Impact of Ethane, Propane, and Diluent Content in Natural Gas on the NOx emissions of a Commercial Microturbine Generator

Transcription

1 Paper # 070IC-0200 Topic: Internal Combustion and Gas Turbine Engines 8 th U. S. National Combustion Meeting Organized by the Western States Section of the Combustion Institute and hosted by the University of Utah May 19-22, 2013 Impact of Ethane, Propane, and Diluent Content in Natural Gas on the NOx emissions of a Commercial Microturbine Generator Andres Colorado 1 Vince McDonell 1 1 UCI Combustion Laboratory, University of California, Irvine, CA Fluid dynamics and chemical kinetics play a significant role in the formation and emission of NOx, particularly in lean premixed gas turbine combustors, where both phenomena must be considered. However, experimental tests conducted on gas turbines generally evaluate the output variables (e.g., the stack emissions) and those results are used to infer, to the extent possible, the physical and chemical processes taking place in the combustor. This approach only gives an external point of view, which can be complimented with the use of simulation tools such as the chemical kinetics models and computational fluid dynamics. In this paper, a reactor network analysis (RNA) of a 60 kw micro gas turbine combustor fueled with mixtures of natural gas-ethane-propane; and natural gas diluted with carbon dioxide are presented. A chemical reactor network (CRN) is developed for the combustor as operated on natural gas, with the purpose of tuning it to match measured exhaust emission levels. Then the CRN is used to analyze the effect of fuel composition on the NOx emissions. The development of the CRN model is guided by reacting flow computational fluid dynamics (CFD). The CFD results define some of the important variables for the CRN, such as the residence time, volume of the reactors, temperature profiles, recirculation and mixing patterns. This strategy also gives insight to the complex phenomena occurring in the combustion chamber otherwise beyond the experiments. The CRN gives a detailed description of the emissions formation pathways while taking into account the mixing patterns obtained with the CFD. Experimental results are available from a previous studies conducted at the UCI Combustion Laboratory. The results obtained with the CRN indicate that the most important mechanism leading the formation of NOx under the operating conditions of this engine is the N 2 O pathway. This is the dominant pathway regardless the fuel composition. The CRN model predicts similar trends in overall emissions levels as those observed experimentally. The results also show that fuel dilution with CO 2 hinders all the NOx routes but has a more significant effect on the thermal and N 2 O routes. The addition of heavier alkanes (propane and ethane) promotes the formation of NOx since the production of O and H radicals are stimulated. Those radicals are involved in the N 2 O mechanism. 1. Introduction Gas turbine systems for stationary applications require compliance with stringent limitations concerning criteria pollutants; especially NO x and CO emissions are limited by the US legislation when the fuel is natural gas. Within this regulation framework, it is important to have access to numerical tools that accurately predict NO x emissions from gas turbines, moreover when the composition of natural gas is variable regarding the content of heavier alkanes and diluting gases such as carbon dioxide (CO 2 ) and nitrogen (N 2 ). Furthermore, the aerodynamics in the combustion chamber and the premixing method of fuel and oxidant play a major role in the chemical reactions, therefore the pollutant formation, as they control the rate and conditions under which active species are mixed together. Consequently the prediction tool must be able to take into account a detailed chemistry set able to predict the effect of fuel composition and the mixing patterns involved inside the combustion chamber. With the current capabilities of computational fluid dynamics (CFD) software, it is possible to predict (relatively well) the main characteristics of the combustion phenomena such as, heat release, flow field and main species concentration. However, it is still very difficult to completely couple the fluid mechanics solver with the full combustion chemistry, especially when dealing with complex geometries in three dimensions and when the fuel composition is not as simple as pure methane. In order to reduce computer power requirements and processing time, CFD simulations for

