How Google Plans to Power AI Data Centers with 24/7 Carbon‑Free Energy by 2030

Google attributes that improvement largely to new renewable projects and long‑term power purchase agreements that started delivering clean electricity to the grids supplying its facilities.

The key shift: hourly carbon‑free energy matching

Google’s 2030 strategy differs from the renewable energy approach most companies historically used.

Traditional annual renewable matching

Many companies claim to run on 100% renewable energy by purchasing enough renewable power or renewable energy certificates over a year to equal their annual electricity use. However, the electricity actually powering their facilities at a given moment may still come from fossil‑fuel sources on the grid.

This system allows renewable generation during sunny or windy periods to offset fossil‑fuel electricity used at other times.

Google’s hourly matching approach

Google’s 24/7 carbon‑free energy goal attempts to eliminate that gap.

The company aims to match electricity consumption with carbon‑free power every hour and in every grid region where it operates.

That requirement forces a different procurement strategy:

• clean power must be available locally rather than somewhere else in the grid
• generation must align with the time electricity is consumed
• additional resources are needed to cover hours when wind or solar output is low

In practice, this pushes investment into a broader mix of technologies and grid strategies, including energy storage, geothermal, advanced nuclear, and flexible data‑center operations.

Evidence the strategy is moving forward

Several recent developments suggest Google is continuing to scale clean electricity procurement even as its energy use grows.

Large renewable procurement pipeline

Google has been one of the world’s largest corporate buyers of clean electricity. From 2010 through 2024 it signed more than 170 agreements for over 22 gigawatts of clean energy generation capacity globally.

New projects coming online

According to Google’s environmental reporting, more than 25 contracted clean‑energy projects came online in 2024, adding roughly 2.5 GW of new capacity to the grids supplying its operations.

These projects played a major role in lowering data‑center emissions even while electricity demand increased.

Long‑term power purchase agreements for data centers

New agreements also continue to expand the clean‑energy supply for AI infrastructure. For example, a set of long‑term power purchase agreements with developer Clearway is expected to provide nearly 1.2 GW of carbon‑free energy from projects in Missouri, Texas, and West Virginia, helping power data centers on those regional grids.

Measured progress toward 24/7 clean energy

Google tracks progress toward its hourly matching goal using a metric called the carbon‑free energy percentage.

Recent reporting indicates the company reached roughly 66% hourly carbon‑free energy across its operations, with several grid regions exceeding 80% coverage.

That shows meaningful progress, though it also highlights the scale of the challenge remaining before the 2030 target.

Operational strategies beyond buying renewables

Procurement alone cannot achieve 24/7 clean power. Google is also redesigning how its infrastructure interacts with electricity grids.

One approach is carbon‑aware computing, which shifts certain workloads to locations or times when cleaner electricity is available.

Another involves demand‑response capabilities that allow data centers to temporarily reduce or shift power consumption during periods when grids are under stress or fossil‑fuel generation is dominant.

Efficiency improvements are also part of the strategy. Google’s environmental research suggests that combinations of training techniques and infrastructure optimizations can reduce the energy required to train AI models by up to 100× and associated emissions by up to 1,000× in some cases.

Where DeepMind’s “AI for the Planet” accelerator fits

Google’s climate strategy extends beyond its own operations.

In May 2026, Google DeepMind launched an Asia‑Pacific accelerator called “AI for the Planet.” The three‑month program supports startups, nonprofits, and research teams building AI solutions for climate and environmental challenges.

Participants receive mentorship and technical support to apply frontier AI models to areas such as:

• nature conservation
• sustainable agriculture
• climate mitigation
• environmental risk analysis

The program reflects a broader thesis inside Google: that AI should not only run on cleaner electricity but also help accelerate climate solutions across industries.

The bottom line

Google’s plan to run on 24/7 carbon‑free energy by 2030 represents a shift in how large technology companies approach renewable energy. Instead of balancing emissions annually, the company is attempting to match its electricity use with clean power hour by hour and region by region.

Progress so far shows both momentum and difficulty. Electricity demand from AI is rising quickly, yet Google reports falling data‑center emissions thanks to new clean‑energy projects, long‑term power contracts, and efficiency improvements.

Whether the company reaches full hourly carbon‑free matching by 2030 remains uncertain, but the strategy is already influencing how large data‑center operators and cloud providers think about powering the next generation of AI infrastructure.