Measuring our carbon output

Reading Time: 2 minutes

The carbon costs of “digital” are not well understood.

How many emails have you received with “Consider the environment before printing this email” in the signature?

How many have you received with “Consider the environment before hitting Reply All to 26 random people” and adding “Thanks”?

For many people, “virtual” means “not real”.

As in, almost literally, non-existent.

Terms like “cloud” don’t help.

The reality is, the internet is the biggest machine humanity has ever built. It circles the entire planet.

A data centre. A dark corridor leads through racks of servers, lights blinking,. A lone figure in the distance pushes a trolley.
A data centre

It’s made of hundreds of thousands of miles of cables. Billions of devices.

Switches. Routers. Modems. Phones. Masts. Computers. Fridges. Cars.

Light bulbs.

(Yeah, you read that right.)

And thousands of data centres that gobble enormous amounts of electricity.

(Data centres also consume huge quantities of water, as does the rest of the tech sector. But let’s just stick with energy consumption for now, to keep things simple.)

Many of these devices, and a lot of this infrastructure, are still powered by electricity generated by burning fossil fuels. That has a knock-on effect in the amount of CO2 we release into the atmosphere.

(A short, text-only tweet results in the release of about 0.2g of carbon.)

Our contribution to the overall effect of this is minuscule. But so is almost every organisation’s, taken individually.

Our web services are hosted on Azure, which claims to be carbon neutral. But much of that is through offsetting, and it’s not clear how effective that is.

So, we’re committed to reducing the carbon impact of all our sites as much as we possibly can.

Even if we can get them powered by 100% renewables, this will help, because it will free up energy to heat people’s homes. Which, let’s be honest, is somewhat more important than running our website.

Any designer will tell you that the first step to a solution to *any* problem is to understand the problem. So we’ve started to gather data. And, as ever, we want to #DesignInTheOpen

So we’ve created a dashboard to illustrate the CO2 our most popular web services create. This is pretty much realtime data. Mobile-friendly version follows. Now we can set targets to reduce it.

Mobile-friendly version:

Graphs are all to the same scale. The comparisons are direct. As we improve things, I’ll update CO2/pv values.

If you care about any of this – and you should care about all of it – follow @gerrymcgovern

 | Website

I'm a service designer in Scottish Enterprise's unsurprisingly-named service design team. I've been a content designer, editor, UX designer and giant haystacks developer on the web for (gulp) over 25 years.

4 Replies to “Measuring our carbon output”

  1. Why does SE’s website generate more CO2 emissions that FBS? I initially wondered if it was directly proportional to traffic, but it doesn’t appear to be so. Is it an ‘accounting’ glitch? Or does FBS have features to reduce carbon emissions?

    1. It’s proportionate to page views, Craig – the graphs show users (blue bars). The SE site tends to generate more pageviews per user than FBS.

      But also FBS was built with performance as a priority. Crucially, it has very few images, and most of them are small – they’re logos – and served as SVG rather than jpeg or png.

      1. Interesting. SE generated 1.62g Co2 per page view and FBS 0.40, so the image thing must make quite a saving

        1. Yes, it makes a huge difference.

          Look at

          How many pictures do you see?

          What do you value? Fast? Or pretty?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.