Nitrogen Oxide Emissions from Premixed Reacting Jets in a Vitiated Crossflow

ABSTRACT

This paper describes nitrogen oxide (NOx) measurements from a reacting jet in crossflow (RJICF). The NOx production of the RJICF is controlled by jet stoichiometry, crossflow temperature and composition, and the mixing rates between the fluid streams. Mixing occurs both pre- and post-flame. Pre-flame mixing refers to mixing between the reactant jet and the crossflow prior to combustion and determines the stoichiometry of burning; it is controlled by degree of flame lifting, LO, and the shear layer vortices. Post-flame mixing refers to mixing of these secondary combustion products with the crossflow; it is controlled by the counter-rotating vortex pair. The literature has clearly shown a monotonic increase in RJICF NOx production with its bulk average temperature rise (ΔT) but also indicated significant dependencies on momentum flux ratio (J), jet stoichiometry (ϕ_jet), and other parameters. Moreover, these parameters are always coupled for a given geometry (e.g., LO varies with ϕ_jet), and the fundamental influence parameters require clarification. To address this, significant effort was spent in this work on differentiating these coupled effects. NOx measurements were obtained from premixed ethane/methane/air jets injected into a vitiated crossflow of lean combustion products, for ΔT values between 75 and 350 K. Data were obtained at two crossflow temperatures (1350 K and 1410 K), two jet geometries, J values from 6 to 40, and ϕ_jet from 0.8 to 8.0. Ethane/methane ratio was varied to influence flame lifting independent of other parameters. Jet geometry was varied to influence shear layer vortex growth rates and, hence, pre-flame mixing rates. Overall, these data are consistent with the idea that NOx emissions are largely controlled by the stoichiometry at which combustion actually occurs, referred to as ϕ_Flame. ϕ_Flame is influenced by ϕ_jet and pre-flame mixing of the jet and crossflow that, in turn, depend upon LO, nozzle geometry, and crossflow temperature. While this result is expected, it manifests itself in complex manners. For example, NOx levels were observed to be nearly independent of ϕ_jet for a range of conditions, due to the coupled dependence of ϕ_jet and LO. Similarly, NOx emissions are a factor of three lower in the nozzle jet geometry relative to a fully developed exit flow, due to enhanced pre-flame mixing. From a practical point of view, the key implication of these results is its suggestion for minimizing NOx production at a given ΔT – designing injection schemes that enhance flame lifting and shear layer vortex growth rates. Finally, there are indications in the data of additional post-flame mixing effects that require further work to clarify.

KEYWORDS:

• NOx emissions
• axial staging
• reacting jet in crossflow
• flame lifting
• hydrodynamic stability

Nomenclature

Table

Acknowledgments

This research was partially supported by the University Turbine Systems Research (contract #DE-FE0025344), contract monitor Dr. Mark Freeman.

In addition, the authors would like to acknowledge Timothy Hardis for his design work on the test section and Divya Sunkara for her assistance with data collection.

Notes

¹ The reduced ϕ_jet ≡ 2ϕ_jet / (1 + ϕ_jet) is used as it is symmetric about unity for lean and rich conditions and is helpful for plotting data with such a large range for ϕ_jet.

² The emissions measurements for these cases were within the error of the associated base NOx measurement. They are plotted as having a normalized ΔNOx value of 0 with the positive portion of their error bars.

³ Note also that the high J cases show that there is a J impact on LO. These observations are consistent with (Kolb et al. 2016) who noted lifting in flames with J values on the order of 60 and above, and a J sensitivity of LO for these higher J values.

Additional information

Funding

This work was supported by the U.S. Department of Energy [DE-FE0025344].