Monte Carlo-informed reliability and availability framework for innovative inspection timelines: a case study of the PA01-BC01 heat exchanger

In nuclear facilities, predictive maintenance is essential to sustain safety, reliability, and availability, particularly for ageing components that are sensitive to downtime. Conventional strategies often rely on reliability indices or fixed timelines, which may not adequately reflect operational disturbances and uncertainty. To address this gap, this study proposes an inspection timeline based on a Monte Carlo-informed Reliability- and Availability-Centered Maintenance (RACM) framework. The framework integrates the Maintenance Priority Index (MPI) with key reliability and maintainability metrics: mean time between failures (MTBF), mean time to failure (MTTF), and mean time to repair (MTTR), to support data-informed decision making. The method is demonstrated on the PA01-BC01 heat exchanger; a 16.5 MW shell-and-tube unit in the GAS multi-purpose reactor’s secondary cooling system. Using Weibull-distributed failure data and Monte Carlo simulations, the component’s MTTF was 799.19 days, closely matching the analytical value of 800.78 days. With an MTTR of 3.84 days, the resulting MTBF was 803.02 days, confirming low failure frequency and high reliability. At a targeted reliability of 0.90, operational availability reached 0.995, underlining its critical role in reactor performance. By weighting reliability (<i>w</i>_<i>1</i> = 0.48) and availability (<i>w</i>_<i>2</i> = 0.52), the MPI-based approach produced innovative inspection timelines ({184, 276, 343} days) that balance downtime risk, labour, and cost more effectively than reliability-only or baseline timelines. This study demonstrates the integration of stochastic reliability and maintainability analyses with maintenance prioritization, providing direct impact for ageing nuclear systems and scalable applicability to other high-reliability industries.

在核设施中，预测性维护（predictive maintenance）对于维持安全性、可靠性与可用性至关重要，尤其针对那些对停机敏感的老化组件。传统维护策略通常依赖可靠性指标或固定检修周期，却难以充分反映运行扰动与不确定性因素。为填补这一研究空白，本研究提出了一种基于蒙特卡洛（Monte Carlo）驱动的可靠性与可用性为中心的维护（Reliability- and Availability-Centered Maintenance, RACM）框架的检修周期规划方法。该框架将维护优先级指数（Maintenance Priority Index, MPI）与核心可靠性及可维护性指标——平均故障间隔时间（mean time between failures, MTBF）、平均失效前时间（mean time to failure, MTTF）以及平均修复时间（mean time to repair, MTTR）——相结合，以支持基于数据的决策制定。本方法以PA01-BC01换热器为验证对象：该设备为GAS多用途反应堆二次冷却系统中的16.5兆瓦管壳式换热单元。通过采用威布尔（Weibull）分布故障数据与蒙特卡洛模拟，该组件的平均失效前时间为799.19天，与解析计算得到的800.78天结果高度吻合。结合3.84天的平均修复时间，最终得到的平均故障间隔时间为803.02天，证实该组件故障频率较低、可靠性较高。当目标可靠性设定为0.90时，系统运行可用性可达0.995，凸显了该组件对反应堆运行性能的关键作用。通过对可靠性（权重w₁=0.48）与可用性（权重w₂=0.52）进行加权，基于维护优先级指数的方法生成了创新的检修周期方案（{184, 276, 343}天），相较于仅基于可靠性的方案或基准检修周期，该方案能更有效地平衡停机风险、人力投入与经济成本。本研究验证了随机可靠性与可维护性分析同维护优先级规划的集成方法，可为老化核系统提供直接的实践支撑，同时也可推广应用至其他高可靠性行业。