We focus on large blackouts in electric distribution systems caused by extreme winds. Such events have a large cost and impact on customers. To quantify resilience to these events, we formulate large event risk and show how to calculate it from the historical outage data routinely collected by utilities' outage management systems. Risk is defined using an event cost exceedance curve. The tail of this curve and the large event risk is described by the probability of a large cost event and the slope magnitude of the tail on a log-log plot. Resilience can be improved by planned investments to upgrade system components or speed up restoration. The benefits that these investments would have had if they had been made in the past can be quantified by "rerunning history" with the effects of the investment included, and then recalculating the large event risk to find the improvement in resilience. An example using utility data shows a 12% and 22% reduction in the probability of a large cost event due to 10% wind hardening and 10% faster restoration respectively. This new data-driven approach to quantify resilience and resilience investments is realistic and much easier to apply than complicated approaches based on modeling all the phases of resilience. Moreover, an appeal to improvements to past lived experience may well be persuasive to customers and regulators in making the case for resilience investments.

翻译：本文聚焦于极端风灾引发的配电系统大规模停电事件。此类事件对用户造成巨大经济损失与严重影响。为量化系统对此类事件的韧性，我们构建了大事件风险度量框架，并阐明如何利用电力公司停电管理系统常规收集的历史停电数据进行计算。风险通过事件成本超越曲线进行定义，该曲线尾部特征及大事件风险可由大成本事件发生概率及双对数坐标图中尾部斜率幅值表征。通过规划性投资升级系统组件或加速故障恢复可提升系统韧性。为量化此类投资的潜在效益，我们采用"历史重演"方法：在历史数据中模拟投资措施的影响，重新计算大事件风险以评估韧性提升程度。基于实际电力公司数据的案例分析表明，实施10%的抗风加固改造与10%的恢复加速措施，可分别将大成本事件发生概率降低12%与22%。这种数据驱动的韧性及韧性投资量化方法具有现实可行性，且相较于需要模拟韧性全阶段的复杂建模方法更易于实施。此外，通过展示投资对历史经验的改善效果，可为向用户和监管机构论证韧性投资必要性提供有力依据。