PEER-TO-PEER POWER SHARING IN DC MICROGRIDS FOR RURAL ELECTRIFICATION

Rabia Khan

doi:10.7273/000006563

Remote rural regions without electricity access suffer from energy poverty and reduced opportunities for the population. Microgrid architectures with optimal planning, design, and operation strategies are essential to meet rural inhabitants’ energy demands. DC microgrids based on photovoltaic panels and batteries are used for remote rural electrification. Centralized islanded systems have shortcomings, i.e., high distribution losses, less efficiency, and are comparatively more expensive than distributed microgrids. The distributed systems comprise independent household prosumers that maywork independently or integrated. The first concept presented in this thesis is a detailed distribution loss analysis of both centralized and distributed microgrid architectures with dynamic load and generation profiles. The distribution loss modeling is extended to low-voltage, low-power islanded DC microgrids. A detailed network loss analysis of four different microgrid architectures is performed using modified Newton-Raphson power flow for DC systems. These architectures include 1) Centralized generation centralized storage (CGCS), 2) Centralized generation distributed storage (CGDS), 3) Distributed generation centralized storage (DGCS), and 4) Distributed generation distributed storage (DGDS), which are implemented with both radial and ring interconnection schemes using time-varying load demand and dynamic PV generation. A comparative distribution loss analysis with various conductor sizes and voltage levels shows that the distributed ring architecture significantly advantages based on low distribution losses, high efficiency, and low voltage drop. It offers an additional feature of scalability and lower capital cost. Secondly, a detailed distribution and conversion loss modeling and analysis is performed for centralized and distributed microgrid architectures using the bus injection method and modified Newton-Raphson power flow method. A comparative power system and power electronic loss analysis for both architectures show that distributed architectures have higher efficiency and lower losses than centralized. Third, the optimal power dispatch and power-sharing among spatially distributed nanogrids are performed to minimize distribution losses and maximize power electronic conversion efficiency in a typical islanded DC microgrid (IDCMG) for rural electrification. A branch flow model is proposed for modeling the power system with DC-DC converters. The optimal power flow is performed by relaxing the original non-convex constraints using second-order conic programming and is implemented on the modified IEEE-14 bus system. This generic framework can be used for optimal energy management in islanded microgrids using the regional solar irradiance information, climate situations, and energy requirements. The key contributions of this dissertation are: i) A comprehensive distribution loss analysis of centralized and distributed microgrid architectures, ii) Developing a mathematical framework and modeling of distribution and power electronic losses, and iii) Optimal peer-to-peer power sharing in DC microgrids for rural electrification.

PEER-TO-PEER POWER SHARING IN DC MICROGRIDS FOR RURAL ELECTRIFICATION

Files and links (1)

Abstract

Metrics

Details