No abstract

and numerous other organizations with whom the three-laboratory team has interacted during this project. The project team would also like to acknowledge the Clean Energy States Alliance (CESA) for very close collaboration in implementing its energy policy and regulations database into the Energy Zones Mapping Tool, as well as Navigant Consulting for its contributions related to demand-side resources.

An ORNL team working on the Energy Awareness and Resiliency Standardized Services (EARSS) project developed a fully automated procedure to take wind speed and location estimates provided by hurricane forecasters and provide a geospatial estimate on the impact to the electric grid in terms of outage areas and projected duration of outages. Hurricane Sandy was one of the worst US storms ever, with reported injuries and deaths, millions of people without power for several days, and billions of dollars in economic impact. Hurricane advisories were released for Sandy from October 22 through 31, 2012. The fact that the geoprocessing was automated was significant-there were 64 advisories for Sandy. Manual analysis typically takes about one hour for each advisory. During a storm event, advisories are released every two to three hours around the clock, and an analyst capable of performing the manual analysis has other tasks they would like to focus on. Initial predictions of a big impact and landfall usually occur three days in advance, so time is of the essence to prepare for utility repair. Automated processing developed at ORNL allowed this analysis to be completed and made publicly available within minutes of each new advisory being released.

The active application of photovoltaic for electricity generation could effectively transform neighborhoods and commercial districts into small, localized power plants. This application, however, relies heavily on an accurate estimation of the amount of solar radiation that is available on individual building rooftops. While many solar energy maps exist at higher spatial resolution for concentrated solar energy applications, the data from these maps are not suitable for roof-mounted photovoltaic for several reasons, including lack of data at the appropriate spatial resolution and lack of integration of building-specific characteristics into the models used to generate the maps. To address this problem, we have developed a modeling framework for estimating solar radiation potentials on individual building rooftops that is suitable for utility-scale applications as well as building-specific applications. The framework uses light detection and ranging (LIDAR) data at approximately 1-meter horizontal resolution and 0.3-meter vertical resolution as input for modeling a large number of buildings quickly. One of the strengths of this framework is the ability to parallelize its implementation. Furthermore, the framework accounts for building specific characteristics, such as roof slope, roof aspect, and shadowing effects, that are critical to roof-mounted photovoltaic systems. The resulting data has helped us to identify the so-called “solar panel sweet spots” on individual building rooftops and obtain accurate statistics of the variation in solar radiation as a function of time of year and geographical location.