Turbulence Radiation Interactions Research Articles

Turbulence-radiation interaction in large eddy simulations of turbulent reactive flows: a review

Turbulence-radiation interaction (TRI) arises from the highly non-linear coupling between fluctuations of radiative intensity and fluctuations of temperature and chemical composition of the medium. It is well recognized that neglecting turbulent fluctuations of the latter scalars when computing the mean radiation field, as is standard in Reynolds-Averaged Navier-Stokes (RANS) simulations, leads to mispredictions of the radiative heat fluxes and the radiant fraction in as much as one order of magnitude. Conversely, the importance of TRI in large eddy simulations (LES)—i.e., the effect of subfilter-scale (SFS) fluctuations on the filtered radiation field—has only started to be investigated in the last decade. The present text reviews the knowledge gathered so far on the subject. The instantaneous and the spatially filtered forms of the radiative transfer equation (RTE) are presented, and the unresolved correlations resulting from the filtering operation are discussed. Models for accounting for SFS-TRI are surveyed, with the most popular ones (namely, probability density function-based methods and empirical approximations) reviewed in detail. An appraisal of the literature on the importance and characteristics of SFS-TRI is performed, with special attention given to turbulent jet diffusion flames and pool fires, which are the two configurations for which SFS-TRI has been studied the most. The main findings of these studies are summarized, and recommendations are given on directions of future works on the subject.

Large eddy simulations of a buoyant turbulent line flame using conditional source-term estimation (CSE)

The objective of this paper is to assess the capabilities of the Conditional Source-term Estimation (CSE) approach applied to a laboratory scaled turbulent buoyant flame without extinction. CSE is coupled with the Large Eddy Simulation (LES) solver, FireFOAM. Tabulated detailed chemistry is included. Radiation is treated using the optically thin model using four different implementations and the effect of subgrid scale (sgs) Turbulence-Radiation Interactions (TRI) is considered. Predictions of time-averaged temperatures and corresponding root mean square (rms) are compared with the experimental measurements at several locations. Further, flame shape, unconditional and conditional species mass fractions, axial velocity and mixture fraction are also examined for qualitative analysis. The predicted temperatures are in good agreement with the experimental data, except very close to the fuel inlet. The predicted temperatures increase more slowly compared to the experimental data in the first 0.09 m above the burner. The temperature rms is in reasonable agreement with experimental data with the peak being overpredicted by approximately 17%. The predicted flame height closely matches the experimental value. The temperature predictions are consistent with previously published results. Best predictions are obtained when the effect of sgs TRI is included. For this flame, the optically thin assumption is found to be valid, in agreement with previous investigations. The calculated radiative fraction is found to be close to the experimental value, but the postprocessed heat release rate is larger than the experimental finding. This first LES-CSE study provides a good foundation for further developments, for example more detailed radiation models, for more complex fire cases.

Investigation of Sub-Grid Scale Turbulence-Radiation Interaction Effects on Turbulence Energy Transport and Varying Thermophysical Properties Using Large Eddy Simulation

Abstract The main objective of this article is to investigate sub-grid scale turbulence–radiation interaction (SGS TRI) effects on SGS turbulence kinetic energy (TKE) fluctuations and varying thermophysical properties in a partially premixed combustion system for a laboratory-piloted methane/air flame. The large eddy simulation approach is employed to simulate the turbulence of the compressible reactive flow. SGS quantities, including turbulent stress and fluxes of enthalpy and species in the sub-grid scale, are computed using the standard Smagorinsky–Lilly model. The radiative transfer equation is modeled by applying the spherical harmonic P1 approximation by considering the radiative heat source related to the SGS TRI contribution. Optically thin fluctuation approximation is utilized to simplify the radiative absorption term. A chemical reaction mechanism comprising 41 steps and 16 species is applied to model methane–air mixture combustion. Diffusion flamelet-generated manifolds are employed to govern the species transport equation. About 87% of TKE is resolved by applying the finest grid consisting of 1,822,580 cells. Impacts of SGS TRI on the spatially filtered density, eddy viscosity, SGS velocity and TKE, overall radiative emission, RMS temperature fluctuations, and nitrogen monoxide (NO) formation are studied. The results reveal that considering SGS TRI in the simulation leads to remarkable discrepancies, particularly in SGS velocity and TKE by 6.70% and 7.40%, respectively. Meanwhile, SGS density and eddy viscosity deviate negligibly in the presence of SGS TRI. Also, the filtered mass fraction of NO reduces up to 17.54% on average by considering TRI.

