Selective Catalytic Reduction (SCR) for the removal of NOx (NO and NO2) This primer includes:

• SCR and Catalyst Basics • SCR Design Considerations • Catalyst Management Trade names and companies mentioned are for illustration and clarification purposes but not for endorsement. 10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

1

SCR and Catalyst Basics

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

2

Catalysis: a modification and especially increase in the rate of a chemical reaction induced by material unchanged chemically at the end of the reaction.

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

3

CHEMISTRY Basic chemical reactions: 4NO + 4NH3 + O2  4N2 + 6H2O 2NO2 + 4NH3 + O2  3N2 + 6H2O NO + NO2 + 2NH3  2N2 + 3H2O

The key reductant is ammonia, the ammonia molecules need to be thoroughly mixed with NOx in the flue gas, this is essential for high NOx removal efficiency for all SCR systems. Thus, the mixing system and ammonia injection grid design is closely related to removal efficiency. In general cold flow modeling and Computational Fluid Dynamics (CFD) modeling are conducted to ensure that the system has the least amount of maldistribution and ash dropout for a range of conditions. Due to maldistribution in NH3 and NOx, a minute amount of NH3 in single digit ppm will exit the catalyst layers, this is usually referred to as ammonia slip or slip. 10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

4

Ammonia (NH3) mixed with NOX Typical SCR Process and Schematic

Ammonia (NH3)

NH3 reacts with NO on an active site

Ammonia Injection Grid (AIG)

Economizer gas outlet with NOX

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

SCR catalyst Exit consists of N2 + H2O. Minor unreacted ammonia /slip 5

SCR Catalysts Most commercial SCR catalysts are referred to as titania-vanadia base catalyst (since the 1970’s). Basic material is similar to ceramic; the main component is titanium oxide (TiO2) with minor components such as tungsten. As for the SCR reaction, the active sites are vanadium oxides (V2O5, V2O3). There are many physical forms: homogeneous (extruded honeycombs), and heterogeneous (plates, corrugated). In addition there are different pitches and/or cell openings for different applications (cleaner low dust flue gas vs. high dust solid fuel boilers.) The catalysts are packaged at different depths in the direction of gas flow. Some applications require multiple layers of catalyst to meet the removal efficiency and expected operating life. Catalysts are sold by volumes in cubic meters (m3). However, the effective specific geometric surface area is used for catalyst design, expressed in m2 per m3. The total surface area, Acat in m2 is derived from specific area (m2 per m3) times total catalyst volume (m3). This Acat is also used to derive the Area Velocity, AV (m/hr), a key design parameter and for activity calculation. AV is defined as Flue Gas Flow rate in standard condition divided by Acat. Note: Internal surface area (pore volume) and pore distribution varies for different manufacturers, the standard measurement is the BET surface area, m2/gm of material. [Brunauer–Emmett–Teller theory/model]. BET surface area is used for quality control as well as for determining the exposed catalyst sample condition. 10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

6

SCR Catalysts, contd.

In addition to physical differences, there are ranges in activity which is reflected by the bulk concentration of vanadium in the formulation. High V concentrations will have higher activity and NOX removal efficiency for a given catalyst volume, however, due to the concomitant oxidation reaction, a small amount SO2 is converted to SO3. In general the ammonia slip will react with SO3 and form a sticky salt, ammonia bisulfate (ABS) due to its high melting point. This is especially troublesome for downstream heat exchange surfaces (lower bulk flue gas/metal temperature) such as air heater baskets and finned tubes.

Note: ABS reaction – NH3 + SO3 + H2O  (NH4)HSO4 (ABS)

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

; usually under [NH3] 9.3 mm

12 to 16

6.5 mm

NA

10.3 mm

16 to 20

6.5 mm

NA

12 mm

gr/dscf

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

11

SCR Catalyst Blocks and Modules

• Catalyst elements arranged and packed in steel frames. o Plate – 2 levels of 8 element boxes o Honeycomb – 72 monoliths o Corrugated – 2 to 3 levels of 8 element boxes.

• Standardized cross-section module o Possible to interchange corrugated and plate element boxes in most modules.

• Possible to interchange catalyst types within reactor • Module height varies with the catalyst monolith heights, different for different catalyst suppliers. • Most applications have a top grid/mesh designed onto the modules.

