Infrared photoacoustic spectroscopy in animal houses: Effect of non-compensated interferences on ammonia, nitrous oxide and methane air concentrations

Reduction of ammonia and greenhouse gas emissions is a major goal in countries where the livestock sector has a high social and economic importance. Estimates of emissions per building and per annum usually have uncertainties higher than the reduction targets. Therefore, mitigation strategies and public funding efficiency will benefit from progress and standardisation in measuring and modelling emissions from livestock buildings. Quantification of livestock emissions requires accurate measurements of trace gas concentrations. Infrared photoacoustic spectroscopy has been used for decades because of its selectivity, reduced drift in calibration, the existence of industrial analysers ease of use and the reduced cost per measurement. Its use can be further improved through better knowledge of measurement uncertainty. Although the infrared bands selected by optical filters are narrow, the risk of interference bias exists due to the high number and diversity of gas species. The emitted gases depend on the management of animals, feed, manure, heating and motorised equipment. Non-compensated interferences between gases induce an overestimation of gas concentrations, and a cascade effect induces underestimation. This study demonstrates two examples in broiler and dairy cow buildings, which lead either to overestimation (e.g. 12 instead of 2 mg [NH3] m(-3) and 4 instead of 1 mg [N2O] m(-3) or underestimation (e.g. 2 instead of 4 mg [CH4] m(-3)). The detection of interference bias and the reduction of uncertainty in concentration measurements are discussed. These results can be used to provide uncertainty estimates for previously published emission factors that were based on infrared photoacoustic spectroscopy. (C) 2013 IAgrE. Published by Elsevier Ltd. All rights reserved.