Mechanisms of the NO 2 Formation in Diesel Engines

Transcription

1 RESEARCH Diesel Engines AUTHORS Mechanisms of the NO 2 Formation in Diesel Engines Dipl.-Ing. Michael Rößler is Research Associate at the Institute of Internal Combustion Engines at the Karlsruhe Institute of Technology (KIT) (Germany). While the formation of nitric oxide in diesel engines has been extensively investigated in the past, the processes of the engine-internal formation of nitrogen dioxide are only hardly known. Although nitrogen dioxide represents only a small proportion of the total nitric oxide emissions, the highly reactive species plays an important role in exhaust gas aftertreatment. The current FVV research work at the KIT aims to develop a basic understanding of the processes and mechanisms in the engine-internal formation of nitrogen dioxide. Prof. Dr. sc. techn. Thomas Koch is Head of the Institute of Internal Combustion Engines at the Karlsruhe Institute of Technology (KIT) (Germany). Dipl.-Chem. Corina Janzer is Research Associate at the Department of Molecular Physical Chemistry at the Institute of Physical Chemistry at the Karlsruhe Institute of Technology (KIT) (Germany). Prof. Dr. rer. nat. Matthias Olzmann is Head of the Department of Molecular Physical Chemistry at the Institute of Physical Chemistry at the Karlsruhe Institute of Technology (KIT) (Germany). molekuul.be Fotolia KIT 70

2 1 NITROGEN DIOXIDE IN THE ENGINE EXHAUST 2 ANALYSIS OF ENGINE-INTERNAL FORMATION PROCESSES 3 TEST ENGINE AND BOUNDARY CONDITIONS 4 INFLUENCE OF ENGINE LOAD AND SPEED 5 PARAMETER VARIATIONS 6 FAST GAS SAMPLING 7 KINETIC MODELING OF NO 2 FORMATION 8 SUMMARY 1 NITROGEN DIOXIDE IN THE ENGINE EXHAUST The latest discussions on pollutant emissions from diesel engines and the consequent urban emission load point out, in particular, the relevance of nitrogen dioxide (NO 2 ) emissions. Their harmful effects on humans and the environment are undisputed. In the case of strong solar irradiation, NO 2 is a relevant precursor substance for the formation of ground-level ozone (O 3 ) and thus, it is regarded as the main cause of photochemical smog. Furthermore, the compound of NO 2 with moisture leads to the formation of nitrous acid (HNO 2 ) or directly to the formation of nitric acid (HNO 3 ). On the one hand, these are particularly detrimental to human respiratory organs and, moreover, as acid rain also damage the ecosystem. On the other hand, NO 2 is very useful for exhaust aftertreatment of modern diesel engines. For example, the highly reactive NO 2 promotes soot oxidation and thus contributes significantly to the passive regeneration of diesel particulate filters. An increased proportion of NO 2 is also advantageous in the aftertreatment of nitrogen oxides using SCR technology because the catalytic efficiency rises considerably with increasing NO 2 concentration. NO 2 that is formed in the engine is used up very quickly in the exhaust tract and is usually not emitted into the atmosphere. But NO 2 can also be formed outside the combustion chamber. At an oxidation catalyst, nitric oxide (NO) may react easily with excess oxygen to NO 2. For this reason, the arrangement of the catalytic components in Engine data single-cylinder OM 936 Stroke bore [mm mm] Displacement [l] Compression ratio [-] 17.4 : 1 Maximum torque [Nm] 220 Maximum engine speed [rpm] 2600 Maximum rail pressure [bar] 2400 TABLE 1 Specifications of the single-cylinder test engine OM 936 ( KIT) the exhaust tract is also elementary for the efficiency of the entire exhaust aftertreatment system. 2 ANALYSIS OF ENGINE-INTERNAL FORMATION PROCESSES During the combustion process, mainly NO is formed. From this, NO 2 can be formed subsequently. The formation mechanisms of NO have already been well investigated and comprehensively described. In contrast, the processes and mechanisms for the engine-internal formation of NO 2 are hardly known. Within the FVV research project Formation/Modeling NO 2, which was founded by the BMWi, the basic understanding for the formation of NO 2 under diesel engine conditions is to be developed using experimental and simulative tools. Extensive experimental investigations were performed on an engine test bench to identify and to analyse the impact of thermodynamic conditions and engine operating parameters on the engine-internal formation processes. Furthermore, a detailed chemical kinetic mechanism is developed for the description of the reaction pathways and the formation processes of NO and NO 2. For the determination of relevant concentration-time curves for the parameterisation and validation of the mechanism, additional investigations on a shock tube were carried out. 3 TEST ENGINE AND BOUNDARY CONDITIONS The test engine used is a single-cylinder research aggregate, based on the medium-heavy Euro VI commercial vehicle diesel engine OM 936 from Mercedes-Benz. The technical data of the test engine are listed in TABLE 1. The test bench design allows individual and reproducible adjustment of all operating parameters. Coolant and engine oil are delivered via separate conditioning circuits and the fuel supply is carried out via an autarkical driven high-pressure pump. A modified engine control unit is used to specify the injection parameters. An external water-lubricated compressor is used to provide the demanded charge pressure. Besides the pressure, also the temperature and the humidity of the charge can be set very precisely over large ranges. An exhaust throttle valve is used to adjust the exhaust gas back pressure and thus to control the pressure gradient between intake and exhaust gas. This allows adjusting near-series operating conditions, corresponding to the turbocharged full engine. Further, the increased exhaust gas back pressure is required to realise the high-pressure EGR. A gas-coolant heat exchanger with a mounted rotary valve is used for cooling and dosing the recirculated exhaust gas. Those components correspond to the engine s serial parts. An emission measurement system AVL AMA4000 and an FTIR of the type MKS MG2030 are used for the analysis of the composition of the exhaust gas. The AMA4000 has a dual-channel chemiluminescence detector (CLD) that is capable of simultaneously measuring the concentrations of NO and NO x using a thermocatalytic NO 2 converter. The concentration of NO 2 then results from the difference between these two measured values. In addition, the NO 2 concentration is measured with a fast infrared analyser. DEGAS (Dynamic Exhaust Gas Analyser System) determines the absorption of NO 2 in the narrowband middle infrared range (wavelength = μm) and allows the dynamic measurement of NO 2 with a sampling rate of up to 200 Hz [1]. MTZ worldwide

