The Sotavento Experimental Wind Farm consists of 24 different wind turbines, placed in a moderately complex orography. It is divided into 5 different technologies with a total of 9 different turbines.
The wind farm’s mission is to evaluate the behaviour of each of the turbines, and provide an ideal environment for the investigation of possible innovations in the development of wind turbines.
Due to the effect of contrails, not all machines receive a wind of equal characteristics, hence the need to determine the wind event on each of the turbines, as the data provided by the casing anemometer contain disturbances by the hood and accelerator effects.
Prior to the construction of the wind farm, a study was made by Garrad Hassam with existing wind data.
The aim was to develop a computer model that determines in real time the incoming wind in each of the wind turbines, considering contrail and orographic effects, in order to know the true wind received by each turbine at the wind farm.
Distinguish the differences between a wind incident and that caught by the wind turbines for each wind direction and speed.
Correlate actual wind incident characteristics in each of the machines, with the measured wind anemometers used in the anemometer towers in the wind farm. This project has been conducted by the Fluid Mechanics Laboratory of the Department of Energy Engineering and Fluid Mechanics of the School of Industrial Engineering of the Polytechnic University of Madrid (UPM).
The data elements used were as follows:
Programs for estimating orographic effects:
- WASP. Is a commercial program developed by RISO, widely circulated.
- UPM. Own program for moderately complex terrain. It is a program based on the known model of Jackson and Hunt. Its applicability limits are similar to WASP, and its foundations may be similar, but it has the advantage of being an open program.
- FLUENT. Is a program for general fluid-mechanical problem solving. It was applied to the wind farm configuration for wind direction reviews. It would be used with astationary character and k-ε type closure for turbulence modelling. It can be applied to complex orographic configurations with recirculation flow and landslides.
Programs for estimating the effect of contrails:
- UPMPARK. It is a program developed in the laboratory of the UPM, which can calculate the flow conditions: it solves the general equations of fluid mechanics with a k-ε closure turbulence modelling. It uses parabolic flow approximation. The terrain is flat and it uses an appropriate algorithm that overlaps terrain effects calculated with any of the above programs and contrails. This program has been extensively validated by comparison with real turbines and wind tunnel experiments.
- PARK. It is a program developed by RISO and is built to WASP. Their degrees of development and validation are well below those of UPMPARK, but can be used primarily to negate results.
The different phases are developed throughout the project are:
- Sotavento provided the UPM: digitised topography, historical wind data and final report by Garrad Hassam
- The Fluid Mechanics Lab of the UPM ETSII conducted a report or critical analysis of the final report of Garrad Hassam
- Sotavento made available to the UPM wind data at the wind farm stations and wind turbine power-anemometry
- The Fluid Mechanics Lab of the UPM ETSII made an initial approach to distinguishing wind incidents in the 24 position against wind records of the period provided
- The Fluid Mechanics Lab of the UPM ETSII developed the wind incident report, and provided for its use with Sotavento computer applications, a program that provides 24 wind values in each wind turbine
This project has been conducted by the Foundation for Industrial Development, counting on the collaboration of the Fluid Mechanics Laboratory of the Department of Energy Engineering and Fluid Mechanics of the School of Industrial Engineering at the Polytechnic University of Madrid.
Team principal investigator was Professor D. Antonio Crespo Martinez.
The project has been completed since 2003.
- November 2002. Discrepancies about Garrad Hassam and study of deviations of this
- November 2002-January 2003: Analysis of our own data (UPM). Checking contrail effects. Variability regarding data of Garrad Hassam
- December 2003: Data handling of wind at the towers installed in the wind farm; interrelating: wind – turbine data – data substation (like the whole wind farm)