this configuration requires a costly hot-side ESP or a flue gas reheating system to maintain the optimum operating temperature.

In tail-end systems (also referred to as cold-side, low-dust), the SCR reactor is installed downstream of the FGD unit. It may be used mainly in wet-bottom boilers and also on retrofit installations with space limitations [30]; however, this configuration is typically more expensive than the highdust configuration due to flue gas reheating requirements. This configuration does have the advantage of longer catalyst life and the use of more active catalyst formulations to reduce overall catalyst cost.

Several issues need to be considered in the design and operation of SCR systems, including coal characteristics, catalyst and reagent selections, process conditions, ammonia injection, catalyst cleaning and regeneration, low-load operation, and process optimization [30]. Coals with high sulfur in combination with significant quantities of alkaline, alkaline earth, arsenic, or phosphorus in the ash can severely deactivate a catalyst and reduce its service life. In addition, the SO3 can react with residual ammonia, resulting in ammonium sulfate deposition in the air preheater and loss of performance.

The two leading geometries of SCR catalysts are honeycomb and plate [47]. The honeycomb form usually is an extruded ceramic, with the catalyst either incorporated throughout the structure (homogenous) or coated on the substrate. In the plate geometry, the support material is generally coated with catalyst. The catalyst commonly consists of a vanadium pentoxide active material on a titanium dioxide substrate.

For optimum SCR performance, the reagent must be well mixed with the flue gas and in direct proportion to the amount of NOz reaching the catalyst. Anhydrous ammonia has been commonly used as the reagent, accounting for over 90% of current-world SCR applications [30]. It dominates planned installations in the United States, although numerous aqueous systems will be installed. Urea-based processes are being developed to address utilizing anhydrous ammonia, which is a hazardous and toxic chemical. When urea CO(NH2)2 is used, it produces ammonia, which is the active reducing agent, by the following reactions:

During the operation of the SCR, the catalyst is deactivated by fly ash plugging, catalyst poisoning, and/or the formation of binding layers. The most common method of catalyst cleaning has been the installation of steam sootblowers, although acoustic cleaners have been successfully tested. Once the catalyst has been severely deactivated, it is conventional practice to add additional catalyst or replace it; however, several regeneration techniques have evolved over the last few years, providing extended service life for catalysts [30]. Low-load boiler operation can be problematic with SCR operation, specifically with high-sulfur coals. There is a minimum temperature below which the SCR should not be operated; therefore, system modifications, such as economizer bypass, to raise the SCR temperature during low-load operation may be required [30].

Selective Non-Catalytic Reduction Selective non-catalytic reduction is a proven, commercially-available technology that has been applied since 1974; over 300 systems are installed worldwide on various combustion sources, including utility applications [30]. The SNCR process involves injecting nitrogen-containing chemicals into the upper furnace or convec-tive pass of a boiler within a specific temperature window without the use of an expensive catalyst. Various chemicals can be used that selectively react with NO in the presence of oxygen to form molecular nitrogen and water, but the two most common are ammonia and urea. Other chemicals that have been tested in research include amines, amides, amine salts, and cya-nuric acid. In recent years, urea-based reagents such as dry urea, molten urea, or urea solution have been increasingly used, replacing ammonia at many

plants because anhydrous ammonia is the most toxic and requires strict transportation, storage, and handling procedures [30]. The main reactions when using ammonia or urea are, respectively:

A critical issue is finding an injection location with the proper temperature window for all operating conditions and boiler loads. The chemicals then need to be adequately mixed with the flue gases to ensure maximum NOZ reduction without producing too much ammonia. Ammonia slip from an SNCR can affect downstream equipment by forming ammonium sulfates.

