共同沟天然气管道泄漏报警实验研究

Experimental study on response of alarm detector to natural gas leakage in utility tunnel

采用甲烷模拟天然气在共同沟(狭长空间)中的微小泄漏,分析甲烷的泄漏量和模型的断面宽度对纵向布置的探头报警时间的影响,提出单宽泄漏量的概念.分析探头位置对报警响应时间的影响,根据实验数据的分布特征,采用抛物线函数对各组实验数据进行拟合,并对单宽泄漏量与抛物线拟合系数的变化关系,进行二次拟合,得出报警响应曲线的一般经验公式.

The experimental background of this study is a real utility tunnel sited in Dameisha, Shenzhen, China, in which a natural gas pipe will be laid. For security, the gas pipe is enclosed in a small channel with an airproof cover, to be separated from other pipelines. Gas leakage detectors are installed in the channel. The object of the experimental simulation is the finite diffusion of weak leakage of natural gas in the closed long-narrow space without air current, which is thought of as a two-dimensional gas diffusion of a steady continuous release point-source in a bounded domain. In order to study the relationship between the response time and the position of detectors, and to provide an experimental basis for the security design, the experiments were performed to simulate and test the diffusion by using methane and detectors. To be realistic, each experimental model of the gas pipe channel is horizontally laid, in which six detectors are setup on the cover of the model. Of nine models used in the experiments, four models are made up of plywood for analyzing the characteristic and main effect factors of the diffusion ; the rest are of plexiglass plate for performing quantitative analysis on experimental data. By comparing the experimental data in two plywood models with equal section width and different section area, it was found that the difference of the correlative alarm time is not significant. It could be induced that the model＇s width is an influencing factor of alarm time, but the area and height are not. The gas leakage flux is the other influencing factor. Experimental results show that the larger leakage flux of methane is, the shorter alarm time is; In contrast, the larger width of model section is, the longer alarm time is. Propsed in the paper is concept of leakage flux per unit width, defined as the ratio of leakage flux of methane to width of model section. The fitting calculations for analyzing experimental data obtained from the plexiglass plate models were carried out by two steps. First, the data were fitted with a parabola function according to their distribution characteristics. Second, the general empirical formulas of alarm response curves were obtained by analyzing the relationship between the leakage flux per unit width and the parameters of the parabola functions.