Direct Measurements of the Volume Flow Rate and Emissions in a Large Naturally Ventilated Building

• Published in: Sensors (Basel, Switzerland) Vol. 20; no. 21; p. 6223
• Main Authors: Janke, David, Yi, Qianying, Thormann, Lars, Hempel, Sabrina, Amon, Barbara, Nosek, Štěpán, van Overbeke, Philippe, Amon, Thomas
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 31.10.2020; MDPI
• Subjects: Accuracy; Aerodynamics; air exchange; Air flow; Atmospheric boundary layer; Barns; Boundary conditions; boundary layer wind tunnel; Carbon dioxide; Computational fluid dynamics; dairy barn; flame ionization detector; Flame ionization detectors; Flow velocity; Foreign exchange rates; Intensive farming; Investigations; laser Doppler anemometer; Methods; Sensors; Systematic errors; Tracer gas; Velocity; Velocity measurement; Ventilation; Wind tunnels
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s20216223

The direct measurement of emissions from naturally ventilated dairy barns is challenging due to their large openings and the turbulent and unsteady airflow at the inlets and outlets. The aim of this study was to quantify the impacts of the number and positions of sensors on the estimation of volume flow rate and emissions. High resolution measurements of a naturally ventilated scaled building model in an atmospheric boundary layer wind tunnel were done. Tracer gas was released inside the model and measured at the outlet area, using a fast flame ionization detector (FFID). Additionally, the normal velocity on the area was measured using laser Doppler anemometry (LDA). In total, for a matrix of 65 × 4 sensor positions, the mean normal velocities and the mean concentrations were measured and used to calculate the volume flow rate and the emissions. This dataset was used as a reference to assess the accuracy while systematically reducing the number of sensors and varying the positions of them. The results showed systematic errors in the emission estimation up to +97%, when measurements of concentration and velocity were done at one constant height. This error could be lowered under 5%, when the concentrations were measured as a vertical composite sample.