Elsevier

Atmospheric Environment

Volume 238, 1 October 2020, 117723
Atmospheric Environment

Effect of industrial flare DREs derived by CFD and WERF on ozone pollution through CAMx simulation

https://doi.org/10.1016/j.atmosenv.2020.117723Get rights and content

Highlights

  • Coupling Dynamic Flaring DREs with CAMx Modeling for Ozone Simulation.

  • Flare DRE corrections established based on both CFD and WERF modeling.

  • Effect of Flare DRE on Regional Ozone Pollution from OGCPI.

Abstract

Flare is the last safety measure for daily operations in oil, gas & chemical process industries (OGCPI). However, an excessive flaring releases large quantity of emissions of VOCs and NOx, which may suddenly enhance local ozone as a secondary pollution. Normally, the flare destruction and removal efficiency (DRE) of 98% or 99% is regulated as the national standard and presumed for industrial practices in the U.S. Unfortunately, real DRE values could be much lower than the standard due to impact factors including various meteorological and operating conditions such as the cross-wind speed, flare jet velocity and heating value of combustion. Thus, it is critically important to explore the sensitivity of the regional ozone impact due to low DREs of OGCPI flare combustions. In this paper, a systematic methodology has been developed to examine ozone impacts due to the low flare DREs, which have never been systematically studied before. The DRE formulas were derived from computational fluid dynamic (CFD) modeling and Water Environment Research Foundation (WERF) results and then employed to recompile the point source emission inventory. After that, comprehensive air quality model with extensions (CAMx) was employede to simulate and quantify local ozone changes impacted by flare emissions of OGCPI. Case studies indicate that the maximum hourly ozone increments due to the low DRE through CFD and WERF modeling is 0.18 ppb and 1.3 ppb, respectively. This study could enrich fundamental understandings of industrial point source emissions and provide the quantitative and valuable support for the ozone pollution caused by OGCPI flare emissions under low DRE instead of standard values.

Introduction

Industrial flaring is to safely combust off-spec, unusable, or unwanted process streams, which might otherwise be harmful to the local environment if directly vented without destructions. The oil, gas and chemical process industries (OGCPI) in the U.S. daily processes millions of cubic feet of hydrocarbon gases (Baukal and Schwartz, 2001; Aalsalem et al., 2018). Thus, a slight decrease in flaring performance will release millions of cubic feet of gaseous emissions into the atmospheric environment. Note that although flaring is the last safety measure for protecting plant personal and equipment in OGCPI, excessive flaring should be avoided as they will generate large amounts of emissions such as NOx (nitrogen oxides), CO2, CO, VOCs (volatile organic compounds) especially for high reactive VOCs (i.e., HRVOCs such as ethylene, propylene, acetylene). For instance, an olefin plant with a capacity of 1.2 billion pounds of ethylene productivity per year can easily flare about 5.0 million pounds of ethylene during one single start-up operation (Xu and Li, 2008). Given the 98% flaring efficiency (TCEQ, 2015), the resultant air emissions include at least 15.4 million pounds of CO2, 40.0 Klbs CO, 7.4 Klbs NOx, 15.1 Klbs hydrocarbons, and 100.0 Klbs HRVOC (Xu et al., 2009). These emissions may cause seriously regional and transient air pollution events as well as negative societal impacts (Ge et al., 2016, 2017, 2018a, 2019). It should also be noted that under adverse meteorological and operation conditions (e.g., strong cross-wind, high jet velocity, or low combustion heating value), the flare destruction and removal efficiency (DRE) can be largely reduced, and thus the portion of unburned species will be significantly increased (Castiñeira and Edgar, 2008; Singh et al., 2012). Among the resultant consequences, one of particular concerns is that the increment of unburned VOCs from flares would elevate the local ground-ozone concentration as a secondary pollution, because ozone is generated by photochemical reactions between NOx and VOCs under solar radiation (Cleveland, 1974).

The ground-level ozone poses detrimental effects on human beings and many other living species. For instance, ozone can irritate respiratory system, which includes asthma aggravation, lung function reduction, and permanent lung damage (Kampa, 2008). Thus, ozone is one of six common pollutants regulated by the Federal Clean Air Act. The U.S. EPA (Environmental Protection Agency of the United States) has set the National Ambient Air Quality Standards (NAAQs) for the ground-level ozone since July 1997. From Oct 1st, 2015, a more stringent ozone standard on the 8-hr average of 70 ppb has been issued (EPA, 2016). Currently, OGCPI flaring practices needs to satisfy the 98% standard value for DRE (American Petroleum Institute, 2008). According to EPA regulations, a 98% DRE or higher should be obtained if the flare operations can be in accordance with 40 CFR Section 60.18 (McDaniel, 1983). Industrial flaring activities (61%) are among the top three HRVOCs emission sources in Texas, USA and thus has much potential to form ozone pollution (Singh et al., 2014). A rapid increase in ozone concentration has been commonly observed at air quality monitoring stations in Houston of Texas, USA. This phenomenon was regarded as a transient high ozone event, which might be induced by industrial flare emissions (Allen, 2017; Ge et al., 2018b).

