Abstract:
Electric power distribution systems with radial nature and improper installation of successive
feeders’ causes voltage drop. In addition, the growth of load demand increases the current drawn
from the source and hence voltage drop and power loss will also be increase. Different methods
used to avoid such problems; one of these methods is using a distributed generation (DG). In this
thesis minimization of power loss and voltage profile improvement in radial distribution system
have been achieved by using an optimal type, size and location of DG. IEEE 33 bus system is used
for comparing analytical and particle swarm optimization with loss sensitivity factor analysis
methods with three load variation cases, and the best method is selected to be applied for
optimization of DG units in agriculture feeder of Jimma city. Load flow analysis of the test system
and existing feeder is performed with forward backward sweep method. When comparing the
results obtained by the two methods, the particle swarm optimization (PSO) method with loss
sensitivity factor analysis in type-1 DG provides an optimal size and location of DG with best loss
minimization and improved voltage profile result than analytical method. In case of full load
condition, the optimal locations of DG found to be bus 13 and 30 with optimal size of 773.4KW
and 632.4 KW by PSO method with percentage reduction for total real and reactive power loss
obtained as 52.73% and 52.56% respectively. The voltage magnitudes in all buses are within
acceptable limit and the minimum per unit voltage is 0.9521 at bus 33. Furthermore, the impact of
DG and the selected method have been analyzed and verified by considering case study at Jimma
old substation. Among the five 15 KV feeders of the substation, Agriculture feeder is considered
and the total real and reactive power loss before optimization was 94KW and 53KVAR, then using
PSO and loss sensitivity factor the optimal location of bus 26 and 40 with optimal size of 782 KW
and 474 KW of DG is obtained. The percentage reduction of loss found to be 43.6% and 42.8% in
real and reactive power loss respectively with good voltage profile improvement. Finally,
Economic analysis using HOMER PRO software also results in optimal combinations of
technologies depending on net present cost (NPC) of the system for the selected feeder. Economic
issues associated with the DG unit installation: such as cost of energy losses (CL) with and without
DG unit have been compared and total saving of 32,324.4 $ is found.
