海底泥浆举升钻井(Subsea Mud Lift Drilling，SMD)系统属于双梯度钻井系统。在钻井过程中遇到天然气水合物层(Natural Gas Hydrate，NGH)时，NGH的分解和二次生成造成了重大的风险。为了预测风险事故，对环境因素和设备因素进行了分析。首先，在建立蝴蝶结模型后，利用映射算法将其转化为贝叶斯网络(Bayesian Networks，BN);其次，考虑不确定性影响，在BN中进行不确定性建模;然后，考虑设备因素的动态不确定性，通过转移概率矩阵将建立的BN转化为不确定性动态贝叶斯网络(Uncertain Dynamic Bayesian Networks，UDBN)。此外，还利用模糊理论和专家判断来量化设备故障的先验概率。结果表明，钻遇NGH层时，钻井风险概率比正常工况下有所增加。由于有天然气水合物的影响，发生风险的概率也变得具有不确定性。
Dynamic risk analysis of the subsea mudlift drilling system encounters natural gas hydrate
The Subsea Mud Lift Drilling(SMD)system is a dual gradient drilling system.The breakdown and secondary production of NGH pose a significant risk when Natural Gas Hydrate(NGH)formations are encountered during subsea drilling.To predict risky incidents,risk analysis methods based on dynamic Bayesian networks and bow-tie models are proposed based on traditional methods such as fault trees and static Bayesian networks.The study also analyses environmental and equipment factors.First,the bowtie model is built and transformed into a Bayesian Network(BN)using FTA's mapping algorithm.Secondly,considering the effects of uncertainty in the probabilistic parameters and uncertainty in the logical relationships between the nodes,a new node U is introduced and uncertainty is modeled in the BN using the Leaky Noisy-OR model.The established BN is then transformed into an Uncertain Dynamic Bayesian Network(UDBN)by transferring the probability matrix considering the dynamic uncertainty of the device factors.In addition,a priori probabilities of failure rates and environmental factors for selected equipment were determined by reference to historical literature and databases.Fuzzy theory and expert judgment are also used to quantify the a priori probability of equipment failure.The risk analysis showed that in the absence of maintenance,the probability of various risks tended to increase over time,and the probability of well leakage reached 0.003 135 on the 20th day.the risk impact of a subsea pump failure is more significant.The results show an increase in the probability of drilling risk when encountering the NGH formation under the three working conditions compared to normal working conditions.Meanwhile,because of the second generation of NGH,the SMD system is also more susceptible to well leaks,stuck drilling,and blow-out risks than under normal conditions.The probability of occurrence of risk also becomes uncertain due to the influence of gas hydrates.
safety engineeringuncertaintyDynamic Bayesian Network(DBN)Subsea Mud Lift Drilling(SMD)drilling riskNatural Gas Hydrate(NGH)
万方数据
维普
