air diffusion flame

air diffusion flame

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1 Available online at www.sciencedirect.com ScienceDi...

2MB Sizes 3 Downloads 47 Views

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame Fengshan Liu a, Yuhua Ai b,*, Wenjun Kong b a

Measurement Science and Standards, National Research Council, Building M-9, 1200 Montreal Road, Ottawa, Ontario K1A 0R6, Canada b Institute of Engineering Thermophysics, Chinese Academy of Sciences, 11 Beisihuanxi Road, Beijing 100190, China

article info

abstract

Article history:

A detailed numerical study was conducted to investigate the effects of hydrogen and he-

Received 5 September 2013

lium addition to fuel on soot formation in atmospheric axisymmetric coflow laminar

Received in revised form

methane/air diffusion flame. Detailed gas-phase chemistry and thermal and transport

20 December 2013

properties were employed in the numerical calculations. Soot was modeled using a PAH

Accepted 24 December 2013

based inception model and the HACA mechanism for surface growth and oxidation. Nu-

Available online xxx

merical results were compared with available experimental data. Both experimental and numerical results show that helium addition is more effective than hydrogen addition in

Keywords:

reducing soot loading in the methane/air diffusion flame. These results are different from

Laminar diffusion flame

the previous investigations in ethylene/air diffusion flames. Hydrogen chemically en-

Soot formation

hances soot formation when added to methane. The different chemical effects of hydrogen

Hydrogen addition

addition to ethylene and methane on soot formation are explained in terms of the different

Chemical kinetics

effects of hydrogen addition on propargyl, benzene, and pyrene formation low in the flames. Crown Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

1.

Introduction

Emission of particulates (soot) from combustion devices burning conventional hydrocarbon fuels into the environment has been identified to be detrimental to human health [1] and suspected to be an important contributor to climate change [2]. These concerns have drawn renewed interests in finding alternative fuels that lead to reduced soot emissions. In this regard, hydrogen has been proposed as a promising clean energy carrier in the long run because of its high energy content on mass basis and environmentally friendly nature,

even though significant challenges remain for its production, storage, and safe handling. These challenges currently prevent the widespread use of hydrogen for power generation and transportation. An attractive and feasible approach to use hydrogen for power generation is to develop blended fuels, such as hydrogen/hydrocarbon or syngas/hydrocarbon mixtures. Previous research showed that these fuels can improve combustion and pollutant emission performance, such as combustion ignitability [3], stability [4], and NOx [5,6] and soot formation [7e12]. To help improve the performance of combustion devices burning these blended fuels it is important to understand the fundamental combustion and pollutant

* Corresponding author. Tel.: þ86 10 82543141; fax: þ86 10 82543019. E-mail address: [email protected] (Y. Ai). 0360-3199/$ e see front matter Crown Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijhydene.2013.12.151

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

2

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

formation characteristics of such fuels. Unfortunately, the combustion and emission characteristics of blended fuels are in general difficult to predict without conducting experimental studies or detailed modeling due to the complex and highly nonlinear nature of gas-phase combustion chemistry as well as soot formation and oxidation chemistry. Much work is still required to better understand the complex chemical kinetics mechanisms of these blended fuels to provide fundamental understanding to support their widespread use. To date there has been many studies in the literature on the influence of hydrogen addition to fuel on soot formation in laminar hydrocarbon diffusion and premixed flames. The following literature review focuses on previous studies of hydrogen addition on soot formation in diffusion flames, since diffusion flames are commonly used in practice. Tesner was among the first researchers to investigate the effects of hydrogen on soot formation. He found that dilution of natural gas by hydrogen slows down the formation of carbon black particles during thermal decomposition [13]. In an investigation of soot formation in laminar cylindrical methane diffusion flame, Tesner et al. [14] showed that addition of hydrogen or nitrogen to fuel reduces soot yield in a methane diffusion flame with nitrogen is more effective than hydrogen. Dearden and Long [15] experimentally investigated the influence of hydrogen and nitrogen addition to fuel on the sooting rate of laminar ethylene and propane diffusion flames established on a WolfhardeParker burner. They found that addition of hydrogen to fuel reduces the sooting rate in both flames and in the case of ethylene hydrogen is more effective than nitrogen. Through measurements of the influence of various additives to fuel on the soot-particle inception strain rate in counterflow C2H4, C3H8, and C4H10 diffusion flames, Du et al. [16] found that addition of H2 to these hydrocarbon fuels substantially reduces the soot inception strain rate. Their results also indicate that addition of helium to fuel is somewhat more effective than H2 in suppressing soot formation and addition of H2 results in a higher flame temperature. The possible mechanisms by which the added H2 suppresses soot formation in diffusion flames were discussed by Du et al. [16] as dilution, preferential diffusion, thermal, and chemical. The possible chemical effect of H2 on soot formation was discussed in terms of the hydrogen abstraction acetylene addition (HACA) mechanism of Frenklach [17]. The chemically inhibiting effect of H2 addition to fuel on soot formation was further investigated experimentally by Gu¨lder et al. [7] in laminar axisymmetric coflow C2H4, C3H8, and C4H10 diffusion flames with and without dilution by H2 and helium. Their experimental results showed that hydrogen is much more effective than helium in suppressing soot formation when added to C2H4, but it does not show additional effectiveness when added to C3H8 or C4H10. By arguing that the addition of hydrogen or helium does not significantly alter the temperature distribution (related to the thermal effect) nor the visible flame height (related to the residence time), they concluded that the additional effectiveness of hydrogen addition to C2H4 compared to that of helium addition is due to the chemical effect of hydrogen. Guo et al. [18] numerically investigated the effect of hydrogen and helium addition to the C2H4 flame experimentally investigated by Gu¨lder et al. [7] using a detailed reaction mechanism and a two-equation soot model. Their numerical results reproduce

