Australia's renewable energy future
Wind Energy: How it works and where is it going?
Tuesday, 2 June 2009
Associate Professor David Wood
David Wood obtained his undergraduate and masters' degrees from Sydney University and a PhD in aeronuatics from Imperial College, London in 1980. From 1981 to 2004 he was on the academic staff of the Faculty of Engineering, University of Newcastle, apart from a year spent as a senior research associate at NASA Ames Research Center, California. In 2004 he formed Aerogenesis to commercialise the small wind energy technology developed over many years of research and development at the university.
The frist Aerogenesis product, a 5 kW wind turbine, is currently being demonstrated in a $495,000 program supported by the Australian Government under the Asia Pacific Partnership for Clean Development. During 2009, Aerogenesis will receive $750,000 from the NSW Renewable Energy Development Program to install 40 of its turbines in NSW. A similar project for 50 to100 turbines has been approved by the government of West Bengal, India. Other projects are under negotiation in India, China and Africa.
Wind energy: How it works and where it’s going
(Click on image for a larger version)
Wind energy, how it works and where is it going? I'm an engineer, so I am going to be telling you more about how it works than where it is going. You will find that this is a very personal view. I think anyone who gave this lecture would end up giving a personal view of wind energy. It's such a vast field that it's hard for any one person to be able to cover the lot of it.
I'm going to talk about large wind turbines, such as the one shown here, and the ones that are being installed at the Capitol Wind Farm near Lake George.
Typically a large wind turbine has an installed cost of about $2 per watt of generating capacity. That price is coming down slightly, but not rapidly.
I'm also going to talk about small wind turbines, largely because that is my particular area of interest. I will preface that by indicating that at the moment we're building machines that have an installed cost per watt of generating capacity about three times that of large turbines. So we've got a fair way to go. I'll be telling you something about that path.
(Click on image for a larger version)
This is a little bit of a history lesson to show my involvement over the years with wind turbine technology. I was at the University of Newcastle for many years. We had a number of prototype turbines at the university. In 1999 we put that one [shown on left of slide] up on the roof of the engineering building at the university. And I have a number of blades from that turbine here, which I shall pass around, [passes blades of the turbine] because I will be asking you which way they rotate and which is the front. And there might be an exam on it.
Since 2005 it has morphed into the company Aerogenesis, even though we keep very close ties with the university.
(Click on image for a larger version)
One of the nice things about small wind turbines is that you can bring lots of bits and pieces with you. Here are a number of generations of blades and some bits of blades there. I've brought a generator with me as well. When you work on 40‑metre wind turbine blades of course it is a little bit harder to bring them along to a lecture.
The topics that I am going to cover – and this is not going to be the particular order of the topics – I am going to talk quite a lot about how wind turbines work. Being an aerodynamicist, you are going to get a little bit of a lecture on blade aerodynamics because it is the blades that drive everything else in a wind turbine, literally, metaphorically and virtually every other way.
To understand how a wind turbine blade works we need to know about lift and drag – I'll talk a bit about that – performance and power curves. Then a pretty big mouthful: multi‑dimensional blade optimisation. I am going to talk about a particular example that we did to come up with this particular blade design. Even though it is a specific example, it is typical of how modern wind turbine blades of all sizes are made.
I'm then going to talk a little bit about generators and controllers, some features of small wind turbines and finish up with some applications and some future trends.
(Click on image for a larger version)
How do we determine how a wind turbine blade works? The standard way of doing that is to use a theory called blade element theory. The idea is that you take a wind turbine blade and you divide it up into a number of so‑called elements. You then determine the power that is extracted from the wind by each of those blade elements and you sum over all the blade elements, and that gives you the total power that is extracted by the blade.
There are some important bits of terminology. Lower-case c is the chord of a blade element, just as we talk about the chord of an aircraft wing or the chord of an aerofoil. And capital R is the tip radius of the blade.That particular blade has been divided into five blade elements. In some of the analyses that I will show you later, we would use 15 to 20 blade elements. Generally, using more gives you higher accuracy.
(Click on image for a larger version)
And the fundamental assumption of blade element theory is stated at the top of this slide. If we consider a blade element at any radius r – and little r is the blade radius which goes from zero at the hub to capital R at the tip – we assume that it behaves as if it was an aerofoil at an angle of attack that is defined by alpha.
