Computational Fluid Dynamics (CFD) Assessment of the Internal Flue Gases Recirculation (IFGR) Applied to Gas Microturbine in the Context of More Hydrogen-Enriched Fuel Use

Abstract

Renewable energy is a promising substitute for fossil fuels when corelated with P2G technology. To optimise P2G efficiency, there is a need to increase hydrogen fraction in the fuel stream. Simultaneously gas microturbines are widely applied in many industry sectors. These devices are often equipped with diffusion combustors. This situation was investigated in this paper. The P2G and gas microturbines may be integrated together in the future leading to the application of hydrogen-enriched fuel. Hydrogen-enriched fuel causes increase in combustion temperature and velocity. In a nonadapted combustor, these phenomena could result in an increase of NOx emissions and risk of material overheating and failure. In order to adapt the combustors for hydrogen-enriched fuel, the concept of autonomous internal flue gases recirculation (IFGR) system was applied to this issue. In this paper, the IFGR system applied to gas microturbine was studied in terms of hydrogen-enriched fuel application. The obtained exhaust gases recirculation ratios were too low to affect the combustion process as it was expected. The observed combustion modifications in the combustor were hardly linked to the air flow modification in the liner, due to IFGR system implementation. After CFD studies, the proposed IFGR system does not seem to provide the expected effects.

1. Introduction

1.1. Energy Crisis and Integration of Power-to-Gas and Gas Microturbine Technologies

In recent years, the world has been confronted with an energy crisis, most recently accentuated by the Russian–Ukrainian war. The clean transition implicates an increase in variable electricity supply from renewable sources. There are various kinds of renewable energy sources: wind turbines, hydroelectric power plants, photovoltaic panels, etc. Renewable sources of energy offer many advantages, such as “unlimited” energy production. The major operational problem of the renewable sources of energy is variable electricity supply depending on the external conditions. This inconvenience could be eliminated by storing the excess energy when the conditions are suitable for energy production and by using this stored energy during high-demand periods. It is possible to transform the excess of energy (electricity) into hydrogen by the hydrolysis process, and then to transform this hydrogen into methane by the methanation process. Hydrogen and methane gases can be mixed together and stored as an energy reserve. This technology is called “Power-To-Gas” (“P2G”) [1]. From an efficiency point of view, the presence of methane in this technology causes a reduction in efficiency, linked to the methanation process. It would be more efficient to use pure hydrogen. However, hydrogen presents many challenges in terms of its storage and application in energy devices [2]. This is the main reason why methane is produced in this technology. To increase the efficiency of the use of renewable energy sources, the participation of hydrogen in renewable fuel, derived from P2G technology, needs to be increased.

Gas microturbine devices are another interesting technology, which is intensively used. These devices may be applied in various industry sectors: in aviation to power drones or as an Auxiliary Power Unit (“APU”) in classical aircraft, in the automotive industry to extend vehicle range (e.g., Pininfarina H600, Jaguar CX75, ect.), to generate energy for houses and/or emergency applications, and in the military industry to power a few models of tanks (e.g., M1 Adams, M1150 Assault Breacher Vehicle, ect.) [3]. The gas microturbine devices have great success in various industry sectors thanks to their advantages, including low emissions, low noise levels, cheap exploitation, limited number of moving parts, etc. [4].

Since there is a focus on P2G and gas microturbines technology research, it can be beneficial to integrate both technologies together and run the microturbine with renewable fuel. To optimize the use of P2G technology, it is necessary to increase the fraction of hydrogen in the fuel blend. That will imply a need to adapt the combustion chamber of the gas microturbine devices capable of accommodating the increased fraction of hydrogen in the reference fuel (here methane). The use of hydrogen-enriched fuel is associated with higher combustion temperatures and higher flame spread speed [5]. Unmodified combustors would be susceptible to material overheating, NOx emissions increase, and flame flashbacks. All of these phenomena could have undesirable effects on the combustion chamber work parameters and its mechanical integrity. In the future, the need to adapt the combustion chamber of gas microturbine devices to more hydrogen-enriched fuel would seem to be necessary.

Gas microturbine devices may be equipped with a premixed or diffusion combustion chamber. The use of premixed combustors is more suitable in terms of NOx emissions reduction. This kind of combustor has recently been developed and applied in modern devices. The inconvenience of this type of combustion chamber is the low flame stability in the combustion zone, as the introduction of hydrogen into the reference fuel could cause flame flashbacks into the air–fuel mixture preparation zone, which could damage this zone. The diffusion combustion chambers are an older generation of combustors. The combustion zone is created by a diffusive flame, extremely rich in the top part of the liner, and then extremely poor in the back part, due to air progressive addition. This kind of combustion is very stable, and it permits easily controlling the power range of the gas microturbine devices. Diffusion type combustors are still applied in various industries and there are many devices equipped with this kind of combustors. In this paper, the diffusion type combustor will be studied.

1.2. Internal Flue Gases Recirculation (IFGR) System Applied to Gas Microturbine Devices

According to the above, there is a need to adapt the classical (diffusion) combustion chamber of the gas microturbine devices in order to reduce combustion temperatures and to move the combustion zone toward the exhaust part of the combustion chamber. In order to propose a combustor modification to solve these problems (linked to the use of more hydrogen-enriched fuel), a literature review on the topic was performed and is presented below.

Felix Guethe et al. [6] investigated the impact of exhaust gas recirculation in a gas turbine, in terms of the modification of combustion parameters. The exhaust gases were recirculated from the gas turbine exhaust system to its air intake. The presence of exhaust gases caused a reduction in the amount of oxygen in the combustion zone of the combustion chamber. At the same time, an increase in the amount of carbon dioxide was indicated. Oxygen is a radical in the combustion zone. Its reduction in the combustion zone results in an inhibition of combustion processes. In the same work, it was indicated that carbon dioxide is not a perfectly inert gas in the combustion zone; the carbon dioxide may react in the combustion with other(s) species. The studies allowed identifying the carbon dioxide reaction in the combustion zone: OH + CO ↔ H + CO₂. Due to this reaction, in the combustion zone, the concentration of the hydrogen radical is reduced, while the concentrations of OH and CO are increased. Another research study (Fengsham Liu et al. [7]) denoted the next reaction in the combustion zone: H + O₂ ↔ O + OH. According to the above-presented research, the presence of carbon dioxide limits the concentration of hydrogen radical in the combustion zone, which, in turn, limits the concentration of O and OH species. The H, O and OH species are radicals and their concentrations in the combustion zone have a non-negligible effect on the combustion process. According to [6,7], the reintroduction of exhaust gases into the combustion zone of the combustion chamber limits the concentration of O, H and OH radicals in the combustion chamber, which has a direct impact on the combustion process. It causes the combustion temperature and laminar flame speed decrease.

