The distribution system of an electric utility, particularly with an overhead infrastructure, can be highly sensitive to local weather conditions. Power outages caused by severe weather events can have major impacts on such operations. Hence, the ability to predict specific events or combination of weather conditions that can disrupt that distribution network with sufficient spatial and temporal precision, and lead time has the potential to enable proactive allocation and deployment of resources (people and equipment) to minimize time for restoration. Near-real-time assessment of observations of current weather conditions may have the appropriate geographic locality, but by its very nature is only directly suitable for reactive response.

We have continued our effort in the New York City metropolitan area working with the emergency management group at a major utility company in applying mesoscale numerical weather prediction (NWP) at the IBM Thomas J. Watson Research Center as part of a project dubbed “Deep Thunder”. We have deployed a coupled modelling system for outage prediction and evaluated it operationally since April 2009. The NWP component is derived from a configuration of the WRF-ARW (version 3.1.1) community model. It operates in a nested configuration, with the highest resolution at two km, utilizing 42 vertical levels. The configuration also includes parameterization and selection of physics options appropriate for the range of geography within the domain from highly urbanized to rural. This includes WSM-6 microphysics (explicit ice, snow and graupel), Yonsei University non-local-K scheme with explicit entrainment layer and parabolic K profile in the unstable mixed layer for the planetary boundary layer, NOAH land-surface modelling with soil temperature and moisture in four layers, fractional snow cover and frozen soil physics, Grell-Devenyi ensemble cumulus parameterization, and the 3-category urban canopy model with surface effects for roofs, walls, and streets. It produces 84-hour forecasts, which are updated every twelve hours.

Such an NWP-based weather forecast is only a prerequisite to the optimization of weather-sensitive business operations. Based upon analysis of historical damage events and anecdotal evidence in the region, it is clear, for example, that storm-driven disruptions of the overhead electric distribution network (e.g., poles and wires) are caused by physical interaction of the atmosphere directly with that infrastructure or indirectly via nearby trees. However, reliable modelling with sufficient throughput for operational utilization of such interaction is neither tractable from a computational perspective nor verifiable from observations. Given that and the relative uncertainty in the processes and data, we approached the outage prediction from a stochastic perspective. Historical observations from a local, relatively dense mesonet (WeatherBug) operated by Earth Networks were analyzed along with data from the utility concerning the characteristics of outage-related infrastructure damage from weather events, as well as the information about the distribution network and local environmental conditions. Data from this mesonet are also utilized to validate and improve the results of the WRF configuration as well as model tuning.

Based upon the analysis of the historical data, one of the key linkages between outage and severe storms is the magnitude of wind gusts. However, as is typical for NWP codes, WRF-ARW does not produce a direct representation of gusts. Therefore, a post-processing model was developed using a Bayesian hierarchical approach. A Quasi-Poisson model was developed to statistically upscale the results of each WRF execution to irregularly shaped substation areas within the utility's service territory. This has enabled the generation of forecasts of the number of jobs to be dispatched to repair storm-induced outages in each area for each of the three full days covered by each WRF execution. This approach also incorporates uncertainties in both the weather and outage data. Therefore, the model provides probabilities of outage restoration jobs per substation, which is associated with visualizations that explicitly depict the uncertainty. It links the physical model outputs with real observations to create a calibrated, statistical representation of gusts. This model post-processes the sustained wind forecast on the topologically regular output generated by WRF based upon the gust observations from the mesonet over time. It is used as one of the inputs to the storm impact prediction for each substation area.

A validation methodology was developed for the weather forecasts component focused on the utility's needs. It was built with the Model Evaluation Tools (MET) community verification package. Temperature, wind, dew point and precipitation data from the aforementioned mesonet were compared to the forecast data as well as the data from the NOAA/NCEP North American Model. In addition, verification is being done for both the weather and outage models with respect to specific weather events that affect utility operations.

We will discuss the on-going work, the overall approach to the problem, some specifics of the solution, and lessons that were learned through the development and deployment. In addition to the evaluation, we will present how the content is being used, including the deployment of customized visualizations. We will also discuss the overall effectiveness of our particular approach for these and related applications, issues such as calibration of data and quantifying uncertainty as well as recommendations for future work.