Acta Polytechnica https://doi.org/10.14311/AP.2022.62.0341 Acta Polytechnica 62(3):341–351, 2022 © 2022 The Author(s). Licensed under a CC-BY 4.0 licence Published by the Czech Technical University in Prague APPLICABILITY OF SECONDARY DENITRIFICATION MEASURES ON A FLUIDIZED BED BOILER Jitka Jeníková∗, Kristýna Michaliková, František Hrdlička, Jan Hrdlička, Lukáš Pilař, Matěj Vodička, Pavel Skopec Czech Technical University in Prague, Faculty of Mechanical Engineering, Department of Energy Engineering, Technická 4, Prague 6, Dejvice, 160 00, Czech Republic ∗ corresponding author: jitka.jenikova@fs.cvut.cz Abstract. This article compares the performance of selective catalytic reduction (SCR) and selective non-catalytic reduction (SNCR) applied on the same pilot unit, a 500 kW fluidized bed boiler burning Czech lignite. A correlation of the denitrification efficiency and the normalized stoichiometric ratio (NSR) is investigated. The fundamental principle of the SCR and SNCR is similar with the same reaction scheme. The difference is in the use of the catalyst that lowers the activation energy of the key reaction. As a result, the reduction is performed at lower temperatures during the SCR method. During experiments, the NSR was up to 1.6 for the SCR method. For the SNCR method, which has a higher reducing agent consumption, the maximum denitrification efficiency was reached for NSR of about 2.5. The efficiency of both secondary methods was investigated. The denitrification efficiency during experiments exceeded 98 % for the SCR method, and the SNCR method, together with the primary measures, reached an efficiency of 58 %. Keywords: SCR, SNCR, fluidized bed boiler, denitrification, deNOX, coal. 1. Introduction Many countries rely on and will have to rely on the combustion of fossil fuels for electricity and heat pro- duction over the next few years. The combustion of fossil fuels is associated with the production of pollu- tants that must be minimized in order to operate the technology with low environmental impacts. Nitro- gen oxides can be identified among typical pollutants and are responsible for acid gas deposition, ozone de- pletion, and health effects on humans. In the field of pollutant reduction, the most important regula- tion is given by the BAT (Best Available Techniques) Reference Document for Large Combustion Plants (LCP) [1] that describes the primary and secondary measures to reduce the release of nitrogen oxides (so- called denitrification) from combustion plants to the atmosphere. These measures, which are usable for flu- idized bed boilers, are summarized in Table 1 with the corresponding general NOX reduction rates efficiencies. This article is focused on the experimental inves- tigation of secondary denitrification measures in a bubbling fluidized bed boiler using Czech lignite as a fuel. Reachable denitrification levels are analysed using the SCR and SNCR technologies. The mitiga- tion of nitrogen oxides is important for more than just meeting the BAT and emission standards. Regarding upcoming trends of lowering CO2 emissions from en- ergy conversion, combustion systems using fossil fuels can be extended by CCS/U technologies, most typi- cally post-combustion systems or oxy-fuel combustion. The reduction in NOX production is crucial for those Primary measures NOX reductionrate [%] Low excess air firing 10–44 Air staging 10–77 Flue-gas recirculation (FGR) 20–60 Reduction of the combustion air temperature 20–30 Secondary measures Selective catalytic reduction (SCR) 80–95 Selective non-catalytic reduction (SNCR) 30–50 Table 1. NOX reduction rates of primary and sec- ondary measures [1]. technologies as well, since a high purity of CO2 and low levels of acid-forming gases (like nitrogen oxides) are required. 2. Nitrogen oxides 2.1. Formation of NOX emissions There are three known mechanisms of nitrogen oxides formation in combustion processes [2–4]: • thermal NOX – oxidation of molecular nitrogen from the oxidant at high temperatures, known as the Zeldovich mechanism, • fuel NOX – oxidation of chemically bound nitrogen in solid fuels, 341 https://doi.org/10.14311/AP.2022.62.0341 https://creativecommons.org/licenses/by/4.0/ https://www.cvut.cz/en J. Jeníková, K. Michaliková, F. Hrdlička et al. Acta Polytechnica • prompt NOX – reactions of molecular nitrogen with hydrocarbon radicals with subsequent oxidation of intermediate products in high-temperature reducing flame zones, known as the Fenimore mechanism. NOX from conventional coal combustion typically con- sists of nitric oxide (NO) and nitrogen dioxide (NO2), where NO is the most dominant with a share of about 90 % and more. The dominating formation mecha- nisms depend