Data Centres and Data Transmission Networks – Analysis - IEA

Data centres and data transmission networks are responsible for nearly 1% of energy-related GHG emissions

Digital technologies have direct and indirect effects on energy use and emissions, and hold enormous potential to help (or hinder) global clean energy transitions, including through the digitalisation of the energy sector.

The data centres and data transmission networks1 that underpin digitalisation accounted for around 300 Mt CO2-eq in 2020 (including embodied emissions), equivalent to 0.9% of energy-related GHG emissions (or 0.6% of total GHG emissions). Since 2010, emissions have grown only modestly despite rapidly growing demand for digital services, thanks to energy efficiency improvements, renewable energy purchases by information and communications technology (ICT) companies and broader decarbonisation of electricity grids in many regions. However, to get on track with the Net Zero Scenario, emissions must halve by 2030.

Energy

Strong efficiency improvements have helped to limit growth in energy demand from data centres globally

Global data centre electricity use in 2021 was 220-320 TWh2, or around 0.9-1.3% of global final electricity demand. This excludes energy used for cryptocurrency mining, which was 100-140 TWh in 2021.3

Since 2010, data centre energy use (excluding crypto) has grown only moderately despite the strong growth in demand for data centre services, thanks in part to efficiency improvements in IT hardware and cooling and a shift away from small, inefficient enterprise data centres towards more efficient cloud and hyperscale data centres. However, the rapid growth in workloads handled by large data centres has resulted in rising energy use in this segment over the past several years (increasing by 10-30% per year). Overall data centre energy use (excluding crypto) appears likely to continue growing moderately over the next few years, but longer-term trends are highly uncertain.

Although data centre electricity consumption globally has grown only moderately, some smaller countries with expanding data centre markets are seeing rapid growth. For example, data centre electricity use in Ireland has more than tripled since 2015, accounting for 14% of total electricity consumption in 2021. In Denmark, data centre energy use is projected to triple by 2025 to account for around 7% of the country’s electricity use.

Global trends in digital and energy indicators, 2015-2021

Sources: Internet users [ITU (2022)]; internet traffic [IEA analysis based on Cisco (2015); TeleGeography (2022); Cisco (2019), Cisco Visual Networking Index]; data centre workloads [Cisco (2018), Cisco Global Cloud Index]; data centre energy use [IEA analysis based on Malmodin & Lundén (2018); ITU (2020); Masanet et al. (2020); Malmodin (2020); Hintemann & Hinterholzer (2022)]; cryptocurrency mining energy use [IEA analysis based on Cambridge Centre for Alternative Finance (2022); Gallersdörfer, Klaaßen and Stoll (2020); McDonald (2022)]; data transmission network energy use [Malmodin & Lundén (2018); Malmodin (2020); ITU (2020); Coroama (2021); GSMA (2022)].

Energy efficiency of data transmission networks has also improved rapidly

Globally, data transmission networks consumed 260-340 TWh in 20214, or 1.1-1.4% of global electricity use. The energy efficiency of data transmission has improved rapidly over the past decade: fixed-line network energy intensity has halved every two years in developed countries, and mobile-access network energy efficiency has improved by 10-30% annually in recent years.

Internet traffic globally was up 23% in 2021, lower than the 40-50% pandemic-driven surge in 2020. GSMA members reported that their network data traffic increased by 31% in 2021 while total electricity use by operators rose by 5%. Data from major European telecom network operators analysed by Lundén et al. (2022) mirror these global efficiency trends. Electricity consumption by reporting companies – representing about 36% of European subscriptions and 8% of global subscriptions – increased by only 1% between 2015 and 2018, while data traffic tripled.

Activity

Emerging services and technologies such as streaming, cloud gaming, blockchain, machine learning and virtual reality are poised to boost demand for data services

Strong growth in demand for data network services is expected to continue, driven primarily by data-intensive activities such as video streaming, cloud gaming and augmented and virtual reality applications. However, these data-intensive services may only have limited impacts on energy use in the near term since energy use does not increase proportionally with traffic volumes. In addition, the average energy consumption of video streaming is fairly low compared with other everyday activities, with end-user devices such as televisions consuming the majority. But if streaming and other data-intensive services add to peak internet traffic, the build-out of additional infrastructure to accommodate higher anticipated peak capacity could raise overall network energy use in the long run.