Mario Ditaranto et al. in [8] investigated the impact of the dilutant heat capacity added to the combustion chamber of the gas turbine on the combustion process. According to this paper, the use of steam is more efficient than the use of nitrogen. The introduction of an inert gas with a heat capacity higher than that of the air into the combustion zone allows limiting the combustion temperature. Exhaust gases contain a part of pure exhaust gases. Pure exhaust gases are composed of carbon dioxide and steam. The heat capacity at a constant pressure of these two species is greater than that of air entering the combustion zone, when the temperature is above 550 K [9,10,11]. In the combustion chamber, in the vicinity of the liner, the temperature ranges from 770 K to 2300 K [12]. Under these conditions, the carbon dioxide and steam heat capacity at constant pressure are higher than that of the air entering the combustor, meaning that the presence of exhaust gases in the combustor permits releasing the same energy (from fuel) with a lower temperature increase than without exhaust gases. The application of exhaust gases recirculation to gas turbine permits the combustion peak temperature and its gradient to be reduced. According to van’t Hoff’s law, if the reaction temperature is reduced, then the combustion rate is reduced. To conclude, it may be denoted that the introduction of exhaust gases into the combustion zone allows reducing the combustion temperature and the flame speed.

Baolo Shi et al. [13] investigations were performed on a modified furnace burner. The burner was equipped with a pipe system that allowed part of the exhaust gases to be taken from the burner outlet and reintroduced into the air–fuel mixing zone inside of the burner. When the exhaust gas pipe system was operating, an important reduction in combustion temperature and NOx emissions was observed. This reduction is linked to the increase in the combustion volume caused by the addition of exhaust gases to the fresh air–fuel mixture.

According to the research mentioned above, the introduction of exhaust gases into the combustion zone in the combustion chamber of a gas microturbine may be an interesting solution for the use of a more hydrogen-enriched fuel. The exhaust gases introduced to the combustion zone act chemically (limiting the concentrations of H, O and OH radical concentrations) and physically (increasing the heat capacity and the combustion volume), allowing the reduction of the combustion peak temperature and its gradient, the reduction of NOx emissions, and the limitation of the flame speed. All these phenomena are suitable for the application of hydrogen-enriched fuel to gas microturbine devices. When exhaust gases recirculation is applied to a gas microturbine combustor, combustion temperature, flame speed, and NOx emissions could be reduced compared to the combustor without exhaust gas recirculation. The addition of hydrogen into the reference fuel could provoke a back to the nominal combustion parameters. In this paper this solution will be investigated and applied to the diffusion type gas microturbine combustor.

To study this solution, autonomous internal exhaust gas recirculation systems were created and integrated into gas microturbine combustor. This system is called Internal Flue Gases Recirculation (“IFGR”). This reference and modified combustors are exposed in the second part of this paper.

Unfortunately, other phenomena may have negative effects on the combustion process when the exhaust gas recirculation system is applied.

In the exhaust gases coming out of a gas microturbine, there is some carbon monoxide. This fact should be taken into consideration. According to M. Jadidi et al. [14] and S. Taamallah et al. [15], the adiabatic combustion temperature of the carbon monoxide is about 2450 K. This combustion temperature is higher than that for methane (2250 K) and even for hydrogen (2400 K). The recirculation of exhaust gases, by the presence of carbon monoxide, may result in an increase of the combustion temperature compared to a case without exhaust gases recirculation.

Another unsuitable phenomenon is that some of the fresh air coming from the compressor will be replaced by hot exhaust gases. This will cause an increase of the enthalpy in the top part of the liner, and it may result in a peak and gradient temperature rise.

Another research study [12] demonstrates a concept of combustion chamber for an aviation gas microturbine with a system able to modify the distribution and the size of the holes present on the combustion liner. This concept could permit the liner shape to be adapted as a function of the gas microturbine operating conditions. This concept highlights the necessity of taking into consideration the repartition of the air entering the liner because it potentially can impact the lambda ration in various sections of the liner. When applying the IFGR system, an air repartition modification may occur and disturb the combustion process. The lambda ratio modification may cause an increase in terms of combustion temperature and NOx emissions.

The last three phenomena show that the IFGR system applied to the gas microturbine combustor may have also unsuitable effects on the combustion process compared to the investigated solution.

According to the literature review and critical thinking, the autonomous internal exhaust gases recirculation (IFGR) system applied to a gas microturbine may have positive or negative effects in the context of applying hydrogen-enriched fuel. The research presented in the rest of this paper allowed assessing the effect of the IFGR concept applied to gas microturbine devices in the context of the use of hydrogen-enriched fuel.

1.3. Novelty of the Research

The potential integration of P2G and gas microturbine technologies in the near future may lead to the use of hydrogen-enriched fuel to power gas microturbine devices. Many of the gas microturbine devices are equipped with a diffusion type combustor. The effects of applying hydrogen-enriched fuel on a nonadapted combustor may cause thermal failure and an increase of NOx emissions. According to the above literature review, the application of an autonomous internal exhaust gases (flue) recirculation (IFGR) system to a gas microturbine combustor may be a potential solution to solve these problems. This solution may be an alternative to other kinds of research focused on the hydrogen combustion, as for example the Moderate and Intense Low-oxygen Dilution (MILD) combustion applied to the gas turbine [16,17,18,19,20,21,22]. The aim of this paper is to numerically assess the IFGR concept applied to gas microturbine diffusion combustor in the context of the use of hydrogen-enriched fuel. A methane-powered diffusion type combustor was designed and then modified by adding an IFGR system, allowing the autonomous recirculation of exhaust gases inside the combustion chamber. The following issues were investigated:

• the possibility of designing an operating autonomous IFGR system for gas microturbine combustor;
• variations in the main operating parameters of the combustor due to the implementation of the IFGR system (the total pressure drop and the exhaust gases total temperature);
• the influence of the IFGR system on the temperature field in the combustion chamber;
• the influence of the IFGR system on CO and NOx concentrations in exhaust gases.

Each of the evaluations was conducted for the hydrogen mass fraction in the reference fuel (here, methane) ranging from 0 to 0.5. This numerical investigation of the IFGR system applied to the gas microturbine is an extension of preliminary studies performed previously and described in [23,24].

The novelty of this study is the numerical assessment of an IFGR system applied to a gas microturbine diffusion combustor powered by hydrogen-enriched fuel.

2. Case Study Combustor and Operating Conditions

To obtain the operational parameters, an in-house reference gas microturbine combustor was designed and respective three-D models were generated. The reference case combustor was assumed to work inside a 40 kW gas microturbine powered by methane. The combustor main operating parameters are listed in

Table 1. The referential gas microturbine combustion chamber work parameters.

