WPCA/DUKE SCR Seminar

1 1NH3 Injection/Gas Mixing and the Effect on Reactor PerformancePresented byKevin RogersThe Babcock & Wilcox CompanyWPCA/ duke SCR SeminarTuesday June 7, 20051:30PM-2 Breakdown(1)Influence Ammonia & NOx Molar Ratio Blending has on Reactor Performance(2)Ammonia Injection & Gas Mixing to Achieve Molar Ratio Distribution GoalsTwo General Areas of Discussion:123456 ABCDEF102030405060 NOx conc. (ppm)PointPortDistribution Contour DescriptionsDistribution goals commonly stated as X% of the Data to fall within Y% of the Mean(Arithmetic Average):Often more than one is applied to single distribution goal ( velocity or NH3/NOx molar ratio at a given location) 80% of points to be within 10% of the Mean 80% of points to be within 20% of the Mean 85% of points to be within 15% of the Mean 100% of points to be within 25% of the MeanSystem blending and reactor performance can be more easily correlated when the distributions are adequately defined by a single of Variance Provides a Single Description of the Degree of MaldistributionCoefficient of Variance (COV, Cv, CV).

2 Standard deviation as a percentage of the arithmetic average Often called the % RMS The Root-Mean-Square of the Deviations ( ), expressed as a percentage of the mean (x)%100xCv = = =niixxn12)()1(1 == Distribution RelationshipNeed to measure many points to capture 3 50 pt or less profiles often have Maximums and Minimums below 3 .6 Relative Effect of Distributions on Reactor Efficiency @ 70% DeNOx05101520 Coefficient of Variation (Cv), %Average NOx Removal EfficiencyVelocityMole RatioTemperature70% NOx Removal@ Uniform of Variation (Cv), %Average NOx Removal EfficiencyVelocityMole RatioTemperature90% NOx Removal@ Uniform ConditionsRelative Effect of Distributions on Reactor Efficiency @ 90% Catalyst Volume vs Inlet NH3/NOxCvand NOxRemoval (@ 2 ppm Ammonia Slip) Molar Ratio Cv, %CVR90_50090_30090_10085_50085_30085_100 80_50080_30080_100 CVR Plt_KR040802_1 Increasing Catalyst Peak Ammonia Slip vs Molar Ratio (NH3/NOx)

3 CvEffect of Inlet NOxGreater as Efficiency Increases0246810NH3/NOx Cv, %Localized Peak NH3 Slip90% DeNOx500 ppm Inlet NOxOut-of-Range above 8% Cv90% DeNOx100 ppm Inlet NOx80% DeNOx500 ppm Inlet NOx80% DeNOx100 ppm Inlet NOxLocal Peaks Caused by High Inlet NH3/NOx at 2 Standard Deviations above AverageConstant Avg NH3 Slip = 2 ppmIncreasing Reactor Outlet Local Peak Influence of Inlet Distributions on Reactor Outlet NOx Cv0102030405060404550556065707580859095 Ammonia-to-NOxCv @ Cv @ Cv @ DeNOx Efficiency, %Reactor Outlet NOx Profile Cv, %Inlet NOx = 300 ppmSlip = 2 ppmIncreasing Reactor Outlet NOx the following ExampleDeNOx Efficiency: 90 %Inlet NOx: 270 ppmSlip: 1 ppmAverage Mole Ratio= + 1/270= NH3/NOx INExcess Reagent= ( ) = NH3/NOx REMOVEDAt an Inlet NH3/NOxCv 5%Maximums could easily range between +2 and +3 Or 2 x 5 = +10% to 3 x 5 = +15%Or x = x = Much More Catalyst to approach 100% @ efficiency not easily achieved thus higher local slipAn Infinite Amount of Catalyst will Not Allow Excess Ammonia > to be Consumed (NOx has been Depleted).

4 12 Assume a Typical Reactor Gas Flow Deviation from MeanFlow Cv 4%.13 Then Consider Two Possible Inlet NH3 Deviation from Mean% Deviation from MeanOne @ Cv 5%Avg Slip PPMA nother @ Cv Slip Increases to PPM to maintain 90% Outlet NOx DistributionsPPMWhen Inlet NH3/NOxCv 5%Avg Slip PPMWhen Inlet NH3/NOxCv Slip Increases to PPM to maintain 90% 46%Cv 69%.15 Resultant Outlet NH3 Slip DistributionsPPMWhen Inlet NH3/NOxCv 5%Avg Slip PPMWhen Inlet NH3/NOxCv Slip Increases to PPM to maintain 90% removalPPMCv 120%Cv 220% Distributions-Less Critical to Performance-Typical Concern - Deposition & Erosion-Cv 15% For Catalyst Approach Generally AcceptableTemperature Distributions-Typically Less Critical to Performance-Sintering & BiSulfate Formation Concerns-Min/Max 20F to 50F Generally AcceptableDistributions other than NH3 Ratio Distributions Less Critical Below 70% DeNOx Efficiency Becoming Rapidly More Important Above 80% 90% Removal of 300 ppm Inlet NOx @ 2 ppm Slip (Not Practical/Possible with NH3/NOx Cv = 10%) Critical for 90% DeNOx and Above Cv<3% for 95% DeNOx Excess Reagent Reduces Sensitivity[slip/inlet NOx] (Gas Fired - High Slip Coal Fired - Low Slip).

