Wind Farm Forecasting Actually Works Really Quite Well, You Know

Recently, Terry McCrann, a columnist at the Herald Sun, wrote some snarky words about Tristan Edis, a writer at Climate Spectator, labelling him a 'warmist, dishonest propagandist'. I still quite enjoy the fact that there are people who are capable using the term 'warmist' without irony.

Part of his criticism centred around a correction Edis had made to a chart showing 1-hour-ahead wind forecasts, against the actual generation of wind farms:

"CORRECTION: The initial version of this article incorrectly stated that the chart showing wind power output forecasts versus actuals illustrated the forecast for 24 hours ahead when it was in fact the forecast for an hour ahead."

Tristan Edis followed up by sourcing the actual 24-hour-ahead forecasts from the Australian Energy Market Operator (AEMO), in this article. I liked the chart, and Tristan was nice enough to throw the raw data my way, which I've loaded into DataWrapper, so you can browse to individual points by hovering over the chart:






Charts are fun, aren't they? Here's another one, from AEMO's 2013 Wind Integration Report:

As you get closer in time to the interval you're forecasting, accuracy increases. AEMO's wind forecasting system lets them know well in advance a good approximation of the wind power output for a certain half-hour interval.

You'll often hear invective from critics of wind farms, trying to paint national fluctuations in wind speed as insurmountable moments of terror - reasons to quake in fear at the prospect of some portion of our total electrical supply coming from variable renewable sources.

They want you to feel scared, because they want you to forget facts. Don't give in - you don't need to do either. We can slice off a huge chunk of our reliance on fossil fuels safely and reliably.

Anti-Renewable Bingo During the Upcoming Heatwave

It's going to be an exceedingly hot week in South-Eastern Australia. Not as hot as a couple of weeks ago, which the Bureau of Meteorology labelled 'one of Australia's most significant heatwaves'. During this week, Victoria experienced its hottest four-day period on record.

From the Bureau's Special Climate Statements

As heat soars, so will electrical demand, as domestic consumers turn on their air-conditioning units. It'll be a critical time for the national market operator, AEMO. It'll also be a critical time for critics of renewable energy. Times of need are, I suspect, times at which deception is easiest, and most readily consumed.

By taking the normal operation of renewable energy sources, and painting this as some sort of 'failure', the critics of renewable energy will avoid having to use logic in their polemics.

In a previous post, I explored the mechanics of this spin in a bit more detail. Since the first heatwave, there have been a few more attempts to label the normal operation of variable fuel types as some type of 'failure'. They'll show up again this week, so here's my handy, quick-reference guide to the tactics you'll see deployed by critics of renewable energy, and the logical fallacies they're plugged in to.

Intentional Focus on a Small Time Period

Wind generation is variable; as wind speeds vary, so does the output of wind farms across the National Electricity Market (NEM). The fact that wind speeds are sometimes high and sometimes low will be proffered as some sort of analytical revelation, brought to the masses by merciless scientific analysis. From the first heatwave comes an example, from The Australian's Environment Editor, Graham Lloyd:

Critics will point to periods when the combined generation output of wind farms is low, and offer this as proof of 'failure'. This decrease in output is easily forecast. Sometimes, wind generates a lot, reducing our need for fossil fuels. Other times, it generates less. This is precisely what we expect, and by painting it as some sort of malfunction, anti-renewable writers will be hoping you quickly forget about the actual quantity of energy contributed to the electricity market from renewable sources - take South Australia's fuel-mix, for instance:


You'll see short periods of low generation labelled as a failure, without an explanation of what 'success' would look like (presumably, the generation of wind energy without the movement of wind). It's a logical fallacy we shouldn't forgive.

Assumption that the Electricity Market Doesn't Exist

"If we'd been relying on wind farms, we would have had multiple blackouts and hundreds, if not thousands of extra deaths" wrote Terry McCrann, in an article for the Herald Sun. Wind energy is one fuel source on a diversified energy system. We use coal, gas, hydro, wind and solar to source electrons. Quite simply, the more energy we source from renewable sources, the less we need to burn coal.

A grid comprised purely of wind exists only in the dark recesses of the minds of those opposed to renewable energy. An allergy to nuance drives this fictional sole-technology scenario, the basis of McCrann's horrible visage, of thousands of deaths due to a single, sinister technology.