2 chemically reactive flows often ignore or greatly reduce the treatment of chemical kinetics. This is done by using reduced mechanisms with fewer reaction steps or even using skeletal mechanisms with reactions in series e.g. CH O 2 CO+2H 2 O CO+0.5O 2 CO 2. Still, the assumption of local chemical equilibrium is not appropriate since it neglects the participation of intermediate radicals or minor species leading to the formation of pollutant emission. Hence, with the purpose of predicting accurately the formation of pollutant species (NO, NO 2 ) it is necessary to couple CFD with a detailed description of kinetics. The concept of modeling the flow field using a network of ideal reactors was first introduced by S.L Bragg in Bragg modeled a premixed flame using a perfectly stirred reactor (PSR) followed by a plug flow reactor (PFR). This model is known as a Bragg cell (Bragg 1953). More recently the Bragg cell has evolved to a more complete simulation tool called reactor network analysis (RNA); the network is built as a set of interconnected reactors (PSR, PFR, and mixers) in series or parallel representing a simplified and equivalent flow field that allows the use of detailed chemistry while reducing the amount of computational time with respect to the direct implementation into a threedimensional CFD code. The flow inputs and outputs are also included in the simplified model, and represent the boundary conditions of the system. In order to implement the RNA, the flow field information is first extracted from a CFD simulation or from in-situ experimental results. In practice, experimentally measured spatially resolved flowfield vector and scalar information may not be readily available. In any case, the methodology has been successfully implemented to the analysis of emission from simple burners, where the fluid dynamics can be represented by two or three reactors in series (e.g., Rutar and Malte 2002), and to more complex system such as furnaces and gas turbines with multiple stream inputs where the recirculation patterns must be modeled with multiple reactors interconnected in parallel (e.g., Falcitelli, Pasini, and Tognotti 2002; Falcitelli et al. 2002; Fichet et al. 2010; Novosselov and Malte 2007). Advanced gas turbine combustion systems must operate within narrow margins in order to meet regulated emission targets (Richards, et al., 2001; Lieuwen, et al., 2008). As a result, these systems are susceptible to instabilities eliminating much of the tolerance to system and operational variation that have been positive attributes of conventional combustion systems in the past. Natural gas composition varies around the United States, and even more so around the world. Furthermore, the variability in natural gas composition is increasing as producers relax their control on the amount of inert gases such as CO 2 and high-molecular fuel species such as ethane, propane, butane and even hexane in the natural gas supply. In a previous experimental study, Hack and McDonell conducted several tests on a commercial micro turbine generator (MTG) operated on different gaseous fuels (Hack and McDonell 2008). The fuels used comprised natural gas with the presence of ethane, propane, and, in some cases, inert gases. A statistically designed experiment was carried out at full load and showed that the presence of higher hydrocarbons leads to systematically elevated NOx levels when compared to 100% methane. Similarly, Effinger et al. conducted experiments on a the same model MTG to analyze the effect of diluting natural gas with carbon dioxide (CO 2 ), their results indicate that higher levels of diluent produce the lowest NOx and highest level of CO (Effinger, Mauzey, and McDonell 2005). However those studies did not address the relative influence of each NOx pathway on the total NO X emissions. In this paper, the effect that fuel variability has on the NOx emissions of a gas turbine is analyzed using RNA. The CRN is able to handle complex chemical mechanisms and can provide significant insight into pollutant formation. Because of its small computational time requirement, the CRN can be used as tool for analysis of combustion systems and can be integrated into combustor design. Additionally, the RNA gives insight about the different pathways that lead to the total emissions of NOx measured during the experiments. In order to present the data results in a consistent manner, the emissions for the different fuel compositions are compared to those emitted when running the turbine on natural gas. All the results are presented at the same power output, air preheating temperature, and adiabatic flame temperature (AFT). Since for a given requested power setting the turbine exit temperature remains constant, it can be expected that the global reaction temperature would remain constant regardless of the incoming fuel heat content. This control strategy is important relative to the mechanism by which fuel composition may impact NOx emissions. It has been shown that NOx is sensitive to fuel composition for the situation where the combustion temperature is held at a fixed level below the point where thermal NOx is predominant (Hack and McDonell 2008). This approach might be useful to in terms of how appliances and equipment might be modified in the field to better address a known change in gas composition, or during the design process to optimize the performance of the turbine regarding pollutant criteria. 