This master's thesis develops a pricing method for spark spread options using a Monte Carlo method. The underlying commodities of interest, natural gas and uranium highlight the prevalence of natural gas power and nuclear power in Canada. To characterize the dynamics of electricity prices and capture specific features they have, two Levy models are proposed: a jump-diffusion model and a time-changed model. Real data are used to calibrate the models, using the daily average market prices for the last five years. We created a method to compute the price of the derivative under realistic modelling conditions using parameters found through the real data. Such models can be used to value the spark spread contracts to mitigate the risk associated the contracts.