The storage of compressed green hydrogen in tanks, utilizing renewable energy sources such as wind and solar power, is explored in this research. Data from 2023, divided into four trimesters, was used to calculate hydrogen storage volumes. For 11 m3 tanks, the volumes stored were 322 m3, 293.34 m3, 278.29 m3, and 285.74 m3, filling 30, 27, 26, and 26 tanks, respectively. Efficiency ranged between 70 and 71.36%, with losses attributed to system inefficiencies. The XGBoost algorithm was applied for predictions, achieving R2 scores of 0.966 for wind power and 0.996 for solar power. Wind speed and global irradiance emerged as key predictors, while temporal features refined daily and seasonal trends. Wind production peaked in winter afternoons, while solar energy reached its maximum during summer noon. Simulated scenarios validated the model’s accuracy under varying conditions. In a summer noon scenario (25 °C, 5 m/s wind, 800 W/m2 irradiance), wind and solar produced 461.81 kW and 156.13 W, respectively. In a winter evening (5 °C, 8 m/s wind, 100 W/m2 irradiance), wind generated 1891.59 kW, and solar produced 22.00 W. These findings emphasize the potential of renewable energy for efficient and sustainable hydrogen storage solutions.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Predictive Analytics for Compressed Green Hydrogen Storage: Combined XGBoost and Engineering Approach to Renewable Energy Optimization

  • Qamar Limami,
  • Fahd Azouz,
  • Soukaina Jaafari,
  • Ahmed Bazzi,
  • Ahmed Khallaayoun

摘要

The storage of compressed green hydrogen in tanks, utilizing renewable energy sources such as wind and solar power, is explored in this research. Data from 2023, divided into four trimesters, was used to calculate hydrogen storage volumes. For 11 m3 tanks, the volumes stored were 322 m3, 293.34 m3, 278.29 m3, and 285.74 m3, filling 30, 27, 26, and 26 tanks, respectively. Efficiency ranged between 70 and 71.36%, with losses attributed to system inefficiencies. The XGBoost algorithm was applied for predictions, achieving R2 scores of 0.966 for wind power and 0.996 for solar power. Wind speed and global irradiance emerged as key predictors, while temporal features refined daily and seasonal trends. Wind production peaked in winter afternoons, while solar energy reached its maximum during summer noon. Simulated scenarios validated the model’s accuracy under varying conditions. In a summer noon scenario (25 °C, 5 m/s wind, 800 W/m2 irradiance), wind and solar produced 461.81 kW and 156.13 W, respectively. In a winter evening (5 °C, 8 m/s wind, 100 W/m2 irradiance), wind generated 1891.59 kW, and solar produced 22.00 W. These findings emphasize the potential of renewable energy for efficient and sustainable hydrogen storage solutions.