the additional chemical effect of hydrogen addition on reducing the soot yield, besides the dilution effect revealed in the results of helium addition, observed experimentally by Gu¨lder et al. [7]. The primary chemical effect of hydrogen addition to C2H4 on soot formation was identified to be the decrease in the hydrogen atom concentration in the soot surface growth regions, which leads to a reduced surface active site number density according to the HACA mechanism for soot surface growth [17,19]. In this regard it is also interesting to mention the work of Arthur [20] conducted in 1950, who found that suppression of hydrogen radical in hydrocarbon flames is accompanied by reduced flame luminosity. More recently, the effect of hydrogen addition to fuel on soot formation has been experimentally investigated in laminar coflow acetylene/air [10] and methane/air [21] diffusion flames. Pandey et al. [10] conducted three sets of experiment: (i) hydrogen was introduced through a separate annular port while keeping the overall equivalence ratio constant, (ii) hydrogen was premixed with acetylene while keeping the overall equivalence ratio constant, and (iii) hydrogen was premixed with acetylene while keeping the acetylene flow rate constant. In all three sets of experiment the amount of the added hydrogen was varied between 0 and 9.69% (mass basis). The experimental results of Pandey et al. showed that there is somewhat significant increase in the flame temperature when hydrogen is added, up to 50 K in the first set of experiment and 100 K in the second and third sets of experiment, while the soot volume fraction decreases substantially as more hydrogen is added in all three sets of experiment. Although the increased flame temperature, as a result of hydrogen addition, is in agreement with the observation of Du et al. in counterflow flames [16], it is in disagreement with the results of Gu¨lder et al. [7], who measured temperatures using CARS and found that the addition of hydrogen to fuel has very little influence on the flame temperatures in laminar axisymmetric coflow C2H4, C3H8, and C4H10 diffusion flames. The potential chemical role of hydrogen in the suppression of soot formation in the acetylene diffusion flame was discussed by Pandey et al. [10] based on the HACA mechanism. Migliorini [21] experimentally investigated the effects of hydrogen and helium addition to fuel in a laminar axisymmetric coflow methane/air diffusion flame by adding up to 40% (volume basis) of hydrogen and helium to fuel while keeping the methane flow rate constant. Her results showed that although addition of hydrogen and helium to fuel reduces the total soot loading in the flame, addition of hydrogen is less effective in soot formation reduction than helium. These results suggest that hydrogen might promote soot formation chemically in comparison to its role of dilution, as revealed in the difference between the results of hydrogen and helium addition. However, how the added hydrogen chemically enhances soot formation remains to be identified by detailed numerical modeling. Although studies conducted so far have established that addition of hydrogen to a hydrocarbon fuel reduces soot formation in diffusion flames, there is inconsistency in the literature with regard to if the added hydrogen plays an additional chemical role in suppressing or promoting soot formation besides its dilution and thermal effects, i.e., if

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

hydrogen addition is more effective in reducing soot formation in comparison to helium addition. For example, addition of hydrogen to ethylene was found to chemically suppress soot formation [7,18]. However, the opposite trend was observed by Tesner et al. [14] and Migliorini [21] when hydrogen was added to methane. In this study detailed numerical calculations were carried out to model the laminar axisymmetric coflow methane/air diffusion flames with hydrogen and helium addition to fuel experimentally studied by Migliorini [21]. The objects of this study are to ascertain if and how the added hydrogen in the fuel stream affects soot formation chemically in a laminar methane/air diffusion flame, besides its dilution and thermal effects, and why the added hydrogen affects soot formation differently in methane and ethylene flames.

2.

Numerical model

2.1.

Governing equations, radiation and soot models

The governing equations in our numerical model are identical to those described in previous publications [22,23] and are therefore not described here. It is important to point out that the gravitational acceleration term was retained in the momentum equation in the stream-wise direction (z, vertically upwards) and the radiation source term was included in the energy equation. The method of correction diffusion velocity described in Ref. [24] was employed to ensure that the net diffusion flux of all species sums to zero in both the radial (r) and the stream-wise directions (z). The scrubbing effect of soot formation on the relevant gaseous species involved in soot production and oxidation was taken into account through the modification of the net reaction rate of these species. Only the thermal diffusion velocities of H, H2, and He (helium) were accounted for using the expression given in Ref. [24]. The rationale for this practice was discussed in Ref. [22]. The source term in the energy equation due to radiation heat transfer was calculated using the discrete-ordinates method (DOM) coupled with the statistical narrow-band correlated-k (SNBCK) method for the absorption coefficients of CO, CO2, and H2O [25]. Radiation contribution due to soot was also included and the details were provided in Ref. [23]. Further details of thermal radiation calculations can be found in Refs. [25,26]. Because the simple acetylene-based semi-empirical soot models [27,28] are in general unable to capture the chemical effects of certain chemically active additives, such as hydrogen considered here, on soot formation, it is necessary to employ a PAH based soot model for soot nucleation and surface growth in this study. In this study the PAH based soot model described in previous studies [23,29,30] was employed. The nucleation rate is assumed to be the collision rate of two A4 (pyrene) molecules in the free-molecular region with a van der Waals enhancement factor of 2.2 [31]. Surface growth and oxidation terms are calculated by the HACA mechanism described in Ref. [19]. All parameters are taken from the study of Appel et al. [19] except the parameter a, the fraction of the reactive soot surface. The value of this parameter is bounded

3

between 0 and 1. In the current soot model the a recommended by Xu et al. [32] is used a ¼ 0:004 expð10800=TÞ

(1)

which takes the value of unity for temperatures below about 1950 K. Therefore, a as expressed in Eq. (1) takes the value of unity almost everywhere in soot formation regions in a diffusion flame. There is currently no well established universal a that works equally well in different flames. In the present modeling practice, unfortunately, the value of a is in fact also strongly affected by a particular combustion mechanism, especially the PAH sub-mechanism [33], used in the modeling. It is noticed that soot particle surface growth due to PAH condensation (here by A4 molecules) on soot particle surface was also accounted for in this study. The condensation rate was calculated by the collision rate between A4 molecules and aggregates [34]. Further details of the soot model can be found in Refs. [29,30]. When used in conjunction with the gas-phase reaction mechanism of Appel et al. [19] the soot model described above performed reasonably well in the prediction of soot properties in a laminar axisymmetric coflow ethylene/ air diffusion flame, e.g. [23,29,30].

