【作者】 张杰；

【导师】 宫敬；

【作者基本信息】 中国石油大学(北京) ， 油气储运工程， 2020， 硕士

【摘要】 管道是陆上油气资源最有效的输送方式。随着服役时间的增长,管道安全事故时有发生,对人民的生命和财产安全、自然环境和经济造成严重威胁。腐蚀是影响油气管道结构完整性和安全运行的主要因素之一。以可靠性为中心的预防性维修决策是目前国际上广泛认可的腐蚀管道管理模式,对保障管道全生命周期安全运行具有重要意义。本文从数据分析的角度,通过建立腐蚀预测模型,评价了腐蚀管道的结构可靠性,并基于可靠性评价结果,提出了腐蚀管道内检测周期和维修策略的优化方法。首先,本文对内检测获得的腐蚀缺陷数据进行了校正,并基于校正数据建立了腐蚀增长模型,预测腐蚀缺陷深度随时间的增长。首先,采用线性方程描述腐蚀缺陷内检测深度与现场测量深度的关系,经贝叶斯推断获得线性方程参数的概率特征,量化测量误差。然后采用幂律方程描述腐蚀缺陷深度随时间的变化,建立多参数层次贝叶斯模型,使用马尔可夫链蒙特卡罗(MCMC)模拟技术得到了幂律方程参数的概率特征。最后,通过对比某天然气管道上腐蚀缺陷的预测深度和实测深度,验证了模型的有效性。其次,针对包含多个腐蚀缺陷的管道系统,建立了基于蒙特卡罗仿真的可靠性评价模型。该模型考虑了小孔泄漏和爆裂两种失效模式,引入了失效数据和相关性对可靠性评价的影响,从而提高了可靠性评价结果的精度,且可靠性评价结果可基于失效数据进行更新。最后,提出了腐蚀管道的内检测周期和维修策略优化方法。采用蒙特卡罗仿真技术计算出管道的可靠性和总成本率,以失效概率阈值为决策变量优化了腐蚀管道的内检测周期。考虑腐蚀管道发生小孔泄漏和爆裂的年失效概率以及维修成本三个目标函数,采用遗传算法对腐蚀管道的维修策略进行了多目标优化。更多还原

【Abstract】 Pipelines are the most effective method of transporting onshore oil and gas resources.With the increase of service time,pipeline safety accidents occur from time to time,posing serious threats to people’s lives and property safety,the natural environment and the economy.Corrosion is one of the main factors affecting the structural integrity and safe operation of oil and gas pipelines.Reliability-centered preventive maintenance strategy is a widely accepted internationally accepted corrosion pipeline management model,and is of great significance for ensuring the safe operation of the pipeline throughout its life cycle.From the perspective of data analysis,this paper evaluates the structural reliability of corroded pipelines by establishing a corrosion prediction model.Based on the results of reliability evaluation,an optimization method for the inspection cycle and maintenance strategy in corroded pipelines is proposed.First,the corrosion defect data obtained by internal inspection is corrected in this paper,and a corrosion growth model is established based on the corrected data to predict the growth of corrosion defect depth over time.The linear equation is used to describe the relationship between the depth of detection within the corrosion defect and the measured depth in the field.The probability characteristics of the parameters of the linear equation are obtained by Bayesian inference,and the measurement error is quantified.Then the power-law equation is used to describe the change of corrosion defect depth with time.A multi-parameter hierarchical Bayesian model is established.The Markov chain Monte Carlo(MCMC)simulation technique is used to obtain the probability characteristics of the power-law equation parameters.The validity of the model is verified by comparing the predicted depth and the measured depth of a corrosion defect on a natural gas pipeline.For the pipeline system containing multiple corrosion defects,a reliability evaluation method based on Monte Carlo simulation is proposed.This method takes into account the two failure modes of small leak and burst,introduces the impact of failure data and correlation on reliability evaluation,improves the accuracy of reliability evaluation results,and reliability evaluation results can be updated based on failure data.Finally,the internal inspection cycle and maintenance strategy optimization methods of corroded pipelines are proposed.Monte Carlo simulation technology was used to calculate the reliability and total cost rate of the pipeline,and the internal detection period of the corroded pipeline was optimized with the failure probability threshold as a decision variable.Considering three objective functions,namely the annual condition failure probability of small leak and burst in the corroded pipeline,and the maintenance cost,the genetic algorithm was used to optimize the maintenance strategy of the corroded pipeline.更多还原