Title | Analysis of Combustion Efficiencies for Industrial Steam-Assisted Flares |

Creator | Marr, Kevin C. |

Contributor | Ponchaut, Nicolas F., and Kytomaa, Harri K. |

Date | 2013-09-25 |

Spatial Coverage | Kauai, Hawaii |

Subject | AFRC 2013 Industrial Combustion Symposium |

Description | Paper from the AFRC 2013 conference titled Analysis of Combustion Efficiencies for Industrial Steam-Assisted Flares by Kevin Marr |

Abstract | Flares are essential in safely managing excess combustible gases that can result from normal operations as well as process upsets and emergency conditions in refineries, chemical plants, and other industrial facilities. In many modern flare designs, steam or air is added to the vent gas to suppress visible emissions per EPA regulations. Several studies have shown that operating flares with too much steam addition can reduce hydrocarbon combustion efficiencies. However, the current EPA regulations do not address the influence of steaming on flare performance. Recently, EPA directed test campaigns to measure combustion efficiencies for industrial-scale flares at several facilities throughout the country. This work analyzes the results from the EPA-directed flare test campaigns. The analysis investigates the effect of vent gas composition, in particular hydrogen concentration, and steam flow on combustion efficiency. Existing methods for estimating combustion efficiency are reviewed and a prediction methodology based on a Monte Carlo method that makes direct use of the test campaign data in the form of a look-up table is proposed. |

Type | Event |

Format | application/pdf |

Rights | No copyright issues |

OCR Text | Show ANALYSIS OF COMBUSTION EFFICIENCIES FOR INDUSTRIAL STEAM-ASSISTED FLARES Kevin C. Marr, Ph.D. kmarr@exponent.com Nicolas F. Ponchaut, Ph.D., P.E. nponchaut@exponent.com Harri K. Kytömaa, Ph.D., P.E. hkytomaa@exponent.com Exponent, Inc., 9 Strathmore Road, Natick, MA 01760, USA ABSTRACT Flares are essential in safely managing excess combustible gases that can result from normal operations as well as process upsets and emergency conditions in refineries, chemical plants, and other industrial facilities. In many modern flare designs, steam or air is added to the vent gas to suppress visible emissions per EPA regulations. Several studies have shown that operating flares with too much steam addition can reduce hydrocarbon combustion efficiencies. However, the current EPA regulations do not address the influence of steaming on flare performance. Recently, EPA directed test campaigns to measure combustion efficiencies for industrial-scale flares at several facilities throughout the country. This work analyzes the results from the EPA-directed flare test campaigns. The analysis investigates the effect of vent gas composition, in particular hydrogen concentration, and steam flow on combustion efficiency. Existing methods for estimating combustion efficiency are reviewed and a prediction methodology based on a Monte Carlo method that makes direct use of the test campaign data in the form of a look-up table is proposed. INTRODUCTION Flares are common safety devices that are used to manage waste gasses that are produced from normal or emergency operations in refineries and other industrial facilities. EPA regulations, 40 CFR 60.18, require that flares operate without visible emissions, i.e. smoke. One common method to suppress smoke in modern flare designs is to add steam to the vent gas upstream of the reaction zone. In the 1980s, several benchmark studies were performed that formed the basis for current EPA regulations [1] [2], [3], [4]. These studies identified several parameters that affect combustion efficiency. They found that when a flare operated near its flame stability limits, the combustion efficiency degraded significantly. They explained that flames were susceptible to perturbations near their stability limits, and that such perturbations result in increased emissions of unburned fuel. Two main conclusions were drawn, which formed the foundation of the current EPA regulations on the operation envelope for flares. The first was that flame stability depends on the heat content of the flare and the exit velocity at blow-out, and the second was that the combustion efficiency was a function of the heat content of the flare. The most significant contribution of these studies was defining an operation envelope where a flare can be assumed to have a 98% destruction efficiency. More recent studies have shown that several additional parameters can have an influence on combustion efficiency, including hydrogen content, excess steam addition and crosswind [5], [6]. These have also shown that even if a flare is operating within the envelope specified by EPA regulations, excessive steam addition can result in destruction efficiencies that fall below the assumed 98% threshold [6], [7]. In light of these newer findings, EPA recently released a report analyzing data from several industrial scale flare tests [8]. Both References [6] and [8] propose thresholds based on modified parameters-namely net heating value and lower flammable limit-that are derived from the gas species in the combustion zone, which includes vent gases and steam. Although EPA threshold parameters provide simple guidelines in part to maintain the flare destruction efficiency above 98%, they do not provide, nor were they intended to provide, quantitative predictions of flare emissions. The scope of this work is to provide a framework to predict emissions from an arbitrary flaring event. Owing to the complexities of flares of differing geometries that receive multiple combustible and inert species, the approach of this work is not to predict an exact amount of emissions, but rather to determine a range of predicted emissions defined by prediction intervals. The analysis in the present work focuses on the net heating value approach rather than the lower flammable limit approach. The current work also investigates the effect of hydrogen content on combustion efficiency, evaluates a previously published hydrogen correction methodology [6], [9], and proposes an alternative hydrogen correction methodology. EPA 2012 REPORT ON PROPERLY DESIGNED AND OPERATED FLARES In the past few years, EPA enforcement of the Clean Air Act has led to the testing of flares at several industrial facilities. In 2012, the EPA issued a report, Parameters for Properly Designed and Operated Flares, that analyzes results from nine elevated flare tests [8]. Table 1 summarizes the different studies analyzed by the EPA report. The measurement techniques used by these studies include extractive measurements from probes (Extractive), Passive Fourier Transform Infrared spectroscopy (PFTIR), and Active Fourier Transform Infrared spectroscopy (AFTIR). The two spectroscopic methods are non-intrusive remote sensing methods that do not require a probe to be positioned downstream of the flare. The hydrogen content of the vent gas varies from none for the two EPA-600 studies and the TCEQ study, and up to 62% for the SDP EPF study. Table 1: Test reports included in EPA analysis Study ID Authors Date % H2 in Vent Gas Test Method EPA-600/2-83-052 McDaniel [1] July 1983 0 Extractive EPA-600/2-85-106 Pohl and Soelberg [3] Sept 1985 0 Extractive MPC TX Clean Air Engineering [10] May 2010 3.1-24 PFTIR INEOS Clean Air Engineering [11] July 2010 0 PFTIR MPC Detroit Clean Air Engineering [12] Nov 2010 7.0-55 PFTIR FHR (AU) Clean Air Engineering [9] June 2011 13-47 PFTIR FHR (LOU) Clean Air Engineering [9] June 2011 20-30 PFTIR SDP EPF Shell Global Solutions [13] Apr 2011 37-62 PFTIR TCEQ Allen and Torres [7] Aug 2011 0 Extractive, AFTIR, PFTIRNET HEATING VALUE OF THE COMBUSTION ZONE GASES The net heating value (NHV) of the flare directly influences the stability, and hence the combustion efficiency of a flare. The net heating value of the fuel is dependent on the specific composition of the vent gases. For a steam-assisted flare, where steam is also present in the combustion zone, a net heating value of the combustion zone gases (NHVCZ) in BTU/scf can be defined as follows [7]: =̇(386.3⁄)+̇(386.3⁄)̇(386.3⁄)+̇(386.3⁄)+̇(386.3⁄) [Eqn 1] where the net heating value, NHV, mass flow rate, ̇, and molecular weight, MW, for the vent gas, pilot gas, and steam are indicated by the subscripts VG, PG and S, respectively. The constant 386.3 is the ideal gas volume in standard cubic feet per lb-mole at 68ºF and 1 atm. Note Equation 1 is specified for English units, where the net heat value is in BTU/scf, the mass flow rate is in lb/hr and the molecular weight is in lb/lb-mole. Figure 1: Measured combustion efficiency for industrial-scale flaresSeveral studies have suggested that the NHVCZ of the flare can be used to estimate the combustion efficiency of the flare [6], [7]. A high combustion efficiency (>98%) can be expected as long as the NHVCZ is larger than a threshold value of 300 BTU/scf. Figure 1 shows a plot of the measured combustion efficiency for the flares considered in the testing studies detailed in Table 1, and clearly shows a sharp decrease in combustion efficiency when NHVCZ < 300 BTU/scf. EFFECT OF HYDROGEN ADDITION Figure 2 shows the same data as Figure 1, but color coded by hydrogen mole fraction in the flare combustion zone, RH2.1 For vent gases consisting of only hydrocarbons, using NHVCZ as a scaling parameter results in a reasonable collapse of the test data. However, the scatter in data increases as the hydrogen content of the vent gases increases, and NHVCZ as defined by Equation 1 does not fully collapse the data. For vent gases with high hydrogen content, the NHVCZ can be less than 300 and still have high combustion efficiency. As discussed previously, flare inefficiency is the result of operating in regimes near stability limits of the flame. Because the reaction rate of hydrogen is significantly higher than hydrocarbons, hydrogen addition to a hydrocarbon based fuel expands the stable operating range of a flare. The increased reaction rate effectively increases the flame speed of the fuel mixture, which results in higher blow-out and blow-off velocities. This causes the combustion efficiency to increase in a manner that is not captured by the heating value of hydrogen, which is low compared to higher molecular weight combustible species. 1 RH2 is the ratio between the molar concentration of hydrogen and the molar concentration of combustible species in the flare combustion zone Figure 2: Effect of hydrogen addition on combustion efficiency for industrial-scale flares. Hydrogen Adjusted Net Heating Value of the Combustion Zone Gases Previous studies by Clean Air Engineering propose a simple methodology to correct for the hydrogen effects referred to as the hydrogen adjusted NHVCZ correction [6], [9] . The method recognizes that the net heating value for hydrogen is significantly less than that of hydrocarbons, which effectively decreases the NHVCZ despite enhancing the flame stability of the flare. To correct for this drop in NHVCZ, the correction method uses an artificially elevated "hydrocarbon equivalent net heat value for hydrogen of 1212 BTU/scf in NHVCZ calculations. The hydrogen adjustment factor of 4.49 was derived from the ratio of the average net heating value at the lower flammable limit of hydrocarbons (48.5 BTU/scf) and that of hydrogen (10.8 BTU/scf). Multiplying the actual net heating value of hydrogen (270 BTU/scf) by the hydrogen adjustment factor (4.49) gives the "hydrocarbon equivalent net heat value of hydrogen". Figure 3 shows the effect of the hydrogen adjusted NHVCZ correction that increases the value of NHVCZ based upon the hydrogen concentration in the vent gas. The data points for hydrogen-bearing vent gases move to the right resulting in a better collapse of the data. Examples of this horizontal shift are illustrated by the purple arrows.Figure 3: Flare test data plotted with respect to the hydrogen adjusted NHVCZ correction. Examples of the resultant shift due to the correction are indicated by purple arrows. The trend lines in Figure 2 show "fan-like" characteristics, whereas RH2 increases, the curve does not only shift to the right along the horizontal axes, but also shifts along the vertical axis. This "fan-like" character implies that a more appropriate collapse of the data can be obtained by shifting, or scaling, both NHVCZ and the combustion efficiency. Generalized Scaling The hydrogen adjusted NHVCZ correction is a simple correction. However, a more general form for hydrogen correction can be defined as follows: =100−1 [Eqn 2] =+2 [Eqn 3] where X and Y are the scaled variables for NHVCZ and the combustion inefficiency, 100-CE, respectively. The X and Y scaled variables can be interpreted as the effective NHVCZ and the effective combustion inefficiency, respectively. In general, the scaling functions f1 and f2 can be functions of any relevant parameters. For the purposes of this work, the scaling functions are assumed to be functions of the hydrogen molar ratio in the flare combustion zone, RH2. This generalized form is based on the observation that increasing hydrogen results in a trend that shifts the curves along both the horizontal and vertical axes as seen in Figure 2. In the present work, the functions f1 and f2 are assumed to be of the form 1=(1−2)(1− 2) [Eqn 4] 2=2 [Eqn 5] where D, E, and F are constant coefficients. Both f1 and f2 are chosen such that when the vent gas does not contain hydrogen, the scaled variables, X and Y, return to their dimensional forms, NHVCZ and 100-CE. In addition, f1 was chosen to have a quadratic form such that a flare burning 100% hydrogen would have a combustion efficiency of 100%, whereas the power law form of f2 is based on the law of mass action as applied to equilibrium reactions [14]. For comparison, the generalized scaling simplifies to the hydrogen adjusted NHVCZ correction when E=942Rcomb [BTU/scf]) and F=1 in f2, and f1 is 1. Rcomb is the molar fraction of combustible species in the flare combustion zone2. The general scaling is obtained by optimizing the coefficients D-F. Table 2 summarizes the expressions for f1 and f2 for the hydrogen adjusted NHVCZ correction and the optimized generalized scaling proposed in this study. Table 2: Correction functions f1 and f2 for the hydrogen adjusted NHVCZ and the generalized scaling corrections Correction f1 f2 No adjustment 1 0 Hydrogen Adjusted NHVCZ 1 942RcombRH2 Generalized Scaling (1-RH2)(1+4.2RH2) 169.520.7 Figure 4 compares the final results of the optimized generalized hydrogen scaling to the hydrogen adjusted NHVCZ correction. In Figure 4, the effective combustion efficiency, 100-Y, is plotted with respect to the effective NHVCZ, X . For both hydrogen correction methods, the correction does not affect data points that do not contain hydrogen-the RH2=0 data points do not shift. The generalized scaling provides a slightly 2 Rcomb is the molar fraction of combustible species in the flare combustion zone, and 952 BTU/scf is the difference between the adjusted net heating value of hydrogen (1212 BTU/scf) and the unadjusted net heating value of hydrogen (270 BTU/scf). tighter collapse than the hydrogen adjusted NHVCZ correction. It is not the intent of the authors to propose a scaling that is optimal for all flares, but rather to provide a framework for future scaling studies. To determine a more complete and robust general scaling, more experimental test data is required. Figure 4: Comparison of the hydrocarbon equivalency and generalized scaling corrections for hydrogen DETERMINISTIC PREDICTION METHOD A simple approach to predict flare emissions is to find a best fit curve for the test data. Assuming an exponential function of the form Ae(-Bx)+C, best fit curves can be determined for both the hydrogen adjusted NHVCZ correction and the generalized scaling. Figures 5 and 6 show the best fit curves for the hydrogen adjusted NHVCZ and generalized scaling corrections. The simplicity of a simple exponential curve fit is appealing as a practical tool. However, other curve fit methodologies can provide a better statistical fit. One such method is to apply a logarithmic transform to the y-axis, which in this case is the combustion efficiency. Another method is to apply a weight-averaged curve fit, where data points are weighed according to some assumed distribution function. The emission of a particular species can be directly calculated from the species-specific destruction efficiency. Although it is common to assume that a 98% combustion efficiency achieves a 99.5% destruction efficiency [15], it is not a reasonable to assume that the 1.5% difference between destruction and combustion efficiency holds for combustion efficiencies less than 98%. Therefore, in the present work, species emissions are calculated by assuming that the species specific destruction efficiency and the combustion efficiency are equivalent. Figure 5: Best fit curve for the hydrogen adjusted NHVCZ correctionFigure 6: Best fit curve for the generalized scaling correction for hydrogen Regardless of the details of the curve fit, deterministic prediction methods have limitations. Because any curve fit assumes a particular functionality, there is an inherent bias based on the selection of the function. In addition, the way the fit is obtained (i.e., using weights or logarithm transforms, etc.) also influences the results. Deterministic approaches based on empirical data are only able to provide an emissions estimate of an "average" flare. To provide insight for cases where the flare may not behave like an "average" flare, determination of upper and lower bounds of the emissions estimate is necessary. Uncertainty analysis for deterministic methods is limited to bounds on the fit coefficients and provides no information on the predictive uncertainty of the aggregate emissions from a flaring event. MONTE CARLO PREDICTION METHOD The scatter in the test data is the result of measurement uncertainty, which for FTIR methods can be significant for low combustion efficiency measurements, and variations from other parameters including flare tip geometry, vent gas composition and crosswind turbulence. Given that the data has unknown measurement errors, using a deterministic approach such as using curve fitted data can be misleading. The Monte Carlo prediction method is perfectly suited for such predictions that rely on pre-existing data. In the present work, the basis of the Monte Carlo prediction method is the measured combustion efficiency EPA data set. The Monte Carlo method can be used for any existing data set regardless of whether a hydrogen correction is applied. The estimated emissions are determined from the measured combustion efficiencies, which are assumed to correlate directly to the effective NHVCZ-for non-hydrogen corrected data, the effective NHVCZ is the actual NHVCZ. The Monte Carlo algorithm is divided in two portions. • First, for a flaring event, the effective NHVCZ values are calculated at a particular time interval (hourly, daily, etc.). For each time period, all data points with effective NHVCZ that are within ± 12.5 BTU/scf (i.e. within a bin size of 25 BTU/scf) are binned into a subset. Instead of assuming that the correct combustion efficiency is the one corresponding to the average of the data points in the subset, the method assumes that all the combustion efficiencies within that bin are possible and equally probable. Figure 7 illustrates a bin that is centered at 350 BTU/scf. • Once the list of possible combustion efficiency values is determined, the Monte Carlo method selects randomly one data point in each bin. In the example shown in Figure 7, each data point in the bin window has an equal probability of being selected if the effective NHVCZ is 350 BTU/scf. Based on this random sampling, the flare emissions are calculated. A single trial would not be meaningful since it would only be one of the possible outcomes, and therefore the process is repeated many times (100,000 times in the present work), which allows the method to determine an expected average outcome, as well as probability ranges around the average.Figure 7: Example of a 25 BTU/scf bin centered at an effective NHVCZ of 350 BTU/scf. Inset plot shows a zoomed in view of the bin. Other than the choice of bin size (and the requirement in our algorithm that a minimum of three data points be in each bins), the method does not make any additional assumptions such as the form of a fit function, and therefore eliminates the bias that a deterministic curve fitting method introduces. In addition, the data scatter associated with the test data is implicitly included in the Monte Carlo scheme, and calculating the emissions estimate. As a result of the random selection of the data points, the Monte Carlo method also directly determines the statistical distribution for the emissions prediction from the multiple random trials (100,000) for each time period of interest. VALIDATION OF THE MONTE CARLO METHOD To validate the Monte Carlo method, one test study is removed from the set of test studies and is used as validation data. A flaring case study is defined by the removed test study where each test condition is assumed to be one hour of a fictitious flaring event. The Monte Carlo method is then applied based on the remaining data to predict the C5+ emissions for the flaring event. Validation case studies were performed for the FHR AU, FHR LOU, and SDP EPF test sets. The two EPA-600 test data and the TCEQ test data were not used as validation studies because they did not contain C5+. The MPC TX and MPC Detroit test data were not used as validation studies because they contain significant amount of nitrogen. The generalized scaling proposed in this paper does not address the effect of nitrogen. However, an analysis similar to the method used to obtain the generalized scaling for hydrogen could be applied to nitrogen and other inert species. For the validation studies, the Monte Carlo method was applied to both hydrogen correction methods-the hydrocarbon equivalency and the generalized scaling. Figures 8-10 compare the prediction results for the Monte Carlo and deterministic method using the hydrocarbon equivalency correction for hydrogen. The varying shades of blue show the probability distribution of the predicted emissions. For example, in Figure 8, the Monte Carlo method predicts that there is a 99% chance that the actual cumulative emissions are below 9.7 tons. Similarly, 1%, 5%, 10%, 90%, and 95% probabilities are also shown. An alternative interpretation is that the dark blue area between the 10% and 90% contours define the 80% prediction interval where there is an 80% probability that the actual emissions fall within the interval. Similarly, the areas between the 5% and 95% probabilities and the 1% and 99% probabilities define the 90% and 98% prediction intervals, respectively. The deterministic method considered is a simple best fit exponential function. Figure 8: Comparison of the Monte Carlo and the deterministic predictions of C5+ emissions for FHR AU test data using the hydrogen adjusted NHVCZ correction. The prediction intervals are shown for the Monte Carlo analysis. Figure 9: Comparison of the Monte Carlo and the deterministic predictions of C5+ emissions for FHR LOU test data using the hydrogen adjusted NHVCZ correction for hydrogen. The prediction intervals are shown for the Monte Carlo analysis.Figure 10: Comparison of the Monte Carlo and the deterministic predictions of C5+ emissions for SDP EPF test data using the hydrogen adjusted NHVCZ correction for hydrogen. The prediction intervals are shown for the Monte Carlo analysis. Neither the Monte Carlo nor deterministic method is able to predict the exact emissions consistently. Although for these validation cases, the cumulative emission predicted by the Monte Carlo method is similar or slightly closer to the actual cumulative emission compared to the emission predicted by the deterministic method, one of the benefits of the Monte Carlo method is that it also provides quantitative prediction intervals. For two of the three validation case studies, the actual emission falls within the 90% prediction interval. For the SDP EPF case study, the actual emission is below the 1% probability limit. In the SDP EPF study, the vent gas had high hydrogen concentrations (up to 62% by volume of the vent gas). The over prediction in the SDP EPF case is due to the hydrogen adjusted NHVCZ correction not being able to sufficiently collapse these high hydrogen points. Figures 11-13 compare the prediction results for the Monte Carlo and deterministic methods using generalized scaling to correct for hydrogen. Similar to Figures 8-10, prediction intervals are indicated by the different shades of blue, and the deterministic method is based on a simple exponential curve fit using generalized coordinates. Figure 11: Comparison of the Monte Carlo and the deterministic predictions of C5+ emissions for FHR AU test data using generalized scaling correction for hydrogen. The prediction intervals are shown for the Monte Carlo analysis. Figure 12: Comparison of the Monte Carlo and the deterministic predictions of C5+ emissions for FHR LOU test data using generalize scaling correction for hydrogen. The prediction intervals are shown for the Monte Carlo analysis.Figure 13: Comparison of the Monte Carlo and the deterministic predictions of C5+ emissions for SDP EPF test data using generalize scaling correction for hydrogen. The prediction intervals are shown for the Monte Carlo analysis. The predicted and measured C5+ cumulative emission is tabulated in Table 3. Table 4 shows the percent error deviation between the measured cumulative emissions and predicted cumulative emissions determined by deterministic and Monte Carlo methods for unadjusted NHVCZ, the hydrogen adjusted NHVCZ corrected, and generalized scaling data. The large errors associated with the unadjusted NHVCZ estimates substantiate the use of a hydrogen correction. Irrespective of the hydrogen correction methodology, the Monte Carlo approach gives more accurate estimates compared to the deterministic approach. For both the FHR AU and FHR LOU validation cases, the generalized scaling and the hydrogen adjusted NHVCZ correction give similarly accurate estimates. However, for the SDP EPF validation case, the generalized scaling performs significantly better than the hydrogen adjusted NHVCZ correction. Compared to the FHR AU and FHR LOU cases, the SDP EPF validation case includes data points with significant hydrogen content. Recall the "fan-like" behavior of the hydrogen trend lines in Figure 2. Because the generalized scaling accounts for this "fan-like" behavior, the generalized scaling reduces the scatter of the higher hydrogen content data points more than the hydrogen adjusted NHVCZ correction, and therefore, results in more accurate emissions predictions for high hydrogen content flares. Table 3: Predicted C5+ cumulative emissions for validation case studies H2 Correction Method Estimation Method C5+ Cumulative Emission (tons) FHR AU FHR LOU SDP EPF Unadjusted NHVCZ Deterministic 12.5 13.4 225 Monte Carlo 12.3 14.9 232 Measured 8.8 7.5 48 H2 Adjusted NHVCZ Deterministic 6.3 8.2 75 Monte Carlo 6.5 8.1 72 Measured 8.8 7.5 48 Generalized Scaling Deterministic 6.0 8.6 52 Monte Carlo 6.8 8.2 49 Measured 8.8 7.5 48 Table 4: Percent error of predicted C5+ cumulative emissions for validation case studies H2 Correction Method Prediction Method Percent Error FHR AU FHR LOU SDP EPF Unadjusted NHVCZ Deterministic 42% 79% 372% Monte Carlo 39% 99% 387% H2 Adjusted NHVCZ Deterministic 29% 10% 57% Monte Carlo 26% 8% 51% Generalized Scaling Deterministic 32% 15% 8% Monte Carlo 23% 10% 4% The validation procedure presented in this work had the benefit of objectively demonstrating the improvements gained from the generalized scaling and the Monte Carlo approaches. However, removing a data set for use as a validation set inherently decreases the accuracy of the Monte Carlo method, or any predictive method, by decreasing the size of the existing data set. Additionally, in the Monte Carlo method, the accuracy of the prediction increases as the number of predicted events increase. The validation case studies used in this study provide only a limited number of events. In practice, the Monte Carlo method would be beneficial in estimating cumulative emissions from longer events that may last over several days.. In practice, care should be taken when applying the Monte Carlo method to flaring events. A determination of the appropriateness of the test data must be made prior to applying the Monte Carlo method to a subject flare. The accuracy of the prediction is directly related to whether a subject flare can be described by the probability distribution of the existing test data. CONCLUSIONS Until recently, analysis of flare efficiency has been limited by the small number of industrial flare testing studies. With the development of remote sensing methods, such as AFTIR and PFTIR, measuring combustion efficiency from industrial scale flares has become more feasible. Analysis of recent flare test data by Clean Air Engineering showed that proper operation of a flare can be expected if the net heating value of the combustion zone gases is larger than approximately 300 BTU/scf, [6], [9], [10] [12]. The EPA report, however, proposes that the lower flammable limit of the combustion zone gases3 should be 15.3% by volume or less. These proposed thresholds, however, are not intended to predict emissions for flares, but rather intended to provide guidelines for efficiently operating flares. Furthermore, one criticism of the EPA 2012 report from the report's review panel is that the threshold values are determined to minimize the number of false positives, where a flare may operate within the threshold envelope but has a lower