skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Premixer Design for High Hydrogen Fuels

Abstract

This 21-month project translated DLN technology to the unique properties of high hydrogen content IGCC fuels, and yielded designs in preparation for a future testing and validation phase. Fundamental flame characterization, mixing, and flame property measurement experiments were conducted to tailor computational design tools and criteria to create a framework for predicting nozzle operability (e.g., flame stabilization, emissions, resistance to flashback/flame-holding and auto-ignition). This framework was then used to establish, rank, and evaluate potential solutions to the operability challenges of IGCC combustion. The leading contenders were studied and developed with the most promising concepts evaluated via computational fluid dynamics (CFD) modeling and using the design rules generated by the fundamental experiments, as well as using GE's combustion design tools and practices. Finally, the project scoped the necessary steps required to carry the design through mechanical and durability review, testing, and validation, towards full demonstration of this revolutionary technology. This project was carried out in three linked tasks with the following results. (1) Develop conceptual designs of premixer and down-select the promising options. This task defined the ''gap'' between existing design capabilities and the targeted range of IGCC fuel compositions and evaluated the current capability of DLN pre-mixer designs when operatedmore » at similar conditions. Two concepts (1) swirl based and (2) multiple point lean direct injection based premixers were selected via a QFD from 13 potential design concepts. (2) Carry out CFD on chosen options (1 or 2) to evaluate operability risks. This task developed the leading options down-selected in Task 1. Both a GE15 swozzle based premixer and a lean direct injection concept were examined by performing a detailed CFD study wherein the aerodynamics of the design, together with the chemical kinetics of the combustion process, were analyzed to evaluate the performance of the different concepts. Detailed 1-D analysis was performed to provide 1-step NOx and 1-step combustion models that could be utilized in CFD to provide more accurate estimates of NOx for more complicated combustion designs. The swozzle results identified potential problems with flame holding, flashback and with adequate mixing. Flame holding issues were further evaluated with laboratory testing to determine under what conditions a jet in cross flow would flame hold. Additional CFD analysis was also performed on fuel injection from a peg to simulate fuel injection off a vane's trailing edge. This task was concluded with a Conceptual Design Review of the two selected design concepts. (3) Optimize design and re-evaluate operability risks. This task extended the analysis of LDI concepts and increased understanding of the optimal design configuration. Designs were selected for subscale combustion laboratory testing and then modeled using CFD to validate CFD methodology. CFD provided a good qualitative match and reasonable quantitative match with the test results. Tests and CFD modeling indicated a path to low NOx combustion with no diluent addition. Different swirler designs were also evaluated and the most promising, a counter rotating swirler, was selected for further evaluation. CFD modeling was performed and the design was optimized to improve mixing. CFD modeling indicated the potential for low NOx combustion without diluent addition. CFD was validated against cold flow testing on a swirler using helium injection in place of hydrogen. Further validation work is still needed to ensure the ability to accurately model the mixing of swirling flows. Entitlement testing was performed on a perfectly premixed H2/N2/air mixture. Results showed that low NOx could be obtained at the temperatures of interest (7FB conditions) with no diluent addition. Results also showed that further NOx reductions might be possible by taking advantage of the very rapid H2 reaction to reduce combustor length and hence residence time. These results also indicate that there is a high probability that a combustor can be built that will meet DOE emissions goals. Based on these results and the promising results for the optimized swirler, the swirler will be the primary concept and the LDI will be maintained in reserve if flameholding issues cannot be resolved.« less

Authors:
; ; ;
Publication Date:
Research Org.:
General Electric Company
Sponsoring Org.:
USDOE
OSTI Identifier:
889756
DOE Contract Number:
FC26-03NT41893
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
08 HYDROGEN; AERODYNAMICS; COMBUSTION; COMBUSTORS; COMPUTERIZED SIMULATION; CONFIGURATION; FLAMES; FLASHBACK; FLUID MECHANICS; HELIUM; HYDROGEN; HYDROGEN FUELS; KINETICS; NOZZLES; SOLVENTS; STABILIZATION; TESTING; VALIDATION

