A mechanistic model for simulating methane emissions from unstirred liquid manure storages

The objective of the study was to develop a mechanistic model of methane (CH 4 ) producing processes in unstirred conditions with potential application for estimating CH 4 emissions from anaerobic manure storage facilities. Although models for describing anaerobic digestion processes are available, they largely relate to anaerobic digesters, and do not directly apply to the prediction of CH 4 emissions from liquid manure storage. Based on extant models, six biochemical steps were described: hydrolysis, acetogenesis, hydrogenogenesis, homoacetogenesis, hydrogenous methanogenesis and acetic methanogenesis, performed by five bacterial groups. The model contains six state variables, and mass flow is mostly generated and quantified using bacterial kinetics. The model was coded in acslX and a fourth-order Runge-Kutta method with an integration step size of 0.05 d was used for numerical integration. The time courses of CH 4 production and volatile fatty acid (VFA) concentration of two laboratory-scale liquid manure storage tanks, both filled with liquid sow manures and running in unstirred and constant 25°C conditions, were well predicted, with correlation coefficients over 0.90. Discrepancies between predicted and measured CH 4 production and VFA concentration were mainly due to random variation of observed data. The model was sensitive to parameters describing hydrolysis and the kinetics of acetogenic and acetate methanogenic bacteria. Simulations based on the Intergovernmental Panel on Climate Change model (Tier II) predicted 260 g CH 4 kg -1 volatile solids (VS, assuming maximum CH 4 producing capacity of 0.48 and methane conversion factor of 80%), whereas the measured value was 78.3 g CH 4 kg -1 VS after 146 d and the mechanistic model predicted 74.8 g CH 4 kg -1 VS. The model developed in this study appears to be better suited to batch manure storage than the IPCC model.