2. Methods 2.1. Microturbine Generator. The Capstone C60 MTG operating with natural gas has been studied extensively in previous work (Hack and McDonell 2008; Effinger, Mauzey, and McDonell 2005). This system was selected as the test platform for this 2

3 investigation because the experimental results were readily available, but also because some questions regarding the mechanisms that lead to the experimental levels of NOx as a function of the fuel variation remained unanswered. Those questions can be answered using the RNA analysis. The C60 has 6 injectors that are fuel staged to enhance engine stability over the entire operational range. The fuel/air premixing occurs within the six individual fuel injectors with a fraction of the total combustion air. This premixing strategy, while successful at mitigating emission, also raised some concerns for tolerance to fuel composition since the higher flame speeds and reaction rates associated with ethane and propane compared to natural gas could result in flashback into the premixer, which would result in diffusion reactions that would generate significantly higher levels of NOx and/or hardware damage. The system is designed to operate fuel staged and features two planes of injection followed by a dilution zone (Figure 1). Testing with natural gas indicated that unfired injectors, during partial load operation, create a quenching effect in the combustor. The 60 kw MTG features a single stage compressor and single stage turbine on a common shaft as well as an annular combustor configuration. The Capstone combustor incorporates a globally swirl stabilized combustion process with fuel injectors that are canted slightly from purely radial orientation. The MTG includes a recuperator that provides preheated air to the combustor at a temperature of 840±10 K, regardless of the turbine load and/or the ambient temperature. The pressure ratio is nominally 4:1 and recuperation of the exhaust gas for preheating the incoming combustion air is incorporated to increase the overall efficiency. Figure 1. C60 combustor outline with injector plane cross section 2.2. Methodology The application of the CRN methodology to the Capstone C60 Microturbine engine is presented in this section. The most important characteristics of the C60 combustor were presented previously. Both CFD and CRN models take into account the air/fuel splits across the six injectors and the distribution of air jets in the dilution zone. The dilution zone includes two air streams, the inner air jets and the outer air jets, those jets cool down the exhaust gases before they are supplied to the turbine. Information regarding air split is proprietary. An intelligently designed chemical reactor network must be an accurate representation of the flow patterns and mixing characteristics of the device it represents. In a CRN, the flow and flame patterns in the combustion volume are divided into zones represented by a reactor. Each reactor requires several inputs such as composition of the reactants, residence time, reactor volume, temperature, pressure, flow rate, distribution of the recirculation patterns. That information can be obtained from previous experimental results and from Computational Fluid Dynamics (CFD) simulations. Since the reaction structure inside the combustion chamber of a turbine is not optically accessible for the experimenter, the CFD results have shown to be particularly suitable to gain details about those reaction structures. For that reason, the flame profiles, volume of the reactors, residence time, recirculation patterns, were obtained from the CFD results. At full load (60kWe) the six nozzles inject a mixture of fuel and air into the combustion chamber. The turbine operating at full load on natural gas (in this case pure CH 4 ) was the model base to set up the chemical reactor network. Important variables such as the residence time, volume of each reactor and temperature after the core of the reactions, 3

