Matthew R. Johnson (Ph.D., P.Eng.), Canada Research Chair in Energy & Combustion, Generated Pollutant Emissions, Associate Professor, Energy & Emissions Research Lab. Mechanical & Aerospace Eng., Carleton University
Robin W. Devillers (Ph.D.), Post Doctoral Fellow, National Research Council of Canada, Measurement Science & Standards
November 19, 2012
# 09-9185-50
Executive Summary
Black carbon is a potent short-lived climate forcer whose net warming effect in the atmosphere is likely second only to that of CO2. Global gas flaring is implicated as a potentially critical source of black carbon emissions, in large part due to the very significant volumes of gas flared annually (which exceed 140 billion m3 according to satellite estimates). However, efforts to date to accurately assess, manage, and mitigate flare generated black carbon have been severely hampered by a dearth of quantitative emissions data for flares, an absence of reliable emission factor models, and a lack of quantitative, in-situ measurement techniques. As discussed in this report, this has the potential to change with the emergence of the sky-LOSA technique for directly quantifying black carbon emission rates in flares under field conditions. The method is based on Line-Of-Sight Attenuation (LOSA) of sky-light which, in combination with image correlation velocimetry and Rayleigh-Debye-Gans theory for Fractal Aggregates (RDG-FA), enables accurate quantification of black carbon mass emission rates in atmospheric plumes of flares.
The primary objectives of the work described in this report were to conduct field measurements of black carbon emission rates at available field sites in Mexico both as a means to collecting some of the first-ever in-situ measurements of black carbon emission rates from flares and also as an opportunity to refine the sky-LOSA technology and test improved hardware and algorithms for the first time. Measurements were ultimately performed at two flare sites in Mexico on December 1st and 2nd, 2011. On December 1st, 2011, sky-LOSA frames were acquired in Punta de Piedra for one of two adjacent 10 m high flare stacks. 14,441 frames were acquired, equivalent to 184 sec of continuous acquisitions split into 7 frame series recorded over a time span of 30 minutes. Images were difficult to analyse because of the presence of trees in the background and because of unstable flare operation. However, soot emission rates were successfully calculated, and the measured flare was found to emit black carbon at an average rate of 0.029 g/s, with instantaneous emission rates (95%-probability interval) ranging from 0.000 g/s – 0.181 g/s, and with soot emission peaks reaching 0.55 g/s. The most striking observation was the wide variability in emission rates of this flare during normal operation in relatively consistent wind conditions.
On December 2nd, 2011, 36,260 sky-LOSA frames were acquired for a 20 m-high flare at a turbocompressor in Poza Rica. An equivalent to 717 sec of continuous measurements were acquired in 13 frame series collected over a 1 h 30 min time span. Although the flare was producing soot emissions that were only barely visible to the human eye, they were readily quantified with the sky-LOSA system. In contrast to the flare measured on December 1, the emission rates were quite stable with an average value of 0.053 g/s and a 95%-probability interval of 0.014 g/s – 0.105 g/s. A detailed uncertainty analysis was performed, and the mean emission rate of 0.053 g/s was quantified with an overall uncertainty of -15.9% to +20.1% that was mostly attributable to uncertainties in soot optical properties.
Because data were not available on the compositions or flow rates of gas being directed to the flares, it was not possible to calculate fuel-mass- or fuel-energy-specific black carbon emission rates. However, a preliminary analysis was performed to assess the context of the field measurement results and make comparisons with ongoing lab-based work aimed at flaregenerated black carbon emissions factor development. Results of this comparison suggest that for an assumed flare gas exit velocity in the range of 4 m/s, the 0.053 g/s soot emission rate measured by sky-LOSA on December 2nd was consistent with lab-scale flare emissions measurements for light fuel mixes representative of methane-rich natural gas. This is a very encouraging result for the prospect of future efforts to combine results of in-situ field measurements using sky-LOSA with controlled lab-scale experiments to significantly improve current emission factor approaches for estimating flare generated black carbon emissions. Moreover, the success in applying sky-LOSA to a very lightly sooting flare, and the high degree of accuracy achieved during measurements, demonstrate the potential of this new technology in supporting future mitigation efforts where quantitative measurements of black carbon emissions reductions are required.