Access to electricity might in some parts of the world seem evident. However, Ethiopia struggles to provide its large and growing population with electricity. Although around all the households in the capital Addis Ababa are connected to the electricity grid, the grid is unreliable and results in daily outages. As the photovoltaic (PV) potential in Addis Ababa on the other hand is great, this thesis examines the feasibility and profitability of decentralized PV adoption with battery and hydrogen storage respectively. Based on an ongoing construction project in the sub-city Yeka, Addis Ababa, a reference building was used to simulate the PV systems with battery and hydrogen storage. Furthermore, a load profile based on time-use diaries was developed and used in the simulations, as data on household electric consumption was non-existent. The load profile resulted in an average daily use of 1341 kWh and a 165 kW peak for all of the 130 apartments in the reference building. The results of the simulations indicated that neither of the two systems were feasible nor profitable to implement on the reference building. The PV-system with battery storage was cheaper and required less installed PV capacity, however the cost of energy for both systems was significantly higher than the current cost of energy in Ethiopia. The installed PV capacity of both systems exceeded the maximum capacity that was feasible on the reference building.

Grid-connected photovoltaic (PV) capacity is increasing worldwide, mainly due to extensive subsidy schemes for renewable electricity generation. A majority of newly installed systems are distributed small-scale systems located in distribution grids, often at residential customers. Recent developments suggest that such distributed PV generation (PV-DG) could gain more interest in Sweden in the near future. With prospects of decreasing system prices, an extensive integration does not seem impossible. In this PhD thesis the opportunities for utilisation of on-site PV generation and the consequences of a widespread introduction are studied. The specific aims are to improve modelling of residential electricity demand to provide a basis for simulations, to study load matching and grid interaction of on-site PV and to add to the understanding of power system impacts. Time-use data (TUD) provided a realistic basis for residential load modelling. Both a deterministic and a stochastic approach for generating different types of end-use profiles were developed. The models are capable of realistically reproducing important electric load properties such as diurnal and seasonal variations, short time-scale fluctuations and random load coincidence. The load matching capability of residential on-site PV was found to be low by default but possible to improve to some extent by different measures. Net metering reduces the economic effects of the mismatch and has a decisive impact on the production value and on the system sizes that are reasonable to install for a small-scale producer. Impacts of large-scale PV-DG on low-voltage (LV) grids and on the national power system were studied. Power flow studies showed that voltage rise in LV grids is not a limiting factor for integration of PV-DG. Variability and correlations with large-scale wind power were determined using a scenario for large-scale building-mounted PV. Profound impacts on the power system were found only for the most extreme scenarios. / Felaktigt tryckt som Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 711