Designing IT workloads to reduce carbon intensity improves a company’s triple bottom line.
This practice reduces environmental impact, improves profitability, and increases job satisfaction for software developers and system architects.
In fact, recent research by Salesforce found that 75% of technologists want to develop software applications that do less harm to the environment but are uncertain about how to get started.
One approach is to develop software that manages the computational workload to use more renewable energy sources, thereby reducing carbon intensity. To understand how this can be done, we need to briefly review the energy mix provided by utility companies.
Utility companies buy energy to meet varying daytime demands. This energy can be divided into dispatchable and intermittent sources.
Dispatchable sources are natural gas, coal, nuclear, and reservoir-based hydroelectric power. These resources are highly reliable because they provide a constant supply of energy. They are referred to as baseload resources.
Intermittent sources, like wind and solar, generate electricity when the wind blows and the sun shines.
To ensure utility companies have enough energy, they buy dispatchable energy from power plants operating at a minimum threshold or baseload, complemented by marginal power plants that respond quickly to the sudden need for increased power.
Globally, wind and solar power are rapidly being added to the energy mix. Since this renewable energy is being added on top of the stable baseload supply of dispatchable energy, there are times when the supply of renewable energy exceeds demand and renewable energy is discarded or “curtailed.” Curtailment isn’t new. It has always existed because electricity generation must equal electricity usage for the grid to work.
Wasting renewable energy isn’t good for the planet. And since renewable energy is generally less expensive than fossil-fueled energy, reducing waste could also improve the utility company’s profitability.
IT specialists can shift the computational workload to locations and times that increase the use of renewable energy and reduce the disposal of excess renewable energy.
For example, depending on the season, a company can transfer the computing workload to a different hemisphere for more sunlight hours. Or, given advances in weather forecasting, they can schedule high workloads later in the day when it’s sunnier or at night when it is windier.
Here are three examples of green IT in practice:
- Google’s carbon-intelligent platform uses day-ahead predictions of how heavily a given grid will use carbon-intensive energy to shift their computing to favor regions with more available renewable electricity.
- Starting with Windows 11, version 22H2, Microsoft’s Windows Update will schedule installations when a higher proportion of electricity is available from lower-carbon sources.
- AWS’s new customer carbon footprint tool helps its customers measure the carbon emissions associated with using AWS services. The tool enables them to analyze emissions over time and forecast how emissions will change as Amazon moves toward powering operations with 100% renewable energy.