A detailed of analysis of joint soot volume fraction/temperature statistics in non-premixed jet flame: Implication for soot emission turbulence/radiation interaction

The main objective of this study is to provide insights on the modelling of soot emission Turbulence Radiation Interaction (TRI) through a detailed analysis of the joint SVF/temperature statistics in turbulent non-premixed flames. The analysis relies on RANS/Transported Probability Density Function (TPDF) simulation of the Sandia ethylene jet turbulent non-premixed flame. RANS/TPDF predictions reproduce with fidelity the exhaustive set of experimental data, namely the flame structure, the soot and temperature statistics including the soot intermittency, and the radiative outputs. The marginal PDF of SVF exhibits the typical shape of an intermittent process and can be decomposed into a non-sooting mode and a sooting mode, with this latter being reasonably represented by a truncated Gaussian PDF. As expected, the marginal PDF of temperature can be reasonably represented by a clipped Gaussian PDF whereas that of the soot temperature follows a significantly narrower PDF as soot is fully oxidized in the fuel rich side of the flame. Analysis of the joint SVF/temperature PDF shows that both SVF/temperature and SVF/soot temperature first-order cross correlations are positive in the soot formation region and becomes strongly negative in the region of strong soot emission. The TPDF predictions are analysed a priori to design a new joint presumed PDF (PPDF) of SVF and temperature through the Gaussian copula. The Gaussian copula-based PPDF depends on six parameters, namely the five related to the marginal PDFs of SVF and temperature added to the first-order cross-correlation coefficient of SVF and temperature. Mean soot emission term computed with the copula-based PPDF agrees very well with the reference TPDF solution over the entire flame and improves significantly the solutions obtained by neglecting either TRI or the correlation between SVF and temperature. These results indicate that the Gaussian copula-based joint PPDF model is a good candidate to interpret soot emission-related experimental data in turbulent non-premixed flames.

Numerical investigation on turbulence-radiation interaction in the UMD turbulent line fires

This work aims to study the turbulence-radiation interaction (TRI) in the UMD turbulent line fires with increasing levels of underventilation until strong local extinction occurs. Large eddy simulations (LES) are performed with a radiative flamelet/progress variable model to account for local extinction events. The effects of the resolved-scale and subgrid-scale (SGS) TRI on global radiant fractions and local radiative power are quantified. Results show that the total TRI makes a considerable contribution to radiation for all oxygen concentrations (XO2) tested in this work. Specifically, in the most under-ventilated case of XO2=13% featuring strong local extinction, more than 95% of the global radiant fraction is induced by the TRI. More interestingly, it is found that on a regular LES grid resolving more than 80% of the turbulent kinetic energy, the resolved-scale TRI plays a leading role in the XO2=21% case, while the SGS TRI becomes dominant in the XO2=13% case. Further investigations reveal that the temperature self-correlation and the absorption coefficient-temperature correlation are the key components for TRI, which is confirmed by the corresponding conditional statistics. This study highlights the importance of including an SGS TRI model to accurately predict radiation fields in LES of fires with local extinction.

A two-zone subgrid flame model for predicting radiant emission from fires

A simple two-zone subgrid temperature distribution is developed to model turbulence-radiation interaction in the Fire Dynamics Simulator. The new approach enforces consistency between the subgrid flame distribution and the heat release rate from the eddy dissipation model. To investigate the performance of the new model, we simulate the University of Maryland turbulent line burner and compare global radiative fraction as a function of coflow oxygen dilution. The results suggest that the model can improve grid resolution dependence in prediction of radiative emissions. Sensitivity of the model predictions to the assumed fuel carbon to CO fraction (a parameter of the two-step fast chemistry scheme) was found to be significant, which is not surprising since this parameter essentially controls the in-flame soot concentration and generally has a first-order effect on resolved emission. Four different model variations were tested, corresponding to varying degrees of complexity in modeling the subgrid correlations for the absorption coefficient. Differences in results between model implementations were on par with the variation in results obtained with different path lengths applied to the gray gas model for the absorption coefficient.

Simulation of Pulverized Coal Combustion Process Considering Turbulence-Radiation Interaction.

