Solar energy is one of the best candidates, particularly in rural areas where electricity supply is considered to be a challenge. Before any solar system installation, a sizing process is required. Sizing involves determining the size of the PV array and the energy storage capacity, considering weather conditions, energy demand and cost. In this context, the research presented focuses on the development of an optimal sizing methodology (Iterative method) based on the development of two models: a reliability model based on the Probability of Energy Deficit (PED) and an economic model based on Total Net Present Cost (TNPC). A case study is conducted to analyse an autonomous system for internal lighting for a research laboratory located at the Ecole Nationale Polytechnique ENP in Algiers, Algeria. The results found are compared with those determined using the conventional (intuitive) method and PVsyst software. These results clearly show the superiority of the proposed iterative method over the other two methods.