Written submission from Dr Felicia H. M. Liu and Professor Karen P. Y. Lai (DCU0050)

 

 

 

Written evidence submitted by

Dr Felicia H. M. Liu and Professor Karen P. Y. Lai

 

Biographies and Research Context

 

        Dr Felicia Liu is Lecturer in Sustainability at the University of York (Environment and Geography) with expertise in sustainable finance and green FinTech, including sustainable investment patterns into data centres.

        Professor Karen Lai is Professor of Economic Geography at Durham University (Geography) with expertise on global production and financial networks, state-firm relations and political economy of data centre growth.

 

This submission draws on international empirical research on data centres and digital infrastructure to identify both beneficial and precautionary lessons for the UK as it seeks to expand data centre capacity. The evidence presented consists of qualitative data informed by extensive engagement with the sector between 2022 and 2025, including over 50 interviews with industry, policy, and finance stakeholders, and participatory workshops held in Singapore and London[1], as well as stakeholder engagement in Manchester. Our submission focuses on Q2 and Q9 from the EAC Call for Evidence.

_________________________________________________________________________

 

Q2: What are the environmental impacts of different types of data centre currently having in the UK and what are the future impacts likely to be? What are the potential short, medium and long-term projections of these impacts?

 

Based on our research, we recommend that policies and guidance regarding the environmental impacts of data centres should be attuned to their varying use cases, ownership models, and design characteristics, as these differences shape not only their current environmental resource use, but also their capacity to reduce environmental impacts over the short, medium, and long term. Rather than pointing to one type of data centre as being more environmentally favourable over others, our research across different geographical contexts shows that they serve different purposes in the economy, and they have different scopes and flexibility to adopt sustainability technologies.

        Hyperscale data centres are typically owned and operated by large technology firms (e.g. Alphabet, Amazon, Microsoft). They are designed for rapid scaling and high-performance computing, including applications such as artificial intelligence and advanced data analytics.

        Short-term impacts: Hyperscale facilities are highly resource-intensive due to their size and energy demands.

        Medium-term trajectory: These operators typically have significant financial and technical capacity to invest in efficiency improvements, including advanced cooling systems, renewable energy procurement, and bespoke design optimisations.

        Long-term outlook: Hyperscalers are the most likely to lead in the adoption and retrofitting of cutting-edge sustainability technologies, particularly where these generate cost savings or reputational benefits. Therefore, while their total resource use remains high, their relative efficiency is likely to improve over time.

        Enterprise data centres serve single organisations, such as financial institutions (especially for crypto or high-frequency trading) or telecommunications providers, but operate at a smaller scale than hyperscalers.

        Short-term impacts: Their environmental footprint is more moderate than hyperscalers, reflecting smaller-scale operations. Depending on use cases (e.g. financial trading), these data centres can be highly location-sensitive, which creates competing land use in cities and adds pressure on the electricity grid and water infrastructure that are already serving high urban demands.

        Medium-term trajectory: These facilities typically face greater budget constraints and fewer economies of scale, limiting their ability to invest in leading-edge sustainability technologies.

        Long-term outlook: Without targeted incentives or regulatory pressure, enterprise data centres may lag behind hyperscalers in efficiency gains, leading to comparatively higher environmental impacts per unit of computation over time.

        Colocation and bare metal data centres serve a wide range of organisations, including governments and corporations, where infrastructure is standardised and shared across multiple clients.

        Short-term impacts: These facilities benefit from economies of scale and high utilisation rates, which can improve efficiency at the system level.

        Medium-term trajectory: Operational constraints (e.g. service-level agreements, priorisation of uptime, coordination across multiple tenants) could limit the adoption of new technologies to minimise service interruptions.

        Long-term outlook: These data centres may face the greatest challenges in improving sustainability and decarbonisation. Their limited flexibility and split incentives between operators and clients can slow the uptake of sustainability measures, potentially making them more resource-intensive over time unless there is coordinated demand or regulatory intervention.

 

Other than types of data centres, their location in different parts of the location could also affect their environmental impacts and future projections.  With climate change scenarios, the UK is likely to experience more days where temperatures exceed 25 degrees (see Table 1), especially in the largest data centre hub in Slough. This could lead to differentiated adaptation responses of data centres in different parts of the country, with data centres located in the south facing higher resource demands for air or water cooling. Without clear policy or regulatory guidance for cohesive industry-wide development, the longer term outlook could take the form of regional disparities in the spatially uneven adoption of sustainable technologies in data centres and their environmental impacts.

 

Table 1. Number of days exceeded 25 degrees celsius in key UK data centre hubs under different climate change scenarios (Source: The authors, based on data from the Met Office Climate Portal)

Location (number of data centres)

2001-2020

+1.5°C scenario

+2.0°C scenario

+3.0°C scenario

Slough (140+)

33.5

37.5

45.9

61.9

Birmingham  (10+)

21.0

25.0

30.9

42.3

Leeds (10+)

13.2

15.5

21.0

30.1

Manchester

(25+)

16.1

17.8

22.8

31.4

Newcastle (10+)

3.1

7.4

11.8

19.0

 

_________________________________________________________________________

Q9: Are there beneficial or precautionary lessons to learn from the impact of data centres outside the UK?

