本文關鍵詞： LNG加氣站 HAZOP 定量分析 偏差臨界值 SDG模型 后果嚴重度分析　出處：《西南石油大學》2015年碩士論文　論文類型：學位論文

【摘要】：LNG加氣站一般建于城市邊緣,一旦發(fā)生事故將造成嚴重的后果,所以采用危險辨識技術對LNG加氣站潛在的危險進行分析,針對存在的危險提出有效的防治措施,防止發(fā)生事故是非常重要的。眾多的危險辨識方法中,危險與可操作性分析(HAZOP)具有系統(tǒng)性、全面性、結構性、細致性等優(yōu)勢,所以,本文運用HAZOP對LNG加氣站進行安全分析。HAZOP自問世以來,一直以定性的風險分析方法廣泛應用于各個領域,但HAZOP定性分析存在諸多不足：不能明確指出所分析參數(shù)偏差的具體數(shù)值,即偏差模糊性大；分析所得結果針對性差,不能使工作人員明確系統(tǒng)的薄弱環(huán)節(jié),從而抓住主要風險。故本文提出對LNG加氣站進行HAZOP定量分析。 本文提出的HAZOP定量分析方法是在定性分析的基礎上進行的,所以文章首先介紹了HAZOP定性分析方法的原理和分析步驟,并以四川某LNG加氣站為例進行分析。為了克服定性分析偏差模糊性大的缺點,本文提出界定偏差臨界值,可根據(jù)LNG加氣站工藝參數(shù)控制指標界定,對于不能由工藝指標確定的,采用HYSYS軟件動態(tài)模擬偏差發(fā)生時對系統(tǒng)的影響,通過其他參數(shù)指標界定偏差臨界值。為使HAZOP分析結果具有針對性,讓人們明確LNG加氣站的薄弱環(huán)節(jié)和主要危害,文章提出結合風險矩陣確定偏差風險等級。偏差風險等級由偏差發(fā)生可能性等級和偏差導致的后果嚴重度等級確定。由于SDG(符號有向圖)與HAZOP推理機制類似,兩種方法有著天然的聯(lián)系,故將SDG與HAZOP結合。首先根據(jù)LNG加氣站工藝流程畫出SDG模型,再以發(fā)生偏差的參數(shù)所在節(jié)點為對象進行拉偏(偏離偏差臨界值),找出所有偏差傳播通路,原因事件概率用基于專家意見的模糊處理方法求得,傳播變量概率用基于貝葉斯網(wǎng)絡的SDG概率計算規(guī)則求得,便可計算偏差發(fā)生概率,進而確定偏差發(fā)生可能性等級。鑒于LNG加氣站工藝流程的特殊性(同一設備用于不同工藝流程中),HAZOP分析的偏差不能以一籠統(tǒng)的值確定,偏差所致的后果也不同,本文提出將偏差后果分為工藝后果(影響工藝生產(chǎn))和安全后果(偏差導致安全事故發(fā)生)兩大類,當存在安全后果時,考慮安全后果,進行安全后果嚴重度評價；當偏差導致工藝后果而未引起安全事故時,進行工藝后果嚴重度評價。參考國內相關條例,并結合LNG加氣站自身規(guī)模和特點進行了事故后果嚴重度等級劃分,以確定偏差后果嚴重度等級。 最后以四川某LNG加氣站為研究對象,進行了HAZOP定量分析,辨識出LNG儲罐超壓是該加氣站最大的潛在危害,符合現(xiàn)場運行情況,驗證了本文提出的LNG加氣站HAZOP定量分析方法的實用性。
[Abstract]:The LNG filling station is generally built on the edge of the city, once an accident occurs, it will cause serious consequences, so the potential hazard of LNG filling station is analyzed by using hazard identification technology, and effective prevention measures are put forward in view of the existing danger. Prevention of accidents is very important. Among many hazard identification methods, HAZOPhas the advantages of systematic, comprehensive, structural and meticulous, so, In this paper, HAZOP is used to analyze the safety of LNG filling station. HAZOP has been widely used in various fields with qualitative risk analysis method since its inception. However, HAZOP qualitative analysis has many shortcomings: it can not clearly point out the specific value of the deviation of the parameters analyzed. That is, the ambiguity of deviation is big, the result of analysis is poor, and the staff can not make clear the weak link of the system, so the main risk can be grasped. Therefore, this paper puts forward the HAZOP quantitative analysis of LNG filling station. The HAZOP quantitative analysis method proposed in this paper is carried out on the basis of qualitative analysis, so this paper first introduces the principle and steps of HAZOP qualitative analysis method. Taking a LNG gas station in Sichuan province as an example, in order to overcome the shortcoming of the ambiguity of qualitative analysis deviation, the critical value of deviation is proposed in this paper, which can be defined according to the control index of process parameters of LNG gas filling station, but can not be determined by the process index. HYSYS software is used to dynamically simulate the influence of deviation on the system, and the critical value of deviation is defined by other parameter indexes. In order to make the results of HAZOP analysis targeted, the weak link and main harm of LNG filling station are made clear. In this paper, it is proposed to determine the deviation risk grade by combining the risk matrix. The deviation risk grade is determined by the possibility class of deviation and the consequence severity grade caused by deviation. Because the HAZOP (symbolic directed graph) is similar to the reasoning mechanism of HAZOP, There is a natural connection between the two methods, so the SDG and HAZOP are combined. Firstly, the SDG model is drawn according to the process flow of the LNG filling station. Then taking the node where the parameter of the deviation occurs as the object to pull the deviation (deviating from the critical value of deviation) and finding out all the propagation paths of the deviation, the probability of the cause event is obtained by the fuzzy processing method based on the expert opinion. The probability of propagation variable can be calculated by using the SDG probability calculation rule based on Bayesian network, and the probability of occurrence of deviation can be calculated. In view of the particularity of the process flow of the LNG filling station (the deviation of the same equipment used in different processes cannot be determined by a general value, the consequences of the deviation are also different. In this paper, the deviation consequences are divided into two categories: process consequences (affecting process production) and safety consequences (resulting in safety accidents). When there are safety consequences, safety consequences are considered and the severity of safety consequences is evaluated. When the deviation causes the process consequence but does not cause the safety accident, carries on the craft consequence severity appraisal. Referring to the domestic related regulations, and combined with the LNG filling station own scale and the characteristic, the accident consequence severity grade is divided. To determine the severity of deviation consequences. Finally, taking a LNG gas station in Sichuan province as the research object, the HAZOP quantitative analysis is carried out, and it is identified that the overpressure of the LNG tank is the biggest potential hazard of the station, which is in line with the field operation. The practicability of the HAZOP quantitative analysis method proposed in this paper for LNG filling station is verified.
【學位授予單位】：西南石油大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：TE88