The chord line that joins the trailing edge to the leading edge of the blade is shown there.
The velocities that the blade sees are the velocity in the direction of the wind, the wind speed at the blade and the rotational component of the blade speed. So what that diagram is meant to show you is if the wind is coming from the audience to me and the blade is rotating like that, then that is a blade element diagram for that particular blade. So you should be able to answer the question 'which direction of rotation?' fairly easily.
Using Pythagoras we get the resultant velocity here and the angle between that resultant velocity and the chord line gives us the angle of attack.
The major parameter that governs the aerodynamics of a blade is a thing that I have defined at the bottom, called the 'tip speed ratio'. This is the circumferential velocity of the blade tip, which is omega, the rate of rotation of the blade times the tip radius divided by the wind speed. Lambda, the tip speed ratio, is important because it controls the aerodynamics. I will explain that in a subsequent slide.
(Click on image for a larger version)
Just a little bit about typical aerofoil lift and drag behaviour. I have done it in terms of the coefficients of lift and coefficients of drag, which I won't define, because I think I'm loading you up fairly heavily anyway. But they are shown here [referring to first chart]. Typically the lift varies linearly with the angle of attack and the drag is relatively low for small angles of attack and then rises rapidly as the angle of attack increases.
What I didn't point out [refers back to previous slide] from the previous slide is that if you take account of this arrow here [top of the diagram] which shows the direction of rotation of that blade element, you will see that there is a component of the lift in the direction of rotation which is pulling the blade around. It is that component of the lift that gives the torque acting on the blade element which can sum overall blade elements to give the torque acting on the blade.
The lift acting on the blade element as an aerofoil gives the torque which translates to the power extracted from the wind. Notice that the drag also has a component, but that component opposes the rotation of the blade. Lift is good. Drag is bad. Almost back to Animal Farm.
So really what is important, as far as the aerodynamics is concerned, is not the individual values of the lift and the drag but the ratio of lift to drag. The higher that ratio the better the aerodynamics of the blade is going to be, because you are going to get a greater net force or torque in the direction of rotation.
Going back to our diagrams on aerofoil lift and drag, I'll now bring your attention to this one here [last chart] which shows the ratio of lift to drag as a function of the angle of attack, which I have defined, and also as a function of the thing called the Reynolds number, which I deliberately haven't defined, again trying not to make this too difficult to absorb in one go. I had to learn all this stuff over about a 10‑year period. So it is a bit easier than one hour.
All you need to know for the moment is, generally speaking, the smaller the turbine the lower the Reynolds number. And what you see in terms of the lift to drag behaviour is that we can get some pretty good values of that ratio.
Here we have a maximum value of about 90. If you go up in Reynolds number, up to values that are typical of large blades, that ratio gets up to 180 or 190. So aerodynamicists have done a really good job in designing aerofoils that give a high lift to drag ratio.
The penalty that we pay is that that ratio is Reynolds number dependent. So when you are someone like me, who is trying to design little blades, then we are living at relatively low lift to drag ratio. So in many ways designing blades for small wind turbines is harder than designing blades for large ones.
(Click on image for a larger version)
The next quantity that I want to talk about before we go on to overall measures of turbine performance is a thing called the blade pitch. As I have defined it here, the blade pitch is the angle between the chord line and the plane of rotation of the blade.
To control the angle of attack, that pitch has to vary along the blade. Because if you have the situation where r times omega is small – which is down near the hub region of the blade – then you need a large pitch to get a low angle of attack.
As you go out along the blade r times omega gets large. Typically tip speed ratios are in the region of 7 to 10 for modern wind turbines. So the effect of going out from the hub to the tip makes a large change to this quantity here, which means that the pitch of a blade has to decrease as you go out along the blade.
If you have a look at those blades being passed around you will see that down near the hub region it is pitched at about 25 degrees. Out near the tip it has about zero pitch. And the reason that we need to control the pitch as the r times omega term changes, is to keep the angle of attack at a value that gives us the maximum lift to drag ratio. So there is a very quick tour of the basic aerodynamics of wind turbines.