It should be noted that industrial flaring DREs could be lower than the standard value due to impact factors such as the cross-wind speed, jet velocity, heating value of combustion zone (HVCZ), and flare design (Pohl, 1984, 1985). Recently, Ge et al. (2016) has studied the ozone impacts due to low DREs of multiple olefin plant start-ups via virtual case studies, where the 8-hr ozone increment under the assumed DREs of 95%, 96%, 97% and 98% have been investigated, respectively. Generally, plant start-up operations would generate larger amounts of flare emissions than those of normal operations; but they have a much less frequency. Thus, it is still interesting to explore the air quality impact from lower DREs under adverse meteorological during OGCPI normal operating conditions. And such relevant studies are still lacking.

In this paper, a systematic methodology has been developed to examine ozone impacts due to the excess VOCs and NOx released from regional OGCPI plants when their DRE values are lower than the presumed national standard caused by adverse meteorological and operating conditions. The formulas considering meteorological and operating conditions were derived from CFD and WERF modeling and employed to predict their effects on flare DREs and thus subsequent ozone formations. This study could enrich fundamental understandings of industrial point source emissions and provide the quantitative and valuable support for the ozone pollution caused by OGCPI flare emissions under low DREs instead of the standard value.

Section snippets

Problem statement

This paper derives the DRE formula for flare combustion associated with cross-wind speed, flare jet velocity, HVCZ, as well as flare design parameters, which are based on both CFD modeling (Jatale et al., 2012; Ge et al., 2018c; Chen and Alphones, 2019) and the studies of WERF (Willi et al., 2013). Next, the elevated point sources from OGCPI emission inventory files (i.e., flare emissions) are extracted and modified based on the derived DRE formula instead of the standard value regulated by

Methodology framework

The general methodology framework for this study has been summarized and showed in Fig. 1. Firstly, both CFD simulations and WERF modeling are employed to obtain DRE profiles, which will be the function of the cross-wind speed, flare jet velocity, HVCZ, flare design and others. Note that the DREs for different elevated point sources will not be the same due to the different cross-wind speed and jet velocity at each location of elevated point source. After that, flare emissions (i.e. VOCs and NOx

Scenario I: DRE results obtained through CFD modeling

CFD simulation shows that four variables affect the DRE of industrial flares: cross-wind speed, jet velocity of flare vent gas, HVCZ, and stoichiometric ratio. Note that the cross-wind speed and jet velocity are available in the air-quality model of the studied ozone episode. The cross-wind speed of each flare emissions can be obtained from the meteorological information, and the jet velocity can be obtained from the emission inventory. Thus, the actual DRE value for each elevated point source

Conclusions

By coupling dynamic flaring DREs of OGCPI plants with CAMx based air-quality modeling and simulation, the ground-level ozone impact associated with meteorological and industrial flare operating conditions have been quantitatively studied in this work. The CFD and WERF modeling based DRE correlations have been investigated respectively. Through case studies, it shows that although the individual DREs through CFD modeling could be significantly changed by the atmospheric cross-wind speed and

CRediT authorship contribution statement

Sijie Ge: Conceptualization, Methodology, Software, Data curation, Writing - original draft. Sujing Wang: Visualization, Investigation. Qiang Xu: Conceptualization, Methodology, Supervision, Writing - review & editing, Funding acquisition. Thomas Ho: Supervision, Writing - review & editing, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

This work was supported by National Natural Science Foundation of China (Grant No. U1903209), the Fundamental Research Funds for the Central Universities (Grant No. 2020QN11), China Postdoctoral Science Foundation (Grant No. 2019M650131), Texas Air Research Center (TARC), as well as Graduate Student Scholarship and Anita Riddle Faculty Fellowship from Lamar University in USA.

References (28)

  • CAMx

    User's Guide: Comprehensive Air Quality Model with Extensions Version 4.53

    (April, 2014)
  • D. Castiñeira et al.

    CFD for simulation of crosswind on the efficiency of high momentum jet turbulent combustion flames

    J. Environ. Eng.

    (2008)
  • D.H. Chen et al.

    Characterization of the incipient smoke point for steam-/air-assisted and nonassisted flares

    J. Air Waste Manag. Assoc.

    (2019)
  • W.S. Cleveland et al.

    Sunday and workday variations in photochemical air pollutants in New Jersey and New York

    Science

    (1974)
  • Cited by (0)

    View full text