 

Beneficial lessons

A key beneficial lesson from our research in Singapore (Southeast Asia’s leading data centre hub facing severe energy, water, and land use challenges) is the importance of early policy alignment between digital infrastructure expansion, digital economic strategy, and sustainability frameworks. Since imposing a moratorium on data centres in 2019, Singapore has brought together key ministries, including the Ministry of Sustainability and Environment and the Ministry of Manpower and statutory bodies (e.g. Infocomm Media Development Authority, Economic Development Board, Public Utilities Board, National Climate Change Secretariat) to devise the Digital Connectivity Blueprint and the accompanying Green Data Centre Roadmap. The former comprehensively considers the economic, infrastructural, labour and skills, as well as environmental priorities in designing Singapore’s digital economic strategy, whilst the latter focuses on engaging with the entire data centre ecosystem, including facility design, computing hardware design, energy sources, and innovation, to highlight opportunities for sustainable transitions. These strategic documents culminated in actionable policies, including the establishment of the Green Mark for Data Centres (which all new-built data centres must comply with), the incorporation of new-built and refurbished data centres into the Singapore-Asia Taxonomy for Sustainable Finance, the Energy Efficiency Grant for the data center sector, and the voluntary Singapore Standard on Energy Efficiency of Data Centre IT Equipment.

Our research shows that these measures provide clarity to developers, investors, and users by establishing a shared framework for understanding the different dimensions of data centre sustainability, thereby reducing ambiguity around what qualifies as a “green” or “sustainable” data centre. These provide a stable foundation for ongoing refinement, allowing environmental standards to be progressively strengthened over time. At the same time, they enhance institutional confidence in green investments, which could help attract new investors with strong sustainability mandates to invest in Singapore’s data centre and digital sector. Complementarily, an energy efficiency grant providing up to SGD 30,000 (approx. GBP17,600) per company to help overcome the upfront capital barriers associated with adopting or retrofitting more sustainable equipment. Singapore’s experience shows that such frameworks serve not only as regulatory tools but also as coordination devices, aligning developers, investors, and local authorities around shared benchmarks. Government grants or other concessional financing support could spur further public-private investments into sustainable technologies and adaptation when there is better clarity and direction for the industry.

Precautionary lessons
Experiences outside the UK point to several risks associated with poorly managed data centre expansion.

First, the use of moratoria, as seen in Singapore, but also parts of the Netherlands and the Republic of Ireland, illustrates the potential for abrupt policy interventions when infrastructure growth outpaces planning and energy system capacity. While often introduced in response to legitimate concerns (e.g. grid constraints), such measures can be highly disruptive. They may undermine investor confidence, delay strategically important infrastructure, and damage the host country’s reputation as a stable destination for digital growth and investment.

Second, there is growing evidence of community resistance. As a result, data centres are losing their social licence to operate. In Dublin, public concern over energy consumption, energy costs, and perceived uneven local benefits has led to significant opposition to new data centre projects. Similar dynamics are emerging in parts of the USA. While public awareness of data centres in the UK remains low, there are emergent campaigns against data centre development within the UK. For example, Foxglove and Global Action Plan successfully brought a case against the government’s approval of a Buckinghamshire data centre, raising concerns over the electricity and environmental implications.

These cases highlight that data centres are no longer perceived as neutral or low-impact infrastructure but are becoming contested. This is especially so when local environmental impacts are visible but economic and social benefits are perceived as unevenly distributed or failing to benefit local communities, and when infrastructure development outpaces public engagement. We therefore recommend that data centre regulations and standards should consider incorporating ‘social sustainability’ as a key criterion to manage the risks of potential political, social, and reputational backlash.

_________________________________________________________________________

Summary

Evidence from our research highlights the need for coordinated and proactive government action to ensure the sustainable growth and development of data centres in the UK. Policies should recognise that hyperscale, enterprise and colocation data centres have distinct roles to play in growing the digital economy, as well as distinctive environmental impacts and differing capacities to adopt sustainability measures, requiring tailored regulation and targeted incentives to support efficiency improvements and reduce long‑term resource use. There should be clear, comprehensive, and forward‑looking sustainability standards that integrate data centre planning within broader digital, energy and financing strategies. At the same time, lessons from countries facing moratoria and rising public opposition underscore the risks of unmanaged expansion, grid pressure and loss of social licence. To mitigate these risks, policies should embed social sustainability, community engagement, and transparent benefit‑sharing to maintain public trust and ensure data centre growth aligns with national environmental and societal objectives.

 

 

 

 


[1] Liu, Felicia H.M., Lai, Karen P.Y., Seah, Bertrand and Chow, Winston T.L. (2025) Decarbonising digital infrastructure and urban sustainability in the case of data centres, npj Urban Sustainability, Vol. 5, Article No. 15; Chow, Winston, Lai, Karen P. Y., Liu, Felicia and Seah, Bertrand (2024) White Paper: Green FinTech and Data Centres in Singapore, April 2024, Singapore: Singapore Green Finance Centre.