Abstract
There are many types of power plants producing different amounts of emissions and utilizing different amounts of water to operate. The requirement of natural resources such as water in this process is an important aspect to be considered. Maintenance of good air quality along with the emissions from power plants is important as well. US Energy Information Administration (EIA) and other related government agencies and organizations provide data sources of power plant electricity generation, greenhouse gas emission, and water usage. This project studied the data and performed data analysis. Based on the complex relationships among electricity generation, emission and water usage, this project used machine learning and deep learning techniques to produce emission and water usage data for the missing years and predict data for future years. For example, this project used the data from the years that have both generation and emission to predict the emission for the years that we only have generation data. Results from deep learning techniques outperforms the machine learning algorithms. In addition, the results were compared with a traditional approach which uses co-efficient to produce derived data for the missing years.