Data network equipment energy use and savings potential in buildings

Abstract

Network connectivity has become nearly ubiquitous, and the energy use of the equipment required for this connectivity is growing. Network equipment consists of devices that primarily switch and route Internet Protocol (IP) packets from a source to a destination, and this category specifically excludes edge devices like PCs, servers and other sources and sinks of IP traffic. This paper presents the results of a study of network equipment energy use and includes case studies of networks in a campus, a medium commercial building, and a typical home. The total energy use of network equipment is the product of the stock of equipment in use, the power of each device, and their usage patterns. This information was gathered from market research reports, broadband market penetration studies, field metering, and interviews with network administrators and service providers. We estimate that network equipment in the USA used 18 TWh, or about 1% of building electricity, in 2008 and that consumption is expected to grow at roughly 6% per year to 23 TWh in 2012; world usage in 2008 was 51 TWh. This study shows that office building network switches and residential equipment are the two largest categories of energy use consuming 40% and 30% of the total respectively. We estimate potential energy savings for different scenarios using forecasts of equipment stock and energy use, and savings estimates range from 20% to 50% based on full market penetration of efficient technologies.

Trends in worldwide ICT electricity consumption from 2007 to 2012

“…We base our estimate on a study by Lanzisera et al [10], which estimates the USA and worldwide electricity consumption of data network equipment in both residential buildings and offices. Their study focuses on IP-based network equipment only, and does not include the electricity used by power or cooling infrastructure.…”

“…These three categories broaden the scope to completely cover all data center network equipment as well. To exclude telecom operator network equipment (which was already covered in Section 2.1), we consider only half of the electricity use reported in [10] for the modular & 10G switches and large routers. We do not have data to more accurately assess the share of operator network consumption in both categories; assuming the share to be half minimizes the potential deviation.…”

“…We do not have data to more accurately assess the share of operator network consumption in both categories; assuming the share to be half minimizes the potential deviation. As the sum of both categories contributes less than 10% of the total electricity use, [7] b Power Usage Effectiveness (PUE) value for 2007 / 2012, from Table 3 c Half of the value specified in [10], to avoid counting the telecom operator share the potential error will be small. We also subtract the data center network equipment electricity consumption (which can be derived from Table 3 in Section 4) to avoid overlap between the categories.…”

“…We consider the following access technologies: cable, Digital Subscriber Line (DSL), Fiber To The Home (FTTH), narrowband, and other broadband (such as satellite). Our scope does not include residential stand-alone wired switches, but they have been estimated in [10] to be only a small contributor. Similarly, we do not include power line communication devices; we estimate 2 that their consumption is around 2 TWh/y, but an in-depth study would be needed for a more accurate evaluation.…”

“…Values for 2012 are extrapolations based on data from previous years. The power consumption per user values for cable, DSL and FTTH are based on [10,19]. For the relatively small number of users accessing the Internet through other broadband technologies we assumed a power consumption comparable to that of the more common broadband technologies.…”