3 ms·

> On an ecological level first of all, whilst digital solutions may be less energy-consuming than paper, they’re by no means energy-free. A short email sent and

> On an ecological level first of all, whilst digital solutions may be less energy-consuming than paper, they’re by no means energy-free. A short email sent and received on a phone produces 0.2 grams of carbon dioxide equivalent (CO2e), while an email with an image or attachment – for instance, a PDF ticket - produces 50 grams of CO2e, according to the great authority on carbon footprint measurement, Professor Mike Berners-Lee. Ok. I need to see the math on this, or else my phone and computer is capable of some sort of physics defying magic. I can send an email with a large attachment on my modest "homelab" a few raspberry pis in, lets be absurdly generous, 10 seconds. So 10 seconds of raspberry pi time, maxed out, 270 watt seconds (probably close to 1/10th that since it isn't maxed out). Reading it on my phone, takes a similar amount of power, realistically less. My phone is perfectly happy to charge off of 2.5 watts so lets say that receiving it takes another 10 seconds. So, being absurdly generous I'm using 295 watt-seconds. If I convert that to kWh I get 0.00008194444444 KWh. I get all my electricity from hydro, but lets say that I was down the road from a coal plant. A coal plant emits 800g CO2 per KWh. That means that the devices that I used to send and receive a large email consumed a whopping 0.06555555556 grams of CO2. I understand that there will be server farms handling my email in the real world, but those server farms are handling billions of other emails, and live and die on power costs, so I don't really think that my email is sucking down power there? Where in the world is it 50g CO2 to send an email (or 62Wh if we convert it back)? What in the world makes you think that a paper ticket, which is literally several grams of carbon and oxygen, has less of an impact? I'm willing to hear that there are people impacted by technology adoption. I'm more than willing to learn about carbon impact. But please don't put together an argument that anyone with any knowledge about the domain can see right through in a blink.

Dalewyn 2y ago
>Where in the world is it 50g CO2 to send an email (or 62Wh if we convert it back)? What in the world makes you think that a paper ticket, which is literally several grams of carbon and oxygen, has less of an impact? I agree with you that the provided example is stupid, but I think there is an argument to be made that going digital can lead to more waste. Take for example digital signage. All those signs, billboards, advertisements, and more that used to be just cardstock paper in ye olde days are now bright, huge LCD screens. I don't want to be the guy signing off on the electric bill, nor consider the geopolitiecological side of where the screens came from.
jiggawatts 2y ago
These "estimates" are often conveniently cherry-picked to support some agenda, on both sides of every political debate. I've dug into a few like these and found that maybe they were valid 20 years ago, and nobody bothered to update the data point. Why? Because that would be inconvenient to their agenda. One that particularly annoys me is some right-wing pundits that keep insisting that solar panels can't recoup the cost of the energy invested into making them. That was true for the panels used to power portable calculators back in the 1980s! Some Ronald Reagan voters got this stuck in their brains and now they are unable to dislodge this factoid without surgical intervention. If you stop and think about it for ten seconds, it would be obviously absurd that there are huge industrial-scale solar power plants springing up by the hundreds globally if they were making a financial loss by selling less power than it took to make the panels! Not to mention that in all forms of power generation (fossil, nuclear, or renewable) the Energy Returned on Energy Invested (EREI) is a key metric that industrialists have been seeking to optimise for over a century now.
mckn1ght 2y ago
I love doing these exercises and would also love to see the math behind CO2e if anyone could provide a helpful link. I’m also thinking about the carbon emissions that go into getting that piece of paper in your hand. The lost carbon capturing capacity of the felled tree, the machinery used to fell, transport and process it, the production of the chemicals used to process the pulp, and finally moving around the finished product. And probably quite a few steps I am missing. Maybe the cars everyone drives to work at the paper plant each day? I’m sure the amortized amount is still small for a tiny piece of paper, but it’s not nothing.
- dghlsakjg 2y ago
  Don't forget that the piece of paper get used for 5 minutes while you walk to the ticket checker, at which point, the entire process is reversed with a janitor, garbage truck, dump/recycling process.
qiqitori 2y ago
My mom heard this statistic somewhere and now doesn't send email anymore, just WhatsApps :/
xg15 2y ago
> I understand that there will be server farms handling my email in the real world, but those server farms are handling billions of other emails, and live and die on power costs, so I don't really think that my email is sucking down power there? Of course it does? The handful of CPU cycles the server spends on it draws power, so will the tiny amount of additional heat generated that has to be moved by the HVAC, repeat that for all the MTAs, proxies, gateways, routers, switches, cell base stations, etc that are involved in the exchange and I can see how the overall consumed power might add up. Of course in reality, it's hard to impossible to trace the total power consumption and CO2 emissions of a single transaction, so I suppose, you'd rather measure in aggregate - e.g. some data center that does nothing but email routing processes x million emails and consumes y kWh per day, so assume a power draw of y/(x*1 mio) kWh per email for that data center -, then multiply by the average number of hops per email, add routing on the lower layers to that, etc etc, and I see how you could feasibly arrive at sure a value.
- dghlsakjg 2y ago
  Sure, I'm not arguing that it doesn't use power, I'm arguing that it is ridiculous to believe that sending an email with an attachment is somehow a very intensive prospect (or that sending an email can really be measured in grams of CO2). I can do it in my house, using extremely inneficient machines for 1/1000 the carbon emisions being claimed. To put it into context, 50g of carbon is more than all of the carbon that will be emitted by a disposable lighter if you burn it until the butane is gone. That is an incredible amount of energy to claim that sending an email costs. Especially when you consider that in the same breath they claim that sending an email without an attached e-ticket uses 1/250th the energy. I just refuse to believe that sending a QR code in a PDF is somehow polluting more than burning all of the butane in a light. of course the real answer is that power isn't free, and emitting carbon costs power. If sending an email cost that much power, each email would cost more than $.01 to send in energy costs. Given the number of emails I can arbitrarily send and receive I wouldn't expect email sending to be so cheap, and I would expect large attachments to be extra.
asynchronousx 2y ago
Thank you for actually putting some numbers out- based on my experience with AWS I’d have to guess most datacenters are actually 10-50 times more efficient with an email than a homelab would be, if measured atomically. It gets tricky because datacenters obviously want to be maxed out as much as possible vs. idle time, but the extra electricity usage is not coming from Exchange servers.

