Improving Design of SCR Systems with CFD Modeling Engineering International Improving Design of SCR...

Reaction Engineering International

Improving Design of SCR Systems with CFD Modeling

B. Adams, C. Senior


DOE Environmental Controls ConferenceMay 16-18, 2006 Pittsburgh, PA


Project Objective

• Demonstrate value of CFD in SCR design and tuning– Improved insight of flue gas properties through

AIG, SCR catalyst, and ductwork – Impact of flue gas profiles and flow rates on SCR

performance– Guidance for matching AIG flows with flue gas

profiles and resulting impact of AIG tuning


Project Approach• Use CFD and 1-D reactor modeling to predict

flow properties and NO reduction in 550 MW coal-fired boiler SCR system– Flow patterns and velocity profiles– Overall pressure drop through duct work– Temperature distribution– NH3 distribution from ammonia injection grid (AIG)– NO reduction through SCR catalyst– Ammonia slip after catalyst– NO reduction sensitivities


Modeling Approach

• Three model components:– Duct flow model– Ammonia Injection Grid

(AIG) submodel– SCR catalyst submodel

Duct Flow Model

SCR Reactor

Duct Flow Model

Duct Flow Model

Detailed AIG Model

- ∆P- Flow distribution

- NH3 distribution

- SCR model inlet properties• NH3 distribution• NO distribution• Flow distribution

- NO reduction- NH3 slip

- ∆P- Flow field- NH3 slip distribution- NO distribution


Inlet conditions to each

catalyst cell

SCR Model – Implementation1-D plug flow model with heterogeneous kinetics for NO reduction by NH3 over vanadia/titania catalysts Solves gas transport and surface reactions along catalystNumber of “active sites” for reaction calibrated with design dataPredicts dependencies of NH3/NO ratio, temperature, residence time and flue gas composition

NH3

NO

> 200

< 100

ppm

> 230

< 150

ppm

Results at SCR Inlet

> 5

< 0

m/sVelocity

Surface Flux

NONH3

N2H2O

NONH3

N2H2O

1-D catalystchannelmodel

Outlet conditions from each catalyst

cell to CFD


SCR GeometriesUnit 3-1 Unit 3-2

Model inlet Model out

Measurement inlet

Measurement out

Flow Direction

Flow Direction

Flow Direction

Model inletModel out

Measurement inlet

Measurement out

Unit 3-2 has shorter run length between turn and AIG and shorter run length between AIG, turn, and guide vanes


Gas Velocity Profile

MCR

30% Load

MCR

Velocity profiles can suggest improvements to guide vane design

Unit 3-1


0

20

40

60

80

100

0 2 4 6 8 10 12 14

Location

Pres

sure

, mm

H2 O

Pressure Drop

0

2

4

6

8

10

12

0 2 4 6 8 10 12 14

Location

Pres

sure

, mm

H2 O

SCR Reactor

SCR ReactorMCR 30% Load

Unit 3-1

Full Load

Part Load

Profiles identify specific contributors as well as total system pressure drop


Flow ComparisonsUnit 3-1 Unit 3-2 Total

Flow, 103 kg/hr 1128.5 1128.5 2257.0MCR ∆P (mm H2O, SCR) 40.0 37.5 -

∆P (mm H2O, N1 to N3 or N4 to N6) 89.1 79.2 -Total ∆P (mm H2O) 94.0 84.0 -

Flow, 103 kg/hr - 694.2 -Model 30% Load ∆P (mm H2O, SCR) - 13.1 -

(July 17, 2004) ∆P (mm H2O, N1 to N3 or N4 to N6) - 27.7 -Total ∆P (mm H2O) - 29.0 -

Flow, 103 kg/hr 431.5 431.5 863.030% Load ∆P (mm H2O, SCR) 5.0 4.9 -(design) ∆P (mm H2O, N1 to N3 or N4 to N6) 11.7 10.4 -

Total ∆P (mm H2O) 12.0 10.8 -

Flow, 103 Nm3/hr 948.3 871.0 1819.2MCR Flow, 103 kg/hr 1260.8 1158.0 2418.9

(July 16, 2004) Design SCR ∆P (mm H2O) 39.8 37.3 -Measured ∆P (mm H2O, N1 to N3 or N4 to N6) 98.4 73.7 -

