Understanding capacity factors

I'll occassionally post some basic concepts which I think are important to have a good understanding of for the industry. The first one is on the often abused concept of 'capacity factors'.

Wind speed distributions can be described by Raleigh and Weibull distributions (A Raleigh distribution is just a simplified Weibull distribution with k=2), with characteristic winds prescribed by the IEC (2005) in three distinct classes. For illustration, the IEC (2005) wind class probability density functions f(u) are:

The actual useable energy captured from the wind resource characterized by the wind distribution is a function of the individual wind turbine’s power curve, which is dependant on the detailed turbine design. Theoretical power curves, p(u), for both pitch-controlled and stall-controlled turbines (DNV/Risø).

Combining the power curve with the wind distribution the energy production can be determined, often expressed in terms of the Annual Energy Production (AEP).

The total energy that would be produced by a wind turbine during a one-year period, assuming a certain distribution of wind speed probability density and assuming 100 per cent availability, is referred to as the potential Annual Energy Production (AEP). The capacity factor is defined as the ratio of actual average power to the rated power measured over a period of time (average power/rated power). The better the local wind conditions, the higher the capacity factor for the turbine at that site. The maximum potential capacity factor is illustrated below for an ideal turbine with no losses.

General rule of thumb is that capacity factors really have to be 30%-40% for the site to be considered a good one, and above 40% is considered excellent (If people are talking about >50% then alarm bells should seriously be ringing!). The capacity factor gives you a good quick understanding of the park performance, however it's important to note the assumptions in the capacity factor calculations; particularly whether achieved availability and maintenance down time is considered.

Additionally, it is important to have an understanding of the turbine design and to what speed the turbines have been optimised for - namely if it is a 'high wind' or 'low wind' version. As there is an increasing cost involved with the rated power of the turbine (associated with the generator, gearbox rating increase), the same turbine nacelle is often marketed with different diameter rotors - larger for 'low wind' sites, and vice verca. Therefore, the same rated turbine will have different capacity factors at a low speed site (the larger diameter rotor will be higher) - it is therefore important to select the right turbine for the right site.

So what are some good achieved capacity factors? For offshore sites the average achieved capacity factor (including availability losses) is around 35% - with some of the great offshore sites off the west coast of Denmark measuring achieved capacity factor of 45% in 2005 (Horns Rev) and 40% in 2004 (Nysted); and Copenhagen's Middelgrund around 42% (check the link for live performance). For onshore sites, average capacity factors around the world range around 20% - 30% with some significant local variation by site. The highest recorded average capacity factor was actually in Australia at 37%, which reflects the excellent wind resources in the country. Conversely, the lowest capacity factors can be found in Germany with average capacity factors of around 17% - however interestingly Germany has the largest amount of installed capacity in the world with 18,428 MW in 2005 compared to around 700 MW in Australia - not a very smart allocation of resources on the surface is it?

Better metrics for the evaluation of projects is to evaluate the actual energy yield - which could be the capital cost/AEP, or simply put the annuakl cost per generated kW.hr. This is of course difficult to forecast, however involves the same assumptioons as when making the initial energy assessment.