Emissions in the Pipeline

The most energy-intensive stages are those that require sustained, large-scale computation: 3D rendering and cloud computing, VFX simulations involving fluids, crowds, and destruction, and large-scale data centre processing. These are the stages where render farms — clusters of high-performance computers processing millions of frames — run continuously, often for weeks. Cooling these systems adds significantly to the energy load: in most data centres, cooling accounts for more than 40% of total electricity use.

Artist workstations, data storage, and re-render cycles sit at the next tier of impact. A 3D modelling workstation draws significant power during active use, and a high-performance rendering workstation can consume approximately 1,500 kilowatt-hours per month. Even idle machines contribute: computers left in standby mode represent ongoing energy waste that adds up across a studio fleet.

At the lower end of the emissions spectrum sit 2D animation pipelines, pre-production work including storyboarding and design, and stop-motion production — which has lower compute demands but its own material footprint.

The key insight is this: a small number of compute-intensive stages account for the majority of pipeline emissions. The most effective reduction strategies focus on those stages — reducing unnecessary iteration, optimizing rendering workflows, managing data intelligently — rather than making small incremental changes everywhere at once.

‍

The Canadian Grid

Canadian animation studios operate in one of the most favourable electricity environments in the world for compute-intensive work. Nationally, approximately 80% of Canada's electricity comes from non-fossil fuel sources, with hydropower alone accounting for roughly 57–64% of total generation.

But this advantage is unevenly distributed — and for animation studios, where your compute happens matters enormously.

Quebec and British Columbia both operate grids that are approximately 95% hydroelectric, making them among the cleanest electricity environments in North America. Render farms and workstation-heavy production in these provinces have relatively low operational emissions per kilowatt-hour consumed.

Ontario sits in a mid-to-low carbon position, powered largely by nuclear energy, hydroelectricity, and some natural gas — significantly cleaner than fossil-fuel-heavy regions but not as low-carbon as the hydro-dominant provinces.

Alberta and parts of Atlantic Canada, with higher reliance on natural gas and in some cases coal, produce substantially higher emissions per kilowatt-hour for identical computational workloads.

For studios making decisions about cloud rendering, server location, or co-production infrastructure, this regional gradient is a core operational variable. A render job processed in Montreal produces a fraction of the emissions of the same job processed in a fossil-fuel-heavy data centre. Location is an environmental decision, not just a logistical one.