on the type of combustor. In the high- temperature systems, e.g. pulverized coal combustion, Zeldovich and Fenimore mechanisms are more dom- inant, while fuel-N oxidation dominates in fluidized beds. The fuel-N mechanism is only weakly depen- dent on the combustion temperature, and there is a proportional correlation with oxygen stoichiometry [5– 8]. In addition to NO and NO2, a significant nitrous oxide production (N2O) can also be observed. N2O is not part of NOX emission limits and the measures for its reduction are not part of BAT, it is a gas of importance due to its high GWP potential [9]. The measured N2O emissions from coal combustion sys- tems (except fluidized bed) as the ratio of N2O/NOX emissions are typically less than 2 percent [4]. For coal-fired fluidized bed combustors, N2O emissions are within the range of 17 to 48 % of overall NOX emis- sions [4]. N2O is produced in fluidized bed boilers due to its dependence on the bed temperature. A higher temperature leads to lower N2O emissions, which is the reverse of the bed temperature dependence of NO formation [10]. The amount of conversion of fuel- bound nitrogen to NO and N2O is considered to be roughly constant as shown by de las Obras-Loscertales et al. [11]. 2.2. Denitrification methods Denitrification is a general term for a NOX limitation. Technologies for NOX reduction can be categorized into primary measures that consist of modifying the operating parameters of combustion, leading to a sup- pression of the formation mechanisms, and secondary measures. Secondary measures represent flue gas treat- ment leading to the reduction of NOX already formed. Those technologies can be used independently or in combinations. 2.2.1. Primary measures The primary measures are typically most effective for the Zeldovich and Fenimore mechanisms, and they are focused on reducing the oxygen available in the com- bustion zone and reducing the peak temperatures. Pri- mary measures technologies include air or fuel staging, low NOX burners, and flue gas recirculation systems. When solid fuels are burned in fluidized bed boilers, the relevant measures to reduce the NOX produc- tion are those that focus on fuel-N-originating NOX. As explained in Section 2.1, the fuel-NOX mecha- nism is mostly dependent on the concentration of oxygen in the primary combustion zone. Therefore, the most effective measures aim only at lowering the stoichiometry of the primary air and not at lower- ing the combustion temperature, since the fluidized bed temperature is inevitably too low for the thermal and prompt mechanism to occur. In particular, the only primary measure, which is not an inherent part of the fluidized bed combustion control process, is the staged injection of combustion air. It is used to achieve the required combustion parameters, such as sub-stoichiometric conditions in the dense bed (which decrease the NOX formation), simultaneous combus- tion of the unburned CO in the freeboard section, and increase of the freeboard temperature for efficient injection of the reducing agent. When secondary air is used to burn unburned CO, no more nitrogen ox- ides are formed in the freeboard section [3, 12–16]. Air staging has been shown to be an effective pri- mary measure for NOX reduction in a fluidized bed boiler; for example, Lupiáñez et al. [13] observed a 40 % reduction in NOX for a 20 % secondary air ratio as compared to NOX without air staging. However, air staging shows insufficient NOX reduction rates to meet the emission limits defined in the LCP directive, and secondary measures have to be applied. 2.2.2. Secondary measures Secondary measures, also called post-combustion methods, represent a group of chemical processes in which already formed nitrogen oxides are decomposed into molecular nitrogen and water vapor using a reduc- ing agent. Typical reducing agents are ammonia and urea solutions. Selective non-catalytic reduction and selective catalytic reduction are the basic secondary methods. Other processes developed to date, such as simultaneous denitrification and desulphurization methods or wet scrubbing, have not been applied on a larger scale [7, 17, 18]. SNCR The selective non-catalytic reduction is a method that reduces nitrogen oxides in the absence of a catalyst. The process is based on the following reaction [18]: 4 NH3 + 4 NO + O2 4 N2 + 6 H2O (1) To achieve a sufficient NO to N2 conversion, the reaction temperature of 900 ◦C is required according to the calculation of the Gibbs energy. The typical temperature window for SNCR in industrial applica- tions is between 850 and 1100 ◦C. When the reducing agent is injected into the low temperature region, the nitrogen oxides do not react