2020, Hempel, Sabrina, Adolphs, Julian, Landwehr, Niels, Willink, Dilya, Janke, David, Amon, Thomas

A reliable quantification of greenhouse gas emissions is a basis for the development of adequate mitigation measures. Protocols for emission measurements and data analysis approaches to extrapolate to accurate annual emission values are a substantial prerequisite in this context. We systematically analyzed the benefit of supervised machine learning methods to project methane emissions from a naturally ventilated cattle building with a concrete solid floor and manure scraper located in Northern Germany. We took into account approximately 40 weeks of hourly emission measurements and compared model predictions using eight regression approaches, 27 different sampling scenarios and four measures of model accuracy. Data normalization was applied based on median and quartile range. A correlation analysis was performed to evaluate the influence of individual features. This indicated only a very weak linear relation between the methane emission and features that are typically used to predict methane emission values of naturally ventilated barns. It further highlighted the added value of including day-time and squared ambient temperature as features. The error of the predicted emission values was in general below 10%. The results from Gaussian processes, ordinary multilinear regression and neural networks were least robust. More robust results were obtained with multilinear regression with regularization, support vector machines and particularly the ensemble methods gradient boosting and random forest. The latter had the added value to be rather insensitive against the normalization procedure. In the case of multilinear regression, also the removal of not significantly linearly related variables (i.e., keeping only the day-time component) led to robust modeling results. We concluded that measurement protocols with 7 days and six measurement periods can be considered sufficient to model methane emissions from the dairy barn with solid floor with manure scraper, particularly when periods are distributed over the year with a preference for transition periods. Features should be normalized according to median and quartile range and must be carefully selected depending on the modeling approach.

2021, Doumbia, E. Moustapha, Janke, David, Yi, Qianying, Zhang, Guoqiang, Amon, Thomas, Kriegel, Martin, Hempel, Sabrina

The tracer gas method is one of the common ways to evaluate the air exchange rate in a naturally ventilated barn. One crucial condition for the accuracy of the method is that both considered gases (pollutant and tracer) are perfectly mixed at the points where the measurements are done. In the present study, by means of computational fluids dynamics (CFD), the mixing ratio NH3/CO2 is evaluated inside a barn in order to assess under which flow conditions the common height recommendation guidelines for sampling points (sampling line and sampling net) of the tracer gas method are most valuable. Our CFD model considered a barn with a rectangular layout and four animal-occupied zones modeled as a porous medium representing pressure drop and heat entry from lying and standing cows. We studied three inflow angles and six combinations of air inlet wind speed and temperatures gradients covering the three types of convection, i.e., natural, mixed, and forced. Our results showed that few cases corresponded to a nearly perfect gas mixing ratio at the currently common recommendation of at least a 3 m measurement height, while the best height in fact lied between 1.5 m and 2.5 m for most cases.

2020, Saha, Chayan Kumer, Yi, Qianying, Janke, David, Hempel, Sabrina, Amon, Barbara, Amon, Thomas

Airflow inside naturally ventilated dairy (NVD) buildings is highly variable and difficult to understand due to the lack of precious measuring techniques with the existing methods. Computational fluid dynamics (CFD) was applied to investigate the effect of different seasonal opening combinations of an NVD building on airflow patterns and airflow rate inside the NVD building as an alternative to full scale and scale model experiments. ANSYS 2019R2 was used for creating model geometry, meshing, and simulation. Eight ventilation opening combinations and 10 different reference air velocities were used for the series of simulation. The data measured in a large boundary layer wind tunnel using a 1:100 scale model of the NVD building was used for CFD model validation. The results show that CFD using standard k-ε turbulence model was capable of simulating airflow in and outside of the NVD building. Airflow patterns were different for different opening scenarios at the same external wind speed, which may affect cow comfort and gaseous emissions. Guiding inlet air by controlling openings may ensure animal comfort and minimize emissions. Non-isothermal and transient simulations of NVD buildings should be carried out for better understanding of airflow patterns.

2021, Nosek, Štěpán, Jaňour, Zbyněk, Janke, David, Yi, Qianying, Aarnink, André, Calvet, Salvador, Hassouna, Mélynda, Jakubcová, Michala, Demeyer, Peter, Zhang, Guoqiang

Ammonia emissions from naturally ventilated livestock buildings (NVLBs) pose a serious environmental problem. However, the mechanisms that control these emissions are still not fully understood. One promising method for understanding these mechanisms is physical modelling in wind tunnels. This paper reviews studies that have used this method to investigate flow or pollutant dispersion within or from NVLBs. The review indicates the importance of wind tunnels for understanding the flow and pollutant dispersion processes within and from NVLBs. However, most studies have investigated the flow, while only few studies have focused on pollutant dispersion. Furthermore, only few studies have simulated all the essential parameters of the approaching boundary layer. Therefore, this paper discusses these shortcomings and provides tips and recommendations for further research in this respect.