India’s data centre sector is navigating a pivotal shift in its energy story from managing soaring power demand to actively shaping a greener, more resilient grid. As artificial intelligence (AI) workloads intensify and hyperscaler commitments deepen, the pressure to decarbonise infrastructure has never been more urgent. Shilpy Dewan, Vice President, Markets, Operations & Digital, Serentica Renewables; Sindhu Sharma, Head – ESG, Nxtra by Airtel; Anvesha Thakker, Partner, Business Consulting and National Lead, Clean Energy, KPMG India; and Nimesh Thakur, Vice President, Hyperscale Sales – India Head, NTT, discussed how India’s data infrastructure is powering and greening up for the decade ahead. Key takeaways from their addresses…
Shilpy Dewan
India’s ambition to emerge as a global data centre hub is being reinforced by policy intent as well as evolving geopolitical dynamics that position the country as an alternative destination for global data flows. However, this growth trajectory is closely tied to a critical factor, the ability to deliver reliable, round-the-clock power at scale.
The challenge of ensuring uninterrupted power supply is not entirely new. Energy-intensive industrial operations such as metal smelting have long operated with inflexible loads that cannot tolerate even momentary disruptions. Traditionally, these facilities relied on captive thermal power plants and remained disconnected from the grid due to reliability concerns. The shift towards green energy in such sectors led to the development of more distributed power solutions. For instance, instead of concentrating renewable capacity in a single geography, generation was diversified across regions, with solar assets in Rajasthan complemented by wind installations in Maharashtra, Karnataka and Andhra Pradesh. This geographic spread allows variations in wind and solar generation profiles to balance each other, enabling a more despatchable renewable power supply.
To further stabilise supply, battery storage systems have been deployed not only at solar sites but also alongside wind assets, which is less common. In addition, renewable capacities are intentionally overbuilt, enabling excess generation to be managed dynamically. This combination enables the delivery of consistent power. In addition, pumped storage has been integrated to address longer-duration gaps, ensuring power availability even when renewable generation is low.
These solutions were initially deployed for large industrial loads in the range of 600-700 MW and are now being seen as applicable to data centres with relatively smaller but similar load profiles.
As these systems become more distributed, managing them across locations becomes critical. Managing multiple generation sources, storage assets and demand centres requires a strong digital backbone, capable of operating at granular time intervals.
While renewable energy has expanded rapidly due to its declining costs and ease of deployment, its limitations are becoming more pronounced. High solar penetration has led to situations where discoms are unable to absorb additional solar power and are unwilling to provide banking during solar hours, as generation already exceeds demand during those periods. This has brought a renewed focus on the need for stable baseload power. Existing coal-based assets continue to contribute, but new coal capacity is no longer as cost competitive. In this context, nuclear energy is emerging as a potential option to support long-term, continuous supply.
Moreover, policy developments have opened limited pathways for private participation in nuclear projects, typically through partnerships with government entities, with fuel management remaining a key constraint. At the same time, small modular reactors (SMRs) are being explored as a future solution. While piloted in parts of Europe, these technologies are not yet available at commercial scale but are expected to evolve over the next five to ten years.
The evolving energy landscape is also prompting a shift in how data centres approach power planning. Earlier strategies often prioritised low-cost solar adoption, with other requirements addressed at a later stage. This approach is now leading to higher costs and operational complexities as storage and balancing needs become unavoidable.
A more integrated approach is, therefore, essential. Energy considerations need to be embedded early in the planning process, rather than treated as a secondary factor. Aligning generation, storage and grid connectivity from the outset can enable more efficient and cost-effective solutions.
Long-term reliability remains a key requirement, with power solutions needing to deliver certainty over extended time horizons. This underscores the importance of experienced partners and the need for at least 10-year certainty in power supply arrangements.
In parallel, there is a need for greater industry participation in policy shaping, with governments being responsive when the right issues are raised at the right time. Grid resilience is also becoming a critical determinant, with considerations around whether to remain dependent on state-level policies or connect to central networks to mitigate risks.
As the sector evolves, lessons from industrial applications that have already navigated similar challenges offer a practical reference point for data centres.
Sindhu Sharma
The approach to sustainability in India’s data centre industry has undergone a clear shift over the past five years. Earlier, the focus was primarily on integrating renewable energy into operations, with discussions centred on sourcing and targets.
This has evolved significantly. Clients and investors are now looking at data centres through an environmental, social and governance (ESG) lens. Environmental considerations remain important, but they form only a part of a broader framework, which also includes social and governance aspects.
The nature of engagement has also changed. Earlier, the focus was on renewable energy sourcing and whether defined targets such as SBTi (science-based targets initiative) were in place. Now, there is increasing emphasis on whether climate-related assessments have been undertaken, including the TCFD (Task Force on Climate-Related Financial Disclosures) analysis, physical risk assessment and transition risk evaluation.
Within the environmental pillar, expectations have expanded beyond renewable energy sourcing. There is a growing focus on demand-side measures such as reducing cooling load and improving operational efficiency. Water management and baseline clarity are also being considered, along with other environmental aspects. Environmental performance is no longer assessed in isolation. It is being evaluated alongside social and governance parameters, reflecting a more integrated approach to sustainability.
Renewable energy sourcing strategies have evolved over time. Initial efforts focused on simple contracts for solar and wind energy. With changes in regulations and the phasing out of banking mechanisms, there is now a need for a more balanced approach.
The strategy has moved towards a mix of solar, wind and battery storage. There has also been a transition from intra-state sourcing to incorporating interstate transmission system capacities, enabling better alignment with data centre load profiles.
The regulatory landscape has also improved. Awareness of open access mechanisms and settlement processes has increased and central guidelines, such as green energy open access regulations introduced in 2022, have been adopted by most states. At the same time, changes in banking mechanisms have influenced sourcing strategies. The progression from yearly banking to monthly and time-of-day banking indicates a shift towards reducing dependence on banking altogether. This requires sourcing strategies that are designed accordingly.
Sustainability initiatives are increasingly being supported by technology-led interventions. There is a focus on integrating internet of things systems to track water utilisation at multiple levels and on implementing tools to monitor waste from origin to disposal, supporting circularity.
Digitalisation is also a key area of investment. AI and machine learning layers have been deployed over building management systems, enabling a shift from preventive to predictive maintenance. This allows assessment of asset life, based on actual conditions rather than predefined timelines.
In addition, new technologies are being explored. This includes fuel cell installations at data centre facilities, with additional capacity being added. There is also continued exploration of different technologies to strengthen the overall sustainability agenda.
A single-source approach is no longer sufficient. Plain renewable energy contracts are not adequate to meet the current requirements. The focus is now on building a diversified sourcing strategy, which includes solar-wind-grid-based supply and battery energy storage. The role of partnerships is critical in this context. Identifying the right developers and solutions is essential to achieving high levels of renewable energy sourcing.
There is also an emphasis on adapting to policy changes rather than depending on them. The availability of multiple solutions makes it possible to move forward, with sourcing strategies aligned to current and future requirements.
Sustainability is being addressed as a comprehensive framework encompassing ESG aspects. While environmental initiatives have gained traction, there is increasing recognition of the need to focus on social and governance dimensions. This includes areas such as resource building, community engagement and gender diversity. Governance plays a critical role across all levels, extending beyond corporate or board-level structures to operational functions. These aspects are seen as important in strengthening the overall sustainability framework.
Anvesha Thakker
India’s data centre market continues to evolve, with energy management emerging as a central theme. While several trends are becoming visible, much of the progress remains at the planning or pilot stage and there is still a long way to go in terms of implementation. The discussion around energy management is broadly split between demand-side optimisation and supply-side considerations.
On the demand side, the focus is on reducing energy consumption within data centres, particularly addressing the high energy requirements of cooling systems. Cooling accounts for nearly 30-40 per cent of a data centre’s total energy load, making it a key area of attention. A range of measures is being explored to address this. Techniques such as hot and cold aisle containment, along with solutions such as liquid immersion cooling, are being piloted, tested or planned. These approaches are aimed at reducing the energy required for cooling, including through improvements in airflow design and cooling systems.
Beyond physical infrastructure, operational optimisation is also being considered. The use of AI presents possibilities for optimising load planning, shifting loads to non-peak hours, and enabling sensor-based adjustments of cooling and airflow. However, the extent of AI deployment and its impact will become clearer as data centres begin operating at scale.
On the supply side, there is significant focus on green energy. However, the requirement is for power that is not only green but also schedulable, reliable and capable of meeting baseload demand, given that the data centre demand is highly inflexible. While policy provisions such as deemed distribution licensing and open access have been introduced, there are still implementation-related challenges that need to be addressed. The ability to enable and transmit large volumes of power remains a key consideration.
In response, most data centres are looking at hybrid renewable energy models, combining solar and wind with battery storage, to replace a significant portion of conventional power consumption.
