基于场景聚类的天然气场站泄漏风险区域研究

Leakage Hazard Area of Natural Gas Stations Based on Scenario Cluster Analysis

肖飞 杨露 王虓 赵学清 谢宗宝

XIAOFei YANGLu WANGXiao ZHAOXueqing XIEZongbao

中国石油西南油气田公司输气管理处, 四川 成都 610200 中国石油大学(华东)机电工程学院, 山东 青岛 266580 中国石油西南油气田公司, 四川 成都 610051

Gas Management Office, Southwest Oil & Gas Field Company, PetroChina, Chengdu, Sichuan 610200, China College of Mechanical and Electrical Engineering, China University of Petroleum (East China), Qingdao, Shandong 266580, China Southwest Oil & Gas Field Company, PetroChina, Chengdu, Sichuan 610051, China

泄漏风险区域划分可为天然气场站的维护管理和应急疏散提供依据，但当前相关研究多基于发生概率最大或损失最严重的单一泄漏场景。针对以上问题，提出了基于场景聚类分析的输配气场站泄漏风险区域划分方法，并以西部某天然气输配气站为例具体分析研究。通过K-means聚类分析筛选出153个代表场景，借助数值模拟得到场站内天然气泄漏扩散后气云浓度分布，最终将场景概率和场站平面图中各网格浓度加权得到泄漏风险区域。与其他风险区域划分相比，提出的方法不仅解决现有单一泄漏场景考虑因素不够全面，普适性差的问题，而且进一步降低了数据冗余度，减少大量工作量和时间。

Leakage hazard area can provide the basis for decision in maintenance management and emergency evacuation of natural gas stations. However, the relevant researches were mostly focused on the single leakage scenario with the highest probability of occurrence or the most serious loss. In view of the above problems, a method based on scenario cluster analysis for the leakage hazard area classification was developed, and a gas transmission station was selected for test. The typical 153 scenarios were selected via K-means cluster analysis, and the gas distribution for different leakage scenarios were calculated from CFD simulations. Finally, a leakage classification area was obtained with the consideration of scenario probability and its gas distribution. By comparison, the leakage hazard area in this paper solved the problems of single scenario, such as bad generality and flexibility. It also improved the efficiency of work by decreasing data redundancy.

10.11885/j.issn.1674-5086.2022.07.22.02