CORDIS proporciona enlaces a los documentos públicos y las publicaciones de los proyectos de los programas marco HORIZONTE.
Los enlaces a los documentos y las publicaciones de los proyectos del Séptimo Programa Marco, así como los enlaces a algunos tipos de resultados específicos, como conjuntos de datos y «software», se obtienen dinámicamente de OpenAIRE .
Resultado final
The decrease in the conversion to NO with increasing fuel-nitrogen contents indicated that the increased presence of nitrogenous species (e.g. NH3, HCN) acts to reduce the NO generated by oxidation of these same compounds. This behavior can be compared with the thermal DeNOx or SNR process where NH3 is added to the exhaust gas after the combustion chamber reducing NO to N2. NH3 and HCN were found for the first time directly emitted from a burning single wood particle. They are important precursors during wood combustion and should be considered in NOx reduction strategies. The HCN/NH3 - ratios seemed to depend on the H/N - ratios of the fuels, which also correlate with the O/N - ratios. The effects of oxygen and temperature on NH3, HCN, NO, and N2O could be explained with the reaction mechanism.
Generally, similar trends were seen if the oxygen partial pressure or the temperature was increased. Higher oxygen partial pressures led to higher combustion rates and higher concentrations of radicals. Higher temperatures gave a similar effect. A maximum conversion to NO was found between 10 and 15 kPa or around 800 °C. This behaviour was explained with the temperature window for NO reduction (refer to thermal DeNOx or SNCR process) with the onset at 900 °C depending on oxygen partial pressure and concentration of radicals. N2O was formed after the extinction of the flame indicating that N2O is rapidly destroyed mainly by the reaction:N(2)O + H ® N(2) (+) OH. The concentration of H-radicals should be especially high during the combustion of the hydrogen-rich biofuels. The thermal destruction mechanism seemed to be of minor importance. The fast reaction and the usually low nitrogen contents and the relatively high NH3 levels (compared to coals) explained the very low N2O concentration levels observed during biomass combustion in large-scale units. The experimental results showed that H2 is the least effective in enhancing NO formation through HCN and NH3 oxidation. It is likely that H2 can be an effective additive for decreasing NO formation. HCN oxidation promoted NO and N2O formation whereas NH3 oxidation formed NO and almost no N2O. Significant reduction of NO was observed when NO was present during the oxidation of HCN and NH3. This is because the presence of NO at lower temperatures favours the destruction routes of NO such as (NCO + NO ® N2O +CO) and (NO + NH2 ® N2 + H2O).
With the modelling work the experimental results of the single particle NO formation rate studies as well as the gas-phase chemistry were studied. The model is based on a detailed mechanism considering radical chemistry. The principle NO and N2O formation paths were evaluated and calculated. NO, N2O, NO2, NH3, HCN, CH4, CO, CO2 were measured at different locations in a 200 kW grate furnace burning chipboard and spruce wood. Additionally, NH3, CH4, and H2 were added at different feeding points. Almost no CO, CH4, N2O, NO2, NH3, and HCN were found at the beginning of the convection pass or at the exit. NO is not significantly reduced in the hot convection pass. Its final concentration is determined in the combustion chamber. Shortly above the glow bed very high concentrations of CH4 (16 - 25 v-%) and to a lesser extent CO (1 v-%) were found locally depending on fuel and operating conditions. In the case of the chipboard high concentrations of NH3 (3800 ppm) were found. HCN was significantly lower (170 ppm). However, in the case of the spruce wood both species NH3 and HCN were almost zero due to the lower secondary to primary air ratio and the much lower nitrogen content. The NO concentrations are at its highest level (70 ppm). With the addition of NH3 before the heat exchanger, NO was reduced dramatically (from 240 ppm to 35 ppm).
However, the degree of reduction was very sensitive to the location where NH3 was added as well as the NH3 - slip. N2O increased only insignificantly (from 5 to 7 ppm). NH3 addition has proven to be a powerful tool to reduce NO. The addition of CH4 and H2 did not lead to a significant reduction of NO.
Results of experiments with a NOx reduced moving grate furnace and a NOx reduced understoker furnace
Air staging and fuel staging have been successfully applied as primary measures for NOx reduction in wood furnaces. For high nitrogen content in the fuel and thus for high NOX emissions, a higher reduction rate is achieved than for low nitrogen content.
The main process parameters in the reduction chamber are excess air ratio, temperature, mixing quality and residence time. With air staging at optimum conditions, app. 50% reduction can be achieved for native wood and approximately 75% for chipboard. Optimum conditions are a slightly under stoichiometric excess air ratio in the reduction zone and a reduction temperature of app. 1100 °C 1200 °C with a residence time of app. 0,5s. The average reduction rate in practice can be reduced significantly due to non-ideal conditions and/or varying operating parameters.
