INFLUENCES OF OXYGEN-ENRICHED AIR ON COMBUSTION CHARACTERISTICS OF INDUSTRIAL FURNACES

Yu-Cheng Chang1, Kuo-Kuang Wu2, Chiun-Hsun Chen1 and Young-Da Chen1 Department of Mechanical Engineering, National Chiao Tung University, 1001 University Road, Hsinchu, Taiwan 30010, R. O. C. 2 Green Energy & Environment Research Laboratories, Industrial Technology Research Institute Bldg.64, 195 Sec. 4, Chung Hsing Road, Chutung, Hsinchu, Taiwan 31040, R.O.C Abstract This experimental investigation adopted commercial gas burner to study the influence of oxygen concentration on the flame structures, heating rates, emissions, fuel consumptions and radiant heat transfers between 21% O2 and 30% O2. The fuel was natural gas. Based on experimental results, the increase of the oxygen concentration led to longer flame length and stronger flame brightness. Increase in the oxygen concentration led to a more rapid heating rate and lesser fuel consumption due to lower levels of the inert gas (N2). Compared with 21% O2, the heating rate was increased by 64.3% at the same fuel flow rate for 30% O2. The fuel consumption from 23.0 m3/hr at 21% O2 reduced to 18.0 m3/hr at 30% O2 when the furnace temperature was maintained 1200 and 5% excess air. The energy saving percentage reached 21.74%. Higher oxygen concentrations yielded higher flame temperature; hence, the NOx concentration increased with increasing oxygen concentration. However, the trend of total NOx emission was decreased. Finally, the radiant heat flux was increase with oxygen concentration. INTRODUCTION Dalton et al. adopted conventional air/fuel burners to carry out oxygen-enriched combustion. They found that air/fuel burners could be adapted to low-level oxygen enrichment (generally lower than 28% O2) with no modifications. Qiu and Hayden investigated oxygen-enriched combustion of natural gas in porous ceramic radiant burners, at oxygen concentrations varying between 21 and 28%. The experimental results indicated that the saving in natural gas was about 22% when oxygen concentration was increased to 28%. Czakiert et al. carried out combustion in an oxygen-enriched atmosphere using a circulating fluidized bed combustor. They used brown coal as the fuel, and the oxygen concentration was set as 21, 40, and 60%. The experimental results showed that the ratio of carbon conversion increased with oxygen concentration. Moreover, oxygen-enriched conditions promoted an increase in the conversion of sulfur present in the fuel. Poirier et al. adopted low-NOx burner to explore the influence of oxygen concentration on fuel consumption and emission. The experimental results indicated a linear decrease in fuel consumption with increase in oxygen enrichment when furnace temperature was maintained at 1100C. NOx emission was relatively constant when oxygen-enrichment levels were below 60% and the stack oxygen concentrations were maintained at the same level. However, NOx emission increased with increasing stack oxygen concentration at all oxygen-enrichment levels. Chen and Axelbaum studied the variations of the extinction strain rate under oxygen-enriched conditions both numerically and experimentally. Oxygen-enriched combustion could yield extremely strong flames. They found that the extinction strain rate for a methane flame increased by more than two times as
1

the oxygen concentration was increased to 100%. This study has used a high speed gas-fired burner to investigate the effects of oxygen concentrations lower than 2130%. The measurements included heating rates, emissions, radiation heat flux and fuel consumptions. EXPERIMENTAL SETUP Figure 1 shows a schematic representation of the experimental apparatus. The dimensions of the internal furnace were 2.58 m long, 2.47 m wide and 1.5 m high. The refractory lining for the furnace walls and roof were ceramic fibre, 25 cm thick, which can resist the high temperature up to 1600 . The bottom of the furnace was made up of firebricks. The burner was North American TEMPEST 4442-5 high-speed gas burner. The flame was ignited by a spark pilot, and the combustion occurred in a premixed mode. Both air and oxygen were let into the gas mixer that had a flowmeter to adjust the oxygen concentration in the combustion chamber. The oxygen source in this study is generated from the liquid oxygen. O2/Air mixtures were produced at chosen molar ratios by mixing metered O2 and air before entering the burner. Flue gas samples were continuously drawn from the stack through a stainless probe. All the flue gas samples were dried and filtered before they enter a gas analyzer. The flue gas compositions (O2, CO, CO2, and NO) were measured by Testo 350-XL gas analyzer. The furnace temperature was measured by a sheathed R-type thermocouple. The thermocouple was 0.5 mm diameter of wires and shield by 17 mm diameter of ceramic tube. The radiative heat loss from the thermocouple was not corrected in the present study. The radiative heat fluxes at the right side wall were measured by Vatell TG9000-2 Radiometer. A Fluke 2680A data-acquisition system recorded the temperature and radiative heat flux values at a recording frequency of 1 Hz. The left side wall of combustion chamber opened an 18.5 cm diameter observed window to view the flame. The flame images were recorded by the Sony TRV-950 video camera. Natural gas, with a heating value of 42 MJ/m3, supplied by the Chinese Petroleum Corporation, was used as the fuel. The flow rate of natural gas, oxygen and air were adjusted to maintain furnace temperature at 1200