4 were extracted from the CFD results of the turbine at full load. Figure 2 shows the temperature contours at the first and second plane of injectors. These contours were obtained for the turbine combustor running at full load. It is possible to see in Fig 2. the six injectors and the characteristic conic section of an attached premixed flame (blue region). Both CFD and CRN simulations assume perfect premixing of fuel and oxidant. Figure 2. Temperature contours of a turbine combustion chamber. The computational model shows the temperature profiles at full load (60kW of electric power). See the two planes of injectors indicated in the figure. The chemical reactor network presented in Figure 3 depicts the flow distribution extracted from the CFD analysis. The first block (or block 1) includes the total mass flow inlet of premixed gas that is necessary to achieve an electrical power output equal to 60kWe. The flow rate of premixed gas was set using the actual thermal efficiency at full power. In the same block, the flow rate is divided into three blocks (Blocks 2, 4 and 5). Since the turbine configuration displays flame symmetry, it was possible to set a block of two perfectly stirred reactors (PSR) to represent each pair symmetrical flames. Each pair of flames is grouped into Blocks 2, 4 and 5, respectively and is represented by two PSRs in series. The first reactor of the series represents the primary reaction zone or the core of the reactions (the reactions taking place inside the blue cone), where the effects of the surroundings are negligible; the second reactor in each block symbolizes a post flame zone, which is a region at high temperature where the first PSR injects its combustion products; additionally, Blocks 4 and 5 account for the effect of the exhaust gases that were produced in Block 2. The splitter in the CRN divides the exhaust products from Block 2 into three flow streams; two of these portions are directed toward the two second PSRs of the fourth and fifth Blocks, in that manner the effect of the exhaust gas that come from the plane of two injectors is accounted with this CRN configuration. Conversely this CRN does not consider an interaction between flames that located within the same plane. From the analysis of the CFD results neglecting the effect of recirculation patterns in the plane of two injectors is a reasonable assumption; however the same assumption may not be as valid for the plane of four injectors where it is expected that the exhaust gases emitted by the flames in the two plane of 2 injectors and the preceding coplanar flame have a significant effect on the combustion process. The exhaust recirculation effect (both from the preceding flame and from the preceding plane) is accounted by using a portion of the exhaust gases from the Block 2. Block 6 represents a post flame region where the streams of all the previous reactors are mixed again before they are diluted with air coming from Block 7 (dilution air inlet). Finally, Block 8 accounts for the mixing process and the reactions after the dilution air is injected. NOx emissions corrected to 15% O 2 are analyzed in the exhaust products of Block 8. 4

5 Figure 3. Combustion System divided into several chemical reactors. 3. Results and Discussion The effect of the addition of ethane and propane to natural gas, and the dilution of natural gas with CO 2 is assessed using the CRN methodology. This section presents the results obtained with the CRN for those fuel mixtures and compare them to the experimental results Higher Hydrocarbons The effect of blending natural gas with ethane and propane on the NOx pathways is examined this section. Two fuel mixtures of methane with heavier alkane gases were assessed with the RNA. The first fuel is a mixture 15% ethane (C 2 H 6 ) balanced with CH 4 and the second blend is 20% propane (C 3 H 8 ) balanced with methane. This represents the expected variation in natural gas content within the United States (NGC+ Interchangeability Work Group, 2005). The fuel composition is presented on a volumetric basis. The experimental results indicate that the presence of higher hydrocarbons in the fuel leads to appreciably higher NOx emissions under the MTG conditions. Similarly, the CRN results display a similar trend. Figure 4 presents the total emissions of NOx as predicted with the CRN model and compared to the experimental results obtained by Hack & McDonell in The emissions are presented in ppm corrected to 15% O 2. For the CRN models, the equivalence ratio was set to guarantee a constant adiabatic flame temperature equal to 1850 K and the temperature of the preheated air was held constant at 835K. The experimental protocol also guarantees a constant temperature at the turbine exit while the temperature of the preheated air remains also constant regardless of the fuel composition. Even though the NOx emissions predicted with the CRN are slightly higher than those measured during the experiments, it is clear from fig 4 that the CRN predicts a similar trend for different fuel compositions. 5