with the NH2 radical due to the low reaction rate and unreacted ammonia leaves the combustor with the flue gas. As a result, the concentration of ammonia in the flue gas increases and it may also be adsorbed on fly ash particles. On the other hand, when the reducing agent is injected above the high boundary of the temperature range, the NH2 radical preferentially begins to react with 342 vol. 62 no. 3/2022 Applicability of secondary denitrification measures on a fluidized bed boiler Figure 1. Possible locations of secondary denitrification measures. oxygen, resulting in an increase in NOX concentra- tion in the flue gas. The efficiency of this method is, therefore, highly dependent on the injection of the reducing agent at the right temperature, which can differ with the used reducing agent. The optimum temperature for the reaction also varies depending on the reducing agent used. For example, for ammonia, it is in the range of 850–1000 ◦C, and for urea, between 950–1100 ◦C [4–6, 17]. SCR The fundamental principle of the selective cat- alytic reduction is similar to that of the non-selective reduction with the same reaction scheme. In this method, a catalyst is used that lowers the activation energy of the key reaction. As a result, the reduc- tion can be performed at lower temperatures, and there is no need to keep the reacting substances for the necessary period of time in the high-temperature region. Depending on the type of catalyst, the tem- perature window can be 250 ◦C to 600 ◦C (for zeolite catalysts) [7], but the most commonly used vanadium pentoxide catalyst in the titanium dioxide carrier has an optimal temperature window of 250–430 ◦C with an achievable denitrification efficiency of more than 90 % [3, 7, 17, 19]. The lower temperature limit is set by the reaction rate and formation and deposition of the ammonium sulphate salt, which may deposit on the catalyst and cause its temporary deactivation. The upper temperature limit is established by physical damage to the catalyst by sintering and by oxidation of NH3 to NO, thus limiting the NOX conversion, and by supersaturation of the catalyst that leads to an excess of unreacted ammonia, which escapes along with the combustion gas [4, 18, 20]. The latest V2O5- based SCR catalysts are produced with the addition of tungsten trioxide (WO3) and molybdenum triox- ide (MoO3), which are added for the expansion of the optimal temperature window and because of their ability to resist catalyst poisoning. These are ap- plied by impregnation on a TiO2 support that has a good resistance to sulfur oxides. This support is coated on the ceramic skeleton of the catalyst body. Vanadium catalysts work the best at temperatures of about 350 ◦C [5]. At lower temperatures, their effi- ciency decreases rapidly and at higher temperatures, corrosion problems arise [2, 7, 19]. The catalyst can be placed at different locations along the flue gas con- duit as shown in Figure 1, and the placement depends on its type and material. It is not appropriate to place the catalyst in the high dust region for the flu- idized bed combustion while using the dry additive desulfurization method because of the high abrasion properties of the present particles. 3. Experimental set-up 3.1. Experimental facility The experimental boiler is located in the CTU lab- oratories in Prague. This pilot unit is a fluidized bed boiler with a thermal output of 500 kW, and its scheme is shown in Figure 2. Fluidization is achieved by primary air together with recirculated flue gas passing through the distrib- utor, which consists of 36 nozzles. The distributor is described in detail in [21] and the boiler in [22]. The combustion chamber has a cylindrical cross section. In the freeboard area, there are 6 thermocouples placed along the height. The secondary air is supplied to the freeboard section by four evenly placed distributors on a perimeter, and each distributor can provide a secondary air inlet at 4 different heights. The heat exchanger is located in the second descending draft of the boiler. The flue gases are sampled downstream of the boiler prior to the cyclone particle separator, and its composition is continuously analysed. In particular, the volumetric fractions of the following components are measured: O2 using a paramagnetic sensor; SO2, NOX, CO2 and CO using NDIR analysers. The boiler can also be operated in oxy-fuel mode. The off gas was also sampled downstream of the deNOX unit and analysed using the multicomponent FT-IR analyser. The SNCR reducing agent distribution line basi- cally consists of two main components: a probe with a spray nozzle and a system for transporting the re- ducing agent to the spray system. The probe