Energy optimization through voltage reduction method in electrical power system with distributed generation

Deployment of smart grid technologies in distribution systems provides many opportunities for utilities. Thus, a large number of utilities intend to explore the capabilites of smart grid technologies to imrove the efficiency, reliability of grid and overcome technical-economical issues of the grid. Conventional control methods need to be improved to address the challenges associated with negative influence of distributed generation (DG). Hence, new control techniques need to be developed for smart grid operations to overcome the problems, provide high quality power for the customers and increase the grid efficiency. This thesis aims to introduce novel approaches for one of the useful techniques employed for distribution network voltage and reactive power optimization called as Volt-VAR Optimization (VVO). The proposed approaches are applied for Conservation Voltage Reduction (CVR) as a part of VVO which provides energy saving by decreasing the voltage magnitudes to the minimum allowable limits without affecting the performance of the end user's devices. The first part of the study introduces distributed reactive power control based VVO/CVR approach in a decoupled way. The objective of the novel VVO/CVR approach is to minimize system power losses by distributed reactive power control while reducing the voltage down to the lower acceptable limits. The second part of the study introduces centralized and decentralized integrated VVO/CVR approach. The control problem is formulated as a mixed integer non-linear programming (MINLP) problem. The foregoing problem is solved through Genetic Algorithm (GA). Applying two different VVO/CVR approaches not only improves the voltage profile along the feeder but also reduces the daily energy demand. Model Predictive Control (MPC) technique is applied to second VVO/CVR problem and it has been seen that the operation/switching times of equipment are reduced by avoiding unnecessary control. The proposed approaches in this thesis are validated on IEEE 13-34 test feeder system with DG and DSTATCOMs assuming steady-state operating conditions and two-way communication network. The formulation is based on a three-phase power flow with variable load demand, different customer types, voltage dependent loads, unbalanced line configurations.