The Cost of Weather Forecast Error in Electricity Prices in Southeastern Pennsylvania
Open Access
- Author:
- Decuollo, Alecia Grace
- Area of Honors:
- Interdisciplinary in Economics and Energy, Business and Finance
- Degree:
- Bachelor of Science
- Document Type:
- Thesis
- Thesis Supervisors:
- George Spencer Young, Thesis Supervisor
Andrew Nathan Kleit, Thesis Supervisor
Jonathan P Mathews, Thesis Honors Advisor
James R. Tybout, Thesis Honors Advisor - Keywords:
- Electricity
Forecast/Observation Weather
Energy - Abstract:
- With the structure of competitive electricity markets, there are several incentives to power producers and consumers to accurately forecast the price of electricity for a given area. Through the use of regional forecast weather data and historical load information, one could potentially predict the next-day load schedule, with the ultimate goal of predicting the price of electricity per kilowatt hour. However, errors in forecast temperature are a major cause of short-term electricity load forecast error, having significant financial consequences. The National Research Council estimates that 25% of the U.S. gross domestic product is weather- and climate- sensitive. This paper examines the effect of the margin of error between forecast and observation weather has on electricity prices and load schedule. Using locational marginal pricing along with historical load data for the PECO region along with forecast and observation weather data, this study conducted on an hourly basis from July 10th, 2010 to April 30th, 2011, will test the effectiveness of several weather attributes in accurately predict electricity prices. In my initial data review, I found significant price fluctuations occurred at extreme temperatures of less than 40°F and greater than 70°F. Additionally, I found that at the temperature deviates from my estimated temperature threshold of 55°F threshold, the margin of forecast error becomes more significant in electricity price.