Tools for Optimizing Gas Turbine SCR Performance

Transcription

1 Tools for Optimizing Gas Turbine SCR Performance L. J. Muzio, T. D. Martz Fossil Energy Research Corp. Laguna Hills, CA Reinhold 213 NOx-Combustion Round Table February 18, 213 Salt Lake City, Utah 1

2 Simple Cycle Gas Turbine SCR Ammonia Injection Grid (AIG) SCR Catalyst Perforated Plate Diffuser Vanes CO Catalyst SCR Performance Parameters: - NO x Reduction - Ammonia Slip Uniform NH3/NOx Profile at Catalyst Inlet is Critical! Tempering Air Flue gas 75-1 F NH3 NH3 Dilution Air 2

3 Cogeneration Gas Turbine SCR No Diffuser Vanes No Perforated Plates No Tempering Air Ammonia Injection Grid (AIG) CO Catalyst SCR Catalyst SCR Performance Parameters: - NO x Reduction - Ammonia Slip Uniform NH3/NOx Profile at Catalyst Inlet is Critical! Steam Tube Banks Flue gas ~55-65 F NH3 NH3 Dilution Air 3

4 Optimizing Gas Turbine SCR Performance Topics Troubleshooting Catalyst Inlet NH 3 /NO x Distribution and AIG Tuning Identifying Flue Gas Bypass (e.g., Leaky Catalyst Seals) Measuring Catalyst Velocity Distribution SCR Impact on CEMs Relative Accuracy Catalyst Management/Measuring Catalyst Activity 4

5 NH3-slip, ppm Troubleshooting Compliance Test NH 3 slip is too high Why? 4 2 Compliance Window NOx, ppm 5

6 NH3-slip, ppm Minimum Reactor Potential Why? Catalyst Activity (K)? Reactor Potential? How active the material is in reducing NOx f(material, geometry) Compliance Test Ability of the catalyst bed to reduce NO x RP= K*A sp *V cat /Q fg NH3 slip=2.5ppm NH3 slip=5 ppm NOx-out=5 ppm Compliance Window NOx, ppm Inlet NOx, ppm Poor NH3/NOx Distribution? Want NH 3 /NO x uniform across the catalyst Local NH 3 /NO x >1=NH 3 slip 6 Flue Gas Bypass? Any bypass by the catalyst increases stack NO x & NH 3

7 Minimum Reactor Potential Reactor Potential Ability of the catalyst bed to reduce NO x RP= K*A sp *V cat /Q fg RP=f(NO x i, DNO x,and NH 3 slip) NH3 slip=2.5ppm NH3 slip=5 ppm NOx-out=5 ppm Inlet NOx, ppm 7

8 NH3-slip, ppm Expected Performance Based on Catalyst Properties Compliance Test Expected (RMS=1%) Compliance Window NOx, ppm Reactor Potential (RP) and is catalyst activity are typically adequate for new or young catalyst Exception: unexpected high temperatures 8

9 NH3-slip, ppm Poor NH3/NOx Distribution? Compliance Test Expected (RMS=1%) RMS=3% Compliance Window NOx, ppm 9

10 NH3-slip, ppm Flue Gas Bypass? With a single test it is difficult to determine the reason for non-compliance Compliance Test Expected (RMS=1%) RMS=3% Bypass NOx, ppm 1

11 NH3-slip, ppm Perform Additional Tests Compliance Window NOx, ppm 11

12 NH3-slip, ppm Poor NH3/NOx Distribution? Compliance Test Expected (RMS=1%) RMS=3% Compliance Window NOx, ppm Poor NH 3 /NO x distribution does not explain NH 3 slip at high NO x levels 12

13 NH3-slip, ppm Flue Gas Bypass? Compliance Test Expected (RMS=1%) RMS=3% Bypass NH 3 slip at high NO x suggests flue gas bypass NOx, ppm 13