under different oxygen concentrations.

The experimental data was

started to record when the furnace temperature was maintained at approximately 1200 for a period of at least 90 minutes, and the duration of record was 90 minutes. In this study, five oxygen concentrations were selected for the experiments21%, 24%, 26%, 28%, and 30%. RESULTS AND DISCUSSION In order to find the critical completed combustion point, we increased the equivalence ratios ( ) gradually from 0.83 until the CO concentration increase sharply. Based on experimental results (not show in this paper), we observed that the critical completed combustion point is in the neighborhood of 0.95 and is not very dependent on the oxygen concentration for all cases. Figure 2 displayed the variations of flame length at different equivalence ratios and oxygen concentrations. The flame length is defined by the constant gray-scale value. Because the flame length under a specific condition is impossible fixed, it is obtained by taking 350 pieces of images to make the average. When equivalence ratio increases from 0.87 to 0.98, the increases of flames length were 10.45, 8.49, 10.11, 10.67 and 9.67 cm for 21%, 24%, 26%, 28% and 30% O2. Furthermore, the flame length varied nearly linearly with the equivalence ratio when equivalence ratio was less than 0.98. At this equivalence ratio, the incomplete combustion occurred. The proportion of decrease of flame length was in the range of 11.58-15.77% when equivalence ratio decreased from 0.98 to 0.87. Moreover, the flame was longer when oxygen concentration was increased. For example, the

flame was extended from 55.88 cm to 73.12 cm as oxygen concentration increased from 21% to 30% for = 0.87 . This is because the high-speed burner is an impinging type that both air and fuel flows are impinged together in front of the burner mouth to mix, and then ignited. Air supply was decreased substantially at constant natural gas flow when using oxygenenriched combustion. The driving force of mixing was the momentums of natural gas and combustion air. Accordingly, the fuel/air mixture was not uniform and it velocity was lower leading to lengthen the flame length. In the industry application, the usages of combustion are generally on the heating and/or heat treatment. In heating process, the requirement may be fast heating, such as ladle furnace, or high temperature, such as glass melting furnace. Figure 3 displays the time elapsed for
heating the natural gas fuel to 1200 at different oxygen concentrations. The fuel supply rate was also maintained at 30.5 m3/hr. The elapsed time was decreased as the oxygen concentration increased. This was due to the reduction in the inert gas, nitrogen, as the oxygen concentration increased. Hence, the energy availability is raised, i.e. the energy loss is reduced. The major energy loss is from flue gas. Diluting N2 concentration can reduce energy loss by flue gas apparently so that the energy availability can be raised, i.e. saving energy. Because of this, the combustion heat was more efficient in increasing the furnace temperature. Moreover, the magnitude of decrease in the elapsed time was more obvious at higher oxygen concentrations. For example, the elapsed time for 24% O2 was 80.8% of the time recorded at 21% O2 (11396 s for 21% O2 and 9209 s for 24% O2); however, for 30% O2, the elapsed time was only 64.3% of that at 21% O2 (7330 s for 30% O2). In other words, using combustion air containing 30% O2 could save 35.7% of the fuel compared to using combustion air with 21% O2. This shortening of the heating time implies savings in terms of fuel. Furthermore, the heating rate was comparable for all O2 concentrations when the furnace temperature was lower than 700 . This phenomenon indicates that the advantage of oxygen-enriched combustion is apparent only in the high-temperature environment during the heating process. In many heat treatment procedures, they are usually limited at specific temperature to perform. Because of this, we performed constant furnace temperature tests to understand the influences of oxygen enriched combustion. The firing rate was adjusted to maintain a constant furnace temperature of 1200 as O2 enrichment was varied. Moreover, the equivalence ratios corresponding to the oxygen concentrations from 21% to 30% were maintained in the neighborhood of 0.95, which is critical completed combustion point. This clearly demonstrated fuel savings gained by O2 enrichment and provided a better basis for comparison of other data. According to the experimental results, the CO concentration was maintained at a low level (< 10 ppm) for all O2 concentrations. This indicated that completed combustion was achieved. Figure 4 illustrates the variations of rates of fuel flow (i.e., fuel consumption) at different oxygen concentrations. As shown in the figure, increasing the oxygen concentration could in fact save energy. For example, the fuel consumption decreased from 23.0 m3/hr at 21% O2 to 18.0 m3/hr at 30% O2 at a constant furnace temperature of 1200 . This implies that 21.74% of fuel could be saved when the oxygen concentration increased from 21% to 30%. The magnitudes of energy saving is essentially related to energy availability. The N2 is an inert gas and carries a lot of energy to the ambient environment so that the energy availability was lower. Oxygen-enriched combustion could overcome this drawback because less N2 was contained in the oxidizer. Excluding the lower levels of N2 contained within the combustion air, the radiation heat transfer was enhanced due to higher concentrations of CO2 and H2O while O2 concentration increased. Figure 5 shows the radiation heat flux distributions per energy input at different oxygen concentrations. The