Out of curiosity, I had a look at how many domestic air-conditioners could have been powered by the output of wind farms, on a five-minute basis, over the course of the heatwave (a domestic air-con uses ~3.08 kilowatts):


At its minimum output, the output of wind farms could still have powered 29,865  air-conditioners. This week, you'll hear people criticising wind, predicated on the assumption that no other generator on the electricity market exists. It's surreal, absurd and wrong.

The Myth of Extra 'Spinning Reserve' 

The fact that wind energy often crowds out fossil-fuel generation is a tricky fact to reconcile with the narrative of inefficacy pushed strongly by critics of renewable energy. South Australia's emissions intensity has been steadily reducing as wind energy begins to comprise a larger percentage of electrical energy in the state:

A mix of reduced demand and increased renewable penetration has led to SA's emissions being down by a third since 2005

To counter this fact, there's a curious little myth that seems utterly invulnerable to contact with reality. It's centered around the concept that even when wind farms are generating electricity, coal and gas fired power stations are, somehow, continually using fuel, as they're being used as 'spinning reserve' - and so, renewable sources provide zero reduction in NEM-wide carbon emissions.

McCrann invokes it in the last line of his opinion piece: "You have to keep extra coal-fired - or gas - stations ticking over, literally 24/7, to be able to supply power when..the wind don't (so often) blow"

There's already some degree of 'spinning reserve' built into the NEM - wind requires no extra, and this 'spinning reserve' is a key component of any energy system, regardless of the penetration of renewable energy. As stated in “GHG [Green House Gas] and Cost Implications of Spinning Reserve for High Penetration Renewables, Technical Assessment Report 73 – March 2008” from the CRC for Coal in Sustainable Development:

"The often held view that operating larger fossil plants at part load to provide spinning reserve would negate the greenhouse gas emission benefits of wind is false"

And, you can read more about an NREL study into the issue here. So, when someone tells you, this coming week, that mysterious, non-specified generators were pumping out carbon despite not generating electricity, think closely on the logic of their claims, and whether they've produced evidence, in the first instance, to support their claim.

The Usage of Really, Really, Really, Really Bad Graphics 

This image, again from McCrann's article, uses a tactic that I suspect we'll be seeing again, in the near future:

There's a few wonderful little components of this attempt to present and explain generation data, that I found amusing. Most obvious is the fact that McCrann chose the day with the lowest quantity of wind generation, and made that chart the biggest. It's clearer when you look at the total energy over the five days - note that the two hottest days are the ones with, by far, the most wind generation:


Another trick is making the (unlabelled) y-axes on differing scales - masking the significant increase in generation that occurred over Thursday and Friday. Confusingly, the chart on the far left seems to have been denied its full axis entirely.

Pointing out that the wind turbine used as the background is actually from Hawaii, not Australia, is an afterthought, at best.

We'll see chart-crimes committed in the name of the invocation of fear around renewable energy. Forget the mechanics of an emotional response to an illogical argument - think about how wind actually operates on the National Electricity Market. You don't need to quake in fear, this week.

----------------------------------------------------------------------------------------

It's odd how predictable these things are. As the penetration of renewable energy increases in the grid, we'll continue to see criticism because renewables haven't instantaneously resulted in the immediate decommissioning of all fossil fuel powered stations. We'll also see continued efforts to distract from the fact that these machines are designed to provide energy, not power. Alone, they can't service the entire energy market, but in a diversified energy system, they're already chipping away at our addiction to fossil fuels.

At the time of writing (08:15 AEST 28/01/2014), wind is providing 40% of South Australia's total generation mix - today, Adelaide maximum temperature is forecast to be 43 degrees celsius:

At times, this will be higher, and at other times, this will be lower. But there's no escaping the fact that renewable energy sources are behaving exactly as we expect them to. If anyone tries to tell you otherwise, take some time to tease apart the logic of their claims - It's likely you'll find a gaping void.

Peeking into the Parallel Universe of Physics Denial

Steve Fielding never failed to deliver. On a fascinating Q&A episode featuring a discomfited Richard Dawkins and the breezy then-Family First Senator, he openly admitted to be being a Young Earth Creationist:



Watching this at the time, I felt an odd thrill knowing that an elected official could hold a belief so brazenly loopy. I found improbably joy in watching Fielding reject basic high school science on national television.

In recent months, an analogous phenomenon has grown in amplitude and frequency, but delivers no thrill. The established rejection of the outcomes of climate science seems to have secured residence in Australian culture, and unsurprisingly, in the rhetorical output of our elected officials. Maurice Newman, Prime Minister Tony Abbott's key business advisor, has penned frequent missives in the masthead of climate change denial, The Australian.

