Taking respite from renewable fiction: Why the numbers do matter

Last Saturday, my Twitter feed was a curious combination of sentiment. During the 2015 Labor party conference, about 90% of the tweets in my feed consisted of genuine dissatisfaction with the Labor party’s policy of turning back asylum seeker vessels. 9% consisted of somewhat hushed celebrations of the Labor party’s climate policies, including a 50% by 2030 renewable energy target, and an emissions trading scheme (1% was a stream of once-amusing but increasingly-unfunny Bronwyn Bishop helicopter memes).

The ALP conference ratified the climate policies unanimously, and Shorten issued a stern challenge to Abbott: “We will not be intimidated by ridiculous, ignorant, fear-mongering scare campaigns that will come….if Mr Abbott wants to make the next election a contest about who has the best policy solution for climate change, I’ve got a three-word slogan for him: Bring. It. On.”

Buried among the inevitably absurd back-and-forth (sourced primarily from the back of envelopes) that’s clogged what might have been, in a parallel universe, a serious policy debate, I noticed that on Saturday, the day of the ALP’s conference, it was a particularly windy day. A little further examination revealed that wind power was its highest output on record, as I showed in yesterday’s graph of the day:

Chart1

Not only did total wind power output peak at 3,378 megawatts at 20:10, the total number of megawatt hours of energy from wind farms was the highest on record. Here are the top ten days, since the start of 2012:

Chart2

When demand dipped, SA came close to having demand match wind output, and excess power was exported to Victoria when demand dipped below total generation:

chart2.5

(To put this in context, South Australian electricity has trended towards net import as the penetration of wind power has grown, but this has to be contextualised against the growth of wind power in Victoria, and the truly impenetrable complexities of interconnector trading, transfer and dispatch, in the electricity market)

As I mentioned yesterday, these moments of high output are important.

Wind speeds vary over time, so wind turbines must be designed to capture the energy in high winds, as well as the more regular middle-range wind speeds. The lay-hater’s interpretation of a 30% capacity factor (or variations in output) is failure or under-performance, but this is simply how the atmosphere works. This is neatly illustrated by a chart that shows the amount of time the NEM wind fleet has spent at different levels of output, packaged into one hundred megawatt blocks, again since the start of 2012:

chart3

When moments like Saturday’s high winds pop into existence with increasing regularity, we know the curve is shifting to the right, up the megawatt output scale – our wind turbines are spending more time at high output. It’s a signal that not only are we capturing infrequent moments of incredibly high wind speeds – we’re also capturing a big bulk of wind at the lower ranges.

I find this incredibly uplifting. That’s a really significant amount of electrical power we’ve captured from the churning currents of the atmosphere – all of which is used instantly to power industry, residences, microwaves, mobile phone chargers. There is so much more room to fill in our energy system for wind power. This tangible, pointedly not-made-up capture and conversation process is such a profoundly soothing balm, when I’m faced with a daily barrage of columnists blurting out fictitious facts and figures.  It’s my only respite, and it’s genuinely exciting.

To illustrate this ‘pushing upwards’ of newly built wind power a slightly different way, consider the same ‘hours at each 100 megawatt level’ chart, but expanded back to 2010 and broken up into each year (note that 2015 is a partial year, so the amount of time is around half of the previous years):

Chart4

Our disconnect from the current and future successes of clean technology is created by the fact we’re desensitised to numbers. Alan Jones’ Qanda debacle, Joe Hockey’s confusion around carbon tax savings, or Tony Abbott’s ‘back of the envelope’ $60bn guess, are small examples in an extremely crowded, expansive and decades-old field of  people simply throwing out half-truths and inventions about renewable energy.

It’s not all bad. Recently, Denmark experienced a similar event, but early in the morning. The news was incredibly popular – around 9,800 tweets were posted about the news (Europe is about 8% wind powered on average; double that of Australia’s percentage). Contained within these moments is a reminder that there are challenges ahead – hence, why wind and solar forecasting is so important. The NEM is a closed system – 140% wind power isn’t possible – so we’re going to need to dwell deeply on the technical complexities of variable output technology (thankfully, we have quite a few experts who are particularly good at confronting these engineering problems).

