Mechanistic models of ammonia release from liquid manure: a review

Mechanistic models of ammonia (NH3) release from liquid manure in 30 available publications are reviewed. They include models of NH3 release in animal houses, from slurry applied fields, from constructed manure stores, from an anaerobic lagoon and from flooded rice and soil systems. The models are decomposed and their general structure is synthesized as a core model combined with some submodels. The theoretical bases of the models are compared. The general model structure, core model and different submodels are discussed. The core model is a model of convective mass transfer. Determination of the convective mass transfer coefficient is essential for modelling NH3 release. This coefficient ranges from 11.7 x 10(-3) to 1.3 x 10(-6) m/s as investigated in laboratory and field experiments. It usually suffices to express it as a function of air velocity at the manure surface and air or manure temperature. Determination of the gaseous NH3 concentration at the manure surface is also essential. This concentration can be calculated using some submodels, which may include Henry's constant, dissociation constant, pH change coefficient, manure production by animals, NH3 generation in manure and NH3 diffusion in manure. The Henry's constant found in the reviewed reports has nine different definitions and units. It is in a state of confusion and a standard Henry's constant is needed. For NH3 in animal manure, the dissociation constant ranged from one-sixth to one-fifth of that for NH3 in water. The pH change coefficient describes a dynamic change of pH in the surface manure as influenced by the releases of carbon dioxide (CO2) and NH3. The manure production submodel is still insufficiently developed. The submodel of NH3 generation in manure is usually neglected. The submodel of NH3 diffusion in manure is based on Fick's law of diffusion. (C) 1999 Silsoe Research Institute.