Differences between measured and reported volatile organic compound emissions from oil sands facilities in Alberta, Canada

Significance Validation of volatile organic compound (VOC) emission reports, especially from large industrial facilities, is rarely attempted. Given uncertainties in emission reports, their evaluation and validation will build confidence in emission inventories. It is shown that a top-down approach can provide measurement-based emission rates for such emission validation. Comparisons with emission reports from Alberta oil sands surface mining facilities revealed significant differences in VOC emissions between top-down emissions rates and reports. Comparison with VOC species emission reports using currently accepted estimation methods indicates that emissions were underestimated in the reports for most species. This exercise shows that improvements in the accuracy and completeness of emissions estimates from complex facilities would enhance their application to assessing the impacts of such emissions. Large-scale oil production from oil sands deposits in Alberta, Canada has raised concerns about environmental impacts, such as the magnitude of air pollution emissions. This paper reports compound emission rates (E) for 69–89 nonbiogenic volatile organic compounds (VOCs) for each of four surface mining facilities, determined with a top-down approach using aircraft measurements in the summer of 2013. The aggregate emission rate (aE) of the nonbiogenic VOCs ranged from 50 ± 14 to 70 ± 22 t/d depending on the facility. In comparison, equivalent VOC emission rates reported to the Canadian National Pollutant Release Inventory (NPRI) using accepted estimation methods were lower than the aE values by factors of 2.0 ± 0.6, 3.1 ± 1.1, 4.5 ± 1.5, and 4.1 ± 1.6 for the four facilities, indicating underestimation in the reported VOC emissions. For 11 of the combined 93 VOC species reported by all four facilities, the reported emission rate and E were similar; but for the other 82 species, the reported emission rate was lower than E. The median ratio of E to that reported for all species by a facility ranged from 4.5 to 375 depending on the facility. Moreover, between 9 and 53 VOCs, for which there are existing reporting requirements to the NPRI, were not included in the facility emission reports. The comparisons between the emission reports and measurement-based emission rates indicate that improvements to VOC emission estimation methods would enhance the accuracy and completeness of emission estimates and their applicability to environmental impact assessments of oil sands developments.