Optimal renewable enegy intergration for an efficient operation of a MV Distribution network. Case study of Goma DRC

Abstract

Integrating VRES to the existing grid still challenging in Democratic Republic of Congo (DRC), where there are a lot of isolated electrical network and some of them like the Goma national distribution network operate with a high level of losses on certain feeders due to frequent extensions. Variable Renewable Source (VRESs) are growing faster in the region and are gradually produced in urban areas and more studies have to be carried out in the sense of reconciling the existing grid to Distributed Generation (DG). Even though a lot of DG optimization algorithms for placement and size in a distribution network have already been proposed with the objective of decreasing system power losses and improving voltage profile, they still suffer from several drawbacks and most of them have been tested only on IEEE test system. Thus, by creating new or improving the existing ones this important issue can be addressed more efficiently and effectively. Some works proposed hybrid methods to solve the accuracy issue of Heuristic methods but they were still applied on predefined test systems. This research work aimed at solving the power losses minimization problem by proposing a two-step Fuzzy Logic (FL) and Particle Swarm Optimization (PSO) methodology so as to allocate and size DGs while testing it on the Goma city MV distribution network. Power loss reduction index were utilized for FL-based DG location process and Backward/Forward sweep load flow model for PSO-based DG sizing process. This results in reducing the searching space and thus, raised up convergence rate while reducing the computation time of the algorithm. The result obtained were compared with those obtained in a previous study in which capacitor were optimally sized and allocate using analytical method. An analysis of the integration level was also done. The results showed that DG placement is more efficient than capacitor placement for the case studied. Indeed, the proposed method presented the highest power loss reduction. The percentage active and reactive power loss reduction was 51.4% each for Route-Sake and 61.2% each for Sotraki feeder. The voltage profile which was out of standards before DG placement was largely improved with 0.959 p.u as lowest buss voltage for RouteSake and 0.94 p.u for Sotraki feeder. When studying DG penetration effect on system power losses and voltage profile, we found that more DG inclusion beyond the optimal number, although it was leading to more voltage improvement, it resulted in a sudden increase in power system losses.