The Labor party still has some work to do, in terms of fleshing out a series of publicly accepted policy mechanisms that might achieve the goal of 50% renewable energy in a short time period. But we can push the percentage of wind power further upwards, with relative ease. For now, I’m going to take refuge in the solid blocks of electricity we’ve sourced directly from the efficient conversion of atmospheric kinetic energy.

Ketan Joshi is an analyst with a leading wind energy developer and operator.

Ketan Joshi is a European-based climate and energy consultant.

Comments

25 responses to “Taking respite from renewable fiction: Why the numbers do matter”

  1. Paul McArdle Avatar

    Thanks Ketan – indeed, challenges ahead (either way).

  2. sunoba Avatar
    sunoba

    Great article, uplifting. Thanks. Got any insight on what that distribution is (in mathematical terms)? Even a mug statistician like me can see it’s not Gaussian.

    1. Ketan Joshi Avatar
      Ketan Joshi

      No idea! Can I use descriptive terminology? I’d call it ‘nice’ and ‘friendly’

    2. Chris K Avatar
      Chris K

      It looks somewhat ‘log normal’….

    3. Chris Fraser Avatar
      Chris Fraser

      Could be a normal distribution positive skewness;
      https://en.wikipedia.org/wiki/Skewness

    4. David Osmond Avatar
      David Osmond

      It’s pretty close to Weibull. Weibull distributions are commonly used in the wind industry, more usually for wind speed distributions, but occasionally for power distributions too: http://www.cleanenergycouncil.org.au/dam/cec/events/2015-wif/Presentations-in-PDF/Ben-Inkster–An-analytical-investigation-of-wind-energy-intermittancy-in-Australia/Ben%20Inkster%3A%20An%20analytical%20investigation%20of%20wind%20energy%20intermittancy%20in%20Australia.pdf

  3. juxx0r Avatar
    juxx0r

    Hi Ketan, good work again. Nice charts.

    Can you do that bottom chart in % hours per year on the Y-axis, then plot it as a series of line graphs, so that we get more emphasis on the rightward movement? Or email me the data and I’ll show you what I mean? juxxor at the gmails.

    1. Ketan Joshi Avatar
      Ketan Joshi

      I’ll just send you the data 🙂

  4. Kevin Brown Avatar
    Kevin Brown

    It is interesting to see that wind power peaked at 20.10 which is within the period of peak demand. Is wind power better at reducing peak demand than big solar which the Coalition Government is championing? Won’t big solar exacerbate the “duck curve” solar output problem.

    Could someone explain why the Coalition and many cross bench senators are so supportive of big solar when it is more expensive than wind and doesn’t provide power at times of peak demand??

    1. Ketan Joshi Avatar
      Ketan Joshi

      As far as I know, wind isn’t correlated in any way with demand – positively or negatively. The atmosphere just spins about, and it doesn’t really care whether we’re flicking on our plasma screens.

      Solar is incredibly predictable, as you might expect, given the fairly reliable nature of the rotation of the Earth and its orbit around our star.

      The two technologies, combined with storage in the near future, provide the best mix, I think. I did a chart a while back here, having a casual play-around with generation profiles:

      http://www.infigenenergy.com/renewable-energy/blog/wind-and-solar-in-australias-energy-future.html

      I also strongly suspect solar is better-loved by prominent government figures mainly because it is less effective at pushing out coal (at the moment) – when this changes in the future, and solar takes over, it’ll be the main subject of hatred, I suspect.

      1. Kevin Brown Avatar
        Kevin Brown

        Hi Ketan

        WR’s wind power data shows it is more reliable and evenly distributed than solar. It contributes more power than solar at periods of peak demand. It is cheaper than solar.

        Why isn’t the wind power industry demanding that the Coalition and cross bench senators explain why they are promoting the more expensive option of big solar over wind power?

    2. juxx0r Avatar
      juxx0r

      The duck curve ramp rate is nothing compared to the nightly spike of doom caused by everyone turning on their hot water systems all at once. Check out the second graph from the top, the midnight spike of doom.

      1. Kevin Brown Avatar
        Kevin Brown

        Hi juxx0r
        I guess that your “nightly spike of doom” when off-peak demand kicks in is not a problem as long as it does not require more poles and wires. I direct my solar PV output into my hot water to minimise my off-peak hot water heating.