Citation Formats

Benjamin P. Lacy, Keith R. McManus, Balachandar Varatharajan, and Biswadip Shome. Premixer Design for High Hydrogen Fuels. United States: N. p., 2005. Web. doi:10.2172/889756.
Benjamin P. Lacy, Keith R. McManus, Balachandar Varatharajan, & Biswadip Shome. Premixer Design for High Hydrogen Fuels. United States. doi:10.2172/889756.
Benjamin P. Lacy, Keith R. McManus, Balachandar Varatharajan, and Biswadip Shome. Fri . "Premixer Design for High Hydrogen Fuels". United States. doi:10.2172/889756. https://www.osti.gov/servlets/purl/889756.
@article{osti_889756,
title = {Premixer Design for High Hydrogen Fuels},
author = {Benjamin P. Lacy and Keith R. McManus and Balachandar Varatharajan and Biswadip Shome},
abstractNote = {This 21-month project translated DLN technology to the unique properties of high hydrogen content IGCC fuels, and yielded designs in preparation for a future testing and validation phase. Fundamental flame characterization, mixing, and flame property measurement experiments were conducted to tailor computational design tools and criteria to create a framework for predicting nozzle operability (e.g., flame stabilization, emissions, resistance to flashback/flame-holding and auto-ignition). This framework was then used to establish, rank, and evaluate potential solutions to the operability challenges of IGCC combustion. The leading contenders were studied and developed with the most promising concepts evaluated via computational fluid dynamics (CFD) modeling and using the design rules generated by the fundamental experiments, as well as using GE's combustion design tools and practices. Finally, the project scoped the necessary steps required to carry the design through mechanical and durability review, testing, and validation, towards full demonstration of this revolutionary technology. This project was carried out in three linked tasks with the following results. (1) Develop conceptual designs of premixer and down-select the promising options. This task defined the ''gap'' between existing design capabilities and the targeted range of IGCC fuel compositions and evaluated the current capability of DLN pre-mixer designs when operated at similar conditions. Two concepts (1) swirl based and (2) multiple point lean direct injection based premixers were selected via a QFD from 13 potential design concepts. (2) Carry out CFD on chosen options (1 or 2) to evaluate operability risks. This task developed the leading options down-selected in Task 1. Both a GE15 swozzle based premixer and a lean direct injection concept were examined by performing a detailed CFD study wherein the aerodynamics of the design, together with the chemical kinetics of the combustion process, were analyzed to evaluate the performance of the different concepts. Detailed 1-D analysis was performed to provide 1-step NOx and 1-step combustion models that could be utilized in CFD to provide more accurate estimates of NOx for more complicated combustion designs. The swozzle results identified potential problems with flame holding, flashback and with adequate mixing. Flame holding issues were further evaluated with laboratory testing to determine under what conditions a jet in cross flow would flame hold. Additional CFD analysis was also performed on fuel injection from a peg to simulate fuel injection off a vane's trailing edge. This task was concluded with a Conceptual Design Review of the two selected design concepts. (3) Optimize design and re-evaluate operability risks. This task extended the analysis of LDI concepts and increased understanding of the optimal design configuration. Designs were selected for subscale combustion laboratory testing and then modeled using CFD to validate CFD methodology. CFD provided a good qualitative match and reasonable quantitative match with the test results. Tests and CFD modeling indicated a path to low NOx combustion with no diluent addition. Different swirler designs were also evaluated and the most promising, a counter rotating swirler, was selected for further evaluation. CFD modeling was performed and the design was optimized to improve mixing. CFD modeling indicated the potential for low NOx combustion without diluent addition. CFD was validated against cold flow testing on a swirler using helium injection in place of hydrogen. Further validation work is still needed to ensure the ability to accurately model the mixing of swirling flows. Entitlement testing was performed on a perfectly premixed H2/N2/air mixture. Results showed that low NOx could be obtained at the temperatures of interest (7FB conditions) with no diluent addition. Results also showed that further NOx reductions might be possible by taking advantage of the very rapid H2 reaction to reduce combustor length and hence residence time. These results also indicate that there is a high probability that a combustor can be built that will meet DOE emissions goals. Based on these results and the promising results for the optimized swirler, the swirler will be the primary concept and the LDI will be maintained in reserve if flameholding issues cannot be resolved.},
doi = {10.2172/889756},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Fri Dec 16 00:00:00 EST 2005},
month = {Fri Dec 16 00:00:00 EST 2005}
}

Technical Report:

