Rural area electrification is one of the challenging issues to be solved by the engineer as it is often constrained by several issues such as geographical condition and underdeveloped infrastructures. As a result, off-grid DC system is expected to answer the need of electricity in these areas.

Solar Home System (SHS) as one of the practical implementation of off-grid DC system has been used widely due to the low operating cost of PV panel which requires no fuel to operate. Besides that, this system offers more modularity rather than conventional power generator due to the absence of frequency and reactive power parameter in the static component. Decentralized control for energy management between component is preferred rather than centralized, so it does not require additional communication lines. Moreover, it allows each component to work independently when there is a failure or disturbance in the distribution lines. Equal power balance for multiple batteries is maintained by the additional controller in each local battery converter. Thus, the batteries rate of charge and discharge will be adjusted on their capacity.

Optimization of the individual SHS is conducted by interconnecting them as an attempt to form a low voltage DC off-grid network. It enables power-sharing mechanism to happen within connected houses. Additional power converters are required to provide bi-directional power flow within coupled houses, and decentralized secondary control needs to be introduced to give an equal independence level between these houses. Direct interconnection on the same voltage level allows the system to be simply coupled through cables, yet the current in the interconnection cable could increase sharply according to the sum of total connected SHS. On the other hand, interconnection in higher voltage level requires additional bi-directional converter between SHS and interconnection cables. However, the current is much lower and, thus provides a better location for high power-consuming loads to be installed.

The fact that the developing world struggles to cover even the most basics of needs, in contrast to the developed world, is tangibly connected to the lack of electricity. As many developing nations, especially the ones located in rural areas, face many difficulties to access the grid, alternative stand-alone solutions such as Solar Home Systems (SHS) have been proposed and used already for a while. The main technical limitation plaguing SHSs is the necessity of including a storage device that mitigates the energy difference between production and demand. Batteries are the components that experience the shortest lifetime and highest cost in stand-alone PV systems such as SHSs.

This study aims to explore and find the most suitable and viable storage option in terms of lifetime and costs for certain load cases concerning SHS applications. In order for this to be achieved, single battery options, as well as Hybrid battery options, were examined.

For the comparison of the different battery technologies with the different Hybrid battery configurations, lifetime and costs had to be taken into account. For this purpose a non-empirical dynamic battery lifetime estimation method was modelled and applied to four different battery technologies and a variety of different Hybrid battery combinations. The battery technologies examined in this thesis are Lithium iron phosphate (LiFePO4), Sealed lead-acid (VRLA), Nickel-Cadmium (NiCd) and Flooded lead-acid. This method takes into account different DOD (Depth of Discharge) levels, Temperature effects and lifetime curves extracted from different datasheets for the different technologies, in order to be able to model the different battery behaviours as accurately as possible.

For the modelling of the different Hybrid battery configurations two different power management schemes were constructed and applied to the dynamic lifetime estimation that is proposed. This method was validated and concluded to be realistic as it was able to capture the different battery characteristics for the different technologies.

For the comparison of the different storage options, a singular function was constructed for both single and hybrid battery options. This function quantifies both upfront and lifetime battery costs while, at the same time, captures the future battery cost trends that are foreseen for different technologies. This function was used as the point of reference in the comparison of every storage option that was tested.

The optimum storage option that minimizes the battery cost function for the two sets of PV-load Case studies used in this project, was found to be a Hybrid battery that consists of LiFePO4 and Sealed lead-acid technology, with different sizes and configurations.

In conclusion, the lifetime method proposed as well as the battery cost function constructed could be easily adopted and used for a variety of different battery technologies in low-power applications. A prerequisite for the above mentioned realization is that cycle life data of the examined technology are either provided or easy to be constructed.