SCR Design Considerations

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

13

SCR Design Parameters

Both new and retrofit applications have the same basic design criteria and considerations. However, some retrofits may have more stringent and available height, width and depth as well as access to the reactor. Projects are structured in many fashions, e.g., SCR system suppliers, EPC contractor, A/E firm as sole owner engineer, A/E firm as partner, catalyst supplier and manufacturer as process design engineer and catalyst supplier, catalyst manfacturer as supplier only with pass through guarantees by A/E firm. Most SCR guarantees are for a specific NOX removal efficiency (ETA, η) and/or Stack NOX emissions for a certain cumulative operating hours. In general this is called ‘life’, even though the catalyst proper is still capable of reducing NOX but not at a high level at the design conditions. Note:

η

= [ (NOX)in – (NOX)out ]/ (NOX)in

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

14

SCR Design Considerations

• Performance Requirements (and Guarantees) • • • • •

NOx reduction (80 – 95%) (or Stack NOX limit, ppmvdc) and the associated operating life (8,000 to 24,000 hours). NH3 slip allowed, 2 to 5 ppm, typically lower values for high sulfur fuel. SO2 oxidation allowed, 0.1 to 1.0% per initial catalyst charge. Pressure drop limit, usually 1 to 1.5” wc per layer.

Note on mercury emission (MATS): SCR catalysts do oxidize elemental mercury to the oxidized form, the latter is readily removed in the wet scrubber system. Certain system suppliers do guarantee mercury oxidation.

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

15

SCR Design Considerations

• Flue Gas Operating Conditions • Flue gas temperature and flue gas flow rate, and operating range (minimum operating temperature for low loads, some employ economizer bypass.) • NOx inlet concentration and flue gas composition • Fuel characteristics (affects catalyst volume) • Fly ash concentration and characteristics (issues with large particle ash/popcorn ash) • Other potential poisons in fuel and from the combustion process (affects catalyst volume due to deactivation.)

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

16

SCR Design Considerations

• SCR Reactor • Initial catalyst charge (e.g., 2 + one spare layer, 3 + 1, etc.) • Reactor size – layers, catalyst depth, modules per layer • Optimize effective cross-section to mitigate erosion potential for high erosional ash, most applications are in the 15 to 17 actual feet per second range. • Plant configuration – high or low dust, AIG only, AIG/Mixers • Governing - flue gas ammonia to NOX distribution entering first layer, a reasonably low maldistributionnote (5%) is required for high removal efficiency. • Sealing for intra- and inter-module contact surface as well as side-walls and grating and floor. Note: The actual local ammonia to NOx mole ratio (stoichiometric ratio, SR) at the catalyst inlet is a key design flue gas parameter for all SCR systems. It is usually normalized to 1.0, thus, NSR, and is expressed in percent, in RMS or Std Deviation divided by the mean value or Co-variance.

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

17

SCR Design Considerations • Governing - flue gas ammonia to NOX distribution entering first layer and its effect on slip and removal efficiency. Effect of maldistribution on removal and slip 120

12

higher maldistrbution will shift removal to a lower value for the same stoichiometry (and catalyst volume.)

10

80

8

60

6

40

4

higher maldistrbution will shift the ammonia slip to a higher value for the same stoichiometry (and Catalyst volume.)

20

ammonia slip, ppm

removal

100

2

0

0 0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

stoichiometry

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

18

Catalyst Management

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

19

SCR Catalyst Management All applications are aimed at providing the design performance with the installed catalyst layer(s) for a period beyond the guaranteed/expected operating life. Should there be a need to replace a layer, the plan should yield the best option and fit in with the outage schedule. The users need to have a comprehensive catalyst management plan (CCMP) in addition to a catalyst replacement plan from the suppliers. Within the industry, the latter is also called ‘Catalyst Management Plan (CMP)’.

A CCMP has many essential components: • • • • • •

Actively study SCR performance trending and evaluation of key SCR indicators. Conduct periodic full load SCR performance test for removal and ammonia slip under design (guaranteed) SCR conditions. Perform SCR reactor outage inspection with documentation of the system: reactor flow devices, AIG, and catalyst (appearance and deposition, seals/bypasses, modules and sidewalls) mapping. Follow the SCR shutdown procedures and SCR catalyst outage protection. Perform periodic sample extraction analysis (either built-in sample log or coring of sample): 1. Catalyst Activity Analysis (deactivation leads to low activity, K) 2. Physical and Chemical Analyses (root causes for deactivation) Review the supplier’s ‘CMP’ with data from the first three items.

