Over the past decade, the roster of familiar brands committing to 100% renewable energy has grown steadily. Facebook, HP, and Visa are among the latest to advance plans to run on 100% renewables, joining the likes of Amazon, Apple, and Microsoft.
Hundreds of cities and states are chasing similar targets, too. At the federal level, the recently announced infrastructure bill includes a long-anticipated mandate to shift all federal buildings to 100% renewables –– a potentially enormous undertaking.
Yet today, almost all of these commitments face a common challenge: it’s nearly impossible to source renewable energy 24 hours a day, 7 days a week from the grid. Even in regions with large supplies of renewables, output varies as clouds roll in or the wind dies down.
When you flip a switch, it’s impossible to know the exact mix of energy flowing at that moment. The grid is engineered to seamlessly match electricity supply — typically generated using a mix of natural gas, coal, nuclear, wind, solar, and hydropower — with demand, second by second.
For companies, governments, and others aiming to go 100% renewable, the intermittency of renewables makes 24×7 matching curiously complex and difficult to achieve.
When 100% isn’t really 100%
Aside from rare cases where a company can isolate a facility’s energy supplies, no large energy buyer has so far been able to crack this 24 x 7 problem. Instead, to balance out the non-renewable share of their power, energy buyers rely on a variety of industry-standard techniques, such as offsets via renewable energy credits.
Credits let a buyer match its power consumption with renewables. By paying a generator for renewable energy that wouldn’t have been otherwise produced, the buyer can own the renewable attribute of that generation. The buyer can then offset their carbon-rich power consumption by securing an equal volume of carbon-free power.
These tradable credits have driven enormous additions of renewable capacity and production to the grid. But they don’t achieve many buyers’ ultimate goal of truly carbon-free power 24 x 7: when renewables are not available, carbon is still emitted, which must then be offset.
This challenge reveals some of the thorny barriers facing efforts to decarbonize. The upshot is that, to date, no big energy user has solved the puzzle of these basic grid constraints. Given the extent of these commitments, and the nature of the challenge, that’s why it’s worth taking a quick look at the first high profile effort to do so.
Google’s bid to solve 24 x 7
Google made waves recently when it announced a commitment to go to true 24 x 7 carbon-free electricity by 2030, on a worldwide basis for all its operations.
From Gmail and YouTube to the ads that fund much of the web, Google powers a huge part of our digital lives. And since 2017, the internet goliath has matched 100% of its power demand with renewables, using a mix of offsets (as described above) and by building its own renewable energy capacity.
For years, Google has ranked among the planet’s top corporate buyer of renewables. In 2019, Google gulped down 12.2 terawatt hours of power globally, roughly twice San Francisco’s demand.
This scale makes the challenge of buying 100% renewables more difficult still. Google operates on every continent, save Antarctica. And every region has varying supplies of power. In the tiny city-state of Singapore, for example, the local grid runs around 3% green, not enough to meet Google’s goals for its data center there.
Intermittency poses a related challenge. In wind-whipped Oklahoma, Google’s Midwest data centers can tap wind power 96% of the time. But when the wind wanes, the fill-in is not renewable.
Averaged all together — around the world and around the clock — Google estimates that in 2019, 61% of its global hourly electricity needs were met directly by carbon-free sources. So to move to 24 x 7 carbon-free energy, Google is chasing a two-pronged strategy high-tech approach, by exploring the potential of shifting the time and/or location of work.
Google’s hourly carbon clocks
Average annual carbon-free energy performance (numbers in blue) at select Google data centers in 2019. Carbon clocks show change by hour for one day in 2019.
By gathering details of hour-by-hour usage across its global operations (see diagram), Google documented the variance of its power supply by time and region. The data is helping Google identify ways to shift select non-urgent tasks — like updating street view photos for its maps — to be processed when the electricity is greenest.
A second option emerged, too: shifting computing tasks not just by time, but by geography, away from a region where renewables were waning, to one where they’re rising. So at dusk on the West Coast, as solar ebbs, computing load might be shifted to data centers in the Midwest, when nighttime wind output usually rises.
This peek under the hood at Google’s efforts offers a case study in how, for big energy buyers, the move to 24 x 7 carbon-free operations is driving technology investment to help dynamically optimize energy use.
As the pandemic recedes, energy demand is on track to rebound, just when more companies and consumers are exploring how to balance the priorities of sustainability, costs, and reliability. And as these dynamics gain momentum, technology is emerging as a vital ingredient to better understand how we use energy and how to seamlessly balance options.