- Eolian Facilities
- Renewable Facilities
- Real Time Data
- Monitored Data
System of reactive power compensation at the Sotavento Experimental Wind Farm
A large number of electrical receivers consume reactive power, i.e. when connected to an AC network they generate a lag between the voltage and the current flowing through it. This mismatch causes the intensity required for operation to be greater and therefore losses during transport of that amount of energy by the network to also be greater, this situation also increasing the work of power lines and transformers.
The power factor is a parameter indicator of the proportion of reactive power which is consumed or generated. The centres of power generation should help avoid this situation by correcting this gap from its installations, i.e. they should consume or generate reactive power required by the system. So, wind farms must work together at power factor correction.
A significant number of wind turbines installed across the country are of asynchronous squirrel cage technology, Sotavento having 19 wind turbines of this type. This technology requires a system of reactive compensation (capacitors) for power factor correction.
Generally the implementation of these systems was offset by the manufacturer and operated individually in each turbine. This was reasonable under the RD 2818, which did not require fine tuning of the power factor, as it allowed for bonuses or penalties that depended on the degree of remoteness to the unitary power factor, with the peculiarity that it was calculated from monthly production data. That is,a wind farm that had delivered a certain reactive amount during the first half of a month and spent the same amount in the second would get the same compensation as another that had maintained a unitary power factor during the month. This obviously does not seem fair because the first caused an unfavourable situation for the network.
But after the enforcement of the RD 436 the opportunity was offered to receive bonuses for a particular use or generation of reactive power calculated from quarter-time data, a possibility that remained with later RD 661.
Implementation of an innovative, reliable and versatile system for controlling the power factor at Sotavento Wind Farm, in order to maximize the reactive energy supplements listed in the regulations at the time of project development.
- Creation of a centralized system to control the power factor preventing each turbine operating individually
- System suitability both for receiving external power factor set points and for programming time slots
- Providing in an accurate and immediate way the power factor assigned to each time
- Developing a systemindependent of the machine manufacturer, to ensureapplicability toanywind farm of squirrel cage wind turbines
- Minimize the wear of electrical elements associated with compensation systems
Under these main goals the system has been able to optimise the reactive power bonus maximizing existing equipment and reducing the number of operations on the capacitor.
In a wind farm with squirrel cage machines, reactive power is compensated by capacitors, a low voltage wind turbine level and a medium-voltage substation. Specifically, the Sotavento Wind Farm had 4.630 MVAr at wind turbine level and two reactive steps of 1.215 MVAr in the substation. Generally this compensation model is not designed, nor is it able to operate for power factor instructions given. The main problems presented by this model of reactive power management were:
- Sometimes the turbines worked with a limited number of capacitors to obtain a unitary power factor from certain levels of active power generation. This deficiency had to be corrected independently from the substation via the medium voltage capacitor
- Regulating steps at medium voltage were too large to fine tune, and were precisely those carrying out the final adjustment
- The connection of the capacitor of the turbines was done exclusively by power segments. Each tranche of power produced by each turbine had associated with it a number of capacitors that were connected to the grid. More active involved more reactive
- Faults on capacitors were not detected. A machine could be operating with a condenser without knowing its status
- It produced staged disconnections without taking into account the safety times
- Thyristor systems, responsible for connecting the capacitors, too often failed causing asymmetric flows by faults in different phases
- There was no information about failures in real-time stages
Proposed for the achievement of objectives and described as a solution to the problems encountered, was the design of a central controller that should manage a set of local regulators would admit external reactive set points, where the central controller developed the following tasks:
- Collecting data using analysers installed in each of the turbines and the substation
- Calculated from these data was the amount of reactive power needed globally to achieve the desired power factor at the point of evacuation
- Analysis and identification of the number of capacitors connected based on reducing the number of operations on them, especially on the medium voltage, to have a longer download. This step must not rule out the capacitors operating, either because they are damaged or because they belong to the wind turbine that is not connected
- Sending the order to the drivers for each machine and substation for these to act on the relevant capacitors
- Study of the initial conditions of the reactive compensation systems of each of the turbines and substation
- Design of a regulatory scheme maximizing existing equipment in order to intervene as little as possible
- Development and simulation of a universal control algorithm in collaboration with the University of Vigo that achieves an optimum balance between maximum bonus and number of operations to be performed on capacitors
- Changes in compensation systems
- Selection and installation of equipment
- Design and installation of communications network
|System of reactive energy compensation at the Sotavento Experimental Wind Farm||Spanish||1,075 Kb|