Skip to main content

The blockchain isn't doomed to become an energy hog

Much ado has been made of blockchain technology's role as an ally of corporate sustainability teams and distributed energy advocates.

Unilever, Nestle and Walmart are testing food safety and provenance applications, Siemens and IBM are developing services, and utilities ranging from Engie to Centrica are backing a much-ballyhooed framework to accelerate grid-related uses.

The organization behind the latter work is Energy Web Foundation (EWF), a partnership between Rocky Mountain Institute and Grid Singularity. On Monday it revealed it has snagged $14 million in funding and more than a dozen more strategic partners, including Duke Energy, Exelon and PG&E, with total backing exceeding $17 million. That's just a drop in the bucket when you consider the money being allocated to blockchain experiments.

For those companies and many others for that matter, the blockchain — the digital verification and transaction system originally created for bitcoin but now promoted as an enabler for dozens of other applications — has become something akin to the Holy Grail.

In the best-case scenarios touted by technologists across the corporate sustainability world, the blockchain promises to scale data collection and transparency dramatically across supply chains or to enable completely new peer-to-peer energy trading markets. Organizations such as the ones listed above are investing in development projects and startups to test out those theories.

The problem is that, at the moment, only about a fifth of the electricity used in the world’s data centers comes from renewable sources, and that's not good enough.
But critics are urging early adopters to step back and consider the side effects — especially where these potentially vast digital networks and services will get their power.

What's got tongues wagging is a blizzard of dire forecasts centered on the amount of power needed to run bitcoin digital currency networks and to validate the transactions that take place. These estimates are jaw-dropping (and widely fluctuating) — anywhere from 1 terawatt-hour to 32 TWh annually, according to one Energy Web Foundation (EWF) blog. In everyday terms, the bigger number is equal to the amount of power used by Denmark in one year.

What that means for blockchain — which is not the same thing as bitcoin — is still a wildcard. But you get the point: We need to pay more attention, collectively.

Know thy data center

Some concern about blockchain energy consumption relates to the ongoing concern over the world's vast and growing vaster reliance on power-guzzling computer servers and data transmission networks. "As our digital world continues to rapidly grow and become more complex, the demand for computing power and data management will continue to rise, and that will require much more energy," said Gary Cook, a senior corporate campaigner at Greenpeace who leads its research on the information technology industry.

Data center operators face a twofold challenge: building more efficiency into how information is processed (through design changes) and ensuring that a growing number of their locations draw from renewable power sources such as solar, wind or hydroelectricity rather than fossil fuels.

"The problem is that, at the moment, only about a fifth of the electricity used in the world’s data centers comes from renewable sources, and that's not good enough," Cook said. "The internet has the potential to serve as a critical foundation for a sustainable economy, but its expansion needs to be powered by clean energy sources that help, not hinder, the crucial challenge of tackling climate change."

With that in mind, businesses investing in blockchain experiments and applications should evaluate the data centers where they choose to run those projects more closely, Cook said, as there's a big difference in how some of the best-known providers perform when it comes to adopting clean power.

For example, one of the biggest hosts of blockchain services for corporate applications, IBM, received only average grades in Greenpeace's January 2017 "Clicking Clean" report when it comes to renewable energy procurement, efficiency and transparency about its policies.

IBM declined to comment about its blockchain energy strategy for this article.

It comes down to prioritizing efficiency

Stanford University researcher Jonathan Koomey, following the bitcoin power consumption debate closely to understand its broader implications, said it's becoming more obvious that the energy efficiency of a given blockchain network will be closely linked to how the applications handle the process of validating information. 

"There are more or less efficient ways to structure this," he said. "There are many different ways to accomplish this."

One question that companies investing in the technology should evaluate is whether using blockchain to manage information — such as validating that a product did, indeed, come from a responsible manufacturer — uses more or less power than the current method of doing so. "Why is doing it this way better than the way they were doing it in the past?" Koomey noted.

There are more or less efficient ways to structure this. There are many different ways to accomplish this.
When it comes to energy-related applications, EWF is studying this closely and prioritizing validation methods that have a "lighter energy footprint" than the one typically used for bitcoin, said Sam Hartnett, an associate with RMI, in a recent blog post.

For its test network, called Tobalaba, the foundation will use a method called Proof-of-Authority (PoA). That means "a trusted set of authorities" will assume the validation role. "Such PoA networks are well suited to regulated industries where entities responsible for maintaining the network (authorities) need to be known, rather than remain anonymous as in mining-based chains like bitcoin and ethereum," Hartnett wrote.

How much energy will that take? EWF estimates that each "node" used to authorize transactions can run on about 78 watts of energy — about the same amount of power needed for an incandescent light bulb. As the network will use 20 nodes, "the upper boundary total energy demand is approximately 1.5 kW — roughly equivalent to a microwave oven."

The power consumption will grow on a linear basis, EWF figures, but the organization also will leverage improvements in hardware efficiency or in shared server applications to manage the "energy burden" as these applications mature, according to Hartnett.

Given that many corporate blockchain experiments are still in the decidedly experimental phase, thoughtful, energy-conscious development should definitely be part of the evaluation equation.

More on this topic