Data centers, house servers and other electronic devices necessary to provide the information and communication technology (ICT) services are pervasive throughout our society. The rapid proliferation of these buildings has received much attention and increased scrutiny [ 1 , 2 ], given that their energy demand has grown to about 1.3% of global electricity use [ 3 , 4 ]. However, great strides have been made recently to reduce the energy associated with providing ICT services by better utilizing data center servers and consolidating those servers in larger, more energy efficient facilities [ 5 ].

Furthermore, ICT services may also yield net reductions in societal energy use and environmental impacts as these services augment or replace traditionally non-ICT activities [ 6 , 7 ], such as utilizing teleconferencing to reduce business travel [ 8 ]. One strategy available with ICT is the dematerialization associated with replacing physical goods with equivalent services provided through the internet [ 9 – 11 ].

Streaming music and video entertainment are examples of this type of dematerialization that have recently experienced rapid growth and have received attention as possibly less energy intensive alternatives to the manufacturing and transportation of physical goods, such as digital video discs (DVDs) and compact discs (CDs) [ 12 – 14 ]. This study aims to better understand the net energy and emissions implications of the growing shift from DVD use in the United States to video streaming services. A life-cycle assessment (LCA) approach is utilized to estimate the energy use and greenhouse-gas emissions of present day (2011) streaming video systems compared to present day physical DVD systems. The methods and results presented in this study illuminate the streaming video system components that most affect life-cycle energy use and emissions, which can help focus future research towards reducing the impact of these components as the widespread shift to streaming video continues.

This study applies the Cloud Energy and Emission Research (CLEER) model [5] to estimate the primary energy use and carbon dioxide equivalent (CO2(e)) emissions associated with replacing in-home DVD viewing with internet video streaming services in the United States. The CLEER model provides a systems-wide perspective of the energy use and CO2(e) emissions across different components of ICT services; from data centers to end-user devices. The model also includes non-IT modules, such as product manufacturing, transportation, and building energy use, to allow for the comparison of services that influence different economic sectors [15].

Input values characterizing present day US home video viewing of both DVDs and internet streaming content were entered into the CLEER model. For many of these inputs, ranges of values were applied to provide a high, low, and base-case input value, and to serve as a sensitivity analysis range. The base-case input values are derived from literature estimates and are intended to represent a typical or median value across a possible range of values within the US. As such, the base-case represents the authors' best estimates of US average energy use and CO2(e) emissions per viewing hour.

When scaled up by total viewing hours, the base-case provides a best estimate of total national energy use and CO2(e) emissions attributable to each viewing method. Given that US average values for each modeling parameter are best estimates with appreciable uncertainties, high and low cases for each modeling parameter were also established. The high and low case values are meant to provide a plausible range for the US average value chosen for each parameter in the base case.

As such, the high and low cases should not be interpreted as extreme bounds on the technically-possible values for each parameter; rather, they should be interpreted as plausible uncertainty ranges for the US average point values chosen for each parameter in the...