2.2.

Chemical kinetic mechanism

To model the effect of hydrogen addition to methane on PAH and soot formation it seems reasonable to consider the use of GRI-Mech 3.0 [35], since this mechanism was developed and optimized for methane combustion. Unfortunately, this mechanism consists of primarily the C1 chemistry with only a small number of reactions for C2 and C3 species added. Another obvious choice for the objective of the present study is the C2 mechanism of Appel et al. [19] (hereafter the ABF mechanism), which contains PAH formation chemistry up to pyrene (A4). This mechanism has been employed previously in many studies of soot formation in laminar coflow ethylene/ air diffusion flames, e.g. [29,30]. During the course of the present study, numerical calculations were first conducted using the ABF mechanism for the laminar pure methane/air diffusion flame, i.e., without hydrogen or helium addition. The predicted peak soot volume fraction is very low at only 0.042 ppm, which is about an order of magnitude lower than the experimental data reported by Migliorini [21] (with the peak value about 0.5 ppm) and Trottier et al. [36] (with the peak value about 0.3 ppm). Although such a poor prediction of soot in the methane/air diffusion flame using the ABF gasphase reaction mechanism and the soot model described above may be attributed to deficiencies in either the gas-phase chemistry or the soot model, or both, it is appropriate to state that the gas-phase reaction mechanism is more likely to be the cause for the significant underprediction of soot in the methane diffusion flame, since it was developed for C2 hydrocarbons, such as ethylene and acetylene, not for methane combustion. In addition, use of the present soot model and the ABF gas-phase reaction mechanism led to reasonably good prediction of soot formation in laminar coflow ethylene/air diffusion flames [29,30]. On the other hand, it is well known that GRI-Mech 3.0 performs quite well for methane

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

4

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

combustion for the reason mentioned above. Based on these considerations, the ABF mechanism was revised by replacing its C1 chemistry subset by that of GRI-Mech 3.0 (hereafter termed the revised ABF mechanism). The resultant mechanism has the same numbers of species and reactions as those in the original ABF mechanism. The performance of the ABF and the revised ABF mechanism was evaluated for the prediction of the burning velocity of CH4/air mixtures. The results are compared in Fig. 1. Also plotted in this figure are the results from GRI-Mech 3.0. It can be seen that the ABF mechanism overpredicts the burning velocity for both lean and rich mixtures, especially for lean mixtures. On the other hand, the revised ABF mechanism underpredicts the burning velocity, especially for mixtures around the stoichiometry (for f about 0.8e1.2). Overall it can be seen that the revised ABF mechanism performs somewhat better than the ABF mechanism in the prediction of the burning velocity of CH4/air mixtures. The revised ABF mechanism was used in this study.

2.3.

Numerical method and modeling conditions

Similar to several previous studies [18,23,29,30], the finite volume method was used to discretize the governing equations. The gaseous species equations were solved simultaneously to effectively deal with the stiffness of the system and speedup the convergence process [18,23,29,30]. The sectional soot equations were solved in the same manner as the species equations due to the stiffness of the system. Further details of the numerical method are given in Refs. [23,29,30]. The thermal and transport properties of gaseous species and the chemical reaction rates were obtained using Sandia’s CHEMKIN [37] and TRANSPORT [38] libraries and the databases associated with the ABF and GRI-Mech 3.0 mechanisms [19,35]. In the study of Migliorini [21] the atmospheric-pressure laminar axisymmetric coflow methane/air diffusion flame was established using a co-annular burner consisting of an

inner fuel tube of 10 mm i.d. (with a thickness of about 0.9 mm) and a co-annular air supply tube of about 80 mm i.d. The flow rates for methane and air were 0.38 l/min and 35.4 l/ min, respectively, and both the fuel and air were delivered at room temperature. Under these conditions, a stable laminar diffusion flame was achieved with a visible flame height of about 65 mm [21]. When different amounts of hydrogen or helium were added to the fuel stream (up to 40% on volume basis, which corresponds to 7.7% and 14.3% on mass basis, respectively), the flow rates of methane and air were kept constant. The laminar flames experimentally investigated by Migliorini [21] were modeled in a domain of 10.46 cm (in the streamwise direction)  4.71 cm (in the radial direction) using 210 (z)  88 (r) control volumes. The wall thickness of the central fuel tube was included in the numerical calculations. A non-uniform mesh was used to save computational time while resolving large gradients within the flame. Very fine grids were placed in the r-direction (with the resolution of about 0.2 mm) and near the burner exit in the z-direction (resolution 0.3 mm). A further refinement of the mesh has negligible influences on the results. The problem of modeling coflow laminar diffusion flames with soot formation using detailed chemistry is very computationally intensive. It would be intractable to obtain the solution with serial processing within a reasonable amount of time. Therefore, it is essential to employ parallel processing for such problems. Details of implementing the distributed memory parallelization with domain decomposition have been discussed in Refs. [29,39]. The system of transport equations in the overall flame model was closed with the ideal gas state equation and the adequate boundary conditions. The inlet velocity profile at the fuel stream was assumed to be parabolic, i.e., u ¼ 2uF[1  (r/ RI)2], where r is the radial position, RI is the inner radius of the fuel tube, and uF is the mean fuel stream velocity, which is 6.78 cm/s for the pure methane flame and increases as hydrogen or helium is added to the fuel stream. A uniform velocity of 60 cm/s was assigned to be air stream outside the boundary layer at the outer surface of the fuel pipe. Inside this boundary layer a boundary layer type velocity profile was assumed. The inlet temperatures for fuel and air were both assumed to be 300 K. Symmetry, free-slip and zero-gradient conditions were assumed at the centerline, the outer radial boundary and the exit boundary, respectively. Details with regard to the implementation of the sectional model can be found in Refs. [29,30]. Iterations in all the calculations conducted in this study were stopped after the maximum relative variation in temperature and soot volume fraction over 500 iterations was less than 1  104.

