Scholarlink

Sustainability faces critical challenges, including the demand for renewable energy and the environmental impacts of pollution. Efficient wastewater treatment, combined with energy recovery, offers a strategic solution. Anaerobic sludge digesters in Wastewater Treatment Plants (WWTPs) convert organic waste into biogas, primarily methane (%CH \(_4\) ), a renewable energy source. This study applies Deep Learning (DL) models to optimize operational efficiency and boost renewable energy production in Portuguese WWTPs. Using time series analysis, it forecasts %CH \(_4\) in biogas production. Models evaluated include Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), Temporal Convolutional Networks (TCN), and Gated Recurrent Units (GRU), employing a multistep recursive approach. CNNs performed best, achieving a low RMSE of 0.061%. These results demonstrate the potential of CNNs to enhance WWTP energy efficiency and contribute to environmental sustainability.

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

Towards Efficient Biogas Production: Deep Learning-Based Methane Forecasting in Anaerobic Digesters of Wastewater Treatment Plants

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Duarte Lucas,
  • Pedro Oliveira,
  • Afonso Bessa,
  • Francisco S. Marcondes,
  • Manuel Rodrigues

摘要

Sustainability faces critical challenges, including the demand for renewable energy and the environmental impacts of pollution. Efficient wastewater treatment, combined with energy recovery, offers a strategic solution. Anaerobic sludge digesters in Wastewater Treatment Plants (WWTPs) convert organic waste into biogas, primarily methane (%CH \(_4\) ), a renewable energy source. This study applies Deep Learning (DL) models to optimize operational efficiency and boost renewable energy production in Portuguese WWTPs. Using time series analysis, it forecasts %CH \(_4\) in biogas production. Models evaluated include Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), Temporal Convolutional Networks (TCN), and Gated Recurrent Units (GRU), employing a multistep recursive approach. CNNs performed best, achieving a low RMSE of 0.061%. These results demonstrate the potential of CNNs to enhance WWTP energy efficiency and contribute to environmental sustainability.