In the Netherlands an odour research program was started to study odour from conventional housing systems and also new low ammonia systems using a new official odour measurement protocol. These measurements should give the basis for the integration of the new systems in the odour regulatory system. The results of the first year program are presented and emphasis is made on the new protocol and calibration of the Pig odour unit evaluation. The protocol evaluation was done for two conventional pig housing systems: fattening pigs (FP) and sows in individual stalls (SP) both on partly slatted floors and for each system four farms were chosen. The experimental design and analysis was done to obtain the variation between and within the farms and also the variation coming from the olfactometric measurements for both systems. The variations between-farm were 22% (for FP) and 2% (for SP) mean percentage deviations, the variation within-farm was 34% (FP) and 19% (SP) and the variation between duplicate samples were 16% (FP) and 23% (SP). For the FP system the odour emission geometric mean was 22.6 odour units/s per animal place for a ventilation rate of 33.3 m3/h per animal. The protocol used prescribes ten sampling days for the sampling of single farm and considering the estimated variation sources, the accuracy levels obtained are of 24.5% for FP and 8.1% for SP. Higher accuracy, 9.5% for FP and 6.7% for SP could be obtained with the same number of odour samples by taking one sample on 20 different farms. On conclusion, all variation sources should considered for each animal category before the protocol design of housing systems evaluation. The resulting sample strategies could then be different between animal categories.

When odor is regulated and controlled, standard protocols should be establish for odor determination of specific housing systems. Those protocols may be quite complex to establish because all variation sources have to be considered for all animal categories. For example, the growing stage in a fattening building will have an impact on those variations (beginning of the period vs. end of the period for all-in all-out).