Sustainability on GCP

Sustainability is an increasingly important consideration in cloud architecture. The environmental impact of data centres — electricity consumption, water usage, and carbon emissions — is significant and growing. Google Cloud has made substantial commitments to sustainability, and the Architecture Framework encourages architects to consider the environmental impact of their design decisions alongside reliability, security, performance, and cost.

---

Google's Sustainability Commitments

Commitment	Details
Carbon-free energy by 2030	Google aims to run all data centres on carbon-free energy 24/7 by 2030
Net-zero emissions since 2007	Google has offset all operational carbon emissions since 2007
Matched 100% renewable energy since 2017	Google purchases renewable energy equal to 100% of its electricity consumption
Water stewardship	Google is committed to replenishing 120% of the water it consumes by 2030
Circular economy	Google designs hardware for longevity and recycles components at end of life

---

Carbon Footprint Reporting

Google Cloud provides tools to measure and report the carbon footprint of your cloud usage:

Carbon Footprint Dashboard

The Carbon Footprint dashboard in the Cloud Console shows:

Metric	Description
Gross carbon emissions	Total CO2 equivalent emissions from your cloud usage
Net carbon emissions	Emissions after Google's carbon offsets (typically zero)
Location-based emissions	Emissions based on the grid electricity mix of each region
Market-based emissions	Emissions after accounting for renewable energy purchases

Carbon Footprint Export

# Export carbon footprint data to BigQuery for analysis
# Configured via the Cloud Console > Carbon Footprint > Export

-- Analyse carbon emissions by project and region
SELECT
  project.number AS project,
  location.region AS region,
  SUM(carbon_footprint.gross_carbon_footprint_kgCO2e) AS gross_emissions_kg,
  SUM(carbon_footprint.electricity_consumption_kWh) AS electricity_kwh
FROM `my-project.carbon_footprint.carbon_footprint`
WHERE DATE(usage_month) >= DATE_SUB(CURRENT_DATE(), INTERVAL 12 MONTH)
GROUP BY project, region
ORDER BY gross_emissions_kg DESC

---

Choosing Low-Carbon Regions

The carbon intensity of Google Cloud regions varies significantly based on the local electricity grid mix:

Region Carbon Intensity

Region	Carbon-Free Energy %	Recommendation
us-central1 (Iowa)	~97%	Excellent — high wind energy
europe-north1 (Finland)	~97%	Excellent — high renewable mix
northamerica-northeast1 (Montreal)	~98%	Excellent — hydroelectric power
europe-west1 (Belgium)	~53%	Moderate
asia-east1 (Taiwan)	~18%	Higher carbon — fossil fuel heavy grid

Region Selection Framework

Priority 1: Meet data residency and compliance requirements
Priority 2: Minimise latency to primary user base
Priority 3: Choose the region with the highest carbon-free energy percentage

You should not sacrifice reliability or compliance for sustainability, but when multiple regions meet your functional requirements, choose the one with the lowest carbon intensity.

---

Efficient Resource Usage

Reducing Compute Waste

Strategy	Impact
Right-size VMs	Over-provisioned VMs consume electricity without delivering value
Use autoscaling	Match capacity to demand — do not run idle resources
Adopt serverless	Cloud Run and Cloud Functions use zero resources when idle
Schedule workloads	Run batch jobs during off-peak hours when carbon intensity may be lower
Delete idle resources	Orphaned VMs, disks, and load balancers consume energy continuously

Reducing Storage Waste