Mobile data traffic is also projected to continue growing quickly, quadrupling by 2027. 5G’s share of mobile data traffic is projected to rise to 60% in 2027, up from 10% in 2021. Although 5G networks are expected to be more energy efficient than 4G networks, the overall energy and emissions impacts of 5G are still uncertain.

Demand for data centre services is also poised to rise, driven in part by emerging digital technologies such as blockchain (particularly proof-of-work) and machine learning. For example, Bitcoin – the most prominent example of proof-of-work blockchain and most valuable cryptocurrency by market capitalisation – consumed an estimated 105 TWh in 20215 20-times more than it used in 2016. Ethereum, second behind Bitcoin in terms of market capitalisation and energy use, consumed around 17 TWh in 2021. In September 2022, Ethereum transitioned from a proof-of-work consensus mechanism to proof-of-stake, which is expected to slash energy use by 99.95%. As blockchain applications become more widespread, understanding and managing their energy use implications may become increasingly important for energy analysts and policy makers.

Machine learning (ML) is another area of demand growth, with potentially significant implications for data centre energy use in upcoming years. While the amount of computing power needed to train the largest ML models is growing rapidly, it is unclear how quickly overall ML-related energy use in data centres is increasing. At Facebook, computing demand for ML training (increasing by 150% per year) and inference (increasing by 105% per year) have outpaced overall data centre energy use (up 40% per year) in recent years. Google reports that ML accounted for only 10-15% of their total energy use, despite representing 70-80% of overall computing demand.

The nature of data centre demand appears likely to evolve over the coming decade. 5G, the Internet of things and the metaverse are likely to increase demand for low-latency computing, increasing demand for edge data centres. User devices such as smartphones – increasingly equipped with ML accelerators – are set to increase the use of ML with uncertain effects on overall energy demand.

Renewable energy

ICT companies are major purchasers of renewable energy, accounting for half of all corporate renewable PPAs

ICT companies invest considerable sums in renewable energy projects to protect themselves from power price volatility, reduce their environmental impact and improve their brand reputation. Hyperscale data centre operators in particular lead in corporate renewable energy procurement, mainly through power purchase agreements (PPAs). In fact, Amazon, Microsoft, Meta and Google are the four largest purchasers of corporate renewable energy PPAs, having contracted over 38 GW to date (including 15 GW in 2021).

Apple (2.8 TWh), Google (18.3 TWh) and Meta (9.4 TWh) purchased or generated enough renewable electricity to match 100% of their operational electricity consumption in 2021 (primarily in data centres). Amazon consumed 30.9 TWh (85% renewable) across their operations in 2021, with a goal of achieving 100% renewables by 2025.

However, matching 100% of annual demand with renewable energy purchases or certificates does not mean that data centres are actually powered exclusively by renewable sources. The variability of wind and solar sources may not match a data centre’s demand profile, and the renewable energy may be purchased from projects in a different grid or region from where demand is located. Renewable energy certificates in particular are unlikely to lead to additional renewable energy production, resulting in inflated estimates of real-world emissions mitigation.

Google and Microsoft have announced 2030 targets to source and match zero-carbon electricity on a 24/7 basis within each grid where demand is located. A growing number of organisations are working towards 24/7 carbon-free energy.

Although a few network operators have also achieved 100% renewables (including BT, TIM and T-Mobile), data transmission network operators are generally lagging behind data centre operators in renewable energy purchase and use. Compared with data centres, which are typically large, centralised and more flexible in location, telecommunication network operators have many sites (with limited flexibility for site selection). As a result, accessing renewable energy is noted as a challenge in many markets, particularly in emerging and developing economies with less well developed energy markets.