The temperature window varies for most of the reducing chemicals used but generally is between 1650 and 2100°F. Ammonia can be formed below the temperature window, and the reducing chemicals can actually form more NOZ above the temperature window. Ammonia has a lower operating temperature than urea: 1560 to 1920°F vs. 1830 to 2100°F, respectively. Enhancers such as hydrogen, carbon monoxide, hydrogen peroxide (H2O2), ethane (C2H6), light alkanes, and alcohols have been used in combination with urea to reduce the temperature window [51]. Several processes use proprietary additives with urea in order to reduce NOZ emissions [52].

The efficiency of reagent utilization is significantly less with SNCR than with SCR. In commercial SNCR systems, the utilization is typically between 20 and 60%; consequently, usually three to four times as much reagent is required with SNCR to achieve NOZ reductions similar to those of SCR. SNCR processes typically achieve 20 to 50% NOZ reduction with stoichiometric ratios of 1.0 to 2.0.

The major operational impacts of SNCR include air preheater fouling, ash contamination, N2O emissions, and minor increases in heat rate. A major plant impact of SNCR is on the air preheater, where residual ammonia reacts with the SO3 in the flue gas to form ammonium sulfate and bisulfate (see Reactions (6-2), (6-75), (6-76)), causing plugging and downstream corrosion. High levels of ammonia slip can contaminate the fly ash and reduce its sale or disposal. Significant quantities of N2O can be formed when the reagent is injected into areas of the boiler that are below the SNCR optimum operating temperature range. Urea injection tends to produce a higher level of N2O compared to ammonia. The unit heat rate is increased slightly due to the latent heat losses from vaporization of injected liquids and/or increased power requirements for high-energy injection systems. The overall efficiency and power losses normally range from 0.3 to 0.8% [30].

Hybrid SNCR/SCR Selective catalytic reduction generally represents a relatively high capital requirement, whereas selective non-catalytic reduction has a high reagent cost. A hybrid SNCR/SCR system balances these costs over the life cycle for a specific NOZ reduction level, provides improvements

4NO + 4NH3 + O2 4N2 + 6H2O 4NO + 2CO (NH2)2 + O2 4N2 + 2CO2 + 4H2O

in reagent utilization, and increases overall NOZ reduction [30]. However, experience with these hybrid systems is limited, as full-scale power plant operation to date has only been in demonstrations. They are discussed here because they have demonstrated NOZ reductions as high as 60 to 70%.

In a hybrid SNCR/SCR system, the SNCR operates at lower temperatures than stand-alone SNCRs, resulting in greater NOx reduction but also higher ammonia slip. The residual ammonia feeds a smaller-sized SCR reactor, which removes the ammonia slip and decreases NOx emissions further. The SCR component may achieve only 10 to 30% NOZ reduction, with reagent utilization being as high as 60 to 80% [30]. Hybrid SNCR/SCR systems can be installed in various configurations, including [30]:

• SNCR with conventional reactor-housed SCR;

• SNCR with in-duct SCR, which uses catalysts in existing or expanded flue gas ductwork;

• SNCR with catalyzed air preheater, where catalytically active heat transfer elements are used;

• SNCR with a combination of in-duct SCR and catalyzed air heater.

Rich Reagent Injection Cyclone burners, with their turbulent and high-temperature environment, are conducive for NOZ production. Methods that cost less than installing SCRs to reduce NOZ production in cyclone-fired boilers have been tested, such as CWSF or biomass cofiring, while others are under development. One such process currently under development is the rich reagent injection (RRI) process, which involves injection of amine reagents in the fuel-rich zone above the main combustion zone at temperatures of 2370 to 3100°F. NOZ in the flue gas is converted to molecular nitrogen, and reductions of 30% have been achieved. The capital costs for an RRI system are consistent with those of SNCR; however, the operating costs are expected to be 2 to 3 times that of SNCR due to increased reagent usage.

