ANALYSIS OF TWO ONSHORE WIND FARMS WITH A DYNAMIC FARM CONTROLLER

摘要

Coordinated or cooperative control of wind farms can increase power output reducing cost of energy (COE) per area. The increase in production depends upon wind conditions, terrain characteristics and wind farm layout. This work develops an assessment methodology for identifying potential wind conditions where coordinated control can increase farm power output. Average normalised power across all directions, average relative power with reference to a specified turbine and average relative efficiency with respect to the turbine producing maximum power were used for identifying the negative impact of wakes on power output. A dynamic farm controller is presented that exploits the benefits of reducing power of the upstream turbine(s) for increasing overall farm production. The dynamic controller uses a modified version of the Jensen model for wind speed deficit calculation. This model calculates the wake decay coefficient on a turbine by turbine basis. The decay coefficient varies according to turbulence intensity inside the farm. Particle Swarm Optimisation (PSO) is used for selecting an optimum set of turbine coefficients of power (Cp). The assessment methodology and farm controller were applied to data from two operating onshore wind farms to determine the benefits of coordinated control. The dynamic farm controller can increase power production up to 10% when compared to the conventional greedy control. The modified Jensen model accurately predicted the wind speed deficit in most of the cases. The case study farms were optimised in under 5 seconds on a simple computer. This makes the dynamic controller very suitable for online real time operations.