Policy

Few policies have been implemented to address the energy and emissions footprint of data centres and networks

While broader electricity decarbonisation policies are playing a key role in reducing operational emissions from data centres and networks, there are only a few existing policies and regulations primarily focused on reducing their energy consumption or emissions footprint. In data centres, these include:

• Regulatory and voluntary schemes to improve energy efficiency at the component level (e.g. servers, data storage, HVAC) such as ENERGY STAR and EU Ecodesign Regulations for servers and data storage products.
• Buildings-based data centre energy efficiency guidance, standards, ratings, certifications and labelling schemes such as the EU Code of Conduct on Data Centre Energy Efficiency, CLC/TS 50600-5-1, BREEAM SD 5068 (United Kingdom) and IGBC Green Data Center Rating System (India).

Some recent policy developments are aimed at improving data collection and transparency:

• The US Energy Act of 2020 calls for an updated study on data centre energy use (following a 2016 report), an open data initiative on energy use for federally owned and operated data centres, and the development of new efficiency metrics.
• A new law passed by the French Senate in November 2021 requires French telecom operators to disclose key environmental indicators to their customers.
• The European Commission adopted amendments to energy statistics regulations in February 2022, which include new data reporting requirements for data centres. Some countries have already started to publish country-level data centre energy consumption estimates in recent years, notably Ireland and the Netherlands.

The European Commission’s proposed Energy Efficiency Directive in its Fit for 55 legislative package includes sustainability reporting requirements for data centre operators. It would also require data centres exceeding 100 kW capacity (threshold to be finalised) to conduct a cost–benefit assessment of waste heat utilisation. In 2021 the Norwegian government proposed requirements for data centres with capacity above 2 MW to explore connections with district heating suppliers.

In China, the government has called for average power use effectiveness (PUE) of 1.25 in the east and 1.2 in the west of the country as part of its Eastern Data and Western Computing project. Major cities in China have minimum PUE requirements for new data centres, including in Beijing (1.4), Shanghai (1.3) and Shenzhen (no subsidies above 1.4).

Other jurisdictions announced temporary restrictions on new hyperscale data centre developments in 2022, including the Netherlands. Singapore recently lifted its 2019 moratorium on new data centres.

Private-sector initiatives

Several initiatives are working to measure, track and reduce the environmental impacts of digital infrastructure

• In January 2021 data centre operators and industry associations in Europe launched the Climate Neutral Data Centre Pact, which includes a pledge to make data centres climate-neutral by 2030 and has intermediate (2025) targets for power usage effectiveness and carbon-free energy.
• The Open Compute Project is a collaborative community focused on redesigning hardware technology to efficiently support the growing demands on computing infrastructure.
• The 24/7 Carbon-free Energy Compact, coordinated by Sustainable Energy for All and the United Nations, includes three data centre operators – Google (a pioneer of 24/7 carbon-free energy), Microsoft and Iron Mountain.
• DIMPACT is a collaborative project convened by Carnstone and the University of Bristol to measure and report the carbon footprint of digital services. DIMPACT participants include some of the largest media companies in the world, including Netflix, the BBC and the Economist.

Recommendations for policy makers

Improve data collection and transparency

Improving data collection and sharing on ICTs and their energy-use characteristics can help inform energy analysis and policymaking, particularly in segments where data are limited or not available (e.g. small data centres). National research programmes can develop better modelling tools to improve understanding and forecasting of the energy and sustainability impacts of data centres and networks.

Governments can also play an important role in helping to develop appropriate indicators to track progress on energy efficiency and sustainability, building on efforts by industry and researchers.

Enact policies to encourage energy efficiency, demand response and clean energy procurement

Governments can be instrumental in implementing policies and programmes to improve the energy efficiency of data transmission networks while ensuring reliability and resilience. Potential policies include network device energy efficiency standards, improving metrics and incentives for efficient network operations, and supporting international technology protocols.

Data centres could also become even more energy efficient, while providing flexibility to the grid. Governments can offer guidance, incentives and standards to encourage further energy efficiency, while regulations and price signals could help incentivise demand-side flexibility. For example, allowing for some flexibility in ancillary service requirements (e.g. longer notice periods, longer response times) may make it easier for data centre operators to participate in demand response programmes.

The ICT sector has been a leader in corporate renewable energy procurement, particularly in North America and Europe. But the Asia Pacific region has lagged behind in terms of renewable energy use due in part to the limited availability of renewables, regulatory complexity and high costs. Regulatory frameworks should incentivise varied, affordable, and additional renewable power purchasing options.