(Click on image for a larger version)
In terms of operating parameters, there are two important ones. There is this one here [top chart] which I call the performance curve. It is a plot of a thing called the power coefficient [against the tip speed ratio]. If you know anything about wind turbines you will know about the power coefficient. Think of it as efficiency. The plot shows why the tip speed ratio is critically important for wind turbine performance. If the tip speed ratio is too small the efficiency will be low.
As the tip speed ratio increases we reach an optimum efficiency. And that optimum efficiency is the tip speed ratio at which all blade elements are operating at their optimum angle of attack. That point [peak of curve] is the primary goal of most wind turbine blade designers. We want to get the maximum efficiency out of our blade. The way we do it is through manipulating the blade element theory, as I discussed before.
Typically, lambda is in the range of 7 to 10. The turbine that these blades come from [showing blade to audience] is a three‑bladed turbine. It has an optimum tip speed ratio of about seven. The turbine that this comes from is a two‑bladed turbine and has an optimum tip speed ratio of close to eight.
When I am talking to other people who are interested in aerodynamics and blade design we spend a lot of time talking about stuff like this. But if you go to the website of wind turbine manufacturers, or you see their brochures, what they actually show is the power curve; the power output as a function of wind speed. Because this is the bang that you get for your buck. This is how much power the wind turbine will produce at any particular wind speed.
There is a wind speed called the cut‑in wind speed [lower graph on slide], which is the lowest wind speed at which power is produced. From there to the rated speed, there is a very rapid increase in the power output. That shows that it is important, as far as possible, to put your turbine in an area that is as windy as possible. Because incremental increase in the wind speed really pays off in terms of the turbine economics.
(Click on image for a larger version)
So that is a very brief, and hopefully not too confusing, introduction to wind turbine aerodynamics. Where are turbines headed? Basically the big turbines are getting bigger and bigger. We're sitting somewhere about here [the dotted line on slide]. The turbines that are going into the Capitol Wind Farm have roughly 80 metre towers with 40‑metre long blades. They were leading edge turbines about four years ago. From 1980 onwards the trend has been to larger and larger turbines, basically for two reasons. One is that there are some economies of scale. So when the technology matures typically the larger the turbine the cheaper it is. Also, a lot of the largest wind turbines are designed for offshore applications in Europe and the US. When you go offshore, installing and maintaining the turbines becomes a much higher cost or proportion of your cost. So you want to use the biggest wind turbine that you can. So that's basically what's driving the large wind turbine sector.
What I'm going to be talking about are turbines down in this region here. In a sense we can view Aerogenesis' activity in small wind turbines in terms of running between the legs of the larger ones. We are trying to fill up the gaps that the big players have left behind them.
(Click on image for a larger version)
I mentioned that the Capitol Wind Farm near Lake George has 80 metre towers and 40 metre long blades. It would be fairly hard to get a tip section of one of those blades into my Mazda 3.
(Click on image for a larger version)
If that is the case why have small wind turbines? This is where the presentation starts to get highly personal. The obvious application is remote area power systems. If you live away from the electricity grid, typically in Australia it costs about $15,000 per kilometre to get the power lines extended to your house. So you don't have to be too far off the grid to have a remote power system – which might include a wind turbine, photovoltaic cells, battery backup, possibly a portable generator – to become economically viable.
The trouble with that market – and it extends to village electrification in Asia and Africa – is that it is relatively small and it is a one‑off sale usually for each application. We're a small company, struggling to get off the ground, so we've actually taken a slightly different tact. We are not ignoring the remote power market but we are also looking very carefully and making some inroads into direct grid connection of small wind turbines.
In other words, small wind turbines doing at small scale exactly the same job that large wind turbines do. As I pointed out, there is a factor of three difference in installed cost. So it is a pretty big challenge. I will tell you a little bit about how we are trying to meet that challenge.
(Click on image for a larger version)
Here is a picture of the three‑bladed turbine that has these blades on it, the ones that are going around the room. That's installed on the roof of the Engineering Building at the University of Newcastle. We have been testing and monitoring that for about 10 years. This is the permanent magnet generator that comes from that turbine. I will talk a little bit about that later. We have had a number of PhD and masters students work on this project.
I am going to talk now about a particular feature of wind turbine performance that we have to worry about for small wind turbines, and that is the low wind speed performance.