14 Catalyst Inlet NH3/NOx Distribution and AIG Tuning 14

15 NH3 Slip, ppm NH3 Slip, ppm NH 3 /NO x Distribution and AIG Tuning 1 RMS=5% RMS=1% RMS=15% RMS=25% 8 New Catalyst Near Catalyst End-of-Life Catalyst Aging NOx Reduction, %

16 Sample Probe Grid Expedites Tuning AIG tuning is difficult without a probe grid at the catalyst exit The costs for installing a probe grid will be recovered in the long run: - No scaffolding or manlift required for testing ($, safety issues) - Reduced test times (no manual probe) - Reduced testing contractor costs - More data 5MW simple cycle SCR With a grid of 5 sample probes (5 x 1) 16

17 Sample Probe Grid Expedites Tuning 17

18 Sample Probe Grid Expedites Tuning 18

19 Sample Probe Grid Expedites Tuning 19

20 FERCo s Multipoint Instrumentation - Samples 48 points in 15 minutes - NOx and O2 2

21 AIG Design Influences Tuning No Adjustments: Some systems have no AIG adjustment valves bad idea! No flexibility to account for 1) duct velocity gradients, 2) duct NOx gradients, or 3) lance-to-lance ammonia flow gradients 1-D: Most systems have one-dimensional adjustability NH3 Vertical adjustability can handle only vertical gradients Multi-Zone: Ideal design: multiple zone adjustability Multiple zones allow treatment of localized gradients 21

22 AIG Tuning Procedure Basis (SCR operated with no local NH 3 slip) NH3 ( NO NOx ) ini in i out i NH x 3 slipi Procedure NH NO 3 x i 1 NO NO x x outi in i 1. Turn off NH 3, obtain NO x outlet profile 2. Turn NH 3 on to produce 5-7% DNO x, obtain NO x outlet profile 3. Use 1 and 2 to calculate local NH 3 / NO x ratios 4. Make contour plot of NH 3 / NO x distribution 5. Adjust AIG and repeat

23 North Wall (ft) AIG Tuning, 1 MW Gas Turbine SCR As-Found NH3/NOx Distribution Adjusted Riverside Springs Unit 4 SCR NH3/NOx NH3/NOx Distribution, Distribution Adjustment #1 Catalyst Inlet One-Dimensional, Vertical AIG NH Adjustments across the width not possible Duct Bottom (ft) RMS = 13%

24 West Wall (ft) West Wall (ft) AIG Tuning, 35MW Combined Cycle SCR As-Found Outlet NOx (ppm) NH 3 consumption reduced by 5% Optimized Outlet NOx (ppm) 6 6 One-Dimensional, Vertical AIG NH Duct Bottom (ft) Adjustments across the width not possible Duct Bottom (ft)

25 Duct Burners Impact AIG Tuning Duct Burners Off (Inlet NO x ppm) Duct Burners On (Inlet NO x ppm) AIG Tough to Tune 2 NH

26 South North South North AIG Tuning, 14 MW Gas Turbine SCR As-found (RMS = 14.7%) Optimized (RMS = 6.5%) Std. Dev=14.7% Std. Dev= 6.5% Vertical AIG, Two Horizontal Zones AIG No way to adjust the center Flow Into Page 26

27 Height (ft) Height (ft) AIG Tuning, 1 MW Gas Turbine SCR, 9 Turn As-Found Normalized NH 3 /NO x Distribution Tuned Normalized NH 3 /NO x Distribution A B C A B C Multiple zones, skewed based on flow modeling Width (ft) Width (ft) RMS = 32% Corners adjusted RMS = 13% 27

28 Measuring Catalyst Velocity Distribution 28

29 Catalyst Velocity Distribution Velocity uniformity typically in the specification Velocity uniformity established during the SCR Design (Cold Flow and/or CFD) Can it be measured at full scale? Difficult; large reactors, long probes, low velocities(~15 ft/sec) Pitot Probes, Thermal Anemometers Pitot Probe dps ~.2 in H2O Velocity profile re-aligns very close to the surface Velocity distribution can be inferred using NOx measurements 29

