Wednesday, January 30, 2008

Global installed capacity beats expectations in 2007!

GWEC have just announced that the global installed capacity of wind energy jumped in 2007 with 20 GW of installed capacity - I've plotted that against other historic data and projections in the chart there. This beats last years expectations (see my previous post), and shows that the wind industry is really going strong. The lead country market was the US with 5.2 GW of capacity, followed by Spain and then China, however Europe continued to lead with 61% of the installed capacity for 2007. More encouraging to me, 150MW was installed in New Zealand - harvesting some of their great natural wind resources with very little government economic incentives.

So how much is 20 GW of wind energy? If we assume an average capacity factor of 25% (which is relatively conservative) - that's about in practical terms:
  • 15,000 individual turbines costing around €30 billion to install
  • 55 TW.hr of additional annual energy production - enough for about 13 million average EU households
  • Additional annual energy sales revenue of roughly €300 million (at EU prices)
No doubt about it, wind energy today is big business, and the market is naturally allocating capital to it's development with relatively minimal government assistance now.

One major assumption here is that the installed capacity of turbines represents only that a turbine at a rated-power is errected at a site - and doesn't consider the commissioning, grid connection, and the actual energy production. Additionally, there is no allowance for the phasing out of older turbines (not that significant now, however will get more and more so). Anecdotal reports are that 25% of new capacity in China is still not connected to the grid due to planning problems (I will try and find a reference for that!).

Thursday, January 24, 2008

Airfoil aerodynamics

After a bit of a break from the European winter back home in Australia for Christmas, here is the first post for 2008, I thought I would start with airfoil aerodynamics and develop that concept over the next few months. I

For current 'state of the art' turbines, let's say 1.5 MW to 3.5 MW, a typical operational Reynold's numbers of around 1-5 million are typical. An indicative airfoil for the power-production area of the blade (the outer half) is the NACA 63415 as seen below (coordinates taken from the public UIUC coordinates database). The NACA five-digit system says (in relation to chord): a maximum camber (curvature) of 6% at 34/2=17% from the leading edge, with a maximum thickness of 15%. This is a typical low speed airfoil, and quite similar to that seen in the glider industry.

The Danish research institution, Risø, have conducted numerous tests on this airfoil, specifically for wind turbine applications, with a good indicative Reynold's number of 1.6 million, some good test data is available from the Risø-R-1280 'aifoil catalogue'. Using this wind tunnel test data, the lift-curve for the NACA 63415 is shown below; CL-max and CL/CD-max are marked for information. In the context of wind turbine blade design, the CL-max value represents the most lift the section can produce prior to the onset of stall, followed by the post-stall region (which is a lot more uncertain than the graph suggests!). The fairly linear lift-curve slope prior to CL-max, and the 'very roughly' linear stall region, are characteristics utilised by the turbine's controller system to regulate power production (this will be the subject of an upcoming post!).


The CL/CD-max value of around 5 degrees is derived from the lift and drag characteristics as shown on the left here. Here, the maximum CL/CD ratio is around 70, which means the airfoil is generating 70 times moer lift than drag at that particular AoA. This is the optimal relationship between lift and drag, and the most efficient lift generation for an airfoil possible. For fixed-wing aircraft (especially gliders!) this is fundamental to gliding performance (actual glide ratios around this value have been achieved!), however for wind turbines this represents the 'sweet spot' where the most amount of energy can be extracted from the wind resource and therefore the turbine's achieved Cp optimised.

The main task therefore for the detailed blade aerodynamic design, as well as the turbine's control system algorithm design, is therefore to optimise the AoA as close to CL/CD-max as possible . When the blade's AoA is outside of CL/CD max, turbine efficiency (Cp) and therefore energy production drop.

This discussion is limited to 2D only, so neglects 3D effects which further complicate the matter! Some further topics I'll write about include: detailed control systems, influence of dirty blades and icing, and changes in Reynold's number (varying the tip speed, and density altitude).