6 Figure 4. Comparison of NOx emissions (CRN and experimental) at constant adiabatic flame temperature and preheated combustion air temperature (AFT= 1850K, preheated air temperature=835k) For low emission systems, combustion temperatures are purposefully maintained at relatively low values (e.g., below 1900 deg K) to avoid the formation of NOx through the most commonly discussed mechanism, known as thermal NOx. Given the relatively low equivalence ratios (~ ) and low reaction temperatures that are considered in the experiments and for the CRN model (max temperature: 1850K), it is not expected that the thermal NOx pathway is responsible for most of the measured NOx. The conditions that lead to high thermal NOx are controlled by using very low equivalence ratios, which leads to a lean reaction at a controlled temperature below 1850 deg K. Since the conditions for the combustion reactions are held below the point where thermal NOx is predominant, it is possible analyze the other pathways that are more sensitive to the fuel composition. Using Chemkin software in combination with the GRI Mech 3.0 and eliminating one by one each NOx pathway included in the full chemistry mechanism, it is possible to determine the relative production of NOx generated by each pathway (See Figure 5). Figure 5. Influence of the different NOx pathways to the total NOx emissions. The nitrous oxide (N 2 O) pathway combined with the thermal NO x pathway are responsible for about ~ 90% of the total NO x emissions, with the N 2 O pathway being responsible for a substantial portion, greater than 53% of the NO x 6

7 measured. This is not surprising, when considering that while thermal NO x may be present, it requires high reaction temperatures, typically above 1900 K, to become dominant. Below this temperature other NO x pathways may play a significant role. Furthermore, the percentage and amount of NO x formed though the N 2 O pathway increases with the addition of propane (Flores, McDonell, and Samuelsen 2003); the N 2 O pathway plays a significant role in the overall levels of NO x emitted from lean premixed reactions with equivalence ratios less the This is also consistent with the fact that the MTG operates at elevated pressure (4 atm) and N 2 O mechanism is dominant at these conditions. Other pathways as the prompt NO x and the NNH are less likely to contribute to the overall emissions of NO x from the system. However, prompt NO x can continue to play a role at reaction temperatures below 1800 K; even so, in lean premixed systems, it has been indicated that prompt NO x is not a major source of NO x until the equivalence ratio exceeds approximately The N 2 O Intermediate mechanism is important in lean premixed combustion. This mechanism is explicated by the following reactions. Equation 1 Equation 2 Equation 3 The pressure influence on NO formed from this mechanism is evidence in Equation 1 with M representing a chemically unchanged third body species. Per Le Chatelier s principle, increasing pressure drives Equation 1 to the right, thus enabling NO formation through the subsequent reactions. Figure 6 presents the relative percentage rise/reduction of the intermediate radicals (compared to pure CH 4 ) leading to the formation of NO x through the N 2 O mechanism. The maximum concentration of intermediate species that lead the formation of NO x through the N 2 O route is found in the post flame reactors (second reactors in the Blocks 2, 4 and 5). It is clear from the figure that the addition of heavier hydrocarbons increases the production of O, H, and N 2 O species; especially the O radical is the most significant since it is the precursor of the N 2 O intermediate formed through the triple collision of N 2 +O+M, as presented in the equation 1. Even though the relative concentration of the H radical is greater when mixing the fuel with ethane, its relative importance is lower compared to the roll that the O radical plays in the N 2 O pathway, consequently the higher concentration of the O radical that is found when the fuel is mixed with propane leads to a higher concentration of N 2 O, which increases the production of NO x through this mechanism. Figure 6. Relative increase reduction of radicals (compared to 100% CH4) leading the formation of NOx through N2O Intermediate mechanism 7