Parameters	Combustor Inlet	Combustor Outlet
$p^{*} \left[\mathrm{P}\mathrm{a}\right]$	324,992	311,992
$p \left[\mathrm{P}\mathrm{a}\right]$	306,584	301,133
$T^{*} \left[\mathrm{K}\right]$	433	1185
$T \left[\mathrm{K}\right]$	426	1175
$c [\mathrm{m}/\mathrm{s}]$	120	155
$\dot{m}=0.251\frac{\mathrm{k}\mathrm{g}}{\mathrm{s}}$ $c_s=[see \mathrm{Table} 4]\frac{\mathrm{k}\mathrm{g}}{\mathrm{s}}$

. The design and description of the reference case combustor are available in Figure 1.

Next, the reference combustor was modified in order to integrate the IFGR system inside. The IFGR system is a set of pipes dedicated to transport a part of exhaust gases from the back part of the liner into the inlet part of the liner. The challenge consists in designing an IFGR system capable of working autonomously. This goal is difficult to reach because of the total pressure drop that occurs in the combustion chamber. Despite this difficulty, two operating IFGR pipe systems were designed for the reference combustor. The two IFGR combustors are presented below.

The first IFGR combustor is based on a set of six pipes located inside the combustor. This IFGR system permits the removal of part of the exhaust gases from the outlet zone of the liner and the introduction of these exhaust gases into the air–fuel mixing zone. Recirculated exhaust gas flow is possible because of the difference between the total pressure at the liner outlet and the static pressure at the top of the liner. In the outlet part of the liner, the recirculation system is collecting the total pressure of the exhaust gases. The total pressure of the exhaust gases in the outlet part of the liner is inferior to the total pressure of gases in the top part of the liner (due to the total pressure drop). Taking into consideration only the static pressure of the gases at the top part of the liner (without the dynamic pressure), this pressure may be locally lower than the total pressure of exhaust gases in the end part of the liner. An adequate design of the recirculating pipes in the top part of the liner diminishes the sensitivity to the dynamic pressure of the environment. The recirculated exhaust gases are driven by this pressure difference. This IFGR system is presented in Figure 2 and is designated as the “Case A” combustor.

The second IFGR combustor is based on modified mixing pipes. The mixing pipes were modified in order to add an exhaust gases collector, and the internal volume of the mixing pipe is designed as a venturi. The exhaust gases are recirculated inside the mixing pipes. The recirculated exhaust gases flow is possible because of the total-static pressure difference (as in case A) and because of the venturi-shaped channel inside the mixing pipes. This design of the IFGR system should be more effective than the case A in terms of recirculated exhaust gases mass flow. This second IFGR system is presented in the Figure 3 and is designated as the “Case B” combustor.

Many other designs of the IFGR system were elaborated and tested in preliminary flow simulations. Only the IFGR systems A and B permitted obtaining autonomous exhaust gases recirculation inside of the liner. This step of research was very time consuming and difficult to achieve.

3. Numerical Methods

3.1. Computational Domain

The tridimensional models of the reference and two IFGR modified combustors were created using Solid Edge and Ansys DesignModeler software [25]. The combustors 3D models were then exported to the Ansys Meshing software in order to generate the mesh. The generation of the combustors mesh was one of the most important steps in modelling the combustion and flow in the combustors. From an accuracy point of view, tetrahedral elements are often used to create meshes of gas microturbine simulations [26,27,28,29,30]. This kind of mesh element is known to be able to fill complex structures while keeping acceptable quality parameter values (skewness, orthogonality, and aspect ratio). Actually, many numerical studies demonstrated the advantage of using another mesh element geometry, the polyhedral elements. The main advantages of polyhedral elements are good accuracy and ability to fill complex geometry with values of quality parameters that are better than those for tetrahedral elements [31,32,33]. According to the above, the polyhedral elements were selected to create the combustors meshes. The size of the mesh elements was limited to a cell length of 0.8 mm. After the volume mesh generation, an improvement of that parameter was achieved by choosing the value of 0.45 as the desired minimum orthogonal quality. After this step, five layers of mesh in the near-war region were added to ensure that the Y+ value of 300 is not exceeded. The created meshes as described above were then applied to the CFD software to perform the desired combustion-flow simulations. According to the literature, the obtained meshes are sufficient to perform accurate simulations [26,27]. The meshes critical quality parameter values are presented in

Table 2. Quality parameters of the meshes.

Case	Number of Cells [Millions]	Maximum Aspect Ratio [-]	Maximum Skewness [-]	Minimum Orthogonal Quality [-]
Reference	5.8	38.3	0.895	0.435
Case A	7.0	35.8	0.900	0.432
Case B	6.4	62.0	0.895	0.200

. The view of the calculation domain and the cross-sections of the combustors meshes are shown in Figure 4.

3.2. Flow and Combustion Mathematical Models

The flow and combustion simulations were performed in the Ansys Fluent CFD software [34]. The following phenomena were modelled: turbulent flow, non-premixed combustion in the gas phase, radiative heat exchange, and pollutant formation (NOx and CO).

Turbulence phenomena that occur in combustors were modelled using the Reynolds Averaged Stoke–Navier (RANS) Realisable k-ε model [12,35]. This model is commonly applied for industrial issues. The main advantages of this model are low computational cost and acceptable results. In order to enhance the near-wall flow phenomena, the enhanced wall treatment was enabled. This option permits resolving the flow on the wall boundary when Y+ is less than or equal to one, or to use the enhanced wall function (for Y+ greater than one). The chosen flow and wall boundary models are a reasonable compromise when the wall phenomena are not of crucial importance, as in this study. The Reynolds’ Average Navier–Stokes (RANS) Equations (1) and (2) are presented below. Equation (3) presents the simplified relationship between the Reynolds stresses and the k and ε variables. Equations (4) and (5) are the transport equations for k and ε variables [34].

$$\frac{\partial \rho }{\partial t}+\frac{\partial }{\partial x_i}\left(\rho u_i\right)=0$$

(1)

$$\frac{\partial }{\partial t}\left(\rho u_i\right)+\frac{\partial }{\partial x_j}\left(\rho u_i{u}_j\right)=-\frac{\partial p}{\partial x_i}+\frac{\partial }{\partial x_j}\left[\mu \left(\frac{\partial u_i}{\partial x_j}+\frac{\partial u_j}{\partial x_i}-\frac{2}{3}{\delta }_{ij}\frac{\partial u_k}{\partial x_k}\right)\right]+\frac{\partial }{\partial x_j}\left(-\rho \overline{u_i^{\prime }{u}_j^{\prime }}\right)$$

(2)

$$-\rho \overline{u_i^{\prime }{u}_j^{\prime }}=f(k,\epsilon )$$

(3)

$$\frac{\partial }{\partial t}\left(\rho k\right)+\frac{\partial }{\partial x_j}\left(\rho k{u}_j\right)=\frac{\partial }{\partial x_j}\left[\left(\mu +\frac{{\mu }_t}{{\sigma }_k}\right)\frac{\partial k}{\partial x_j}\right]+G_k+G_b-\rho \epsilon -Y_M+S_k$$

