Elsevier

Fuel

Volume 224, 15 July 2018, Pages 32-46
Fuel

Full Length Article
Laminar flame properties of C1-C3 alkanes/hydrogen blends at gas engine conditions

https://doi.org/10.1016/j.fuel.2018.02.167Get rights and content

Highlights

  • New parameter HC-ratio for modeling laminar flame speed of C1-C3 alkanes blends.

  • Laminar flame speed of C1-C3 alkanes blends are similar for a constant HC-ratio.

  • Laminar flame speed correlations for C1-C3 alkanes, H2, EGR blends.

  • Le Chatelier’s blending law for CH4/H2 shows good agreement up to 30 vol% H2.

  • Flame thickness correlation for C1-C3 alkanes, H2, EGR blends.

Abstract

The use of fuel blending is encouraged in order to achieve more flexibility in gas engines. In order to design such engines effectively, relevant information about the laminar flame speeds and laminar flame thickness are necessary. Hydrodynamic and thermo-diffusive instabilities at gas engine conditions prevent the acquisition of reliable data experimentally. One-dimensional numerical simulations with detailed chemistry can be a solution. A huge database of laminar flame speeds is generated covering a broad range of gas engine applications, pressure (p) 0.1–20 MPa, fresh gas temperature (Tu) 300–1100 K, air-fuel equivalence ratio (λ) 0.9–2.5, methane 100–60 vol%, ethane 0–40 vol%, propane 0–40 vol%, hydrogen 0–30 vol% and exhaust gas recirculation (EGR) 0–30 m%. The detailed reaction mechanisms GRI 3.0 and AramcoMech 1.3 are used for the generation of flame speed data for the mentioned conditions. A laminar flame speed correlation for 100% hydrogen extending up to elevated pressure and temperature conditions is developed. A blending law based on Le Chatelier’s rule is investigated. It is observed that the HC-ratio has a very determining effect for the laminar flame properties of different C1-C3 alkane blends. Based on this observation, it was possible to derive a very efficient correlation for both laminar flame speed and laminar flame thickness for the group of natural gas blends with methane, ethane, propane and hydrogen, and as well as including relevant EGR, which corresponds within 7% accuracy to the calculated database of about 73,000 points. The developed laminar flame speed correlation is incorporated in an engine process simulation code. It is validated with the measured in-cylinder pressure traces from the single cylinder research engine experiments for different gas blends and EGR ratios.

Introduction

Gas engines are gaining increased interest because of economical advantages and reduced emissions. Natural gas allows CO2 emission reduction in the range of 20% compared to conventional fuels like gasoline and diesel due to the lower C/H ratio. Furthermore, gas engines operate without soot and for lean operation conditions with relatively low NOx emissions. The main constituent of natural gas is methane, with small amounts of heavier hydrocarbons, mainly based on C2-C4. This fraction can significantly vary depending on the source of natural gas up to 18 vol% [1]. The influence of these compounds has a significant effect on the reactivity of natural gas [2]. Further, processes such as gasification of coal or biomass can produce fuels with high hydrogen content and other components up to 38 vol% that are more reactive than methane [3].

Apart from the availability of natural gas surrogate mixtures, blending of natural gas with reactive gases such as ethane or hydrogen is an attractive option to the gas turbine and engine industry. Since it can extend the flammability limits and can increase the flame speeds at extremely fuel lean conditions [1], [4]. Further, the addition of hydrogen reduces pollutant emissions [5], [6]. Natural gas blending with higher hydrocarbons such as ethane and propane as well as hydrogen can be an interesting alternative. Furthermore, the use of these mixtures has been subject to a very large number of studies in relation to engine applications but also in relation to the gas supply industry where there have been a number of significant studies recently [7], [8].

Heavy duty gas engines in particular operate on fuel lean conditions with diluted mixtures. This can be achieved either with excess air, leading to a “lean burning” mode, or with exhaust gas recirculation (EGR), being sometimes called as “lambda = 1 + EGR” engine [9], [10]. Both dilution methods reduce the combustion temperatures which lead to lower NOx emissions and to lower heat losses to the walls, hence, the efficiency of the gas engine is increased.

However, the dilution is limited to a certain extent as the reactivity and laminar flame speed decreases, which can reduce the burn out, it can lead to combustion instabilities and to flame quenching. Near this limit, the unburned hydrocarbon emission (UHC) increases and the engine efficiency decreases.