is cooled by water to prevent the reducing agent from boiling before it is sprayed. Compressed air is introduced in front of the nozzle orifice to improve the atomiza- tion of the supplied reducing agent. It is possible to place the probe in various inspection holes in the combustion space of the boiler and thus change the height of the injection of the reducing agent. For the experiments, secondary air inlets at a height of 343 J. Jeníková, K. Michaliková, F. Hrdlička et al. Acta Polytechnica Figure 2. Fluidized bed boiler scheme. Properties “As received” Properties “Dry ash free” LHV Water Ash C H O N S [MJ/kg] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] [wt. %] 18.5 25.0 9.3 72.3 6.3 19.0 1.1 1.3 Table 2. Proximate and ultimate analysis of the fuel 550 mm above the fluidized bed were used to achieve the optimal temperature window. The catalyst for the SCR method has dimensions of 160 mm × 160 mm × 1260 mm. The reduction of NOX in the flue gas is carried out by means of ammonia, which is dosed into the flue gas stream before the reactor itself. The technology is connected to the output of a cyclone separator from the fluidized bed boiler. Dedusted flue gases at a temperature of 150– 180 ◦C are heated in an electric heater to the required temperature of 250–300 ◦C. The amount of flue gas that passes through the reactor at a speed of 4.5 m/s through the catalyst is approximately 150 Nm3/hour. It is necessary to inject the ammonia gas into the flue gas for the reaction on the catalyst surface. The stoichiometric amount of ammonia is approximately 0.012–0.016 kg/h. The catalyst itself is a honeycomb- type based on vanadium pentoxide (V2O5) with the addition of tungsten trioxide (WO3) and molybdenum trioxide (MoO3). Doping is applied by impregnation of the TiO2 supporting body and is used to improve the mechanical stability and chemical resistance of the catalyst, which is related to the widening of the optimal temperature window. 3.2. Fuel and reducing agents Lignite from the coal basin of North Bohemia was used as fuel for the experiments. Its proximate and ultimate analysis is shown in Table 2. The size of the coal particles was less than 10 mm. Pure ammonia was used as the SCR reducing agent and AdBlue (a chemically highly pure aqueous solution of synthetic urea – 32.5 % wt. urea) was used for the SNCR. 3.3. Methods The normalized stoichiometric ratio is the proportion of the molar ratio of the reducing agent and nitrogen oxides at the beginning of the denitrification process. The range of variables measured was as follows. A de- tailed description of the variables is given in Table 2, where the O2 concentrations are related to 6 % vol. of O2 in dry flue gas. For the SNCR method: • NSR values from 0.55 to 3.47, • application of primary measures (flue gas recircula- tion and air staging), • temperature for reducing agent injection from 880 to 950 ◦C, • the average time in one setting was 45 min. For the SCR method: • NSR values from 0.29 to 1.6, • application of primary measures (flue gas recircula- tion), • catalyst temperatures from 259 to 299 ◦C, • the average time in one setting was 74 min. Individual states were set maintaining a constant temperature while the injection of the reducing agent was gradually changed. The urea solution was chosen 344 vol. 62 no. 3/2022 Applicability of secondary denitrification measures on a fluidized bed boiler Figure 3. Summarization of the experimental results for SCR and SNCR. Figure 4. Achievable NOX level for SCR and SNCR methods in NSR for the same BFB boiler. *BAT-AELs (mg/Nm3) for NOX emissions from combustion of black and/or brown coal into the atmosphere for the new combustion plants with the total rated thermal output < 100 MWth of 100–150 mg/mN3 [23]. for the SNCR due to the properties of ammonia, which has very strict storage rules, and therefore is not used in large industrial plants for the SNCR method. Nev- ertheless, the fundamental reaction scheme for SCR and SNCR and both used reagents remains similar. Furthermore, the experiments performed are closer to the practical results due to the size of the experimental boiler. 4. Results and discussion Measured data are presented in Table 3 and Figure 4. More detailed data with their variations are listed in Appendix. The CO emissions were also measured, however, no significant correlation between CO and NOX was observed as can be seen in Figure 3. 