5 18 High Efficiency ( 90%) = More Stringent Requirements Increased AIG Flow Uniformity Increased Dispersion @ AIG Increased Mixing Energy Increased Flue Length Sacrifice of Velocity Profiles to Improve NH3/NOxProfile Blending of Peak Slip Downstream of Reactor Tolerance to Mixing Error Rapidly Reduced for High DeNOx Removal Efficiency (Cv drift 5% 8% is a problem) Ammonia Injection & Gas Mixing Technology Why Mix ? Technology Fundamentals AIG & Mixer Design Field of BlendingDecreasing Scale of SegregationDecreasing Intensity of SegregationDecreasing Intensity of is Dominated by Three Primary Distribution How well ammonia is deposited over the space or profile of an initial NOx distribution (Scale of Segregation) and Folding- How larger regions of varying concentrations are being thinned and spread across the area via macroscopic turbulence (Scale of Segregation) Diffusion Occurring simultaneously throughout and necessary for the final an approach to complete homogeneity influenced & facilitated by the increased surface area brought about by 1& 2 (Intensity of Segregation)AIG & Mixing of Initial Ammonia Dosing The relative matching of regional or local ammonia flow to the regional or local flue gas flow -(Fairly stable flow profile contours through the plane of ammonia injection typically allow for stable flow dosing)

6 Dosing- The relative matching of regional or local ammonia flow to the regional or local NOx Concentration -(NOx profile contours are often not as stable with regard to the positioning of high a low concentration regions and can thus be difficult to adjust to).AIG & Mixing of Initial Ammonia Dosing Cont Dosing Defined by the spacing of the injection points and thus the injection point quantity: (Initial Scale of Ammonia Segregation) the number improves the final blend to a point of diminishing few reduces system effectiveness and few can over-sensitize placement of injection points relative to mixer vane & Mixing Mixing ApproachRapid Full Duct Dispersion AIGE fficient Use of Duct Length with Rapid Shear, Controlled Turbulence and Duct Macro Flow DevelopmentTest Stands From Simple to Complicated Size, Cost, Testing & Run Times More Complicated for Analysis of Arrangement Performance (the performance of devices in complex arrangements)

7 Predictive Formula Structured Array of Test Programs Initial Application & OptimizationField Effectiveness0510152025303540-1012345678 L/DCO Tracer Concentration COV, %No Mixer '1 = '3 = '7 = No MixerForced of Bend Design012345678910111213140 10203040506070 Equivalent Centerline Distance, ftNOx COV, %No MixerMixer_Std BendMixer_Modified BendAIGM ixerBendHoodCatalyst a b of Injection Point Quantity0204060801001201401601802000 10203040506070 Equivalent Centerline Distance, ftNH3 COV, %Na = pts/m2Na = pts/m2Na = pts/m2 AIGM ixerBendHoodCatalystMixer Inlet Location Unit Molar Ratio Drift(w/o Static Mixing)90%87%83%85%87%85%0246810120 10203040506070 Reactor Outlet NOx COV, %Inlet NH3/NOx Molar Ratio COV, % Unit Molar Ratio Drift(with Static Mixing)93%92%90%92%90%0123450 10203040506070 Reactor Outlet NOx COV, %Inlet NH3/NOx Molar Ratio COV, % Based Sensitivity Analysis051015202505101520253035 Inlet NOx COV, %NH3/NOx COV, %No Forced MixingWith Forced MixingSystem Configuration: +2 LMixer ConsiderationsForced Mixing for Stable UniformityStability W/O Forced Mixing: Molar Ratio COV 2% 8% 15% With Forced Mixing: Molar Ratio COV 2% 5%Important Optimization Parameters Initial Dosing (AIG Design, NOx profile, Flow Profile) Injection Nozzle Quantity Directivity of Mixing Effectiveness System Length vs Energy Efficiency Arrangement Design (Dampers, Exp/Con, Bends, Hoods, etc.)

8 33 Design Considerations Cont d NOx removal duty (Inlet loading, % removal, allowable slip) NOx & Temperature Profiles Allowable draft loss Straight flue lengths available Bends & overall flue geometry AIG inlet gas flow profile Quantity of injection points Mixer Placement Flow correction prior to catalyst .34 Poor Arrangements for High RemovalSimple Close-Coupled Arrangements Will Not Workfor High NOx Removal SCR ProjectsInsufficient distance to inject and blend for very low and stable molar ratio maldistributionsShort opportunity to blend peak local slip concentrations prior to for Promotion of MixingAIG LocationMixer LocationAll mixing considered complete prior to flow straightening into the catalystGood mixing distance prior to bendFlow correction as required prior to the AIG (Often perforated plate and/or vanes)Permanent sample grid LocationGreater opportunity for blending of peak local slip concentrations prior to the Field TuningIncrease ammonia flow in region of higher gas flow through the AIG for improved initial ammonia dosingDecrease ammonia flow in region of lower gas flow through the AIG for improved initial ammonia dosingWhen units are designed with static mixers, this simple initial tuning trial, based on model study results for the flow profile through the AIG, is often sufficient to obtain an acceptable Field TuningReactor Outlet Distribution 275 MW275 MW350 (ppm) LoadFinal LoadNH3 Flow% Reduction123456 ABCDEF102030405060 NOx conc.

9 (ppm)PointPortDistribution Contour Field TuningGiven:270 ppm Inlet ppm ammonia slipMd = + = NOx Cv Target 44% - 46% @ inlet NH3/NOx Cv 5% NOx Cv is Solving for the inlet NH3/NOx Cv @ outlet NOx Cv is yields an inlet NH3/NOx Cv 3% .39 Field Uniformity Performance Mixer DesignOtherOtherB&WB&WB&W Removal Efficiency, %87%86%92%93%89% SCR Outlet NOx COV18%23%16%10%24% SCR Inlet NH3/NOx Relative Mix Relative Shock Site