3 RESEARCH Diesel Engines 4 INFLUENCE OF ENGINE LOAD AND SPEED Complying with current emissions directives requires the adaptation of various operating parameters such as EGR rate, air-fuel ratio λ and also injection parameters (e.g. injection time, fuel pressure) for the entire engine operating range. As a consequence, the resulting pollutant concentrations in the exhaust gas are significantly characterised by these specific application measures. But to consider the dependency of the nitrogen oxide emissions from engine load and speed without any interfering effects, imposed by those applicational settings, the engine is operated without exhaust gas recirculation. Further operating parameters are only adapted to a lesser extent, FIGURE 1. Under these conditions, the NO x emissions reach their maximum concentration at the highest engine loads and low engine speeds. Here, the high combustion temperatures and the low process velocities ensure ideal conditions for the formation of thermal NO and are therefore decisive for the total NO x emissions. For the formation of NO 2, additionally, the oxygen concentration in the combustion chamber plays an important role. The maximum NO 2 concentration is recorded at medium engine load and low engine speed. Under these conditions, an adequate amount of NO is formed in the combustion chamber which, in combination with excess oxygen and a sufficiently large time window, promotes NO 2 formation. At higher loads and thus decreasing λ, the NO 2 concentration also drops significantly. And especially at high speeds, the NO 2 emissions are reduced to a minimum. However, it is generally conspicuous that NO 2, in this case, represents only a very small proportion of the total NO x emissions. In most cases, this is less than 4 %. FIGURE 1 Engine maps of the concentrations of NO and NO 2 without EGR ( KIT) 72 FIGURE 2 Impacts of injection timing and rail pressure on the NO und NO 2 emissions (partial load: 55 Nm at 1400 rpm) ( KIT)