4.1. Reducing agent excess As shown in Figure 4, there is a significant difference between the excess of reducing agent needed to achieve the same NOX reduction efficiency for the SCR and SNCR denitrification methods. As can be seen from Figure 4, the experiments confirmed that for SCR, a lower excess of the reducing agent is needed to reach the same NOX level after the secondary denitrification method because the catalyst reduces the activation energy of the chemical process. According to [24], the reducing agent stoichiometry for SNCR has the optimum values between 1.5–2.5. Experiments showed the highest efficiency values when using NSR around 2.5, but it is always necessary to monitor the ammonia slip in the flue gas, which is sub- ject to the BAT-associated emission level. According to BAT, the level of NH3 emissions into the air from 345 J. Jeníková, K. Michaliková, F. Hrdlička et al. Acta Polytechnica Measu- rement no. Tempe- rature NOX before SNCR NOX after SNCR NOX conversion NSR NH3 slip [◦C] [mg/Nm3] [mg/Nm3] [%] [mol/mol] [mg/Nm3] 1 959 321 250 22.1 1.00 no t m ea su re d 2 238 26.0 2.10 3 228 29.0 3.10 4 209 35.1 4.10 5 883 459 392 14.7 0.55 6 303 34.1 1.10 7 267 41.9 1.65 8 257 43.9 2.20 9 242 47.2 2.64 10 886 408 356 12.9 0.62 11 293 28.2 1.30 12 249 39.0 1.89 13 179 56.3 2.50 14 170 58.4 3.00 15 868 424 388 8.7 0.43 0.3 16 343 19.2 0.87 0.2 17 306 28.0 1.30 0.1 18 284 33.0 1.70 0.2 19 269 36.7 3.47 0.4 20 948 461 450 2.3 0.57 0.0 21 437 5.2 1.12 0.0 22 426 7.7 1.70 0.0 23 393 14.7 2.70 0.8 24 925 498 458 8.0 0.41 0.1 25 450 9.5 0.81 0.0 26 410 17.7 1.23 0.0 27 364 26.9 1.97 0.3 Measu- rement no. Tempe- rature NOX before SCR NOX after SCR NOX conversion NSR NH3 slip [◦C] [mg/Nm3] [mg/Nm3] [%] [mol/mol] [mg/Nm3] 1 260 692 544 21.0 0.29 0 2 705 386 45.0 0.59 0 3 698 305 56.0 0.75 0 4 593 132 77.6 0.91 0 5 533 31 94.1 1.17 0.1 6 613 12 98.0 1.29 0.9 7 290 518 168 67.4 0.81 0.2 8 519 158 71.1 0.83 0.3 9 504 79 85.2 0.99 0.2 10 528 124 76.7 1.04 0.2 11 487 47 90.4 1.09 0.2 12 477 18 96.2 1.28 0.3 13 486 36 93.0 1.35 0.0 14 300 520 144 72.4 1.00 0.1 15 531 75 85.9 1.21 0.1 16 519 40 92.3 1.43 0.1 17 539 31 94.1 1.60 0.1 Table 3. Summarization of the experimental results for SCR and SNCR. 346 vol. 62 no. 3/2022 Applicability of secondary denitrification measures on a fluidized bed boiler the use of SCR and/or SNCR is < 3–10 mg/mN3 [1]. In the case of plants that burn biomass and operate at variable loads, the upper end of the BAT-AEL range is 15 mg/mN3 [1]. Some of the unreacted am- monia is converted to ammonia salts and bound to fly ash, which then exhibits an unwanted odour and could become unapplicable in future use. In addition, leachable ammonia salts can restrict its application as well. Stricter requirements are therefore required by fly ash buyers who use it in the construction industry. Therefore, large combustion sources require values as low as 7 mg/mN3 for NH3 in the flue gas after ESP, although ammonia can be smelled already at values of 5 mg/mN3. For these reasons, the feeding of higher amounts of the reducing agent and thus a higher NSR is not desirable. The NOX reduction efficiency of more than 90 % can be reached with the NSR greater than 1.1 for the SCR method. Further increase in the reducing agent feed is not desirable because the catalyst would be supersat- urated and excess unreacted ammonia would escape along with the flue gas, causing the above-mentioned problems with the ash utilisation. In general, the SCNR method requires higher doses of the reducing agent for the same NOX level required in the flue gas. 4.2. SCR and SNCR efficiency The NOX reduction efficiency varied between 2 % and 58 % for the SNCR method and between 21 % and 98 % for the SCR method. The lower value corre- sponds to the lowest injection rate of the reducing agent for both methods. For the SNCR method, the best results were achieved for temperatures between 880 and 890 ◦C and NSR between 2.2 and 3.0 when efficiency reached 44 to 58 %. For the SCR method, efficiencies higher than 90 % were reached for all cat- alyst temperatures, while the NSR’s were between 1.1 and 1.6. The results agree with the BAT con- clusions as stated in Section 1. From Table 2, it can be seen that the primary measures reduce the input NOX concentrations for the SNCR method to values between 321 and 498 mg/Nm3. The primary measures on this boiler are used mainly to increase the temperature in the freeboard section and thus to ensure that the optimum temperatures for NOX reduction are reached. The SCR method was tested with flue gas recirculation (in order not to lower NOX emission but to keep the combustion process stable) and the initial NOX concentrations ranged between 477 and 705 mg/Nm3. The nitrogen emissions con- centrations after the denitrification measures were between 170 and 458 mg/Nm3 for the SNCR method and between 12 and 544 mg/Nm3 for the SCR method (the highest value is for an insufficiently low NSR of 0.3). 