Using AI to plan for climate variability in renewable energy systems

Harnessing solar and wind energy for sustainable power generation is contingent upon the unpredictable nature of weather. Apart from seasonal fluctuations and hourly changes, energy planners must also factor in rare and intense weather occurrences. Conventional planning techniques often struggle to accurately reflect these events, leading to either excessively costly or unreliable systems.

In response, researchers sought to devise a more efficient approach to planning energy systems that can withstand genuine climate variability. They developed an artificial intelligence model trained on 30 years of wind and solar data collected from Pingtan, China. This model generated thousands of realistic weather scenarios, encompassing both average conditions and extreme events.

The researchers utilized these scenarios to ascertain the most economical and dependable blend of energy technologies. Their Integrated Energy System comprised solar panels, wind turbines, batteries, hydrogen production and storage facilities, and a connection to the national power grid. By evaluating the system's performance across all AI-generated scenarios, the optimal configuration was identified.

The optimization revealed that the most effective renewable energy system is a synergistic combination of multiple technologies. Wind and solar serve as the primary sources of electricity, while batteries regulate short-term variations in supply and demand. Hydrogen storage provides backup during prolonged periods of low renewable generation, and the grid functions as an additional safety measure.

The key insight is that batteries and hydrogen have complementary functions: batteries excel at managing daily variations, whereas hydrogen is more suited for storing energy over weeks or months. By combining these technologies, a dependable and cost-effective renewable energy system is achieved.