This particular turbine, if you determine its power curve, if you measure the power versus wind speed, you will see that it has a cut‑in wind speed of 3.5 metres per second. That's pretty good for a small wind turbine. The main reason it is that value is that we use a permanent magnet generator. This is a generator that has permanent magnets attached to the stator.
The magnets require a positive torque to get the generator rotating. It's a very small torque in absolute terms, but it's enough to stop the blades turning until a wind speed of about 4.5 metres per second.
If the blades are turning, if the wind is blowing strongly but then starts to slow down, the blades will keep rotating until about 2.5 metres per second. So what we measure as the cut-in wind speed is actually the average of those two.
We've actually paid a lot of attention – and I'm talking about a research and development program that's gone on over 10 years – in order to understand the cut-in wind speed, and look at ways in which we can reduce it.
(Click on image for a larger version)
Low wind speed performance is important for small wind turbines. Particularly for the remote power application the turbine goes where the power is needed. It doesn't necessarily go where it is windy.
A turbine that starts quickly is ready to harness the available power. And, not to be laughed at, if a customer knows that the wind is blowing and their blades aren't rotating, they'll be on the phone to you. So it's actually in your interest to sell wind turbines that have good low wind performance.
Something else that makes life very difficult for us is that small wind turbines typically have no pitch adjustment. Large wind turbines do have pitch adjustment. That means that as the wind speed is low the blades are pitched into the wind. As the wind picks up the blades pitch out.
I only know of one small wind turbine that has that arrangement, and that turbine is no longer made because it was far too expensive. The consensus is that at small scale you can't afford to do pitch adjustment.
So that means that when this blade is stationary and trying to start rotating the angles of attack on the blade elements are very high. So the lift is low and the drag is high. And getting it started is a major, major exercise.
(Click on image for a larger version)
This figure shows exactly that difficulty. It shows a typical starting sequence that we measured on the turbine with these blades [indicating the small blades].
On the left‑hand axis is the wind speed. So you can see that the wind speed is varying. It starts at about 2 metres per second and then goes up to somewhere between 4 and 6 metres per second, with an average of about 4 metres per second. The axis along the bottom is time. The other vertical axis is the rotor speed, the rpm of the blade. The blades start at rest; so zero rotor speed at low times. We are measuring how the rotor accelerates.
You can see that it is just sitting there doing nothing at 2 metres per second, which is not unusual. Then there is a gust of wind that goes from 2 to about 6 metres per second. That's enough to kick the blades around.
You can see the very noisy looking trace there. It is our measurements of the blade speed. But once the blades have started rotating you will see that it takes 86 seconds from initial rotation till the blades reach operational speed where power extraction starts.
During that 82 seconds there is no power extraction. So we are wasting the energy that is in the wind. So not only is this a long time to get going, but it's a waste in time. You'll see also from that figure that we've come up with a blade element analysis of the starting performance of that blade which does remarkably well. There's three different predictions shown here, and all I'll say about those is that their difference is the way in which we treat the lift and the drag at high angles of attack.
If we get that right we actually have very good predictive capability of the starting performance of the blade. So about five years into a 10‑year research program we got to this point. This is taken from a PhD thesis that was submitted and accepted in 2004.
(Click on image for a larger version)
But how do we use that in terms of blade design? Now I'm getting into the area of multi‑dimensional blade optimisation, which occurs in different ways for blades of all sizes.
What we observe from our analysis of the starting is that it for the torque that starts the blade; surprisingly at first sight, it’s actually generated near the hub region of the blade. So when the blade is starting it's the torque on this section of the blade that gets it going. When it is at operational speed, the power producing torque is coming from this region here [indicating the region near the blade tip].
I have made a comment there, that I won't elaborate, that we need wide chord and thin aerofoils near the hub. I will show you why that is the case later.
That we can identify different regions of the blade for those two different performances gives us hope that we can actually do some form of multi‑dimensional optimisation of the blade design. We should be able to optimise a design both to make the blade start rapidly as well as have high efficiency.
To give you an indication of what is involved there is the fundamental blade design first of all requires selecting the aerofoil that you are going to use for the blade section and then after that working out the chord and the twist of the blade. So that's what's required in blade design.
