Rising AI Data Centres Revive Polluting Peaker Power Plants

AI Growth Meets India’s Energy Challenge Artificial Intelligence (AI) is transforming India at a rapid pace. From digital payments and smart governance to healthcare diagnostics and education platforms, AI is becoming the backbone of the country’s digital future. At the heart of this revolution are AI data centres—massive facilities that store data, run complex algorithms, and power AI models 24/7. However, behind the promise of innovation lies a growing environmental concern. The explosive rise of AI data centres is placing unprecedented pressure on electricity systems. In India, this surge in power demand is forcing utilities to rely once again on dirty ‘peaker’ power plants—high-polluting power stations meant only for short-term use during peak demand. This trend raises serious questions about India’s climate goals, clean energy transition, and sustainable digital growth.  

What Are AI Data Centres? AI data centres are large-scale facilities that house thousands of servers, processors, and cooling systems. These centres are different from traditional data centres because AI workloads require: Extremely high computing power Continuous electricity supply Advanced cooling systems Low-latency, high-speed connectivity 

Popular AI applications such as machine learning, generative AI, facial recognition, and language models consume enormous amounts of electricity. A single AI-focused data centre can use as much power as a small town. India’s Data Centre Boom India is fast becoming a global data centre hub due to: Rapid digitalisation Growing internet users Expansion of cloud services Government initiatives like Digital India Rising demand for AI services 

Major data centre hubs are emerging in Mumbai, Chennai, Hyderabad, Bengaluru, and Noida. While this growth boosts the economy, it also intensifies electricity demand.  

Understanding ‘Peaker’ Power Plants 

What Are Peaker Plants? Peaker power plants are electricity-generating units that operate only during periods of high demand, such as: Heatwaves Evening peak hours Sudden industrial load spikes 

They are usually switched on when renewable energy or base-load power plants cannot meet immediate demand. Why Are Peaker Plants Dirty? Most peaker plants in India run on: Coal Diesel Natural gas 

These fuels are expensive and highly polluting. Peaker plants: Emit high levels of carbon dioxide (CO₂) Release nitrogen oxides and particulate matter Are less energy-efficient Have higher per-unit emissions than regular plants 

They were designed for emergency use, not continuous operation.  

How AI Data Centres Are Bringing Peaker Plants Back Massive and Continuous Power Demand AI data centres operate 24/7, unlike many industries that follow fixed working hours. Their electricity demand is: Constant Predictable but very high Sensitive to outages 

Even a few minutes of power disruption can cause massive data losses. Grid Stress During Peak Hours India’s power grid already faces stress due to: Rising air conditioner use Urbanisation Industrial growth Extreme weather events 

When AI data centres add to this load—especially during peak hours—utilities are left with limited options. Renewable energy alone cannot always fill the gap due to: Solar power declining at night Wind power variability Limited battery storage capacity 

As a result, peaker plants are switched on more frequently.  

Environmental Impact of Restarting Peaker Plants Rising Carbon Emissions India has committed to reducing carbon intensity and achieving net-zero emissions by 2070. However, increased use of peaker plants: Pushes up coal and gas consumption Undermines renewable energy gains Increases overall emissions 

AI, often promoted as a tool for sustainability, ironically becomes a contributor to climate change. 

Air Pollution and Public Health 

Peaker plants worsen air quality, especially in urban and industrial regions where data centres are located. This leads to: Higher respiratory illnesses Increased healthcare costs Poor air quality index (AQI) levels 

Cities like Mumbai and Delhi already struggle with pollution, and additional emissions intensify the crisis.  

Economic Costs of Peaker Power High Cost of Electricity Electricity generated from peaker plants is much more expensive than renewable energy. These costs are eventually passed on to: Consumers Businesses State electricity distribution companies (DISCOMs) 

This increases electricity tariffs and financial stress on power utilities. Burden on State Utilities Many state DISCOMs are already financially weak. Frequent reliance on peaker plants: Raises fuel import bills Increases operational losses Worsens debt levels 

This threatens long-term power sector stability.  

Why Renewable Energy Alone Is Not Enough—Yet India has made impressive progress in renewable energy, especially solar and wind. However, challenges remain: Intermittency Problem Solar works only during daylight Wind depends on weather conditions Output fluctuates unpredictably 

AI data centres need round-the-clock power, which renewables cannot always guarantee. Limited Energy Storage Battery storage technology in India is still developing. Large-scale batteries are: Expensive Limited in capacity Not widely deployed 

Without storage, renewables cannot replace peaker plants during demand spikes.  

India’s Climate Commitments at Risk India’s international commitments include: Reducing emissions intensity of GDP Expanding non-fossil fuel capacity Increasing renewable energy share 

The return of peaker plants due to AI demand creates a policy contradiction. While digital infrastructure expands, climate goals face setbacks.  

What Can Be Done? Sustainable Solutions for India 

1. Green Data Centres 

India must promote green data centres that use: 100% renewable energy contracts On-site solar power Energy-efficient cooling systems 

Mandatory green certification standards can reduce dependence on fossil fuels.  

2. Investment in Energy Storage Battery energy storage systems (BESS) are crucial. Government support through: Subsidies Viability gap funding Public-private partnerships 

can help scale up storage solutions and reduce peaker plant usage.  

3. AI for Energy Efficiency Ironically, AI itself can help solve the problem. AI-driven systems can: Predict electricity demand Optimise data centre workloads Reduce peak-time consumption 

Smarter energy management can ease grid pressure.  

4. Time-Based Electricity Pricing Introducing time-of-day tariffs can encourage data centres to shift non-critical tasks to off-peak hours. This reduces peak demand and reliance on peaker plants.  

5. Stronger Policy and Regulation The Indian government can: Set emission limits for data centres Penalise excessive fossil fuel use Encourage renewable power purchase agreements (PPAs) 

Clear rules can align digital growth with sustainability.  

Role of State Governments and Urban Planning States hosting major data centres must integrate: Clean power availability Grid upgrades Environmental impact assessments 

Urban planning should ensure data centre clusters are supported by renewable-rich grids.  

Global Lessons for India Countries like: Denmark Netherlands Singapore 

are enforcing strict green data centre norms. India can learn from these models by balancing innovation with environmental responsibility.  

The Bigger Question: Is AI Sustainable? AI is often seen as the future, but its environmental footprint is growing. Without careful planning, AI-driven growth could: Increase fossil fuel dependence Worsen climate change Deepen energy inequality 

Sustainability must become a core part of India’s AI strategy.  

Balancing Digital Growth and Clean Energy AI data centres are reshaping India’s economy, governance, and society. But their rising electricity demand is forcing dirty peaker power plants back into service, threatening environmental progress. India stands at a crossroads. With the right policies, investments, and innovation, the country can power its AI revolution using clean energy. Without them, digital growth may come at the cost of climate stability and public health. The challenge is clear: India must ensure that the future of AI is not powered by the mistakes of the past.