(4)

$$\frac{\partial }{\partial t}\left(\rho \epsilon \right)+\frac{\partial }{\partial x_j}\left(\rho \epsilon u_j\right)=\frac{\partial }{\partial x_j}\left[\left(\mu +\frac{{\mu }_t}{{\sigma }_{\epsilon }}\right)\frac{\partial \epsilon }{\partial x_j}\right]+\rho C_{1}S\epsilon -\rho C_2\frac{{\epsilon }^2}{k+\sqrt{\nu \epsilon }}+C_{1\epsilon }\frac{\epsilon }{k}{C}_{3\epsilon }{G}_b+S_{\epsilon }$$

(5)

When considering a gas (micro)turbine combustor, the major mode of heat transfer is the radiation. In order to accurately simulate the combustion process, it is necessary to implement the radiation model. In the presented study, the Discrete Ordinate model (DO) [12,27] was utilized. This model permits considering the mixture of gases as a grey body. Steam and carbon dioxide, present in the combustion zone and in exhaust, are the main radiation absorbers and emitters. In order to accurately model the absorption coefficient, the Weighted Sum of Grey Gases Model (WSGGM) was applied [12]. The WSGGM approach uses experimentally obtained data [36]. The wall emissivity coefficient was set as unity (like black body), which is a common recommendation when modelling radiation in the gas (micro) turbine combustor [37].

The non-premixed steady diffusion flamelet model [38] was selected to model the turbulent combustion processes. This model considers that the flame in the combustion zone is a set of small laminar flames called “flamelets”. The use of this model permits considering the combustion kinetic effects and the turbulence phenomena (via the strain rates). Before running the calculation, a look-up table is created. This look-up table permits establishing the mean mass fraction of species, density and temperature in the combustion zone as a function of the mean mixture fraction, its variance, enthalpy and strain rate. For these last parameters, transport equations are solved during the simulation. Because of using the look-up table, the computational cost is strongly reduced compared to full chemistry calculations (just mixture fraction and mixture fraction variance transport equations are solved). This model permits, at the same time, obtaining accurate results. The application of this combustion model is a good compromise between computational cost and accuracy. The look-up table use is described by Equations (6)–(8). The mean mixture fraction and its variance transport equations are presented by Equations (9) and (10) [34].

$$\overline{Y_i}=\left\{\begin{array}{c}{g}_1\left(\overline{f},\overline{{f^{\prime }}^2},\overline{\chi }\right), \chi \ne 0 \\ g_2\left(\overline{f},\overline{{f^{\prime }}^2},\overline{H}\right), \chi =0\end{array}\right.$$

(6)

$$\overline{T}=g_3(\overline{f}, \overline{{f^{\prime }}^2}, \overline{\chi },\overline{H})$$

(7)

$$\overline{\rho }=g_4(\overline{f}, \overline{{f^{\prime }}^2}, \overline{\chi },\overline{H})$$

(8)

$$\frac{\partial }{\partial t}\left(\rho \overline{f}\right)+\nabla ·\left(\rho \overrightarrow{\nu }\overline{f}\right)=\nabla ·\left(\frac{{\mu }_l+{\mu }_t}{{\sigma }_t}\nabla \overline{f}\right)+S_m+S_{user}$$

(9)

$$\frac{\partial }{\partial t}\left(\rho \overline{{f^{\prime }}^2}\right)+\nabla ·\left(\rho \overrightarrow{\nu }\overline{{f^{\prime }}^2}\right)=\nabla ·\left(\frac{{\mu }_l+{\mu }_t}{{\sigma }_t}\nabla \overline{{f^{\prime }}^2}\right)+C_g{\mu }_t·{\left(\nabla \overline{f}\right)}^2-C_d\rho \frac{\epsilon }{k}\overline{{f^{\prime }}^2}+S_{user}$$

(10)

In order to generate the look-up table for the non-premixed combustion model described above, there is a need to use a combustion kinetic mechanism. For this study, the GRI-MECH 3.0 chemistry mechanism was selected [39,40]. This mechanism considers 53 species and 325 chemical reactions. Depending on the hydrogen participation in the fuel, the number of flamelets evolves from 65 to 347. Based on theses flamelet sets and individual probability density functions (PDF), the look-up table was generated for the combustion model.

In the consecutive step the boundary conditions were set. The inlets (fuel and air) were set as “mass flow” and the outlet was set as “pressure outlet”. The values of these boundary conditions were directly taken from the combustor design calculation parameters described in Table 1. At the domain’s walls, the no-slip condition was selected (no velocity on the wall). In the vicinity of the walls, the molecular viscosity affects the flow more than the turbulence phenomena.

Table 3. Implemented general boundary conditions.

Designation of the Boundary Condition	Type	Parameters
Air inlet	Mass Flow Inlet	Mass flow = 0.251 kg/s|Turbulent Intensity = 15%|Turbulent Viscosity Ratio = 10|Total Temperature = 433.834 K|Mean Mixture Fraction = 0|Mixture Fraction Variance = 0
Fuel inlet	Mass Flow Inlet	Mass flow = 0.004874 kg/s|Turbulent Intensity = 15%|Turbulent Viscosity Ratio = 10|Total Temperature = 300 K|Mean Mixture Fraction = 1|Mixture Fraction Variance = 0
Exhaust	Pressure Outlet	Static Pressure = 0 Pa|Turbulent Intensity = 15%|Turbulent Viscosity Ratio = 10|Backflow Total Temperature = 300 K|Mean Mixture Fraction = 0|Mixture Fraction Variance = 0
Wall	Wall	Stationary Wall|No Slip|No Heat Exchange|Internal Emissivity = 1|Opaque Wall|Diffuse Fraction of Radiation = 1
Operating conditions	-	Operating pressure = 301,133.803 Pa|Gravity off

summarizes the selected parameters for the boundary conditions.

The assumption of constant fuel enthalpy was assumed for the combustors when modifying fuel composition. The fuel energy supplied in the reference case (for methane) was naturally set as the reference enthalpy of the fuel flow. For pure methane, it is necessary to supply a fuel mass flow of 0.004874 kg/s. Equation (11) was used to calculate the mass flow rate of methane–hydrogen fuel blend given the lower heating values of both gases (LHV_CH4 = 50 MJ/kg and LHV_H2 = 120 MJ/kg [1]). Fuel mass flow rates for various hydrogen participations in the fuel are given in Table 4.

The pressure-based solver was selected for the calculations, with a second-order discretization with the pressure–velocity coupled methods [27,41]. The next step was to run the simulations. The calculations were considered to be converged when (i) the curves representing residuals were stable or with slight oscillations around a constant value (ii) the before and after combustor pressure ratio was stable, without perturbations, (iii) the mass flow of the exhaust gases leaving the combustor was stable, without perturbations and (iv) the relative difference between the mass flows at the inlet/outlet of the combustor was below 0.1%. The reached residuals at the end of the calculations were between 10⁻² and 10⁻⁴.