(Click on image for a larger version)
This is where the fun begins. This is stuff that we started about five or six years ago. How do we do multi‑dimensional optimisation? There are a number of different ways of doing it. We have chosen to use a technique called computational evolution. Computational evolution is a computer program that mimics the process of evolution. It is a burgeoning area in engineering and science optimisation studies, because with these techniques you can tackle optimisation problems that you can't in any other form. Ours is not a particularly difficult problem. In fact, it is quite easy. Also, the method that we use is quite easy.
Our aim is to design blades with high efficiency and low starting time using a technique of computational evolution.
Usually we use a population of 2000 blades and we randomly assign chord and twist distributions to those blades. We end up with some very strange looking blades from doing that. We then, for each of those blades, we breed a subsequent generation. We have techniques to select parents and techniques to determine how those parents pass on their equivalent of genetic information.
I haven't given any details of that because there's about a squillion different ways of doing this, and I don't want to get bogged down in the details of reproductive processes, because I don't think it really matters as far as our optimisation is concerned.
So we breed a potential next generation. We then look through our potential next generation and decide whether the new blade is fitter than the old blade. We're doing this on the basis of survival of the fittest.
The way we assess the fitness is a combination of minimising the starting time and maximising the power coefficient or the efficiency. That is done using the equation that I have shown there [indicating equation on the slide]. The parameter a can be varied from zero to one, as I will show you.
Then if a member survives 20 generations we kill it. So if you suffer from hubris, this is almost as good as a computer game.
Then we go back to step two [shown on the slide] and repeat that, typically 200 times. We let the blades evolve over 200 generations. So we set up a Galapagos Island of small wind turbine blades. We go away and we sail back after 200 generations to see what's happened.
(Click on image for a larger version)
What happens is shown as an artist's impression by this drawing here. Each of the final population of the blades will have a particular efficiency and a particular starting time. Each of those red squares is meant to indicate a member of the final population. What actually happens in practice depends on the value of a – no. It happens that for virtually all values of a the blades actually converge to be very tightly bunched. That doesn't make for a good diagram.
What we are interested in is the blue line here, which is called the Pareto Front. The Pareto Front has a very technical definition that I won't go into, but basically it defines the optimum blades that we get out of this process.
On the right‑hand side there is something that we have observed every time we've done an exercise like this – and we have done many, many of these – that the most efficient blade is nearly always the slowest to start. I don't have a good explanation for why that's the case, but in practice that is.All the other blades, all the less efficient blades have progressively better starting performance. So what we do with blade design is trade off. This is where the multi‑dimensional optimisation comes in. If we design our blades only to maximise power extraction efficiency they are going to perform badly at high wind speeds. So we actually give up a little bit of power extraction capability for a very large decrease in the starting time. That's basically how we design our blades. We think that that's good enough to warrant patenting the blade. We are in the process of doing that.
(Click on image for a larger version)
Then we have to make them.
(Click on image for a larger version)
This is expensive because we machine the moulds for the blades on a computer‑controlled milling machine. We make the blade moulds in two halves, so there is a top half and a bottom half.
The picture on the right shows the method of manufacture of one half of the blade by a technique called vacuum infusion. This technique is used by the vast majority of large wind turbine blade manufacturers as well.
The idea is that we lay the fibre glass used for reinforcement into the mould, using up to nine layers of fibre glass. Then we cover it with this blue stuff which helps to release the blade [after manufacture], and our secret ingredient, a resin runner which promotes the flow of resin through the mould – it is actually shade cloth – and then we cover it all with a vacuum bag. Its technical name is tacky tape; it seals the vacuum bag down onto the mould.
You can't see in this slide here, but the resin comes in over this side [bottom left of image on right]. And up the top here you can see two exit tubes [near the person] from the mould, which suck the resin out of the mould. The darker green area is where the resin already is and the lighter green area is where it is going.
So we make up a little pot of resin, feed it in through the tube here and suck it out at the top. After about 20 minutes the resin has gone throughout the blade and when it cures we have a blade half.
Here [holding up a mold] is a blade half as it comes out of the mould. I won't try and hand this around. You can come and have a look at this later, if you would like to. You can see the resin tube where most of the resin flowed through here. You can see one of the outlet ports here and the other one is down there.
