Carbon cost of google search 1000 times less than boiling a kettle
January 13, 2009
A few days ago, the BBC new website published this article:
http://news.bbc.co.uk/1/hi/technology/7823387.stm
which states that 2 searches on google uses “as much carbon dioxide as boiling a kettle”. Now I’m pretty sure this is a load of cobblers, it doesn’t make sense and is disputed by google who say a search only uses 1/35 of this amount.
Firstly I have a problem with the recognition that this study is getting by being front page technology news on BBC. Its been reposted all over the internet and taken as the truth. We all like to trust the BBC implicitly so I find reporting an article which is obviously hocum as a bona fide study causes all kinds of trust issues.
Secondly I think some basic calculations using my own electricity monitor could prove that the study by US physicist Alex Wissner-Gross is flawed. The article suggests that a good proportion of the energy use in the search is by the use of a standard home computer for the search. My home computer draws 96W of power and a search takes 0.24 secs. So the energy use for a typical search on my computer is 0.0000064KWh. As a rough estimate I will calculate the total energy use for a search as twice this – to allow for the power needed in google’s data centres (using big power hungry servers for a tiny fraction of a second). So I have a grand total of 0.0000256KWh as the total energy use for 2 google searches.
Compare this to boiling my kettle with the minimum amount of water (about 1 cup). Power use is 2.007KW while boiling and the time taken to boil the kettle is 55 seconds. So total energy used in boiling the kettle for one cup of tea is 0.0307KWh. Final results are that boiling a kettle takes rougly 1000 times more energy than doing 2 google searches.
Now I know that this may not be the perfect calculation but I think my sums are relatively sound. Obviously where I am missing information is on how much power is used at google’s end. Even allowing for my calculation to be out by several factors I could not get anywhere near the result suggested. Now I am not a physicist but i was once a mathematician, but I’m pretty sure my ’study’ won’t get the same recognition as Alex Wissner-Gross!
Entry Filed under: going backwards. .
4 Comments Add your own
Leave a Comment
Some HTML allowed:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <pre> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>
Trackback this post | Subscribe to the comments via RSS Feed
1.
Alex Andronov | January 13, 2009 at 4:51 pm
It didn’t pass the sniff test (or whatever you like to call it). It just sounds wrong. And your post makes me very pleased.
I wondered if they were factoring in every computer between yours and the google machine and thinking you were using them exclusively.
2.
Liam Martin | January 13, 2009 at 5:33 pm
This is interesting. I had similar thoughts as you when I first read the article.
Google have stated that the power used in their datacentre is negligable compared to the power used by your home PC. I.E your own PC would consume more power to complete the search than Google’s datacenters do.
This makes sense, as even though they have hundreds of thousands of servers – and all will be more powerful than the average home PC – they are all processing millions of searches. So the power used per individual search is minute.
However, think about this, if you didn’t use Google to find the information you are looking for quickly, you would quickly consume significantly more energy trying to find the information by other means. Finding stuff on the internet would take a lot longer… thus meaning your PC is in use longer, you’re using more processing power etc.. etc..
… oh dear… surely by the time i’ve finished writing this comment I’ve consumed enough energy to boil 20 odd kettles!
(I don’t think so!).
Cheers,
Liam
3.
fourstar | January 13, 2009 at 9:58 pm
Alex Wissner-Gross is full of hot air.
4.
Alan | April 1, 2009 at 7:05 pm
I wonder if you could do me a favour whilst you have your mathematical head on please. My question is this; How much power was used….
1. In conducting what was most likely a highly suspect study (as you rightly state).
2. Was used by his PC to write the ‘findings’
3. In related telephone calls and emails promoting his ‘findings’
4. Used by the BBC (and other media) in broadcasting it.
5. In the resulting Google searches by people hoping to expose it for the nonsense it is by finding a well reasoned post such as this. (as well as the re-posting on both side of the argument to a multitude of blogs)
6. By the time I spent writing this comment!
I’m sure there must be plenty of steps – but I think I’ve made my point. Puts it into perspective doesn’t it.