After the simulation was completed, an additional post-processing step was performed in order to calculate the NOx and CO mass fractions. The unsteady diffusion flamelet post-processing model was enabled in order to determine theses species mass fractions at the outlet of the combustion chamber [34]. The unsteady diffusion flamelet model is dedicated to model combustion processes whose state is significantly different from the equilibrium state (slow reactions). The model is primarily used to model nitrogen oxides (NOx), and the slow oxidation of carbon monoxide (CO). In order to use this model, the calculation results from the steady diffusion flamelet model must be available. The model is used at the post-processing calculation stage. In order to determine the combustion products resulting from slow chemical reactions (out of equilibrium), the probability marker is used, which determines the probability of new flamelet structures (that account for out-of-equilibrium reactions occurring) at a given time interval in a given location of the computational space. The equations used to generate flamelets in the steady diffusion flamelet model, in combination with the probability markers described above, are used to generate the corrected mass fractions of the chemical elements. Three new “slow” flamelets need to be created (any additional flamelet was not increasing the reliability of the calculations). The post-processing calculations were considered achieved when the probability marker of each new flamelet reached a strongly zeroed valued after reaching a maximum value at the outlet of the combustor.

$$c_s^{{{i\%\mathrm{H}}_2}_{mass\_fraction}}=\frac{{LHV}_{\mathrm{C}{\mathrm{H}}_4}\times c_s^{{{0\%\mathrm{H}}_2}_{mass\_fraction}}}{{LHV}_{{\mathrm{H}}_2}\times {{\mathrm{H}}_2}_{mass\_fraction}+{LHV}_{\mathrm{C}{\mathrm{H}}_4}\times \left({{1-\mathrm{H}}_2}_{mass\_fraction}\right)}$$

(11)

Table 4. Fuel mass flows as function of hydrogen mass percentage in fuel.

H2 Mass Percentage [%]	Fuel Mass Flow [kg/s]
10	0.004275
20	0.003808
30	0.003432
40	0.003124
50	0.002867

Table 4. Fuel mass flows as function of hydrogen mass percentage in fuel.

H2 Mass Percentage [%]	Fuel Mass Flow [kg/s]
10	0.004275
20	0.003808
30	0.003432
40	0.003124
50	0.002867

4. Results and Discussion

4.1. Exhaust gas Recirculation Rate

In this part, the mass flow of recirculated exhaust gases will be assessed. Two parameters were created to allow this assessment. The first is the IFGR parameter, and the second is the IFGR* parameter. The definitions of these parameters are given by formulae nr (12) and nr (13). According to the above formulae, the IFGR and IFGR* parameters were calculated for the combustors in cases A and B (equipped with the IFGR systems). The graphs of these parameters are presented in Figure 5 and Figure 6.

$$IFGR=\frac{Recirculated exhaust gas mass flow}{Air mass flow passing through mixing pipe}\times 100$$

(12)

$$IFGR \ast =\frac{Recirculated exhaust gas mass flow}{Exhaust gas mass flow leaving combustor}\times 100$$

(13)

According to above two calculations, the proposed IFGR systems are working and exhaust gas recirculation is performed. The behaviour of the exhaust gases mass flow will be discussed below taking into consideration the combustor design differences and the fuel composition modifications.

Firstly, the case B combustor permits achieving more important exhaust gases recirculation than the case A combustion chamber, independent of fuel composition. The IFGR values for case B are superior to those in case A. Similar observations may be achieved for the IFGR* values. This result was expected because of the design differences between the two IFGR systems. The IFGR system in case B uses the total-static pressure difference and the venturi-shaped channel in order to achieve exhaust gases recirculation, while the IFGR system in case A uses only the total-static pressure difference.

Secondly, for the case A IFGR system, the recirculated exhaust gases mass flow is more constant when adding hydrogen into fuel than for the IFGR case B. This difference may also be justified by the differences in the design of the IFGR systems. In case A, the IFGR system consists of pipe systems added quasi-out of the liner. The fuel property modifications when hydrogen is added have limited impact on the mass flow of recirculated exhaust gases. In case B, the IFGR system is directly connected to the venturi-shaped mixing pipes. When fuel property modifications occur (when adding hydrogen) the flow phenomena inside these mixing pipes are affected, and it may cause modification of the mass flow of recirculated exhaust gases. When hydrogen is added to the fuel, the mass flow of the recirculated exhaust gases decreases in case B. Despite these observations, the case B IFGR system permits achieving more important exhaust gases recirculation than the case A IFGR system, independent of fuel composition.

It must be noticed that it is possible to autonomously recirculate a part of exhaust gases in gas microturbine combustors applying the IFGR system and that the case B IFGR system is more effective than the case A IFGR system in terms of the recirculated exhaust gases mass flow.

4.2. Impact of the IFGR System on the Total Pressure Drop

In this part, the impact of applying IFGR systems on the total pressure drop in the gas microturbine combustion chamber will be assessed. The IFGR systems are additional parts in the combustion chambers. This may potentially increase the total pressure drop in the combustion chamber, which is an unsuitable phenomenon. The total pressure drop through the combustors was calculated according to Equation (14). The total pressure drop is presented in Figure 7.

$$∆p^{*}=\frac{{p_2}^{*}-{p_3}^{*}}{{p_2}^{*}}\times 100$$

(14)

The IFGR systems implies an increase in the total pressure drop, independent of fuel composition. In the reference case, the total pressure drop ranges from 10.1% to 10.2%, in the case A from 10.6% to 10.7% and in the case B from 10.8% to 11.0%. The total pressure drop in the reference case is a standard value for gas microturbine combustors. In case B, the total pressure drop is greater than in case A, independent of fuel composition. The difference between the reference and IFGR combustors in terms of total pressure drop does not exceed 1%, which is an acceptable value. When the fuel composition is modified, the total pressure drop is not significantly impacted. The IFGR systems caused a greater but acceptable total pressure drop than in the reference combustor. The total pressure drop does not seem to be impacted by fuel composition, which is a suitable observation.

4.3. Impact of the IFGR System on the Temperature Field

In this part, the combustion temperature homogeneity, the maximum combustion temperature, and the exhaust gases total temperature will be successively treated. The aim of this part is to assess the effects of the IFGR systems on the combustion temperature parameters.

Firstly, the combustion static temperature homogeneity will be discussed. To perform this assessment, the temperature maps, in representative combustion sections, for methane and 50% hydrogen mass fraction in the fuel mixture, are represented in Figure 8 and Figure 9. Additionally, the uniformity index for the static combustion temperature in this cross-section were calculated according to Formula (15). The uniformity index for the combustion static temperature are presented in Figure 10.