In this paper, the combustion process of a 350 MW opposite pulverized coal combustion boiler is studied using a combined field test and numerical simulation. Considering the turbulence-radiation interaction (TRI) effect caused by turbulent particle pulsation during the combustion process, the TRI model is modified by introducing optical factors. The interaction between the optical thickness and the radiation pulsation of pulverized coal particles during the combustion process is studied to learn the influences of turbulent particle combustion on radiative heat transfer. According to the analysis of the TRI influence on pulverized coal combustion, the temperature of the cross section decreases with TRI. Compared with the traditional radiation model calculation, the time-average temperature of the cross section is reduced by 49.85 K, and the deviation from the experimental data is reduced by 1.54%. It is proved that the temperature distribution in the furnace considering TRI is closer to reality. In the numerical simulation of the TRI sensitivity to the optical thickness of the particles, it is found that with TRI, the increase of optical thickness will strengthen the effect of turbulent kinetic energy in the combustion area. Due to the increase in particle density, the content of the main radiation medium produced by combustion, H2O and CO2, increases and the radiation effect in the furnace is enhanced. In the fire area with dense particles, the average temperature of the cross section increases by 28 K. Hence, the change in optical thickness causes the change of TRI parameters, and the TRI effect becomes more significant with the increase of particle load.

Assessment of turbulence–radiation interactions in spectral radiative emission–absorption modeling

Assessment of turbulence–radiation interactions in spectral radiative emission–absorption modeling

The importance of accurately modelling soot and radiation coupling in laminar and laboratory-scale turbulent diffusion flames

Ultrafine and highly light absorbing soot particles have been known to be harmful to human health and a major climate forcer. Understanding the mechanisms of soot formation and developing reliable and robust soot models are essential to design efficient and clean combustion systems. Thermal radiation is often the dominant heat transfer mode in both turbulent and laminar flames and affects soot formation through its influence on temperature. Significant progress has been made in modelling soot formation, thermal radiation transfer, and turbulence radiation interaction in hydrocarbon flames over the last three decades. However, the optically thin approximation has still been used to simplify thermal radiation transfer modelling and turbulence radiation interaction has often been neglected in soot formation modelling without proper justification. In this paper, soot models developed for laminar and laboratory scale turbulent jet diffusion flames, non-gray radiative heat transfer models, and turbulence radiation interaction models are reviewed. The aim of this review is to highlight the importance of thermal radiation heat transfer in both laminar and turbulent diffusion flames and turbulence radiation interaction in turbulent jet diffusion flames. Use of the optically thin approximation is often not justified even in laminar diffusion flames. This is particularly true in diffusion flames at elevated pressures or microgravity. Although the prediction of soot in both laminar and turbulent flames is primarily influenced by soot kinetic model, it is also significantly affected by aerosol dynamic model, thermal radiation heat transfer model, and turbulence radiation interaction model. It is important to accurately model soot aerosol dynamics and non-gray thermal radiation transfer in laminar diffusion flames and additionally to adequately model turbulence radiation interaction in turbulent flames for the purpose of validating reaction mechanisms and soot kinetic models.

Large-eddy simulation of lab-scale ethylene buoyant diffusion flames: Effects of subgrid turbulence/soot production interaction and radiation models

Large eddy simulation of the 15-kW ethylene buoyant diffusion flames burning under O2/N2 oxidizers with oxygen concentrations, XO2, of 0.209, 0.168 and 0.152, investigated experimentally at FM Global, are performed by using a flamelet/presumed filtered density function (FDF) combustion model, an acetylene/benzene-based soot production model and the rank-correlated full-spectrum k (RCFSK) with non-grey soot as radiative property model. The presumed FDF considers the spatial intermittent nature of soot and the coupling between soot quantities and mixture fraction. Subgrid-scale (SGS) absorption turbulence/radiation interaction (TRI) is neglected whereas the filtered emission term is modelled from the presumed FDF approach. Model predictions reproduce with fidelity the flame structure, the soot production dynamics and the radiative loss for the three flames. Model results show that soot and radiating gas have a similar contribution to radiation for the flame burning in air and the contribution of radiating gas becomes increasingly dominant with reducing XO2 due to a decrease in soot volume fraction (SVF) throughout the overall flames. A sensitivity study is performed on the influence of the modelling approach of SGS turbulence/soot production interaction (TSI) and radiative heat transfer. Model predictions show that neglecting the coupling between soot quantities and mixture fraction overestimates substantially soot oxidation, resulting in significantly lower soot production and radiative loss. For the present LES where more than 80% of temperature variance is resolved, disregarding the SGS emission TRI has also a significant impact with an overprediction of mean and rms SVF by a factor of two. On the other hand, the approximation of grey soot is valid for these lab-scale flames characterized by moderate optical thicknesses and the use of a non-grey weighted-sum-of-grey-gases (WSGG) model coupled to the grey soot approximation represents an alternative to the full-spectrum k-distribution models.

