© 2022 The Author. ORCID 0000-0001-8208-3644 Determining desktop computer energy consumption and GHG emissions reduction delivered by power management peripherals Justin Sutton-Parker* University of Warwick, Computer and Urban Science Department, Coventry, CV4 7AL, United Kingdom Abstract The objective of this research is to determine the reduction in desktop computer electricity consumption and concomitant use phase greenhouse gas emissions enabled by a peripheral keyboard device capable of incremental power management. The value of the research is to examine if such a device can support information sustainability strategies designed to address the 1% of global greenhouse gas emissions already generated by end user computing. In doing so, adoption at scale would support the United Nations Environment Programme initiative to combine existing technology, such as computing, with innovation to reduce societal emissions. Identifying that desktop computers require a peripheral keyboard as standard, remain significant within commercial environments and consume between 3-5 times more electricity than notebooks, the keyboard is tested in such a context. Using analytics, asset management software and survey techniques, 417,880 business computers are profiled to determine substantiated parameters used to calculate the impact of the peripheral device. These include install base device types, age categorisation, number of attached displays and existing power management settings. To determine accurate electricity consumption measurements before and after the keyboard is introduced, a field experiment is also conducted. Specifically, adhering to internationally recognised test conditions, legacy (>6 years old), mid-point (4-5 years old) and new (1-3 years old) desktop computers are measured using watt-metres. The results are applied accordingly to the profile parameters and results relating to annual kilo-watt hour values for standard operation and the keyboard’s three energy management settings. In order to highlight the positive environmental impact, the energy values are subsequently converted to scope 2 use phase carbon dioxide equivalent greenhouse gas emissions values for three major geographical regions. The findings determine that the reduction of energy consumption and concomitant emissions ranges from between 40-43% on average per year of use. In a global context, universal diffusion of such a device is calculated to avoid greenhouse gas emissions that would otherwise require the sequestration capacity of a mature forest the size of the Netherlands. © 2022 The Author. ORCID 0000-0001-8208-3644 Keywords: Computer carbon footprint; computer energy efficiency; human-computer interaction; computer greenhouse gas emissions. * Corresponding author. Tel.: +44-(0)7976-818-530. E-mail address: Justin.Sutton-Parker@warwick.ac.uk 1. Introduction International standards [1, 2, 3] determine that end user computers will experience four use-phase modes including off, sleep, idle (long and short) and the active state. The power draw (watts) required by the computer for each modes differs as the device experiences lesser or greater workloads. As an example, a notebook computer may require 0.4W in off, 0.5W in sleep, 1.9-4W in idle and between 6.3W-11.3W in active depending upon activities being undertaken [4, 5]. The percentage of time (hours) per year spent in each state determines the mode weighting used to generate a use profile [6]. Multiplying this outcome by the relevant power draw for each mode determines the annual electricity consumption measured in kilo-watt hours (kWh). The calculation is represented as follows:   = 8760 1000  (   +    +  ூ஽௅ா   ூ஽௅ா +  ூ஽௅ா   ூ஽௅ா +   )