Is Sydney’s heat decoupling from demand spikes?

If you’re something like me, you don’t do quite so well in the heat. I was born in London; I’m at my happiest pulling up a thick collar, pushing through wind and snow, with a scarf waving in the wind behind my neck.

So, you might be wondering why Sydney’s been through a spell of fairly serious heat the past few weeks. Not extreme heat, mind you – just persistent, unseasonal warmth with no cold breaks, and no rain.

This article in The Conversation explains it quite neatly:

“Over recent weeks, there have been extended periods with a high-pressure ridge over eastern New South Wales and weak low-pressure troughs inland. That is creating mostly stable atmospheric conditions for Sydney (that is, no rain) and funnelling warm air from central Australia over the metropolitan region”

The article explores a range of other factors, but mentions climate change, too:

“If we don’t consider Sydney’s recent weather in isolation, then it is clear that climate change is playing its role here as well. Both land and ocean temperatures have warmed. And those warmer offshore sea surface temperatures keep our nights warmer, particularly in coastal cities.

Warm, hot conditions are becoming more likely, while cool outbreaks are becoming less so. The whole of southeastern Australia has experienced fewer cold outbreaks in recent years (although they do still happen).”

Out of curiosity, I pulled some data from the Bureau of Meteorology’s amazing ‘Climate Data’ tool. Here a fairly simple illustration of just how well above average the last few weeks have been:

February in particular has had a run of hot nights – something which tends to compel the usage of air-conditioners. We know air-conditioners tend to increase power bills, but I feel we we’re not generally cognisant of just how much they push up energy demand – here’s a quick chart I made last year showing the relationship between NSW demand and temperatures – a relationship coupled by the usage of air-conditioners.

This is pretty interesting – there’s a distinct link between hot days in Sydney and broader NSW demand, which makes sense. So has Feb-March 2016 seen an uncommon increase in state-wide demand, to match Sydney’s uncommon run of hot days? Below, I’ve used Global Roam’s NEM Review tool to pull some historical data, and compare it to the past few weeks:

Curiously, demand isn’t much higher or lower than the past five years. Keep in mind this comparison is slightly different – averages from 2000 to 2015, rather than temperature averages from 1859 to 2015 (like-for-like also shows average 2016 temps above the 2010-2015 average, though).

Regardless, there’s a novel disconnect, here. Sydney’s  unusually high temps aren’t driving unusually high demand.

As the article in The Conversation points out, this spell of heat isn’t affecting all of NSW, which would partly explain why there hasn’t been so great an impact on state-wide demand levels. But, I suspect rooftop solar plays a big part here. Here’s the shifting daily profile for January to February, from 2000 to 2016, in a dizzying three dimensional chart gif. Notice how there’s a camel hump in the middle? NSW’ daily demand profile hit a peak in mid 2010, but has been decreasing since then:

from here

To illustrate in a slightly less nauseating way, this is January+February average NSW demand:


This summer’s Sydney heat, and the lacklustre response from state-wide demand, is part of a curious flattening of demand.

This is pretty good news for people like me, who really can’t stand the heat. I rent, but even if I didn’t, I probably couldn’t fit an air-conditioning system (and consequent electricity consumption) into my budget. In the past, price spikes caused by air-conditioning-driven demand were paid for by everyone – regardless of their air-con usage.

Now, demand is lower, and there are fewer price spikes to be covered by consumers, driven partly by solar. It’s still hot, and I’d still prefer to be wafting down an icy street in a scarf, but at least some pressure is being taken off the system to deal with power-hungry air-conditioning systems.

  1. Interesting work, Ketan

    At the end of summer 2014-15 I posted this analysis of the trend in summer demand in NSW (and will look to do the same in April 2016, not least because I need to see who’s won the BBQ prize):

    In the distribution curve I included in that post I highlighted 3 particular factors driving change in demand:
    1) Reducing minimums – with the closure of Kurri Kurri aluminium being the big one in NSW, but better insulation also probably factoring
    2) Reducing “normal day” afternoon demand – with solar PV being the usual suspect here, especially because it seemed to happen in step-change from 2010-11 to 2011-13 (which I think might correlate with the 60c gross FIT feeding frenzy)
    3) Variability of peak demand at the top end.

    We saw the highest demand through the normal summer period on 25th February, as noted here:
    and I listed some of the contributing factors there.

    Hence I was very interested in what you noted, above.



    1. Thanks for that – trying to pick apart the factors behind demand is always quite fun, as I found during my time at Infigen. I think the human behaviour component of it fascinates me the most – you can tell so much about society and cultural norms from something as simple as daily demand.

      It’s a bit of a shame that we can’t get electricity demand broken down by, say, postcodes – that would be pretty cool.

      I plan to go into it more in a future post, but I’m particularly fascinated by how increasing public awareness of electricity demand might actually drive change in personal consumption levels – if you had a glowing globe in your home that warned you when electricity was very $$$, would you casually flick off your washing machine or hot water system, and set it to run later? It’s an interesting concept.


  2. Great piece of work, have you tried to see the effect that humidity has on your results?
    For example you could plot the wet bulb glass temperature or some temperature function that includes humidity.

    I suspect this could clean up the data on the right hand side. Air conditioner performance can vary quite significantly for different amounts of water vapour present in the air.


  3. From what I’ve seen at the individual business and household level is a big shift over the last 4 years, especially the last 2 years, in the depth to which they now ‘participate’ in coping with consumption and demand. SME’s have been driven to address demand and capacity profiles because of better help from Auditors and Consultants addressing a languishing opportunity. Similar things have been happening, at an improved rate, at the residential level.
    The heat-disconnect thing can be partly explained by people becoming accustomed to heat and can tolerate applying less HVAC This breaks down very quickly when gaps in the heat-coping cycle trigger the HVAC over-response on sudden or extreme days & nights.
    More and more monitoring and automation, technology in general, combined with education and information appear to be showing good aggregated results, and the energy productivity plan will help expand that.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s