To enable an optimum temperature for different fuel humidity, measures are needed to influence the temperature. Flue gas re-circulation and partial heat extraction in the primary chamber have been applied to lower the temperature. The amount of re-circulated flue gas can be adapted in function of the temperature to control the temperature in the reduction zone. Since the highest NOx reduction is achieved at the lowest possible overall excess air, CO/Lambda control with set point optimization was applied to ensure optimum conditions for NOx reduction. Fuel staging shows an even higher potential of NOX reduction than air staging.
Furthermore it seems to be less sensitive to the excess air ratio in the reduction zone and it reaches a significant NOX reduction at temperatures app. 200 °C lower than air staging. Since app. 25% to 35% of the fuel has to be fed as secondary fuel, two feeding systems are needed.
Selective non catalytic reduction (SNCR): Selective non catalytic reduction SNCR has been applied as secondary measure for NOX reduction in wood furnaces using urea as reducing agent. The main process parameters are the molar ratio of reducing agent, the temperature in the reduction chamber, the mixing quality, and the residence time. The conditions for optimum NOx reduction and acceptable amount of side products with standard SNCR technique are:
- The temperature in the SNCR chamber has to be inside a temperature window of app. 80100 °C, i.e. app. 850 °C = T = 930 °C measured by thermocouple in the SNCR chamber. Urea injection has to be stopped for temperatures outside the temperature window. To avoid exceeding temperatures at low excess air ratio, a partial heat extraction before injection of reducing agent is needed.
- The mean residence time in the reduction chamber should be = 1 s and a good mixing between reducing agent and flue gas is needed.
- The molar ratio should be 1.5 = n = 2.0 which demands an accurate denox process control.
- To ensure good results in practical operation, stable combustion conditions for temperature and excess air ratio is needed which demands for an accurate combustion process control. Under optimum conditions, a NOx reduction rate between 50 % (n1) to 80 % (n2) can be achieved with an acceptable amount of side products (sum of NH3, HNCO and N2O < 20 mg/Nm3 at 11 %O2) for a molar ratio n = 2.
If the residence time is too short, the mixing quality insufficient or the temperature too low, significant amounts of side products, i.e. N2O, NH3 and HNCO are emitted, while no significant increase of HCN emission was found.Selective catalytic reduction (SCR): Selective catalytic reduction SCR with urea has been applied as secondary measure for NOX reduction in a moving grate furnace. The main process parameters for the SCR process are the molar ratio of reducing agent, the catalyst temperature, and the space velocity. The catalyst type has to be selected for the expected gas temperature, while the space velocity influences the maximum NOX reduction and the emission of side products. To reach optimum NOX reduction with low emissions of NH3 and HNCO, the urea injection must be adapted to the NOX reduction performance of the catalyst and the average molar ratio should not exceed the design value (here n = 0.75). With the investigated catalyst at a molar ratio of n = 0.75, a temperature of 330 °C 370 °C and a space velocity of app. 20000 h1 (i.e. after one of two catalyst elements) a NOX reduction of app. 75 % was reached with low emission of side products (NH3, HNCO and N2O < 10 mg/Nm3 at 11 %O2). Further results show, that reduction rates > 90 % can be reached at space velocities between 5000 h1 10000 h1 (design value 8000 h1), while the time of operation during the test runs is not sufficient for a final conclusion of the long term behaviour. The catalyst has a NH3 storage capacity, which decreases with increasing temperature. Therefore higher urea supply or higher NO concentrations in the raw gas are acceptable for a certain period of time (several minutes up to a maximum of app. 1 hour at low temperature). If the molar ratio exceeds the design value for a longer period of time, or the temperature is too low, HNCO and NH3 are emitted in significant concentrations.
Therefore an appropriate control system for the SCR process should be applied to avoid the emission of side products. The measurement of the catalyst temperature, the NO concentration in the clean gas (permanently), the NO in the raw gas (permanently or periodically), and the flue gas flow rate (which is a function of the actual heat production rate) allows an adequate denox control strategy.
Investigations on the influence of the primary air input on the NOx emissions results show, that decreasing of primary air leads to a significant reduction of NOx emissions in the range of 50% of the previous value. The reduced primary air input of course causes lower reaction velocity in the solid phase of the fuel, resp. in the glow bed. Thus, for reaching the same heat power, the retention time of the glow bed must be increased, which leads to a bigger glow bed volume and therefore to a bigger volume of the combustion chamber. Hitherto results led to the expectation, that for reaching 50 % of NOx reduction, approximately doubling of glow bed volume resp. primary combustion chamber volume is necessary. Furthermore, the combustion air needed in total for complete combustion in the desired heat power range has to be put into the furnace by secondary or even tertiary air control in order to avoid CO- or CxHy-emissions. The investigations have been carried out with chipboard chips, chipboard sawdust, and spruce wood chips as fuel. The results show in general, that a reduction of primary air can lead to a significant decreasing of NO emissions. The decreasing is relatively high if fuel with high nitrogen content (chip boards, chipboard sawdust) is used (reduction in the range of 50 %). The decreasing is relatively lower, if fuel with low nitrogen content (spruce wood chips) is used (reduction in the range of 30 %). However the absolute level of NO emissions when burning low nitrogen containing fuel is much lower than when burning high nitrogen containing fuel. Under certain conditions the dependency of NO concentration on the primary air input seems to be inverted if the NO concentration is expressed in terms of ppm instead of mg/Nm3. This could be important for realizing a control system based on a NO measurement: The measurement results (ppm) in this case must be converted into mg/Nm3.
