Developing the 85Kr tracer gas technique for air exchange rate measurements in naturally ventilated animal buildings

• Published in: Biosystems engineering Vol. 109; no. 4; pp. 276 - 287
• Main Authors: Samer, M., Müller, H.-J., Fiedler, M., Ammon, C., Gläser, M., Berg, W., Sanftleben, P., Brunsch, R.
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.08.2011; Elsevier
• Subjects: Agricultural machinery and engineering; Agronomy. Soil science and plant productions; air; ammonia; animals; Biological and medical sciences; buildings; carbon dioxide; emissions; Fundamental and applied biological sciences. Psychology; Generalities. Biometrics, experimentation. Remote sensing; linear models; methane; nitrous oxide; regression analysis; summer; a; b; c; MP; hpu; AER; Bq; a 2; b 1; a 1; TG; AU; c 2; ANOVA; b 2; c 1; GBq; Z; RM; TFL; MeV; Bioengineering; Engineering; Tracer technique; Agriculture; Air; Livestock farm; Natural ventilation
• ISSN: 1537-5110; 1537-5129
• DOI: 10.1016/j.biosystemseng.2011.04.008

Experiments were performed to study air exchange rates (AER) occurring in naturally ventilated dairy buildings during summer seasons 2006 to 2010. A tracer gas technique (TG) for AER measurements was developed. The AERs were determined by decay of radioactive tracer Krypton-85, and CO 2-balance used as the reference method (RM). During each experiment, continuous measurements of gaseous concentrations (NH 3, CO 2, CH 4 and N 2O) inside and outside the building and 85Kr tracer gas experiments were performed. The combined factors investigated were release over feeding table (a 1) or over the manure alley (a 2), average α-values (b 1) or the sum of impulses (b 2), selected radiation counts (c 1) or all radiation counts (c 2). The results were compared using Pearson correlation analysis, developing a linear regression model, and testing the differences between the factor combinations and the RM using an ANOVA model. There were differences between impulses (Pr > |t| = 0.0013), where the sum of impulses showed better results than the average α-values. Although there was no difference (Pr > |t| = 0.344) between the readings of the radiation counts, it was considered that by using all the readings of the radiation counters it was more representative and easier to calculate the AER. The best factor combinations, having the highest coefficient of determination values R 2, and the most reliable parameter estimates, were: a 1b 2c 2 ( R 2 = 0.94; 1.63 ± 0.14); and a 1b 1c 1 ( R 2 = 0.97; 2.03 ± 0.11). The gaseous emissions, subject to the RM, were 3.9, 19, 1656, and 0.96 g h −1 AU −1 (AU is animal unit of 500 kg) for NH 3, CH 4, CO 2 and N 2O respectively. ► Air exchange rate measurements compared by correlation analysis, linear regression, and ANOVA. ► Factor combinations are the release locations of the tracer gas and the calculation procedures. ► Release locations were over the manure alley and the feeding table. ► Sum of impulses lead to better results than the average α-values. ► Best factor combinations included release over feeding table.