4 FIGURE 4 Results of the in-cylinder measurement of the NO 2 concentration by using fast gas sampling (partial load: 55 Nm at 1400 rpm) ( KIT) FIGURE 3 Impacts of NO 2 and EGR on the NO und NO 2 emissions (partial load: 51 Nm at 1800 rpm) ( KIT) rail pressure and start of injection. Although an increase in NO concentration also causes an increase in the NO 2 concentration, the proportion of NO 2 in the total NO x emissions decreases. Accordingly, the highest NO 2 /NO x ratios are always recorded at the lowest NO x emissions. Solely the increase in λ shows a deviating behaviour. While all other measures, particularly the exhaust gas recirculation, primarily influence the NO emissions, a change of λ acts more strongly on the nitrogen dioxide. Thus, an increase of λ also leads to an increase in the NO concentration. However, the NO 2 concentration increases even more. The comparison of the different dilution variations in FIGURE 3 shows that the influence of the EGR rate on NO 2, compared to the influence of λ, is negligibly small. 5 PARAMETER VARIATIONS The variation of the different operating parameters and influencing variables shows that significantly higher NO 2 /NO x ratios can also occur [2]. In the course of these investigations, the influences of different injection parameters, the charge dilution with fresh air and exhaust gas, as well as the charge state (temperature, humidity) for different operating points are considered. Across all variations, it can be seen that any measurable effect on NO x emissions always coincides with the change in both, NO and NO 2 emissions. Thus, an increase in NO also results in a simultaneous increase in NO 2. FIGURE 2 illustrates this by an exemplary variation of the 6 FAST GAS SAMPLING In order to analyse engine-internal formation processes, exhaust gas samples are taken directly from the combustion chamber using a gas sampling valve. A pressure sensor bore, which is located in the cylinder head at the level of the piston bowl lip between the inlet and outlet sides, serves as an access to the combustion chamber. The gas sampling is applied at defined points of time during the working cycle. The solenoid valve opens in every tenth working cycle with a duration of 1.5 ms. By varying the time point of the gas sampling, the course of the concentration of NO 2 during the phases of compression and expansion can be analysed, FIGURE 4. The measurement data show that the NO 2 concentration MTZ worldwide

5 RESEARCH Diesel Engines increases immediately after the start of combustion. During combustion, NO 2 increases further but then decreases again during expansion. During the later expansion, the measured values of NO 2 correspond to the NO 2 concentration of the emitted exhaust gas. This leads to the assumption that the formation of NO 2 appears to be completed at the end of combustion, and no further NO 2 formation occurs during the later expansion and the charge cycle. Even in the externally recirculated exhaust gas, the NO 2 concentration does not seem to change, as the measured values before the start of the combustion show. 7 KINETIC MODELING OF NO 2 FORMATION For the kinetic modelling, a reaction mechanism was developed that is able to describe both fuel combustion and formation of NO x. n-heptane was selected as a model fuel due to its moderate molec- FIGURE 5 Influence of λ and NO X on the NO 2 emissions (above) and the NO 2 /NO X ratio (below) illustration of experimental results from a variation of λ and EGR (left) and the results from the kinetic simulations (right) ( KIT) FIGURE 6 Correlation of NO 2 or NO 2 /NO X with NO from various parameter variations: experimental results (left) and simulation results (right) ( KIT) 74

