Monday, January 15, 2007

What if extending Daylight Saving didn't save electricity?

Much of what you thought you knew about daylight saving may well be no longer true.

And two academics from the University of California Berkeley have used evidence from Australia has proved it.

Whatever we have thought about daylight saving personally - whether the change confused them, whether it faded our curtains or upset our cows - we have generally agreed that it is a way of saving electricity. The most widely cited guesstimate in the US is that moving the clocks forward by one hour cuts the use of electricity by one percent.

The US inventor Benjamin Franklin made the case early. He said in 1784 that daylight saving was a way of liberating people from the "smoky, unwholesome and enormously expensive light of candles” in order that they might have "the much more pure light of the sun for nothing”...

New US leglislation will bring forward the start of daylight saving from
this year, wind the clocks back one month earlier in March instead of April,
as a means of reducing power consumption and greenhouse gas emissions.

But two economists from the University of California Energy Institute, Ryan
Kellogg and Hendrik Wolff
are sceptics. They note that the studies
suggesting that daylight saving cuts power consumption are by now more than
30 years old and predate the widespread use of air conditioners.

So beginning a few years ago they looked to the only country they knew of
that had actually performed a natural experiment with extending daylight
saving - Australia in the lead up to the 2000 Olympics.

In that year NSW, the ACT and Victoria began daylight saving two months
earlier than normal, at the end of August instead of October. But South
Australia, also a daylight saving state, did not.

To work out what would have happened to power consumption had daylight
saving not been extended during those months, they examined the electricity
records for South Australia.

To work out the difference that extending daylight saving actually made they
examined the electricity records for Victoria. (NSW and the ACT were
excluded from the study in order to make sure that the economists were not
measuring the effect of the nearby Olympic Games. The weeks during which
the Games actually took place were also excluded from the study.)

What Kellogg and Wolff found in a paper just published in the US by the Centre for the Study of Energy Markets was that while the pattern of energy
consumption in South Australia was pretty much as normal in 2000, the
pattern in Victoria was radically different. There was a new very sharp
peak in power consumption at around 7.00am each morning as Victorians got
out of bed in darker colder conditions than they were used to in August and
September and turned on extra lights and heaters and lights. So severe and
unexpected was the new early morning peak in electricity usage that the
wholesale price spiked to around 15 times its normal level as the system
initially failed to keep pace.

Kellogg and Wolff estimate that bringing forward the start of daylight
saving in Victoria actually increased its total demand for electricity
by 0.34 per cent.

As they delicately state their findings:

All told, our results indicate that claims that extending daylight saving time will significantly decrease energy use and greenhouse gas emissions are at best overstated, and at worst carry the wrong sign. In particular, a long, two-month, extension is more likely than not to increase electricity consumption.

They say they expect the US and in particular the big energy-guzzling region of central California to behave much the same way as did Victoria, which has a similar latitude and climate.

Daylight saving may save no electricity and no
greenhouse gasses whatsoever.


Andrew Leigh said...

Neat empirical strategy, but I'm surprised at the premise. Notwithstanding the name, I'd always thought that daylight saving was about lifestyle, not energy use. Maybe someone should use those Deakin happiness surveys, and see whether Victorians were happier than South Australians over this period.

Andrew Bartlett said...

Interesting research, particularly given Queensland's perennial debate about whether to adopt daylight saving (and I guess WA's current trial of it). I'd often thought that the widespread presence of air conditioners these days might mean higher energy consumption if Qld adopted daylight saving.

I wonder if the same impact identified in this study for extending daylight saving would apply across a whole summer - on first glance I can't see why it wouldn't.

I suppose 0.34 per cent isn't a lot, but if we're looking for every emission saving possible - as we probably should be - then perhaps a bit of extra research on this would be worthwhile.

I assume it would be possible to make some rough comparisons with past years (taking into account underlying growth) on energy consumption in Perth since their trial started.

David Jeffery said...

Well, that may be true for Victoria in August but I don't think it sheds much light on Queensland and WA going for daylight saving starting at the end of October.

If the reason for the change observed in that study was people getting up while it was still dark and cold and turning on heaters etc, that simply has no application to WA and Queensland summers.

Hopefully some useful data will come out of the current WA trial but it's hard to generalise the findings from that Victorian study - it seems very location-specific.

Post a Comment