Is the net effect of the internet on the Earth’s environment positive or negative?
The rub of the internet is that it is a collection of data centers filled with computing gear, networks that weave across continents, and a growing amount of battery-powered devices; all of these things need energy to operate. The disturbing part is that the energy consumption of the internet will only grow as the population hits 9 billion in 2050, and all of these people get connected to the internet.
But on the flip side of that energy suck is the idea that the internet can make processes and systems significantly more energy efficient, from transportation to shopping to the electricity network itself. Sustainability wonks call that dematerialization, or replacing atoms with bits. A study
called Climate 2020 found that information and communications technology could reduce greenhouse gas emissions from other sectors of the economy, below business-as-usual growth, by 15 percent.
Other than that seminal report, there’s been a trickle of
research that has reached conclusions along the lines of the notion that buying digital music online is a lot more energy efficient than driving to the store and buying a CD. Data center energy guru Jonathan Koomey, who’s a research fellow at the Steyer-Taylor Center for Energy Policy and Finance and Stanford University, has led a bunch of this research, particularly around how the trend toward
cloud computing has increased the energy efficiency of the internet. The web sharing economy is another much talked about trend that is
indirectly making the use of goods (like cars and apartments) more efficient.
But all of this research is so new, and these issues are so complex, that answering that question — is the internet green or not? — is a very difficult one. After the day long Google event called
How green is the internet? I have a lot more questions than answers.
More good than bad
Throughout the day of Google’s event it became clear that there are significant gaps in knowledge and in research. One of the problems for researchers has been getting access to really detailed industry data. The leaders of the internet industry have only just started to think about these issues, and are slowly warming up to the idea of giving their energy data to third party researchers. Google only
revealed its total electricity use publicly a year and a half ago.
But if you ask that direct question — does the internet have a more positive or negative effect on the environment — there’s some researchers who are already leaning toward the positive camp. Koomey said at the Google event that he thought the preoccupation with the electricity use of the internet was misplaced. The overall system effects are more important than the direct electricity use, said Koomey, noting “Moving bits not atoms can have a major effect on efficiency.”
A research fellow from the Center for Industrial Ecology at the University of Coimbra in Portugal, Vlad Coroama, echoed those thoughts at the Google event: “Sending bits is usually much more efficient than sending atoms.” Carnegie Mellon Professor Scott Matthews, who’s done research showing how e-commerce is more energy efficient than shopping at physical retail outlets, wondered if the introduction of digital goods and e-commerce could have such a large impact that it could reduce carbon emissions and energy use compared to the current physical processes by a factor of ten.
But the complexity of the internet as a system, and how it effects people’s behavior, makes calculating numbers around this very difficult. For example, the process of driving to the store to buy a physical book is clearly more energy intensive than hitting a button to buy a Kindle single on your Kindle. But then you also have to take into account more complex questions like the embodied energy of making the Kindle, and the idea that when consumption is so easy and efficient, you might buy and download exponentially more e-books.
Throughout the day at Google, many noted, like Google’s Urs Hoelzle, senior vice president of technical infrastructure, that problems and opportunities need to be looked at from an entire systems perspective. Looking at just one aspect of the equation doesn’t give the entire picture. A couple people at the event also noted to me that one of the universities or research centers should be acting as more of a research hub for this type of data.
Even some of the data that is out there is difficult to prove. Professor Coroama’s number — that it takes an average of 0.2 kWh/GB to send data across the internet — was head-scratching to some, and several people noted to me that the number seemed high.
Then there’s the soft effects of the internet on the planet that don’t have to do with energy consumption at all. The high level visionary speakers — both Gore and Schmidt — focused more on the internet’s ability to open up access to information and organize people, which could be used for environmental, and climate-fighting, causes.
Gore said that the digital revolution and the explosion of data are some of the most powerful tools that can be used to help solve the climate crisis. It’s hard to quantify such soft effects, but they could still be very powerful.
Keep it that way
The main issue now will be as internet access grows, mobile phones connected to the web proliferate and internet companies build ever more data centers, how does the industry maintain sustainable growth so that the equation doesn’t flip, and so that the internet doesn’t start to have a negative effect on the environment? There’s going to be 9 billion people on the planet by 2050 that could have a handful of connected devices each, and some of them will be spending their lives immersed in digital data 24/7.
That’s exactly why I thought Google’s Summit was so important: As a way to shine a spotlight on the issue, address its complexity, and see how the problem can be solved as the internet grows. Essentially designing the growth of the internet with sustainability in mind. Google can be one of the leaders of this discussion, as can other internet companies like Apple, Microsoft, and Facebook in addition to hardware companies like Cisco, IBM, and Intel.
Going forward, I’d like to see a hub grow at a university or research center that can act as a collection point to draw together this type of research, and also to help validate it. I’d also like to see more mainstream attention on this topic of the intersection of the Internet and the environment. At the Google event, it was invite-only and had about 100 people that had been thinking about these topics for years. This topic is important enough that is needs more mainstream attention and discussion.
