Online Analysis of Gas Composition for Explosion Tests at Elevated Pressure

[[abstract]]摘要 本研究提出一套線上氣體分析系統，來量測在高壓條件下氣體爆炸前後的氣體組成，此系統利用線上的自動進樣系統來取樣導入氣相層析儀-熱導偵測器(GC-TCD)作定性及定量之分析，以期能藉由可靠的數據來了解判斷爆炸界限及爆炸測試氣體組成變化量與爆炸昇壓之間的關係。 在爆炸測試系統上，延用壓力平衡及雙重容器設計之理念，加裝了點火器、混合槽以及快速的數據擷取系統，成為一簡單、安全又具有高壓量測爆炸界限之設備下。在分析系統上，首先將爆炸系統以儀器連線方式將氣體管線連接至緩衝裝置降壓固定流速，再由自動進樣系統將樣品導入氣體進樣閥，利用閥轉動改變氣體走向，分別進入氣相層析儀中PLOT-Q (porous layer open tubular column - 30m×0.53mm)及PLOT-5A (Molecular sieves 5Å - 30m×0.53mm)之不同性質分離管柱中，經適當的層析條件分離後，進入熱導偵測器(TCD)分析定量，此線上分析系統經驗證所得CO、CO2、O2、CH4等待測物之檢量線線性均可大於0.9950以上，線性範圍均可涵蓋待測物，準確度約在101.4±8.7〜105.4±5.9%，精密度則在2.9- 4.3%範圍內，所獲得之方法偵測極限約可達0.01%，整個系統經不同來源之氣體驗證其可靠性約在99-106 %。 本研究在高壓0.6MPa條件下，進行CO/CH4(2：8)及純CH4爆燃測試，得到CO/CH4(2：8)在空氣中之爆炸上限為空氣濃度75.8%及純CH4爆炸上限為空氣濃度78.5%，且發現在CO/CH4(2：8)爆燃測試中，空氣66.47〜74.52%有微小的昇壓，其分析結果顯示也有微小的CO2生成，但量小於可定量範圍(QDL<0.1%)，且無CO生成，推論是由於鎳鉻絲點火器產生高溫火花的瞬間，使部份甲烷與週遭局部空氣起反應，而不是甲烷的部份氧化反應，又因氧氣的不足使得無法將火熖傳播出去造成真正的燃燒，僅造成微小的壓力波傳遞。故此區域可排除在爆炸範圍之外。而在空氣濃度為76.00%時，爆燃產生大昇壓的分析結果中有明顯的CO及CO2生成量，證明為真正燃燒行為。 在一連串的實際爆炸氣體樣品分析數據中，統計其在本系統之條件下所得到的重複分析之數據均在落於 2.90-3.51%之範圍內，顯示本系統於實際樣品之分析時有其一定之穩定度，此可知本系統在實際爆炸分析之應用上有其快速、正確之適用性。[[abstract]]ABSTRACT In this study, an online gas measurement system is developed to measure the gas composition before and after explosion tests. This system uses an auto-sampling system to sample gases into a gas chromatography coupled with thermal conductivity detector (TCD) which capable of qualitative and quantitative analysis. The data will be useful for determine the flammability limits and the relationship between composition change and explosion pressure rise. In the explosion testing system, the pressure balancing and double containment vessel design is adapted. With an igniter, a premix vessel, and a fast data acquisition system, the system is a simple and safe apparatus that is capable of testing explosion at elevated pressures. In the analytical system, a gas sampling line is first connected from the explosion system to a buffer to reduce pressure and confine flow rate. The sampled gas is routed to a gas feed valve. The valve alternates the sampled gas into two different columns, PLOT-Q (Porous layer open tubular column - 30m×0.53mm) and PLOT-5A (Molecular sieves 5Å-30m×0.53mm), for separating all gases before entering the TCD. This online system is validated to have linearity greater than 0.9950 for CO, CO2, O2, and CH4. Accuracy was within 101.4±8.7〜105.4±5.9%. Precision was within 2.9- 4.3%. The method detection limit was found to be around 0.01%. The whole system is validated with different sources of gases and its reliability is found to within 99-106 %. Explosion tests for CO/CH4 (2:8 mixture) and pure CH4 were performed at pressure of 0.6 MPa. It is found that the upper flammability limits of CO/CH4 (2:8) mixture and pure CH4 lie at air concentration of 75.8% and 78.5%, respectively. Small pressure rises were found in CO/CH4 tests where small amount of CO2 were also found but their concentrations were less than quantifiable range (QDL