"I also accept carbon dioxide is a greenhouse gas. The trouble is, I cannot reconcile the claims of dangerous human CO2 emissions with the observed record. I admit it. I am not a climate scientist. That said, I have closely followed this debate for more than two decades, having been seasoned originally by the global cooling certainty of the 1970s."

states Newman, in one of several op-eds for the paper. When I read writing like his, I ponder what the current media landscape would look like if it were culturally acceptable to reject the science of, say, x-ray crystallography, or organic chemistry. There's as much scientific disagreement on climate change as there is on the theory of evolution by natural selection; rejecting the research outcomes of a particular field of science, wholesale, is insane no matter what field you nominate.

Perhaps there's some parallel universe, where another branch of scientific inquiry has been deemed the victim of a systematic misinformation campaign, to be draped with the suffocating blanket of fabricated controversy. As it happens, prying open a window into this parallel universe isn't as hard as you might think.

Recently, I reinstalled Greasemonkey for Firefox, and found a script that searches for words on a page, and replaces them with alternate text. I customised the script to replace all instances of 'climate science' with 'physics' - you can see all the variations I used here. Through this lens, we can step into a world where denying all of physics is perfectly acceptable.

Let's begin with Wikipedia's page for Physics Denial:

"Physics denial is a set of organized attempts to downplay, deny or dismiss the scientific consensus on the extent of physics, its significance, and its connection to human behavior, especially for commercial or ideological reasons....Physics denial has been associated with the fossil fuels lobby, industry advocates and free market think tanks, often in the United States. Some commentators describe physics denial as a particular form of denialism"

The script seems to miss occasional words for some reason, but you get a feel for what the world would be like if physics threatened the income stream of fossil fuel lobby groups. One of Australia's key Physics Skeptics groups has a particularly great website:

And, of course, these physics skeptics frequently troll groups dedicated to spreading information about physics, on Twitter:

There's enough content on the Galileo Movement to fill four or five blog posts, but let's move on to the Institute of Public Affairs, a think-tank with links to tobacco companies and conservative politics - they tackle the weaknesses of physics head on:

That search returned some not-unlikely headlines like 'Climate Denial Think Tank's Plans To Abolish Physics'. Then there's the pithily titled No Carbon Tax Physics Sceptics Political Party:

There's Joanne Nova's deep-set hatred of physicists who 'believe there is evidence of physics', and their attempts to induce 'physics scares' by going on expeditions:

Coalition MP Dennis Jensen is an outspoken critic of physics - burdened with an actual education in material science:

In this bizarro world, there's pushback against the Physics Skeptics Movement too, with sites on how to deal with the growing influence of Physics Denial:

And scholarly attempts to examine the nature of Physics Denial:

Alongside the chaotic world of physics denial, our primary scientific organisations quietly continue to conduct research and present clear information on physics on their sites:

---------------------------------------------------------------------------------

The rejection of climate science seems to have been normalised in the arenas of public discourse. If I were to reject the views of experts on the causal relationship between taking sleeping pills and being tired, I'd immediately pay a price amongst my peers. The social cost of wholesale scientific denial has been quietly revoked by a cabal of commentators and groups dedicated to normalising a format of denial that in some parallel universe, we'd rarely forgive.

I think we ought to shift the balance. Let's start invoking that social cost, whenever someone decides to arbitrarily reject the outcomes of an entire field of scientific inquiry. Abbott's systematic denial of scientific consensus is, in terms of sheer insanity, on par with Fielding's gleeful creationism.

The Laws Of Physics Not Consciously Aware of Human Behaviour, Reports The Australian

There's a clever way of criticising something, without needing to adhere to logic or reality. For example:

"I purchased this vehicle recently. Unfortunately, after a week using the vehicle, I found that it was unable to fly. I deem this vehicle to be a failure"

Set your own criteria. If you really want the subject of your criticism to fail, intentionally shape those criteria to guarantee failure.

It's this fallacy that drives criticism of the generation output of wind energy on hot days. As with all fallacies, it's convenient, simple and palatable.