30 Local Outlet NOx, ppm raw Local Outlet NOx, ppm raw Local Outlet NO x VS. NH 3 Injection Rate NO x variations due primarily to NH 3 /NO x maldistribution Ammonia Injection Rate, lb/hr NO x variations due to velocity maldistribution pt-1 pt-2 pt-3 pt-4 pt-5 pt-6 pt-7 pt-8 pt-9 pt-1 pt-11 pt-12 pt-13 pt-14 pt-15 pt-16 pt-17 pt-18 pt-19 pt-2 dno x =1-e -K/Av Am m onia Injection Rat e, lb/hr 3

31 Identifying Flue Gas Bypass 31

32 Flue Gas Bypass Improper perimeter seals Improper seals between catalyst modules Difficult to diagnose just looking at overall performance A likely region of bypass is the roof 32

33 NOx, ppm Flue Gas Bypass Data should fall on a single line defined by the catalyst activity pt-1 pt-2 pt-3 pt-4 pt-5 pt-6 pt-7 pt-8 pt-9 pt-1 pt-11 pt-12 pt-13 pt-14 pt-15 pt-16 pt-17 pt-18 pt-19 pt NH3/NOx, scaled by ammonia injection Plot Local NH3/NOx rather than ammonia injection rate 33

34 NO-out, ppm NO-out, ppm Flue Gas Bypass, 5 MW Gas Turbine SCR As-Found Repaired NH3/NOx NH3/NOx 34

35 SCR Impact on CEMs Relative Accuracy 35

36 CEMs Relative Accuracy NH 3 /NO x gradients at the catalyst inlet create NO x gradients at the exit of the catalyst Close proximity of the stack CEMs to the catalyst exit on a gas turbine simple cycle SCR system may result in a stratified NO x profile at the CEMs It may be necessary to tune the AIG specifically to pass the Relative Accuracy Test 36

37 SE NW SE NW Relative Accuracy Example As Found (RA=-25%) Tuned (RA=-6.6%) 18 Test 1 SW 18 Test 9 SW NE NE 37

38 Catalyst Management 38

39 Catalyst Management What is Catalyst management? Keeping track of catalyst activity to ensure continued compliance How is Catalyst Management Done? Periodically determining the activity of the catalyst in the reactor Laboratory analysis (if a sample can be obtained) In situ analysis (later discussion topic) Utilize catalyst management software for planning Why is it Done? Forecast when catalyst additions or replacements are necessary Provide sufficient lead time to procure catalyst (6-9 months) 39

40 NH3 Slip, ppm K/Ko Catalyst Activity Degrades With Time NH3 Slip-Model K/K , 4, 6, 8, 1, 12, Operating Hours Monitoring ammonia slip is also critical 4

41 In Situ Catalyst Activity Measurements 41

42 In Situ Catalyst Activity Measurement* Traditional Lab Measuremett Typically one per year Currently no test protocol K Lab = -A Vdesign ln(1-δno x ) FERCo s CatalysTrak * in situ measurement No outage required K In-situ = -A Vactual ln(1-δno x ) full locally * Patented Process 42

43 CatalysTraK In Situ Test Modules 43

44 In Situ vs Laboratory Activity Measurements K/Kvendor Insitu Vendor 3rd Party 44

45 Summary Troubleshooting Obtain a curve of NH3 slip vs Outlet NOx Can help distinguish poor NH3 distribution vs bypass AIG Tuning Important activity Easy to do with a permanent probe grid installed downstream of the catalyst Flue Gas Bypass and Velocity Profiles Can be determined with NO x measurements CEMs Relative Accuracy SCR tuning may impact stack NO x distribution Catalyst Management/Catalyst Activity Requires knowledge of Activity vs Time CatalysTrak allows online determination of activity 45

46 Questions? 46