Anatomy of a Wind Myth

August 23, 2010

Crossposted from my work blog at EmbraceMyPlanet, this is lengthy examination of the bizarre world of the anti-wind industry. This article is written in my professional capacity.

There are plenty of myths around wind power – variability, noise, costs and so on. But where do they come from? We’re going to analyse one particular myth to find out.

The myth we’re going to look at is one that’s been hanging around anti-wind sites for some time – we’ll call it the No Displacement myth. It’s the belief that because the variability of wind power results in fossil fuel power plants needing to raise and lower their output rapidly to compensate for the variations in wind power, the extra carbon emissions caused by this actually outweigh the emissions displaced by using wind power.

A key current advocate of this myth is the website MasterResource. MasterResource is a free-market energy blog with a range of high-profile contributors, including Robert Bradley Jr., an adjunct scholar of the libertarian think-tank The Cato Institute and a former Director of Public Policy Analysis at Enron. The ideological underpinnings of this site can be demonstrated in Bradley’s belief in the economic doctrine of ‘resourceship’, which claims that resources are not limited by the quantity present in the Earth but rather the ingenuity of man to extract them. This means, in essence, that if peak oil becomes a reality it’ll be because we weren’t intelligent enough to find more – not that there was no more to find. We’ll leave it to the reader to decide on the merits of this doctrine.

MasterResource has published a series of posts by former management consultant Kent Hawkins on the No Displacement myth. They deal with two studies which ostensibly show that the emissions caused by power plant ‘cycling’ (i.e. raising or lowering output to match wind generation) are greater than the emissions that would’ve been caused if the plant was left to generate that amount itself.

The first study is from the Netherlands, and is by C. le Pair and K. De Groot. It’s hosted Dr. le Pair’s personal website, which bears an anti-wind logo at the top, and doesn’t appear to be peer-reviewed. The study raises the valid point that inefficiencies are caused by cycling power plants, which isn’t under dispute. The study claims that these inefficiencies outweigh carbon savings from wind when they rise over 2.5% if wind power makes up more than 5% of the amount of electricity that’s used to meet demand. Their maths, in this regard, seems correct – but the problem arises when they try to prove that these inefficiencies actually happen in practice. They admit that they were unable to find data on this subject, but that doesn’t stop them trying to estimate what the efficiency drop required to make using wind turbines to generate electricity result in more carbon emissions would be.

Their discussion section goes through a series of complex-looking equations to find that the drop in efficiency necessary is exactly equivalent to the amount of electricity the wind turbines actually produce. This seems rather obvious. They then produce a table which purports to demonstrate that a relatively small drop in efficiency for the overall system would result in a significant drop in efficiency for the power plant that the wind turbines displace. This table was compiled with help from… Kent Hawkins.

How to make your own Wind Myth #1: Think Ouroboros – Refer to articles that you helped to write as evidence for your myth!

We’ll go into the roots of Hawkin’s calculator that the Dutch study relied on in a moment. The second study in Hawkin’s series of blog posts is one carried out by Bentek Energy, on behalf of the Independent Petroleum Association of Mountain States. The third page of the report bears the disclaimer: “BENTEK DOES NOT WARRANT THE ACCURACY OR CORRECTNESS OF THE REPORT OR THE INFORMATION CONTAINED THEREIN.”

Bentek focus on the experience of the PCSO company in Colorado, which has added significant amounts of wind power capacity to its electricity grid since 2007. Since that time, the heat rate (the amount of fuel used per unit of electricity generated) has increased by around 1% across the entire set of PCSO’s coal plants. Bentek claim that this disguises the situation at individual plants, and presents two graphs of heat rate from Cherokee Plant in 2006 and 2008. They claim that these graphs indicate increased variability in heat rate – but this would be more convincing if both graphs used the same scale. As it is, the 2008 one looks a little stretched.

The next part of the report focuses on two scenarios from a PCSO training presentation. Both show a rapid variation in wind power over the course of a few hours, leading to coal electricity production being quickly ramped up and ramped down. Why these particular scenarios were picked from the presentation is unclear; there are four scenarios given in the presentation, but the other two show much less dramatic shifts in wind power.

Older coal plants like Cherokee have difficulty coping with their emissions following cycling – their emissions control systems are predicated on maintaining a constant boiler temperature. Every coal generator in Colorado is over 30 years old – Cherokee is over 50. This means that following cycling, emissions control systems can frequently fail and lead to a significant increase in emissions of various pollutants – and this is what Bentek claims happens on the two days it examines. Using the amount of pollution emitted by Cherokee over the full day (rather than just the pollution emitted during the wind variation event), Bentek claims that the extra emissions released by the plant as a consequence of cycling are more than are displaced by using wind. They then claim that this is a strong justification for replacing the coal plant with a new gas plant. This is what the plant’s owners are actually doing, which must come as welcome news to the gas companies who commissioned the report.