Laminar flame speed SL is fundamental not only for the reaction progress but also for the design of any combusting system, as it is determinant for combustion phenomenon such as flame stabilization, quenching, turbulent flame structure and plays a vital role in several aspects of combustion processes in spark ignition engines and/or gas turbines [10], [11]. Ranzi et al. [12] reviewed experimental data of laminar flame speeds of CH4/air mixtures at different temperature and pressure conditions available in literature and emphasized on scatter in the measurements. Commonly used experimental methods for the measurement of laminar flame speeds are, such as the spherical flames in constant volume or constant pressure chambers [13], [14], counter flow flames [15] and flat flame adiabatic burner (heat-flux method) [2], [16].

Laminar flame speeds of pure hydrogen have been measured by several groups at atmospheric conditions of 1 atm and 298 K for a broad range of air-fuel equivalence ratios [17], [18], [19]. Kerjci et al. [20] conducted closed vessel experiments for pure hydrogen at 1 atm for higher temperatures up to 443 K. However, measurements for unstretched SL of pure hydrogen at elevated pressures are extremely difficult to achieve [21]. This is due to the higher sensitivity of hydrogen to thermo-diffusive and hydrodynamic instabilities leading to unstable cellular flame structures. The same problem is relevant for high hydrogen blended fuels [22].

Yu et al. [23] published measured laminar flame speed data of CH4 blended with H2 up to 50 vol% using the counterflow method. Several research groups have published measurement data of CH4/H2 blends [24], [25], [26] up to 40 vol% H2. Hu et al. [27] conducted CH4/H2 blending laminar flame speed experiments using the combustion bomb method. This data was later theoretically investigated by Al-Mughanam et al. [28], where different blending laws were investigated. Dirrenberger et al. [2] published CH4/H2 blending laminar flame speeds using the heat flux method. As per the authors knowledge, only Plichta et al. [1] conducted methane/hydrogen blending from 50 to 90 vol% H2 at elevated pressures up to 5 atm using the combustion vessel method.

Few research groups have published laminar flame speed data for realistic natural gas mixtures [2], [3], [8], [16]. Dirrenberger et al. [2] investigated pure methane, ethane, propane, butane and natural gas surrogate mixtures consisting of different proportions of ethane and propane up to 82 vol% methane. Kochar et al. [16] conducted laminar flame speed experiments for pure methane and binary blends of CH4/C2H6 and CH4/C3H8 at elevated pressures up to 10 atm and 650 K. Nillson et al. [7] measured laminar burning velocities of hydrocarbon blends of relevance to natural gas combustion, with addition of 0, 10, 35 and 50% hydrogen using the heat flux method.

For conditions relevant to gas engines, where the in-cylinder pressures can reach up to 20 MPa and the fresh gas temperatures can be higher than 1000 K, the experimental data is very difficult to obtain due to the effects of flame stretch and instabilities. Instead, detailed one-dimensional numerical simulations can be conducted, including the detailed reaction mechanism as well as the diffusive transport processes inside the reaction zone [10], [29], [30], to obtain the laminar flame speed database for the gas blends at engine relevant conditions. Ratzke et al. [10] created a tabulated laminar speed database using the GRI 3.0 mechanism [31] for methane at gas engine conditions as a function of four parameters, λ, Tu,p and YEGR (mass fraction of EGR). Tabulated data, however, leads to very expensive interpolation procedures, especially when the number of parameters increases, as in the situation of fuel blends. This is even more important when calculations within 3D CFD simulations are anticipated [11].

Mathematical correlations for the laminar flame speeds would allow calculations with significantly reduced computational cost. These correlations can be derived from the database obtained from 1D numerical simulations using detailed reaction mechanism appropriate for the engine conditions [22], [32], [33]. For gas blends, the blending effect has a non-linear evolution effect on the laminar flame speed. This leads to complicated mathematical correlation equations to capture this behaviour [22].

