Embodied energy in computers

[DidelisDiskas]

As i strive to be more environmentally conscious, i started thinking about the whole carbon footprint of the electronics i buy. Since computers, smartphones and their parts comprise most of my hardware purchases (ever since i got rid of my car), it dawned on me that looking only at my personal use of the device and it's use of energy does not paint the whole picture, so i googled a bit and found this article:

http://www.lowtechmagazine.com/2009/06/embodied-energy-of-digital-technology.html

It claims that: "Considering an average life expectancy of 3 years, this means that the total energy use of a computer is dominated by production (83% or 7,329 megajoule) as opposed to operation (17%). Similar figures were obtained for mobile phones."

I was shocked by these numbers (although the 7.329 megajoule number seems a bit sketchy, since 17% of that would only be 0.5 kw/h? of energy use in 3 years), so i scrolled down to the comments and someone posted this:

https://www.one-report.com/download.html/2011/shared/library/0692-00006713.pdf

In this document the author claims that: "The contribution from the use phase represents between
65% (for China) and 47% (in Europe) of total impacts" (this is whit a 4 year cycle of use in mind)

So even though the use cycle in this document is 1 year longer, the numbers are much more equally divided between the production and use of the device. But it would be interesting to read more work on this topic, since i don't want to be buying much more efficient (and less powerful) devices every few years (instead of buying a powerful one for 5-6 year use cycle) and actually contribute more to global pollution by doing so. Does anybody have any thoughts on this, or any interesting links to studies and articles on this topic, since it doesn't seem to be brought to attention at all while discussing the energy consumption of a device?

 

[[DHT]Osiris]

I wonder how much is also involved in the recycling/disposal, for instances not involving just dumping it in a landfill.

Also, I wonder how many computers (higher level electronics in general) end up being used beyond their '4 year lifespan' as donations/3rd world countries/whatever as opposed to ending up in landfills, that could skew the numbers further into operations.

 

[DidelisDiskas]

I wonder how much is also involved in the recycling/disposal, for instances not involving just dumping it in a landfill.

Also, I wonder how many computers (higher level electronics in general) end up being used beyond their '4 year lifespan' as donations/3rd world countries/whatever as opposed to ending up in landfills, that could skew the numbers further into operations.

The second paper i quoted says: "End of life (with the assumption of 75% recycling rate) has a contribution of between 4-9%, reducing the total carbon footprint of the laptop by ca. 30kg of CO2eq in each scenario. Transport to recycling has but a minor contribution, less than 0.1% of the total impact."

So not that much either way? Although i guess that's much better than nothing.

 

[Torn Mind]

Good to see that you have become more enlightened. Too often, the environmental movement is extremely short-sighted and ignorant of less obvious things to do when it comes to saving the environs. Such as dumpster diving. Or not realizing that cars' metal is guaranteed to be recycled once the junkyard is finished with it.

permies.com, now richsoil.com, awakened me. Although I don't quite agree with his viewpoint on CFLs nor would I believe everything on that site, it offers a valuable perspective.

 

[[DHT]Osiris]

The second paper i quoted says: "End of life (with the assumption of 75% recycling rate) has a contribution of between 4-9%, reducing the total carbon footprint of the laptop by ca. 30kg of CO2eq in each scenario. Transport to recycling has but a minor contribution, less than 0.1% of the total impact."

I personally think that 75% number is very liberal, it's probably closer to 40-50%, though I don't have any data to back that up. I just find it very unlikely that 75% of breadboards end up at a specialized recycling facility yet only 70% of cardboard gets recycled when they pick it up at your curb once or so a week.

Though having said that, if the contribution is only 4-9% that's a fairly minimal difference either way. Interesting numbers regardless.

 

[VirtualLarry]

Interesting subject. Would those papers you linked, suggest recycling Core2-era machines, or keep using them, until they die?

 

[[DHT]Osiris]

Would those papers you linked, suggest recycling Core2-era machines, or keep using them, until they die?