4.3. SCR and SNCR efficiency correlation with temperature There is a clear effect of temperature on the SNCR method at the feeding point of the reducing agent. In Figure 5, it can be seen that higher efficiencies are achieved at temperatures up to about 900 ◦C, and with increasing temperature, the efficiency decreases at the same NSRs. This corresponds to the theoretical knowledge as mentioned in Section 2.2.2, where the optimal temperature window for urea is said to be between 850 and 1000 ◦C. In contrast, in the case of the SCR method, the correlation with temperature is minimal and the effi- ciency basically depends only on the NSR, as shown in Figure 6. 0 10 20 30 40 50 60 70 80 90 100 0.0 1.0 2.0 3.0 4.0 5.0 øη [ % ] NSR [-] 868 °C 883 °C 886 °C 925 °C 948 °C 959 °C Figure 5. Correlation of SNCR efficiency and tem- perature. 0 10 20 30 40 50 60 70 80 90 100 0.0 0.5 1.0 1.5 2.0 ø η [ % ] NSR [-] 259 °C 261 °C 289 °C 290 °C 299 °C Figure 6. Correlation of SCR efficiency and temper- ature. 4.4. SCR and SNCR comparison for the same NOX input concentrations A specific comparison was made for the same NOX input concentrations. The inlet nitrogen oxides con- centration was kept at 500 mg/Nm3, and no primary measures were used with the purpose of lowering ni- trogen oxides levels. Different NSRs were used and the trend of NOX reduction according to the NSR can be seen in Figure 7. 347 J. Jeníková, K. Michaliková, F. Hrdlička et al. Acta Polytechnica 0 100 200 300 400 500 600 0 0.5 1 1.5 2 N O X [m g/ N m 3 ] NSR [-] SCR SNCR Figure 7. Comparison of the SCR and SNCR meth- ods used within the same BFB boiler. The maximum NSR for the SNCR method was 1.97. From other experimental results, we can assume that with a greater injection of the reducing agent and a lower temperature of the fluidized bed, lower nitrogen emissions could be achieved. In general, from the results it can be seen that a lower excess of the reducing agent is needed for the SCR method. The decision of whether to use the SCR or SNCR denitrification method depends on the final required level of nitrogen oxides and on the consideration of investment and operating costs. 5. Conclusions The experimental results show the denitrification pos- sibilities applied on the fluidized bed boiler with a thermal output of 500 kW. The size of the experi- mental equipment is the biggest benefit of performed experiments. The combustion of various fuel types and the generation of emissions have already been in- vestigated, but those are mainly experimental reactors with a diameter of 100–150 mm and laboratory-made flue gas mixtures [16, 25]. Initial NOX concentrations in the experimental boiler with lignite combustion range from 321 mg/Nm3 to 705 mg/Nm3. The correlation of the denitrification efficiency and the NSR was investigated. It has been found that the SCR needs a lower NSR than the SNCR method to reach the same efficiency. The lower need for the reducing agent corresponds to the higher efficiency of the method because the catalyst reduces the activation energy of the reaction. In particular, NSR up to 1.6 was used for the SCR method. With higher NSRs, the ammonia slip could become too high and the ash could be degraded in the practical use of High-Dust catalysts. In contrast, the SNCR method has a higher reducing agent consumption and the best denitrification results were achieved for NSR around 2.5. For the efficiency of both denitrification methods, the results are as follows. The SNCR method, together with primary measures (flue gas recirculation and air staging), reaches the efficiency of 58 % and the efficiency of the SCR method exceeds 98 %. Acknowledgements This work was supported by the project from Re- search Center for Low-Carbon Energy Technologies, CZ.02.1.01/0.0/0.0/16_019/0000753 which is gratefully acknowledged. List of symbols BAT best available techniques CTU Czech Technical University FT-IR Fourier transform infrared spectroscopy LCP large combustion plants MoO3 molybdenum trioxide NDIR non-dispersive infrared NH3 ammonia NSR normalized stoichiometric ratio NO nitric oxide NOX oxide emissions NO2 nitrogen dioxide N2O nitrous oxide SCR selective catalytic reduction SNCR selective non-catalytic reduction TiO2 titanium dioxide V2O5 vanadium pentoxide WO3 tungsten trioxide References [1] T. Lecomte, J. Ferrería de la Fuente, F. Neuwahl, et al. Best Available Techniques (BAT) reference document for large combustion plants. EUR 28836 EN. Publications Office of the European Union, 2017. https://doi.org/10.2760/949. [2] J. Vejvoda, P. Machač, P. Buryan. Technologie ochrany ovzduší a čištění odpadních plynů. University of Chemistry and Technology, 2003. ISBN 