Scalable DC Microgrids for Rural Electrification

Abstract

Access to electricity is one of the key factors indicating the socio-economic status of any community. Reliable and adequate provision of electricity is mandatory for improved standards of living including better health, education, transport, agriculture and employment opportunities. Unfortunately, according to International Energy Agency, over 1.1 billion people around the world lack access to any electricity out of which 85 percent reside in rural areas of developing world. Electrification of these remote rural communities through national grid interconnection is not economically feasible for many developing countries due to high cost associated with the development of generation, transmission and distribution infrastructure. Alternatively, DC microgrids implemented with distributed generation and low voltage distribution are becoming very popular for low cost rural electrification. However, current implementations are largely suboptimal due to high distribution losses associated with their centralized architecture and their inability to support high power community loads. In this work, a novel distributed DC microgrid architecture which allows a scalable approach with minimal upfront investment to fulfill rural electricity needs along with the provision of higher powers for communal loads and beyond subsistence provisioning of electrical power is proposed. The architecture is capable to work entirely on solar energy with power delivery capability to individual consumers and added inherent ability to integrate resources to power up larger loads for communal/commercial applications. The proposed microgrid architecture consists of a cluster of multiple nanogrids (households), where each nanogrid has its own PV generation and battery storage along with bi-directional connectivity to the microgrid. Thus, each nanogrid can work independently in islanded mode along with the provision of sharing its resources with the community through the bidirectional converter. In the proposed architecture, the bi-directional power flow capability is implemented through a modified flyback converter. A decentralized control methodology is also proposed to ensure a communication-less, yet coordinated control among the distributed resources in multiple nanogrids. The microgrid is evaluated for optimal distribution voltage level, conductor size and interconnection scheme between nanogrids using Newton-Raphson analysis modified for DC power flow. Various scenarios for power sharing among the contributing nanogrids and communal load power allocation are analyzed from operation and control prospective to validate the architecture and its performance. Further, an optimal framework for the planning of distributed generation and storage resources in each nanogrid with respect to time varying profiles of region-specific temperature and irradiance is also presented to ensure the better resource utilization. A scaled version of the proposed architecture is implemented on hardware, while the efficacy of control methodology is validated on MATLAB/Simulink and hardware in loop facilities at microgrid laboratory in Aalborg University. The proposed distributed architecture along with decentralized control can be considered as a promising solution for the future rural electrification implementations in developing regions.