3.

Fig. 1 e Comparison of the predicted burning velocity of CH4/air mixtures using three reaction mechanisms: GRIMech 3.0, ABF, and revised ABF.

Results and discussion

Numerical calculations were conducted for nine diffusion flames: the pure methane flame, four flames with hydrogen addition at 10, 20, 30, and 40% of the methane flow rate (volume basis), and four flames with helium addition, again at 10, 20, 30, and 40% of the methane flow rate, following the experimental work of Migliorini [21].

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

5

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

3.1. Effect of hydrogen and helium addition on temperature

distributions along the flame centerline (not shown) show that the temperature increases at a slower rate and the peak temperature occurs at a higher location, i.e., the flame becomes taller, as more H2 is added to the fuel stream. This trend can also be observed from temperature distributions shown in Fig. 2 and is in agreement with the experimental results of Migliorini [21]. Fig. 3 shows the temperature distributions for 0, 20%, and 40% helium addition. It is evident from Fig. 3 that the peak flame temperature decreases with increasing amount of He in the fuel stream. By adding 20% and 40% He to the fuel stream the peak temperature decreases by about 18 and 33 K, respectively. Similar to the flames with H2 addition, the peak temperature in the flames with He addition also occurs in the annular region mentioned earlier. Examination of the centerline temperature distributions in the pure CH4, 20%, and 40% He addition flames (not shown) reveals that the centerline temperature becomes increasingly lower, but the flame height (defined as the centerline location where the temperature reaches its peak) remains almost unchanged as more He is added to the fuel stream. At the flame tip, the temperature in the 40% He addition flame is about 30 K lower than that in the pure CH4 flame. Such small effects of hydrogen or helium addition up to 40% to the fuel stream (less than about 30 K) are difficult to verify experimentally, since such temperature differences are within or close to the uncertainties of available temperature measurement techniques.

The effect of hydrogen addition on temperature is shown in Fig. 2 for 0, 20%, and 40% H2 addition. The predicted peak temperature in the pure CH4 flame is 1995.2 K, Fig. 2(a). With 20% and 40% H2 addition to the fuel stream, the peak temperature increases by about 8 and 19 K, respectively, Fig. 2(b and c). Overall the effect of adding moderate amount of hydrogen up to 40% (note that this corresponds to 7.7% on mass basis) does not significantly affect the flame temperature. A close examination of the temperature distributions reveal that the peak temperature occurs in the annular region at about r ¼ 0.5 cm and in between about z ¼ 1.5e2.5 cm in the pure CH4 flame and between z ¼ 0.8e3.5 cm in the 40% H2 addition flame. Nevertheless, it may not be possible to detect such a small temperature increase experimentally. In this sense the predicted temperature distributions shown in Fig. 2 are in qualitative agreement with the experimental results of Migliorini [21], who obtained soot temperature distributions in these flames using a two-dimensional two-color flame emission technique and showed that addition of H2 up to 40% has a negligible influence on temperature distribution. The insignificant influence of adding moderate amount of hydrogen up to 40% on temperature is also in agreement with the experimental study of Gu¨lder et al. [7] conducted in a coflow ethylene diffusion flame of a similar size. Temperature

8 7 6 5 4 3 2 1 0 0

0.5

1

r (cm)

1.5

(c) 40% H2 add. Peak: 2014.6 K

9

9

8

8

7

7

6

6

5

5

z (cm)

T 2050 1980 1910 1840 1770 1700 1630 1560 1490 1420 1350 1280 1210 1140 1070 1000 930 860 790 720 650 580 510 440 370 300

9

z (cm)

(b) 20% H2 add. Peak: 2003 K

z (cm)

(a) pure CH4 Peak: 1995.2 K

4

4

3

3

2

2

1

1

0

0 0

0.5

1

r (cm)

1.5

0

0.5

1

1.5

r (cm)

Fig. 2 e Predicted temperature distributions in the pure CH4 flame, (a), 20% H2 addition flame, (b), and 40% H2 addition flame, (c). Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

6

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

8 7 6 5 4 3 2 1 0 0

0.5

1

(c) 40% He add. Peak: 1962 K

9

9

8

8

7

7

6

6

5

5

z (cm)

T 2050 1980 1910 1840 1770 1700 1630 1560 1490 1420 1350 1280 1210 1140 1070 1000 930 860 790 720 650 580 510 440 370 300

9

z (cm)

(b) 20% He add. Peak: 1976.6 K

z (cm)

(a) pure CH4 Peak: 1995.2 K

4

4

3

3

2

2

1

1

0

1.5

r (cm)

0 0

0.5

1

1.5

r (cm)

0

0.5

1

1.5

r (cm)

Fig. 3 e Predicted temperature distributions in the pure CH4 flame, (a), 20% He addition flame, (b), and 40% He addition flame, (c).

