Sizing and placement of distributed generation in electrical distributions system using conventional and heuristic optimization methods

Distribution Generation (DG) has gained increasing popularity as a viable element of electric power systems. DG, as small scale generation sources located at or near load center, is usually deployed within the Distribution System (DS). Deployment of DG has many positive impacts such as reducing transmission and distribution network congestion, deferring costly upgrades, and improving the overall system performance by reducing power losses and enhancing voltage profiles. To achieve the most from DG installation, the DG has to be optimally placed and sized. In this thesis, the DG integration problem for single and multiple installations is handled via deterministic and heuristic methods, where the results of the former technique are used to validate and to be compared with the latter's outcomes.;In the deterministic solution method, the sizing of the DG is formulated as a constrained nonlinear optimization problem with the distribution active power losses as the objective function to be minimized, subject to nonlinear equality and inequality constraints. Such a problem is handled by the developed Fast Sequential Quadratic Programming method (FSQP). The proposed deterministic method is an improved version of the conventional SQP that utilizes the FFRPF method in handling the power flow equality constraints. Such hybridization results in a more robust solution method and drastically reduces the computational time. In a subsequent step, the placement portion of the DG integration problem is dealt with by using an All Possible Combinations (APC) search method. Afterward, the FSQP method's outcomes were compared to those of the developed metaheuristic optimization method.;The difficult nature of the overall problem poses a serious challenge to most derivative based optimization methods due to the discrete nature associated with the bus location. Moreover, a major drawback of deterministic methods is that they are highly-dependent on the initial solution point. As such, a new application of the PSO metaheuristic method in the DG optimal planning area is presented in this thesis. The PSO is improved in order to handle both real and integer variables of the DG mixed-integer nonlinear constrained optimization problem. The algorithm is utilized to simultaneously search for both the optimal DG size and bus location. The proposed approach hybridizes PSO with the developed FFRPF algorithm to satisfy the equality constraints. The inequality constraints handling mechanism is dealt with in the proposed Hybrid PSO (HPSO) by combining the rejecting infeasible solutions method with the preserving feasible solutions method. Results signify the potential of the developed algorithms with regard to the addressed problems commonly encountered in DS.;The unique structure of the radial distribution system is exploited in developing a Fast and Flexible Radial Power Flow (FFRPF) method that accommodates the DS distinct features. Only one building block, bus-bus oriented data matrix is needed to perform the proposed FFRPF method. Two direct descendent matrices are utilized in conducting the backward/forward sweep employed in the FFRPF technique. The proposed method was tested, using several DSs, against other conventional and distribution power flow methods. Furthermore, the FFRPF method is incorporated within Sequential Quadratic Programming (SQP) method and Particle Swarm Optimization (PSO) metaheuristic method to satisfy the power flow equality constraints.