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

20

SCR Catalyst Management Example - supplier’s ‘Catalyst Management Plan’ EXAMPLE -- Expected Replacement strategy 'Catalyst Management Plan'

1.20 add spare 1 layer

1.00

Relative Actvity

14 Relative Activity NH3-slip

add spare 2layer

replace layer 1 replace layer 1

12 10

0.80

8

0.60

6

0.40

4

0.20

2

0.00 0

8,760

17,520

26,280

35,040

NH3-slip, ppm

1.40

0 43,800

Hours 10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

21

SCR Catalyst Management

• SCR performance trending and evaluation (e.g., CEMS data, operating NOx removal, ammonia consumption, ammonia slip (if equipped), pressure differential.) • Most systems should not have sudden changes in performance at the same load, deactivation should be gradual. • Abrupt changes should be immediately investigated (e.g, low removal efficiency, ammonia consumption, ammonia slip, ammonia in ash, increase in pressure drop.) • Full SCR Reactor performance test for removal and ammonia slip under design (guaranteed) SCR conditions. The results should be compared to the design and guaranteed values, certain corrections from test to design may apply by using supplier’s performance correction curves. • SCR performance should exceed that of design (guaranteed) parameters during the initial/acceptance performance test. • SCR performance should meet the design (guaranteed) parameters during the ‘end of life’ performance test. • Non-performance should be discussed with the supplier. 10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

22

SCR Catalyst Management Catalyst Activity (K note) Analysis • Full bench is recommended due to the larger sample and less wall effect. • Micro-reactor is also used, however, extrapolation and certain correlations entered into deriving the activity. • There are only few independent test labs (e.g., SRI, FERCo, EON.) • Most manufacturers and regenerators have their own test lab, both for quality control and aftermarket analysis. • In general the deactivation rate and actual remaining activity is expressed in K/Ko, where K is the initial known activity from the manufacturer. Note: K is defined as -AV * ln (1 - η ) Relative Activity, K = e-a x t,

where: a is the deactivation rate, t in 10,000 hours

Only two ‘standards’ (Test Protocols) are widely accepted and followed, VGB ad EPRI. (There are no SCR test standards/protocol such as ASME, ASTM,…) 10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

23

SCR Catalyst Management Catalyst Activity (K note) Analysis, contd.

Application – High Dust

Time, hr

Relative Activity, Kt/Ko

PRB firing

16,000

0.65 – 0.70

Lignite firing

16,000

0.50 – 0.55

Bituminous with 10 – 20% bio-fuels co-firing

24,000

0.70 – 0.75

Bio-fuels firing

10,000

0.30 – 0.60

24,000

0.85 – 0.95

Application - Low Dust

Note: Compare full size performance and test sample activity to this table and the supplier’s plan/curve. 10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

24

SCR Catalyst Management Example - supplier’s ‘Catalyst Management Plan’, typical strategy is to provide Design NOX removal as catalyst layers deactivate, note the change in operating ammonia slip through the management cycle. EXAMPLE -- Expected Replacement strategy 'Catalyst Management Plan'

1.20 add spare 1 layer

1.00

Relative Actvity

14

Relative Activity NH3-slip

replace layer 1 replace layer 1 add spare 2layer

12 10

0.80

8

0.60

6

0.40

4

0.20

2

0.00 0

8,760

17,520

26,280

35,040

NH3-slip, ppm

1.40

0 43,800

Hours 10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

25

SCR Catalyst Management Example – activity, K/Ko trend and the value of sample test results. EXAMPLE - Expected Relative Activity (deactivation) and Actual Test Sample results 1.10

