Modeling and assessing variability in Use Phase Energy of online multimedia services

Recent studies of the sources and extent of variability in life cycle assessments of the climate impact of information and communication technology (ICT) have highlighted a need to further investigate the uncertainty related to the energy consumption during ICT operation. We use an improved, more accurate model to analyze the energy footprint of the audience of a major online newspaper, which allows for the development of tools to monitor the energy consumption of ICT in real time. We identify the footprint of consuming news with various types of access devices and quantify the impact of behavioral parameters on the overall footprint. Previous estimations of national and global consumption assume average values for all subcomponents. Our results indicate that previous analyses based on average figures for laptops or desktop PCs predict values that are too high for national or global digital media power consumption. Additionally, we identify which components contribute most to the total use phase energy consumption and which should be focused upon in any life cycle assessment of digital services. We find the origin data center to contribute between 4 to 46% of the total energy budget when reading news articles and between 2 and 10% when watching video content. Similarly, we find that user devices contribute between 6 and 90% and 0.7 and 77% when consuming articles or video content, respectively.
(Note: A pre-press version of this text is available at http://sympact.cs.bris.ac.uk/papers/jie-prepress.pdf. The final and authoritative version will be published by the Journal of Industrial Ecology beginning of 2013.)