This implies that adding wind to an electricity grid that contains lots of old coal plant would not result in many emissions savings. Unfortunately for Bentek’s analysis, the emissions from Colorado’s electricity generating plant have been decreasing since wind came onto the system, by nearly the same amount as the added wind capacity. Wind events of the magnitude used by Bentek’s report are rare – far more common sources of cycling are daily changes in demand levels, which you can see in Figure II-7 in Bentek’s report. Indeed, Bentek did not present us with the changes in generation for the days they analysed beyond the times of the wind event, making it very difficult to claim that only that event caused the rise in emissions. It’s therefore not clear at all that cycling caused by wind leads to a net rise in emissions even from old coal plants. However, this doesn’t stop Kent Hawkins from claiming that using gas to balance wind leads to a net increase emissions.

How to make your own Wind Myth #2: Refer to articles that disagree with you as evidence for your myth and hope no-one checks your references

Hawkins uses a methodology similar to that used by Warren Katzenstein and Jay Apt in their article ‘Air Emissions Due To Wind And Solar Power’. Somehow he neglects to mention that Katzenstein and Apt find that renewable energy does result in a net carbon emission decrease – although not as much as has been previously claimed. Katzenstein and Apt’s article relies on a model containing four wind farms, a solar plant and a single gas turbine generator. For comparison’s sake, there are 51 gas plants in the UK, many containing multiple generators, and 264 operational wind farms.

Using data from a very small sample of wind farms presents a problem for their model – wind power variability from a small geographic model is very different from wind power variability from widely distributed wind farms, as this report for Greenpeace shows. Hawkin’s response to this issue as raised by Michael Milligan in ‘Wind Power Myths Debunked‘ fails to get to grips with the reasons why this has an impact on emissions rising as a consequence of variability. He claims that more wind on the system results in more absolute variability from wind – which is trivially obvious, as there’s more scope for the power output to go up and down. However, what’s relevant in calculating the increased emissions from cycling plant is the speed and the intensity with which wind events happen – as the Bentek report showed. Ramping up fossil fuel plants more quickly results in increased emissions, whereas being able to increase or decrease generation more slowly reduces the heat rate of plant. That’s why a small-scale model won’t properly capture the real impact of wind on cycling emissions.

How to make your own Wind Myth #3: Imply that small-scale models are the same as the real world

To demonstrate how wind power works in the UK, consider this wind speed map. It is immediately obvious that not only is the wind always blowing somewhere, but also that it is not the same everywhere at the same time – which Hawkin’s assertion that wind power is stochastic implies. Rather, wind is a series of flows of varying intensity. To illustrate how this affects the variability of wind generation, imagine you’re throwing tennis balls at a small-scale turbine, causing it to turn round. The frequency by which you throw tennis balls will cause different amounts of generation from the turbine. If you suddenly start throwing them at a faster rate, then the turbine will produce a generation ‘spike’.

Now imagine you’re throwing tennis balls at a series of turbines, one behind the other. Each ball hits the blade of each turbine in turn. If you suddenly increase the speed at which you’re throwing, it won’t cause an instantaneous increase in all the turbine – rather, it’ll cause generation to increase on a slower curve. It’s this system-wide slower increase that gas turbines will have to compensate for – not the sudden spike of generation from an individual turbine. That’s one of reasons why a wide geographic spread of wind farms is so important, as the Danish experience shows.

Given that no real-world data appears to support Kent Hawkin’s contention, what inspired him to take this approach to wind power? According to a piece he put up on MasterResource in February, his derived the information supporting his work from a document put out by the Renewable Energy Foundation called ‘Reduction in Carbon Dioxide Emissions: Estimating the Potential Contribution from Wind-Power’, by former senior Esso manager David White. The article makes the same claims – and same mistakes – as the above, by raising the power plant cycling issue but making no quantitative assessment of its impact. It fails to do so because it claims insufficient research has been carried out into this field – although a UKERC report on the same subject identifies a wide variety of papers available when REF’s 2004 report was originally released.

REF is an odd organisation. Despite its name, the majority of its activity involves disseminating anti-wind propaganda. Its funding comes from a variety of wealthy anonymous donors, only a few of whom have been publicly identified. Those we know about include the property magnate Vincent Tchenguiz, whose Consensus Business Group is the major supporter of the core funding of REF. Mr Tchenguiz has previously publicly discussed the likelihood of increased arms spending by Western governments to guard against the enormous floods of people fleeing countries rendered uninhabitable by global warming. He aims to make himself indispensable to arms companies that will be recipients of this increase in spending by functioning as a co-investor on ‘flow-back’ investments these companies are compelled to make with countries who purchase their products.

Curtailing the development of the most mature renewable technology – wind – will result in higher carbon emissions, increasing the likelihood of dangerous climate change.