Measurements and model calculations have been carried out on a 31 MWth district heating Boiler in Trollhätan (S). The upper part of the furnace (above the secondary air inlet) was much easier to model than the lower part of the furnace. The model calculation in the upper part of the furnace agrees very well with the measurements in both cases examined. The secondary air jets determine the gas flow up to the upper part of the furnace, and the gas entering this part is well mixed. This creates a suitable situation for the CFD-calculation and this is also reflected in the modelling result.
Experiments with a pot furnace were carried out to simulate combustion on a grate using different fuels with different particle sizes and moistures. The secondary air rate was constant and the primary air rate was varied. The fuel bed was ignited at the top and the ignition front propagated down against the airflow. Peaks in the NO formation were detected.
These peaks were usually attained at the end of the devolatilization stage, when the flame reached the bottom of the bed and just at the end of the combustion, when practically all carbon has been combusted. The NO formation decreased steadily, when a batch of wood pellets was combusted, which is considered to be due to NO reduction in the reaction with the char layer above the flame zone. This char layer is increasing, if the moisture content of the fuel is not high, as the ignition front is progressing down. For other fuels, such a steady decrease of NO formation were not found, but usually NO formation increased steadily or remained at constant level during the devolatilization stage even the char layer, where NO could be reduced, was increasing. Measurements of the bed height, temperatures in the bed and ignition velocities were carried out. NO formation in the horizontal stoker burner could be reduced with the air staging and further more by using flue gas recirculation.
However, the gas residence time in this small grate (nominal maximum power 200 kW) sets the main limit to NO reduction. Significantly greater NO reduction can be obtained with partial load conditions, if the burner is operated so that all the grate area is covered with the fuel, since the primary air rate per grate area can reduced and the gas residence time can be increased. By fuel drying before combustion, the drying zone is reduced and grate area is released, for which lower primary air rate can be used resulting to lower NO. Also less fuel is needed per MJ for dryer fuel, which naturally leads to less NO/MJ. By additional changes in the present configuration (which are not made during the present project) the NO emissions are expected to decrease further-the burner diameter/length (and the grate area) could be increased and lower primary air rate per grate area be used, which reduces NO-as a future option, replacing the primary air with highly preheated re-circulated flue gas in the whole zone where drying and devolatilization takes place is expected to reduce NO.
Some of these changes are not limited to this special construction, but are usable in other grate combustion applications operating in the cross flow mode as well.
A model for the kinetics of pyrolysis including nitrogen species was developed and incorporated to a model simulating the devolatilization of a wood particle. A computer program was developed to calculate the release of volatiles, steam (from moisture) and nitrogen from the particle.
Comparison of model calculations to measurements gave a good agreement. Correlations for the ignition distance at the bed surface and for the velocity of the propagation of the ignition front against the airflow were developed. The release of the volatile nitrogen from the fuel is directly proportional to the ignition velocity, when steady conditions are reached. Simplified NO mechanisms presented in the literature were compared to calculations against the predictions of the comprehensive reaction mechanism. Model for the combustion of wood particles including NO formation on a grate, where fuel bed and air are in cross flow was developed. In order to have NO reduction in the char layer, NO must firstly be formed from the intermediates in the flame zone below the char layer. In a calculation example it was found that major part of the decrease in the total fuel nitrogen (NO, HCN, NH3) flowing through the char layer was due to NO reduction by the char and not by the reaction of NO with the intermediates NH3 and HCN. This was due the short residence time and rapid decrease of the gas temperature above the flame zone from 1300 to 1140 K due to the endothermic gasification of char mainly with H2O.
However, the temperatures can be regulated by the air rate to some extent, which could lead to better conditions for the gas phase reduction of NO. Measurements and modelling show that when the air rate increases, the temperature of the flame zone below the char layer increases reaching a maximum at a specific air rate. Measured maximum temperatures in the fuel bed varied between 900...1630 K depending on the air rate and the moisture content of the fuel.
Buscando datos de OpenAIRE...
Se ha producido un error en la búsqueda de datos de OpenAIRE
No hay resultados disponibles