than expected combustion efficiency [16]. Minimizing the false positives results in an increase in false negatives, where the combustion efficiency is high despite being outside the threshold envelope, which presents challenges with respect to reporting of emissions. In the current work, more sophisticated prediction methodologies are presented. A deterministic approach based on curve fitting and a Monte Carlo approach based on random sampling of the existing test data are proposed and applied in the form of several validation case studies. These validation case studies were carried out by removing one test study and using the remaining test data in a predictive manner for the recorded flare composition of the removed test study. The advantage of the Monte Carlo approach is that prediction intervals can be determined directly. The expected emissions obtained by 3 The EPA 2012 report estimates the lower flammable limit of the combustion zone gases by molar averaging the lower flammable limits of all species in the combustion zone. Details of the calculation methodology, which accounts for effects of inert species, can be found in [8]. Monte Carlo were more accurate than those resulting from deterministic analysis of fitted data. The validation scheme presented here inherently reduces the effectiveness of the Monte Carlo method by removing data points. The accuracy of the Monte Carlo method improves as the size of the data set and the number of predicted events increase. A frame work for a generalized scaling was also presented. Compared to the previously published hydrogen adjusted NHVCZG correction, the generalized scaling resulted in a tighter collapse of the data points. The Monte Carlo method can be applied to both hydrogen correction methods; however, better results are obtained using a general scaling-in particular for high hydrogen content flares. For flares with high hydrogen content, the authors recommend using the generalized scaling for predicting emissions as it collapses the high hydrogen data points better and gives more accurate estimates. The proposed generalized scaling also provides a framework to incorporate effects from other physical parameters that were not considered in this study. The scaling functions can be defined for additional parameters, which include crosswind, variable flare tip geometries, inert species and steam injection location. REFERENCES [1] M. McDaniel, "Flare efficiency study," EPA-600/2-83-052, 1983. [2] J. Pohl, R. Payne and J. Lee, "Evaluation of the efficiency of industrial flares," EPA-600/2-84-095, 1984. [3] J. Pohl and N. Soelberg, "Evaluation of the efficiency of industrial flares: flare head design and gas composition," EPA-600/2-85-106, 1985. [4] J. Pohl and N. Soelberg, "Evaluation of the efficiency of industrial flares: H2S gas mixtures and pilot assisted flares," EPA-600/2-86-080, 1986. [5] P. Gogolek and A. Caverly, "Emissions from elevated flares-a survey of the literature," CanmetENERGY, 2009. [6] S. Evans, "Insights from passive FTIR flare performance testing," in AFRC Industrial Flares Colloquium, Houston, TX, 2011. [7] D. Allen and T. V.M., "TCEQ 2010 Flare Study Final Report," PGA No. 582-8-86245-fy-09-04, The Unversity of Texas at Austin, 2011. [8] US EPA Office of Air Quality Planning and Standards, "Parameters for properly designed and operated flares," US EPA, 2012. [9] Clean Air Engineering, Inc., "PFTIR test of steam-assisted elevated flares-Port Arthur," Flint Hills Resources Port Arthur, LLC, 2011. [10] Clean Air Engineering, Inc., "Performance test of a steam-assisted elevated flare with passive FTIR," Marathon Petroleum Company, LLC, 2010. [11] INEOS ABS (USA) Corporation, "Passive Fourier transform infraed technology (FTIR) evaluation of P001 process control device at the INEOS ABS (USA Corporation)," INEOS ABS, Addyston, OH, 2010. [12] Clean Air Engineering, Inc., "Performance test of a steam-assisted elevated flare with passive FTIR-Detroit," Marathon Petroleum Company, LLC, 2010. [13] Shell Global Solutions (US) Inc., "Shell Deer Park Refining LP Deer Park Refinery East Property flare test report," Shell Gobal Solution (US) Inc., Houston, TX, 2011. [14] C. Law, Combustion Physics, New York: Cambridge University Press, 2006. [15] C. Baukal, The John Zink Combustion Handbook, New York: CRC Press, 2001. [16] "Flare Peer Review Panel Comments," 2012. [Online]. Available: http://www.epa.gov/airtoxics/flare/2012flarepeerreviewmemo.pdf. |

ARK | ark:/87278/s6j994k8 |

Setname | uu_afrc |

Date Created | 2014-10-14 |

Date Modified | 2014-10-14 |

ID | 14355 |

Reference URL | https://collections.lib.utah.edu/ark:/87278/s6j994k8 |

Copyright © J. Willard Marriott Library. All rights reserved.