8 3.2. Diluted Alkane Fuels The tests were conducted to quantify the relative impacts of CO 2 on C60 emissions. CO 2 and N 2 act as diluents when present in the fuel. The presence of CO 2 is expected to have a greater impact on reaction temperatures emissions than N 2 because it has a greater specific heat. As a result, it might be expected that each species may impact emissions somewhat differently. CO 2 was used as the only diluent during the testing. All the tests involved an extensive characterization of MTG emissions performance with various levels of CO 2 in the natural gas based fuel supply. More details about the experiments can be found in (Effinger, Mauzey, and McDonell 2005). MTG operation remained stable through the 100% load with CO 2 concentrations up to 45%. The experimental results indicate that higher levels of diluent produce the lowest NO X and highest level of CO. This is due to reduced combustion temperatures caused by the addition of fuel diluent and the depletion of OH radicals that form NOx through the thermal route. The observed inverse trend is a wellknown relationship of the dependency of emissions formation on temperature. Similarly, the relative importance of each NOx pathway was analyzed using RNA. Figure 7 shows the predicted and measured values of NOx (both presented in ppm by volume and corrected at 15% O 2 ). The tests indicate that the introduction of a small amount of diluent into the combustion zone will decrease the rate of thermal NOx production. This is the physics behind the injection of water or steam and of lean combustors. By diluting the fuel with 8.8% CO 2 a reduction of NO x emissions by 52% (from 8.4 to 4.3 ppm) was achieved on the C60. The NOx depletion reaches an asymptote as the volumetric concentration of CO 2 in the fuel increases from 36 to 44%. On the other hand, the CRN captures the same NOx reduction trend but does not capture precisely the slope of the reduction. At a higher concentration of CO 2 the CRN becomes less accurate. Figure 7. NOx emissions as a function of the concentration of CO2 in the fuel (Tests and CRN results). The CRN results presented in Figure 8 show that the addition of CO 2 hinders the formation of NO x through the thermal and N 2 O routes, while the NNH and prompt mechanisms are not significantly affected. Similarly to what was found for the alkane base fuels, the conditions of pressure, equivalence ratio, and control strategy adopted in the MTG promotes the formation of NOx through the N 2 O intermediate species mechanism. These results are consistent with measured NOx levels for diluted fuels or oxidants. The addition of CO 2 lowers the average reaction temperature in the combustor which promotes a reduction of the thermal NO x. Also, a higher concentration of diluting gases reduces the probability that a molecule of N 2 react with O and a third body, since the molecule of CO 2 may obstruct the collisions that promote the formation of NO x through the N 2 O route. Figure 9 presents the relative percentage rise/reduction of intermediate radicals (compared to pure CH 4 ) that are produced in the reactors when diluting the fuel with CO 2. Fig 9 shows that the radicals O and H involved in the intermediate N 2 O route are depleted with the addition of CO 2, which hinders the production of the intermediate N 2 O molecule that is the principal molecule in the formation of NOx through the N2O route. 8

9 Figure 8. Influence of the different NOx pathways to the total NOx emissions for mixtures methane-carbon dioxide. Figure 9. Relative increase/reduction of radicals (compared to 100% CH4) leading the formation of NOx through N 2 O Intermediate mechanism Finally, NO 2 emission levels were analyzed using the CRN model. It has been observed that, at low temperature conditions such as at start up and low load, brown exhaust plumes can results due to emission of NO2 from the combustion system. Figure 10 shows the CRN results for NO 2 as a function of the CO 2 concentration in the fuel mixture. It is believed that HO 2, which is formed in low temperature regions, can oxidize NO present in higher temperature regions. NO 2 is related to the HO 2 radical that can be formed by the third body reaction H+O 2 +M HO 2 +M. The concentration of HO 2 in the CRN revealed that the presence of diluents such as CO 2 does not lead to a difference in the generation of HO 2. However, the presence of diluents reduces the concentration of other radicals such as H and OH, and dilutes the O 2 that is necessary in the third body reaction that leads to NO 2. It is important to highlight that the CRN indicates that the concentration of NO 2 in the exhaust represents less than 8% of the total NO x emissions. 9