Flow, 103 Nm3/hr 576.2 522.1 1098.330% Load Flow, 103 kg/hr 766.1 694.2 1460.3

(July 17, 2004) ∆P (mm H2O, N1 to N3 or N4 to N6) 44.1 34.1 -

• Predictions consistent with the measured data

• Differences likely due to:– Inlet flow profile– Different flow

rates– Measurement

locations/grid– Model accuracy


NO Reduction Dependencies• NO reduction depends on:

– NH3/NO ratio – insufficient ammonia (ratio<1.0) will not reduce all available NO (NO remaining)

– Residence time – insufficient residence time will not allow NH3 enough time to react with NO (NO and NH3 remaining)

• NH3 slip depends on:– NH3/NO ratio – too much ammonia (ratio>1.0) will reduce all

NO (NH3 remaining)– Residence time – insufficient residence time will not allow

NH3 enough time to react with NO (NO and NH3 remaining)


NO Concentration ProfileUnit 3-1 vs 3-2 at MCR

Flue Gas

Flue Gas

Both cases show biased NOx profiles


NH3 Concentration ProfileUnit 3-1 vs 3-2 at MCR

Flue Gas

Flue GasBoth cases show biased NH3 profiles


SCR Model Predictions INLET OUTLET INLET OUTLET

NH3, ppm @ 6%O2 Dry 171 4.0 171 4.2MCR, Fully Open AIG NO, ppm @ 6%O2 Dry 210 42.7 210 43.0

NO Reduction, % - 79.6 - 79.5

NH3, ppm @ 6%O2 Dry 171 0.0 171 0.230% load, Fully Open AIG NO, ppm @ 6%O2 Dry 210 38.8 210 38.9

NO Reduction, % - 81.4 - 81.4

30% load with higher O2 NH3, ppm @ 6%O2 Dry - - 171 1.5Fully Open AIG NO, ppm @ 6%O2 Dry - - 210 40.2

NO Reduction, % - - - 80.8

Unit 3-1 Unit 3-2

• Ammonia slip likely due to higher than design velocities in catalyst (lack of residence time)

• Limited residence time also reduces NO reduction• Will more uniform NH3/NO profiles improve performance?


Outlet NO Profile at MCR

Unit 3-1: measured

Unit 3-1: predicted

Unit 3-2: measured

Unit 3-2: predicted

A1 A81

5

B8 B1

4

3

2

1

5

4

3

2

A1 A81

5

B8 B1

4

3

2

1

5

4

3

2

STD = 16.54%Average = 54.2 ppm




Measurement taken with fully open AIG nozzles (NO ppm, dry)


NO Exit Profile ResultsComparison of predicted and measured NO profiles

MCR(outlet), measured on 07/14/04

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

Ratio of m easured NO to average inle t NO

Rat

io o

f pre

dict

ed N

O to

ave

rage

inle

t NO

Unit 3-1

Unit 3-2

Measurement taken with AIG nozzles fully open


AIG Tuning• Background: Modeling results with fully open AIG

nozzles indicate a bias in the flue gas flow in the AIG region, leading to NH3 biasing at the catalyst inlet

• Objective: Use CFD modeling to simulate AIG tuning and show how CFD can be used to guide AIG in the SCR duct (Unit 3-2)

• Approach: Use zonal AIG biasing to achieve more uniform NH3/NO ratio (0.814)


Using CFD For Tuning

• CFD can show profiles through system

• CFD can be used to trace source of profile

• Streamlines correlate local NO at outlet with corresponding NH3/NO entering the catalyst, and ultimately with the corresponding AIG nozzles


NH3 Concentration ProfileUnit 3-2 at MCR

Flue Gas In

Flue Gas Out

Flue Gas In

Flue Gas Out

Fully Open AIG Nozzles (Uniform NH3 injection)

AIG Tuning (Biased NH3 injection)


