An efficient numerical algorithm, which employed the finite volume , finite difference , Monte Carlo and proper orthogonal decomposition ( POD) method, was proposed to simulate the stochastic fluctuation of oil temperatures. The effects of outlet oil temperature, flow rate, pressure, buried depth, soil temperature. soil thermal conductivity, oil viscosity and oil density on the stochastic fluctuation of oil temperatures were investigated through Sobol global sensitivity induces. The results show that the deviations of the mean oil temperatures between numerical simulations and field measurements were lower than 0. 1 ℃ , and the standard deviations vary from 0. 006 ℃ to 0.023 ℃ . When considered the following 4 factors: outlet oil temperature, throughput, soil temperature and oil viscosity, case studies on the China West crude pipeline show that the sum of first order global sensitivity achieved as 77. 44％ on Sipu station and 80. 86％ on Hexi station. Therefore, it is acceptable to consider these 4 factors when simulating the fluctuation of inlet oil temperatures.%综合采用有限容积法、有限差分法、Monte Carlo算法和POD算法建立埋地热油管道沿线油温的随机数值模拟算法,使用Sobol全局敏感性指标进行敏感性分析,综合评价出站油温、流量、压力、埋深、埋深处自然地温、土壤导热系数、油品黏度和密度的随机波动对管道沿线油温波动的影响.计算结果表明:进站油温模拟结果与现场油温均值偏差在0.1℃以内,标准差的偏差为0.006～0.023℃;出站油温、流量、埋深处地温和油品黏度4个参数的不确定性对四堡进站油温波动的敏感性指标之和为77.44%,河西站的该指标为80.86%,进站温度的随机数值模拟中主要考虑这4个参数的不确定性即可.
展开▼