80-708-0517-X. [3] F. Normann, K. Andersson, B. Leckner, F. Johnsson. Emission control of nitrogen oxides in the oxy-fuel process. Progress in Energy and Combustion Science 35(5):385–397, 2009. https://doi.org/10.1016/j.pecs.2009.04.002. [4] C. T. Bowman. Control of combustion-generated nitrogen oxide emissions: Technology driven by regulation. Symposium (International) on Combustion 24(1):859–878, 1992. https://doi.org/10.1016/S0082-0784(06)80104-9. [5] J. Hemerka, P. Vybíral. Ochrana ovzduší. Czech Technical University in Prague, 2010. ISBN 978-80-01-04646-3. [6] P. Machač, E. Baraj. A simplified simulation of the reaction mechanism of NOx formation and non-catalytic reduction. Combustion Science and Technology 190(6):967–982, 2018. https://doi.org/10.1080/00102202.2017.1418335. [7] P. Forzatti. Present status and perspectives in de-NOx SCR catalysis. Applied Catalysis A: General 222(1-2):221–236, 2001. https://doi.org/10.1016/S0926-860X(01)00832-8. 348 https://doi.org/10.2760/949 https://doi.org/10.1016/j.pecs.2009.04.002 https://doi.org/10.1016/S0082-0784(06)80104-9 https://doi.org/10.1080/00102202.2017.1418335 https://doi.org/10.1016/S0926-860X(01)00832-8 vol. 62 no. 3/2022 Applicability of secondary denitrification measures on a fluidized bed boiler [8] K. El Sheikh, M. J. H. Khan, M. Diana Hamid, et al. Advances in reduction of NOx and N2O1 emission formation in an oxy-fired fluidized bed boiler. Chinese Journal of Chemical Engineering 27(2):426–443, 2019. https://doi.org/10.1016/j.cjche.2018.06.033. [9] IPCC, 2007. Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team and Pachauri, R.K. and Reisinger, A. (eds.)]. IPCC, Geneva, Switzerland, 2007. ISBN 92-9169-122-4. [10] L. E. Aamand, B. Leckner, S. Andersson. Formation of nitrous oxide in circulating fluidized-bed boilers. Energy & Fuels 5(6):815–823, 1991. https://doi.org/10.1021/ef00030a008. [11] M. de las Obras-Loscertales, T. Mendiara, A. Rufas, et al. NO and N2O emissions in oxy-fuel combustion of coal in a bubbling fluidized bed combustor. Fuel 150:146–153, 2015. https://doi.org/10.1016/j.fuel.2015.02.023. [12] M. Vodička, J. Hrdlička, P. Skopec. Experimental study of the NOX reduction through the staged oxygen supply in the oxy-fuel combustion in a 30 kWth bubbling fluidized bed. Fuel 286:119343, 2021. https://doi.org/10.1016/j.fuel.2020.119343. [13] C. Lupiáñez, L. I. Díez, L. M. Romeo. Influence of gas-staging on pollutant emissions from fluidized bed oxy-firing. Chemical Engineering Journal 256:380–389, 2014. https://doi.org/10.1016/j.cej.2014.07.011. [14] M. de las Obras-Loscertales, A. Rufas, L. de Diego, et al. Effects of temperature and flue gas recycle on the SO2 and NOx emissions in an oxy-fuel fluidized bed combustor. Energy Procedia 37:1275–1282, 2013. https://doi.org/10.1016/j.egypro.2013.06.002. [15] T. Czakiert, Z. Bis, W. Muskala, W. Nowak. Fuel conversion from oxy-fuel combustion in a circulating fluidized bed. Fuel Processing Technology 87(6):531–538, 2006. https://doi.org/10.1016/j.fuproc.2005.12.003. [16] W. Moroń, W. Rybak. NOx and SO2 emissions of coals, biomass and their blends under different oxy-fuel atmospheres. Atmospheric Environment 116:65–71, 2015. https://doi.org/10.1016/j.atmosenv.2015.06.013. [17] X. Cheng, X. T. Bi. A review of recent advances in selective catalytic NOx reduction reactor technologies. Particuology 16:1–18, 2014. https://doi.org/10.1016/j.partic.2014.01.006. [18] F. Gholami, M. Tomas, Z. Gholami, M. Vakili. Technologies for the nitrogen oxides reduction from flue gas: A review. Science of The Total Environment 714:136712, 2020. https://doi.org/10.1016/j.scitotenv.2020.136712. [19] Y. GAO, T. LUAN, T. LÜ, et al. Performance of V2O5 WO3 MoO3/TiO2 catalyst for selective catalytic reduction of NOx by NH3. Chinese Journal of Chemical Engineering 21(1):1–7, 2013. https://doi.org/10.1016/S1004-9541(13)60434-6. [20] L. Olsson, H. Sjövall, R. J. Blint. A kinetic model for ammonia selective catalytic reduction over Cu ZSM 5. Applied Catalysis B: Environmental 81(3-4):203–217, 2008. https://doi.org/10.1016/j.apcatb.2007.12.011. [21] J. Hrdlicka, P. Skopec, F. Hrdlicka. Trough air distributor for a bubbling fluidized bed boiler with isobaric nozzles. In Proceedings of the 22nd International Conference on Fluidized Bed Conversion, vol. 1. Turku, Finland, 2015. ISBN 978-952-12-3222-0. [22] P. Skopec, J. Hrdlička, J. Opatřil, J. Štefanica. NOX emissions from bubbling fluidized bed combustion of lignite coal. Acta Polytechnica 55(4):275–281, 2015. https://doi.org/10.14311/AP.2015.55.0275. [23] Prováděcí rozhodnutí komise (EU) 2021/2326. Úřední věstník Evropské unie L 469:1–81, 2021. https://eur- lex.europa.eu/legal-content/CS/TXT/PDF/?uri= CELEX:32021D2326&from=EN. [24] M. Mladenović, M. Paprika, A. Marinković. Denitrification techniques for biomass combustion. Renewable and Sustainable Energy Reviews 82:3350–3364, 2018. https://doi.org/10.1016/j.rser.2017.10.054. [25] Y. Hu, S. Naito, N. Kobayashi, M. Hasatani. CO2, NOx and SO2 emissions from the combustion of coal with high oxygen concentration gases. Fuel 79(15):1925–1932, 2000. https://doi.org/10.1016/S0016-2361(00)00047-8. 