NH3/NO Ratio ProfileUnit 3-2 at MCR

Flue Gas In

Flue Gas Out

Flue Gas In

Flue Gas out




NO Concentration ProfileUnit 3-2 at MCR

Flue Gas In

Flue Gas OutFlue Gas Out

Flue Gas In




Predicted Effects of AIG Tuning

NH3/NO RatioSTD = 8.1%

Average = 0.814

NH3/NO RatioSTD = 1.3%

Average = 0.814

NOSTD = 25.7%

Average = 43.0 ppm

NOSTD = 4.7%

Average = 42.5 ppm

UniformNH3+Air Flow

BiasedNH3+Air Flow

Catalyst Inlet

MeasurementOutlet

Dry, 6% O2


SCR Model Predictions

• Why didn’t the NO reduction improve with a more uniform NH3/NO ratio at the SCR reactor inlet?– More uniform NH3/NO ratio improved reduction in low ratio regions,

but decreased reduction in high ratio regions– Average residence time unchanged, still too short to allow full

conversion to 81.4%– Both results were close to the optimal reduction specified in the

catalyst design for this load, NH3/NO ratio, and catalyst size (80%)

AIG ConditionsINLET OUTLET INLET OUTLET

NH3, ppm @ 6%O2 Dry 171 4.2 171 3.8NO, ppm @ 6%O2 Dry 210 43.0 210 42.5

SCR Inlet NH3/NO Mole Ratio 0.814 0.814

NO Red, % 79.50 79.71NH3 Slip, ppm @ 6%O2 Dry 4.2 3.8

Unit 3-2: MCRUniform Tuning


SCR Model: Effect of Gas Velocity on NO Reduction

Velocities plotted are from actual duct velocities

Average vel: 4.56 m/s

All non-blue colors represent velocities

that could lead to suboptimal NO

reduction

Velocity profile at SCR catalyst inlet

76

77

78

79

80

81

82

0 2 4 6 8

Catalyst Channel Gas Velocity, m/s

NO

Red

uctio

n, %

Suboptimal NO reduction

Average velocity: 4.56 m/s

2.2 m/s

81.4%

> 6

< 2.2

m/s

At MCR, catalyst design does not allow sufficient residence time to react all NH3 with NO, even with optimal NH3/NO ratio


Predictions at Higher NH3/NO Ratios

75

80

85

90

95

0.80 0.85 0.90 0.95 1.00 1.05

NH3/NO Ratio

NO R

educ

tion,

%

Fully Open AIGTuned AIGDesign Data

0

5

10

15

20

25

0.80 0.85 0.90 0.95 1.00 1.05

NH3/NO Ratio

NH

3 Sl

ip, p

pm (d

ry, 6

% O

2)

Fully Open AIGTuned AIGDesign Data

NH3/NO Ratio vs NO Reduction NH3/NO Ratio vs NH3 Slip

MCR LoadMCR Load


0

5

10

15

20

25

0.00 0.50 1.00 1.50 2.00 2.50

Residence Time, s

NH

3 S

lip, p

pm (d

ry 6

% O

2) Fully Open AIGTuned AIG

Predictions at Higher Residence Times

Residence Time vs NH3 Slip

MCR

90

91

92

93

94

95

96

97

98

99

100

0.00 0.50 1.00 1.50 2.00 2.50

Residence Time, s

NO

Red

uctio

n, %

Fully Open AIGTuned AIG

Residence Time vs NO Reduction

NH3/NO = 1.0

NH3/NO = 1.0

MCR


What Can Model Do• Predict performance and identify potential design

problems before design/installation– Predict NO reduction and ammonia slip (measurements verify

accuracy for the same NH3/NO ratio, load, catalyst design)– Guide how much NH3 needed to reach NO reduction goal– Guide amount of catalyst needed for NO reduction goal– Suggest duct designs to provide more uniform flow

• Aid tuning and troubleshooting– Illustrate tuning impacts

• Guidance on how to tune AIG injection • Determine if tuning worthwhile for a particular design

– Identify measurement locations for testing and/or analyzer


Thank You

[email protected]

Improving Design of SCR Systems with CFD Modeling Engineering International Improving Design of SCR...

Documents

Transcript of Improving Design of SCR Systems with CFD Modeling Engineering International Improving Design of SCR...