6 ular size and a cetane number (CN 54) [3] similar to that of diesel fuel. The complete reaction mechanism with 668 chemical species and 6654 reactions consists of an n-heptane oxidation mechanism [4] and a sub-mechanism for NO x formation [5-9]. The sub-mechanism was reduced for the actual range of our experimental conditions. In this way, the number of species could be decreased by 17 and the number of reactions by 985. All simulations [10] were performed with this reduced mechanism, where homogeneous reaction mixtures and adiabatic-isobaric conditions were assumed; transport processes were not included. Extensive reaction path analyses revealed that NO 2 is exclusively formed via NO, where the most important reaction is the direct oxidation of NO by peroxy radicals according to: Eq. 1 NO + RO 2 NO 2 + RO combustion chamber. The simulation points out that particularly the direct oxidation of NO with peroxy radicals plays an important role. In addition, low engine speeds promote the NO 2 formation, as more time is available for the formation process. Thus, all measures for influencing the NO x emissions also indirectly affect the NO 2 formation. However, since those measures primarily influence the NO formation, the proportion of NO 2 usually decreases with higher NO x concentrations. In contrast, any change in NO 2 has a stronger effect on the formation of NO 2 than on NO. Therefore, an increase in the oxygen concentration leads to an increase in NO 2 emissions with a simultaneous increase in the NO 2 /NO x ratio. The results from the fast gas sample from the combustion chamber as well as from the chemical kinetic simulations support the assumption that NO 2 is formed directly during combustion. Accordingly, no further formation processes in the exhaust tract without any interaction with catalytic components are to be expected. Here, R represents a hydrocarbon radical or a hydrogen atom. Another reaction pathway proceeds via the reaction sequence: Eq. 2 NO + OH + O 2 HNO 2 + O 2 HO 2 + NO 2 As peroxy radicals are relatively stable intermediates, comparably high RO 2 concentrations can be built up in the cooler regions of the flame. These peroxy radicals react with NO that is formed via the Zeldovich mechanism in the hotter regions of the flame and transported by diffusion into the cooler regions [11]. In the simulations, we obtained significantly higher concentrations of HO 2 than of RO 2 and consequently a dominance of the reaction: Eq. 3 NO + HO 2 NO 2 + HO In the simulations, high absolute concentrations of NO 2 were obtained whenever a large amount of NO x was formed in a large excess of air (λ >> 1). High NO 2 /NO x ratios, however, were found for rather low absolute concentrations of NO x. These results are consistent with the findings of the experimental investigations as is shown in FIGURE 5. Concordant parameter variations in the experimental studies and the kinetic simulations also revealed a very good agreement between the respective amounts of NO 2 and the NO 2 /NO x ratio as a function of the NO amount, FIGURE 6. It turned out that the parameters with the strongest influence on the simulation results were the EGR rate, the temperature at the beginning of the combustion, and λ. The maximum amount of NO 2 varies with λ and is located in the range λ = 1.3 to 1.8 depending on the burning conditions. Furthermore, a re-increase occurs for a large excess of oxygen. High temperatures at the start of combustion and low EGR rates lead to enhanced NO 2 formation due to the higher resulting temperatures in the combustion chamber and the correspondingly enhanced NO formation. 8 SUMMARY The presented investigations work out the relevant influence parameter and process variables for the engine-internal NO 2 formation. Both the experimental and the reaction kinetic studies show that the NO 2 formation in diesel engines is mainly influenced by the NO concentration and the quantity of excess oxygen in the REFERENCES [1] Brunner, R.; Lambrecht, A.; Herbst, J.; Heubuch, A.; Jacob, E.: Quantenkaskaden-Laserspektrometer für die schnelle und artefaktfreie Abgasanalyse. 7. Internationales Forum Abgas- und Partikelemissionen, Ludwigsburg, 2012 [2] Rößler, M.; Velji, A.; Janzer, C.; Koch, T.; Olzmann, M.: Formation of Engine Internal NO 2 : Measures to Control the NO 2 /NO x Ratio for Enhanced Exhaust After Treatment. SAE paper , SAE World Congress Experience, Detroit, 2017 [3] Yanowitz, J; Ratcliff, M. A.; McCormick, R. L.; Taylor, J. D.; Murphy, M. J.: Compendium of experimental cetane numbers. In: Technical Report, National Renewable Energy Laboratory (2014) [4] Mehl, M.; Pitz, W. J.; Westbrook, C. K.; Curran, H. J.: Kinetic modelling of gasoline surrogate components and mixtures under engine conditions. In: Proc. Combust. Inst. 33 (2011), pp [5] Faravelli, T.; Frassoldati, A.; Ranzi, E.: Kinetic modelling of the interactions between NO and hydrocarbons in the oxidation of hydrocarbons at low temperatures. In: Combustion Flame 132 (2003), pp [6] Contino, F.; Foucher, F.; Dagaut, P.; Lucchini, T.; D Errico, G.; Mounaïm- Rousselle, C.: Experimental and numerical analysis of nitric oxide effect on the ignition of iso-octane in a single cylinder HCCI engine. In: Combustion Flame 160 (2013), pp [7] Dayma, G.; Ali, K. H.; Dagaut, P.: Experimental and detailed kinetic modelling study of the high pressure oxidation of methanol sensitized by nitric oxide and nitrogen dioxide. In: Proceeding Combustion Instution 31 (2007), pp [8] Glaude, P. A.; Marinov, N.; Koshiishi, Y.; Matsunaga, N.; Hori, M.: Kinetic Modeling of the Mutual Oxidation of NO and Larger Alkanes at Low Temperature. In: Energy Fuels 19 (2005), pp [9] Manion, J. A. et al.: National Institute of Standard and Technology. NIST Chemical Kinetics Database, [10] Cuoci, A.; Frassoldati, A.; Faravelli, T.; Ranzi, E.: OpenSmoke++: An object-oriented framework for the numerical modelling of reactive systems with detailed kinetic mechanisms. In: Computer Physical Communication 192 (2015), pp [11] Bowman, C.T.: NO x Formation and Models. In: Encyclopedia of Automotive Engineering. Chichester, John Wiley & Sons Ltd. (2014), pp THANKS The research project Bildung/Modellierung NO 2 (project number 1173) was funded by the Bundesministerium für Wirtschaft und Energie (BMWi) and supervised by the Forschungsvereinigung Verbrennungskraftmaschinen e.v. (FVV). The authors would like to thank for the financial funding and the support of the supervising working group under the chairman Dr. Frank P. Zimmermann, Daimler AG. We would also like to thank Dr.-Ing. Amin Velji, Institute of Internal Combustion Engines at the Karlsruhe Institute of Technology, for his contribution to the research project as well as to this publication. MTZ worldwide