What trying to travel low carbon tells us about emissions in games

What trying to travel low carbon tells us about emissions in games
The XPT Countrylink, image CC BY-NC-SA 2.0: John Ward

I spent January interstate, visiting friends, on a sort of open-ended trip, no firm return date. When it came time to plan my trip back from Melbourne to Sydney I wanted to see what it was like to do the climate-friendly thing and avoid flying home. I also wanted to see how much CO2 I would save, and just how hard (or easy) it might be.

Australia is a big country – but the major cities are along a fairly narrow stretch of the east coast. You might think this is a prime candidate for high speed rail, but in fact the rail links between major cities are have suffered from serious underinvestment (perhaps even a bit of neglect), with the only thing even close to high-speed rail a few 'tilting trains' on certain routes in Queensland that top out at about 160 km/h (not even tru high speed rail).

The history of high speed rail in Australia is interesting, but it’s largely about one plan being proposed, failing to make progress, so another is made, it fails to make progress, and so on, rather than actually anything getting built. For GTG readers outside Australia, one of the major relevant factors in this is that the flight from Melbourne-to-Sydney (only 713 km – hardly worth flying you might think) now has the dubious distinction of being the 5th busiest in the world – and from memory, at some point pre-covid it might have even been 2nd or 3rd? Whatever its ranking, it’s absurdly busy and as a result, its creating absurd flight emissions. The other four busiest routes are also distinct from the Sydney to Melbourne route, all being in South East Asia (more dense populations), involve crossing bodies of water (Seoul/Jeju Island; Tokyo/Sapporo; Tokyo/Fukuoka) or longer and crossing more mountainous terrain (Hanoi/Ho Chi Minh City). Now that France has led the way in banning short haul flights (where fast rail alternatives exist, at least), It seems almost inevitable that some sort of rail is going to be necessary in most country’s net zero future – but Australia in particular is a long, long way off.

The three main options for getting from Sydney to Melbourne then are by air, by car, and by rail. Let’s compare the three, see what sort of emissions we can expect.

According to this tool by The Guardian, the total CO2 from a return trip by air between Sydney and Melbourne is about 165 kg CO2 (and that figure more than doubles if you include the effects of radiative forcing – which I touched on a little bit last year). I flew down to Melbourne from Sydney, so in my case I was only avoiding a single direction of travel, so we should halve that figure, giving us 82.5 kg of CO2. We can also make an estimate ourselves using figures from OurWorldInData which provides the gCO2e figure for various modes of transportation per km. A straight line 713km trip by plane using this figure gives a fair bit higher result: 181.8 kg CO2e. I suspect that is probably the more realistic figure (and might be accounting for more of those pesky radiative forcing effects).

If I did the same journey via car, which Google maps says is 876 km by road (using the same start and end points, just for parity’s sake – from Southern Cross Station in Melbourne to Central station in Sydney) what would my emissions have been? My last car, a 2004 Toyota Prius, allegedly had a highway fuel efficiency of about 3.9 L/100km – but it was old and a bit worn out, so lets assume 4 L per 100km – which works out to needing about 35L for the whole trip. I know from experience that it uses quite a bit more than that, so lets round it up to more like 45 L to account for non-perfect conditions, using aircon, etc. The UK Govt’s list of GHG conversion factors gives a figure of 2.16 kg CO2e per L of petrol combusted (assuming an average biofuel blend – your typical E10 ethanol product, I presume) which gives about 97.2 kg CO2e which sounds about right. As noted in the summary on the YourWorldInData page:

If you’re travelling moderate distances (<1000 kilometers or a domestic flight within the UK), then flying has a higher carbon footprint than a medium-sized car. If the distance is longer (>1000 kilometers or an international flight), then flying would actually have a slightly lower carbon footprint per kilometer than driving alone over the same distance.

As Melbourne to Sydney is still under that 1000km rage this fits with our results so far then. Still, not everyone doing the trip is in a car as economical as the humble Toyota Prius: the OurWorldInData figures suggest using a figure of 192 g per km for a single driver trip in a more typical vehicle, or 168.19 kg CO2e for the trip. That’s hardly much better than flying.

There don’t seem to be any obvious data sources for Australian diesel train emissions (in fact they don’t seem to be monitored at all???) but the UK office of rail and road provides a figure of 47.5 g CO2e per passenger km (and the OurWorldInData figure for UK National Rail gives the almost identical figure of 41 g). I can’t find a figure for the exact distance of the rail route, but Google maps shows its a bit of a deviation from the straight line path of the flight, and more like the route the highway takes. So lets round it up to the same distance as the road then, and make it 876km – which gives a figure of 41.61 kg CO2e for the entire journey. That’s slightly affected by how full the train is, how many services run, etc. but it’ll do for our back of the envelope numbers.