In the present work, a laminar flame speed correlation is developed as a function of seven parameters: air-fuel equivalence ratio, pressure, temperature, mole fractions of ethane, propane, hydrogen and EGR ratios for gas engine conditions. This 7D matrix is reduced to a 4D matrix, where the gas composition information is incorporated in terms of the hydrogen to carbon ratio (HC-ratio or HC) in a block-structured pattern. This will be explained in the following sections. Complications in the mathematical correlations due to the blending effects, are taken into account, in which the coefficients are calculated in a pre-processing manner. The coefficients in the present work are entirely the functions of mixture composition, i.e., they contain the information regarding the air-fuel equivalence ratio (λ), mole fractions of ethane (XC2H6), mole fractions of propane (XC3H8), mole fraction of hydrogen (XH2) and the exhaust gas recirculation (YEGR) in the methane mixture. Since most gas engines operate with premixed combustion, hence the composition is already known. This enables the laminar flame speed to be just a 2D function of pressure and temperature for the whole simulation. Even in the case of scavenged prechamber ignition, it is not a concern, since the gas composition variation lasts for just a few crank angle degrees. In addition, an SL correlation for pure hydrogen is developed for engine conditions as a part of this block-structured approach. Details of this approach is discussed in further sections.

Section snippets

Influence of HC-ratio

Warnatz [29] suggested the reactions Ḣ+O2=O¨+ȮH and CO+ȮH=CO2+Ḣ, as the most fundamental reactions for the combustion of any hydrocarbon fuel up to 16 carbon atoms (with the dots indicating the unsaturated free electrons of the radicals). The other major sensible reaction being Ḣ+O2(+M)=HȮ2(+M). Kochar et al. [16] conducted sensitivity analysis for their experiments for pure methane, ethane, propane and the blends 60% CH4/40% C2H6 and 60% CH4/40% C3H8 at 298 K and 5 atm. It is seen that

Choice of mechanisms

The choice of reaction mechanism is based on the relevance of the mechanism for gas blends of CH4/C2H6/C3H8/H2/air/EGR at gas engine conditions. For the generation of a huge laminar speed database, the numerical consistency and computational cost (depending on the size of the mechanism) needs to be taken into account. The validation of the mechanisms with experimental data is not performed in this work, instead the choice is based on the validation found in literature. However, the created SL

Laminar flamespeed correlation

One dimensional detailed numerical simulations are performed with Cantera [30]. This code calculates the detailed local reaction kinetic processes as well as the detailed transport of heat and species based on the relevant transport data [30] and thermo-physical properties [30]. In order to ensure the Soret and Dufour effects, especially for the cases with pure H2 [21], the multicomponent diffusion is considered.

The one dimensional numerical simulation with detailed chemistry allows the

Experiments of single cylinder research engine

Single cylinder research engine test bench measurements are performed at Daimler AG to obtain the in-cylinder pressure traces for different gas-blend compositions and the mixture dilution. The engine is operated with natural gas (97 vol% CH4) and further blends of CH4/C2H6 and CH4/H2 at various dilution rates. Table 5 shows the basic specifications of the test engine, details of the engine set up and experimental procedure can be found in [54].

In all experiments the cooled recirculated gas

Laminar flame thickness

The laminar flame thickness δL determines the location of operating regime on the Borghi-Peters diagramm [11]. Hence, the flame thickness is a very important parameter to define the extent to which turbulence influences the flame stretch. Several combustion models even require the value of δL [11], [32]. Due to computational efficiency needed in engine simulations, correlations for δL are helpful. δL correlations for hydrocarbon fuels have been proposed in literature [22], [56]. δL can be

Conclusions

A numerical approach to evaluate the laminar flame properties for gas mixtures is developed in this work. Laminar flame speed correlations are developed for gas blends and validated with the single cylinder engine measurements. The main results are summarized as follows:

  • 1.

    The laminar flame speeds for gas blends of C1-C3 alkanes up to 60 vol% CH4 can be represented with a single parameter HC-ratio. The gas blends of CH4/C2H6/C3H8 have similar laminar flame speeds for a constant HC-ratio.

  • 2.

    For a gas

Acknowledgements

The authors thank the Forschungsvereinigung Verbrennungskraftmaschinen (FVV) e.V for the financial support for the ongoing project “30 bar IMEP in gas engines” in collaboration with the Lehrstuhl für Verbrennungskraftmaschinen (G. Wachtmeister, M. Prager, S. Eicheldinger) TU Munich.

References (58)

  • C. Sun et al.

    Dynamics of weakly ststretch flames: quantitative description and extraction of global parameters

    Combust Flame

    (1999)
  • U. Gerke et al.

