The decomposition of electricity spot prices: a study of the Alberta and Pennsylvania, New Jersey, and Maryland power markets

Date
2007
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This thesis is a study of the stochastic characteristics exhibited in the Alberta and Pennsylvania, New Jersey, and Maryland (P JM) power markets. Since electricity spot prices are notoriously difficult to model, we have developed a method which breaks the price into a stochastic or noise component and a deterministic component. This simplifies the complexity of modeling and allows for a richer study of each component. We present a robust method using the fast Fourier transform (FFT) for removing deterministic cycles from the spot price data. This allows for the study of the actual stochastic component within the data. Three models are then proposed for modeling the stochastic component of the de-cycled data. One of these models is in the form of a stochastic differential equation (SDE) which includes mean­reverting behaviour. The second is again mean-reverting but it also includes jumps. For both of these SDEs we were able to formulate a method to obtain a solution to the characteristic function and thus can obtain the density function, which is particularly helpful when working with non-standard models. The third model is a non-differential equation that uses random variables to model the spot price of electricity. Each of these models are then tested using real hourly spot market data from the Alberta and PJM markets for January 1, 2002 to December 31, 2005.
Description
Bibliography: p. 74-77
Keywords
Citation
Lyle, M. R. (2007). The decomposition of electricity spot prices: a study of the Alberta and Pennsylvania, New Jersey, and Maryland power markets (Master's thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/1400
Collections