Alongside the Australian Financial Review and Andrew Bolt, Graham Lloyd, Environment Editor at The Australian reported on the low levels of wind generation that occurred last Wednesday, during a week-long heatwave moving across South-Eastern Australia

"The doldrums that stopped wind power production about midday on Wednesday coincided with warnings from the Victorian government and the National electricity market operator that electricity users faced the possibility of blackouts" 

Note carefully how Lloyd places those two facts in front of the reader, letting them  join the two together: Wind farms weren't generating electricity. There was a potential shortfall of electricity supply. Make up your own mind. Pointedly excluded is the actual text of AEMO's press release (Update 24/01/2014 - dead link, referred to saved PDF of their press release), in which they explicitly state the reasons for the potential load-shedding:

"A combination of intense weather, high electricity consumption and some unplanned outages on the generation network has triggered the potential for load shedding to occur in parts of South Australia and Victoria."

Briefly mentioned by Lloyd is the fact that the unplanned outage of AGL's Loy Yang A coal-fired power station contributed significantly to difficulties in meeting demand:

"Dr Napthine said the problem was exacerbated on Wednesday by one of the four generators at Loy Yang A power station breaking down, and the Basslink cable between the mainland and Tasmania not operating at full capacity for technical reasons."

reported The Age, on the 16th of January. As it happened, I managed to snap a screenshot from market monitoring software, showing the unplanned outage:

So, should we be tearing down coal-fired power stations, because they sometimes go offline? No, we shouldn't, and the same logic applies to wind energy.

According to Lloyd, the fact that wind farms were not generating electricity on a hot day constitutes a 'failure'. Logically, the only way Lloyd could hold this belief is if he assumes that wind speeds are somehow 'meant' to be higher during periods of peak demand - perhaps the laws of physics that govern atmospheric movement are somehow consciously aware of how much electricity human beings are consuming. Facetiousness aside, Lloyd obviously feels that, as a fuel type, wind has no place in a diversified energy system. This isn't based on sound logic.

Wind speed varies according to a variety of things, including pressure, local conditions, temperature, the rotation of the Earth and seasonality. It's never been claimed that the movement of the atmosphere will precisely match electrical demand. But, at times, wind generation is high, and at those times, we can ditch fossil fuels and rely on the output of renewable energy.

Without picking particular intervals, let's have a look at wind generation the whole working week - Monday 00:00 AEST to Friday, 23:59:59 AEST (non-daylight savings time):

This pattern is precisely what we expect from generators subject to variations in wind speed. Here's the same data, with the intervals discussed by Lloyd in the third paragraph of his article highlighted:

When we look at the unfiltered data, it becomes clear that the articles declaring the 'failure' of wind energy are hoping desperately to clamp blinders on their readers - presumably, the reason they've focused on a short time period on Wednesday.

The total output of wind energy varied quite a bit over that time period. Thursday and Friday were much better; Wednesday wasn't so great. Here's the kicker: every single half-hour output of wind was forecasted days in advanced by AEMO. Through their own proprietary wind forecasting system, they plan ahead, and issue warnings accordingly. I'd be interested to see if Lloyd can show where the 500 megawatt coal-fired power outage was forecast.

We need to wean ourselves off our addiction to fossil fuels. One might posit that they're cheaper, but they're not - we pay the price down the line, and it's not worth it.

Some desperately need you to feel panic and anger, when you chance upon the increasingly successful integration of renewable energy into the electricity market - overseen by a capable, logical energy market operator.

Generation by fuel type in South Australia - sourced from AEMO's Market Management Systems Database. 

It's not a wholesale, instantaneous replacement of thermal generation, but it's regularly crowding out fossil fuels - South Australia being an excellent example of this. It doesn't happen 100% of the time, but it happens regularly, and AEMO forecasts this with great accuracy.

In the future, Lloyd might do well to take a more nuanced approach to the complex world of the National Electricity Market.

Want to have a play with the raw data? Click here for the Excel file - but be careful, it's 112 megabytes; make sure your computer can handle it.

Update 21/01/2014 - I've been mucking around with the raw data, and have created some additional charts from the week using DataWrapper - see below. Enjoy! And, get in touch with me if there are any mistakes, on Twitter.

Wind generation in South Australia, daily, as a percentage of total generation in the state:

Wind generation in South Australia, at a 5-minute resolution, as a percentage of total generation in the state - note that this peaked at 42.3% of total South Australian generation at 05:00 AEST, Friday 17/01/2014:

Renewable Energy's % contribution to the entire NEM - keep in mind that installed capacity is a small % of total installed generation, so even if wind and hydro were consistently at full capacity, this would still be a small %'age of total generation: