There is a growing opposition to data centres globally: Shylesh Muralidharan, Portulans Institute

Working on AI-native climate intelligence, Shylesh is focused on building verification frameworks for carbon markets, energy transition compliance, and sustainability data infrastructure.

Shylesh has over two decades of experience at the intersection of climate technology, industrial decarbonisation, and digital transformation, with senior leadership roles at SLB, Amazon Web Services, Engie Impact, Schneider Electric, and Capgemini Consulting. He holds a Master’s in Engineering Management from MIT, where he was a System Design and Management Fellow, and also studied Global Strategic Management at Harvard Business School.

Shylesh spoke to indianexpress.com on how communities across the world are alarmed at the growth of data centres, their impact on water and power systems, and the need for greater transparency in location policies. Edited excerpts:

Venkatesh Kannaiah: Tell us about your journey to the Portulans Institute.

Shylesh Muralidharan: As a student at MIT, I did a small project on data centres and energy efficiency. Since then, I have built energy sourcing platforms and worked on clean energy supply platforms at Schneider Electric, Engie Impact, and, more recently, Amazon.

During my earlier work at Amazon, I was focused on energy and clean tech as part of sustainability mandates. It revolved around procurement, alignment with net-zero goals, and resource optimisation at an organisational level.

My focus for a long time was on energy. Then it started expanding into the broader sustainability mandate, which meant looking at water consumption, waste recycling, and incorporating circularity into all the elements being used to build infrastructure. They were all relevant pieces that applied to data centres as well.

Data centres are among the hottest and fastest-moving infrastructure investments in history. Some reports say that data centre construction is poised to surpass general office construction.

With the Portulans Institute, known for its Network Readiness Index, I am now working on building a Data Center Readiness Index.

Venkatesh Kannaiah: Tell us about the Data Center Readiness Index.

Shylesh Muralidharan: It is early days. The research question is this: how ready is a country, city or town for data centre development?

We look at requirements of power and water, water availability, network connectivity, regulatory clarity for designing policy for data centres, community sentiment, and information requirements of governments and communities to decide on locating data centres.

The idea is to bring all these variables into one comparable scale and build an index around it.

It is a multi-year research effort. We have the digital plumbing ready to extract data if it is available. The US is a bit mature in terms of government records. But that may not be the case elsewhere. The idea is that within a year, we would have the information ready, at least for the US.

There is also a lot of information available globally. The International Energy Agency produces data across regions using satellite data. NASA too produces satellite data, or the GRACE dataset, useful for understanding water stress and groundwater extraction. There are global datasets, and we are trying to make that readily available on the platform.

We think that with a Data Center Readiness score, governments, corporations and citizens will have sufficient data to talk to one another and resolve relevant issues. Cities and towns can also attract better projects and terms rather than just those based on market incentives.

Venkatesh Kannaiah: Tell us about your startup and the problems it seeks to solve.

Shylesh Muralidharan: The startup is different from the Readiness Index, as it deals with the same issue in a much more in-depth manner, incorporating proprietary data with many more parameters.

We are looking at whether we can drop a pin anywhere on planet Earth and extract all the parameters and domain variables required for a data centre location.

There are issues of fibre connectivity, ease of accessing grid energy, labour availability, quality of labour, and even proximity to colleges. We have come up with over 12 domains relevant for a data centre, and we have extracted over 500 parameters across these domains. They all trace back to citable sources, public or otherwise, and give a picture of everything that one needs to know about locating a data centre.

There is information asymmetry too. The big data centre builders have large teams doing such work. And a government official who has to design policy on data centre location works with just a land zoning map. We are looking to bridge this gap.

Venkatesh Kannaiah: Give us an overview of how communities in the US and globally are reacting to the setting up of data centres.

Shylesh Muralidharan: There is a growing opposition to these projects. A report by the Data Center Watch in the US shows that about 75 projects worth around $130 billion were blocked or delayed in the first quarter of 2026.

Two years ago, there were about 50 or 60 data centre moratorium bills in the US. Now, almost every state in the US, or sometimes even at a county level, is looking at restricting investments in data centre projects.

In Ireland, data centres consume over 20% of the national electricity, and there are severe restrictions now on getting onto the grid, especially in and around cities like Dublin. There are increasing restrictions in Singapore and Amsterdam.

Communities are weighing very local resource consumption, with very abstract benefits. People are worried that data centres will suck up all the water and power in their area. They also feel that the large tech companies will make most of the money.