So the three travel modes have the following footprints (using the higher figure for each):

Air – 181.8 kg CO2e

Road – 168.19 kg CO2e

Rail – 41.61 kg CO2e

So at minimum I managed to save 126kg CO2e! That’s reasonably substantial, and makes me really think really quite seriously about my methods of travel between the two cities in the future. However, there’s still a lingering issue. The train journey actually sucked – like it was seriously, so bad. Easily one of the worst journeys I have taken in years. This makes it substantially harder to choose in future – even if I know it’s going to save me 120 kgs of CO2e!

What made it so hard? Firstly, sitting down for 12 hours at a stretch anywhere is hard. We know that it’s not good for your health, and there’s not really much you can do to stretch your legs, except go for a walk up and down the length of the train or visit the buffet cart (which was very, very mid). It was physically quite gruelling, and then once I arrived in Sydney I had to get another train to actually get home from Central Station (which should have only taken an hour, but instead took me 3, including a train and a bus, thanks to network issues). I left my friends apartment to go to Southern Cross station at 7.15am and didn’t get home until 11pm. For most people, that’s beyond inconvenient, it’s practically an impossible choice. Imagine travelling with young children for that long.

On the plus side for the train, it is very cheap – it was $72 for an economy ticket, whereas the cheapest flight I could find was closer to $200. But because I was on the train for so long I ended up spending more on (pretty average, lets be honest) food and meals than I would have if I had flown. So that started eating into the value of the ticket. I did, however, get to get some work and reading done – but the train does rattle around a lot, it’s definitely not a smooth trip. And worse, most of the time is spent without phone reception (and thus, without internet). Its probably in all honesty easier to get work done flying than it is on the train. Driving yourself, of course, also doesn’t let you get any work done, but it does have the added convenience of leaving you all to yourself, to listen to music, podcasts, etc. in some comfort, and you don’t have to deal with Sydney trains.

So on the whole, even knowing the different CO2 footprints of each mode, the major deciding factor still comes down to time – the entire flight takes about 90mins, plus another hour or so for getting through security, checking bags, etc. so perhaps 3 hours, in the most ideal situation, plus a bit more for travel to/from airports. The drive from Sydney to Melbourne usually takes about 9 hours, and can take you door to door, but you can’t really multitask. The train takes about 11-12 hours, you might be able to do some work, you might not, and then you still have to get home.

Still – if you travel by train you get to see a couple of wind farms. Including this turbine with a broken blade. Exciting!

A picture from the train window, grass and trees in the foreground, a wind farm in the distance, with five turbines in a row, one with a blade missing from the top
Is it burnt? Is this the turbine that spectacularly caught fire a few months ago? 

My conclusion from this experience then is this: we have to start making the low-emissions options much, much easier to choose, make them more appealing than the alternatives. This applies to travel, it applies to consumer goods, and it applies to games. Low-emissions options in every area need to be made so attractive, so ubiquitous that it doesn’t even register as a choice: it needs to be the default option. In the past, orthodoxy would have said the way to do this was through a carbon tax or similar. The world seems to have moved beyond such simplistic market thinking, recognising the limitations and inequities of such an approach. Instead, I think (I hope?) we are able to take a more holistic approach – perhaps informed by UX type research, social science research, and all sorts of other disciplines that contribute to understanding the complexity of human decision making.

The truth is that almost no one makes everyday choices based solely around what the low-emissions choice is – and its not reasonable to hope that everyone could. Those that do have shown exactly how hard it is. Giving up flying in a world accustomed to it has real costs. But there shouldn’t be – especially after what we have learned from the pandemic experience.

So what does a program of making low-emissions options in games appealing look like? I think that’s the million dollar question. We’ve got a few glimmers of it already: the Xbox carbon aware update is a good example, favouring and designing for super low energy consumption in gaming devices is another, slowing down the hardware upgrade cycle and disclaim photorealism is perhaps the final frontier. There will be others – some no one has even thought of or considered yet.

Feel free to reach out if you want some ideas on what this might look like in your next game, in your business or area.


Thanks for reading Greening the Games Industry. Apologies for the near-radio silence the past two weeks, there has been A Lot happening in my life of late and its just thrown the schedule out. Hopefully back on track.