So we make each blade half in that way. And then we machine foam cores. I brought one of the cores along with me. You can have a look at that later. We glue the two blade halves together with the foam core in between. Here is a piece of a blade that is made that way. It produces an extremely strong blade.
(Click on image for a larger version)
If you are making a large wind turbine blade the technique is very similar. The only difference is that they don't use foam to fill up a gap. They use a box structure which is sometimes called a shear web. There is a whole variety of different methods to do that. So having made our blades we have to ensure that they are safe. We have to test them.
(Click on image for a larger version)
This shows a comparison between our 5 kW blade design, at the top, and one of the largest of the current large wind turbine blades, the 61.5 metre LM Glassfiber blade. You can see in the tip region the big blade comes to a sharp point. That is basically to minimise noise due to the tip vortex production. All large wind turbine blades are shaped like that to minimise the tip noise. For large wind turbine blades the multi‑dimensional optimisation is often balancing performance against minimisation of noise.
The big difference then is down in this region of the blade here [indicating the region near the hub]. Remember that's where the starting performance comes from, so we have much wider chord in that area than does the larger blade. So that is just the comparison.
(Click on image for a larger version)
We go through an extensive test program. This is what we have to do to test the blades. The one on the left is called the bird test. If you have wooden blades and a wooden tail fin, you have to make sure that the local population of sulphur crested cockatoos doesn’t chew them. No bird was killed during the filming of this. We also do static tests. Here's an example, probably the test of one of the blades that is being passed around.
It’s [indicating middle photograph] a graphic demonstration that timber is a wonderful material for wind turbine blades. Large blade manufacturers do very much the same sort of thing [indicating right hand photograph].
(Click on image for a larger version)
Then we do fatigue tests. You can see the yellow blade holder and one of our 5 kilowatt blades on the machine. This mechanism is shaking it up and down. We simulate a full 20‑year lifetime of our blades in that accelerated fatigue testing program.
This [holding up a blade to the audience] is an earlier generation of our blades. The very first one that we fatigue tested. You can see a massive fatigue crack down near the root section of the blade that occurred after about the equivalent of seven years. That's another one of those blades after a big storm. You can see it failed in virtually the same place that the fatigue crack started. That's a pretty graphic demonstration of the need to do extensive fatigue testing of blades.
(Click on image for a larger version)
Moving on now to relatively boring topics for aerodynamicists: generators and controllers. I'll go very quickly through this. Virtually all turbines – this is a generic situation for wind turbines – produce variable frequency and variable voltage AC, usually three‑phase. That typically gets rectified, and in our case boost converted for grid connection, to produce standard 240 volts, 50 cycle AC. We go through an inverter to connect to the grid.
I am conscious of time, so I won't say anything more about that, but I would be happy to answer to the best of my ability any questions on that later.
(Click on image for a larger version)
Moving on to our turbine. That’s it, sitting on an 18‑metre tower. This is our first demonstration turbine funded with a significant grant under the Asia Pacific Partnership for Clean Development.
(Click on image for a larger version)
These are the sort of applications that we are looking at. This is a new area where small wind turbines can be used in appropriate parts of the urban landscape. Not every part, but appropriate parts to direct grid connect and produce electricity for the grid. As I said before, we are not avoiding the traditional market but we have headed in this direction partly because we have been successful in getting a large grant from the much maligned New South Wales Government to subsidise the installation of 40 turbines by mid next year in these sorts of environments. And we have, hopefully, a similar project in West Bengal for a similar number of turbines.
What is needed to promote these applications? There are two main justifications. One is that if you use distributor generation – that is, placing around your electricity grid a large number of small scale renewable generators – generally speaking, you improve the quality of electricity that is delivered. There is a number of fairly technical arguments behind that, most of them are electrical engineering ones, so I am not all that familiar with them.
Also, if there is enough uptake of renewable generation by this method, that delays or avoids the need for future coal‑fired power stations. In order for the technology to succeed, in order for this to happen, we need a feed‑in tariff regime. A feed‑in tariff, if you don't know, is a premium on the buy‑back price for particularly small scale renewable energy generation. That is usually mandated by state or federal government.
