Variation analysis of natural ventilation and air temperature within a greenhouse cluster using a large-scale 3D CFD modelling

Abstract

There is an increasing number of centralized vegetable productions that appear as clusters of greenhouse rows and columns. Differences in temperature distributions and ventilation rate are not well-known and quantified between these greenhouses. Very few greenhouse microclimate models consider the influence of surrounding greenhouse shading on natural ventilation and the variation caused by it on mass and heat transfer. This paper presents a three-dimensional greenhouse cluster model that covers a total area of 3 ha for an interior and exterior flow field, including 22 greenhouses, which is solved using the finite volume method in conjunction with a supercomputer. One of the major contributions is a proposed optimal mesh design methodology for large-scale greenhouse cluster models. It allows for precise control over the quality and size of the mesh elements. Results show that the simulated MAE (Mean Absolute Error) at the lower/upper vents air speed measurement points and 15 distributed temperature measurement points in a greenhouse are respectively 0.19 m s⁻¹, 0.24 m s⁻¹ and 1.67 ℃. The model was also used to assess the variation of natural ventilation rate and temperature distribution that depends on greenhouse specific position inside a farm. This work contributes to refining greenhouse models and control methods, while also serving as a potential reference for both researchers and practitioners in crop yield and quality prediction.