1.00

Relative Activity, low deactivation application

0.90

test sample RA 0.80 HIGH DEACTIVATION samples

High Deactivation Application

0.70

Poly. (Relative Activity, low deactivation application) 0.60

0.50 -

5,000

10,000

15,000

20,000

25,000

time, hours

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

26

SCR Catalyst Management Catalyst Sample Physical Analyses • Physical analysis usually includes a determination of internal surface area (BET), in m2/gm. • The results will support whether the loss in activation is due to a decrease in active surface and/or other mechanisms impeding the reactants from reaching the active sites. Catalyst Sample Chemical Analyses • Bulk elemental analysis as well as acid soluble elemental analysis is useful to determine foreign elements in the exposed catalyst. • ICP is one technique to determine a panel of interested elements (for example, known poisons - alkalis, phosphorus, arsenic (coal units), chromium.) • Scanning EM is also used to scan the surface for poisons and blinding material (e.g., calcium sulfate), this diagnostic technique is also applied to cut sections so as to understand the profile of different poisons and its deposition gradient inside the catalyst.

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

27

SCR Catalyst Management

Actions: • Non-performance could be in the catalyst proper with low activity and/or the process condition, one common problem is the localized maldistribution of ammonia and NOX. In short, there is either not enough ammonia molecules to react with NOX molecules and/or too many ammonia molecules present for too few NOX molecules. • AIG tuning involves the testing and adjusting the injection grid ammonia flows to achieve design maldistribution. • When the total catalyst volume (and its reactor potential) is below the required performance, actions such as root cause analysis, tuning, replacement and/or regeneration options should be evaluated as soon as possible; third party consultants and catalyst suppliers should be engaged to ensure the best course has been chosen. • If the performance is marginal, certain short-term measures such as de-rating and inlet NOX tuning, and AIG tuning should be considered in order to extend the run-time for the next planned or unplanned catalyst replacement.

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

28

SCR Catalyst Management – Action: •

AIG tuning considerations • Need to collect basic SCR design information including the design maldistribution, removal and flue gas conditions, AIG design and CFD/flow model results. • Ensure that SCR outlet test provisions are available, best to have some type of installed grid, if not, then test probes should be assembled so as to provide an adequate grid. (Most end-users do not have inlet sample grid. Although turning off the ammonia feed is the best way to obtain the inlet NOX profile and value, this is not done in practice since it will violate the permit in most circumstances. Some form of inlet NOX is usually available, if not, a single inlet probe to monitor the process condition is a good alternative.) Check: individual flow control devices are operable; zone or point flow indications are operating and indicating (e.g., orifices, Magnehelic.) • Test unit should be at design load, removal efficiency should be slightly below guaranteed, and SCR operates at stable removal rate. • Test should be conducted within a reasonable time frame so as to lower the temporal changes, lower test duration will improve the accuracy of the spatial distribution and the final analysis for the ammonia to NOX distribution. Then the test protocol will require multiple analyzers to sample multiple individual points simultaneously. • Collect and convert raw data so as to derive the maldistribution value. AIG tuning is an iterative process involving adjusting ammonia flows and evaluating the results. (Examples below.)

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

29

SCR Catalyst Management – Action: example AIG tuning

•

AIG tuning considerations • Example 1. As-found Distribution. • Without actual inlet NOX distribution, one can assume uniform distribution, as reflected in the second section with 40 ppm for all points. Example: with an outlet NOX of 10 ppm and inlet NOX of 40 ppm assuming no slip (at less than max removal), the inlet ammonia equals to (4010) 30 ppm. Local stoichiometry = 30/40 or 0.75.