The rather sad situation in Australia is that there are a number of jurisdictions that have feed‑in tariffs – Queensland, Victoria, South Australia and probably Western Australia – that apply only for PV [photovoltaics]. Australia is the only country in the world where this daft policy exists. I think it is partly to do with the muscle of the PV industry in Australia. But it is also to do with the fact that the small wind turbine industry has been largely silent on the issue.
The only decent feed‑in tariff legislation in Australia is in the ACT. It is a legislation that recognises all forms of renewable energy generators and has a high tariff markup. On the small scale it is nearly four times the normal buyback price. So if you put PV on your roof, for example, in the ACT then you will get that significant buyback price.
The other nice thing about the ACT legislation is that it scales down – that factor scales down as the generator size increases. It also means that turbines like ours can be used for schemes like community power schemes and things like that. In other words, it is not limited to residential applications, like most of the other feed‑in tariffs.
(Click on image for a larger version)
What we say is that when we lobby the appropriate governments – and we have been doing that to the best of our ability, so far fairly unsuccessfully, but you have to try these things – is that at the moment both small wind and PV are not mature technologies. Their cost per watt of installed capacity is high, as I pointed out for wind at the beginning of the lecture.
In order for us to get going we need a relatively higher renewable energy or feed‑in tariff. We also say that regime will build a market for us. If we build a market our production volume increases. As our production volume increases the cost per unit turbine goes down.
So it is possible, and highly reasonable, for the government to say yes, we will give you a generous feed‑in tariffs starting now – 3.88 times the normal buy‑back price for the ACT – but beware, if you do it next year the tariff will be lower or the factor will be lower. The ACT legislation actually has that option in it.
That makes it much more politically acceptable. But it also is a highly reasonable thing, because if we can get to mass production and reduce our production costs significantly then the cost of our product goes down.
We are looking for some help to get us going here, so that in 5 or 10 years' time, we will be down somewhere about here [on the graph], at which point we can say that our technology is mature.
(Click on image for a larger version)
A lot of people have contributed to the research and development and the commercialisation that I have been talking about. I have listed most of them here. I would also like to thank the relevant governments for their support, which has been absolutely critical to our progress. And finally, I would like to thank you people for voting in a very sensible government in this jurisdiction to legislate a feed‑in tariff, which I hope will mean that you will see appropriate installations of our five kilowatt turbine over the years to come. Thank you.
Discussion
Chair (Sue Meek): Thank you, David. I would like to invite you to put questions to David after that fascinating talk. Could you please wait for the microphones?
Question: Thank you, David. Three questions. Very direct ones. Where are your production costs today? Where are the installation costs today? I know we are stripping down, and where do you think, with decent production, your future production costs can go to?
David Wood: I'm not going to answer all those questions because that is very sensitive commercial information that I would be giving out. The $6 per installed watt capacity comes from an anticipated small volume production cost, retail cost, of the turbine of $22,000. And based on our experience with installing the turbine at Newcastle University, installation costs of the order of $8,000, bring it up to $30,000. Of course installation costs are highly dependent on how close you are to the electricity grid and also the soil type, in terms of the foundations. So that is only an estimate.
The general rule of thumb with volume production is that if you double the volume of production then your production costs go down by about 20 per cent. So it is not unreasonable to expect in the longer term that that $22,000 can come down to at least $15,000.
Question: Can I ask you two questions? You didn't mention that the new renewable energy target which takes over from the MRET(Mandatory Renewable Energy Target), hopefully later this year, will have banding in it where small generation of lots of types of renewables can get a multiple of the renewable energy certificates between 5:1 down to 2:1 over the first five years of the scheme. My first question is: is that likely to give much of a lift to small scale wind generation? And my second question is: with large wind it is obvious that is becoming increasingly dominated by big companies like Vestas and Siemens and there has been a shake out of smaller large‑wind manufacturers who have basically gone to the wall and been folded into the larger manufacturers, leaving the impression that there is not much room in large wind manufacturing for new start‑up companies to get market share. Is that sort of consolidation happening in the small‑wind sector, or is it still fairly kind of competitive and not too monopolistic yet?
David Wood: Very good observations. The RECs [Renewable Energy Certificates] actually do help us. Just to give you an example, with the project in New South Wales to install 40 of our turbines, we are obviously looking for partners to take those turbines. Most of the three‑quarters of a million dollars that the state government is providing for this project goes to subsidise the turbine. So it will mean that the partners will pay an install cost of the turbine of about $15,000.