Studies show that a large data centre can increase the average temperature in a specific place, which will impact the flora and fauna around it. Researchers say that species migration might happen inadvertently.

Large data centre infrastructure investments are moving too fast for communities. Hence, they are requesting more disclosure before projects are approved.

In India, we still run primarily on coal-based power, and when you lock in a data centre project which can run for 20- 25 years or longer, you are also locking in emission-intensive power, which would impact clean energy goals.

Venkatesh Kannaiah: How is a data centre different from a coal plant or a steel plant?

Shylesh Muralidharan: One way to think of it is that a data centre is an industrial facility wearing office-building clothes. While a coal power plant or a steel plant faces scrutiny under environmental laws, appraisals, mandatory public hearings and impact studies, large data centres do not have any of this, despite having the same impact.

In some geographies, people see cement and steel as critical to a country’s infrastructure development and GDP, and their impact is felt right away.

Communities opposing data centres are primarily doing so because the entire process is opaque. Their request is to make disclosures mandatory before starting these projects.

For example, mines have to file closure plans before they actually design closure. Power plants have to file decommissioning reports with governments before they are decommissioned. Data centres don’t have any of those because it’s a new space. People are still figuring out what the reporting requirements are.

Data centre developers shouldn’t think of transparency as a penalty or a slowing down of the process, because being transparent is the only way to gain the trust of the community.

Venkatesh Kannaiah: Where do you think all this power for data centres is going to come from?

Shylesh Muralidharan: Data centres are going to draw more power than cement, steel and a few other core infrastructure sectors in the next five years.

It is also not easy to get additional power onto the grid quickly. There are interconnection challenges. It’s not only about generation. You need to have transmission and distribution infrastructure in place.

Countries like India are converging in terms of cost for clean energy. Clean energy in India is one of the cheapest in the world.

Companies like Amazon are turning into power utilities to make sure they have better control over the power they consume.

Small modular nuclear plants are getting popular. Usually, it is to satisfy demand of 300 MW or lower. I wouldn’t say it is available at scale. There are a lot of safety concerns, again, from communities.

Scientifically, nuclear energy is a fairly safe technology, but you know that community sentiments don’t necessarily work on scientific reality.

You can’t turn off fossil fuel overnight. You still need them to keep it going. But you can’t also lock in fossil fuel for the next 20 years and then hope to have a clean transition.

Distributed data centres instead of one large data centre is a theme that is under consideration. Some people are now turning unused office space from COVID into data centres. However, I haven’t seen a thorough study of the economics of any distributed data centre yet.

There is also talk of orbital or space data centres. We don’t know their economics yet.

Venkatesh Kannaiah: Should India be getting on to the data centre bandwagon?

Shylesh Muralidharan: Yes. India generates roughly one-fifth of the world’s data, but only hosts about 3% of global data centre capacity.

And with AI, you’re looking at data localisation, which most countries are now insisting on. There is going to be a digital public infrastructure stack, 5G, AI workloads. So this is the right thing for India to do.

India should absolutely build to be ready for the future. But the question is: can it build with the discipline and disclosure that other countries are now learning?

Venkatesh Kannaiah: What should India work on before getting on the data centre bandwagon?

Shylesh Muralidharan: There is a need for a national-level data centre policy. The idea is to make sure one is transparent with regard to all aspects of data centres. How much power is being drawn, who’s paying for grid upgrades, who’s paying for water resources? All these need to be clear.

We need to have reporting and compliance requirements both for water discharge and water consumption.

No shovel before full disclosure should be the policy. Make transparency the rule from the start, rather than it becoming an afterthought.

One also needs to pre-qualify data centre location zones. We need to identify zones with infrastructure, pre-assessed for grid capacity, water availability, water treatment plants, connectivity and community consent.

Also, one must make sure that India is committing to building out the whole process with clean energy.

Venkatesh Kannaiah: If you had to choose between cheaper electricity, abundant water and renewable electricity when selecting a new data-centre location, which would you prioritise, and why?

Shylesh Muralidharan: I would choose renewable electricity. Because it’s the only variable where you cannot engineer around.

Water is still an easier problem to solve through engineering. There is a process called closed-loop cooling. There are technologies where, even if you are in a water-stressed region, you can solve it by designing effectively around it. That does not mean it should be located or designed around water-stressed regions.

Cheap electricity is a trap because, a lot of times, it is not clean and heavily subsidised at various points.

In India, cheap and renewable electricity are converging. India has some of the cheapest renewable energy in the world. So, if you can pair cheap, clean energy with storage, then you get round-the-clock supply.