Today is Earth Day!

I thought I’d re-post the results of an analysis I did last year that compared the manufacturing energy needs versus the operating energy needs of a computer.  The upshot: unless you currently hardly ever use your computer, you’ll save more energy by upgrading your current computer (instead of replacing it) than you will by switching it off more often.

Our goal, as it was presented to us, was to save energy by convincing computer users to switch off (or enable power managmenet on) their computers when not in use.  However, we found that the worst offenders — those users with multiple computers switched on almost all of the time — typically were already aware of the energy cost of their actions. They knew about power-saving strategies and behaviors, but they had made conscious decisions that convenience trumped waste for always-on computing.  The light users — those who turned on their computers only occasionally — were already conserving energy, so there was no behavior to change. We went back to the drawing board and started kicking around new ideas.  Eventually, we formed a new hypothesis: perhaps we could save energy not by turning computers off but by reducing their production.  I set out to bring data to the discussion.

Using data from an IEEE article and a CodingHorror post, I found that it takes about 2000 kWh of energy to manufacture a computer and about 193 kWh of electricity to operate that computer continuously for a year.  Moreover, the data show that a computer is typically replaced after 3.0 years of use. I dug deeper.

What if, instead of replacing that computer after 3.0 years, the user were to add more RAM and (assuming it’s a laptop) give it a new battery?  That upgrade should give the computer another year of useful life, a total of 4.0 years, which means that 25% fewer computers would be purchased.  In energy terms, that means that 2000 kWh* 25% = 500 kWh would be saved.  Subtract the 85 kWh that would be required to manufacture the replacement RAM and battery, and there’s a 500 kWh – 85 kWh = 415 kWh savings.  That’s the same amount of energy as it would take to run the computer for 415 kWh / 193 kWh/yr = 2.2 years!

The user would need to go from running his computer 100% of the time to about 25% of the time to save the same amount of energy, which is a change that we found users were unwilling or unable to make. In contrast, we found users to be very receptive to the idea of delaying new computer purchases for a year via upgrading existing machines. The delayed-purchase approach also has the advantage of effecting real savings for both heavy computer users and light computer users.  The users also liked the fact that they would save money and reduce landfill waste by upgrading their existing systems instead of purchasing new ones.  Clearly, delayed purchases are a better energy-savings strategy than promoting power-savings modes to consumers.

A problem with our approach was that it was for a project sponsored by Climate Savers Computing, which receives financial support from major PC manufacturers.  Although the Climate Savers reps at our presentation seemed to like our idea, I imagine the managers might be skittish about recommending actions that would decrease sales for their patrons.

You all probably remember the video we made to advocate upgrading PCs instead of replacing them, but I thought I’d post it again.  Marilyn and Dan are the actors, and I was the director: