Onshore wind seems to be the lowest cost renewable energy source (perhaps excluding hydro – I haven’t looked into the costs of hydro because it is mostly “tapped out” in developed countries).
Wind power (onshore) is a mature technology – when you buy a wind turbine and install it, you know it’s going to work, and you can have some expectation, at least across a wind farm of many turbines, of your O&M costs. You can have a reasonable expectation that it will run for perhaps 20 years. (Of course, you can’t have certainty on any item like lifetime or O&M costs, but this is true of any piece of equipment).
You also know – based on meterological data for the location – roughly how much energy it will produce. This is the capacity factor – the percentage of energy produced vs the “nameplate” value (the nameplate tells you the output if the wind is blowing at the maximum value).
So if you install a 2MW turbine in some parts of the UK or Ireland, or coastal regions of Europe, you might get 30% of that as annual output – 2000kW x 8760 hours x 30% = 5.3M kWh (written another way – 5.3 GWh annually). If you install the same turbine in some parts of Europe, or other parts of the UK, you might get a little over half that – 3M kWh (3GWh).
You can have some confidence in the annual energy production in advance.
Depending on the regulatory structure in the country/state in question you can have a reasonable idea of how long the process of approvals will take, and the grid connection costs (we’ll come back to grid connection later).
If we compare this with building a gas plant, or a coal-fired or nuclear power station then wind is more “modular” and there is a lot less project risk – if you want to produce say 500 MW annually (4,380 GWh) from a given technology then you could build a 550 MW gas plant (expecting certain downtime each year). From design through to startup might take a few years and there are all kinds of “little” problems that can cause significant delays. All of these can be mitigated one way or another, but many plants are late. It’s just the reality of complex projects.
And by the time of startup, the gas price (your fuel) might have doubled in cost from when the design commenced (also it might have halved). Most “expert” predictions of future gas and oil price ranges are only accurate when the price doesn’t change much, so large fluctuations are difficult to deal with. It’s a bit like predicting the weather tomorrow will be same as today. You are mostly right but how useful is that prediction?
To get the same output of 4,400 GWh from wind farms you need to install around 1,000 2MW wind turbines, depending on location (this calculation assumes 25% average output – 2GW x 8,760 x 0.25). You don’t have to wait 5 years before your investment starts producing energy that you can charge for. You can install around 20 a month for 4 years and be producing energy from month 1.
So there are a lot of project benefits for wind. On the downside for wind, you have to stump up most of your cash at the start, so you are a little more dependent on interest rate risk than a gas plant builder.
There are three main downsides to wind:
- It’s not “dispatchable” so it doesn’t create baseload power – another way to say this is that wind power at significant grid penetrations gets very little “capacity credit” – something needs to provide power when the wind is not blowing, or not blowing much – we looked at this in IV – Wind, Forecast Horizon & Backups and in Renewable Energy I
- A lot of places where you might want to install wind farms there is no transmission grid, so there is a cost which is not usually factored in to wind power costs. Building transmission grids is expensive –VIII – Transmission Costs And Outsourcing Renewable Generation. Another way to look at it is you are constrained to put the wind farms where the wind blows best, rather than at a convenient point on the grid. The same is true of nuclear power, of course – for reasons that are unclear to me they are mostly get built a long way from big cities. Gas and coal power stations can have more flexibility. But the ideal place for a wind farm is often on top of an inaccessible hill with no transmission line for 100km.
- Wind has a very low energy density, so requires a lot of land. We will come back to this important point in a future article. This is why Europe has high projections for offshore wind despite major problems with offshore.. In places with high wind and low population density like, say, Oklahoma this is not so much of an issue.
Of course, many people don’t like wind farms cluttering up the countryside but I’m just going to ignore that. Many people don’t like roads or telegraph poles or coal-fired power stations or nuclear power stations or changing the color of phone boxes (what are they?) or the large quantity of birds killed each year by cats..
This series is about more practical energy considerations like how a grid works, how much power can be produced, what it costs, and so on.
Not hurting peoples’ feelings is for another series.
This numbers I’ve extracted come with quite a margin of error. If you buy one 2MW wind turbine you might pay $X. If you buy 10 you might get a 5% discount. If you buy 500 you might only pay 75% of $X per turbine. One year the prices will be lower because of exchange rate fluctuations and raw material costs. The next year they will be higher. If you negotiate better you might pay 15% less than the next guy for the same quantity in the same month. And so on. This is true for all purchasing. There is no “one price” in real life for most items.