Using a flamelet generated manifold method in transported probability density function modeling of soot formation and thermal radiation

The simple semi-empirical precursor soot model of Brookes and Moss based on the soot number density and soot mass concentration is adopted in a transported probability density function (PDF) method for turbulent diffusion flames. The gas phase chemistry is described by a flamelet generated manifold (FGM) based on the mixture fraction, progress variable and enthalpy loss. The accuracy of the FGM method is validated by using flamelet solutions that are not included in the generating set of the FGM. To account for the radiative heat transfer in the flames, we use a non-gray weighted sum of gray gases model for the gas radiation and a gray soot radiation model. Turbulence–radiation interaction is closed at the level of the optically thin fluctuation approximation and the Reynolds averaged radiative transfer equation is solved by means of a discrete transfer method. The proposed modeling approach is applied in simulations of two turbulent non-premixed methane–air flames at one bar and three bar pressure, respectively. Predictions of the mean temperature and mean soot volume fraction are in good agreement with the measurements in the one bar flame. In the higher pressure flame the mean soot volume fraction is over predicted. For this flame, simulation results using the semi-empirical model of Lindstedt provided better agreement with the measurements. The main difference between the Brookes and Moss model and the Lindstedt model is the nine-times increased soot particle agglomeration rate of the latter. When using the same increased agglomeration rate parameter in the Brookes and Moss model the results become virtually identical. The negligible molecular diffusion of the soot was accounted for by neglecting mean molecular diffusion of the soot variables and by greatly reducing their micro-mixing. The effect of this differential soot diffusion on the mean soot volume fraction is found to be small, but it is significant for the variance of the soot volume fraction.

Local contributions of resolved and subgrid turbulence-radiation interaction in LES/presumed FDF modelling of large-scale methanol pool fires

The present study aims to quantify individual contributions to turbulence-radiation interaction (TRI) in LES of large methanol pool fires with diameters ranging from 1m to 4m. These simulations feature a steady laminar flamelet (SLF)/presumed Filtered Density Function (FDF) combustion model and the Rank Correlated Full Spectrum k-distribution (RC FSK) for spectral gas radiation. This quantification has a fundamental interest to, on the one hand, understand how these contributions affect the radiative heat transfer from the flame to the surroundings that controls fire growth in industrial fires and, on the other hand, to provide guidance for the development of simplified models for TRI. This analysis is accomplished by exercising a well-validated fire model and using a scaling strategy that preserves the relative contributions of resolved-scale and subgrid-scale (SGS) scalar fluctuations. Large-scale non-luminous pool fires feature large optical thicknesses that increase significantly with the pool diameter. As a result, applying the optically-thin fluctuation approximation (OTFA) leads to a noticeable overestimation of the global radiative loss. The fire plume dynamics is governed by the puffing process whose frequency decreases with the pool size, and, in turn, affects the levels of turbulent intensity. As a result, emission TRI increases in the continuous flame and decreases in the high intermittent flame regions. The contribution of SGS emission TRI to global emission TRI becomes progressively dominant as the pool size increases. The largest contributions to both resolved-scale and SGS emission TRI are the temperature self-correlation and the absorption coefficient-temperature cross correlation, although this latter becomes negligible at the subgrid scale for pool diameters larger than 1m. In addition, the SGS absorption-coefficient self-correlation is also found to be negligible.

