When comparing sequences without stacking, the modified side-stream sequence (MSSS) performs better when RCB is smaller (less than about 8.4); the double-column side-stream sequence (DCSSS) performs better when RCB is larger (greater than about 8.4). (2) When the relative volatility of the heavy key and the entrainer is too high, side-stream sequences may be uneconomical even if the remixing effect is substantial. (3) Considering sequences with column stacking, the ordinary stacked sequence (OSS) performs better when RCB is smaller (less than about 2.2) and the stacked double-column side-stream sequence (SDCSSS) performs better when RCB is larger (greater than about 2.2). (4) Column stacking is more effective when RCB is small, and side streams are more effective when RCB is large. (5) The stacked side-stream sequence is recommended for intermediate values of RCB (2.6<RCB<8.4). For RCB<2.6, the benefit of the side-stream is minimal and a sequence with stacking only is probably preferred for simplicity. For RCB>8.4 the benefit of stacking is minimal and a sequence with a side-stream only is probably preferred for simplicity. (6) The dividing-wall column (DWCU) is not economical when RCB is small and is economical but still less attractive than side-stream sequences when RCB is large.