NOx Control in Fluidized-Bed Combustion

The fluidized-bed combustion (FBC) process described in Chapter 5 (Technologies for Coal Utilization) inherently produces lower NOx emissions due to its lower operating temperature (i.e., bed temperature of ~1600°F). Also, the bed is a reducing region where available oxygen is consumed by carbon, thereby reducing ionization of nitrogen. Additional combustion modifications or flue gas treatment for NOx control, discussed previously in this chapter, can also be employed. Techniques currently used for FBC include reducing the peak temperature by flue gas recirculation (FGR), natural gas reburning (NGR), overfire air (OFA), fuel reburning, low excess air (LEA), and reduced air preheat [23]. Post-combustion control is also used, including SCR and SNCR, which achieve 35 to 90% NOZ reductions. Also, low nitrogen fuel can be used (e.g., sawdust), thereby reducing the amount of fuel nitrogen available. Injecting sorbents into the combustion chamber or in the ducts can reduce NOZ by 60 to 90% [23].

NOx Control in Stoker-Fired Boilers

Control of NOZ in stokers (specifically, traveling-grate and spreader stokers) include abatement methods to reduce the peak temperature, to reduce the residence time at peak temperature, and to chemically reduce the NOx , in addition to using low-nitrogen fuels and injecting a sorbent [23]. In traveling-grate stokers, the peak temperature can be reduced by FGR, NGR, combustion optimization, OFA, LEA, water or steam injection, and reduced air preheat, thereby achieving 35 to 50% NOZ reduction. Air or fuel staging, which reduces the residence time at peak temperature, can achieve 50 to 70% NOx reduction, while using SCR, SNCR, or fuel reburning technologies can achieve 55 to 80% NOZ reduction. Sorbent injection, which can achieve 60 to 90% NOx reduction, and the use of fuels with low nitrogen content are technologies also employed. NOx technologies used for spreader stokers are similar to traveling-grate stokers but achieve slightly different results. FGR, natural gas reburning, low-NOx burners, combustion optimization, OFA, LEA, water or steam injection, and reduced air preheat temperature are control options to reduce peak temperatures that can achieve 50 to 65% NOx reductions. Air or fuel staging or steam injection, which reduces the residence time at peak temperature, can achieve 50 to 65% NOx reductions, while using SCR, SNCR, or fuel reburning technologies achieves 35 to 80% NOZ reductions. Additional NOZ reduction technologies include sorbent injection, which can achieve 60 to 90% reductions, and using lower nitrogen fuels.

Economics of NOx Reduction/Removal

The costs for NOx reduction/removal techniques are site and performance specific, thus making it difficult to compare generalized system costs. These techniques depend on several factors, including degree of retrofit difficulty, unit size, uncontrolled NOZ levels, and required NOZ reduction [30]. This section summarizes costs for the various systems using published data.

Low-NOx Burners Wu [30] reported that the capital costs for a low-NOz burner are in the range of $650 to $8300/MM Btu. The operating costs can range from $340 to $1500/MM Btu. The levelized costs can vary from $240 to $4300/short ton of NOx removed, with the average cost being closer to the lower end of the range [53].

Furnace Air Staging The costs for furnace air staging are similar to those for low-NOz burners [30]. The capital costs range from ~$8 to $23/kW, and the levelized costs range from $110 to $210/short ton of NOx removed. If furnace air staging is combined with low-NOx burners, the capital costs can increase to $15 to $30/kW, while the levelized costs remain relatively unchanged. Retrofits of furnace air staging in tangentially-fired boilers are generally more expensive than those in wall-fired boilers: $11 to $23/kW and $5 to $11/kW, respectively.

Flue Gas Recirculation The capital costs for conventional flue gas recirculation is similar to that for low-NOx burners and overfire air: $8 to $35/kW [53]; however, capital costs of induced FGR, a design derivative of conventional forced flue gas desulfurization, have been reduced to $1 to $3/kW.