3.2.

Effect of hydrogen and helium addition on soot

The modeled and measured distributions of soot volume fraction without and with 30% hydrogen and 30% helium addition are shown in Figs. 4 and 5, respectively. Both the predicted and measured soot volume fraction distributions shown in these figures indicate that addition of either hydrogen or helium reduces soot loading with helium being more effective than hydrogen. It can also be observed from these two figures that the predicted soot appears at a much higher location than the experiment in both the annular region and the centerline region. The model predicts that the peak soot volume fraction occurs in the annular region of the soot shell, Fig. 4, while the measured high soot volume fractions appear in the centerline region, Fig. 5. The measured peak soot volume fractions are much higher than the modeled ones by a factor of 5e6 for the pure CH4 flame, a factor of 10 for the 30% hydrogen addition flame, and a factor of 20 for the 30% helium addition flame. The measured distributions show that there is significant amount of soot in the centerline region between z ¼ 4.6 and 6.4 cm, while the model predicts that soot only occupies a narrow region of about only 8 mm thick along the flame centerline around z ¼ 6 cm. Although the revised ABF mechanism leads to a higher peak soot volume fraction in the pure CH4 flame by a factor of 2 than the ABF mechanism, it predicts an overall similar soot distribution to the ABF mechanism.

To make a quantitative comparison of the influence of H2 and He addition to fuel on soot loading, the cross sectional area integrated soot volume fraction along the flame height is evaluated. The cross sectional area integrated soot volume fraction is defined as ZN Fv ðzÞ ¼ 2p

fv ðr; zÞrdr

(2)

N

The influence of hydrogen addition to fuel on the integrated soot volume fraction along the flame height is shown in Fig. 6. Although the predicted cross sectional area integrated soot volume fractions are much lower than the experimental values (by about an order of magnitude) for the reasons discussed earlier, the model qualitatively predicts the two main features of the effect of hydrogen addition shown in the experimental results, Fig. 6(a). First, Fv(z) decreases with increasing amount of H2 addition. Secondly, the visible flame heights, defined by the centerline location where soot vanishes, increases slightly with increasing amount of H2 addition. The model predicts somewhat higher visible flame heights compared to the experiment (65e70 mm vs. 70e75 mm). It is also evident from Fig. 6 that the model predicts a much stronger impact of hydrogen addition on Fv than the experiment. The influence of helium addition to fuel on Fv(z) is displayed in Fig. 7. Again, the model qualitatively reproduces the

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

7

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

fv, ppm 0.075 0.072 0.069 0.066 0.063 0.06 0.057 0.054 0.051 0.048 0.045 0.042 0.039 0.036 0.033 0.03 0.027 0.024 0.021 0.018 0.015 0.012 0.009 0.006 0.003 0

7 6

z (cm)

5 4 3 2 1 0 0

0.5

1

(c) 30% He addition Peak: 0.0221 ppm

8

8

7

7

6

6

5

5

z (cm)

8

z (cm)

(b) 30% H2 addition Peak: 0.0537 ppm

(a) pure CH4 Peak: 0.0977 ppm

4

4

3

3

2

2

1

1

0

0 0

0.5

r (cm)

r (cm)

1

0

0.5

1

r (cm)

Fig. 4 e Comparison of the calculated soot volume fraction distributions without and with 30% hydrogen and 30% helium addition.

influence of helium addition on the reduction of soot volume fraction. However, the model again significantly overpredicts the effect of helium addition on soot loading reduction compared to the experiment. The influence of helium addition on the visible flame height is also well reproduced, i.e., addition of helium up to 40% has no impact on the visible flame height, though the model predicts somewhat higher flame height by about 4 mm. To further quantify the effect of hydrogen and helium addition on the soot loading, the total amount of soot loading in the flame defined below is evaluated ZN ST ¼

Fv ðzÞdz

(3)

0

which is the integration of soot volume fraction over the whole flame volume. The relative effectiveness of H2 and He addition to fuel on soot suppression in the coflow CH4 diffusion flame is illustrated in Fig. 8 by plotting the normalized total soot loading (by that of the pure CH4 flame) at different levels of hydrogen and helium addition. Although the present model correctly predicts the soot loading reduction by adding hydrogen and helium to the fuel stream, it significantly overpredicts the soot suppression effect of these additives, as already noticed from Figs. 6 and 7, especially in the case of helium addition. Nevertheless, Fig. 8 shows that the soot model used here qualitatively reproduces the relative effectiveness of H2 and He addition on soot suppression revealed experimentally, i.e., H2 is significantly less effective than He to soot suppression in the coflow CH4/air diffusion flame. This

result is in contrast to that obtained previously by Gu¨lder et al. [7] and Guo et al. [18], who demonstrated that H2 is significantly more effective than He on soot suppression when added to fuel in a laminar coflow C2H4/air diffusion flame. To explore why hydrogen addition is less effective on soot suppression than helium in the methane/air diffusion flame the effect of hydrogen and helium addition to fuel on species involved in soot inception and surface growth are examined next.

