Testing a simple model of gas bubble dynamics in porous media

Bubble dynamics in porous media are of great importance in industrial and natural systems. Of particular significance is the impact that bubble-related emissions (ebullition) of greenhouse gases from porous media could have on global climate (e.g., wetland methane emissions). Thus, predictions of future changes in bubble storage, movement, and ebullition from porous media are needed. Methods exist to predict ebullition using numerical models, but all existing models are limited in scale (spatial and temporal) by high computational demands or represent porous media simplistically. A suitable model is needed to simulate ebullition at scales beyond individual pores or relatively small collections (<10(-4) m(3)) of connected pores. Here we present a cellular automaton model of bubbles in porous media that addresses this need. The model is computationally efficient, and could be applied over large spatial and temporal extent without sacrificing fine-scale detail. We test this cellular automaton model against a physical model and find a good correspondence in bubble storage, bubble size, and ebullition between both models. It was found that porous media heterogeneity alone can have a strong effect on ebullition. Furthermore, results from both models suggest that the frequency distributions of number of ebullition events per time and the magnitude of bubble loss are strongly right skewed, which partly explains the difficulty in interpreting ebullition events from natural systems.