Effect of oxygen-rich combustion on soot formation in laminar co-flow propane diffusion flames

Highlights

• Effect of oxygen-rich on soot in propane flames was numerically studied.

• Oxygen-rich promotes soot formation in flames.

• Oxygen-rich increases PAHs concentration near the nozzle.

• Soot increase ascribes to the increase in PAHs near the nozzle.

Abstract

Oxygen-rich combustion is a new type of clean combustion technology with important application prospects. In this work, the effects of oxygen-rich combustion on soot formation in the propane/(O₂+N₂) laminar flow coaxial jets diffusion flame were numerically investigated by using the detailed gas-phase chemical reaction model with the mechanism of tetracyclic aromatic hydrocarbons and the complex thermodynamic properties and transport characteristics parameters. Soot surface growth follows the hydrogen-abstraction-carbon-addition (HACA) model. A hybrid gas-phase mechanism was adopted, which contains a DLR-based polycyclic aromatic hydrocarbons (PAHs) formation, growth model and a gas-phase model. Results show that the oxygen-rich combustion has a great influence on the flame temperature, especially the high temperature region. With the increase of oxygen concentration, the soot formation region of flame broadens and the maximum of soot volume fraction increase from 3.95 ppm to 10.87 ppm. The extra oxygen makes PAHs increased around the nozzle, leading to larger rate in early soot nucleation and surface growth, eventually more soot yield.

Introduction

Soot is an important product of traditional fuel combustion. It is not only harmful to human health, but also has a great impact on the environment [1], [2]. For instance, soot particles will move to various respiratory organs along with human breathing, which will seriously affect the human respiratory system. Even soot particles carrying toxic substances will lead to canceration of human organs. In addition, soot is also responsible for photochemical smog, such as London smog incident and Los Angeles photochemical smog incident. Therefore, how to reduce the soot formation is an important research topic in the field of combustion, especially in the field of internal combustion engines. In order to accurately and conveniently explore soot formation mechanism, most scholars choose the axially symmetric diffusion flame of hydrocarbon fuel as the research object [3], [4], [5]. The process of soot nucleation, surface growth, coagulation, agglomeration and oxidation in diffusion flames is very complex [6]. In addition, the soot formation mechanism also has an important influence on the synthesis of other carbon nanomaterials in flames, such as fullerenes [7], carbon nanotubes [8], [9], etc. Although there are many hypotheses about the reaction path of soot formation in each stage, there is no unified mechanism of soot formation at present. Therefore, different methods and technologies are still needed to optimize the simulation and detection soot formation.

At present, there are two kinds of soot detection methods, including contact sampling analysis and non-contact optical diagnosis. There are three kinds of contact sampling methods for soot, i.e., thermophoresis probe sampling method [10], [11], SiC fiber deposition method [12], [13] and dilution extraction method [14]. Almost all non-contact measurements are based on optical methods [15], [16], which can obtain soot parameters such as soot volume fraction, average particle size and particle number density. As another aspect of soot research, combustion numerical simulation has become an efficient and economical universal means. The advantage of numerical simulation is that it does not need to spend a lot of money to purchase experimental equipment, and can simulate some specific conditions that impossible achieved by experiments and avoid the limitations of experimental measurements. Therefore, using numerical simulation to study the numerical density distribution and volume fraction distribution of small molecule gaseous products, PAHs and soot during soot formation is extraordinary for controlling soot emissions. In our previous works [17], [18], [19], we numerically studied the effects of carbon dioxide, hydrogen, elevated initial temperature and pressure on the combustion characteristics of the flame. To date, various empirical and semi-experienced soot models are also widely used for laminar flame calculation [20], [21], [22]. The detailed soot model developed by Frenklach et al. [23], [24] contains 600 elementary reactions and 200 components, but is not suitable for multidimensional flame soot simulation.

On the other hand, propane flames are widely used to study soot formation characteristics, mainly focusing on the effects of fuel structure, flame temperature, gas addition (such as oxygen, hydrogen, etc.) and pressure on soot [25], [26], [27]. Bento et al. [28] investigated the effect of pressure on soot formation and the structure of the temperature field in coflow propane/air laminar diffusion flames, and reported that the overall soot temperature decreased with increasing pressure, and as the pressure increased, the temperature gradient increased. Sahu et al. [29] studied the distribution of intermediate hydrocarbons (CH₄, C₂H₂, C₂H₄ and C₂H₆) and soot in propane/air triple flames, and found that the peak concentration of intermediate hydrocarbons occurred in the internal premixed flame and decays rapidly. It should be noted that the peak soot volume fraction of flame occurred above the peak acetylene concentration, which can be inferred that the formation and consumption of acetylene correspond to the formation mechanism of soot in flames. Henríquez et al. [30] investigated the effect of oxygen index on soot formation in laminar coflow propane diffusion flames by experimental methods. They concluded that with the increase of oxygen volume fraction, soot growth and oxidation rate increased, and the maximum soot volume fraction and the overall peak soot volume fraction also increased. In a recent work, Qiu et al. [4] investigated the soot volume fraction and number density in a laminar n-heptane/n-butanol flame by experimental and numerical methods, and found that n-butanol addition to n-heptane decreases the molar fraction of A4 and C₂H₂, resulting in the decrease of inception and HACA rates, and consequently, the total density and mass fraction decrease simultaneously. In a recent work, Wu et al. [31] found that adding carbon dioxide to propane diffusion flame inhibited the total amount of soot and PAHs as well as the size of soot particles, and also inhibited the conversion of PAH to soot. Compared with ethylene flame, the decrease of primary particle size of soot is mainly due to the shortening of growth residence time, which indicates that the decrease of surface growth rate mainly ascribes to the decrease of surface area of soot particles. Interestingly, Wang et al. [32], [33] have done lots of research on soot formation mechanism from biomass pyrolysis.

Previous studies have focused on the effects of fuel structure, flame temperature, gas addition (such as oxygen, hydrogen, etc.) and pressure on soot, but there are few studies on the soot formation mechanism in propane laminar diffusion flames in an oxygen-rich atmosphere from the perspective of simulation. In our previous works [13], [34], the effects of oxygen concentration on the radial and axial soot formation of methane and ethylene flames were systematically experimentally investigated, respectively, and the synergistic effect of methane addition on soot formation in ethylene flames was analyzed. The soot formation mechanism needs to be further explored in propane laminar diffusion flames. Herein, the effects of oxygen enriched on soot formation in the propane/(O₂+N₂) laminar flow coaxial jets diffusion flame were numerically investigated by using the detailed gas-phase chemical reaction model with the mechanism of tetracyclic aromatic hydrocarbons and the complex thermodynamic properties and transport characteristics parameters. The aim of this work mainly modeled the temperature, soot distribution and gas product distribution in the flame axial and radial directions, and discussed the corresponding relationship between them.