Here I’m just trying to put a stake in the ground so we can get an idea what wind energy costs.
Blanco (2009): the capital cost of installing a 2MW (nameplate) wind turbine ≈ €2.4M, of which just over 70% was the ex-works cost of the turbine. (In 2009 this was about US$ 3.4M with the exchange rate quite high, at €1 = $1.4, vs today around €1 = 1.11).
As explained in the previous sub-section, wind energy is a capital-intensive technology, so most of the outgoings will be made at this stage. The capital cost can be as much as 80% of the total cost of the project over its entire lifetime, with variations between models, markets and locations. The wind turbine constitutes the single largest cost component, followed by grid connection.
After more than two decades of steady reductions, the capital costs of a wind energy project have risen by around 20% over the past 3 years. The results of our survey show that they are in the range of 1100–1400 €/kW for newly-established projects in Europe.
And The Economics of Wind Energy from 2009 has similar data:
So again installing a 2MW turbine and connecting it to the grid costs around €2.4M.
If we look at the capex cost of this wind turbine over 20 years in a many parts of Europe with a capacity factor around 15%, we see that it produces 53M kWh (2000 x 8760 x .15 x 20), so ignoring the cost of capital, a capex cost of 4.5 c€ per kWh. In Ireland and many parts of the UK, with a capacity factor around 30%, we get 2.3 c€ per kWh. This kind of cost is also written as €23/MWh – €45/MWh.
Operations and maintenance cost vary of course. Current estimates seem to be around 1-1.5 c€/kWh or €10-15/MWh. Some of these costs are “fixed” in that they are legal or regulatory so cannot be tied to the energy output while others are clearly related to the energy output (replacing parts, etc). And there is a lot of variability in all of these costs.
So to put this in a different perspective, for our 2MW turbine, running at an attractive 30% capacity in a high wind location, the O&M cost is around €53,000 – €79,000, and over 20 years (again ignoring cost of capital which increases the cost of initial payments vs later maintenance costs) this equates to €1.1M – €1.6M, which is not insignificant in light of the capital cost.
Of course, the long term maintenance costs are quite unclear, as there isn’t a lot of 20-year data on wind turbines, and no long term data at all for current generation products.
The 20-year life is also a value that exists more for planning purposes than a real consideration of actual lifetime. Many conventional power plants were given something like a “30-year life” yet are still operating 50 years on. In those cases, the “lifetime” was more for planning and purposes of financial measurement, rather than the belief that after 30-years they would fall apart. And the “30-year” plants still operating after 50 years may have had a number of expensive refits during that period.
If we sum it up in a “proper financial metric” like Levelized Cost of Energy (LCOE), we need to include the cost of capital and take into account the capacity factor. And then take a view on the number of years the wind turbine will operate.
All this does is obscure the costs, as anyone used to trying to compare the cost of different types of power will attest. So we will stick with raw numbers for now. It makes it easier to compare other forms of energy generation that we will look at in subsequent articles.
The International Renewable Energy Agency (IRENA), 2012 had higher costs:
Installed costs in 2010 for onshore wind farms were as low as USD 1,300 [€1100] to USD 1,400/kW [€1200] in China and Denmark, but typically ranged between USD 1,800/kW and USD 2,200/kW [€1500-€1800] in most other major markets. Preliminary data for the United States in 2011 suggests that wind turbine costs have peaked and that total costs could have declined to USD 2,000/kW for the full year (i.e. a reduction of USD 150/kW compared to 2010). Wind turbines account for 64% to 84% of total installed costs onshore, with grid connection costs, construction costs, and other costs making up the balance..
At this time the exchange rate was around €1=$1.20, I added the € cost in  brackets at that exchange rate.
The US NREL 2011 Cost of Wind Energy Review has $2,100 per kW installed cost. Converted to Euro at the prevailing rate we get about €1,500 per kW. So for our 2MW turbine (reviewed earlier) the US cost would be (in Euros) €3M instead of the €2.4M.
Their operating expenses are in a different format. Here a 2MW turbine would cost $70,000 per year to operate and maintain, or (in that year) about €50,000 – a similar number to the lower end of the range given in Blanco (2009).
They also provide a nice graphic showing (to me at least) why producing LCOE (levelized cost of energy) values is not particularly helpful:
Figure 4 – Any value you like!
The report comments further:
Although the reference project LCOE for land-based installations was observed to be $72/MWh, the full range of land-based estimates covers $50–$148/MWh.