maximum radiation heat flux occurs at a distance of 120 cm from the burner. The measuring point is located in the posterior region of flame, where the area possesses higher temperature. The radiation intensity is proportional to the four power of temperature; consequently, the radiation heat flux is increased. The second highest radiation heat flux is at 210 cm distance from the burner. Because the radiation heat flux is obtained not only from the combustion gases but also from the posterior region of burner. Moreover, the radiation heat flux per energy input was increased with oxygen concentration for every measuring point. For example, the radiation heat flux per energy input was from 0.109 Wcm-2kW-1 to 0.156 Wcm2 kW-1 at second measuring point (at 120 cm from the burner) when oxygen concentration was increased from 21% to 30%. The increased percentage reached 43.1%. The concentrations of CO2 and H2O produced were increased because of dilution of N2 concentration under the fixed air/fuel ratio when increasing of oxygen concentration. Furthermore, the CO2 and H2O are good radiation emitters and their emissivities are proportion to their concentration. Because of this, the total gas emissivity was increased when the oxygen-enrichment level is upgraded; in turn, causes the radiation intensity to become stronger. Moreover, the fuel consumption decreased almost linearly with increasing O2-enrichment level at constant furnace temperature. This trend was the same as the experimental results of Poirier et al. Figure 6 shows the variations of NOx concentrations, corrected to 6% O2, at different oxygen concentration. The NOx emission is defined in this study as the sum of NO and NO2 emissions. The NOx emission increases as the oxygen concentration is enhanced. For example, the NOx concentration increases 1.26 times (from 38 to 48 ppm) as the oxygen concentration is increased from 21 to 26%. However, the NOx concentrations were nearly the same when oxygen concentration was increased to 30% continuously. Generally, the maximum proportion of NOx produced during combustion was thermal NOx, whose production is directly associated with the temperature and the oxygen concentration. Moreover, the thermal NOx shows exponential growth with temperature. Owing to this, even a minor variation in temperature can cause a huge increase in NOx emission. In addition, it can be found that the NOx emission is more sensitive to the excess oxygen prevalent at higher oxygen concentrations during the experimental process. Increase in excess oxygen by even a small amount causes a sharp increase in NOx emission. Because of the effects of temperature and excess oxygen, NOx emission increases obviously with the oxygen concentration. Although the NOx concentration increases with oxygen concentration, the total NOx emission does not show an increase similar to that of the concentration. Because the total NOx emission is a product of the volume concentration of NOx and the flue gas volume flow rate, these two effects increase in NOx concentration and decrease in amounts of flue gascompeted with each other.

CONCLUSIONS
This experimental investigation adopted commercial high-speed gas burner to study the influence of oxygen concentration on the flame characters, combustion behaviors, emissions, fuel consumptions and radiant heat transfers in the range of 2130%. The influences of oxygen concentration on the heating rate, emissions, and fuel consumption were examined. The general conclusions drawn from the results of this work were as follows: 1. The flame length varied nearly linearly with the equivalence ratio when equivalence ratio was less than 0.98. However, the flame length was increased with the oxygen concentration when the equivalence ratio was maintained the same. 2. The heating rate was increased with oxygen concentration. The time of heating to 1200

at 30% O was only 64.3% of that at 21% O .