10 Figure 10. NO 2 emissions as a function of the concentration of carbon dioxide in the fuel 4. Conclusions The MTG operating conditions such as high pressure, lean equivalence ratio ( ), and the temperature control strategy promote the formation of NOx mainly through the N 2 O intermediate species mechanism and the thermal route. The nitrous oxide (N 2 O) pathway combined with the thermal NO x pathway are responsible for about ~ 90% of the total NOx emissions, with the N 2 O pathway being responsible for a substantial portion, greater than 53% of the NO x measured. The reactor network analysis (RNA) indicated that the addition of heavier hydrocarbons increases the production of O, H, and N 2 O species; especially the O radical is the most significant since it is the precursor of the N 2 O intermediate formed through the triple collision of. Even though the relative concentration of the H radical is greater when mixing the fuel with ethane, its relative importance is lower compared to the roll that the O radical plays in the N 2 O pathway, consequently the higher concentration of the O radical that is found when the fuel is mixed with propane leads to a higher concentration of N 2 O, which increases the production of NO x through this mechanism. The addition of CO 2 lowers the average reaction temperature in the combustor which promotes a reduction of the thermal NO x. Also, a higher concentration of diluting gases reduces the probability that a molecule of N 2 reacts with O and a third body, since the molecule of CO 2 may obstruct the collisions that promote the formation of NO x through the N 2 O route. Additionally, the RNA indicates that the radicals O and H involved in the intermediate N 2 O route are depleted with the addition of CO 2, which hinders the production of the intermediate N 2 O molecule that is the principal molecule in the formation of NO x through the N 2 O route. Acknowledgements The support of the California Energy Commission (Contract ; Marla Mueller Contract Monitor) is gratefully appreciated. Discussions with Phil Malte, John Kramlich, Megan Karalus, and Igor Nossolev regarding the application of the RNA were very helpful. References Bragg, S. L Application of Reaction Rate Theory to Combustion Chamber Analysis. Aeronautical Research Council Pub. London, UK. Effinger, Mark, Josh Mauzey, and Vince McDonell Characterization and Reduction of Pollutant Emissions Froma Landfill and Digester Gas Fired Microturbine Generator. In Proceedings of GT2005, 1 7. Reno, NV, United states. Falcitelli, M., S. Pasini, N. Rossi, and L. Tognotti CFD+reactor Network Analysis: An Integrated Methodology for the Modeling and Optimisation of Industrial Systems for Energy Saving and Pollution Reduction. Applied 10

11 Thermal Engineering 22 (8) (June): doi: /s (02) Falcitelli, M., S. Pasini, and L. Tognotti Modelling Practical Combustion Systems and Predicting NOx Emissions with an Integrated CFD Based Approach. Computers & Chemical Engineering 26 (9) (September): doi: /s (01) Fichet, Vincent, Mohamed Kanniche, Pierre Plion, and Olivier Gicquel A Reactor Network Model for Predicting NOx Emissions in Gas Turbines. Fuel 89 (9) (September): doi: /j.fuel Flores, R. M., V. G. McDonell, and G. S. Samuelsen Impact of Ethane and Propane Variation in Natural Gas on the Performance of a Model Gas Turbine Combustor. Journal of Engineering for Gas Turbines and Power 125 (3): 701. doi: / Hack, Richard L, and Vincent G McDonell Impact ofe Thane, Propane, and Diluent Content in Natural Gas on the Performance of a Commercial Microturbine Generator. ASME Conference Proceedings 2005 (4725X) (January 1): Hack, Richard L., and Vincent G. McDonell Impact of Ethane, Propane, and Diluent Content in Natural Gas on the Performance of a Commercial Microturbine Generator. Journal of Engineering for Gas Turbines and Power 130 (1): doi: / Lieuwen, T., McDonell, V.G., Santavicca, D. & Sattelmayer, T., Burner Development and Operability Issues Associated with Steading Flowing Syngas Fired Combustors. Combustion Science and Technology, Volume 180, pp NGC+ Interchangeability Work Group, White Paper on Natural Gas Interchangeability and Non-Combustion End Use, s.l.: NGC+ Interchangeability Work Group. Novosselov, Igor, and Philip Malte Development and Application of an Eight-step Global Mechanism for CFD and CRN Simulations of Lean-Premixed Combustors. In Proceedings of GT2007, Montreal, QC, Canada. Richards, G. et al., Issues for Low-Emission, Fuel-Flexible Power Systems. Progress in Energy and Combustion Science, Volume 27, pp Rutar, Theodora, and P Malte NOx Formation in High-Pressure Jet-Stirred Reactors with Significance to Lean- Premixed Combustion Turbines. Journal of Engineering for Gas Turbines and Power 124: