Search In this Thesis
   Search In this Thesis  
العنوان
Enhancement of Distribution Systems Performance Using Microgrids \
المؤلف
Mohamed,Reham Ahmed Osama Mohamed
هيئة الاعداد
باحث / ?ريهام احمد اسامه محمد محمد?
مشرف / ?المعتز يوسف عبد العزيز?
مشرف / ?رانيا عبد الواحد سويف?
مناقش / ?السعيد عبد العزيزعثمان?
تاريخ النشر
2020
عدد الصفحات
194p.:
اللغة
الإنجليزية
الدرجة
الدكتوراه
التخصص
الهندسة الكهربائية والالكترونية
تاريخ الإجازة
1/1/2020
مكان الإجازة
جامعة عين شمس - كلية الهندسة - قسم هندسة القوى والالات الكهربية
الفهرس
Only 14 pages are availabe for public view

from 244

from 244

Abstract

This thesis introduces a novel approach for the optimal construction of interconnected microgrids in the distribution level having various distributed energy resources. The IEEE Std.1547.4-2011 recommended that the operation and control of distribution networks can be enhanced by dividing these networks into a number of microgrids. The suggested planning scheme combines the essential requirements for microgrids to function with high efficiency during connection with the larger grid and with high possibility of islands’ success in stand-alone mode of operation.
The optimal partitioning of distribution networks involves finding the locations of cut-set lines that define the microgrids’ borderlines for different objectives. To ensure robustness of design, the partitioning mechanism in this thesis optimizes three objective functions. The first objective involves maximization of microgrids’ self-adequacy. During grid-connected operation, minimizing the imbalance of generation and loads can minimize the energy exchange between the microgrids and increase their self-sufficiency. The second objective involves maximization of the probability of successful islanding of microgrids. If interruption of loads within an islanded microgrids is kept to a minimum, the cost of power outages for utilities and customers can be diminished along with an improvement in the overall system’s security. An indicator is defined which takes into consideration the requirements for a successful operation of an island. The third objective function combines both the first and the second objective simultaneously to attain the maximum possible benefits.
To solve the optimization problem, the population-based backtracking search algorithm is adopted which is an iterative method that was recently proposed to overcome the shortcomings of evolutionary algorithms. The complex computation and the use of many parameters are the major drawbacks of such algorithms. The backtracking search algorithm has a single parameter that controls the search’s amplitude and direction to provide the required balance between exploration and exploitation processes, i.e., local and global search procedures, in order to improve problem solving capability.
Due to the hourly variation nature of loads and the intermittent characteristics of distributed generation units, the modeling process of system components is carried out probabilistically using multi-state variable formulation where typical probability density functions are utilized for the modeling of the stochastic variables behavior. The probabilistic backward-forward sweep power flow and other graph related approaches are adopted in this work.
In order to apply the adopted methodologies, we chose the Pacific Gas & Electric distribution network to be the test system. Simulation results demonstrate the effectiveness of considering a combined objective in the partitioning process. In comparison with other strategies present in the previous literature, the proposed framework results in more self-sufficient and successful islands assessed in terms of voltage constraints, real power adequacy, and reactive power adequacy. Next, the effects of installation of both energy storage and reactive resources on the optimal infrastructure are examined. Simulations show a significant improvement of the self-adequacy and islanding success capability of the constructed microgrids. A comparison with other microgrid design objectives applied in previously published researches reveals that the resultant design is sensitive to the system’s reliability, security and economic requirements. Finally, sensitivity analysis showed the robustness of the optimal design against system variations, i.e., increase in load and generation levels up to a certain acceptable level. Hence, the planning framework introduced in this thesis can be reliable and efficient against future network upgrades for a long period of time.