The dirty mobile telecommuters

Telecommuting can be great. It can also be loaded with many lies, one of which I’m going to partially debunk here: environmental benefits.

Here’s typical imagery of a telecommuter, the kind that can be found from a plethora of presentations and articles about telecommuting (this particular image coming from here):

There are so many things wrong with that picture that it’s hard to know where to start. For example, why does the guy wear pants & a tie? He’s at the beach with nobody else in sight; I can imagine one or two more comfortable outfits. Why doesn’t he have sunglasses on a sunny day? Why is he pretending to work on the laptop, because you wouldn’t be able to do that in bright sunlight anyway? But for the sake of the argument, let’s assume he can get actual work done as a mobile telecommuter at the beach.

What isn’t apparent from the picture is that he’s also likely polluting the environment. A lot. And it doesn’t have to be that way.

This matters, because environmental benefits have been touted as one important benefit of telecommuting at least since 1994 [3]; yet it turns out that the carbon footprint of mobile data is 34g of CO2 per transmitted megabyte [1]. But just how much is that?

Let me use myself as an example. We have a beach a short bike ride away, so let’s say I decide to work from there instead of taking the 15km train ride to the city. The carbon emissions from electric rail are around 30-100g of CO2 per passenger kilometer [4, 5]; using an estimate of 75g/km the emissions from my normal commute would therefore be about 2,250g, or a little over 2kg of CO2.

I’ve logged my normal network usage during a typical workday, and it’s around 750-1,500 MB/day; say I use a gigabyte per day. To get a conservative estimate, let’s also say the wireless network emissions are “only” 30g/MB instead of the 34g from the above-mentioned paper. Here’s what it looks like, with a commute by car (a regular car with CO2 emissions of 200g/km) thrown in for good measure:

Yes there are a bunch of disclaimers here; 1) there will be emissions from data used at work, too, but the mobile network estimate is only of the mobile leg – not the fixed part, which essentially remains the same so I’ve excluded that, 2) this does not include building maintenance/HVAC/possibly increased electricity usage emissions at the office, 3) telecommuting from a fixed-Internet connection produces a vastly better result and 4) many other things – but I still think it serves as a useful way to highlight the fact that your mobile data is not actually very clean.

So next time you skip the trip to the office and instead work from a beach, a park or a cafe and let your 3G/4G connection scream away freely, keep in mind that you’re not saving the environment. You’re destroying it.

There has recently been considerable chatter about the cleanliness of our ever-growing ICT infrastructure, whether it is NY Times [7] or Greenpeace [2] criticizing our cloud infrastructure (and rightly so) or the service providers highlighting every effort (again, rightly so) they make to improve things, from Apple’s solar electricity generation to HP’s net zero energy data centers [6]. The fact, however, remains that pretty much nobody in the industry has cleaned up their business even close to the required levels – especially since demateralization and virtualization enabled by ICT is used as an environmental argument, I think our industry is morally obligated to lead the change into renewables. This applies to all polluters-by-proxy, from global service providers to network providers and everyone in between.

Let me emphasize that I am in no way, shape or form against telecommuting or even mobile telecommuting for that matter. Quite the opposite. I merely want to highlight that until the telecommunication industry cleans up their (our?) act, you should be aware that that gigabyte of data you so easily chew through on your smartphone is creating emissions that can be equivalent to driving hundreds of kilometers.

Time for the industry to clean up their act or stop pretending it’s helping the environment. I would prefer the former, because done properly, we could really become a net-positive-impact industry in an environmental sense as well.


[1] Schoenen et al: Green communications by demand shaping and user-in-the-loop tariff-based control

[2] Greenpeace: Clean our Cloud

[3] Richard Johnson: Ten Advantages to Telecommuting

[4] Transport Direct

[5] NSW Transport City Rail: Carbon emissions

[6] HP: HP Unveils Architecture for First Net Zero Energy Data Center

[7] NY Times: Power, Pollution and the Internet

This entry was posted in Energy, Environment, mobile and tagged , , , , , , , . Bookmark the permalink.

6 Responses to The dirty mobile telecommuters

  1. Obakesan says:

    very interesting examination


  2. Pellicle says:

    I haven’t run your numbers myself, but I have a few ‘back of the envelope’ calculations which seems to make this incorrect. For instance, if I drive my car 100Km I’ll need 10L of fuel. So just based on costs of energy for a 100Km journey, so is data on phone’s so sponsored that a gig of data should cost $15? I mean there must be some piggy backing going on there which is ignored? Are those figures assuming that there aren’t thousands of txn’s being carried as well as the specific gig you are tracking?

  3. sim says:

    Ok, I’m not sure I follow your reasoning there; the energy (costs) to move a car vs the energy for the mobile networks? But first of all, the average efficiency of an internal combustion engine is around 18%, while even the worst coal power plants have an efficiency roughly double that. Also keep in mind that the industry wholesale rates for energy are far lower than what we as consumers pay.

    But if your point was in terms of the emissions delta of whether you use that gigabyte or not, i.e. what are the additional emissions, in that you’d be right that you wouldn’t get the above figures as most of the equipment required for your transmission would be running anyway. So what I use is an average carbon footprint per MB figure. But that’s the case with things like trains as well (though less so with cars) – the marginal “cost” is likely to be relatively low.

  4. Andrew says:

    Interesting. What is the best in class for cellular network operation? In general, we in Australia make good use of solar for our networks (e.g. but we can do better.

    That said, I think it is difficult to do robust environmental calculus. There are so many factors at play that coming up with a definitive “X is better than Y” is elusive unless there are multiple orders of magnitude in the difference. For example, for telecommuting, if it were to be considered as a permanent alternative to going to work in a city office, then we could also count the reduction in future office space construction to accommodate more city workers. In that case, I would expect telecommuting to come ahead of city-based working.

  5. James says:

    Thanks for the figures! I’ve had a hard job tracking down good figures on CO2e for mobile data, thank you for pointing me to that IEEE report (sadly, I don’t have IEEE access, but will try my engineer friends).

    For comparison, wired data transmission clocks in at 5kwh/GB to 13kwh/GB, depending on who you believe. So about 9g. That’s data center + transmission, I believe. [ref:

    Another take (from a corporate think-tank) is that “internet” now is 830MT, and will be 1.4GT CO2e by 2020 [
    At the same time, ICT is meant to shave 8GT off the global emissions total, presumably mostly by webex and cloud computing.

    Greentouch are doing a bunch to reduce the 4g footprint, through a whole bunch of research projects. More efficient 4g base-stations, spread-spectrum doohickeys, black magic, etc. More here:

    My own interest is from the design end – trying to make sure that websites server super-lean pages to mobile customers especially, to save time, data costs and scary amounts of CO2e. Mobile first design helps []

    • sim says:

      Thanks for the comment & the links James! Very interesting stuff.

      Glad to hear someone’s paying attention to keeping websites lean 😉 It seems to me that much of the “mainstream media” sites in particular have become an ungodly mess loaded with so much stuff it degrades even the desktop experience and is just a waste.

Leave a Reply

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