‘Regulating Data Centres is not optional but strategically essential’
AI Data Centres have become a foundational pillar of modern economies, enabling digital services, Artificial Intelligence and cross-sector innovation. However, their rapid expansion has significant implications for electricity systems, water resources and emissions trajectories. Without effective regulations and governance, rising energy demand from Data Centres can strain power grids, increase electricity costs for households and businesses, intensify water stress – particularly in high-ambient regions – and elevate GHG emissions.
According to the UN Environment Programme, in 2024, global Data Centres consumed 415 TWh of electricity, about 1.5% of total global supply, and this demand is projected to double by 2030. In fossil fuel-based grids, this increases GHG emissions at a time when the world is nearing the 1.5 degrees C Paris Agreement threshold. In addition, according to the UNEP, wholesale electricity prices have risen by up to 267% in areas near large Data Centres, and planned AI Data Centres could account for 5-10% of peak electricity demand in major European markets. Large water use is also a concern, particularly where water-based cooling is used in hot or water-stressed regions.
For this reason, regulating Data Centres is not optional; it is strategically essential. Implementing energy efficiency standards, mandatory reporting obligations and benchmarking mechanisms ensures that digital infrastructure growth remains aligned with Sustainable Development Goals (SDGs), national decarbonisation goals, grid stability and sustainable development priorities. Effective regulation enhances transparency, strengthens ESG integration and embeds accountability within the sector.
Regulating Data Centres directly interconnects with ‘SDG 7 on Affordable and Clean Energy’ and ‘SDG 13 on Climate Action’. Under SDG 7, governments are committed to ensuring access to affordable, reliable, sustainable and modern energy. Unmanaged growth in high-density digital infrastructure can strain generation capacity and transmission networks, potentially affecting energy affordability and reliability. Aligning Data Centre expansion with energy efficiency standards, renewable energy procurement and grid planning supports cleaner energy systems and long-term cost stability. At the same time, SDG 13 calls for urgent action to combat climate change and its impacts. As Data Centres become larger and more energy-intensive, particularly with AI workloads, their emissions profile becomes increasingly relevant to national climate targets and decarbonisation pathways.
By setting clear performance thresholds and resource management requirements, governments can ensure that Data Centres support and accelerate the transition to a low-carbon economy rather than undermining climate and energy objectives. In this context, sustainability-focused regulation safeguards economic competitiveness and environmental resilience while enabling responsible digital expansion.
Measures to be considered to regulate Data Centres are as follows:
Enforcing energy efficiency standards
Cooling can account for 40-50% of a Data Centre’s energy use. Setting strong Power Usage Effectiveness (PUE) thresholds reduces total electricity demand, lowering emissions and decreasing the need for new fossil fuel-based generation capacity.
| For example: If a Data Centre has an IT load of 50 MW and operates at a PUE of 1.8, its total power demand is 90 MW (50 MW × 1.8). If regulation requires a PUE of 1.4, the same facility would require 70 MW instead of 90 MW. This 20 MW reduction is equivalent to avoiding the continuous output of a small power plant. Over a year, this saves approximately 175 GWh of electricity, reducing emissions and lowering the need for additional fossil fuel-based generation capacity. |
Mandating Renewable Energy procurement
If regulations require a minimum share of Renewable Electricity (e.g., Power Purchase Agreement (PPAs) or on-site generation), Data Centres already consuming ~415 TWh globally in 2024, can directly stimulate large-scale solar, geothermal and wind deployment. Because they are long-term, creditworthy buyers, they de-risk Renewable Energy projects and accelerate clean capacity build-out.
Requiring Demand Response (DR) participation
Data Centres can provide grid flexibility by shifting non-critical workloads or using on-site storage. This supports higher Renewable Energy penetration by stabilising grids during peak variability.
Imposing water efficiency metrics
In water-stressed regions, regulating cooling technologies reduces freshwater withdrawals, protecting natural resources. Water Usage Effectiveness (WUE) reporting is the measurement and disclosure of how efficiently a Data Centre uses water in relation to its IT energy consumption. By imposing WUE reporting, we can increase the water efficiency of cooling systems.
| For example: WUE = Annual Water Usage (litres) / IT Equipment Energy (kWh) If a Data Centre uses 50 million litres of water per year, and it consumes 100 million kWh of IT energy, its WUE would be 0.5 L/kWh Lower WUE = higher water efficiency |
Linking grid connection to sustainability criteria
Conditioning grid approvals on renewable sourcing and efficiency compliance ensures that new AI-related loads do not outpace clean generation capacity.
| For example: A developer proposes a new 100 MW AI Data Centre connection to the national grid. Under the Data Centre regulation, grid approval is granted only if: At least 80% of the facility’s electricity is secured through renewable Power Purchase Agreements (PPAs), and The design meets a maximum PUE of 1.3. As a result, the project must contract new solar and wind capacity before connecting, ensuring that its electricity demand is matched by clean generation rather than increasing reliance on fossil fuel-based power. This prevents AI-driven load growth from outpacing Renewable Energy deployment and protects national decarbonisation targets. The Power Purchase Agreements (PPAs) can be carried out with a long-term contractual arrangement between an electricity buyer (e.g., a Data Centre operator) and a Renewable Energy generator (such as a solar or wind farm). Under a PPA: The Data Centre commits to purchasing electricity (or the renewable attributes of electricity) for a fixed term, often 10-20 years. The agreement provides revenue certainty to the Renewable Energy developer. It helps finance and enable new Renewable Energy capacity. There are two common types: Physical PPA: The Data Centre directly receives electricity from the Renewable Energy plant through the grid. Virtual (Financial) PPA: A financial contract, where the Data Centre supports Renewable Energy generation but may not physically receive that specific power. PPAs are widely used by hyperscale Data Centre operators to secure Renewable Electricity at predictable prices while accelerating clean energy deployment. |
AI Data Centres are here to stay, and from a policy perspective, regulating them is vital to ensuring alignment with SDG 7 and SDG 13. As AI-driven facilities expand, enforceable measures, such as strict energy efficiency standards, mandatory renewable energy procurement, DR participation and water efficiency reporting, provide the most immediate and cost-effective means to reduce emissions, protect electricity affordability, and safeguard grid and water resources. Regulating Data Centres can also transform them into strategic contributors to clean energy deployment by leveraging their scale to accelerate renewable investment, enhance grid flexibility and support low-carbon energy systems.
Unregulated growth in Data Centres demand can increase system costs and risk national climate commitments, making it essential to embed sustainability requirements into Data Centres governance, so that digital infrastructure expands in a resilient, efficient and climate‑aligned manner.
The writer is Director of Energy and Sustainability at Engineering Sustainable Futures (ESF). She may be reached at angela@esfmena.com.