The article analyzes changes in the legislation on the rules of electricity generation from renewable sources and the introduction of penalties for imbalances, which encourages producers to improve the forecast of electricity generation and modernization of existing power plants by installing energy storage systems.
Possible connection points and charge/discharge device (CDD) converter topologies for an energy storage system are analyzed and the converter that meets technical requirements of the system are selected. As a result, the connection from the direct current side has a number of advantages: simpler CDD structure and control principle, in comparison with alternating current; no galvanic separation between input and output. Converter analysis showed, that due to a high operating voltage, usage of resonant topologies is undesirable and the absence of galvanic separation makes bridge converter usage impractical. Therefore, to solve this problem, a bidirectional converter without galvanic separation with hard switching is proposed. To reduce the level of dynamic and static losses, it is advisable to use a modular topology converter with alternating phases.
The operating modes of such a converter at a given error of the weather forecast are analyzed. To improve quality of the generated electricity, it is expedient to use a power stabilization mode. Due to the higher values of charge/discharge currents, as well as higher energy density compared to acid-lead, a lithium-ion battery was chosen. According to the selected operation mode, its minimum capacity was calculated. When using a minimum battery capacity, due to the difference between the maximum discharge and charge current of the battery, a mode of partial power stabilization is possible. This mode is used only when the forecast error is more than 52% in the charging mode.
A charge/discharge device were designed for a 50 kW SMA Sunny Tripower CORE1 inverter and 20*315W LP156*156-M-60 solar panels connected in series. The control principle for such CDD is described. Control algorithm can be divided into four stages: obtaining the predicted solar radiation power in the forecast interval; predicted illumination power conversion into electric power; predicted power calculation and the amount of energy that will be generated and transferred by the solar station to the regulator in the forecast interval; power setpoint stabilization on the forecast interval based on the proportional-integral (PI) control law.
To verify the obtained theoretical relations, converter parameters with a typical input data were calculated. Model of the converter was created in the MATLAB® Simulink® environment and its operability was checked.