Artificial intelligence (AI) plays a central role in environmental studies and recently, its usage in carbon capture technology has considerably increased. Carbon capture technology is responsible for tackling climate change by trapping carbon dioxide emissions produced in power plants. However, the current systems are not efficient and consume considerable amounts of energy.
In this light, researchers from the University of Surrey focussed their studies on enhancing the efficiency of CO2 capture through AI. Their employment of AI algorithms resulted in a 16.7% increase in CO2 capture and a significant 36.3% decrease in energy consumed from the UK’s national grid.
The heart of this system is a packed bubble column (PBC), an interface between freshwater containing crushed limestone and flue gases filled with CO2. This interaction transforms CO2 into bicarbonate. Machine learning techniques were applied to create surrogate dynamic models which forecasted and optimized the CO2 capture rate and power consumption of the reactor. These models were then trained based on data obtained from physics simulations. In addition, LSTM-based models were utilized for predicting wind energy availability and future CO2 concentrations in flue gas.
Researchers noted the rigid conventional nature of carbon capture systems that typically run at a constant rate. They found teaching the system to make minor adjustments led to significant energy savings and simultaneously increased carbon capture. Although this research focuses on enhanced weathering, the findings provide valuable insights applicable to other carbon capture systems. The model can assist in capturing and storing CO2 more effectively, with less energy needed.
By using these predictive models, the algorithm altered the amount of water pumped, depending on variables like CO2 levels and wind speeds. In doing so, energy is conserved during times of reduced CO2 or decreased wind power. In a month, there was a 16.7% increase in carbon dioxide capture rates using this system as compared to traditional methods. The dependence on renewable energy drastically fell from an average of 92.9% to 56.6%.
These findings demonstrate how AI can be used in carbon capture technologies and help address issues arising from varying CO2 levels. This research contributes significantly to the pursuit of the UN’s sustainability goals and offers hope for a more environmentally-friendly future. With further refinements, this technique could make significant contributions towards global sustainability efforts. As our planet seeks resolutions to climate change, this research offers hope and anticipates a cleaner environment for future generations.
Full credit for this research is given to the scholars of the project. For more information about this and other projects, join our online communities on Reddit, Facebook, Discord, and LinkedIn. If you appreciate our work, be sure to sign up for our newsletter and join our Telegram channel.
