201233430 六、發明說明: 【發明所屬之技術領域】 本發明是有關於一種分析裝置異常的偵測方法,特別 是指一種克勞斯製程之尾氣分析儀異常的侦測方法。 【先前技術】 體脫程⑺aus Pr。⑽)是非常重要的硫化氮氣 體脫硫技術,十分廣泛地用^丨1 卞刀璜泛地用於例如石油煉製、天然氣加工 …油砂處理、煤氣化、冶煉、 理製程後產生含有硫化氫的醆氣。’,、、業設施’以處 進行程是將酸氣中1/3的硫化氯與空氣中的氧氣 Γ1ΓΓ而得到二氧化硫,之後與剩下的硫化氣反應 生成硫’由於氧化反應時的空氣 齑供庙“里疋里要的關鍵’空 應太夕或太低皆會降低硫的轉化效率。 由於酸氣的成分、:曹疮 法/辰度在線上無法完全分析出來而無 2由U組成直接推算空氣的最佳進氣量,因此,目1 的作法是利用克勞斯贺- 月] 分析其成分、濃度’由尸i的尾氣以尾氣分析儀 比值而H " 毛礼中的硫化氫與二氧化硫之濃度 持續進行克勞斯製程。"冑值控制4的進氣量以 轉化尾氣分析儀是克勞斯製程維持硫的 無法持續有當尾氣分析儀故障或不準時,則 于3有政地進行克勞斯 分析儀異常的手P㊉U有伯测尾氣 、煉焦廢水中硫氧化生產效率低落、爐溫失控 α物異常增加而導致生化持酸化等重 201233430 法’在尾氣 有效地持續 大損失為此’業者一直在尋求有效的解決方 分析儀異常時能迅速查覺、應變,而能穩定、 進行克勞斯製程。 【發明内容】 因此本發明之目的,即在提供—種可以偏測克勞斯 製程之尾軋分析儀異常的偵測方法。 ,於是’本發明克勞斯製程之尾氣分析儀異常的價測方 法’該克勞斯製程以酸氣與空氣產製硫,並於該克勞斯製 程進行中對產生的尾氣用尾氣分析儀分析其成分濃度後, 依據尾氣中硫化氫與二氧化硫的濃度比值而得到一回饋值 ,進而以該回饋值控制空氣的進氣量以持續進行該克勞斯 製程’該克勞斯製程之尾氣分析儀異常的_方法包含— 水分濃度得到步驟、—模型建立步驟,及—判斷步驟。 該水分濃度得到步驟操取酸氣的多數物性參數得到酸 氣中的水分濃度。 該模型建立步驟以預設的自回歸階數與滑動平均階數 建立-自回歸滑動平均模型,掏取卜最近時點的回饋值 前-預定時段内的多筆回饋值,並連同該水分濃度代入該 自回歸滑動平均模型後得到一建議值。 該判斷步驟是當該建議值與該最近時點的回饋值之差 超出預定範圍時,判定該尾氣分析儀異常。 ,本發明之功效:藉由建立該自回歸滑動平均模型進行 進氣量的預測,並以該水分濃度作為誤差修正,使得到的 該建議值更精準地判斷尾氣分析儀是否異常,並在修復尾 201233430 氣分析儀的期間, 行該克勞斯製裎。 以該建議值控制空氣之進氣量而持續進 【實施方式】 有關本發明之前述及其他技術内容、特點與功致,在 、下配e參考圖式之二個較佳實施例的詳細 清楚的呈現。 將可201233430 VI. Description of the Invention: [Technical Field] The present invention relates to a method for detecting an abnormality of an analysis device, and more particularly to a method for detecting an abnormality of an exhaust gas analyzer of a Claus process. [Prior Art] Body off process (7) aus Pr. (10)) is a very important sulfuric acid nitrogen desulfurization technology, which is widely used for example in petroleum refining, natural gas processing, oil sand treatment, coal gasification, smelting, and chemical processing. Helium from hydrogen. ',,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, For the temple, the key to 'Liu Lili' should be too low or too low to reduce the conversion efficiency of sulfur. Because of the composition of sour gas, Cao sore method / Chen Duo can not be fully analyzed on the line and no 2 consists of U Directly calculate the optimal intake air volume, therefore, the goal of the method is to use Klaus He-month] to analyze its composition, concentration 'from the tail gas of the corpse i to the exhaust gas analyzer ratio and H " The concentration of sulfur dioxide continues to carry out the Claus process. "The value of the intake of the devaluation control 4 to convert the exhaust gas analyzer is the Claus process to maintain the sulfur can not continue. If the exhaust gas analyzer fails or is not on time, then there is 3 The hands of the Claus analyzer are abnormal, the P-U has the measured tail gas, the sulfur oxidation production efficiency in the coking wastewater is low, the furnace temperature is out of control, and the α substance is abnormally increased, resulting in the biochemical