KINETIC MECHANISM INVESTIGATION ON NO x REMOVAL WITH HYDRAZINE HYDRATE AT MODERATE TO HIGH TEMPERATURES HONG Liu 1 , Yin Lijie 1 , CHEN De-zhen 1 , WANG Du 2 1.Thermal & Environmental Engineering Institute, Tongji University, Shanghai, China; 2. Institute of Energy & Environment Engineering, Shanghai University of Electric Power, Shanghai, China Introduction Selective non-catalytic reduction (SNCR) De-NO x technology has been used to reduce nitrogen oxides (NO x ) emission from flue gas for nearly 30 years [1] . In the SNCR process, gaseous or solution reagent (ammonia, urea etc.) is injected into the hot exhaust gases of stationary combustion systems in the presence of certain level of oxygen, initiating a sequence of reactions that convert NO x to molecular nitrogen (N 2 ) [2] . But the SNCR De-NO x process works only in a narrow temperature window roughly between 1100K and 1400K. In practice combustion systems, NO x reduction is often desirable at lower temperatures [3] . Hydrazine hydrate (N 2 H 4 ·H 2 O) is a kind of intensive reducer, which is widely used as rocket propellant, pesticide material and deoxidizer of boiler etc. The N 2 H 4 -NO-O 2 reaction has been studied by Azuhata et al. on a laboratory scale, the result showed that NO was reduced to N 2 and H 2 O by N 2 H 4 in the temperature range of 773-873K and the presence of O 2 prohibited the reduction of NO x [4] . J. B. Lee et al. [5] tested the NO x reduction with hydrazine in a pilot-scale reactor, and they founded that hydrazine had an effective de-NO x effect in temperature range of 800-950K, which was about 300K lower than that of ammonia. However, little information on complete kinetics mechanism of hydrazine-based NO x reduction is reported till now. Zhang et al. [6, 7] proposed the rough kinetics mechanism of NO x reduction with hydrazine and founded that the effective temperature range is 763-928K by experiment, but the NO reduction predicted by calculation is much greater than that of measured by experiments. In 1975, Lyon et al. first discovered SNCR process and applied for the relevant patent [8, 9] ; subsequently, many detailed kinetics models were presented. In 1989, a so-called Miller & Bowman (M&B) model was summarized, which includes 53 species of reactants and 251 reactions, covering the reactions between NH 3 and NO in SNCR process and the NO formation reactions in combustion of hydrocarbons. M&B model was later modified by Miller and Glarborg in 1999 [10] to describe the whole SNCR process between NH 3 and NO more precisely. However the kinetics mechanism of reactions of hydrazine and NO should be different. Konnov et al. [11] proposed a decomposition mechanism of hydrazine in inert gases, including 11 species and 51 elementary reactions. Reaction mechanisms of hydrazine decomposition and oxidation have also been proposed [12, 13] , however most of the rate coefficients in the model were approximately estimated. A more detailed kinetics mechanism for hydrazine-based SNCR process was proposed and verified by comparison with the experimental results in our previous researches [14] , and the model provided 24 species and 113 elementary reactions, matching well with experimental data in respect of NO reduction trends versus temperature. But there is distinct gap between the computational results and the experimental data,
13
Embed
KINETIC MECHANISM INVESTIGATION ON NOx … · KINETIC MECHANISM INVESTIGATION ON NO x ... Figure 2. de-NO efficiency of SNCR experiments by hydrazine Simulation ... and the mixing
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
KINETIC MECHANISM INVESTIGATION ON NOx REMOVAL WITH
HYDRAZINE HYDRATE AT MODERATE TO HIGH TEMPERATURES
HONG Liu1, Yin Lijie
1, CHEN De-zhen
1, WANG Du
2
1.Thermal & Environmental Engineering Institute, Tongji University, Shanghai, China;
2. Institute of Energy & Environment Engineering, Shanghai University of Electric Power, Shanghai,
China
Introduction
Selective non-catalytic reduction (SNCR) De-NOx technology has been used to reduce
nitrogen oxides (NOx) emission from flue gas for nearly 30 years [1]
. In the SNCR process, gaseous or
solution reagent (ammonia, urea etc.) is injected into the hot exhaust gases of stationary combustion
systems in the presence of certain level of oxygen, initiating a sequence of reactions that convert NOx
to molecular nitrogen (N2) [2]
. But the SNCR De-NOx process works only in a narrow temperature
window roughly between 1100K and 1400K. In practice combustion systems, NOx reduction is often
desirable at lower temperatures [3]
. Hydrazine hydrate (N2H4·H2O) is a kind of intensive reducer,
which is widely used as rocket propellant, pesticide material and deoxidizer of boiler etc. The
N2H4-NO-O2 reaction has been studied by Azuhata et al. on a laboratory scale, the result showed that
NO was reduced to N2 and H2O by N2H4 in the temperature range of 773-873K and the presence of
O2 prohibited the reduction of NOx [4]
. J. B. Lee et al. [5]
tested the NOx reduction with hydrazine in a
pilot-scale reactor, and they founded that hydrazine had an effective de-NOx effect in temperature
range of 800-950K, which was about 300K lower than that of ammonia. However, little information
on complete kinetics mechanism of hydrazine-based NOx reduction is reported till now. Zhang et al.[6,
7] proposed the rough kinetics mechanism of NOx reduction with hydrazine and founded that the
effective temperature range is 763-928K by experiment, but the NO reduction predicted by
calculation is much greater than that of measured by experiments.