Fuel Staging (Reburn) The capital costs for reburn technology depend on the size of the unit, ease of retrofit, control system upgrade requirements, and, for natural gas reburn, availability of natural gas at the plant [30]. The retrofit costs are typically about $15 to $20/kW for natural gas, coal, or oil reburn, excluding the cost of any natural gas pipeline. The operating costs for a reburn retrofit are mainly due to the differential cost of the reburn fuel over the main fuel. For coal reburn, this cost is zero, but reburn fuels such as natural gas or oil are usually more expensive than the main fuel. This differential, however, can be offset by reductions in SO2 emissions, ash remediation and disposal, and pulverizer power. The levelized cost for reburn is -$110 to $210/short ton of NO x removed [16].

Cofiring Cofiring of CWSF is not commercially used at this time. Biomass cofiring, on the other hand, is currently being demonstrated at several plants and commercial operations are being performed at seven utilities. The capital costs for biomass cofiring range from $175 to $250/kW [43].

Process Optimization The total turnkey installation cost for an advanced combustion control system ranges from $150,000 to $500,000 [30]. It is possible to achieve moderate cost reductions on a per-unit basis for similar units at the same power plant site. The size of the unit typically has little impact on the cost of a system.

Selective Catalytic Reduction (SCR) The capital costs for an SCR system depend on the level of NOx removal and other site-specific conditions, such as inlet NOx concentration, unit size, and ease of retrofit and range from $80 to $160/kW [54]. The capital costs of an SCR system include [30]:

• Catalyst and reactor system;

• Flow control skid and valving system;

• Ammonia injection grid;

• Ammonia storage;

• Ducts, expansion joints, and dampers;

• Fan upgrades/booster fans;

• Air preheater changes;

• Foundations, structural steel, and electricals;

• Installation.

The operating costs can vary from $1500 to $5800/MM Btu, and the levelized cost can range from $1800 to $10,900/MM Btu [30]. The operating costs include [30]:

• Pressure drop changes;

• Unburned carbon change;

• Catalyst replacement;

• Vaporization/injection energy requirements;

• Other auxiliary power usage.

Selective Non-Catalytic Reduction Selective non-catalytic reduction is less capital intensive than SCR. The cost of an SNCR retrofit is $10 to $20/kW, whereas incorporating SNCR into a new boiler typically costs $5 to $10/kW [30]. The difference is due to the costs associated with modifying the existing boiler to install the reagent injection ports. The operating costs associated with the reagent, auxiliary power, and potential adverse plant impacts are of the order of $1 to $2 mills/kWh. The levelized costs average ~$1000/short ton of NOZ removed. A new, single-level approach to SNCR—SNCR trim— offers 20 to 30% NOZ reduction at about half the cost of conventional SNCR and is being tested by the EPRI [55]. SCNR trim has low operating costs, equivalent to only about $850/short ton of NOZ removed.

Other Flue Gas Treatment Processes Limited data are available for hybrid SNCR/SCR systems as they are still in the demonstration phase. A levelized cost estimate for a 500 MW boiler with 50% NOZ reduction is ~$5800/short ton of NOZ removed [30]. Similarly, cost data on the rich reagent injection process, which is under development, are not available.

Hybrid Flue Gas Treatment and Combustion Modifications A combination of flue gas treatment with combustion modification is increasingly being used. This technology provides higher overall NOZ reductions and can be more cost effective than stand-alone technology for the same level of NOZ control [30]. The costs of SCR can be reduced when it is used in combination with combustion modifications such as low-NOz burners and overfire air [30]. Capital costs are lowered because combustion modifications lower the inlet NOZ concentration, which reduces the catalyst volume, support systems, and installation cost of SCR. In addition, operating costs are lower due to reductions in catalyst replacement and reagent consumption. SNCR can be combined with low-NOz burners or gas reburn. SNCR and gas reburn have comparable economics at the same level of NOZ reduction; however, combining the two technologies considerably lowers costs while achieving a slightly higher NOZ reduction. An example of annual costs, reported by Wu [30], are ~$1140, $1120, and $730 per short ton NOZ removed, respectively, for urea SNCR, gas reburn, and urea SNCR/gas reburn.

0 0

Post a comment