According to a visual assessment, independent of fuel composition, case B appears to present greater homogeneity than the case A combustor, which is higher than in the reference combustor. This effect of IFGR systems is suitable. According to visual analysis, the most intense combustion processes occur in the top part of the liner. The top part of the liner is the inhomogeneity source. Visual observations of the maps are in good correlation with the uniformity index of the static combustion temperature. Independent of fuel composition, the best static temperature homogeneity is achieved in the B IGFR system. At the location where the exhaust gases are recirculated, the combustion process (combustion temperature) is not especially impacted (according to the visual analysis of temperature maps eight and nine). Despite of the exhaust gases recirculation, the obtained effects may be caused by an air flow modification in the liner. In

Table 5. Inlet zone and exit zone lambda vs. flammability limit lambda.

Case	Liner Inlet Part Lambda [-]	Lowest Combustion Lambda [-]	Combustion Status	Liner Exit Part Lambda [-]	Highest Combustion Lambda [-]	Combustion Status
R00H2	0.390	0.592	No combustion	2.96	1.99	No combustion
R10H2	0.390	0.510	No combustion	3.07	3.44	Combustion
R20H2	0.397	0.442	No combustion	3.16	4.66	Combustion
R30H2	0.403	0.384	Combustion	3.23	5.69	Combustion
R40H2	0.411	0.334	Combustion	3.30	6.57	Combustion
R50H2	0.421	0.291	Combustion	3.36	7.33	Combustion
A00H2	0.399	0.592	No combustion	2.96	1.99	No combustion
A10H2	0.400	0.510	No combustion	3.07	3.44	Combustion
A20H2	0.408	0.442	No combustion	3.16	4.66	Combustion
A30H2	0.416	0.384	Combustion	3.23	5.69	Combustion
A40H2	0.423	0.334	Combustion	3.30	6.57	Combustion
A50H2	0.433	0.291	Combustion	3.36	7.33	Combustion
B00H2	0.238	0.592	No combustion	2.96	1.99	No combustion
B10H2	0.246	0.510	No combustion	3.07	3.44	Combustion
B20H2	0.253	0.442	No combustion	3.16	4.66	Combustion
B30H2	0.259	0.384	No combustion	3.23	5.69	Combustion
B40H2	0.264	0.334	No combustion	3.30	6.57	Combustion
B50H2	0.268	0.291	No combustion	3.36	7.33	Combustion

(below), the lambda values are presented in the top and back parts of the liner. According to these data, the combustion zone in the top part of the liner is the most oxygen-depleted, independent of fuel composition. For methane fuel, in case B, the top liner excess air level is 0.238, while in the reference case it is 0.390 and in case A it is 0.399; for half-mass hydrogen–methane fuelling, the top liner excess air level in case B is 0.268, while in the reference case it is 0.421 and in case A it is 0.433. The excess air level differences between the reference and A cases are much lower than between the reference and B cases. The air depletion in the top part of the liner in the B case could justify lower combustion rate in top part of the liner and better temperature uniformity. The case A air access in the top part of the liner is comparable to that in reference combustor. It could justify a closer combustion temperature uniformity index between the reference and A cases. The visual observations are in correlation with the representation of the uniformity index of the static temperature in the analysed cross-section (Figure 10).

In terms of participation of hydrogen in the reference fuel, the addition of hydrogen results in an increase of the hot-spot temperature in the combustors, independent of the combustor design. Based on the temperature maps (Figure 8 and Figure 9), when the hydrogen is added, the size of the hot-spot temperature increases (especially in the top parts of the liner) and inhomogeneity in combustion temperature increases. These phenomena are not desired. The visual observations are in correlation with the representation of the uniformity index of the static temperature in the analysed cross-section (Figure 10).

The best combustion static temperature homogeneity, in the representative cross-section of the liner, is obtained in case B, independent of fuel composition. This increase in homogeneity in case B may be justified by exhaust gases recirculation, but especially by the air flow modification due to the implementation of the IFGR system. When the hydrogen participation in the fuel increases, the homogeneity of the combustion temperature decreases due to the increase of the temperature hot-spots in the top part of the liner.

$${UI}_{area}^{plane}=1-\frac{{\sum }_{i=1}^N\left[\left(\left|{T_{face}}_i-T_{average}\right|\right)\times A_i\right]}{2\times T_{average}\times {\sum }_{i=1}^N\left[A_i\right]}$$

(15)

Figure 10. Area-weighted static temperature uniformity index for selected cross-section.

Figure 10. Area-weighted static temperature uniformity index for selected cross-section.

Secondly, the maximum combustion static temperature will be analysed (Figure 11).

The case B IFGR system permits obtaining the lowest maximum combustion static temperature compared to the reference and A cases, independent of fuel composition. According to this observation, the maximum combustion static temperature (for methane fuelling in reference combustor) is achievable for the B combustor with a hydrogen mass participation in the fuel in the range from 0% to 10%. The use of the B IFGR system makes it possible to obtain operating conditions similar in terms of temperature to those in the reference combustor with methane firing, with an addition of hydrogen (not exactly determined). This effect was researched and is suitable. The reduced maximum combustion temperature in case B may be correlated with the lambda value in the top part of the liner. On the basis of the temperature maps (Figure 8 and Figure 9), the hot-spots in all the combustors are located in the top part of the liner. The reduction of air access in the top part of the liner causes a reduced combustion rate. The most important air depletion occurs in the B case combustor, as described above. The air flow modification in the case B permits justifying the maximum temperature reduction compared to the reference and A cases. In the reference and A cases, the lambda values in the top part of the liner are comparable, and the combustion maximum temperatures in these two cases are also comparable. The above analysis permits observing that the most important combustion temperature reduction is obtained for the case B IFGR system. This achievement may be justified by the modification of the air flow, in this the IFGR case, compared to the reference case.

Figure 11. Combustion maximum static temperature in combustors.

Figure 11. Combustion maximum static temperature in combustors.

In terms of fuel composition, independent of the combustors design, when the hydrogen fraction increases in the fuel, the maximum combustion static temperature also increases. Its observation was expected but it is not suitable for correct combustor operation. The maximum increase is about 120 K.

The maximum combustion static temperature is reduced in case B compared to the reference and A cases. The application of the B IFGR system makes it possible obtaining similar operating conditions in terms of temperature than in the reference combustor with methane firing, with an addition of hydrogen in the range of 0% to 10% (not exactly determined). This temperature reduction in case B is due to the air flow modification compared to the reference and A cases. Air flow modification is due to the implementation of the IFGR system in the combustors. The addition of hydrogen to fuel results in an increase in terms of the maximum combustion static temperature, which is not a suitable phenomenon.

