# To Print or Not to Print?

During the summer, I was musing on Vannevar Bush’s ideas presented in his paper As We May Think.  Having the attention span of a…  where was I?  Oh, yes.  So my mind wandered to thinking about by how much our carbon footprint could be reduced by switching to a paperless office.  The answer surprised me and I mentioned it idly to my research supervisor, Steve Easterbrook.  He suggested I publish it on a blog and we discussed the creation of a software tool to help present the argument more clearly.  I’ve finally worked up the nerve to start up this blog to do the former and the latter is in the works.  So here is my first post detailing how much carbon dioxide is emitted by reading a page of text on a computer instead of printing it out.  Ladies and gentlemen, start your stopwatches!

Warning!  This is a back-of-the-envelope calculation!

See a simplified version of this argument in graphical form.

## Assumptions

A1:  Computers are discarded before they stop working
A2:  Monitors are discarded before they stop working
A3:  Most documents in homes/offices are printed on a laser printer1
A4:  Printers are discarded only when they stop working
A5:  Printer fuser breaks after ~100000 sides
A6:  Printer toner cartridges are used until empty
A7:  Each cartridge lasts for 2000 sides
A8:  Only new cartridges are used; no recycled ones are used
A9:  Fossil fuels required to build a printer are about the same as building computer
A10: Printer weight is 5kg/printer
A11: Printer uses 500W while printing
A12: Only uncoated paper from non-recycled paper is used
A13: There is no proactive tree planting
A14: Computer/monitor consuming 0W when reading paper printout (e.g., reading on the subway)
A15: Printers print at 8 pages per minute
A16: All the CO₂ from paper manufacturing from tree logging to consumer is the same as transporting it 12000km (i.e., we can lump in all the other CO₂ from manufacturing paper by inflating the emissions due to transportation).  This is the same as transporting the paper across the full length of the Trans-Canada Highway 1.5 times.
A17: Printers only use 20 lb copy paper
A18: We can ignore the embodied energy of creating a truck
A19: Employees of factories and power plants generate a net of 0 CO₂; they would produce this CO₂ regardless of where they were employed
A20: Computers used for reading documents/printing would have been purchased, even if they were not used for reading documents/printing
A21: Printers are turned off when not in use2
A22: Printing is always double sided and both sides of the page are used.
A23: No misprints
A24: Computer is running at idle power while viewing documents on a screen
A25: Low power computer uses 80W (based on measurements around the lab in which I work, looking at computers in the lower quartile of consumption — much lower consumptions are possible)
A26: 24″ LCD monitor uses 50W at medium brightness (LCDs with LED backlights use quite a bit less power)

## The Facts

F1:  Toner cartridge manufacturing and distribution result in 4kg of CO₂
F2:  Producing 100 kWh in Ontario results in 46 kg of CO₂
F3:  Producing a computer results in 10x its weight in CO₂ waste
F4:  One tree can be turned into ~8000 sheets of paper
(F5):  One tree absorbs 22kg CO₂/year
F6:  A medium truck produces 90g CO₂/T/km
F8:  1 ream of paper is 2000 A4 sheets3
F9:  A paper’s weight is the weight of one ream
F10: 1 metric ton/1000kg = 1
F11: 1000g/kg = 1
F12: 60minutes/hour = 1
F13: 1kWh/1000W/hour = 1

## The Reasoning

SC1:  We make the simplifying assumption from A3 that all printers are laser printers
SC2:  We make the simplification from SC1+A5 that printers are replaced every 100000 sides
SC3:  We make the simplification from F7 that pulp mills produce 0 CO₂
C1:  A1+A2+A20 imply that viewing information on a computer does not contribute to waste aside from energy use
C2:  A17+F9 imply that a ream of 20 lb copy paper weighs 20 lb
C3:  F8+C2 imply that 2000 sheets of 20 lb paper weigh 9.1kg
C4:  F7+A19 imply that the manufacturing of paper results in 0g CO₂
C5:  F6+A16+C4+F8+F10 imply that paper transportation results in 5g of CO₂/sheet
C6:  A10+F3+SC2+F11 imply that the amortized cost of a printer accounts for 0.5g of CO₂/side
C7:  F1+A7+F11 imply that toner accounts for 2g of CO₂/side
C8:  F2+A15+F12+A11+F13+F11 imply that electricity accounts for 0.5g of CO₂/side
C9:  C6+C7+C8 imply that printing process results in 3g of CO₂/side

## The Upshot

C10:  C4+C5+C9 imply that printing 1 sheet of paper results in 6g of CO₂/side
C11:  C1+C2+F2+A25+A26+F13+F11+A24 imply that using a computer results in 60g of CO₂/hour
FC:  C9+C11+A23+A14+A21 imply that if one spends 6 minutes reading a page, it is more CO₂-efficient to print it out than to read it on a desktop computer

## The Conclusion

How often does one spend 6 minutes or more reading a page?  Perhaps not that often.  On the other hand, there may be some papers that one reads over and over again or perhaps shares with others that will result in a cumulative reading time of over 6 minutes.  In such situations, printing will result in less CO₂ emitted.  Remember, this comparison only deals with CO₂ and none of the other ickies produced during the paper-making process or computer equipment manufacturing.

## Notes

Note that in this analysis, I have systematically given reading on a computer an advantage!  If one uses recycled paper and toner, paper’s CO₂ impact drops by about 50% and you get 3 minutes to read a page to reach break-even (remember, these numbers change depending on how clean/dirty your energy source is).  Further, one need not replace an entire printer when only the fuser breaks.  Of course, one could also use a lower power computer.  I’ve measured the power consumption of the laptop I use, a first generation MacBook Air, to be 10W with Wi-fi on and the screen at medium brightness.  That’s 8% of the hypothetical system in the measurements above, allowing me to read 13 times more slowly and spend about 40 minutes reading a page instead of printing on recycled paper with reused toner to break even.  On the other hand, a more power hungry system results in less time to break even.  A dual 24″ monitor setup on a more powerful computer with a high-end graphics card would require reading at a rate of more than 1 page/minute to be more CO₂ efficient than printing out the page.
1. Consumer inkjet printing is more expensive, so those that require high volumes of printing are probably using laser printers. []
2. Standby for laser printers is quite high — in the 10-50W range. []
3. There are 500 17″x22″ sheets of uncut paper in a ream as used for measuring the basis weight of paper.  This yields 2000 sheets when cut. []

### 2 comments to To Print or Not to Print?

• textibule

I read this article on my computer screen. If I knew before reading what I know now, I would have printed it.

• Alecia

Finally, the proof I needed. *thumbs up*