Dalewyn 2y ago

>Where in the world is it 50g CO2 to send an email (or 62Wh if we convert it back)? What in the world makes you think that a paper ticket, which is literally several grams of carbon and oxygen, has less of an impact? I agree with you that the provided example is stupid, but I think there is an argument to be made that going digital can lead to more waste. Take for example digital signage. All those signs, billboards, advertisements, and more that used to be just cardstock paper in ye olde days are now bright, huge LCD screens. I don't want to be the guy signing off on the electric bill, nor consider the geopolitiecological side of where the screens came from.

jiggawatts 2y ago

These "estimates" are often conveniently cherry-picked to support some agenda, on both sides of every political debate. I've dug into a few like these and found that maybe they were valid 20 years ago, and nobody bothered to update the data point. Why? Because that would be inconvenient to their agenda. One that particularly annoys me is some right-wing pundits that keep insisting that solar panels can't recoup the cost of the energy invested into making them. That was true for the panels used to power portable calculators back in the 1980s! Some Ronald Reagan voters got this stuck in their brains and now they are unable to dislodge this factoid without surgical intervention. If you stop and think about it for ten seconds, it would be obviously absurd that there are huge industrial-scale solar power plants springing up by the hundreds globally if they were making a financial loss by selling less power than it took to make the panels! Not to mention that in all forms of power generation (fossil, nuclear, or renewable) the Energy Returned on Energy Invested (EREI) is a key metric that industrialists have been seeking to optimise for over a century now.

mckn1ght 2y ago

I love doing these exercises and would also love to see the math behind CO2e if anyone could provide a helpful link. I’m also thinking about the carbon emissions that go into getting that piece of paper in your hand. The lost carbon capturing capacity of the felled tree, the machinery used to fell, transport and process it, the production of the chemicals used to process the pulp, and finally moving around the finished product. And probably quite a few steps I am missing. Maybe the cars everyone drives to work at the paper plant each day? I’m sure the amortized amount is still small for a tiny piece of paper, but it’s not nothing.

dghlsakjg 2y ago
Don't forget that the piece of paper get used for 5 minutes while you walk to the ticket checker, at which point, the entire process is reversed with a janitor, garbage truck, dump/recycling process.

qiqitori 2y ago

My mom heard this statistic somewhere and now doesn't send email anymore, just WhatsApps :/

xg15 2y ago

> I understand that there will be server farms handling my email in the real world, but those server farms are handling billions of other emails, and live and die on power costs, so I don't really think that my email is sucking down power there? Of course it does? The handful of CPU cycles the server spends on it draws power, so will the tiny amount of additional heat generated that has to be moved by the HVAC, repeat that for all the MTAs, proxies, gateways, routers, switches, cell base stations, etc that are involved in the exchange and I can see how the overall consumed power might add up. Of course in reality, it's hard to impossible to trace the total power consumption and CO2 emissions of a single transaction, so I suppose, you'd rather measure in aggregate - e.g. some data center that does nothing but email routing processes x million emails and consumes y kWh per day, so assume a power draw of y/(x*1 mio) kWh per email for that data center -, then multiply by the average number of hops per email, add routing on the lower layers to that, etc etc, and I see how you could feasibly arrive at sure a value.

dghlsakjg 2y ago
Sure, I'm not arguing that it doesn't use power, I'm arguing that it is ridiculous to believe that sending an email with an attachment is somehow a very intensive prospect (or that sending an email can really be measured in grams of CO2). I can do it in my house, using extremely inneficient machines for 1/1000 the carbon emisions being claimed. To put it into context, 50g of carbon is more than all of the carbon that will be emitted by a disposable lighter if you burn it until the butane is gone. That is an incredible amount of energy to claim that sending an email costs. Especially when you consider that in the same breath they claim that sending an email without an attached e-ticket uses 1/250th the energy. I just refuse to believe that sending a QR code in a PDF is somehow polluting more than burning all of the butane in a light. of course the real answer is that power isn't free, and emitting carbon costs power. If sending an email cost that much power, each email would cost more than $.01 to send in energy costs. Given the number of emails I can arbitrarily send and receive I wouldn't expect email sending to be so cheap, and I would expect large attachments to be extra.

asynchronousx 2y ago

Thank you for actually putting some numbers out- based on my experience with AWS I’d have to guess most datacenters are actually 10-50 times more efficient with an email than a homelab would be, if measured atomically. It gets tricky because datacenters obviously want to be maxed out as much as possible vs. idle time, but the extra electricity usage is not coming from Exchange servers.