Now the reason I say that is that if you then do the net present value costing, however you want to do it, of the wind turbines with the current buy‑back pricings in New South Wales, then the value that you get from the WRECs, which, as you said, is multiplied for small‑scale generators, actually tips it from just being unfinancial to just being financial.
So it is actually pretty significant for us at the moment. It is not enough to make our turbine competitive without a significant subsidy, which is what a feed‑in tariff should do, but it does help.
The issue with start‑up companies for large wind – you have hit the nail on the head – when we started in wind energy 20‑odd years ago we made the conscious decision not to be involved with large wind turbines as our primary objective for that reason. That was an obvious reason, even 20 years ago, that the Europeans were basically doing it very well and unlikely to be beaten. There is a start‑up company in New Zealand that is trying to develop a large wind turbine, and as far as I know they are struggling more than we are.
Now the situation with small wind turbines is quite different. It is different partly because the level of technology for small wind turbines is still relatively poor. The way I describe it is that where we are with small wind turbines now corresponds to where the large wind turbine companies were about 15 or 20 years ago in terms of their technology.
Just to finish up, there are actually a large number of start‑up companies in the large wind industry. They are all in China and India. They are getting by both through a combination of government support for them and also because they are doing their own multi‑dimensional optimisation. They are trading off low labour costs against sophisticated manufacturing methods. I have just come back from China. I visited five large wind turbine manufacturers. Their general product level is comparable to what the Europeans were doing about 10 years ago.
So it is possible, but I think you need the immediate market to justify that. We don't have that.
Question: I wonder if you could comment on a Dutch company called Home Energy International, which is producing a very novel type of small wind generator which they call Energy Balls, where the turbine tips are curved back onto the rotor, in a sort of eggbeater type design. And they claim they are 40 per cent more efficient than the sort of turbines that you are producing, produce much less noise and have a cut‑in speed at two metres per second?
David Wood: I don't know that particular company, but I would be highly skeptical of those claims. The chief advantage of vertical access wind turbines, in other words, ones that rotate either vertically or in the horizontal plane, is that they are easy to attach to buildings. You will probably realise that I said nothing about building-mounted turbines in my talk. That is not our area of interest. Most of the leaders in that area use Vertical Access Machines. And the main reason is simply that it is easier to attach to a building. I have never seen any reasonably detailed and believable measurements of vertical access wind turbine efficiencies that come anywhere near that of the conventional propeller turbine. And there are very good theoretical reasons why they can't.
Question: I would like to ask a question about a possible future use of wind power. The French have designed recently a compressed air car for zero emission urban transport, which has a range of about 100 kilometres. Can you envisage a technology where wind power could be used for storing wind power in cylinders, so that you could use it for urban transport, and particularly the small area for rural transport in tractors and so on?
David Wood: In principle, yes. Because I think that's also relevant to ideas about the hydrogen economy as well, that you use renewable energy generators in general to produce hydrogen. There is no reason why wind turbines can't be used to compress air. The advantage of those sorts of technologies is that provided you have enough back‑up you can live with the vagaries of the wind.
So, as long as you have got enough compressed air cylinders topped up then it doesn't matter so much if the wind doesn't blow or the sun doesn't shine for a couple of days. Yes, I think in principle those sorts of ideas are eminently reasonable for renewable generators of all types and all scales.
Question: Mine is a slightly different question. I'm curious to know why the design of the blades for power generation is so different from those for aircraft?
David Wood: Well, they're not really. Because the basic requirement of maximising lift to drag is the same for an aircraft wing as it is for a wind turbine blade. Perhaps a better comparison is between a wind turbine blade and a propeller blade, and actually wind turbine blades look pretty much like propeller blades. They are sort of inverted and back to front. And it is very easy to use exactly the same blade element analysis for propellers as it is for wind turbines.
Really, the major difference between an aircraft wing and a wind turbine blade is that if this was a wing of an aircraft the velocity that that wing sees is pretty well constant as you go out along the wing.
If that is a wind turbine blade then the rotational velocity here is zero and at 7 to 10 times the wind speed up here. But the basic principles are very similar. Aerofoil lift and drag is a key component of wing design, propeller design and wind turbine blade design.
