dc.description.abstract |
The attention of many researchers has recently been on the best ways to integrate
Distributed Generation (DG) into conventional centralized electrical power distribution
systems, particularly in the context of the smart grid idea. This is due to its reputation
as a viable remedy for the lack of electric power supply. To optimize the environmental,
financial, and technological advantages of the integration of DG units for distribution
network operators, it is crucial to determine their ideal position and size. The main
objective of this study was to develop and simulate an optimization system for the
placement and sizing of distributed generation units in electrical power distribution
networks for power loss reduction and voltage profile improvement. The specific
objectives were to model and develop the load flow algorithm and codes; develop a
meta-heuristic optimization algorithm and codes that selects the best location and size
of the DG unit; simulate the nested load flow and optimization algorithms and codes on
MATLAB and analyze the effectiveness of the developed algorithm via testing on the
standard IEEE 33-bus radial electrical power distribution benchmark network. The
Backward-Forward Sweep (BFS) technique was employed in the load flow modeling
because it maximized the radial structure of distribution systems. The optimization
algorithm was developed based on the Multi-objective Particle swamp optimization
(PSO) meta-heuristic technique due to its effective global searching characteristic. The
line and load data for the IEEE 33-bus test network, a cutting-edge benchmark for
contemporary power distribution networks; were obtained from the Power Systems
Test Case Archive- a secondary data source. For this network fed by a synchronous
generator, the chosen base MVA (Mega Volt Amp) was 10 MVA and the base voltage
was 12.66 kV. The total active and reactive power demands were 3.715 MW and 2.300
Mvar respectively. The simulation was done on the R2021a version of
MATLAB/Simulink. The total real and reactive power losses obtained from base case
simulation without the placement of any DG unit in the network were obtained as
201.893 kW and 134.641 kvar respectively while the per unit (p.u) average bus voltage
was 0.9485 p.u. After the optimal allocation of one, two, three, and four DG units, the
total real power loss (in kW) in the network was reduced by 140.89, 173.89, 189.89,
and 195.89 respectively while the total reactive power loss (in kvar) reduced by 86.64,
114.64, 124.64 and 128.64 respectively. Likewise, the per unit average bus voltage
improved by 0.0376p. u, 0.0458p.u, 0.0480p.u and 0.0498p.u respectively. Also, the
decrease in the total real and reactive power losses and the improvement in bus voltage
profiles varies proportionally with the number of DG units optimally placed. In
conclusion, the results show that the total real power loss and the total reactive power
loss of the network were significantly decreased; and the voltage profile of the system
was drastically enhanced by incorporating DG units at predetermined buses. The
developed algorithm is recommended for application in a real electrical power
distribution network for more efficient integration of new distributed generation units in
the current electrical power distribution networks. |
en_US |