How to make your own Wind Myth #4: It never hurts to find a rich businessman who stands to benefit from your myth

Mr Tchenguiz isn’t the only public figure associated with REF. The organisation’s first chairman – who only stepped down in February of this year – was Noel Edmonds. He joined REF following a prospective development near his home in Devon when the organisation was formed in 2004. We can therefore pinpoint the origin of the No Displacement wind myth quite accurately – it comes from Crinkley Bottom. I’m sure there’s a joke in there somewhere…


Lib Dem Voice have put a poll asking Lib Dems to vote on Huhne’s new nuclear stance. The party’s official line is that new nuclear power is not necessary to combat climate change, and is a danger in itself. Since starting work in the energy industry, I’ve realised that’s almost certainly daft. However, a lot of older activists are passionately opposed to nuclear power, and Huhne’s sop to them has been to say that new nuclear plants can be built – but that the state won’t subsidise them.

This is a good policy, although Huhne’s been lambasted from the right for it. We’re shutting down a lot of old plant – both nuclear and fossil fuelled – over the next ten years, and we have an energy deficit coming up. DECC’s Pathways document, released at the same time as Huhne’s Energy Statement, attempts to map out our options to overcome this, while ensuring that we reach our 2020 carbon reduction targets.

Politics: the Right on Energy

The right tends to love nuclear as a solution to carbon targets and the power gap – read the rantings of Roger Helmer MEP for a surprising instance of a Conservative praising the French way of doing things. The reasons given are typically economic – they believe nuclear is cheaper than any of the alternatives. The problem is that this nonsense – as I discuss on the Embrace blog. The cost of installing new nuclear capacity is above onshore wind, and comparable to offshore once the financing costs for nuclear are factored in. It’s therefore not clear that nuclear is cheaper – especially as the taxpayer is liable under treaty for the costs of a nuclear accident.

Sorry, let me say that again: the taxpayer is liable under treaty for the costs of a nuclear accident.

That’s quite a subsidy you’ve got going on there. I’m fairly sure the taxpayer wouldn’t have to pay the costs for a wind turbine falling over.

On that subject, the right loath wind turbines in particular, largely because they’re a totemic symbol of the green movement intruding on their rural idylls where they’ve been consuming resources happily for decades without reality intruding at all. Nuclear allows them to shove all that dirty ‘production’ somewhere out of site, so they can continue pretending that an economy based on ever-expanding consumption has no impact on the world. James Delingpole in particular has a weird pathological obsession with them, viewing them as some sort of Martian intruder into the Arcadian landscapes of his youth. In the north, we’ve been digging up our landscapes for centuries to power our industry. To me, views like Delingpole’s are several centuries out of date.

Mechanics of Energy

If we’re looking for the most economic solution to cross the power gap and reach our emissions targets, you can’t just say ‘Wind Turbines!’ or ‘Nukes!’. This is because powering the UK is an incredibly complicated business presided over by the unsung heroes of our day-to-day life, the National Grid. They have to deal with daily fluctuations in power demand of between 20-25GW (to put that in perspective, the biggest generator in the UK, Sizewell B, produces about 1.2GW). To do this, they need two broad types of power stations: baseload plants and ‘peaker’ plants. The baseload supply the electricity that’s constantly demanded – around 40GW. The ‘peaker’ plants supply electricity to meet peak demand – around 20-25GW, as mentioned. Different types of plant are easier to ramp up and down than others – for example, gas can be switched on and off very fast, while it’s unsurprisingly difficult to tone down a nuclear reactor. The French model Roger Helmer advocates uses 80% nuclear, requiring some of their nuclear plants to run in ‘load-following’ mode, which is difficult and expensive to achieve. They have to do this because putting too much power into the grid at once would produce an effect analogous to a power surge in your house, except across the entire country. It has to be carefully balanced – which is why I said National Grid really were unsung heroes.

Even with load-following nuclear plant, the French regularly dump a lot of their excess electricity onto international markets to avoid overloading the grid – there’s a 1.2GW pipe under the Channel they use to give us electricity. They can do this because they’re heavily interconnected with their neighbours, reducing the risk of overload. We’re not. No-one serious is taking Roger Helmers’ position.

Wind turbines are good for peak power – their seasonal & daily output broadly follows demand trends, with a significant amount of variability. However, because they’re variable, they’re less good at satisfying peak demand. Typically they’re balanced with fast reserve plant, like gas.

You should be able to see the outline of a future energy mix from the above, and this is what I suspect Huhne’s policy is aiming at. Nuclear will receive a subsidy in the form of a carbon tax that will make all low carbon sources of electricity competitive. It will receive the mammoth public liability subsidy that no-one likes to talk about. It probably won’t get an additional subsidy along the lines of renewable energy. If it’s economic after all this, it will be built – but those building it will probably be required to set up a special bond for decommissioning, to pay the eventual clear-up costs. Something similar is required of mining projects, and constitutes a good model for dealing with economically useful but environmentally damaging projects.

How then should I vote?

Go for the top option on the poll – nuclear has enough subsidy to be built with the carbon tax and liabilities factored in. Reserve the right to change your mind if a privately-funded decommissioning bond isn’t required. In that event, a massive expansion of renewables is the most cost-effective option.

Isn’t telling people how to vote on a poll illiberal?

Yes. But so are blackouts. They constrain your ability to choose to have the lights on.