3.3. Effect of hydrogen and helium addition on PAH and acetylene Besides temperature it is important to analyze the effects of hydrogen and helium addition on PAHs (A1 and A4) and acetylene, since A1 (benzene) is the first aromatic ringstructure species and A4 and acetylene are soot inception and surface growth species. The effect of 30% H2 and 30% He addition to the fuel stream on the radial distributions of A1 mole fraction at four heights low in the flame are shown in Fig. 9. This figure shows that addition of H2 or He results in lower mole fractions of A1 compared to those in the pure CH4 flame. In addition, addition of He is more effective than H2 in lowering the A1 concentration. Fig. 10 displays the radial distributions of A4 mole fraction without and with 30% H2 and He addition at three heights low in the flame. Similar to the effects of H2 and He addition on A1 shown in Fig. 9, addition of either H2 or He lowers the A4 concentrations at these flame heights with the addition of He being more effective. Because the addition of either H2 or He

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

8

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

7 6

z, cm

5 4 3 2

z, cm

fv, ppm 0.5 0.475 0.45 0.425 0.4 0.375 0.35 0.325 0.3 0.275 0.25 0.225 0.2 0.175 0.15 0.125 0.1 0.075 0.05 0.025 0

1 0

(b) 30% H2 addition Peak: 0.543 ppm 8

(c) 30% He addition Peak: 0.454 ppm 8

7

7

6

6

5

5

z, cm

(a) Pure CH4 Peak: 0.582 ppm 8

4 3

3

2

2

1

1

0 0

0.5

1

4

0 0

r, cm

0.5

r, cm

1

0

0.5

1

r, cm

Fig. 5 e Comparison of the measured soot volume fraction distributions without and with 30% hydrogen and helium addition.

0.16 Area integrated soot, ppm cm

2

(a) Exp. data [21]

0.14 0.12 0.10 0.08 0.06 0.04 0.02

z, cm

0.00 1

2

3

4

5

6

7

8

5

6

7

8

0.018 Area integrated soot, ppm cm

2

(b) Numerical results

0.016 0.014

Pure CH4 flame

0.012

10% H2 addition

0.010

20% H2 addition

0.008

30% H2 addition

0.006

40% H2 addition

0.004 0.002 0.000 1

2

3

4

z, cm Fig. 6 e Influence of hydrogen addition on the cross sectional area integrated soot volume fractions: (a) experiment [21], and (b) numerical results.

has very small influence on temperature, the soot nucleation rate low in the flame without and with H2 and He addition follows the same trend as the A4 mole fraction distributions shown in Fig. 10. Although soot nucleation contributes a small fraction of the total soot loading compared to soot surface growth, it is the bottleneck in the overall soot formation process. The radial distributions of acetylene mole fraction without and with 30% H2 and He addition at the three heights low in the flame are shown in Fig. 11. With a 30% He addition to the fuel the acetylene mole fractions are significantly reduced. When a 30% H2 is added to the fuel, the peak values of acetylene mole fraction at these heights are only slightly reduced, but the radial profile of acetylene concentration is somewhat broadened, which becomes more pronounced with increasing the flame height. Soot surface growth rate is dependent on acetylene (through the HACA mechanism) and A4 (through PAH condensation) concentrations, soot surface area, and the number of dehydrogenated sites per unit surface area csoot, which is an important parameter in the HACA surface growth sequence and its expression is given in Ref. [29]. Examination of csoot distributions in the flames of pure CH4 and 30% H2 and He addition (not shown) reveals that csoot in the 30% H2 and 30% He addition cases is about 10% higher and lower than that in the pure CH4 flame, respectively. Based on the numerical

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

9

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

0.16

2.2e-4

0.12 0.10 0.08 0.06 0.04 0.02 z, cm

0.00 1

2

3

4

5

6

7

8

Pure CH4 30% H2 addition

1.8e-4

30% He addition

1.6e-4 1.4e-4 z = 3 cm

1.2e-4 1.0e-4 8.0e-5 6.0e-5

z = 2 cm z = 1 cm

4.0e-5

0.018

(b) Numerical results

2

Area integrated soot, ppm cm

z = 4 cm

2.0e-4

Mole fraction of benzene

Area integrated soot, ppm cm

2

(a) Exp. data [21]

0.14

0.016 0.014

Pure CH4 flame

0.012

10% He addition 20% He addition 30% He addition 40% He addition

0.010 0.008

2.0e-5 0.0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

r, cm Fig. 9 e Radial distributions of benzene mole fraction at four flame heights without and with 30% addition of hydrogen or helium to the fuel stream.

0.006 0.004 0.002 0.000 1

2

3

4

5

6

7

8

z, cm Fig. 7 e Influence of helium addition on the cross sectional area integrated soot volume fractions: (a) experiment [21], and (b) numerical results.

results obtained in this study, the chemical effect of H2 addition can be summarized as follows. Given the effects of H2 and He addition on temperature, A4 and acetylene concentrations, and csoot, it is therefore clear that both soot nucleation and surface growth rates are higher in the case of H2 addition. This is why the soot loading is higher with H2 addition than with He addition, though the addition of either H2 or He suppresses soot formation. Because the added H2 is actively involved in combustion chemistry, its participation in soot chemistry is also expected. The added H2 affects soot formation mainly through the following pathway. First, the added H2 promotes H and OH radical concentrations through the chain branching

reaction H2 þ O 4 H þ OH. The enhanced H radical directly affects the soot surface growth process through the HACA sequences. On the other hand, the added H2 and the resultant higher H radical concentration lead to higher concentrations of acetylene and propargyl. The latter is mainly responsible for higher concentration of benzene through the recombination reaction.

3.4. Different effects of hydrogen addition to CH4 and C2H4 on soot formation To under why the added H2 plays a different chemical effect on soot formation in laminar methane/air and ethylene/air flames three additional runs were conducted: pure C2H4 coflow diffusion flame, a 30% H2 addition to C2H4, and a 30% He addition to C2H4. In these calculations, the ethylene flow rate was half of the methane flow rate to keep the carbon mass flux constant. The present numerical results also show that helium is indeed more effective than hydrogen in

2.6e-8

0.8

Mole fraction of A4

Normalized soot yield

1.0

0.6

0.4 H2 dilution, Exp. 0.2

2.4e-8

Pure CH4 flame

2.2e-8

30% H2 addition

2.0e-8

30% He addition

1.8e-8 z = 3 cm

1.6e-8 1.4e-8

z = 2 cm

1.2e-8 1.0e-8 8.0e-9

H2 dilution, Model

6.0e-9

He dilution, Exp. He dilution, Model

4.0e-9

z = 1 cm

2.0e-9

0.0

0.0

0

10

20

30

40

Percentage of addition Fig. 8 e Comparison of the predicted and measured normalized total soot loadings in the coflow CH4 diffusion flame at different amount of H2 and He addition up to 40%.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

r, cm Fig. 10 e Radial distributions of pyrene mole fraction at three flame heights without and with 30% addition of hydrogen or helium to the fuel stream.

