Home Blog GreenTech and Sustainable IT Carbon-aware computing: scheduling workloads on green g...
🌱 GreenTech and Sustainable IT May 16, 2026 12 min read

Carbon-aware computing: scheduling workloads on green grids

GreenTech and Sustainable IT Enterprise Guide 2026 SCALE D2C D2C Technology GreenTech and Sustainable IT Enterprise Guide 2026 SCALE D2C D2C Technology

Carbon-aware computing is the practice of scheduling computational workloads to run when and where electricity grids are cleanest — shifting flexible compute to low-carbon windows without any reduction in work completed. It is the single highest-impact, lowest-disruption action enterprises can take to reduce the carbon footprint of their digital operations. A workload that runs at the same time every day regardless of grid conditions misses carbon reduction opportunities of 20–60% that are available simply by running it at a different time or in a different region.

What Is Carbon-Aware Computing?

Carbon-aware computing means designing software and infrastructure to respond to the real-time carbon intensity of the electricity grid — doing more computation when electricity is generated from renewables, less when it comes from fossil fuels. It is not about doing less work overall; it is about choosing when and where to do flexible work based on when the grid is cleanest.

Carbon-Aware Computing — Definition
The practice of scheduling flexible compute workloads — batch jobs, ML training runs, database maintenance, video encoding, backup operations — to execute during periods when the electricity grid in the relevant region has low carbon intensity (high renewable generation, low fossil fuel use). Carbon-aware computing leverages real-time and forecast grid carbon data from APIs like Electricity Maps and WattTime to make scheduling decisions that reduce carbon emissions without sacrificing work throughput or user experience.

Two Types of Carbon-Aware Computing

⏰ Temporal Shifting
  • Run the workload at a different time when the grid is cleaner
  • Example: Run ML training at 2am when California's grid is 80% solar/wind, not 5pm when gas peakers are running
  • Applies to: batch jobs, training runs, backups, maintenance tasks, video encoding
  • Carbon reduction potential: 20–40% for typical deferrable workloads
🌍 Spatial Shifting
  • Run the workload in a different region where the grid is currently cleaner
  • Example: Route batch processing to eu-north-1 (Nordic hydro) rather than us-east-1 (coal mix)
  • Applies to: region-flexible batch compute, CDN origin selection, DR workloads
  • Carbon reduction potential: up to 4× difference between cleanest and dirtiest regions

Carbon Intensity APIs

APICoverageUpdate FrequencyForecastPricing
Electricity Maps70+ countries, 500+ zonesHourly, near-real-time24–72 hour forecastsFree tier + paid plans
WattTimeUS, EU, AU, expanding5-minute marginal emissions24-hour forecastFree basic + paid enterprise
National Grid ESO (UK)Great Britain grid30-minute updates96-hour forecastFree public API
Carbon Aware SDKMultiple providers unifiedDepends on providerYesOpen source

Implementation: Making Your Workloads Carbon-Aware

Carbon intensity difference between eu-north-1 (Nordic hydro — ~18gCO₂/kWh) and ap-southeast-1 (Singapore coal mix — ~450gCO₂/kWh) — spatial shifting delivers the largest carbon wins
40%
Average carbon reduction achievable by temporally shifting deferrable compute workloads to low-carbon windows within the same day — with zero impact on work throughput
Free
Cost of implementing carbon-aware scheduling for most deferrable workloads — the Carbon Aware SDK is open source, Electricity Maps has a free tier, and scheduling changes require only hours of engineering time
01
Step 1
Catalogue Your Deferrable Workloads

Audit every scheduled job in your infrastructure. Classify each as: (a) real-time — cannot be deferred (user-facing requests, critical processing), or (b) deferrable — can be shifted within a time window without user impact (ML training, batch reports, backups, data sync, video processing, database maintenance). Most enterprises find 30–50% of their compute is deferrable. Connect this to your DevOps asset inventory.

Workload classificationDeferral time windowsImpact assessment
02
Step 2
Integrate Carbon Aware SDK

The Carbon Aware SDK (from the Green Software Foundation) provides a unified interface to multiple carbon intensity data providers. Add it to your job scheduler as a carbon-aware wrapper: instead of schedule("2:00am"), use schedule_carbon_aware(within="24h", duration="2h") — it finds the lowest-carbon 2-hour window within the next 24 hours automatically. Works with any language or scheduling system via REST API or SDK.

Carbon Aware SDKScheduler integrationTime window config
03
Step 3
Kubernetes: Deploy Kube Green and Carbon Aware Scheduler

Deploy Kube Green to automatically scale down non-production Kubernetes namespaces outside business hours — saves 40% of non-prod cloud carbon. For batch workloads on Kubernetes, use the Carbon Aware KEDA Operator to scale based on real-time carbon intensity — scale up when grid is clean, scale down when dirty. All changes fully reversible and compatible with existing Kubernetes infrastructure.

Kube GreenCarbon Aware KEDANon-prod scale-down
04
Step 4
Measure and Report Carbon Savings

Instrument your carbon-aware scheduler to log: scheduled time, actual execution time, grid carbon intensity at execution vs. at original time, carbon saved per job. Aggregate into a monthly carbon savings report. Connect to your GreenOps Grafana dashboards alongside cost and performance metrics. This data is valuable for ESG reporting and for building the business case for further GreenOps investment.

Carbon savings loggingESG reporting dataGreenOps dashboards
Implement Carbon-Aware Computing

Carbon-aware scheduling is one of the highest-ROI sustainability interventions available to enterprise engineering teams — and one of the easiest to implement. Our DevOps and digital transformation teams implement carbon-aware computing programmes for enterprises targeting net-zero software operations. Book a free advisory session to design your carbon-aware computing strategy.

Frequently Asked Questions

End-to-end GreenTech and Sustainable IT strategy, implementation, and optimisation for enterprise and D2C brands. Contact us for a free consultation.

Strategy projects: 4–8 weeks. Full implementation: 3–12 months. ROI typically within 12–18 months.

Yes — D2C brands to enterprise. View our pricing.

GREENTECH AN

Ready to Implement GreenTech and Sustainable IT?

Our specialist team delivers measurable ROI from GreenTech and Sustainable IT programmes for enterprise and D2C brands.

Free Audit