acidification and the like. 201233430 Method 'effectively lasts in the exhaust gas For the sake of this, the industry has been seeking an effective solution to the abnormality of the analyzer, and can quickly detect and adapt, and can stably and carry out the Claus process. [Inventive content] Therefore, the object of the present invention is to provide The detection method of the abnormality of the tail-rolling analyzer of the Claus process is measured. Then, the method of measuring the abnormality of the tail gas analyzer of the Claus process of the present invention is the method of producing sulfur from sour gas and air. And in the process of the Claus process, the exhaust gas is analyzed by the exhaust gas analyzer, and then the feedback value is obtained according to the concentration ratio of hydrogen sulfide to sulfur dioxide in the exhaust gas, and then the air intake amount is controlled by the feedback value. The _ method for continuing the process of the Claus process of the Claus process includes: the water concentration obtaining step, the model establishing step, and the - determining step. The water concentration is obtained by the step of taking the majority of the acid gas The physical property parameter obtains the water concentration in the acid gas. The model establishment step is established by the preset autoregressive order and the sliding average order-autoregressive moving average model. Obtaining a plurality of feedback values before the feedback value of the recent time point and the predetermined time period, and obtaining the recommended value together with the water concentration into the autoregressive moving average model. The determining step is when the recommended value and the latest time point are When the difference between the feedback values exceeds the predetermined range, it is determined that the exhaust gas analyzer is abnormal. The effect of the present invention: the intake air amount is predicted by establishing the autoregressive moving average model, and the water concentration is corrected as an error, so that the obtained The recommended value is more accurate to determine whether the exhaust gas analyzer is abnormal, and during the repair of the tail 201233430 gas analyzer, the Claus system is used. The recommended value is used to control the air intake amount and continue to enter [Embodiment] The foregoing and other technical contents, features, and advantages of the present invention are set forth in the detailed description of the preferred embodiments of the present invention. Will be available
在本發明破詳細描述之前,要注意的是,在以下的說 月内谷中,類似的元件是以相同的編號來表示。Before the detailed description of the present invention is broken, it is to be noted that in the following description, similar elements are denoted by the same reference numerals.