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

10

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

0.010 z = 3 cm

Mole fraction of C2H2

0.008

0.006

z = 2 cm

0.004

z = 1 cm

0.002

Pure CH4 30% H2 addition 30% He addition

0.000 0.0

0.1

0.2

0.3

0.4

0.5

0.6

r, cm

3. Addition of either hydrogen or helium to methane suppresses soot formation, but addition of helium is more effective. 4. Compared to the case of helium addition, hydrogen addition causes additional chemical effect in enhancing soot formation besides its dilution and thermal effect. The chemical effect associated with hydrogen addition on soot formation is attributed to higher soot nucleation and surface growth rates due to higher concentrations of pyrene and acetylene. 5. The different chemical effects of hydrogen on soot formation when added to fuel in methane and ethylene diffusion flames were found to be the different effects of hydrogen addition on soot nucleation in the very early stage of soot formation.

Fig. 11 e Radial distributions of acetylene mole fraction at three flame heights without and with 30% addition of hydrogen or helium to the fuel stream.

Acknowledgments suppressing soot formation in the ethylene diffusion flame, in agreement with the previous numerical study of Guo et al. [18] and the experimental findings of Gu¨lder et al. [7]. A detailed comparison between the effects of H2 and He addition to fuel in the methane and ethylene flames on various flame properties indicates that addition of H2 leads to higher temperature, higher H radical concentration, and higher C2H2 concentration than He addition in both flames. Unlike H2 addition to methane, however, H2 addition to ethylene produces less C3H3, less A1, and less A4 than He addition. Therefore, the different chemical effects of H2 addition to methane and ethylene on soot formation can be attributed to its different effects on soot nucleation process.

4.

Conclusions

Numerical calculations were carried out to investigate the effects of hydrogen and helium addition to fuel on soot formation in a laminar coflow methane/air diffusion flame using detailed physical models, a revised C2 chemistry model with PAH formation up to pyrene, and a PAH based soot formation model. Although the predicted soot volume fraction displays large discrepancies with the experimental results in its distribution and peak value, numerical results are in qualitative agreement with available experimental data on the effects of hydrogen and helium addition on the visible flame height, temperature distribution, and the relative effectiveness of hydrogen and helium addition on soot loading reduction. The following conclusions can be drawn based on the present study: 1. Addition of hydrogen to the fuel stream leads to a taller visible flame height, while addition of helium to fuel up to 40% has no influence on the visible flame height. 2. Addition of hydrogen results in a slightly delayed temperature rise in the centerline region and a slight increase in the peak temperature. Addition of helium slightly lowers temperatures.

We would like to thank Dr. Francesca Migliorini of CNR IENI Unita` Operativa di Supporto di Milano, Italy, to generously make her experimental data available. Drs. Ai and Kong thank the National Natural Science Foundation of China for financial support through contract No. 50936005. Financial supports to Drs. Ai and Kong from the 863 Program of China under grant number 2011AA050606 and 973 Program of China through contract number 2014CB239603 are gratefully acknowledged.

references

[1] Pope III CA, Dockery DW. Health effects of fine particulate air pollution: lines that connect. Air Waste Mange Assoc 2006;56:709e42. [2] Jacobson MZ. Control of fossil-fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming. J Geophys Rev 2002;107. ACH 16e21/16-22. [3] Gersen S, Anikin NB, Mokhov AV, Levinsky HB. Ignition properties of methane/hydrogen mixtures in a rapid compression machine. Int J Hydrogen Energy 2008;33:1957e64. [4] Leung T, Wierzba I. The effect of hydrogen addition on biogas non-premixed jet flame stability in a co-flowing air stream. Int J Hydrogen Energy 2008;33:3856e62. ¨ L. The effect of [5] Guo H, Smallwood GJ, Liu F, Ju Y, Gu¨lder O hydrogen addition on flammability limit and NOx emission in ultra-lean counterflow CH4/air premixed flames. Proc Combust Inst 2005;30:303e11. [6] Shin B, Cho Y, Han D, Song S, Chun KM. Hydrogen effects on NOx emissions and brake thermal efficiency in a diesel engine under low-temperature and heavy-EGR conditions. Int J Hydrogen Energy 2011;36:6281e91. ¨ L, Snelling DR, Sawchuk RA. Influence of hydrogen [7] Gu¨lder O addition to fuel on temperature field and soot formation in diffusion flames. Proc Combust Inst 1996;26:2351e8. [8] Choudhuri AR, Gollahalli SR. Combustion characteristics of hydrogen-hydrocarbon hybrid fuels. Int J Hydrogen Energy 2000;25:451e62. [9] Kumar P, Mishra DP. Experimental investigation of laminar LPG-H2 jet diffusion flame. Int J Hydrogen Energy 2008;33:225e31.