If the operating vs creation costs are legit, it would suggest recycling *everything* as much as possible (which is a reasonable argument regardless, you'll never recoup the creation costs but you can hedge down the TCO by using as long as possible). Barring a completely revolutionary change in how stuff works which reduces the power costs so much as to make it cost more to run the object than buy a new one and let it run at a lower power usage (after x ROI, whatever x happens to be).

 

[dave_the_nerd]

It claims that: "Considering an average life expectancy of 3 years, this means that the total energy use of a computer is dominated by production (83% or 7,329 megajoule) as opposed to operation (17%). Similar figures were obtained for mobile phones."

Yup. Which is why you should build PCs that are overpowered enough to last for 6+ years instead.

Interesting subject. Would those papers you linked, suggest recycling Core2-era machines, or keep using them, until they die?

Recycle. But more importantly, no fair giving people "new" PCs with cell phone CPUs in them that will need to be replaced in 18 months.

And disposable tablets need to die.

 

[DidelisDiskas]

Interesting subject. Would those papers you linked, suggest recycling Core2-era machines, or keep using them, until they die?

That's one of the questions i also have. It's kind of difficult to find some study based, concise explanations, that would actually do the numbers on these kinds of scenarios. As i'm not sure how much the manufacturing and the parts have changed and what influence they have on the new computers having more, or less embodied energy than those core2 machines:

Fig. 5. Carbon footprint contributions from components of the motherboard

If i'm not color blind, the ram seem to be a major culprit in this equation, so i wonder if anything about that has changed, is ram manufactured from different materials now, is it manufactured more efficiently? I guess one part of this would be to just not buy new products all the time :/ .

Yup. Which is why you should build PCs that are overpowered enough to last for 6+ years instead.

Well that's what i want to find out, since those papers are not exactly up to date and are conflicting with each other, i would like to know how, if at all this has changed today and if we include the much higher tdp of a pc (and it's components being much larger also), does it offset the embodied energy + power consumption that 2-3 new devices, albeit more efficient, bring with them. One would think that with all the new initiatives against global warming and polluting, there would be more information on this stuff...

 

[DidelisDiskas]

I personally think that 75% number is very liberal, it's probably closer to 40-50%, though I don't have any data to back that up. I just find it very unlikely that 75% of breadboards end up at a specialized recycling facility yet only 70% of cardboard gets recycled when they pick it up at your curb once or so a week.

Though having said that, if the contribution is only 4-9% that's a fairly minimal difference either way. Interesting numbers regardless.

Well, for what it's worth, here is a rather short and concise article about this issue:

http://ifixit.org/recycling

 

[dave_the_nerd]

Well that's what i want to find out, since those papers are not exactly up to date and are conflicting with each other, i would like to know how, if at all this has changed today and if we include the much higher tdp of a pc (and it's components being much larger also), does it offset the embodied energy + power consumption that 2-3 new devices, albeit more efficient, bring with them. One would think that with all the new initiatives against global warming and polluting, there would be more information on this stuff...

Higher max TDP (and "larger components"? Not sure what you mean there) doesn't correspond to higher power use. A new-today PC idles at a much lower power draw than a new-ten-years-ago system, even if some of the parts have higher TDPs.

There's plenty of information out there about some of this, but it's not generally easily/quickly available because for most individual consumers, it doesn't really make a huge difference, so nobody has gone to the trouble of making pretty charts and so on. People who buy computers by the truckload worry about power use but are pretty much insulated from the manufacturing side of the environmental impact.

http://www.networkworld.com/article...ergy-than-running-the-devices-they-build.html
http://www.dell.com/learn/us/en/usc...bon-footprint-whitepaper.pdf?c=us&l=en&s=corp

etc.

Of course, the more efficient we make the computers themselves, the larger the % of the lifetime impact is in manufacturing. Basic accounting principles: your biggest savings are usually found in your biggest line items. So the manufacturing side is probably where the majority of progress is yet to be made.

