Determining penetration limit of central distributed generation topology in radial distribution networks

• Main Author: Suliman, Mohamed Saad Abdelgadir
• Format: Thesis
• Language: English
• Published: 2021
• Subjects: Electric power distribution; Electric network topology; Distributed generation of electric power
• Online Access: http://psasir.upm.edu.my/id/eprint/92804/1/FK%202020%20106%20UPMIR.pdf; http://psasir.upm.edu.my/id/eprint/92804/

Distributed generation has become one of the major electric power system elements. The advantages of utilizing distributed generations in power systems include economic, environmental, and technical benefits. The optimum utilization of distributed generation units offers potential benefits to the electric systems such as network reliability, peak loads reduction, voltage support, and power quality improvement. Improper utilization of distributed generation units in distribution networks lead to frequency variations, raise system power losses, voltage deviation, and altering the fault current value. The potentials of renewable energy sources are categorized based on theoretical, geographical, technical, and economical potentials. The geographic potentials are related to the implementation area, which shall be usable, sufficient, and stable to host the renewable energy sources, particularly photovoltaic solar plants sites are restricted with legal and technical constraints. Distribution network operators are practicing various topologies to align the optimal geographic sites with the optimal points of connection in the distribution networks. These topologies include the central photovoltaic solar plants, which consolidate the optimal distributed generation capacity at one central location, while the power are transferred to multiple optimal locations. On the other hand, the conventional scientific allocation methodology accommodates the optimal size of distributed generation directly to next to the optimal location. Although the scientific research community have investigated the optimal allocation of renewable energy sources from various perspectives that involve sophisticated theoretical, geographical, technical, and economical multi-objective functions, however it lacks a fundamental evidence that directly compares the conventional bus dedicated topology versus the central distributed generation topology on a typical distribution network using a typical methodology. In addition, the applied distributed generation topology directly affects the network penetration limit, which influence network operational limits consequently. Therefore, the study proposed a comparison between the conventional bus dedicated distributed generation topology and the central distributed generation topology. The optimal sizing and allocation of distributed generation problem is based on active power loss reduction and voltage profiles improvement. The scope involved deterministic load flow formulation to obtain the essential power system parameters of the optimal distributed generation allocation. The load flow is performed using the Newton-Raphson method. On the other hand, to test the network operational limits when uncertainties of the photovoltaic generation and load demand are included, the probabilistic load flow was simulated using Monte Carlo Simulation method. The beta probability density functions were used to model the photovoltaic generation, while the normal probability density functions were used to model the load demand. The effectiveness of the proposed topology was validated on IEEE 33 and 69-bus distribution networks. Biogeography based optimization method was formulated to solve the optimal allocation problem, then manual method has been applied to accommodate the central unit. The manual accommodation of the optimally sized central unit was preferred to be applied, which removes the contradictions of comparing two different optimization allocation methodologies. The biogeography based optimization method has been proven to have better performance than artificial bee colony, genetic algorithm, particle swarm optimization, hybrid of particle swarm optimization and constriction factor approach, and hybrid of ant colony optimization and artificial bee colony methods in terms of active power loss reduction. Meanwhile, the central distributed generation unit topology was proved to have better performances over bus dedicated distributed generation topology and the results showed 6.25% and 14.7% higher active power losses reduction in the central topology of IEEE 33 and 69 bus distribution networks respectively. The voltage profiles, distributed generation capacity required, and the penetration limit have shown better performances on the central distributed generation topology over the bus dedicated distributed generation topology. Furthermore, the probabilistic boundaries at minimum, mean, and maximum of power loss reduction, penetration levels, and voltage profiles have shown better performances when the central distributed generation topology is applied.