Big Data, Data Centers, and their Environmental Impacts Infographic

Explosive Growth of Data:

• New data created, captured, copied, and consumed globally has risen from 1.2 ZB in 2010 to 149 ZB in 2024 
• Data is growing at a compound annual growth rate of 20–43% in the U.S. (27% globally), with doubling of data creation occurring every two years within many of the largest organizations 
• 100 million+ servers now operate worldwide
• Drivers include cloud adoption, mobile devices, streaming, e‑commerce, and AI 

Rapidly Rising Energy Consumption 

• Energy consumption by US data centers grew from 60 TWh (2016) to 366 TWh (2025) 
• The electrical load is projected to triple again by 2028, reaching 6.7–12% of U.S. electricity consumption 
• On a global scale, data centers were estimated to account for about 1.5% of electricity consumption in 2024 and projected to be about 3% in 2030 
• Data center impact on national electricity consumption is geographically concentrated — VA, CA, TX, OH, IL, NY 

Water Demand & Infrastructure Impacts

• Water is consumed at two points in data center operation: 1) in cooling of data centers and the servers within them 2) in generating the electricity that data centers consume 
• Average data center uses  450,000 gallons/day
• Hyperscale centers can use millions of gallons/day 
• Data centers primarily draw upon potable water from municipal or regional water utilities which can strain existing infrastructure or require expensive upgrades

AI As A Major New Load Driver

• AI is reported to require more energy than traditional internet uses, tapping potentially hundreds or thousands of servers in far flung locations to generate answers 
• AI training clusters can each require 10,000 watts
• AI applications accounted for 10-20% of data center power consumption in 2024 and by 2028, AI may represent over half of data center electricity use 
• A Chat GPT query requires 10-12% more energy than a Google search , while the least efficient image generation model consumes over 500 times the power as the most efficient

Cooling Technology Shifts

• While evaporative cooling dominates data center operations, it demands substantial water use and produces contaminated runoff that can exceed standard municipal treatment capabilities. 

Direct-to-Chip Liquid Cooling

• Liquid-cooled plates placed directly on processors 
• Heat carried away through circulating coolant 
• Reduces both energy and water use compared to air cooling 

Immersion Cooling

• Servers submerged in non-conductive cooling fluids 
• Highest cooling efficiency and lower energy use 
• Eliminates need for traditional cooling water 

Cost & Infrastructure Tradeoffs

• Liquid cooling systems are significantly more expensive 
• Immersion cooling can cost ~50%+ more than direct-tochip systems 
• Advanced systems (boiling fluids) increase costs further 

Environmental & Safety Risks

• Risk of fluid leaks or spills Some cooling fluids contain hydrocarbon oils or PFAS chemicals 
• Requires biocontainment and spill protection systems 
• Some fluids are flammable, adding fire safety concerns 

Reducing Environmental Impacts of Data Centers 

Data Centers are Here to Stay

• Why they matter: Create jobs, economic growth, and tax revenue. Enable cloud services, streaming, global connectivity, and secure data storage.
• The Challenge:  Extremely high energy and water use. Can strain local utilities and infrastructure. Contribute to higher electricity costs and water stress.

Who Can Reduce the Impact? 

• Businesses & Large organizations, Local governments, individual users
• Reducing data centers requires action at every level

Impact Reducing Strategies

Businesses & Large Organizations:

Choose sustainable data center providers 

• Powered by renewable energy 
• High energy efficiency (PUE close to 1.0) 
• Low water use (lower WUE) 
• Uses advanced cooling, energy, and water-saving technology 
• Transparent & independently monitored water and power use 

Improve data storage practices 

• Use tiered data retention 
• Delete outdated, duplicate, or unnecessary data 
• Move rarely used data to lower‑energy storage (tape or cold storage) Compress files to reduce storage needs 

Local Government Units:

Key questions for data center proposals 

• What is the scale and purpose (AI? crypto mining?) 
• Expected annual energy and water use 
• Will renewable energy be used? 
• What cooling technology will be employed? 
• Water sources, wastewater volume, and discharge plans 
• Energy & water efficiency measures 
• Commitment to transparent, independent monitoring 
• Backup power plans (generators, batteries) 
• Spill containment (for liquid cooling) 
• Noise mitigation plans 

Individuals

Everyday choices matter 

• Streaming uses energy beyond your home—turn it off when not in use 
• HD streaming = 8× more carbon emissions than standard definition 
• Downloading frequently viewed content uses ~75% less energy than re‑streaming
• Open browser tabs continuously consume data and energy 
• Cloud storage consumes 24/7 energy and water—more than local storage 

Reduce digital waste 

• Think before uploading files or sharing with multiple recipients 
• Compress files before emailing or storing 
• Don’t automatically include entire email threads when replying 

Big Picture Takeaway 

Smarter data use + better policies + sustainable infrastructure = lower environmental impact