參閱® ι,ι發明克勞斯製程之尾氣分析儀異常的債測 方法的帛較佳實施例,如圖丨所示該克勞斯製程以 酸氣與空氣產製硫,並於該克勞斯製程進行中對產生的尾 氣用尾氣分析儀分析其成分濃度後,依據尾氣中硫化氫二 氧化硫的濃度比值而得到一回饋值,進而以該回饋值控 制空氣的進氣量以持續進行該克勞斯製程,該克勞斯製程 之尾氣分析儀異常的债測方法包含—水分濃度得到步驟上上 、一模型建立步驟12,及一判斷步驟13。 該水分濃度得到步驟Π是擷取酸氣的溫度、壓力,與 流ι值後,得到酸氣中的水分濃度。 忒模型建立步驟丨2以自回歸階數配合滑動平均階數而 建立自回歸滑動平均模型(Autoregressive moving average model,ARMA),之後,擷取自一最近時點的回饋 值刖一預疋時段範圍内之多筆回饋值,並連同該水分濃度 代入該自回歸滑動平均模型後得到—建議值,較佳地以 例如動差估计法(Method of moments)、最大概似估計法 201233430 (Maxi麵 likelih〇od estimates,M.L E.)等方式來估算 模型中參數的數值,而建立該自回歸滑動平均模型。 該判斷步驟13是當該建議值與該最近時點的回饋值之 差超出預定範圍時,判定該尾氣分析儀異常。 本發明克勞斯製程之尾氣分析儀異常的偵測方法,藉 由建立該自回歸滑動平均模型以預測空氣的進氣量,利用 該最近時點的回饋值前—預定時段範圍内之多筆回饋值作 為預測的依據,並且因為酸氣中水分濃度會影響與酸氣進 订反應的空氣之量而使得進氣量必須隨之調整,所以,以 該水分濃度進行誤差修正,使得刭 供判斷尾氣分析儀是否異常。 "β t加精準而 :閱圖2,本發明克勞斯製程之尾氣分析儀異常的_ :實際㈣測試的結果,如圖2所示,橫轴為測 且縱轴為空氣的進氣量。在評估測試期間内的回饋 實相^應的建議值的誤差之平均值為找且標準差為心, 確貫此達到預期的功效,並 氣分析儀正常作動且回饋值正確。在已先確認尾 方法Γ-圖第-3較:圭發實=勞:製程之尾氣分析儀異常-貞測 不…例與該第一較佳實施例相類似,所 例、一疋,當判定該尾氣分析儀異常時,該第二較佳實施 以一暫時應變步驟2應變處理。 控制Γ氣Γ進變步,1:模型建立步…得到的建議值 、 '進軋罝以持續進行克勞斯製程,較佳曰 該建議值控制空氣之進氣量使該克勞斯製程持續進^ 201233430 同時,檢視相關儀表確認尾氣分析儀是否確實 a =確?、或修復尾氣分析儀期間,以該建議值控制二 =運Γ該克勞斯製程在該尾氣分析儀修復期間能 持續進灯運作,達到降低克勞斯製程異常發 加克勞斯製程穩定性的目的。 並且增 :上:述’本發明克勞斯製程之尾氣分析儀異 =以自回歸滑動平均模型預測空氣之進氣量,更利用See ® ι, i for the preferred embodiment of the debt test method for the exhaust gas analyzer of the Claus process, as shown in Figure 该, the Claus process produces sulfur from sour gas and air, and In the process of the process, the exhaust gas is analyzed by the exhaust gas analyzer, and then the feedback value is obtained according to the concentration ratio of the hydrogen sulfide and sulfur dioxide in the exhaust gas, and then the air intake amount is controlled by the feedback value to continue the Crow. The process of determining the abnormality of the exhaust gas analyzer of the Claus process includes a step of obtaining a water concentration, a step of establishing a model 12, and a step of determining 13. The water concentration obtaining step is to extract the temperature and pressure of the acid gas, and to obtain the water concentration in the acid gas after the value of the liquid.忒Model establishment step 丨2 establishes an autoregressive moving average model (ARMA) with an autoregressive order and a sliding average order, and then extracts a feedback value from a recent time point within a pre-emption period The multiple feedback values, together with the water concentration, are substituted into the autoregressive moving average model to obtain a suggested value, preferably, for example, Method of moments, most approximate estimation method 201233430 (Maxi surface likelih〇 Od estimates, ML E.) and other methods to estimate the values of the parameters in the model, and establish the autoregressive moving average model. The determining step 13 is to determine that the exhaust gas analyzer is abnormal when the difference between the recommended value and the feedback value of the most recent point exceeds a predetermined range. The method for detecting abnormality of the exhaust gas analyzer of the Claus process of the present invention, by establishing the autoregressive moving average model to predict the intake air amount of the air, and utilizing the feedback value of the most recent time point in the range of the predetermined time period The value is used as the basis for the prediction, and because the concentration of water in the acid gas affects the amount of air reacted with the acid gas, the amount of the air must be adjusted accordingly. Therefore, the error is corrected by the concentration of the water, so