Finally, the exhaust total temperature will be treated. The total temperature is one of the most important operating parameters when dealing with gas (micro) turbine combustor. This parameter is presented in the Figure 12.

Independent of fuel composition, the exhaust total temperature is lowest in B case compared to the reference combustor. In case A, the exhaust total temperature is close to that of the reference case. In the reference combustor for methane fuelling, the exhaust total temperature is 1235 K, while the lowest exhaust total temperature is 1217 K (for B at half-mass hydrogen–methane fuelling). The total temperature drop in the exhaust is about 1.5% between the reference operating combustion and the most disadvantageous case. This parameter modification is acceptable. The IFGR systems do not significantly impact the exhaust total temperature, which is a positive observation. The observed reduction of the total temperature of the exhaust in case B may be justified by the air accessibility in the top part of the liner. In case B, the accessibility to air (oxygen) in the top part of the liner causes a move of the combustion zone in the direction of the back part of the liner, compared to the reference and A cases. This movement was initially reached to avoid overheating and damaging the top part of the liner. This movement of the combustion zone can also cause incomplete combustion due to the reduction of the combustion volume and duration. When methane fuel is applied, the air–fuel mixture at the end of the liner is too poor for combustion to occur, but since 10% of the mass participation in the fuel, the air–fuel mixture at the end of the liner may still undergo combustion (Table 5). It means that the reduction of combustion volume and time in case B results in a more incomplete combustion than in reference and A cases, resulting in lower energy release from the fuel to the air/exhaust gases. Since the first-studied addition of hydrogen into the reference fuel, combustion may occur even in the back part of the liner, which may damage the turbine material and be the source of incomplete combustion (lowering exhaust total temperature).

Independent of the combustor design, the addition of hydrogen in the fuel results in an acceptable drop in terms of the exhaust gases total temperature. This phenomenon may be justified by the increased participation of steam in the exhaust gases, due to the increased hydrogen participation in the fuel. It permits increasing the exhaust gases heat capacity and reducing the exhaust gases temperature.

The exhaust total temperature is not especially impacted by the implementation of the IFGR systems. A negligible exhaust total temperature drop may be noticed when hydrogen is added into the fuel.

To conclude the combustion temperature analysis, it must be noticed that the B IFGR system permits achieving the best temperature homogeneity and the lowest maximum combustion temperature; additionally, it does not impact especially the exhaust total temperature. The use of the B IFGR system makes it possible to obtain operating conditions in terms of temperature similar to that in the reference combustor with methane fuelling, with the addition of hydrogen. The observed effects of the B IFGR system are mostly related to the modification of air flow in the liner and not to the recirculated exhaust gases. The addition of hydrogen to the fuel causes a possibility of combustion even at the outlet of the combustor, which is potentially dangerous for the turbine operations. In terms of the combustion temperatures, the IFGR system seems not to be useful, and obtained results could be obtained via the liner holes size and location modifications. The amount of recirculated exhaust gases is too low to have an expected impact on combustion processes. In terms of temperature analysis, the IFGR systems are not suitable.

4.4. Impact of IFGR System on the Air Accessibility in the Top and Back Zones of the Liner

In this part, the lambda values in the top and back parts of the liner will be treated. In order to establish the lambda values in the top and back parts of the liner, the oxygen mass flow at the outlet of the mixing and IFGR pipes (in the top part of the liner) and at the end of the liner, the fuel total mass flow and the oxygen need for stochiometric combustion for various hydrogen participation in the fuel were used [42,43,44]. The “liner inlet part lambda” was established taking into consideration the oxygen from the mixing pipes and the IFGR pipes systems. The “liner exit part lambda” was established taking into consideration the oxygen present in the exhaust gases leaving the combustion chamber. To establish the minimum and maximum combustion lambda, the flammability ranges of the fuels and the volume participation of hydrogen in the reference fuel were applied [42,43,44,45]. On the basis of these studies, it was possible to determine if combustion was theoretically possible in the top/back part of the liner, taking into consideration the oxygen accessibility in these parts of the liner. This issue is theoretical, because the flow in the liner is turbulent and combustion may occur even in location “out” of the flammability range. The status “combustion” means that the combustion may fully occur—with intense heat release, while the status “no combustion” means that the combustion should be inhibited—with low heat release. The analysis results are presented in Table 5.

Firstly, the top part of liner combustion condition will be treated. According to the above, the combustion in the top part of the liner does not occur until 20% of mass hydrogen in the fuel for the reference and A cases. Since 20% hydrogen in the fuel, the combustion may occur in the top part of the liner, the zone where the combustion is not supposed to occur. Only in the B IFGR case the combustion is eliminated for all the hydrogen studied participation. This effect was reached by recirculating exhaust gases. No combustion in the top part of the liner in case B is due to the air-flow modification and not to the recirculated exhaust gases.

Secondly, in the end part of the liner, since the first addition of hydrogen into the reference fuel, there is the possibility of flame. The lambda value, for a given firing mode, is the same for the reference and A/B cases, and the amount of air and fuel applied to the globality of the combustor is the same. It means that there is a possibility of flame at the inlet of the turbine since the first hydrogen addition. This observation is not suitable because it could damage the turbine. It is just a possibility because the increased rate of combustion when adding hydrogen would provoke combustion to occur in the liner top part.

The B IFGR system implementation caused a limitation of the combustion in the top part of the liner for all studied hydrogen participation in the fuel. In the reference and A cases the top part of the liner is free of flame until 20% of hydrogen in the fuel. Since the first addition of hydrogen to the fuel, combustion becomes potentially possible at the inlet of the turbine, not a suitable phenomenon. All observed modifications in combustion at the top and back of the liner are due to air flow modification when applying the IFGR system and not to recirculated exhaust gases. The IFGR systems seem not to be useful, and the same effects could be obtained by modifying the liner holes size and location.

4.5. Impact of the IFGR System on CO and NOx Concentrations in Exhaust

In this part, the CO and NOx concentrations in the exhaust gases will be discussed. The CO and NOx concentrations were calculated, respectively, according to the Formulas (16) and (17). The CO and NOx concentrations in the exhaust gases are, respectively, presented in Figure 13 and Figure 14. Firstly the CO concentrations will be discussed and then the NOx concentrations in the exhaust gases.

$$\mathrm{C}\mathrm{O}=\frac{{\mathrm{C}\mathrm{O}}_{mole\_fraction}}{1-{{\mathrm{H}}_2\mathrm{O}}_{mole\_fraction}}\times \frac{21-15}{21-{\mathrm{O}}_2}\times {10}^6$$

(16)

$${\mathrm{N}\mathrm{O}}_{\mathrm{x}}=\frac{{\mathrm{N}\mathrm{O}}_{mole\_fraction}+{\mathrm{N}{\mathrm{O}}_2}_{mole\_fraction}}{1-{{\mathrm{H}}_2\mathrm{O}}_{mole\_fraction}}\times \frac{21-15}{21-{\mathrm{O}}_2}{\times 10}^6$$

(17)

Firstly, the CO concentrations in the exhaust gases will be treated. Independent of fuel composition, the CO concentrations in the B case are more important (approximatively twice) than in the reference and A cases. In case B, in the top part of the liner, an air access reduction occurs compared to the other cases. This reduction in air access causes an inhibition of the temperature in the top part of the liner (as explained above). Reduced combustion temperature and oxygen access result in higher production of CO in the top part of the liner, compared to the reference and A cases, where air access is more important in this part of the liner. The air flow modification allows justifying the CO evolution in the combustors.

