Water use varies enormously depending on server efficiency, utilisation, cooling systems, climate and the electricity supply. Its central implication is important: no single technological solution guarantees minimum water consumption in every location. Site-specific planning is essential.
Odisha is standing at the crossroads of an ambitious industrial transformation. However, while pursuing such transformation, the environmental consequences of major projects must be carefully assessed.
The state government’s approval of the Adani Group’s proposed ₹1.04 lakh crore artificial intelligence (AI) data centre park at Naraj, near Cuttack, promises to put Odisha on India’s digital infrastructure map. The proposed 1-gigawatt facility, planned over approximately 250 acres, represents one of the largest investments in India’s data centre sector. The government deserves credit for attempting to diversify Odisha’s industrial economy beyond mining, metals and minerals. However, an investment of this magnitude also demands an equally ambitious assessment of its environmental consequences.
The central question is not whether Odisha needs technological advancement. It undoubtedly does. The question is whether economic development should be pursued without first establishing how much water and electricity such a gigantic facility will consume and whether these resources can be supplied without jeopardising the interests of local communities.
The Economic Promise of a Digital Revolution
A modern data centre is the backbone of the digital economy. It houses thousands of servers that store information, process transactions and support cloud computing, artificial intelligence, online banking, digital governance and numerous other services. As India expands its digital economy, demand for such infrastructure is bound to grow.
The proposed Adani project could attract ancillary investments in telecommunications, fibre-optic networks, power infrastructure, equipment maintenance and technology services. It could also strengthen Odisha’s ambition to develop a technology hub at Naraj, attracting other IT, semiconductor and electronics companies.
A large investment can generate substantial construction activity, create demand for skilled professionals and broaden the state’s tax and industrial base. It may also improve digital infrastructure and encourage the development of specialised training institutions.
However, there is a critical distinction between the size of an investment and its employment potential. A data centre is highly capital-intensive and increasingly automated. It does not employ as many people during routine operations as a traditional factory with a comparable investment might. The number of jobs created will depend on the project’s actual design, the extent of local procurement and the development of associated industries.
Consequently, the government’s employment claims, expected revenue, investment incentives and long-term economic benefits must be assessed transparently. Odisha should negotiate measurable commitments on local recruitment, skills development, ancillary industries and public returns rather than treating the headline investment figure as sufficient justification.
The Water Question That Cannot Be Ignored
The most pressing environmental concern is water. Data centres generate enormous amounts of heat because their servers operate continuously. Cooling systems must maintain temperatures within safe limits to prevent equipment failures and ensure uninterrupted operations.
Some facilities use cooling towers and evaporative cooling systems, which consume water as it evaporates. Others employ air cooling, closed-loop liquid cooling or alternative technologies that can significantly reduce freshwater requirements.
According to the United Nations Economic Commission for Europe (UNECE), a relatively small, 1-megawatt data centre using traditional evaporative cooling can consume approximately 25.5 million litres of water annually. If that illustrative rate were applied to a fully utilised 1,000-megawatt facility, annual consumption would reach 25.5 billion litres. This is not a prediction of the Adani project’s actual water requirement; its consumption will depend on cooling technology, operating conditions, water recycling and utilisation levels. Nevertheless, the calculation demonstrates why the project’s water budget must be disclosed before water allocations are finalised.
The proposed facility’s location makes the issue particularly important. Naraj lies near the Mahanadi river system, which is vital to Cuttack’s drinking-water supply, agriculture and the wider regional economy. The Mahanadi Bachao Andolan forum has called for water security to be protected before water is allocated to the proposed data centre and other industrial developments at Naraj. Its concerns deserve a scientific response, not dismissal as opposition to development.
A river may appear abundant during the monsoon, but water availability during summer and drought conditions is a different matter. Water planning must account for seasonal flows, existing withdrawals, drinking-water requirements, irrigation demand and the ecological needs of the river.
The government must therefore publish an independent basin-level water assessment before committing freshwater to the project. The company should also disclose its expected daily and annual water consumption, the source of its water supply, its recycling targets and its arrangements for treating wastewater.
Electricity, Emissions and Ecological Costs
Water is only one part of the environmental equation. A 1-GW data centre would represent a very large electricity load. Although actual demand will depend on how the facility is developed and operated, supplying power reliably could require major investments in generation, transmission and grid infrastructure.
