AAAR 36th Annual Conference October 16 - October 20, 2017 Raleigh Convention Center Raleigh, North Carolina, USA
Abstract View
A General Uncertainty Analysis for Measurements of Black Carbon Emissions from Gas Flaring Using Sky-LOSA
BRADLEY CONRAD, Matthew Johnson, Carleton University
Abstract Number: 279 Working Group: Instrumentation and Methods
Abstract Gas flaring is a ubiquitous practice in the oil and gas industry where unwanted gases are destroyed in an open flame, frequently from a vertically-raised stack. With estimates of global flaring on the order of 5 trillion cubic feet per year, anthropogenic emissions of climate-relevant species during flaring activities are of notable concern. Flare emissions include black carbon (BC), a PM2.5-species that has been recognized as an important short-lived climate forcer. Despite the climatic importance of BC and the scale of gas flaring, only recently have direct measurements of in-field BC emission factors for gas flaring been obtained. These data were acquired using sky-LOSA, an imaging technique enabling the quantification of BC emission rates from flares, made possible through the consideration of visible light attenuation through the BC-laden atmospheric plume.
To extend the knowledge-base of flare-generated BC emissions, dissemination of a general sky-LOSA procedure to make the implementation of measurements available to third parties is necessary; to this end, a general uncertainty analysis (GUA) under a Monte Carlo framework is vital. The primary goal of the GUA is to quantify the effect of various input variables on the uncertainty of computed BC emission rates. Relevant input variables include flame-generated BC aggregate properties, ground-level irradiance by the sun and sky, and camera position with respect to the sun and horizon. Through the latter, a sky-LOSA user has indirect control over measurement uncertainty. Consequently, the GUA can be used to define a generalized procedure for sky-LOSA data acquisition enabling a user to meet desired uncertainty thresholds as a function of ambient conditions. Key results and implications from the GUA will be presented, including practical limitations, achievable uncertainties, and proposed future work to reduce measurement uncertainty.