Modelling of MILD combustion in a lab-scale furnace with an extended FGM model including turbulence–radiation interaction

The flamelet generated manifold (FGM) model is suitable for moderate or intense low oxygen dilution (MILD) combustion provided the flamelets underlying the manifold include the effects of strong dilution by products of the fuel/oxidizer mixture. Here we propose such an extended model based on the use of non-premixed flamelets diluted at the airside and develop its application to non-adiabatic combustion in a lab-scale furnace. The extended model is referred to as diluted air FGM (DA-FGM) model. In the DA-FGM model in addition to mixture fraction, progress variable and scaled enthalpy loss, one additional controlling parameter named air dilution level, is introduced leading to a four-dimensional lookup table for laminar flames. For turbulent flames also variances of mixture fraction and progress variable are taken into account as independent variables leading to a six-dimensional table. Using a RANS approach implemented in OpenFOAM-2.3.1, the DA-FGM model has been applied to MILD combustion of Dutch natural gas in a lab-scale furnace operated at a thermal power 9 kW and at equivalence ratio 0.8. Radiation is described using a weighted-sum-of-gray-gases (WSGG) model. The validation study is mainly done using a grey WSGG model with TRI taken into account. The relative importance of including turbulence radiation interaction (TRI) and spectral treatment of radiative transfer is also studied. The predicted velocity and temperature statistics are in good agreement with the experimental LDA and CARS data provided not only the mixture fraction fluctuations but also the progress variable fluctuations are taken into account.

Assessing subfilter-scale turbulence–radiation interaction in a large-scale ethanol pool fire

Assessing subfilter-scale turbulence–radiation interaction in a large-scale ethanol pool fire

Modelling turbulent heat flux accounting for Turbulence-Radiation Interactions

The present work investigates the modeling of turbulent heat transfer in flows where radiative and convective heat transfer are coupled. In high temperature radiatively participating flows, radiation is the most relevant heat transfer mechanism and, due to its non-locality, it causes counter intuitive interactions with the turbulent temperature field. These so-called Turbulence-Radiation Interactions (TRI) largely affect the temperature field, modifying substantially the turbulent heat transfer. Therefore, in the context of modeling (RANS/LES), these interactions require a closure model. This work provides the inclusion of TRI in the modeling of the turbulent heat transfer by adopting a unique approach which consists in approximating the fluctuations of the radiative field with temperature fluctuations only. Based on this approximation, coefficients of proportionality are employed in order to close the unknown terms in the relevant model equations. A closed form of all radiation-temperature-velocity correlation is explicitly derived depending on the chosen turbulent heat transfer model. This model is applied to a standard two-equation turbulent heat transfer closure and used to reproduce results obtained with high-fidelity DNS simulations. While a standard approach (i.e., neglecting TRI) is not able to correctly predict the DNS data, the new model’s results shows exceptional agreement with the high-fidelity data. This clearly proves the validity (and the necessity) of the proposed model in non-reactive, radiative turbulent flows.

Influence of turbulence-radiation interaction on radiative heat transfer to furnace wall and temperature distribution in large-scale industrial furnaces enveloping hydrocarbon flame

Influence of turbulence-radiation interaction on radiative heat transfer to furnace wall and temperature distribution in large-scale industrial furnaces enveloping hydrocarbon flame

Assessing individual time-averaged emission and absorption correlations for large-scale turbulent pool fires

A fundamental investigation of turbulence-radiation interaction (TRI) is presented in this paper, focused on the correlations that arise in the time-averaging of the radiative absorption and radiative emission. The analyses are based on transient data generated from fully-coupled, high-resolution large eddy simulations of a set of turbulent, large-scale pool fires. Coefficients that quantify the magnitude of the individual correlation are identified and, using the transient data, they are computed throughout the computational domain and compared to one another. Results show that turbulent fluctuations lead to an increase in both the mean emission and the mean absorption. For the latter, the increase is particularly pronounced in the hot gas plume region. In this region, the cross-correlation between the local absorption coefficient and the incident radiation accounts for no more than 15% of the total absorption, and disregarding it leads to negligible errors in the prediction of the mean radiation loss, which points to the applicability of the optically thin fluctuation approximation. Among the correlations in the time-averaged emission term, the temperature autocorrelation is the most important one, followed by the absorption coefficient-temperature cross-correlation. However, radiative transfer calculations considering different closure methods for the emission term evidence that accounting for the temperature autocorrelation alone is not sufficient for an accurate solution of the mean radiation field. Instead, all emission-TRI correlations must be considered for a proper representation of the mean radiative emission.

