Flaring is the process of combusting unwanted flammable gases in an open-atmosphere, turbulent, non-premixed flame. The practice of flaring is ubiquitous in the fossil fuel and petrochemical industries. Satellite data suggest that global flare gas volumes exceed 146 billion m3 annually, an amount equivalent to approximately 5% of global natural gas consumption. Within Canada, flared volumes have risen sharply in recent years, much of them associated with the rapid development of unconventional oil and gas resources through the widespread adoption of hydrofracturing (i.e., “fracking”) as a standard production method for tight-gas recovery, and the added upgrading associated with converting oil sands bitumen to fuels. In Alberta alone, overall reported flare volumes increased by 66% between 2009 and 2013 (AER, ST60B, 2014). Significant flaring occurs in all parts of Canada, whether it is associated with offshore production in Newfoundland and Nova Scotia, refineries and upgraders in New Brunswick, Quebec, Ontario, Saskatchewan, and Alberta, or upstream productions sites in Manitoba, Saskatchewan, Alberta, British Columbia, and the Northwest Territories.
The widespread prevalence of flaring both in Canada and globally raises important concerns for climate change, air quality, and human and animal health impacts, all of which contribute to fossil fuels being labelled a “dirty” energy source. The magnitude of global flaring makes it an important source of CO2 emissions, but additional climate forcing effects from methane and black carbon (BC) emissions are very poorly quantified. The total warming effect of BC in the atmosphere is now understood to be second only to CO2 (Nature Geosciences, 2008), and recent studies have directly implicated flaring as a critical source of BC emissions and the most important anthropogenic source in the Arctic.
On a local level, emitted particulate matter less than 2.5 microns in size, and more specifically BC, are linked to human mortality and significant health hazards. Thus, emissions from flares are both a global and a local concern. Quantitative details on the origins and properties of emitted particles are essential for quantifying climate and health impacts and designing appropriate mitigation strategies to enable cleaner fossil fuels.
Despite the magnitudes of gas flaring globally, data and models to accurately predict their emissions are critically lacking. Most pollutant inventory data for flares are based on crude, single-value emission factor estimates (i.e., g of a particular species per kg of gas flared), which lack sophistication to account for intrinsically wide variations in flare gas composition, flow rates, flare size, wind conditions, and flare design. Even more troubling, recent analysis has revealed that emission factors for flares used in North America, and similarly cited by other jurisdictions globally, are typically based on questionably relevant source information (e.g., measurements on enclosed systems burning landfill gas used as basis for emissions reporting from large, industrial flares in the energy industry). In one notable case, this re-examination uncovered a decades-old factor of ~1600 error in magnitude.
As fossil fuels have shifted toward unconventional sources and operating practices have changed, these issues have become even more acute. For example, potential adverse impacts of chlorinated species entering the flares stream during the flowback process at hydrofractured gas wells are only starting to be investigated. Given the dominance of hydrofracturing as a modern upstream oil and gas production method and the magnitudes of associated flaring (Tyner & Johnson, 2014), this level of uncertainty in emitted species and rates limits our ability to develop effective regulations with which to guide responsible development of unconventional resources. At the other end of the spectrum of flare applications, during refining and upgrading, emissions from air- and steam-assisted large industrial flares have also been a significant source of controversy. Lack of data and understanding of tradeoffs between gas and solid-phase emissions has led to unprecedented U.S. Department of Justice rulings and U.S. EPA enforcement actions against several operators of flares in the U.S. [35, 36, 39]. There is now a growing consensus within the U.S. EPA and the Texas Commission on Environmental Quality (TCEQ) that flares operated to ensure full removal of visible soot emissions may result in lower combustion efficiencies and increased emissions of unburned fuel. Both the U.S. EPA and the Arctic Council have specifically identified global gas flaring as a critical source of BC emissions for which uncertainties are high, measured data are lacking, and predictive models are non-existent (Arctic Council, 2011).
Critical knowledge gaps over flare emissions persist, primarily due to the complexity and scale of the research problem, which is exacerbated when the flare stream originates from unconventional extraction processes or bitumen upgrading. The uncontrolled nature of the turbulent combustion processes is simultaneously essential to the behavior of the system and confounding for the design of simple experiments. In addition, flares operate over an extremely wide range of scales and flow regimes, involve complex and potentially two-phase fuel chemistries, and produce a variety of important emissions that can be very difficult to measure in an open system, may include important engineering complications related to use of steam and air injection in industrial settings, and ultimately present a diversity of difficult research challenges that require a range of complementary skill sets to solve.