Slashdot Mirror

← Back to Stories (view on slashdot.org)

The Environmental Impact of Google Searches

Posted by timothy on from the compounding-wild-ass-guesses dept.

paleshadows writes "The Times Online reports that researchers claim that each query submitted to Google has a quantifiable impact. Specifically, two queries performed through a desktop computer generate about the same amount of carbon dioxide as boiling a cup of tea. From the article: 'While millions of people tap into Google without considering the environment, a typical search generates about 7g of CO2 [whereas] boiling a kettle generates about 15g [...] Google is secretive about its energy consumption and carbon footprint. It also refuses to divulge the locations of its data centers. However, with more than 200m Internet searches estimated daily, the electricity consumption and greenhouse gas emissions caused by computers and the Internet is provoking concern. A recent report [argues that] the global IT industry generate[s] as much greenhouse gas as the world's airlines — about 2% of global CO2 emissions.'" Google makes an interesting focus for such claims, but similar extrapolations have been done before about, for instance, the energy costs of sending a short email.

3 of 516 comments (clear)

  1. I don't buy it... by quibbler · · Score: 4, Interesting

    I'd like to see the in-depth math on this, I don't buy these numbers, its smells of environmental-shock-value reasoning... Example - if they are dividing the total power used by google by the number of searches, that would only be applicable if google were working at 100% capacity and if *all* they did was searches...

    This is kinda like the Greenpeace founder who hated nuclear power till they read a freaking book. Boo.

    1. Re:I don't buy it... by LeDopore · · Score: 5, Interesting

      You're right. Here's some math:

      250g water in a cup of tea.
      Specific heat of water = 4186 J/kg/(degree C). (see http://en.wikipedia.org/wiki/Specific_heat_capacity).
      80 Celsius degree change from room temperature to boiling.

      To boil a teacup's worth of water, therefore it takes ~80 kJ.

      For this to be twice the energy consumed with one search, that's ~40 kJ per search.

      If a search takes Google about 100 ms, that means Google would be using 400 kW while responding to your search. That feels like it's about 3 orders of magnitude too high. It's possible that the original researchers got Calories and kCal confused.

      --
      Expected time to finish is 1 hour and 60 minutes.
  2. An old email relating to carbon footprint of data by Anonymous Coward · · Score: 5, Interesting

    I'm on this advisory group of 6 people and we wanted to participate in a 2 day conference by flying a representative there or through video conferencing. For some reason the carbon footprint argument was used IN SUPPORT of flying because of that recent news about data-centres being polluters. There was news that IT are going to be the 2nd largest cause of pollution in a few years, and therefore flying was somehow comparably damaging to IT.

    I thought that this was against common sense, but it was surprisingly difficult to understand the difference. If an ISP wanted to 'go green' what kind of carbon offset would they need to invest in, per Gig? I found a Harvard study[***] on banner ads that seems to be applicable to internet traffic in general.

    It's difficult to quantify and compare the two scenarios[*] but flying to London and back releases about 4,000 kilos of CO2[**] whereas sending 10G of data (video conferencing of youtube-quality video for 16 hours to 7 people) releases about about 100 kilos of CO2[***] + 30 kilos to run 7 computers for two days. While the plane's CO2 cost is only in terms of fuel (and not airports or surrounding infrastructure) the data CO2 from the Harvard study[***] is inclusive of wider infrastructure. Also planes releasing CO2 into the upper atmosphere do more damage than CO2 being released on the ground due to Radiative forcing.

    One interesting thing from the Harvard study relates to Moores Law, "we calculate that energy intensity of the internet declined by approximately an order of magnitude from 2000 to 2006. While energy use approximately doubled in that time period, data traffic grew by more than a factor of 20". Now I know that Moores Law is purely about transistor chip density so please don't misunderstand me -- I just mean that as computers and networks get faster the energy needed for 1 gig of traffic will decrease.

    So it's about 4000 kilos for flying ONE PERSON vs 130 kilos of video conferencing FOR ALL PEOPLE.

    [*] because of course it depends on how wide you consider the effects. Flight pollution should of course include airport pollution but how far do we go? Does it include power company polution for the power needed in the airport? It seems that a lot of IT studies are wider in scope than that of flight.
    [**] http://www.cheap-parking.net/flight-carbon-emissions.php for flying half way around the world and back.
    [***] Harvard Study on CO2 for data: http://www.imc2.com/Documents/CarbonEmissions.pdf

    ps. In New Zealand? Sign up to http://CreativeFreedom.org.nz