A key issue in energy planning is the need to look at data centre demand at a location-specific level rather than only from an aggregate national perspective. While the overall demand from data centres may appear limited at the country level, the impact at the state or city level can be substantial.
For instance, in Maharashtra, projected data centre demand represents a significant share of the peak demand. Within Mumbai, the expected increase in data centre load could lead to a sharp rise in the peak demand, in a city where transmission constraints already exist and bringing in additional power is challenging.
This highlights the importance of planning for preferred locations, transmission infrastructure and visibility of power availability at a more granular level. Moreover, for new entrants, power planning and policy engagement with state governments are critical to ensuring smooth implementation and cost visibility.
From a policy standpoint, there is a need for states to develop comprehensive data centre-related energy guidelines and frameworks, rather than having fragmented approaches. A consolidated framework can provide visibility not only to developers but also to policymakers and decision-makers, enabling a clearer understanding of goals and coordinated action. There is also a need for interdepartmental coordination bodies to support infrastructure planning and implementation. Such mechanisms can also support the adoption of emerging technologies, including hydrogen-based solutions, although these are likely to play a role over a longer time frame.
Partnering with reliable renewable energy providers is equally important, particularly given that hybrid systems combining solar, wind and battery storage are still evolving in India and require strong execution capabilities.
On the demand side, optimising infrastructure and leveraging AI can support better load management. There is also a need to explore grid interactivity, including the ability to adjust loads in response to grid conditions, support peak reduction and where feasible, return power to the grid.
Data centre development requires careful energy planning alongside infrastructure development. It is important to assess impact at a disaggregated level and address grid-related constraints specific to each location. Ensuring that power planning is undertaken with this level of detail can help identify and address potential bottlenecks early in the process.
Nimesh Thakur
Sustainability in India’s data centre sector is increasingly being treated as a core operational priority rather than a peripheral commitment. The shift reflects a deeper alignment between business requirements and environmental responsibility, with power consumption emerging as the central concern in driving this transition.
The move towards sustainability began with a conscious decision to invest in renewable energy at a time when the concept was largely viewed as a compliance requirement. Instead of treating sustainability as a checklist item, the focus shifted to integrating green power directly into data centre operations. This approach stems from the fundamental nature of data centres as high energy consumers. Managing and optimising power usage is intrinsic to operations, placing an implicit responsibility on operators to minimise wastage. Alongside technological improvements aimed at reducing the PUE (power usage effectiveness), sustainability efforts are being anchored in the sourcing and management of energy itself.
A key element of this strategy has been the development of self-captive renewable energy capacity. Installations spanning Maharashtra, Karnataka and Tamil Nadu collectively account for over 500 MW of capacity. These deployments combine wind energy, solar power and battery storage, forming a hybrid model designed to improve reliability, while increasing the share of green energy.
While the long-term ambition has been to achieve 100 per cent renewable energy usage by 2030, evolving policy conditions, particularly around energy banking, have introduced complexities. As a result, increasing storage capacity is becoming critical to ensuring consistent green energy utilisation, especially given the intermittent nature of renewable sources.
Policy intent around renewable energy is seen as strong, with governments actively working to facilitate adoption. However, challenges persist at the implementation level, particularly due to variations across states.
Further, inconsistencies in open access regulations remain a key concern. Fluctuations in banking policies and the costs associated with wheeling charges further complicate the economics of renewable energy procurement. These factors create uncertainty for operators attempting to scale green energy usage across multiple regions.
The exploration of alternative energy solutions is gaining traction as data centre campuses scale in size. SMRs are one such area under evaluation, with the potential to generate power at the facility or campus level. However, the deployment of such technologies is closely tied to the development of supportive policy frameworks. With large campuses in markets such as Mumbai reaching capacities of 1-2 GW, the question of integrating SMRs becomes increasingly relevant.
The absence of clear policies around their deployment presents a challenge, particularly given the long lead times associated with both data centre construction and energy infrastructure development. Establishing regulatory clarity in the near term is seen as essential to enabling future adoption, especially as projects typically take several years to materialise.
Sustainability is no longer positioned as an optional initiative but as an integral component of data centre operations. It is increasingly viewed as a business imperative, closely linked to long-term growth and operational efficiency.
At the same time, it reflects a broader shared responsibility, requiring coordination between industry stakeholders and policymakers. As the sector continues to expand, the alignment of technological advancements, energy strategies and regulatory frameworks will play a defining role in shaping its sustainability trajectory.
