Full Session with Abstracts
Power systems in rural communities are much more vulnerable to extreme hurricane impacts due to the fact that it is the structurally weaker and prone-to-aging poles and conductors that are mostly installed in the rural distribution grids. Moreover, due to the remote and geospatial sparseness of customer distribution, the recovery efforts usually take more elongated time than otherwise it would be for recovering from urban power outbreak. In the recent 2017 Hurricane Harvey, Texas’s power infrastructure was hit significantly with about 300,000 customers who lost power during the peak of the storm. With the recovery proceeded rapidly, however, the rural areas were expected to be remained in outage for weeks. To achieve an analytical insight of rural power resilience, this paper presents a fully theoretic framework for quantifying the resilience of rural power grids due to an extreme hurricane impact. The framework starts with a probabilistic fault-tree analysis that is converted from a prescribed rural power-distribution network with a general-tree graphical topology. Various sources of uncertainties including wind field and structural (poles/conductors) aging will be included. Probabilistic distributions of loss of customers will be formulated. Considering the geospatial distance dependency and the introduction of a ‘resourcefulness’ model parameter, probabilistic recover functions and resilience measures are defined. To further quantify the degree of resilience considering the effect of resourcefulness and geospatial distances, a set of information theoretic metrics are proposed, in which any variable resilience measures (with probability distributions) are compared against the ideal resilience measured with a linear distance-effect function and a perfect resourcefulness. Numerical studies considering different rural power-grid topologies and different categories of hurricane hazards are investigated in this paper to demonstrate the validity of the proposed methodology.