NOX (ppm) at the Outlet test grid Probe 1

2

3

4

5

6

7

8

Point 1

10

2

6

6

10

3

2

9

Point 2

7

2

6

5

8

4

1

6

Point 3

9

2

6

5

10

7

1

9

Point 4

7

1

5

7

9

5

2

9

LOCAL "inlet NH3 at individual point, ASSUMING no ammonia slip condition

30

38

34

34

30

37

38

31

33

38

34

35

32

36

39

34

31

38

34

35

30

33

39

31

33

39

35

33

31

35

38

31

mathematically the local stoichiometry ratio (SR) WILL THEN BE

0.75

0.95

0.85

0.85

0.75

0.925

0.95

0.775

0.825

0.95

0.85

0.875

0.8

0.9

0.975

0.85

0.775

0.95

0.85

0.875

0.75

0.825

0.975

0.775

0.825

0.975

0.875

0.825

0.775

0.875

0.95

0.775

with an average Stoichiometry of

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

40 ppm

'most PROBABLE' INLET NOX AVG.=

0.858594

30

mathematically the Normalized SR (NSR) distribution WILL THEN BE

SCR Catalyst Management – Action: • AIG tuning considerations Point 1 • Example 1, contd. Point 2 • Local SR is then normalized as Point 3 shown in the table. Example: Point 4 0.75/0.8586 = 0.874 • Mapping shows high ammonia flows to Lances 2 and 7, these should be decreased. Should also increase flow to Lance 5. • After changing flows, wait for 1 hour before testing for outlet profile, again the process is iterative and the same calculation applies after each adjustment. ‘Final’ result is shown in Example 2.

Probe 1

2

3

4

5

6

7

8

0.874

1.106

0.990

0.990

0.874

1.077

1.106

0.903

0.961

1.106

0.990

1.019

0.932

1.048

1.136

0.990

0.903

1.106

0.990

1.019

0.874

0.961

1.136

0.903

0.961

1.136

1.019

0.961

0.903

1.019

1.106

0.903

with an average Stoichiometry of

1

sample standard deviation of

0.084

OR A MALDISTRIBUTION OF

8.4%

example - NSR distribution Point 1

Point 2

Point 3

Point 4 1

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

0.850-0.900

2

3 0.900-0.950

4 0.950-1.000

5

6 1.000-1.050

7 1.050-1.100

8 1.100-1.150

31

SCR Catalyst Management – Action:

example AIG tuning - 2 NOX (ppm) at the Outlet test grid

• AIG tuning considerations • Example 2: as-left, successful tuning test results due to the available lances for the targeted zones.

Probe 1

2

3

4

5

6

7

8

Point 1

7

4

6

6

6

4

5

7

Point 2

5

7

5

5

7

4

7

6

Point 3

5

6

6

5

7

7

7

7

Point 4

6

4

5

7

6

7

5

6

'most PROBABLE' INLET NOX AVG.=

40 ppm

LOCAL "inlet NH3 at individual point, ASSUMING no ammonia slip condition 33

36

34

34

34

36

35

33

35

33

35

35

33

36

33

34

35

34

34

35

33

33

33

33

34

36

35

33

34

33

35

34

mathematically the local stoichiometry ratio (SR) WILL THEN BE 0.825

0.9

0.85

0.85

0.85

0.9

0.875

0.825

0.875

0.825

0.875

0.875

0.825

0.9

0.825

0.85

0.875

0.85

0.85

0.875

0.825

0.825

0.825

0.825

0.85

0.9

0.875

0.825

0.85

0.825

0.875

0.85

with an average Stoichiometry of

10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

0.853906

32

mathematically the Normalized SR (NSR) distribution WILL THEN BE Probe

SCR Catalyst Management – Action: • AIG tuning considerations • Example 2 after tuning, contd. • This tuning iteration has lowered the maldistribution from 8.4% to 3% as calculated below. The 3% is a very low value, usually 5% is used for a well designed AIG and mixing system. The ‘final’ mapping results should be documented. • AIG valve positions should be recorded and ‘locked’ in position. • The end result of AIG tuning will be higher removal for the same ammonia slip value, also this will effectively ‘lengthen’ the ‘life’ of the catalyst since it can now meet higher removal efficiency with the same process conditions. 10 Commerce Drive Pelham, AL 35124 Phone(205)453-0236 Facsmile(205)453-0239 www.innovativecombustion.com

1

2

3

4

5

6

7

8

Point 1

0.966

1.054

0.995

0.995

0.995

1.054

1.025

0.966

Point 2

1.025

0.966

1.025

1.025

0.966

1.054

0.966

0.995

Point 3

1.025

0.995

0.995

1.025

0.966

0.966

0.966

0.966

Point 4

0.995

1.054

1.025

0.966

0.995

0.966

1.025

0.995

with an average Stoichiometry of

1

sample standard deviation of

0.030

OR A MALDISTRIBUTION OF

3.0%

example - NSR distribution Point 1

Point 2

Point 3

1 0.850-0.900

2

3 0.900-0.950

4 0.950-1.000

5

6 1.000-1.050

7 1.050-1.100

8

Point 4

1.100-1.150

33