349 https://doi.org/10.1016/j.cjche.2018.06.033 https://doi.org/10.1021/ef00030a008 https://doi.org/10.1016/j.fuel.2015.02.023 https://doi.org/10.1016/j.fuel.2020.119343 https://doi.org/10.1016/j.cej.2014.07.011 https://doi.org/10.1016/j.egypro.2013.06.002 https://doi.org/10.1016/j.fuproc.2005.12.003 https://doi.org/10.1016/j.atmosenv.2015.06.013 https://doi.org/10.1016/j.partic.2014.01.006 https://doi.org/10.1016/j.scitotenv.2020.136712 https://doi.org/10.1016/S1004-9541(13)60434-6 https://doi.org/10.1016/j.apcatb.2007.12.011 https://doi.org/10.14311/AP.2015.55.0275 https://eur-lex.europa.eu/legal-content/CS/TXT/PDF/?uri=CELEX:32021D2326&from=EN https://eur-lex.europa.eu/legal-content/CS/TXT/PDF/?uri=CELEX:32021D2326&from=EN https://eur-lex.europa.eu/legal-content/CS/TXT/PDF/?uri=CELEX:32021D2326&from=EN https://doi.org/10.1016/j.rser.2017.10.054 https://doi.org/10.1016/S0016-2361(00)00047-8 J. Jeníková, K. Michaliková, F. Hrdlička et al. Acta Polytechnica A. Appendix SNCR Measu- rement no. Tempe- rature NOX before SNCR NOX after SNCR NOX conversion NSR O2 NH3 slip [◦C] [mg/Nm3] [mg/Nm3] [%] [mol/mol] [%] [mg/Nm3] 1 959 ± 5 321 ± 31 250 ± 12 22.1 ± 2.3 1.00 5.0 ± 0.2 no t m ea su re d 2 238 ± 9 26.0 ± 1.0 2.10 6.5 ± 0.7 3 228 ± 14 29.0 ± 0.8 3.10 7.1 ± 0.2 4 209 ± 9 35.1 ± 1.7 4.10 6.6 ± 0.2 5 883 ± 10 459 ± 29 392 ± 33 14.7 ± 7.2 0.55 6.9 ± 0.6 6 303 ± 14 34.1 ± 3.0 1.10 5.9 ± 0.3 7 267 ± 15 41.9 ± 3.3 1.65 5.7 ± 0.3 8 257 ± 14 43.9 ± 3.4 2.20 5.7 ± 0.3 9 242 ± 10 47.2 ± 2.3 2.64 5.9 ± 0.3 10 886 ± 8 408 ± 22 356 ± 25 12.9 ± 3.3 0.62 5.3 ± 0.4 11 293 ± 29 28.2 ± 3.6 1.30 4.6 ± 0.4 12 249 ± 23 39.0 ± 3.1 1.89 4.5 ± 0.4 13 179 ± 37 56.3 ± 3.9 2.50 4.7 ± 0.7 14 170 ± 11 58.4 ± 0.6 3.00 3.6 ± 0.3 15 868 ± 20 424 ± 33 388 ± 25 8.7 ± 3.6 0.43 6.5 ± 0.8 0.3 ± 0.1 16 343 ± 19 19.2 ± 1.7 0.87 5.8 ± 0.5 0.2 ± 0.1 17 306 ± 15 28.0 ± 1.1 1.30 5.5 ± 0.6 0.1 ± 0.1 18 284 ± 13 33.0 ± 1.2 1.70 5.1 ± 0.4 0.2 ± 0.1 19 269 ± 22 36.7 ± 2.0 3.47 5.5 ± 0.5 0.4 ± 0.2 20 948 ± 29 461 ± 7 450 ± 22 2.3 ± 3.4 0.57 6.4 ± 0.5 0.0 ± 0.1 21 437 ± 38 5.2 ± 1.2 1.12 5.4 ± 1.8 0.0 22 426 ± 27 7.7 ± 2.1 1.70 4.3 ± 0.3 0.0 23 393 ± 12 14.7 ± 0.8 2.70 4.4 ± 0.4 0.8 ± 0.3 24 925 ± 11 498 ± 8 458 ± 12 8.0 ± 2.6 0.41 7.0 ± 0.3 0.1 ± 0.1 25 450 ± 8 9.5 ± 1.5 0.81 7.2 ± 0.3 0.0 26 410 ± 12 17.7 ± 2.6 1.23 7.2 ± 0.2 0.0 27 364 ± 43 26.9 ± 8.8 1.97 7.3 ± 0.8 0.3 ± 0.2 Table 4. Experimental results for SNCR. 350 vol. 62 no. 3/2022 Applicability of secondary denitrification measures on a fluidized bed boiler SCR Measu- rement no. Tempe- rature NOX before SCR NOX after SCR NOX conversion NSR O2 NH3 slip [◦C] [mg/Nm3] [mg/Nm3] [%] [mol/mol] [%] [mg/Nm3] 1 260 ± 2 692 ± 74 544 ± 58 21.0 ± 9.9 0.29 ± 0.02 10.1 ± 0.5 0 2 705 ± 60 386 ± 46 45.0 ± 7.3 0.59 ± 0.03 10.5 ± 0.9 0 3 698 ± 50 305 ± 38 56.0 ± 6.5 0.75 ± 0.03 10.6 ± 0.7 0 4 593 ± 54 132 ± 20 77.6 ± 3.0 0.91 ± 0.07 9.8 ± 0.7 0 5 533 ± 70 31 ± 15 94.1 ± 2.5 1.17 ± 0.12 9.7 ± 0.8 0.1 ± 0.2 6 613 ± 73 12 ± 4 98.0 ± 0.9 1.29 ± 0.09 10.5 ± 0.9 0.9 ± 0.2 7 290 ± 1 518 ± 40 168 ± 24 67.4 ± 4.3 0.81 ± 0.06 9.8 ± 0.6 0.2 ± 0.1 8 519 ± 19 158 ± 11 71.1 ± 1.8 0.83 ± 0.04 9.5 ± 0.5 0.3 ± 0.1 9 504 ± 19 79 ± 9 85.2 ± 1.7 0.99 ± 0.05 9.3 ± 0.5 0.2 ± 0.1 10 528 ± 45 124 ± 24 76.7 ± 3.6 1.04 ± 0.15 10.7 ± 1.4 0.2 ± 0.1 11 487 ± 28 47 ± 9 90.4 ± 1.9 1.09 ± 0.06 8.4 ± 0.4 0.2 ± 0.1 12 477 ± 46 18 ± 7 96.2 ± 1.4 1.28 ± 0.10 8.4 ± 0.6 0.3 ± 0.1 13 486 ± 18 36 ± 8 93.0 ± 1.6 1.35 ± 0.05 9.0 ± 0.5 0.0 ± 0.1 14 300 ± 1 520 ± 37 144 ± 17 0.4 ± 2.4 1.00 ± 0.06 9.2 ± 0.5 0.1 ± 0.1 15 531 ± 15 75 ± 5 85.9 ± 1.0 1.21 ± 0.06 9.4 ± 0.2 0.1 ± 0.1 16 519 ± 15 40 ± 7 92.3 ± 1.2 1.43 ± 0.07 9.3 ± 0.4 0.1 ± 0.1 17 539 ± 30 31 ± 6 94.1 ± 1.2 1.60 ± 0.09 9.5 ± 0.4 0.1 ± 0.1 Table 5. Experimental results for SCR. 351 Acta Polytechnica 62(3):341–351, 2022 1 Introduction 2 Nitrogen oxides 2.1 Formation of NOX emissions 2.2 Denitrification methods 2.2.1 Primary measures 2.2.2 Secondary measures 3 Experimental set-up 3.1 Experimental facility 3.2 Fuel and reducing agents 3.3 Methods 4 Results and discussion 4.1 Reducing agent excess 4.2 SCR and SNCR efficiency 4.3 SCR and SNCR efficiency correlation with temperature 4.4 SCR and SNCR comparison for the same NOX input concentrations 5 Conclusions Acknowledgements List of symbols References A Appendix