Local Chapter Activities
Presidents of local USAEE Chapters are encouraged to share information about your local events and activities by submitting write-ups with photographs of member participation. Please contact the Editor for submission guidelines.
Lehigh University Student Chapter (Bethlehem, PA)
Kevin Forbes' Presentation on the Accuracy of Wind and Solar Energy Forecasts
(L to R): Lehigh University Student Chapter Officers: Guilherme Junqueira Perticarari, Vice President; Fernando Fontes, Treasurer; Jeff Kasle, President; and Jade Van Streepen,Secretary;with Prof. Kevin Forbes (USAEE Distinguished Lecturer).
On November 5, 2015, USAEE Distinguished Lecturer Kevin Forbes discussed the accuracy of wind and solar energy forecasts and the prospects for improvement at Lehigh University in Bethlehem, Pennsylvania. Forbes has been an active member of USAEE for over 20 years and is serving as the Vice President of Membership. He is also currently an associate professor of economics at The Catholic University of America. The Lehigh chapter of USAEE is grateful to the national organization for making this event possible through the Distinguished Lecturer Series.
In his presentation, Forbes discussed why forecasting is important, how accurate it really is, and what can be done to improve the accuracy. Among his key points:
Energy forecasting is important because the stability of the power grid is enhanced with more accurate forecasting and blackouts have high societal costs. There is also a great price difference between real time price of energy and day-ahead price of energy so accurate energy forecasting is important for cost forecasting.
The literature on forecast accuracy has agreed that wind energy forecasting is fairly accurate. Most studies that have been conducted have used a capacity weighted root mean squared error, however, rather than an energy weighted root mean squared error. When using the capacity weighted term, which is a metric endorsed by the NREL, most of the previous studies have found an error less than 10 percent. Forbes and his coauthors found errors over 20 percent studies using the energy-weighted root mean squared error. Their consensus is that onshore wind energy is moderately predictable, but not as predictable as some of the literature might suggest.
Forbes also used a mean squared error skill score, which compares the accuracy of the forecast to the persistence forecast, to assess the accuracy of wind, solar, and load forecasting. In general, he found that day-ahead load forecasts are much more accurate than day-ahead wind and solar forecasts, with positive mean squared error skill scores for load in most cases and very negative mean squared error skill scores for wind and solar in most cases. One possible reason for these large errors in wind and solar forecasting, as discussed by Forbes, is that meteorologists have historically focused on temperature rather than cloud cover and wind speeds. This is the case because cloud cover and wind speeds are much more volatile and harder to predict, but they are also large factors to be considered in wind and solar energy forecasting.
Forbes compared the energy weighted root mean squared errors of forecasting different types of conventional energy generation in Great Britain with the root mean squared error of forecasting wind generation in great Britain to assess the differences. It was found that the forecasting error for coal is 2.5 percent, combined cycle gas turbines is 5.6 percent, nuclear energy is 7.4 percent, and the one hour ahead forecasting error for wind energy is 18 percent. It should be noted that the error for wind energy forecasting includes balancing actions.
Because of these forecasting errors for wind and solar energy and the intermittency that stems from inaccurate forecasting, society is not getting the full benefit of green energy. In order to improve these predictions in the long run, major advances in meteorological research needs to be done in order to improve our understanding of the heat trapping properties of greenhouse gases which affect wind speeds. There is room for improvement in the short run, however. Forbes and his team have developed a model that exploits the systematic nature of existing day-ahead forecast errors for wind and solar. Using this model, Forbes was able to make much more accurate wind and solar forecasts with positive mean squared error skill scores and energy weighted root mean squared errors less than those not using the model.
About USAEE-Lehigh University Student Chapter: Our chapter is based out of Bethlehem, Pennsylvania, home to Lehigh University. Our goals are to introduce students to energy economics research, provide networking for the public and private sector, and provide resources for research and career opportunities. There has been substantial interest and attendance in on-campus events among students, despite our school’s relatively small size. Student interests include renewable energy, advanced energy technology, microgrids and smart-grids, oil and gas and energy policy assessment. For example, more than 40 students attended and actively participated in our most recent event with Kevin Forbes.
Our chapter hopes to expand beyond Lehigh and serve as a forum for students in the Lehigh Valley and around Philadelphia, who are interested in energy, sustainability or economics. In order to work towards this goal, our chapter plans to host the Energy Outlook, a marquee event for students and professionals. This event will likely take place in Spring 2016, and we are currently looking for presenters. Please feel free to contact us if you may be interested in presenting. Also, we would like to show appreciation to Alberto Lamadrid, our faculty advisor and Professor of Economics, and David DeAngelo, an experienced industry professional and long-time USAEE supporter, who have been critical to this newly-founded chapter’s success.
--- Contributor: Jeff Kasle, President, USAEE-Lehigh University Student Chapter