2 2

3.

The fuel consumption rate was decreased linearly with increase of oxygen concentration when furnace temperature and equivalence ratio were maintained at 1200 and 0.95, respectively. The energy saving could reach 21.74% when the oxygen concentration increased from 21% to 30%. The radiation heat flux per energy input was increased with oxygen concentration because the CO2 and H2O are good radiation emitters and their emissivities are proportion to their concentration.

4.

ACKNOWLEDGEMENTS
The author would like to thank the Bureau of Energy, Ministry of Economic Affairs of the Republic of China for financially supporting this research.

REFERENCES
Baukal, C. E., Oxygen-Enhanced Combustion, CRC Press, New York, 1998. Dalton, A. I., Tyndall, D. W., Oxygen enriched air/natural gas burner system development, NTIS report PB91-167510, Springfield, VA, 1989. Joshi, S. V., Becker, J. S., Lytle, G. C., Effects of oxygen enrichment on the performance of air-fuel burners, in Industrial combustion technologies, Ed. by Lukasiewicz, M. A., American society of metals, Materials park, OH, 1986. Qiu, K. and Hayden, A.C.S., Increasing the efficiency of radiant burners by using polymer membranes, Applied Energy 86 (2009) 349-354. Czakiert, T., Bis, Z., Muskala, W., Nowak, W., Fuel conversion from oxy-fuel combustion in a circulating fluidized bed, Fuel Processing Technology 87 (2006) 531-538. Luo, S. Y., Xiao, B., Hu, Z. Q., Liu, S. M. and Guan, Y. W., Experimental study on oxygenenriched combustion of biomass micro fuel, Energy 34 (2009), 1880-1884. Poirier, D., Grandmaison, E. W., Lawrence, A. D., Matovic, M. D., Boyd, E., Oxygenenriched combustion studies with the low NOx CGRI burner, IFRF Combustion Journal 2004, Article Number 200404. Chen, R. and Axelbaum, R. L, Scalar dissipation rate at extinction and the effects of oxygenenriched combustion, Combustion and Flame 142 (2005) 62-71. Bejarano P. A. and Levendis Y. A., Combustion of coal chars in oxygen-enriched atmospheres, Combustion Science and Technology 179 (2007) 1569-1587. Murphy, J. J. and Shaddix C. R, Combustion kinetics of coal chars in oxygen-enriched environments, Combustion and Flame 144 (2006) 710729. Kennedy, L. A., Saveliev, A. V., Bingue, J. P. and Fridman, A. A., Filtration combustion of a methane wave in air for oxygen-enriched and oxygen-depleted environments, Proceedings of the combustion Institute 29 (2002) 835841. Ho, K. K., Yongmo, K., Sang, M. L. and Kook, Y. A., Studies on Combustion Characteristics and Flame Length of Turbulent Oxy-Fuel Flames, Energy and Fuels 21 (2007), 1459-1467. Hamzeh, J. K. and Mohammad, H. S., Heat transfer and energy analysis of a pusher type reheating furnace using oxygen enhanced air for combustion, Journal of Iron and Steel Research 17 (2010), 1217.

Figure 1: Schematic diagram of the experimental apparatus.

180 160

Flame Length (cm)

140 120 100 80 60 40 20 0 1.00 1.02 1.04 1.06 1.08 1.10 1.12

21% 24% 26% 28% 30%

1.14

1.16

Eexcss Air Ratio

Figure 2: Variation of flame length with equivalence ratios.

Furnace Temperature (oC)

1400 1200 1000 800 600 400 200 0 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 21% 24% 26% 28% 30%

Time (s)

Figur

e 3: Heating times for different oxygen concentrations.

Fuel Flow Rate ( m3 hr )

30 25 20 15 10 18 20 22 24 26 28 30 32

-1

O2 (%)

Figure 4: Variation of fuel consumption at different oxygen concentration.

Radiation Heat Flux (W / kW cm )

-2 -1

0.4 21% 24% 26% 28% 30%

0.3

0.2

0.1

0.0 0 75 90 105 120 135 150 165 180 195 210 225

X Axis (cm)

Figure 5: Radiant heat flux profile of the right side wall.

52

NOx at 6% O2 (ppm)

50 48 46 44 42 40 38 36 20 22 24 26 28 30 32

O2 (%)

Figure 6: Variation of NOx concentration at different oxygen concentration.