The International Energy Agency’s Energy and AI report estimated global data centre electricity consumption at around 415 terawatt-hours in 2024, with demand expected to rise sharply as AI expands. Electricity consumption by data centres contributes indirectly to emissions when the power is generated from fossil fuels. The environmental footprint also includes the construction of buildings, the manufacture of servers, batteries and cooling equipment, backup generators, and the eventual disposal of electronic waste.
Renewable energy can reduce emissions, but a claim that a facility will use renewable power is not enough by itself. The government should establish whether clean electricity will be available when required, how intermittency will be managed and whether the project will impose additional costs on the electricity system.
There is also a water-energy connection. Thermal power generation can consume water, meaning that a data centre’s indirect water footprint depends partly on its electricity supply. The overall impact must therefore be evaluated across the entire system, not just within the facility’s boundaries.
Why the World Is Becoming Cautious
Odisha should pay attention to the international debate. Governments and local authorities have begun restricting data centre development where electricity, water, land or community interests are under pressure.
In the United States, New York State introduced a proposed one-year construction moratorium covering data centres drawing 50 megawatts or more, as reported in July 2026. The objective was to develop stronger standards for assessing their environmental effects. In California, residents of Monterey Park voted in June 2026 to ban data centres in their city following opposition to a proposed facility.
Amsterdam, meanwhile, imposed restrictions on new data centre development and expansions through at least 2030. Ireland effectively restricted new data centre grid connections around Dublin from 2021 because of electricity-system concerns. Those restrictions were subsequently eased, with new connections subject to conditions, including on-site power generation. These cases demonstrate that governments are reconsidering where and under what conditions data centres should be built. They do not establish that entire countries have universally banned such projects.
The expert consensus is not that data centres should never be built. Rather, their environmental impact depends heavily on location, technology and regulation.
A June 2025 study by researchers associated with Lawrence Berkeley National Laboratory found that water use varies enormously depending on server efficiency, utilisation, cooling systems, climate and the electricity supply. Its central implication is important: no single technological solution guarantees minimum water consumption in every location. Site-specific planning is essential.
The World Economic Forum has likewise highlighted circular water management, reuse and recycling as important ways to reduce the sector’s water footprint. These measures can help, but their effectiveness depends on local conditions and the design of the facility.
What Odisha Must Demand Before Proceeding
Odisha should neither roll out the red carpet for this data centre project nor abandon it outright. The government must, however, ensure that the public interest is protected through enforceable conditions before granting the necessary approvals.
First, an independent water-impact assessment should establish whether the Mahanadi system can sustain the proposed water withdrawal during the driest months without affecting drinking-water supplies, irrigation or ecological flows.
Second, the company should be required to prioritise treated wastewater and recycled water wherever technically feasible, adopt water-efficient cooling systems and publish audited figures for freshwater consumption.
Third, the government should disclose the project’s electricity requirements, renewable-energy commitments, carbon emissions, backup power arrangements and plans for managing electronic waste.
Fourth, environmental scrutiny should be proportionate to the project’s scale. India’s Ministry of Environment, Forest and Climate Change clarified in August 2026 that data centres do not automatically require environmental clearance solely because they are data centres. However, qualifying building and construction projects, as well as township and area-development projects, do require prior clearance under the applicable rules. The precise requirements must therefore be established for the proposed development rather than assumed either way.
Finally, local communities, water experts, environmental scientists and independent institutions must have access to relevant information and an opportunity to scrutinise the project’s plans. Water allocations and incentives should be conditional on compliance, with penalties for violations.
Development Must Not Come at the Expense of Survival
The Adani data centre could become a landmark in Odisha’s technological journey. However, its success cannot be measured solely by the amount of money invested or the prestige associated with attracting a major corporate group.
Water security, affordable electricity, environmental protection and employment are equally important measures of development. If the benefits are private while the environmental costs are borne by ordinary citizens, the project will raise serious questions about the state’s development priorities.
Odisha has every right to aspire to become a hub of artificial intelligence and digital infrastructure. However, it must also demonstrate that modernisation and environmental responsibility can coexist.
The government should welcome investment with open arms, but not with closed eyes. Before the servers start humming at Naraj, the people of Cuttack deserve a clear answer to one fundamental question: will this technological leap strengthen their future, or will it place an avoidable burden on the water resources on which their lives depend?
The real test of Odisha’s development model is not how much investment it attracts, but how much prosperity it creates without sacrificing the resources essential to future generations.