    Derivation of burning velocities of premixed hydrogen/air flames at engine-relevant conditions using a single-cylinder compression machine with optical access

    Int J Hydrogen Energy

    (2010)
  • S. Bougrine et al.

    Numerical study of laminar flame properties of diluted methane-hydrogen-air flames at high pressure and temperature using detailed chemistry

    Int J Hydrogen Energy

    (2011)
  • G. Yu et al.

    Laminar flame speeds of hydrocarbon + air mixtures with hydrogen addition

    Combust Flame

    (1986)
  • F. Halter et al.

    Characterization of the effects of pressure and hydrogen concentration on laminar burning velocities of methane-hydrogen-air mixtures

    Proc Combust Inst

    (2005)
  • F. Coppens et al.

    The effects of composition on burning velocity and nitric oxide formation in laminar premixed flames of CH4 + H2 + O2 + N2

    Combust Flame

    (2007)
  • R. Hermanns et al.

    Effects of temperature and composition on the laminar burning velocity of CH4 + H2 + O2 + N2 flames

    Fuel

    (2010)
  • E. Hu et al.

    Experimental and numerical study on laminar burning velocities and flame instabilities of hydrogen-air mixtures at elevated pressures and temperatures

    Int J Hydrogen Energy

    (2009)
  • J. Vancoillie et al.

    A laminar burning velocity and flame thickness correlation for ethanol-air mixtures valid at spark-ignition engine conditions

    Fuel

    (2012)
  • E. Hu et al.

    Laminar flame speeds and ignition delay times of methane-air mixtures at elevated temperatures and pressures

    Fuel

    (2015)
  • F. Egolfopoulos et al.

    Laminar flame speeds of methane-air mixtures under reduced and elevated pressures

    Combust Flame

    (1989)
  • K. Bosschaart et al.

    The laminar burning velocity of flames propagating in mixtures of hydrocarbons and air measured with the heat flux method

    Combust Flame

    (2004)
  • S. Verhelst et al.

    Laminar and unstable burning velocities and markstein lengths of hydrogen-air mixtures at engine-like conditions

    Proc Combust Inst

    (2005)
  • S. Tse et al.

    Morphology and burning rates of expanding spherical flames in h2/o2/inert mixtures up to 60 atmospheres

    Proc Combust Inst

    (2000)
  • V. Di Sarli et al.

    Laminar burning velocity of hydrogen-methane/air premixed flames

    Int J Hydrogen Energy

    (2007)
  • E. Hu et al.

    Experimental and numerical study on laminar burning characteristics of premixed methane-hydrogen-air flames

    Int J Hydrogen Energy

    (2009)
  • F. Dinkelacker et al.

    Structure of locally quenched highly turbulent lean premixed flames

    Symp (Int) Combust

    (1998)
  • Plichta D, Mathieu O, Petersen E, Bourque G, Curran H, Burke S, et al. Laminar flame speeds of natural gas blends with...
  • P. Dirrenberger et al.

    Measurements of laminar flame velocity for components of natural gas

    Energy Fuels

    (2011)
  • Cited by (15)

    • Effects of natural-gas blending on ignition delay and pollutant emission of diesel fuel for the condition of homogenous charge compression ignition engine

      2022, Fuel
      Citation Excerpt :

      It requires more information on burning of each fuel as well as a mixture, i.e., dual fuel, based on their physical and chemical properties. The major component of natural gas, i.e., methane, has been studied by many researchers, who measured ignition delay time [14–19], species concentration profiles [20–24], and laminar flame speed [17,25–29] of methane. It led to reliable detailed chemical mechanisms for methane.

    • Kinetic study of the effects of hydrogen blending to toluene reference fuel (TRF)/air mixtures on laminar burning velocity and flame structure

      2020, Fuel
      Citation Excerpt :

      Because hydrogen addition increases the mole fraction of H, it enhances the reactions of TRF. On one hand, H radicals play the dominant role in the H-abstraction of TRF; on the other hand, H radicals are an important chain carrier that can enhance the chain reactions of TRF and the radical pool [59,60]. Therefore, although the addition of hydrogen reduces the mole fraction of fuel, the chemical effect of hydrogen has an important effect on TRF, which can directly prompt the active radical pool.

    View all citing articles on Scopus
    View full text