Regardless of the combustor designs, the CO concentrations decrease when hydrogen is added to the reference fuel. This result was expected, because when hydrogen is added to the reference fuel, the fuel contains less carbon. Another interesting observation is that the CO concentrations in the exhaust difference between the B and reference/A cases decreases when hydrogen is added to the fuel. When hydrogen is added to the fuel, the flammability range of the fuel increases, and the combustion volume increases, resulting in comparable combustion conditions in all design cases.

The case B combustor has higher CO concentrations in exhaust compared to the reference and A cases, independent of the hydrogen addition into the fuel. The reference and A cases CO concentrations are comparable for all fuelling modes. The CO concentrations in the exhaust are justified by the modification of the air flow in the liner and not by the reburning effect of the recirculated exhaust gases by the IFGR system. In terms of the CO concentrations in the exhaust gases, the IFGR systems are not suitable; similar effects may be obtained by the liner holes size and location modifications.

Then the NOx concentrations in the exhausts will be discussed. Independent of fuel composition, the NOx concentrations in case B are greater than in the reference and A cases. When comparing Figure 11 (Tmax) and Figure 12 (exhaust gases temperature), case B always has lower temperature than the others cases, which should permit a lower NOx generation than in the other cases. The NOx production is not only linked to the maximum temperature occurring in the liner but also to the temperature field. The exhaust temperature is very close between all cases, even in case B, so this element may not strongly impact the NOx generation. The NOx concentrations in the exhausts in the reference and A cases are comparable. In the reference and A cases, the hot-spots are located in the top part of the liner (according to the temperature maps eight and nine), while in case B, the hot-spot is inhibited in the top part of the liner. As in all the cases, the same amount of energy is provided from fuel into the air passing across the liner, and the absence of intense hot-spots in case B in top part of the liner makes obligatory the mean temperature in the central and end parts of the liner to be higher than in the other cases. In the reference and A cases, the NOx is intensively produced in the hot-spots, while in case B, the NOx are produced less intensively than in the reference/A case hot-spots, but in greater volume at a higher mean temperature. The NOx concentrations in the exhaust are linked to the temperature behaviour, which is connected to the air flow passing across the liner. The NOx concentrations in the exhaust are justifiable by the air flow modification when implementing the IFGR systems. Recirculated exhaust gases appear not to have effects on NOx production.

Independent of the combustors design, when adding hydrogen into the fuel, the NOx production in reference and A cases is quasi stable (from 67 ppmvd15%O₂ to 72 ppmvd15%O₂). In case B, NOx production is greater and evolves from 83 ppmvd15%O2 to 73 ppmvd15%O₂. In all cases, the evolution of NOx concentrations in exhaust is not strongly impacted by the increase in participation of hydrogen in the fuel. When hydrogen is added, the hot-spots are more intense, whereas the other liner zones must be cooler (principle of energy conservation). In the intense hot-spots, the NOx production increases, while in the others liner locations, the NOx production decreases. This balance leads to quasi-constant NOx production while hydrogen is added to the fuel.

Higher NOx concentrations in exhaust gases are met in case B, independent of the fuelling mode. The NOx production in all the combustors is quasi-constant. The modifications in NOx concentrations observed, especially in case B, may be justified by air flow modification due to IFGR system implementation and not by the action of the recirculated exhaust gases. The IFGR systems do not seem to be useful in terms of NOx concentrations in exhaust gases.

5. Conclusions

In order to conclude the conducted studies presented in this paper, the following points are made:

• There is a possibility of performing an autonomous internal flue (exhaust) gas recirculation using the adequate IFGR system.
• The case B IFGR system permits recirculating more exhaust gases than the case A IFGR system.
• The IFGR systems do not impact/alter especially the combustor main operating parameters (total pressure drop and the exhaust total temperature), which is a positive observation.
• The case A IFGR system presents a lower combustion temperature gradient than the reference case, and the maximum combustion temperature in case A is comparable to that of the reference case, independent of fuel composition,
• The case B IFGR system permits reducing the combustion temperature gradient and the maximum combustion temperature compared to the reference and A cases, independent of fuel composition.
• The application of the B IFGR system makes it possible to obtain similar operating conditions in term of temperature than that in the reference combustor with methane firing, with an addition of hydrogen in range from 0% to 10% (not exactly determined).
• In terms of temperature parameters, the case B IFGR system presents the best results.
• The addition of hydrogen into the fuel causes a temperature gradient degradation and the maximum temperature to increase.
• The CO and NOx concentrations in exhaust gases are comparable between the case A and the reference case, independent of fuel composition.
• The CO and NOx concentrations in the exhaust gases are greater in case B than in the reference and A cases, independent of fuel composition.
• In terms of CO and NOx concentrations in the exhaust gases, the case B IFGR system is not suitable and the case A IFGR system has no impact compared to the reference combustor.
• The CO production in all the studied combustion chambers decreases when hydrogen participation increases in the fuel.
• The NOx production in all the studied combustion chambers is not especially impacted by the hydrogen increase in the fuel.
• The observed and described modifications in terms of combustion temperature and CO/NOx concentrations in exhaust may be justified by the lambda ratio modifications between the IFGR cases and the reference case.
• The lambda modification in IFGR cases, especially in case B, is due to the air flow modification in the liner due to the IFGR system implementation.
• The combustion processes modifications in the IFGR cases are due to the air flow modification in the liner.
• The recirculated exhaust gases have no visible impact on the combustion processes.

According to the above, it must be noticed that the IFGR systems are operational (permit recirculating a part of exhaust gases). Unfortunately, autonomous IFGR systems have too poor exhaust mass flow recirculation to effectively impact the combustion processes. All observed modifications in terms of combustion processes in the IFGR cases are due to air flow passing across the liner modifications and are not due to recirculated exhaust gases action. The proposed IFGR systems are not useful; the modifications observed in combustion processes could be achieved by modifying the size and location of the liner holes, without an IFGR system application. The proposed IFGR system is not suitable for gas microturbine combustors.

The studies could be continued with a forced IFGR system, which could allow more exhaust gases to be recirculated, as proposed in [12]. The MILD technology needs also to be studied because it presents potential improvement in the gas microturbine combustor domain.