Carbon capture from power plants could be climate friendlier and consume less energy thanks to AI modeling, according to a study from the University of Surrey.   

Researchers there adjusted a capture system based on a real coal-fired power station. Their aim was to self-optimize the CO2 capture process in a renewable energy system via enhanced weathering of calcite with fresh water in a packed bubble column (PBC) reactor.

Through that process, CO2 is captured by bubbling the flue gas through fresh water containing limestone in the reactor, converting the CO2 into bicarbonate and storing it in the ocean.

But since it takes energy to pump the water and the CO2, the capture system had its own wind turbine – but when the wind wasn’t blowing, it took energy from the grid.

Using AI, researchers taught a model system to predict what would happen so it could pump less water when there was less CO2 to capture, or when less renewable energy was available.

Specifically, two deep learning models were considered to capture the dynamics of the PBC reactor: a long short-term memory network (LSTM) and a two-stage multilayer perceptron network (MLP).

Data-driven models based on LSTM were developed to predict wind energy (R2: 0.908) and inlet flue gas CO2 concentration (R2: 0.981) using publicly available datasets.

A multi-objective NSGA-II genetic algorithm was then applied that utilized the inlet flue gas CO2 concentration and wind energy predictions to pre-emptively self-optimize the reactor process conditions (i.e., superficial liquid flow rate and superficial gas flow rate) to maximize the CO2 capture rate and minimize non-renewable energy consumption.

Schematic representation of (A) enhanced weather packed bubble column reactor and (B) the model predictive control framework developed to maximize enhanced weathering reactor CO2 capture rate, whilst simultaneously minimizing non-renewable energy consumption (Source: Reaction Chemistry & Engineering (2023). DOI: 10.1039/D3RE00544E).

Researchers said the model could capture 16.7% more carbon dioxide over a one-month operation while using 36.3% less energy from the National Grid.  

See the full body of research, which has been published in the journal Reaction & Chemistry Engineering

This post appeared first on Power Engineering.