Variability analysis of renewable power generation in complex terrain and the contribution of the spatio-temporal synergies for a resilient power supply. Methodology development and Ecuador case study

Abstract

South American countries highly rely on hydropower for their electricity supply and the deployment of remaining untapped hydropower potential is the cornerstone of national generation capacity expansion plans. However, given the dependency of hydropower on water availability, risks associated with weather and climate variability could jeopardize the security of electricity supply of these countries. In the case of Ecuador, according to government plans, fossil fuel thermal power plants will continue playing an important role in meeting the demand during the dry periods. A more sustainable and resilient strategy would be to diversify the power mix focusing on exploiting the complementarities between hydropower and other variable renewable energies, such as wind and solar. However, the deployment of these technologies in the country is still at a very early stage and there are some challenges to be tackled. Due to the varying nature of solar and wind resources, the optimal planning and deployment of solar and wind power that could potentially complement hydropower requires detailed knowledge of the spatial and temporal variability of the resources. Unfortunately, long-term, high-quality solar irradiance and wind speed measurements are generally scarce and sparsely distributed, challenging the characterization of solar and wind resources at a country level. The complex climatic characteristics and topography of Ecuador represent another challenge to better understand the spatio-temporal dynamics between renewable resources, as well as the potential synergies of solar and wind power generation to compensate hydropower during the dry periods. The main research goal of this dissertation is to develop tools and data to support the optimal planning of a more sustainable and resilient power system in Ecuador by systematically investigating the spatio-temporal variabilities and synergies of renewable resources in the complex terrain of Ecuador. For this purpose, climate data, machine learning techniques, and power system modeling tools are used. The lack of solar and wind resource data is addressed by processing satellite-derived solar irradiance data and by using numerical weather prediction models to simulate wind resources. The generated meteorological datasets for Ecuador have a temporal resolution of one hour and a spatial resolution of 3 x 3 km. They comprise 21 years (1998–2018) of solar irradiance data and 14 years (2005–2018) of bias-corrected wind speed and wind direction data at a turbine hub heigh of 80 m. A novel methodology to characterize the spatio-temporal variability of gridded solar and wind resource datasets is proposed and demonstrated. Spatial functional data analysis (sFDA) is used to identify spatial subregions with similar intra-annual variability patterns of solar radiation and wind speed. Finally, the ability of geographically-dispersed photovoltaic (PV) and wind power systems to reduce the power output variability and provide reliable power generation is assessed by using power system performance models. This dissertation provides the first comprehensive spatio-temporal characterization of solar radiation and wind speed in Ecuador. The identified subregions from the sFDA regionalization approach are the basis for the assessment of the complementarity between solar and wind to water resources of existing and planned hydropower plants. One of most important finding is that solar and wind resources have a strong spatio-temporal complementary behavior with water resources from both the Amazon and Pacific basins. This demonstrates that the seasonal variability of hydropower can be compensated by geographically-dispersed PV and wind power systems. Another important finding is that the joint operation of PV and wind power systems from different subregions reduces the intrinsic variability of each resource. Wind power at a high-resource site (52% capacity factor) paired with PV from different subregions can provide the highest level of firm capacity (up to 5.5% of the combined capacity) for 87.5% of the time in a year. Furthermore, wind power from subregions with high resources stabilizes PV power output at diurnal timescales during the windy months (June–September), suggesting that both technologies could serve as baseload during this period, thus reducing the requirements for energy storage. The identified operational benefits of the spatio-temporal synergies among renewable power generation may provide economic incentives to increase the participation of PV and wind power in the Ecuadorian power mix. These findings demonstrate that solar and wind power can play an important role in shaping a more sustainable and resilient power system in the country. The insights gained in this dissertation, as well as the provided data and tools, will support power sector planners and decision-makers in the development of strategies for the optimal expansion of solar and wind power technologies to complement hydropower and to reduce the dependencies on fossil fuel thermal power. This dissertation contributes to the ongoing discussion on renewable energy complementarities in the region and the proposed methodology can be transferred to other countries to support optimal capacity expansion planning. Furthermore, the methods and data developed in this dissertation provide the groundwork for further research into energy system modeling.