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 1

[10] Pandey P, Pundir BP, Panigrahi PK. Hydrogen addition to acetylene-air laminar diffusion flames: studies on soot formation under different flow arrangements. Combust Flame 2007;148:249e62. [11] Park SH, Lee KM, Hwang CH. Effects of hydrogen addition on soot formation and oxidation in laminar premixed C2H2/air flames. Int J Hydrogen Energy 2011;36:9304e11. [12] De Iuliis S, Maffi S, Migliorini F, Cignoli F, Zizak G. Effect of hydrogen addition on soot formation in an ethylene/air premixed flame. Appl Phys B 2012;106:707e15. [13] Tesner PA. Formation of dispersed carbon by thermal decomposition of hydrocarbons. Proc Combust Inst 1958;7:546e53. [14] Tesner PA, Robinovitch HJ, Rafalkes IS. The formation of dispersed carbon in hydrocarbon diffusion flames. Proc Combust Inst 1960;8:801e6. [15] Dearden P, Long R. Soot formation in ethylene and propane diffusion flames. J Appl Chem 1968;18:243e51. [16] Du DX, Axelbaum RL, Law CK. Soot formation in strained diffusion flames with gaseous additives. Combust Flame 1995;102:11e20. [17] Frenklach M. On the driving force of PAH production. Proc Combust Inst 1988;22:1075e82. ¨ L. A numerical study on [18] Guo H, Liu F, Smallwood GJ, Gu¨lder O the influence of hydrogen addition on soot formation in a laminar ethylene-air diffusion flame. Combust Flame 2006;145:324e38. [19] Appel J, Bockhorn H, Frenklach M. Kinetic modeling of soot formation with detailed chemistry and physics: laminar premixed flames of C2 hydrocarbons. Combust Flame 2000;121:122e36. [20] Arthur JR. Some reactions of atomic hydrogen in flames. Nature 1950;165:557e8. [21] Migliorini F. Study of combustion process of hydrogenhydrocarbon mixtures. Ph.D thesis. Politecnico di Milano; 2008. ¨ L. A numerical [22] Guo H, Liu F, Smallwood GJ, Gu¨lder O investigation on soot formation in ethylene/air diffusion flames. Int J Comput Fluid Dyna 2004;18:139e51. [23] Liu F, He X, Ma X, Zhang Q, Thomson MJ, Guo H, et al. An experimental and numerical study of the effects of dimethyl ether addition to fuel on polycyclic aromatic hydrocarbon and soot formation in laminar coflow ethylene/air diffusion flames. Combust Flame 2011;158:547e63. [24] Kee RJ, Grcar JF, Smooke MD, Miller JA. A Fortran program for modeling steady laminar one-dimensional premixed flames. Report SAND85e8240. Sandia National Laboratories; 1985. [25] Liu F, Smallwood GJ. An efficient approach for the implementation of the SNB based correlated-k method and its evaluation. JQSRT 2004;84:465e75.

11

[26] Liu F, Guo H, Smallwood GJ. Effects of radiation model on the modeling of a laminar coflow methane/air diffusion flame. Combust Flame 2004;138:136e54. [27] Leung KM, Lindstedt RP, Jones WP. A simplified reaction mechanism for soot formation in nonpremixed flames. Combust Flame 1991;87:289e305. ¨ L. Effects of gas and soot [28] Liu F, Guo H, Smallwood GJ, Gu¨lder O radiation on soot formation in a coflow laminar ethylene diffusion flame. JQSRT 2002;73:409e21. [29] Zhang Q. Detailed modeling of soot formation/oxidation in laminar coflow diffusion flames. Ph.D Thesis. University of Toronto; 2009. [30] Zhang Q, Guo H, Liu F, Smallwood GJ, Thomson MJ. Modeling of soot aggregate formation and size distribution in a laminar ethylene/air coflow diffusion flame with detailed PAH chemistry and an advanced sectional serosol dynamics model. Proc Combust Inst 2009;32:761e8. [31] Frenklach M, Wang H. Detailed mechanism and modeling of soot particle formation. In: Bockhorn H, editor. Soot formation on combustion, mechanism and models. Berlin: Springer-Verlag; 1994. pp. 165e92. [32] Xu F, Sunderland PB, Faeth GM. Soot formation in laminar premixed ethylene/air flames at atmospheric pressure. Combust Flame 1997;108:471e93. [33] Dworkin SB, Zhang Q, Thomson MJ, Slavinskaya NA, Riedel U. Application of an enhanced PAH growth model to soot formation in a laminar coflow ethylene/air diffusion flame. Combust Flame 2011;158:1682e95. [34] Park SH, Rogak SN, Bushe WK, Wen JZ, Thomson MJ. An aerosol model to predict size and structure of soot particles. Combust Theory Model 2005;9:499e513. [35] Smith GP, Golden DM, Frenklach M, Moriarty NW, Eiteneer B, Goldenberg M, et al. http://www.me.berkeley.edu/gri_mech/. [36] Trottier S, Guo H, Smallwood GJ, Johnson MR. Measurement and modeling of the sooting propensity of binary fuel mixtures. Proc Combust Inst 2007;31:611e9. [37] Kee RJ, Rupley FM, Miller JA. Chemkin-II: a Fortran chemical kinetics package for the analysis of gas-phase chemical kinetics. Sandia Report SAND89e8009. Sandia National Laboratories; 1989. [38] Kee RJ, Dixon-Lewis G, Warnatz J, Coltrin ME, Miller JA. A Fortran computer code package for the evaluation of gasphase multicomponent transport properties. Sandia Report SAND86e8246. Sandia National Laboratories; 1986. [39] Zhang Q, Guo H, Liu F, Smallwood GJ, Thomson MJ. Implementation of an advanced fixed sectional aerosol dynamics model with soot aggregate formation in a laminar methane/air coflow diffusion flame. Combust Theory Model 2008;12:621e41.

Please cite this article in press as: Liu F, et al., Effect of hydrogen and helium addition to fuel on soot formation in an axisymmetric coflow laminar methane/air diffusion flame, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2013.12.151