rethink sustainability
How green data will power the energy transition
In 2010, the global population created, captured, copied and consumed two zettabytes of digital data1 – enough to store two trillion feature films. By 2021, that figure had risen to 120 zettabytes.2 And by 2030, as we become ever more connected and increasing aspects of our lives shift online, forecasts suggest we will generate one yottabyte of data annually3 – one thousand trillion (or one quadrillion) feature films’ worth.
‘Big data’ – the near-constant stream of information we hand over both voluntarily and unwittingly, and the computing power to analyse it – has become so big it can be hard to envisage. A gargantuan and rapidly growing quantity of information, data is the lifeblood of the digital age, enabling a personalised service in everything from dating to financial planning.
Big data is also playing a key role in creating efficiencies and catalysing innovations that will drive the transition to a socially and environmentally sustainable economy, including helping to accurately and quickly sort discarded materials for recycling, and optimising food production and distribution to reduce waste. However, as the Digital Revolution merges with the Sustainability Revolution, concerns are growing over the sustainability of big data itself.
Omnipresent data
Big data touches every aspect of our daily lives. For every internet search, every online purchase, even when simply travelling from place to place carrying a mobile phone – new data is generated and added to the global store. Genetic testing that unravels our DNA and predicts our susceptibility to particular diseases relies on the power of big data analytics.4 So, too, does the social media algorithm that serves you a video of a cat falling off a sofa.
Though some big data applications may seem frivolous, they all serve a purpose. For the companies that employ big data analytics, the effectiveness of their algorithms can be the key to their commercial success. In 2006, for instance, in their search for an improved algorithm to predict users’ film ratings (in order to make personalised film recommendations) Netflix launched the ‘Netflix Prize’, offering USD 1 million to the team that could come up with an algorithm that bested their own.
Netflix has continued to invest in their big data analytics – today they track every moment that each of their 278 million subscribers5 spend on the platform. The data they gather is analysed by machine learning and artificial intelligence (AI) to decide what films and series to produce, whether to renew them, and who to recommend them to.6 Across the industry, it is thought that Netflix’s expertise when it comes to analysing big data is the key to their rapid rise to becoming one of the five biggest media firms in the world7.
Read also: The metals of the future: how critical is nickel?
Intelligent energy
The same technology that has transformed business models across multiple industries is also being harnessed in the transformation of our energy systems. In many places, renewable electricity is now the first port of call when energy is needed, but the variable nature of renewables means that even where solar and wind generators have been rolled out at scale, baseload power must still come from traditional power stations.
Much of this can be solved by ‘smart grids’ that use real-time data to dynamically adjust power flows by analysing and predicting the wind and sun and monitoring and anticipating consumer usage. When the weather is favourable, smart grids divert excess renewable energy to storage systems, ready to be released at moments of high consumer demand. In the European Union, it is estimated that, by 2040, the growth of data-driven smart grids linked to energy storage systems could reduce solar and wind ‘down-time’ from 7% to 1.6%, potentially reducing CO2 emissions by 30 million tonnes8, the equivalent of taking 20 million medium-sized cars off the road9.
On the demand side, smart appliances that incorporate data-driven technologies can enable consumers and organisations to contribute to this transition. For instance, intelligent electric vehicle (EV) charging technologies have the potential to add flexibility to the grid by using data analytics to draw power more quickly when supply is high and consumer demand low. Depending on EV adoption rates, smart EV chargers could save anywhere between USD 100 billion and USD 280 billion on the need for additional electricity infrastructure between 2016 and 2040.
Data-driven technologies can also be used to turn EVs into a nationwide battery pack, with owners being paid to feed electricity back from their vehicles to the grid at times of high consumer demand, or to optimise the grid feed-in from home solar panels.10 However, big data requires an infrastructure all of its own – namely, data centres and vast computing power, both of which are highly energy intensive.
What powers big data?
Data centres have always been energy-hungry – but we haven’t always needed so many of them. In 2018, global data centre energy use amounted to 205 terawatt-hours (TWh) of electricity, a mere 11 TWh higher than in 2010.11 But by 2022, that number had more than doubled to an estimated 460 TWh, accounting for almost 2% of all global electricity demand. And this will continue to accelerate, with data centre electricity consumption expected to reach anywhere from 620 to 1,050 TWh by 2026 – equivalent to a new country the size of Germany.12
This exponential growth has been driven by the rise of cryptocurrency mining and increasingly data-reliant technologies like artificial intelligence (AI), both of which require more work than older digital technologies. A typical Google search, for instance, consumes around 0.3 watt-hours (Wh) of electricity, whereas a typical ChatGPT query requires around 2.9 Wh – the equivalent of powering a typical LED light bulb for more than 17 minutes.13
To date, the transition to renewable energy has failed to keep pace. In 2020, Microsoft pledged to become carbon-negative by the end of the decade. By 2023, driven primarily by its forays into AI, Microsoft had actually increased its emissions by 30%.14
Read also: Back to the future: the rising role of metals recycling in the energy transition
Clean energy, green data
In places, policymakers are now stepping in. The EU has already introduced regulations around data centre energy efficiency standards and reporting requirements, while Chinese regulators have mandated that all data centres owned by public organisations must run on 100% renewable energy by 2032.15 To get ahead of the potential for strict new regulations, and to meet their own corporate commitments, a number of major firms are now looking for ways to wean themselves off fossil fuels.