The largest factor, and the reason why a generic cost per MWh for wind and solar is a useless number, is it depends where it is located. High wind, lower LCOE. Low wind, higher LCOE. I doubt anyone would have come up with “LCOE” if energy generation had been dominated by wind and solar in the past. They would have come up with something like “LCOE per % capacity factor”. The “reference value” of $72/MWh was at a capacity factor of 37% (note 2), a value rarely seen in new European installations. I think the average for the UK (from memory, not checking) is around 30% currently, and it has gradually increased over the past few years. Many parts of continental Europe have capacity factors below 20%.
The NREL report also shows the formula for LCOE (for those interested). I’m assuming that the reason they have nearly $80/MWh for a 30% capacity factor (figure 4) – vs our figure of €23+€10/MWh = $46/MWh in that year – is due to the 25% higher capital cost along with introducing the cost of capital.
This illustrates an important point with renewables – from country to country and region to region there will be very large differences in their ability to convert to high penetrations of renewables.
One further point to be noted from these data points is that we can’t always assume the costs over the next few years will go down – as outlined in Renewable Energy I. Due to high demand, the capex cost of wind power increased for a few years.
As far as I can tell, the above costs are all free of subsidies.
The grid connection costs have been considered in the capex costs, but these pre-suppose that transmission is available, or paid for by “someone else”. In some countries like Spain, this (paid for by someone else) has been true. As wind power grows, moving the costs to the grid operator becomes more problematic. If you connect to an existing transmission line and add 10 MW this is probably fine. Once you add 500MW at peak wind periods you might overload that transmission line and a $500M upgrade may be needed. On the other hand, if you are lucky, you might be replacing conventional capacity on an existing transmission line and no upgrade will be needed.
So it should be clear that this is one of the wild cards. Each case is different, but in most cases there will be substantial cost to be incurred – once wind power becomes significant, which of course is the idea.
In the last article we looked at building long transmission costs and as a massive over-simplification suggested that a cost of $1BN per 1GW per 1000km was a handy guide. So, if we build a large series of wind farms to replace a 500MW gas-fired power station, it will be something like 2GW nameplate. If we want to add 2GW into our transmission line at peak times, and it’s 500km long we can expect to incur a cost of $1BN.
Note that there will be other complications – see V – Grid Stability As Wind Power Penetration Increases.
If we compare the transmission estimate of $1BN with building 2GW nameplate capacity of wind power – a capex cost around €2.4BN – we see it will be significant.
Wind power is a mature form of energy generation with fairly well-known costs, minimal risks, and the benefit of not being subject to fuel costs with large fluctuations. Any cost analysis is always out of date but at least here we can see approximate values for capital costs and for ongoing costs – and the basis for these values.
Depending on where you locate your wind turbine you can get a factor of 3 change in annual output so €/MWh and $/MWh are not useful metrics without a location guide.
A good estimate from a few fairly recent studies is:
- Capital cost = €1.2M per MW or $1.5M per MW of nameplate (including grid connection costs, but excluding bringing a transmission line to the area). So to convert that to cost per energy produced you divide that cost by the capacity factor (which depends on location and might be 15% in a poor location -40% in a prime location) / 8760 hours in the year / number of years you expect your turbine to operate – and you get € or $/MWh (excluding cost of capital)
- Ongoing O&M costs = €10-15/MWh
We will look at the costs of other forms of energy in subsequent articles.
Articles in this Series
Renewable Energy I – Introduction
Renewables II – Solar and Free Lunches – Solar power
Renewables III – US Grid Operators’ Opinions – The grid operators’ concerns
Renewables IV – Wind, Forecast Horizon & Backups – Some more detail about wind power – what do we do when the wind goes on vacation
The economics of wind energy, Isabel Blanco, Renewable and Sustainable Energy Reviews (2009)
The Economics of Wind Energy, European Wind Energy Association (2009)
RENEWABLE ENERGY TECHNOLOGIES: COST ANALYSIS SERIES – Wind Power, International Renewable Energy Agency (IRENA), 2012
2011 Cost of Wind Energy Review, S. Tegen, E. Lantz, M. Hand, B. Maples, A. Smith & P. Schwabe, National Renewable Energy Laboratory (2011)
Note 1: From Blanca 2009:
Note 2: NREL report says:
The annual average wind speed chosen for the reference project analysis is 7.25 meters per second (m/s) at a 50-m height above ground level (7.75 m/s at hub height). This wind speed is representative of a Class 4 wind resource (7−7.5 m/s) and is intended to be generally indicative of the wind regime for projects installed in moderate quality sites in the “heartland” (Minnesota to Oklahoma).