Support the utilisation of waste heat from data centres

Waste heat from data centres could help to heat nearby commercial and residential buildings or supply industrial heat users, reducing energy use from other sources. Waste heat arrangements should be assessed on a site-by-site basis and include a range of criteria including economic viability, technical feasibility, offtaker demand, and impact on energy efficiency. Given the high costs of new infrastructure, proximity to users of waste heat or existing infrastructure is needed to ensure that waste heat is actually used.

To overcome potential barriers to waste heat utilisation, such as achieving sufficiently high temperatures and contractual and legal challenges, policy makers, data centre operators and district heating suppliers need to work together on adequate incentives and guarantees.

Recommendations for the private sector

Collect and report energy use and other sustainability data

ICT companies can help energy researchers and policy makers better understand how changing demand for ICT services translates into overall energy demand by sharing reliable, comprehensive and timely data. For example, data centre and telecommunication network operators should track and publicly report energy use and other sustainability indicators (e.g. emissions, water use). Cloud data centre operators should provide robust and transparent tools for their customers to measure, report and reduce the GHG emissions of cloud services.

Industry groups that collect self-reported energy and sustainability information from members (e.g. Bitcoin Mining Council) should share underlying data and methodologies with researchers to increase the credibility of their sustainability claims.

Commit to efficiency and climate targets and implement measures to achieve them

ICT companies should set ambitious efficiency and CO2 emission targets and implement concrete measures to track progress and achieve these goals. This includes alignment with the ICT industry’s science-based target to reduce GHG emissions by 45% between 2020 and 2030.

For data centre operators, this includes following energy efficiency best practices, locating new data centres in areas with suitable climates and low water stress, and adopting the most energy-efficient servers and storage, network and cooling equipment. All companies along the ICT value chain must do their part to increase system-wide efficiency, including hardware manufacturers, software developers and customers.

Increase the purchase and use of clean electricity and other clean energy technologies

Several major data centre and telecom network operators have set and/or achieved targets to use 100% clean electricity on an annual matching basis. More ambitious approaches to carbon-free operations can have even greater environmental benefits, specifically by accounting for both location and time. ML and other digital technologies can help achieve such goals by actively shifting computing tasks to times and regions in which low-carbon sources are plentiful.

In cooperation with electricity utilities, regulators and project developers, data centre operators investing in renewable energy should identify projects that maximise benefits for the local grid and also reduce overall GHG emissions. This could also include the use of emerging clean energy technologies such as battery storage and green hydrogen to increase flexibility and contribute to system-wide decarbonisation.

Invest in RD&D for efficient next-generation computing and communications technologies

Demand for data centre services will continue to grow strongly, driven by media streaming and emerging technologies such as artificial intelligence, virtual reality, 5G and blockchain. As the efficiency gains of current technologies decelerate (or even stall) in upcoming years, more efficient new technologies will be needed to keep pace with growing data demand.

Reduce life cycle environmental impacts

In addition to their operational energy use and emissions, data centres and data transmission networks are also responsible for “embodied” life cycle emissions, including from raw material extraction, manufacturing, transport and end-of-life disposal or recycling. Companies should ramp up efforts to reduce embodied emissions across their supply chains, including devices and buildings.

Data centres and data transmission networks also pose other environmental impacts beyond energy use and greenhouse gas emissions, such as water use and the generation of electronic waste. Companies should adopt technologies and approaches to minimise water use, particularly in drought-prone areas.

Additional resources

The IEA Energy Efficient End-use Equipment Technology Collaboration Programme – Electronic Devices and Networks Annex project aims to ensure that the next generation of network-connected devices uses electricity as efficiently as possible.

The author would like to thank Arman Shehabi (Lawrence Berkeley National Laboratory), Brian Denvir (Google), Daniel Schien (University of Bristol), Devon Swezey (Google), Jake Oster (Amazon), Jens Malmodin (Ericsson Research), Simon Hinterholzer (Borderstep Institute), Steven Moore (GSMA), Stijn Grove (Dutch Data Center Association), and Vlad Coroama (Technische Universität Berlin) for their helpful comments on earlier drafts of this report.