Data centre electricity consumption in Europe is expected to double between 2021 and 2030. However, according to a report by BloombergNEF16, this fast-growing demand need not jeopardise the energy transition. By purchasing more renewable energy or investing in dedicated, purpose-built solar arrays and wind farms, data centres could move away from fossil fuels and advance the energy transition in the process.
Microsoft began this journey when it signed a USD 10 billion deal to buy renewable power17 and committed to transitioning its data centres to 100% renewable energy.18 Google has even turned to AI to reduce its own carbon footprint, using DeepMind to cut the electricity demand of its data centre cooling systems – which are highly energy intensive – by 40%.19
Read also: Energy’s forgotten giant – grid infrastructure plays catch-up as renewables gather pace
Leading the Electrification Revolution
Data centres could play another, more fundamental role by plugging in to national grids to provide rapid response power flexibility via onsite backup batteries, and by ‘load-shifting’. According to Michael Kenefick, lead author of the Bloomberg NEF report, “Data centres can be part of the solution for achieving higher renewable energy penetrations in Europe. Their on-site resources…could be brought to bear to help support the grid. And computing tasks could be shifted to times – or locations – of high wind and solar resource.”
Report contributor Karina Rigby, president, Critical Systems, Electrical Sector at power management firm Eaton, said, “Data centre facilities are unique and comparable to microgrids in the opportunities they offer through their computing power and physical infrastructure, particularly the vast amounts of battery energy storage…This study highlights the huge untapped potential of data centre flexibility to deliver economic, regulatory and climate benefits.”
For investors, this shift to green data is set to present new opportunities. Many tech companies are increasingly focussed on so-called “hyperscale” data centres, massive facilities designed to scale efficiently as demand grows, while incorporating the latest energy-efficient technologies and practices. The global market for this new breed of data centres is expected to double between 2023 and 2026 to a value of USD 212 billion.20
As the need to fund the expansion of data centres continues to grow, investors will play an increasingly important role in tipping the scales towards a sustainable approach. At Lombard Odier, we believe the ‘Electrification Revolution’ – the transition to an electrified economy powered by renewables – represents one of the most fundamental economic shifts since the Industrial Revolution. And we believe that green data, powered by clean energy, will be at the heart of it.
1 Data growth worldwide 2010-2025 | Statista
2 Idem
3 www-file.huawei.com/-/media/corp2020/pdf/giv/industry-reports/computing_2030_en.pdf
4 Moving back to the future of big data-driven research: reflecting on the social in genomics | Humanities and Social Sciences Communications (nature.com)
5 Netflix: number of subscribers worldwide 2024 | Statista
6 Netflix and AI: How Artificial Intelligence Powers Streaming Success | by Jeyadev Needhi | Jun, 2024 | Medium
7 The World’s Largest Media Companies 2022: Netflix Falls In The Ranks After Subscriber Loss, Disney Climbs To No. 2 (forbes.com)
8 Digitalization & Energy (IEA report)
9 Chart: Miles Apart: Car CO2 Emissions | Statista
10 Digitalization & Energy (IEA report)
11 ICT Sector Electricity Consumption and Greenhouse Gas Emissions – 2020 Outcome (Malmodin et al.)
12 Electricity 2024 (IEA research)
13 Electricity 2024 (IEA research); LED Light Bulb Energy Calculator: Watts and kWh (energybot.com)
14 Microsoft’s AI obsession is jeopardizing its climate ambitions (The Verge)
15 Electricity 2024 (IEA research)
16 Data Centers Set to Double Their Power Demand in Europe, Could Play Critical Role in Enabling More Renewable Energy | BloombergNEF (bnef.com)
17 Rush for data centers creates US solar hotspots | Reuters
18 AI's looming climate cost: Energy demand surges amid data center race (Nikkei Asia)
19 Electricity 2024 (IEA research)
20 Electricity 2024 (IEA research)
Important information
This document is issued by Bank Lombard Odier & Co Ltd or an entity of the Group (hereinafter “Lombard Odier”). It is not intended for distribution, publication, or use in any jurisdiction where such distribution, publication, or use would be unlawful, nor is it aimed at any person or entity to whom it would be unlawful to address such a document. This document was not prepared by the Financial Research Department of Lombard Odier.
Read more.
share.