 

[DidelisDiskas]

Higher max TDP (and "larger components"? Not sure what you mean there) doesn't correspond to higher power use. A new-today PC idles at a much lower power draw than a new-ten-years-ago system, even if some of the parts have higher TDPs.

I did not describe what i meant all that well here (sorry for the confusion). I was thinking about a scenario of buying a long lasting pc, versus a laptop every other few years. Even if a pc idles, we still pair them with a much larger screen that can eat up more energy than a laptop idling and not to mention that when the pc idles it eats more energy than an idling laptop. So i was thinking if including the laptop energy savings and the fact that pc components are bigger (and thus have more parts and more environmental footprint), does that offset their manufacturing footprint, or does this 6+ year pc still win out in having expended less total pollution in it's life cycle.

1 PC (total pollution) vs 2-3 Laptop (total pollution)

There's plenty of information out there about some of this, but it's not generally easily/quickly available because for most individual consumers, it doesn't really make a huge difference, so nobody has gone to the trouble of making pretty charts and so on. People who buy computers by the truckload worry about power use but are pretty much insulated from the manufacturing side of the environmental impact.

http://www.networkworld.com/article...ergy-than-running-the-devices-they-build.html
http://www.dell.com/learn/us/en/usc...bon-footprint-whitepaper.pdf?c=us&l=en&s=corp

etc.

But most individuals don't care about using less electricity and saving energy (to a point where it really hits their wallet), yet there are all these videos and articles about how we should use more efficient appliances, however not many of them touch on this subject, when this might actually be the most important part in having less of an impact on environment.

Of course, the more efficient we make the computers themselves, the larger the % of the lifetime impact is in manufacturing. Basic accounting principles: your biggest savings are usually found in your biggest line items. So the manufacturing side is probably where the majority of progress is yet to be made.

What is a "line item"?

 

[dave_the_nerd]

What is a "line item"?

A category of expenses.

Say this is my abridged monthly budget:

• $800 rent
• $200 groceries
• $200 gas
• $100 car insurance
• $50 cell phone bill

Rent is one line item. Groceries are one line item. And so on.

And theoretically, it's easier to save $100 on rent than it is to save $100 on groceries or gasoline.

Applying the same idea to computers and environmental impact, (making up numbers here, and ignoring different types of chemical pollution, but hopefully it makes sense anyway) if it takes 10MW of power to build a computer and 2MW of power to run a computer for 3 years, then if you want to shave 1MW of power off the total impact of the computer, it's probably easier to make the manufacturing process 10% more efficient than it is to cut the power draw of the computer in half.

 

[dave_the_nerd]

I did not describe what i meant all that well here (sorry for the confusion). I was thinking about a scenario of buying a long lasting pc, versus a laptop every other few years. Even if a pc idles, we still pair them with a much larger screen that can eat up more energy than a laptop idling and not to mention that when the pc idles it eats more energy than an idling laptop. So i was thinking if including the laptop energy savings and the fact that pc components are bigger (and thus have more parts and more environmental footprint), does that offset their manufacturing footprint, or does this 6+ year pc still win out in having expended less total pollution in it's life cycle.

Ah. I see.

My bet is yes; using a desktop PC for 6 years vs. a new laptop every 2-3 years, the desktop will more than likely win. Same number of components (CPU, chipset, motherboard, etc.) even if they're in a smaller case. And a desktop rig is more likely to be fixed if a part breaks, rather than laptops that tend to get replaced completely.

It depends on not just the size of things, but what they're made out of - various chemicals and pollutants that get used in chip manufacturing, circuit board manufacturing, plastics, etc. Is a steel-cased desktop computer "greener" than a plastic one? I honestly don't know. So I'm just guessing yes.

The increased power use from the consumer may be a concern, but it depends on how your power is generated, too. (Solar vs. coal, etc.) Consumers will tend to care more about that, since they're the ones paying for it, but it doesn't really matter in the ecological sense.