【參考文獻】

相關期刊論文 前10條

1 李漾;周昌玉;張伯君;;石油化工行業(yè)可接受風險水平研究[J];安全與環(huán)境學報;2007年06期

2 付建民;趙東風;陳國明;周樂平;張斌;;石油化工裝置HAZOP分析技術概率定量化研究[J];安全與環(huán)境學報;2008年06期

3 趙林度;基于Petri網(wǎng)的自動HAZOP分析和故障診斷集成(英文)[J];Journal of Southeast University(English Edition);2003年03期

4 石玉美,顧安忠;天然氣液化流程中的汽液相平衡計算[J];低溫工程;1997年03期

5 萬志文,華賁,陸明亮;淺談危險性可操作研究[J];廣東化工;2005年01期

6 朱云斌;黃曉明;常青;;模糊故障樹分析方法在機場環(huán)境安全中的應用[J];國防科技大學學報;2009年06期

7 蘇偉;劉小娟;;LNG小型站安全設計[J];光盤技術;2009年11期

8 姜巍巍;趙文芳;李奇;李俊杰;;HAZOP風險分析在環(huán)氧乙烷罐區(qū)的應用[J];工業(yè)安全與環(huán)保;2007年02期

9 齊福強;趙云勝;何華剛;;LPG罐區(qū)火災爆炸事故危險性分析[J];工業(yè)安全與環(huán)保;2008年01期

10 邢培育;陶剛;;基于模糊信息融合的HAZOP技術應用[J];工業(yè)安全與環(huán)保;2011年12期

，

本文編號：1549642