In 1975, Lyon et al. first discovered SNCR process and applied for the relevant patent [8, 9]
;
subsequently, many detailed kinetics models were presented. In 1989, a so-called Miller & Bowman
(M&B) model was summarized, which includes 53 species of reactants and 251 reactions, covering
the reactions between NH3 and NO in SNCR process and the NO formation reactions in combustion
of hydrocarbons. M&B model was later modified by Miller and Glarborg in 1999 [10]
to describe the
whole SNCR process between NH3 and NO more precisely. However the kinetics mechanism of
reactions of hydrazine and NO should be different. Konnov et al. [11]
proposed a decomposition
mechanism of hydrazine in inert gases, including 11 species and 51 elementary reactions. Reaction
mechanisms of hydrazine decomposition and oxidation have also been proposed [12, 13]
, however most
of the rate coefficients in the model were approximately estimated. A more detailed kinetics
mechanism for hydrazine-based SNCR process was proposed and verified by comparison with the
experimental results in our previous researches
[14], and the model provided 24 species and 113
elementary reactions, matching well with experimental data in respect of NO reduction trends versus
temperature. But there is distinct gap between the computational results and the experimental data,
which was caused by the fact that the operation condition of the experimental reactor was far from
perfectly stirred reactor (PSR) that was adopted in the simulation model.
In this study, a revised kinetic model for the N2H4-NO-O2 reaction is proposed to verify the
kinetics model, the computational results were compared with the experimental results; and the
practical operation conditions of the experimental reactor was involved in the computational
simulation process.
Experiment
Experiment set-up
The experimental system of NOx reduction by N2H4·H2O is shown in Fig 1. The reactor wall is
insulated to keep a nearly stable temperature distribution in the reactor. The total flow rate of the flue
gas varies from 62.8 to 98.2 Nm3h
-1. N2H4·H2O solution is sprayed into the stream of combustion
products through two atomizing nozzles installed in a pilot-scale pipe flow reactor. In order to
measure the dosage of de-NOx reagent, normalized stoichiometric ratio (NSR) is defined as:
x
x
NOandreagentbetweenratiomoletricstoichiome
NOandreagentbetweenratiomoleactualNSR (1)
Three thermocouples are installed at 1m length intervals along the reactor axis to measure the
temperature of the flue gas in the reactor. The flue gas contained about 150ppm of CO, 10% of H2O
and 3% of CO2. N2 is added into the chamber of the furnace in order to regulate the O2 pressure of
the flue gas, O2 content of the flue gas descends with the increase of temperature and varies from
9.8% at 1271K to 15.8% around 780K. In this temperature range, more than 90% of NOx exists in
the form of NO [17]
, so NO was used to replace NOx here. The inlet concentration of NO varies from
320 to 580ppmv and increases with the increase of temperature. During the experiments, the NSR
value varies from 1.5 to 4.0 considering the incomplete contact between flue gas and de-NOx reagent.
A KM9106 type gas analyzer (Kane, UK) is used to measure concentrations of NO, NO2 and O2 etc.
A PGM-7800 multiple gas analyzer (RAE, USA) is used to monitor concentration of NH3 to check
the “ammonia slip” and a PGM-7240 VOC analyzer (RAE, USA) is used to monitor N2H4 slip.
Figure 1. Schematic diagram of SNCR experimental set-up
Experiment results and discussion
The de-NO efficiency η is calculated by
in
out
NO
NO
][
][1 (2)
Where [NO]in and [NO]out are volume fractions of NO at inlet and outlet respectively.
Fig 2 shows the de-NO efficiency versus temperatures when NSR=4.0. With the increase of
temperature, the de-NO efficiency represents bimodal characteristics, where the lower temperature
window is from 820K to 960K, and the higher temperature window is from 1260K to 1330K. Fig 2
also shows the influence of NO concentration on the de-NO efficiency, it’s found that when the NO
concentration increases from 360ppmv to 811ppmv, the temperature windows changes little. This
means that the NO concentration has little effect on the de-NO efficiency.