Carbon Constrained Interviews

It is easy to not realise that even our digital activities have a carbon impact.

Sending emails, writing blog articles, checking you video door bell, communicating on Slack or social media all create carbon emissions.

These carbon emissions derive from the energy used to mine the components, build and ship the devices we use – aka ’embodied emissions’ – to  the energy expended when we use them – ‘operational emissions’.

Estimating the Digital Carbon Footprints of Video

One of the bigger culprits of of digital lifestyle, is video.

Video creates a lot of data, and often requires additional hardware and energy to produce. Video is also one of the highest viewed content types, contributing to significant traffic volumes. Social Media video alone tops 70% of mobile network traffic, and is expected to grow another 10% by 2028.

However, when you’re watching a video or using Zoom for calls it can be difficult to remember that, or even care about it when engrossed in the latest thing to hit your feed.

The Digital Carbon Gauge

Digital Carbon Gauge: Conveying the digital carbon emissions of a video call

For this reason, I decided to begin a series of “carbon constrained” interviews with thought leaders in the world of digital carbon.

Initially I thought I’d focus on keeping these interviews short, and bound by a 7 minute window. But then I realised it would be best to relate them directly to carbon emissions.

With this in mind I created the realtime video Carbon Gauge.

In this series of videos I will ask my guests to convey their important messages while keeping within 200g of carbon emissions (which is approximately the same amount as generated from charging 25 smartphones)

Digital Carbon Gauge Methodology

Determining how much carbon is generated during a video call is an incredibly complex affair.

It is somewhat easier than calculating the emissions of stream video (such as Netflix or YouTube), because you don’t have to consider content delivery networks (CDNs), but it is still fraught with assumptions, inaccuracies and unknown factors.

However, just because something is difficult or potentially inaccurate doesn’t mean we shouldn’t try. So try I have.

Primarily, my goal here is to visualise the carbon emissions of video and raise awareness. As such, I am tolerating and acknowledging the fact that the methodology is inaccurate, yet aiming for something that is “indicative” – e.g. gives you a sense of the issue.

Calculating the carbon emissions of video calls

With that in mind, the methodology I used involves the following highly simplified steps:

1. Take the amount of data generated during the video call.
2. Convert the amount of data into energy use.
3. Convert energy use into carbon emissions.

Step 1 is relatively easy to approximate from the video call platform’s documentation.

Step 2 is significantly more complex and involves a lot of assumption and complex methodology. I looked towards a comprehensive piece of work from some of the pioneers of this calculation.

Step 3 seems straightforward, but is yet again fraught with complexity. Once you have an energy value, carbon emissions can be calculated using a carbon intensity factor. However, this factor fluctuates throughout the day, varies from country to country and even office to office. For my gauge I settled on the real-time UK National Grid carbon intensity at the time of recording the interviews.

If you want to know more about digital carbon footprints, then you’re in the right place. I’ve written a great deal on this topic, so pop over to the 7 Articles On Digital Carbon post, or read the seminal report on Exploring Digital Carbon Footprints.

Examples of known limitations

To get a sense of some of the known compromises made during this methodology, consider the following issues:

• My office runs mostly on batter power. That batter is mostly topped up by solar, unless the sun hasn’t shone for a while. In those cases I top it up from the grid at a time where the grid’s intensity is low(ish).
• Sometimes I use my laptop that has been charged at another office.
• This doesn’t include any embodied carbon calculation.
• The people I interview will being using energy from their own provider, which is not taken into consideration
• The video conference platform will be using yet another energy source
• The energy cost of data transmission is taken into consideration in the conversion from data to energy (Step 2), but doesn’t reflect the locality or energy provider points mentioned above.
• Storage of the video is not considered
• Broadcasting (when viewed) is not included.

So is it worth it?

With all those known inaccuracies, let alone the unknown, you might wonder if this activity is worth it.

My opinion is yes, it is. (Well, of course I’d say that, right?)

Seriously though, my goal is to get you to think about carbon emissions during digital activities in the same way you might think about it when turning off lights, or taking the bus, or choosing a vegetable based menu option.

It might seem a small thing to do, but if we all reduce our video use just a little bit each time we can have a cumulatively much larger impact.

So go ahead, watch the videos, and imagine the gauge when you’re pointlessly waiting in a team video call for something to happen. Would it hurt you to turn of the video until everyone was there?

Open to comments and feedback. Let me know your thoughts.

Carbon Constrained Interview with Ismael Velasco

Here is the first in the series. I hope you enjoy it.