This project was undertaken as a course project for the course Distributed Intelligent Systems in the fall '09. Here Olivier Monod and I performed detailed analysis using simulations to analyse the performance of threshold based and market based techniques for deciding between a cheap but highly variable energy source (sun, wind, etc.) or an expensive but stable source (fossil fuel, nuclear energy, etc.) while attempting to meet the demand and minimizing the cost of the energy used. This task becomes harder as the noise in the system increases and fair estimates of the demand and the supply are no longer available. Very interesting results were obtained and are described in the following documents: