CleverFarmCleverFarmCleverFarm was an EU JOULE funded project (ERK6-CT-1999-00006).
The system includes short-term prediction of the wind farm electricity production, video camera surveillance of the wind farm, information on the current running conditions and wind turbine fault prediction. This combination enables the wind farm owner to effectively plan the maintenance of the wind farm, and, furthermore, the short-term predictions add increased value to the wind farm produced electricity.
The system is implemented in Java™
using a modular approach. It gives authorized persons access to the system from local networks and the internet. The need for such a system is
realized from the fact that many wind farms in the future will be located at remote sites, for instance off-shore. A utility or operator of such wind farms can therefore use the system for maintenance planning, taking into account the weather conditions at the wind farm to allow for safe working conditions for the maintenance crew. The fault predictions can be used to
prioritize the maintenance schedule. Furthermore, the maintenance crew will have on-line access to the system's status and history at all times via the internet.
We have decided to open-source the code, but need to straighten out the legal issues involved. If you're interested, drop me an email.
Status:
Two wind farms have been instrumented in Denmark and Germany, the initial CleverBeans
are installed, and the central servers and the communications
is set up. Data acquisition works. The project as such is over - here is the
publishable final report.
On a related note, our member Gram&Juhl
has a success story from another farm, where they could detect the failure of a
gearbox bearing already one year in advance. Read about it in the final
report, or the Madrid poster/paper!
Papers:
Papers from the follow-up projects:
Questions (and maybe even answers):
Gregor.Giebel@Risoe.DK
Last touched up 20-07-2006 .