Techno-economic Analysis of Grid-connected Photovoltaic Systems Under Sahelian Climate: the Case of Burkina Faso

Abstract

Solar energy has emerged as a promising alternative for global socio-economic development, especially in the abundant sunshine of the Sahelian climate zone. Solar photovoltaic (PV) systems have become more popular for converting sunlight into electricity. As a result, Burkina Faso already has many PV power plants linked to the national grid, and more projects are being planned. However, the injection of solar PV system energy into the power grid has been identified as a potential source of disruptions. Can these grid-connected solar PV systems effectively address Burkina Faso’s persistent issue of prolonged load shedding? The extensive potential for PV interconnection necessitates a careful examination of regulations limiting the injection of surplus energy. Therefore, there is a critical need to investigate the technical and economic efficiency of Grid-connected Solar PV Systems (GCPVS) implementation in Burkina Faso. Accordingly, it is necessary to conduct thorough performance research in a given climate, like Burkina Faso, using a combination of qualitative and quantitative performance metrics. This research should evaluate energy efficiency and material appropriateness, including module and inverter types. Furthermore, doing a feasibility study that covers various climatic locations in Burkina Faso would help evaluate the solar energy capacity and the corresponding expenses of generating power from GCPVS using PVSyst and Python. The analysis of the power system's behaviour when incorporating PV power flow through MATLAB R2022b/SIMULINK/PSAT software will identify specific injection points. To keep the grid stable for the best control of PV integration, it might be helpful to come up with a good forecasting method that uses deep learning models like Gate Recurrent Unit (GRU) and Long Short-Term Memory (LSTM). The first results proved that the months with high solar radiation were the most energy-productive, indicating a direct correlation between solar irradiance and energy generation. The PV plant had the highest conversion efficiency during the rainy season (July and August) and the lowest during the hot season (March and April) when module temperatures reached about 47°C. Efficiency declined from 12.29% in 2019 to 12.10% in 2021. The performance ratio ranged from 80.73% in 2019 to 79.36% in 2021, and the capacity factor from 19.89% to 19.33%. The final yield was 4.89 h/d in 2019, 4.61 h/d in 2020, and 4.92 h/d in 2021, indicating a performance decline over time. The study highlights the need for improved and consistent module cleaning systems to enhance effectiveness and underscores the utility of PVGIS-SARAH2 for estimating solar irradiation. Furthermore, the economic feasibility of installing a 50 MW GCPVS in each climatic zone was compared using PVSyst and Python. The results showed that Dori, in the Sahelian zone, v demonstrated superior economic feasibility with an LCOE of $0.0825/kWh compared to $0.0853/kWh and $0.0859/kWh in Ouagadougou and Bobo-Dioulasso, respectively. This information is crucial for policymakers and private investors considering investments in GCPVS installations in the Sahel region. Furthermore, a simulation of the national electrical grid identified suitable junctions for the integration of photovoltaic solar power facilities. The installation of a 40 MWp solar PV generator in low-voltage buses led to the identification of weaker buses and an improvement in overall network voltage. However, inappropriate voltages at certain locations beyond the normal voltage range, which should be within ± 5% of the nominal voltage according to the grid code, were identified. This will necessitate network restructuring for a more efficient injection of solar photovoltaic flux. Finally, the study on deep learning techniques, specifically LSTM and GRU, for forecasting the PV power output from the Zagtouli plant involved three evaluation metrics: Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and coefficient of determination (R2). The RMSE evaluation criteria gave 10.799(LSTM), 11.695(GRU) and 10.629(LSTM-GRU) giving the LSTM-GRU model as the best for RMSE evaluation. The MAE evaluation provided 2.09, 2.1, and 2.0 for the LSTM, GRU, and LSTM-GRU models respectively, showing that the LSTM-GRU model is superior for MAE evaluation. The R2 criteria similarly showed the LSTM-GRU model to be best with 0.999 compared to 0.998 for LSTM and 0.997 for GRU. It becomes evident that the hybrid LSTM-GRU model exhibits superior predictive capabilities compared to the other two models. These results indicate that the hybrid LSTM-GRU model has the potential to reliably predict the solar PV power output. It is therefore recommended that the authorities in charge of the solar PV Plant in Ouagadougou should consider switching to the deep learning LSTM-GRU model. These approaches aim to enhance the profitability of solar photovoltaic power plants, contributing to improved energy supply for industrialization and supporting the development of vulnerable populations in the Sahel region.