Monday, 15 January 2007
Weather influences on power outages in New Hampshire: Development of a Web based decision making tool
Exhibit Hall C (Henry B. Gonzalez Convention Center)
Since the fall of 2004, the Plymouth State University (PSU) meteorology program and Public Service of New Hampshire (PSNH), a major electric utility in New Hampshire, have participated in a partnership to investigate ways in which PSNH could make better operational decisions when faced with possible weather related power outages. The primary goal of the partnership was to identify forecastable meteorological trends that occurred during major outages (100 or more areas without power) and then develop a tool to bring the results of the research into day-to-day operations. In Phase I, Michael Nahmias (undergraduate student at PSU) developed an 8-year climatology of weather events that caused major outages. The results of this research identified four main weather systems that cause major outages: winter storms; wind storms; thunderstorms; and heat waves. For each type of weather system there were several common characteristics that were identified. For example, 90% of winter storm outages occurred when the temperature was between 28 and 34 °F. In Phase II of the project (completed by June 2006), the main objective was to use the results from Phase I and integrate them with actual meteorological data to provide a web based application that would serve PSNH as a tool for identifying conditions that may cause outages ahead of time. Content for this tool would be specific to the needs of PSNH and conform mainly to the New Hampshire and New England region.
PSU and PSNH worked closely to develop the content of the web site. The web site contains five main sections: 1) links to real-time meteorological observations; 2) links to forecast weather data; 3) a decision making tool; 4) a reference section; and 5) the results of the climatological research. The real-time section contains links to weather maps and images of current conditions around NH and the New England, as well as some national and global resources. The forecast section contains maps from various model sources with forecasts for conditions out to 48 hours. In the decision making tool, real-time and forecast weather data are combined with the results of the Phase I research into an interactive tool which leads the user through a decision tree. The user is prompted to answer questions about the forecast weather situation in relation to the criteria for each weather event. At the terminal end of each branch of the decision tree, a categorical likelihood of a major outage is given (likely, possible, and unlikely) and the relevant course of actions for the PSNH personnel is displayed. The references section contains definitions, keys, and general information that may be helpful to PSNH users and assist them with understanding all of the maps, meteorological data, and vocabulary that are provided in the website. Lastly, the results section provides a short recap of the Phase I research as a reference for the user. An overview of the website will be presented at this conference.