After a career working in the information and communication technology (ICT) sector, I was surprised by press reports suggesting that ICT might be contributing to climate change. For example, one article in the New Scientist suggested a typical internet server produced similar emissions to driving an SUV, an influential book by written by Mike Berners Lee calculated that frequent mobile phone use could cause as many carbon emissions a year as four domestic flights, the Shift Project predicted that video streaming services like Netflix have an enormous carbon footprint, and one technical report suggested ICT might consume up to 50% of electricity production by 2030.
This topic came up in a STAP (WinACC’s scientific and technical panel) meeting, but none of us had a feel for the scale of ICTs carbon emissions contribution. Since I had a background in ICT I volunteered to investigate.
STAP reports aim to provide an informed and reasonably digestible summary of climate change related information. Providing references is especially important so that others can form their own opinions based on the same information. For those of us with less of a research background/ability this usually involves a lot of time finding and managing sources of information and wrestling what we have found into a coherent structure!
The ideal STAP report would be a link to an article which addresses the subject. Publications like the New Scientist or topical web sites like CarbonBrief have excellent articles summarizing current topics, but these are often focussed on a single aspect that has been in the news. Academic review pieces are often narrowly focussed and aimed at other researchers. Hence, STAP sometimes produces its own report. It would be interesting to know if similar groups are also writing their own reports.
My report, written with help from other STAP members, looked at some of the claims mentioned above and related articles and research papers. Like many attempts to estimate carbon emissions from a given source, researchers use many approximations and extrapolations as well as varying definitions about what emissions count as ICT-related. Thus any conclusions have a high degree of uncertainty.
The consensus is that the carbon footprint of ICT is significant. It is estimated to be similar to aviation at around 2% of worldwide carbon emissions. However, the runaway growth predicted by some seems to have been held in check and ICT-related emissions have remained largely stable for the last 10 years. This is because the expected exponential growth in demand for IT services has been balanced by significant technological advances, especially in electronic technology, mobile communications and cloud computing.
New applications like the ‘internet of things’ and Bitcoin-related technology along with lack of regulation of the internet may yet see significant growth in electricity use by ICT. Existing applications can also use more resources – for example as video quality improves. Some of this growth should be offset by the reduced carbon intensity of electricity as the proportion of electricity from renewable sources increases. ICT is also a key part of much energy saving technology such as video conferencing.
Personally, I was surprised that ICT was a significant contributor to carbon emissions, and it seems that the alarming reports raised valid concerns. Maybe it is worth remembering that our brains use about 20% of the body’s energy. The main message seems to be that people should be aware that ICT has a carbon footprint, so it is worth turning off devices and being aware of their power consumption – for example large TV monitors.
However, using ICT can often improve efficiency – especially in the area of eliminating unnecessary journeys through video conferencing, using GPS to avoid getting lost etc. ICT can be used to enhance life such as interacting with others or organizing outdoor activities. At the government level initiatives such as device efficiency ratings have a role to play and, since ICT all runs on electricity, support for renewables.
I have realized it is important to know when to stop digging. There are many areas not covered in the report I would personally find interesting – investigating the physical limits of energy use in communication, how to reduce spam traffic, the role of software technology – but those are for another day.
Robin Speed
STAP Member, WinACC