Numerical assessment of the effect of inflow turbulators on the thermal behavior of a combustion chamber

This work numerically investigates the effects of turbulators at the air and fuel (methane) inlets on the thermal behavior of a combustion chamber. Conservation equations for mass, momentum, energy, gaseous chemical species, soot, and temperature fluctuation variance in cylindrical axysimmetric co-ordinates were solved using the finite volume method. Chemical reaction rates were computed through the Arrhenius-Magnussen model, with two-step combustion reaction. The turbulence closure model, to compute the turbulent viscosity, was the standard k-?. The modelling of turbulence-radiation interactions considered the absorption coefficient-temperature correlation and the temperature self-correlation. The radiative heat source was calculated using the discrete ordinates method, considering the weighted-sum-of-gray-gases model with the superposition method to compute the radiation from the gaseous species and soot. The effect of inlet turbulators was studied by varying the turbulence intensity (TI) of both inlet streams (air and fuel), encompassing mild to severe turbulators (TI = 3%, 6%, 15%, and 20%). The results showed that temperature and radiative heat source fields, and heat transfer rates on the chamber wall and radiative fraction were importantly affected by the different turbulators intensities (e. g. radiative fraction was increased from 20.6% to 32.8% when the TI was varied from 3% to 20%). Comparisons of results obtained when turbulence-radiation interactions modelling was neglected in relation to results obtained when turbulence-radiation interactions modelling was computed showed that turbulence-radiation interactions influenced the thermal field (temperature and radiative exchange) in a similar way independently of the turbulator intensity (e. g. radiative fraction decreased 20% when turbulence-radiation interactions modelling was neglected, for both turbulators intensities).

Large Eddy Simulation of medium-scale methanol pool fires - effects of pool boundary conditions

This article reports Large Eddy Simulation (LES) of 30 cm diameter methanol pool fires in order to investigate the effects of the burner boundary conditions on the pool dynamics. The numerical model involves state-of-the-art subgrid scale (SGS) sub-models for mixing, combustion and turbulence-radiation interaction with all the model constants computed dynamically. The non-adiabatic steady laminar flamelet (SLF)/presumed filtered density function (FDF) combustion model is used whereas the radiation model combines the Rank Correlated Full Spectrum k-distribution (RCFSK) for spectral radiation with the finite volume method (FVM) as radiative transfer equation (RTE) solver. The baseline case considers a burner located one-pool diameter above the floor and a fuel lip height of 1 cm as specified in the experiments used for model validation. Model predictions for puffing frequency, mean and rms temperature and velocity, mean molar fractions of major species, radiative loss to the surrounding and radiative and total heat feedback to the fuel surface are in good agreement with the available experimental data. Two other burner boundary conditions are designed. The first disregards the fuel lip height while keeping the burner located one-pool diameter above the floor. The second modifies the first configuration by assuming that the burner rim is floor flush. Model results show that the burner boundary conditions affect significantly the flow structure and the pool fire dynamics by altering the flame base instability near the edge of the pan. A non-zero fuel lip height produces significantly wider and shorter flames which affect the heat feedback toward the fuel surface whereas altering the air entrainment at the pool basis enhances substantially the puffing frequency. This shows that the experimental burner boundary conditions have to be reproduced scrupulously for relevant model validations.

Assessment of subfilter-scale turbulence-radiation interaction in non-luminous pool fires

The objective of this work is to investigate the effects of Turbulence-Radiation Interaction (TRI) in Large Eddy Simulation (LES) of ethanol pool fires ranging from medium to large sizes. LES of three pool fires are performed using a steady laminar flamelet (SLF)/presumed Filtered Density Function (FDF) combustion model and the Rank Correlated Full Spectrum k-distribution (RC FSK) for spectral gas radiation. Subfilter-scale (SGS) absorption TRI is neglected whereas the filtered emission term is modelled from the presumed FDF approach. The baseline case is the 0.5 m diameter pool fire investigated experimentally by Fischer et al. (Combust. Flame 70 (1987) 291–306). The two other synthetic pool fires were generated by scaling the pool diameter by factors of 2 and 4 while maintaining the fuel mass burning rate per unit area unchanged. Predictions in terms of temperature, mole fractions of radiating species, flame puffing frequency and radiant fraction are in good agreement with the available experimental data. Model results show that TRI accounts for at least 80% of the radiative loss. For LES that resolves more than 80% of the temperature variance, the SFS TRI contributes for a non-negligible part to radiative loss and its contribution increases significantly with the pool size. Neglecting it affects significantly the predictions of the fire plume structure and radiative heat flux even for medium-scale pool fires.