Multi-Objective Solar–Wind Portfolio Optimization for Energy Yield and Low-Output Persistence: A Climate-Informed Assessment in Eastern Anatolia, Türkiye

Seda Aktürk, A. Aktürk

Abstract

Renewable capacity planning must consider not only mean energy yield but also the frequency and persistence of low-generation periods. The main contribution of this study is a province-specific, three-objective solar–wind portfolio optimization framework that treats energy yield, low-output frequency and low-output persistence jointly and tests the selected portfolios on independent weather years. The framework was applied to 14 provinces of Eastern Anatolia, Türkiye, using hourly ERA5 and ERA5-Land data for 2020–2024 to optimize the photovoltaic share and spatial wind capacity allocation against the mean capacity factor, the percentage of hours below 0.05 and the longest consecutive low-output event. Across 840 independent runs, the non-dominated sorting genetic algorithm III (NSGA-III) outperformed the decomposition-based MOEA/D in 12 of 14 provinces. In the independent 2023–2024 test period, Maximum Energy portfolios, solar-only in all provinces, achieved a mean capacity factor of 0.177, versus 0.137 and 0.110 for Balanced and Frequency Priority portfolios, whereas low-output frequency fell only from 54.7% to 51.6% and 49.7%. Fewer low-output hours did not shorten persistent events: the mean longest event increased from 20.5 h to 29.5 h, and the frequency ranking reversed at a threshold of 0.10. Bias correction reduced the mean absolute wind speed bias of ten CMIP6 models from 0.70 to 0.04 m s−1. An indicative cost screening showed that adding wind at the reanalysis-derived capacity factors raised the levelized cost from about 41 to 63–88 USD MWh−1. Because a 3-percentage-point reduction in low-output hours cost 22% of the energy yield, yield, low-output frequency and event persistence should be assessed jointly under explicit low-output definitions rather than inferred from mean output.

Source: semanticscholar · PDF

Read the AI summary and key takeaways for traders on WOBR Quant Research.


Open the interactive page on WOBR AI → · WOBR.AI home