Save / Share:
  • Due to increasingly stringent air quality requirements stationary power gas turbines have moved to lean-premixed operation, which reduces pollutant emissions but can result in flashback. Flashback can cause serious damage to the premixer hardware. Curtailing flashback can be difficult with hydrocarbon fuels and becomes even more challenging when hydrogen is used as the fuel. The two main approaches for coping with flashback are either to design a combustor that is resistant to flashback, or to design a premixer that will not anchor a flame if flashback occurs. Even with a well-designed combustor flashback can occur under certain circumstances, thus itmore » is necessary to determine how to avoid flameholding within the premixer passageways of a gas turbine. To this end, an experiment was designed that would determine the flameholding propensities at elevated pressures and temperatures of three different classes of geometric features commonly found in gas turbine premixers, with both natural gas and hydrogen fuel. Experiments to find the equivalence ratio at blow off were conducted within an optically accessible test apparatus with four flameholders: 0.25 and 0.50 inch diameter cylinders, a reverse facing step with a height of 0.25 inches, and a symmetric airfoil with a thickness of 0.25 inches and a chord length of one inch. Tests were carried out at temperatures between 300 K and 750 K, at pressures up to 9 atmospheres. Typical bulk velocities were between 40 and 100 m/s. The effect of airfoil’s angle of rotation was also investigated. Blow off for hydrogen flames was found to occur at much lower adiabatic flame temperatures than natural gas flames. Additionally it was observed that at high pressures and high turbulence intensities, reactant velocity does not have a noticeable effect on the point of blow off due in large part to corresponding increases in turbulent flame speed. Finally a semi empirical correlation was developed that predicts flame extinction for both natural gas and hydrogen flames.« less
  • As simulation capability improves exponentially with increasingly more cost effective CPUs and hardware, it can be used ?routinely? for engineering applications. Many commercial products are available and they are marketed as increasingly powerful and easy to use. The question remains as to the overall accuracy of results obtained. To support the validation of the CFD, a hierarchical experiment was established in which the type of fuel injection (radial, axial) as well as level of swirl (non-swirling, swirling) could be systematically varied. The effort was limited to time efficient approaches (i.e., generally RANS approaches) although limited assessment of time resolved methodsmore » (i.e., unsteady RANS and LES) were considered. Careful measurements of the flowfield velocity and fuel concentration were made using both intrusive and non-intrusive methods. This database was then used as the basis for the assessment of the CFD approach. The numerical studies were carried out with a statistically based matrix. As a result, the effect of turbulence model, fuel type, axial plane, turbulent Schmidt number, and injection type could be studied using analysis of variance. The results for the non-swirling cases could be analyzed as planned, and demonstrate that turbulence model selection, turbulence Schmidt number, and the type of injection will strongly influence the agreement with measured values. Interestingly, the type of fuel used (either hydrogen or methane) has no influence on the accuracy of the simulations. For axial injection, the selection of proper turbulence Schmidt number is important, whereas for radial injection, the results are relatively insensitive to this parameter. In general, it was found that the nature of the flowfield influences the performance of the predictions. This result implies that it is difficult to establish a priori the ?best? simulation approach to use. However, the insights from the relative orientation of the jet and flow do offer some guidance for which approach to take. Overall, the results underscore the importance of model ?anchoring? (i.e., ?tuning? the model to provide ?reasonable? agreement with a well characterized geometry/flow). Finally, the results obtained have been carefully compiled into a standalone database following a standard format that is contained in an Appendix. This database is thus available for use by others for CFD modeling evaluations.« less
  • The embrittiement of high strength steels due to the action of H introduced by Cd-electroplating was studied in sustained-load, rotating beam fatigue, and bending tests. Strength levels from 180,000 to 300,000 psi as suitable for the various steels were examined for a variety of initial conditions of stress concentration. All steels were found to be embrittled in some measure after Cd-plating and this embrittlement could not be fully eliminated, as determined in the bend test, through the baking treatment used. The improvement in properties which did result from baking was promoted by a redistribution and not an elimination of Hmore » from the steel. Failure promoted by Cd-plating is affected by the experimental conditions and has been discussed at length in the report. In the H bearing zone a crack is initiated and then depending on the experimental conditions may propagate to failure of the cross section through overloading. Crack development is apparently dependent, in part, on the composition and is minimized by reduction in C content or by an increase in Si content. Bath the sustained-load and bend tests are suitable tests for evaluation of H embrittlement in ultra-high strength steels. The rotating beam fatigue test is a relatively insensitive test of H embrittlement, but can be used to provide an excellent measure of the ''static'' notch strength of the steel. (auth)« less
  • OAK B188 Initial Screening of Thermochemical Water-Splitting Cycles for High Efficiency Generation of Hydrogen Fuels Using Nuclear Power There is currently no large scale, cost-effective, environmentally attractive hydrogen production process, nor is such a process available for commercialization. Hydrogen is a promising energy carrier, which potentially could replace the fossil fuels used in the transportation sector of our economy. Fossil fuels are polluting and carbon dioxide emissions from their combustion are thought to be responsible for global warming. The purpose of this work is to determine the potential for efficient, cost-effective, large-scale production of hydrogen utilizing high temperature heat frommore » an advanced nuclear power station. Almost 800 literature references were located which pertain to thermochemical production of hydrogen from water and over 100 thermochemical watersplitting cycles were examined. Using defined criteria and quantifiable metrics, 25 cycles have been selected for more detailed study.« less
  • OAK B188 High Efficiency Generation of Hydrogen Fuels Using Nuclear Power