FactCheck: Why Osborne is calling the kettle green
“IT is an increasingly heavy user of energy – the typical visit to Facebook uses as much energy as boiling a kettle.”
George Osborne, 9 November 2012
The background
Several readers have asked us to adjudicate on the one burning political question that you just can’t ignore this week.
Does it really use as much electricity to catch up with your friends on Facebook as it does to boil a kettle?
That’s what George Osborne asserted in his first big speech on science to the Royal Society.
More than anything else the Chancellor said in the speech – which set out his vision for Britain’s high-tech industries but controversially avoided the thorny topic of onshore wind farms - this is the burning question that is exercising people on Twitter and internet forums.
Several commentators have claimed that the chancellor was hopelessly wrong - out by an order or magnitude. But we think he’s got it roughly right. Here’s why.
The analysis
This can only be a back-of-envelope effort, but here goes.
Facebook estimated its annual energy costs at 532 million kwh (kilowatt hours) in 2011. About 400 million people were visiting the site every day last year. That’s 146 billion visits per year.
So the energy cost per visit is 532 million divided by 146 billion. That works out at 0.0036kwh. All those zeroes make our eyes hurt so for the sake of convenience we’re going to convert that to a more convenient unit of energy – the kilojoule (kj).
Facebook’s energy consumption is just over 13kj per visit.
How much energy do you and I burn when we check our Facebook page? Well, our energy usage depends on whether we’re using a smartphone, laptop or desk computer, all of which use vastly different amounts of power.
US energy saving guru Michael Bluejay (“Mr Electricity”) has done a lot of the homework for us. He reckons a desk computer and a monitor can guzzle more than 300 watts, while small laptops use less than a tenth of that.
We’re going to accept his estimate of 15 to 60 watts for most laptop computers and use the midpoint of that range – 37.5 watts.
Facebook says the average length of a visit to their site is 23 minutes and 20 seconds. An average laptop running for that length of time will burn 52.6kj.
So if we add Facebook’s costs (just over 13kj) to the end-user’s costs we get nearly 66kj per visit.
But this will undoubtedly be an underestimate. What about all the other equipment in between Facebook HQ and your laptop? Wifi router, local servers etc…
We don’t know how much all this costs and frankly don’t know where to begin. Let’s just remember that our final Facebook figure must be a bit on the low side.
Time for a cuppa
Heating a litre of cold water from room temperature (20C) to boiling point (100C) ought to take about 335kj.
Actually, to be strictly accurate you need to know the exact amount of water, the starting temperature, the specific heat capacity of the material it’s made from, the efficiency of the device and other factors. Others have put the figure at between 240kj to nearly 600kj depending on varying factors.
If Mr Osborne is doing the right thing by the planet and only boiling enough for one cup of tea, say 250ml instead of a litre, this brings the figure down to perhaps 80kj per brew.
The verdict
Our guesstimate is that a Facebook visit uses just under 70kj of energy per visit, and a single cup of tea might need just over 80kj, assuming that you’re on a laptop and you boil as little water as possible.
That’s remarkably close, especially since that first figure is a (cough) conservative one.
In truth, many of the assumptions we’re making are pretty vague, and if you change any of them you could alter the final numbers significantly.
But the chancellor is clearly in the right ballpark, certainly within an order of magnitude of accuracy. Long may it continue.
By Patrick Worrall



There are 10 comments on this post
Problem.
You are assuming that all the energy used by the computer is because it’s connected to facebook.
1. Most of the energy usage is the computer working in the background regardless of whether it’s on a site or not.
2. With most people leaving computers on a lot and just flicking to pages the increase in power consumption to go onto a particular site is minimal.
Surely the energy usage (for items left on) that is because of facebook etc is minimal and therefore the calculations are working on a false assumption that people have switched on just to use facebook.
(or I may be wildly wrong!)
Like or Dislike:
0
0
I would argue that most people do not have some water standing around in their kitchen at 20C room temperature ready for making the tea. I would argue that most of us draw fresh water from the mains supply for the purpose whether for lack of planning ahead or for better tasting tea. That water will be considerably colder than 20C even in the warmest part of summer.
Like or Dislike:
0
0
This Facebook nonsense is at best a red herring and at worst disinformation from a man whose scientific credentials are zero. Laptops use about, let’s say 40W(from your figures) – kettles are usually about 1kW (kiloWatt, 1000 watts, so a little bit more then). 1 kW is 1kJ/s (kilojoule per second) so it also depends on how long you use the appliance for. To compare boiling a kettle (a couple of minutes) to surfing the net on Facebook is extraordinary. It sounds like some b****** an aide has come up with and you claim that you have proved it is true. You can always get things to agree unscientifically, like I’ll be twice as (half as) old as you at some point. So what? It don’t mean owt.
Like or Dislike:
0
0
Please give the conversion figure from K/wh to Joules as , many, including me are not used to using the J every day. Also why bother to convert? All the figures started in K/wh and it is something we are familiar with. -, it just over complicates the calculation
May be you chose the Calorie because “A Calorie is the amount of heat needed to raise 1 liter of water by one degree C”
Like or Dislike:
0
0
I would like to see more discussion on this serious topes, not enough science when looking at the world problems.
My computer heat reduces the other energy input for apartment heat I need in the winter bt increases it in the summer with the , I hate to say I do use some air-conditioning in the summer ( I am in a hot part of China).
But the heating is a communal source and may be gas, not sure, , more to look at here…..I am sure others will have a view and expect it will be better informed than mine.
Like or Dislike:
0
0
Hi English Prof. 1kWh = 3600000J. Converted so we wouldn’t be talking about tiny numbers that are difficult to read on the page eg. 0.005277778 vs 0.015555556 kwh.
Like or Dislike:
0
0
What a total waste of time and energy in compiling this report.
To many assumptions too just like George Osborne and his forecasting of the economy.
Like or Dislike:
0
0
The comparison is flawed as the infrastructure of Facebook is taken into account but not the infrastructure needed to supply power to the kettle and indeed making the kettle and getting the water to the tap.
Like or Dislike:
0
0
There are many omissions and assumptions in this analysis that I think render it uncalculable.
To all the above I’d like to add for example that the comparison talks only about boiling a kettle. Few people drink plain boiled water, so the comparison should consider the energy costs for processing and transporting the tea, coffee, chocolate, milk, sugar, lemon and/or whatever else is used to make the beverage in question. Then what energy does it take to filter water to make it drinkable and to pump it around the system to our houses? And then what will it take to gather and process the resultant waste water that, ahem, we ourselves will produce after we’ve finished with the brew?
And anyway, I’ve never been on Facebook!
Like or Dislike:
0
0
Could it be George that IS correct on at least this calculation? Seems likely since he’s NOT forecasting anything.
Like or Dislike:
0
0