that the exhaust gas is judged. Is the analyzer abnormal? "β t plus precision: read Figure 2, the exhaust gas analyzer of the Claus process of the invention is abnormal _: the actual (four) test results, as shown in Figure 2, the horizontal axis is measured and the vertical axis is the air intake the amount. In the evaluation test period, the average value of the error of the recommended value of the actual phase is found and the standard deviation is the heart, and the expected effect is achieved, and the gas analyzer is normally operated and the feedback value is correct. In the case where the tail method has been confirmed first - Fig. 3 - comparison: Guifa Shi = labor: Process exhaust gas analyzer abnormality - speculation is not... The example is similar to the first preferred embodiment, the example, a trip, when determining When the exhaust gas analyzer is abnormal, the second preferred embodiment is strain treated by a temporary strain step 2. Control the enthalpy change step, 1: model build step... get the recommended value, 'into the rolling 罝 to continue the Claus process, preferably 建议 the recommended value to control the air intake of the air so that the Claus process continues Into ^ 201233430 At the same time, check the relevant instrument to confirm whether the exhaust gas analyzer is indeed a = indeed? During the repair of the exhaust gas analyzer, the recommended value is used to control the two processes. The Claus process can continue to enter the lamp during the repair of the exhaust gas analyzer, thereby reducing the stability of the Claus process and the stability of the Claus process. the goal of. And increase: above: "The tail gas analyzer of the Claus process of the present invention is different = the self-regressive moving average model is used to predict the air intake amount of the air, and the utilization is further utilized.
進# 丁&差為正而使得到的建議值更加精 準,並且經實際評估測試的結果驗證’該建議值確實達到 預期的功效,故確實能達成本發明之目的。 實達 处惟以上所述者,僅為本發明之較佳實施例而已,當不 =以此限定本發明實施之範圍’即大凡依本發明中請專利 I巳圍及發明說明内容所作之簡單的等效變化與修飾,皆仍 屬本發明專利涵蓋之範圍内。 【圖式簡單說明】 圖1是-流程圖,說明本發明克勞斯製程之尾氣分析 儀異常的偵測方法之-第-較佳實施例; 圖2疋一杈軸為測試的時間點且縱軸為空氣的進氣量 ’曲線圖5兑明圖i的該第一較佳實施例實際評估測試的 結果;及 圖3是一流程圖, 儀異常的谓測方法之一第 說明本發明克勞斯製程之尾氣分析 —較佳實施例。 201233430 【主要元件符號說明】 11 水分濃度得到步驟 13 12 模型建立步驟 2 判斷步驟 暫時應變步驟The difference between the positive and negative values is more accurate, and the result of the actual evaluation test verifies that the recommended value does achieve the intended effect, so that the object of the present invention can be achieved. The above is only the preferred embodiment of the present invention, and is not intended to limit the scope of the practice of the present invention, that is, the simplicity of the invention and the content of the invention. Equivalent variations and modifications are still within the scope of the invention. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flow chart showing a method for detecting an abnormality of an exhaust gas analyzer of the Claus process of the present invention - a preferred embodiment; FIG. 2 is a time point of testing and The vertical axis is the intake air amount of the air'. FIG. 5 is a result of the actual evaluation test of the first preferred embodiment of FIG. 5; and FIG. 3 is a flowchart, one of the methods for predicting the abnormality of the instrument. Exhaust gas analysis of the Claus process - preferred embodiment. 201233430 [Explanation of main component symbols] 11 Water concentration obtaining step 13 12 Model establishment step 2 Judgment step Temporary strain step