        1. Chris Fraser Avatar
          Chris Fraser

          I would consider doing that just because i may want to find additional appliances for use as storage. But in the meantime, i’m using a timer to turn on a 16A water heater at 2am (the 2am spike of eternal balance and wind utilisation).

    3. WR Avatar
      WR

      From 1 April 2014 to 31 March 2015, the average wind capacity factors for turbines on the NEM at 3-hour intervals were:

      3am 30.4%

      6am 29.9%

      9am 28.3%

      midday 28.8%

      3pm 30.5%

      6pm 30.6%

      9pm 31.1%

      midnight 30.8%

      So, on average, wind energy was distributed fairly evenly throughout the day during that year-long interval. It didn’t seem to strongly favour any particular time of day. I would be surprised if data from other years showed anything different to this.

      Watch my Youtube video and use the associated links to download a simple interactive renewable energy model that will allow you to see how different amounts of wind and PV could be used to supply large amounts of RE to southern Australia. https://www.youtube.com/watch?v=axxcYcumJjg

      1. Kevin Brown Avatar
        Kevin Brown

        WR. Thank you for those figures, The even distribution of wind energy output that you have illustrated simply confounds my lack of understanding as to why the Coalition and anti-wind cross bench senators want to hobble wind power and promote big solar that will excerbate the duck curve solar output problem? Can anyone provide an explanation?

        1. Steve159 Avatar
          Steve159

          I think it’s that the LNP will attack the most effective and efficient form of RE that most threatens coal (“follow the money”)

          Next in their sights is roof-top solar (that’s impacting demand as well).

          Last is large-scale solar because relatively speaking its not (yet) as pervasive.

          So I guess their thinking is “let’s support large-scale solar because there’s bugger all of it, and it’s got relatively long lead times, so then we look like we’re supporting renewable energy. Mind you, if large-scale gets up we’ll have to belt it as well, so our mates in coal don’t lose their shirts.”

      2. Ketan Joshi Avatar
        Ketan Joshi

        Yeah, that is quite interesting. Thanks for sharing!

    4. Alastair Leith Avatar
      Alastair Leith

      SolarCST with thermal storage holds some of that power over until what is now an evening peak instead of a midday peak. It dispatches (when we actually get some built) into the peak markets and can store for a few hours or even 24 hours depending on the specs of the plant built.

  5. Kevin Brown Avatar
    Kevin Brown

    Hi Ketan

    Your NEM generation graph for 25/07/2015 shows solar output peaking at around 20,000mWh at about 18.05. How is this possible? Doesn’t solar output peak around midday?

    1. Ketan Joshi Avatar
      Ketan Joshi

      I might have not labelled them clearly – large-scale solar is only a tiny sliver in that chart, that you can just see appearing around midday, underneath the red band (wind).

      The values are ‘stacked’, meaning the upper bounds of the stack is the total, and each band shows the volume for each. The units are in megawatts, five minute-resolution data.

      1. Kevin Brown Avatar
        Kevin Brown

        Hi Ketan

        Thanks for clarifying this but total solar output still appears to be at a maximum around 10.05 and 18.05? How come?

        1. Alastair Leith Avatar
          Alastair Leith

          I think you are looking at Gas output. I was tricked by that too at first. There are three “green” colours used in the Graph but the stack order is the same as the Key order so it’s definitely Gas that is the band which runs day and night and peaks at those times you mention.

  6. Paul Turnbull Avatar

    When moments like Saturday’s high winds pop into existence with increasing regularity, we know the curve is shifting to the right, up the megawatt output scale – our wind turbines are spending more time at high output. It’s a signal that not only are we capturing infrequent moments of incredibly high wind speeds – we’re also capturing a big bulk of wind at the lower ranges.

    I find this incredibly uplifting.

    So do I! Thanks for pointing it out and using such potent graphs.

  7. onesecond Avatar
    onesecond

    Hi Ketan, if you find this uplifting, you might want to look here from time to time.

    http://www.agora-energiewende.de/de/themen/-agothem-/Produkt/produkt/76/Agorameter/

Get up to 3 quotes from pre-vetted solar (and battery) installers.