Inconsistencies in sulfur dioxide emissions from the Canadian oil sands and potential implications

Chris A Mclinden,Nicolay Krotkov,Nicolas Theys,Paul A Makar,Vitali Fioletov,Pascal Hedelt,Diego G Loyola,Can Li,Cassandra Brown,Debora Griffin,Nolan Dickson,Andrew Kovachik,Cristen L F Adams,Xiaoyi Zhao

doi:10.1088/1748-9326/abcbbb

Abstract

Satellite-derived and reported sulfur dioxide (SO2) emissions from the Canadian oil sands are shown to have been consistent up to 2013. Post-2013, these sources of emissions data diverged, with reported emissions dropping by a factor of two, while satellite-derived emissions for the region remained relatively constant, with the discrepancy (satellite-derived emissions minus reported emissions) peaking at 50 kt(SO2) yr−1 around 2016. The 2013–2014 period corresponds to when new flue-gas desulfurization units came on-line. Previous work has established a high level of consistency between at-stack SO2 emissions observations and satellite estimates, and surface monitoring network SO2 concentrations over the same multi-year period show similar trends as the satellite data, with a slight increase in concentrations post-2013. No clear explanation for this discrepancy currently exists. The implications of the discrepancy towards estimated total sulfur deposition to downwind ecosystems were estimated relative to 2013 emissions levels, with the satellite-derived values leaving the area of regional critical load exceedances of aquatic ecosystems largely unchanged from 2013 values, 335 000 km2, and reported values potentially decreasing this area to 185 000 km2.

Highlights

Space-based sensors have been widely used to derive or constrain emissions of air pollutants since the mid1990s (e.g. Streets et al 2013)
Oil sands emissions data Reported SO2 emissions in the oil sands include hourly emissions measured by continuous emissions monitoring systems (CEMS) on the main stacks (AG 1998), which emit most of the SO2 at upgrading facilities, as well as engineering estimates for emissions from other sources SO2, such as flaring
In the second half of 2013, additional scrubbers came online at SML with the impact of these scrubbers being the inferred cause for the reported factor of three decrease beginning in 2014, and a factor of two decrease in overall emissions from the upgraders. This decrease in reported emissions is not matched by a decrease in ambient SO2 levels

Summary

Introduction

Space-based sensors have been widely used to derive or constrain emissions of air pollutants since the mid1990s (e.g. Streets et al 2013). Direct approaches have been developed, in which the satellite observations are paired with winds from a meteorological reanalysis (Beirle et al 2011, Fioletov et al 2015). This newer approach is best suited for short(er)-lived species, but it has proved useful for intermediate (CO; Pommier et al 2013) and longlived (CO2; Nassar et al 2017) pollutants. The ability of OMI to capture annually varying emissions has been demonstrated for many sources globally, including power plants (Fioletov et al 2016b), smelters (Ialongo et al 2018), and volcanos (McLinden et al 2016b)

Methods

Results

Conclusion