Today’s combustion equipment market poses significant challenges with the rapidly changing fuels landscape, stricter emissions regulations and tight economic constraints. In the first paper of this series, we discussed how Computational Fluid Dynamics (CFD) alone is not providing the combustion simulation value required in today’s business environment because of CFD’s inherent limitations in handling complex combustion chemistry. In this paper, we will describe how the simulation of real fuel behavior has been achieved through recent advances in our understanding of detailed combustion chemistry and pollutant emissions formation.

1. A WHITE PAPER BY REACTION DESIGN<br />RECENT ADVANCES IN FUEL CHEMISTRYPERMIT NOVEL DESIGN APPROACHES<br />Today’s combustion equipment market poses significant challenges with the rapidly changing fuels landscape, stricter emissions regulations and tight economic constraints. In the first paper of this series, we discussed how Computational Fluid Dynamics (CFD) alone is not providing the combustion simulation value required in today’s business environment because of CFD’s inherent limitations in handling complex combustion chemistry. In this paper, we will describe how the simulation of real fuel behavior has been achieved through recent advances in our understanding of detailed combustion chemistry and pollutant emissions formation.<br />Let us begin by describing how fuel models are used in combustion simulation. We typically think of combustion simply as a single step where fuel and oxygen burn to produce carbon dioxide and water with some pollutants. The reality is that when fuel burns, it undergoes a transformation consisting of thousands of reactions with shortly lived radical species. We are interested in these reactions and radical species because they dictate combustion performance such as ignition, extinction and pollutant formation. However, combustion CFD is limited to using severely simplified fuel models in order to provide a solution in a reasonable amount of time as computational time increases exponentially with model size. While this approach has proven successful in some design applications, combustion CFD with severely simplified models does not have the ability to address today’s key design issues that are related to emissions formation and combustion stability issues such as ignition and flashback. But, there is no method available in CFD that allows the inclusion of more accurate fuel chemistry at a reasonable compute cost. <br />How have engineers dealt with this deficiency? They have used severely simplified models that have been “fitted” to experimental data; an example being the GRI-Mech model for natural gas NOx emissions. This is an acceptable approach as long as your specific conditions are similar to those “fitted” conditions. Often this is not the case as the model is fitted to specific conditions that may not represent the conditions of today’s ultra low NOx combustion technologies. Advanced combustion devices have substantially different conditions due to increases in flue gas recirculation (FGR), lean premixing and other combustion staging techniques. As an example, the GRI-Mech model was developed and validated using 1990’s era boiler technology and works quite well with those applications. However, it has been proven deficient in predicting NOx from modern gas turbines and burner technologies. Still, many of today’s gas turbine and burner CFD simulations continue to use GRI-Mech derived models that cannot yield sufficient accuracy to justify the elimination of experimental tests. <br />Another approach used to deal with the issue of insufficient chemistry detail in CFD is the use of combustion models in conjunction with look-up tables (or cheat sheets) or progress variables that simplify the complex chemistry. This allows the use of more detailed chemistry models but the simplifications that occur in the combustion model prevent accurate prediction of emissions and combustion stability. <br />Reaction Design has significant experience in the use of accurate combustion chemistry through its Model Fuels Consortium (MFC), an industry led effort, advised by academic experts. Our experience with the use of accurate combustion chemistry is similar to that of many researchers in academia, government and industry. This experience shows that you cannot accurately represent ignition, extinction and pollutant formation using the severely simplified chemistry models that are required for practical CFD run times. Ignition delay is a critical parameter that is required in order to determine the location of the flame and combustion stability. The ignition delay results of the GRI-Mech model for natural gas is compared against a more accurate model created from the MFC database are shown in REF _Ref280349266 Figure 1. The X-axis on the figure is the inverse of temperature so low flame temperatures present in modern combustion devices are represented on the right side of the curve. The GRI-Mech model fails to predict ignition delay accurately for pressures that are typical of gas turbine combustors. More importantly, the ignition delay predicted by GRI-Mech shows its greatest error in conditions that are present in modern ultra low NOx combustion on the right side of the curve. GRI-Mech proves successful only at 8atm for higher temperatures that are typical of older combustion system designs but not likely to be found in today’s combustion designs. <br />Let’s look at some NOx results comparing GRI-Mech to modern chemistry models as shown in REF _Ref280349896 Figure 2. In this figure, we compare NOx predictions over a pressure range comparable to that of most gas turbine combustors. As with the ignition delay results, GRI-Mech is most successful in NOx predictions at low pressures that correspond to the conditions present in older boiler and burner designs (i.e. high flame temperature and low pressure). As has been reported by researchers and engine OEM’s recently, GRI-Mech under predicts NOx in today’s designs because it lacks the details required to reflect all of the ways that NOx can be produced. The modern MFC chemistry model effectively predicts NOx over this pressure range showing how recent advances in combustion kinetics can substantially improve emissions predictions over older severely simplified models.<br />Flame speed is another key measure of the quality of a fuel mechanism. Let’s look at how modern MFC chemistry models compare to existing simplified models in REF _Ref280361729 Figure 3 for liquid Jet-A fuel. In this figure, we present flame speed results for a publicly available model that appears on the surface to be accurate. However, this model substantially under predicts flame speed compared to experimental data. The modern MFC model shows much better agreement to experimental data. What does this mean if the model under predicts flame speed? It means that combustor conditions and designs considered to be unstable and not tested may actually be stable. It also means that conditions that were not considered to pose a flashback risk may actually trigger flashback. In the first case it negatively affects product quality and in the latter it leads to increased experimental costs.<br />In this paper, we have demonstrated that accurate combustion simulation requires a level of detail that cannot be run using today’s CFD regardless of how many processors you throw at the job. What is required is an alternate technique that allows the ability to incorporate the required level of accuracy in the model for accurate combustion analysis. This will be the topic of our next whitepaper where we will describe how some combustion system designers are substantially improving the value of combustion simulation to their organization.<br />Figure SEQ Figure ARABIC 1: Modern chemistry models (MFC) match ignition delay data and trends better than GRI-Mech(Petersen, E.L., et al., Proc. Comb. Inst., Vol. 31, 2007, pp. 447-454)<br />Figure SEQ Figure ARABIC 2: Modern chemistry models (MFC) predict NOx formation much better than GRI-Mech(Thomsen, Kuligowski and Laurendeau, Combustion and Flame 119:307-318, 1999)<br />Figure SEQ Figure ARABIC 3: Significant improvement in flame speed prediction with modern MFC chemistry model for Jet-A fuel.<br />