CONTROL OF NO% EMISSIONS FROM U.S. COAL-FIRED ELECTRIC UTILITY BOILERS R.K. Srivastava and R.E. Hall U.S. Environmental Protection Agency Office of Research and Development National Risk Management Research Laboratory Air Pollution Prevention and Control Division Research Triangle Park, NC 27711 Prepared for: All-Russian Thermal Engineering Institute (VTI) 80th Anniversary Conference Moscow, Russia, October 9-10, 2001 ABSTRACT Recently, several regulations have been promulgated in the U.S. requiring reductions in emissions of nitrogen oxides (NOx) from electric utility boilers. To comply with these regulations, state-of-the-art NOx control technologies have been applied to a large number of coal-fired U.S. utility boilers. This paper reviews these technologies and their applications. In general, NOx control technologies are categorized as being either primary control technologies or secondary control technologies. Primary control technologies reduce the formation of NOx in the primary combustion zone. In contrast, secondary control technologies destroy the NOx present in the flue gas from the primary combustion zone. Primary control technologies being used in the U.S. are low NOx burner (LNB) and overfire air (OFA). Data reflect that these technologies have been used on 177 boilers and have resulted in NOx reductions between 33 and 48 percent, on average, from 1990 emissions levels. The secondary NOx control technologies in use on U.S. coal-fired utility boilers include reburning, selective non-catalytic reduction (SNCR), and selective catalytic reduction (SCR). More than 100 boilers either have used, or will use, these technologies to achieve the desired NOx reductions. The NOx reductions achieved, or projected, at these applications range from 20 to more than 80 percent. In the last 3 years, SCR has been chosen as the preferred secondary technology at numerous U.S. coal-fired utility boilers. Current data indicate that 79 boilers either use, or will use, SCR for NOx control. 1 ------- INTRODUCTION Emissions of nitrogen oxides (NOx) are associated with a variety of environmental concerns including increasing ground level ozone, formation of acid rain, acidification of aquatic systems, forest damage, degradation of visibility, and formation of fine particles in the atmosphere.1 Such concerns have resulted in a need to reduce these emissions in the United States (U.S.) and elsewhere. In order to implement controls efficiently, it is important to determine which sources are significant emitters of NOx. Shown in Figure 1 is the contribution to NOx emissions in 1997 from each of the applicable source categories in the U.S.2 It is evident from these data that stationary combustion sources (electric utility, industrial, and other combustion sources) accounted for a significant portion, about 45 percent, of these emissions. Moreover electric utilities accounted for about 26 percent of NOx emissions and comprised the largest emitting source category within stationary sources. Based on these data, reduction of NOx emissions from stationary sources, particularly electric utility sources, needs to be considered in efforts undertaken to address the environmental concerns associated with NOx. Recently, a number of regulatory actions have been taken in the U.S., focused on reducing NOx emissions from stationary combustion sources. These actions include the Acid Rain NOx regulations3'4, the Ozone Transport Commission's NOx Budget Program5, revision of the New Source Performance Standards (NSPS) for NOx emissions from utility sources6, and the Ozone Transport rulemakings7. Control technology applications necessarily play a key role in the formulation and implementation of air pollution reduction strategies. The current focus on reduction of NOx from stationary combustion sources establishes a need to review current information on pertinent control technologies. This paper reviews the technologies for controlling NOx from coal-fired power plants. The review not only includes the established commercial technologies that are being used in the U.S., but also examines those that can be considered to be relatively new or in an advanced stage of development. There are several reasons for focusing this review on coal- fired power plants. First, data are available from technology applications at such plants. Second, it is more cost-effective to control NOx from large sources and, as such, it is expected that the technologies would be applied to such sources. Third, based on data in Reference 2, coal-fired power plants account for approximately 90 percent of the NOx emissions from the U.S. electric utility industry. It is expected that this review will be useful to a broad audience including;(l) individuals responsible for developing and implementing NOx control strategies at sources, (2) persons involved in developing NOx and other regulations, (3) state regulatory authorities implementing NOx control programs, and (4) the interested public at large. Moreover, persons engaged in research and development (R&D) efforts aimed at improving the cost-effectiveness of controls may also benefit from this review. Finally, this review will also be useful for technology applications on large coal-fired industrial boilers, which are quite similar to electric utility boilers. 2 ------- REGULATORY OVERVIEW The 1990 Clean Air Act Amendments (CAAA)8 authorize EPA to establish standards for a number of atmospheric pollutants, including NOx. Two major portions of the CAAA relevant to stationary source NOx control are Titles I and IV. Title 1 established National Ambient Air Quality Standards (NAAQS) for six criteria pollutants, including ozone. Title IV includes provisions designed to address acid deposition resulting from emissions of NOx and sulfur dioxide (S02) from electric power plants. Table 1 presents an overview of the regulatory actions affecting NOx sources. NOx reduction requirements under Titles I and IV are discussed below. Title I NOx Requirements Title I of the CAAA of 1990 included provisions designed to address both the continued nonattainment of the existing ozone NAAQS and the transport of air pollutants across state boundaries. These provisions also allow downwind states to petition for tighter controls on upwind states that contribute to their NAAQS nonattainment status. In general, Title I NOx provisions require: (I) existing major stationary sources to apply reasonably available control technologies (RACT); (2) new or modified major stationary sources to offset their new emissions and install controls representing the lowest achievable emissions rate (LAER); and (3) each state with an ozone nonattainment region to develop a State Implementation Plan (SIP) that, in some cases, includes reductions in stationary source NOx emissions beyond those required by the RACT provisions of Title I. Ozone Transport Commission (OTC) NOx Budget Program5. Section 184 of the CAAA delineated a multi-state ozone transport region (OTR) in the northeast and required specific additional NOx and volatile organic compound (VOC) controls for all areas in this region. Section 184 also established the OTC for the purpose of assessing the degree of ozone transport in the OTR and recommending strategies to mitigate the interstate transport of pollution. The OTR consists of Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont, parts of northern Virginia, and the District of Columbia. The OTR states confirmed that they would implement RACT on major stationary sources of NOx (Phase I), and agreed to a phased approach for additional controls, beyond RACT, for power plants and other large fuel combustion sources (Phases II and III). This agreement, known as the OTC Memorandum of Understanding (MOU) for Stationary Source NOx Controls was approved on September 27, 1994.9 All OTR states, except Virginia, are signatories to the OTC NOx MOU. The MOU establishes an emissions trading system to reduce the costs of compliance with the control requirements under Phase II (which began on May 1, 1999) and Phase HI (beginning on May 1, 2003). The OTC program caps summer-season (May 1 - September 30) NOx emissions for all 13 OTC jurisdictions at approximately 219,000 tons in 1999, and 143,000 tons in 2003, which represent approximately 55 and 70 percent reductions in NOx, respectively, from the 1990 baseline emission level of 464,898 tons. The actual reductions during the 1999 season, however, reflect participation by only 8 of the 13 jurisdictions. This subset includes Connecticut, 3 ------- Delaware, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, and Rhode Island. NOx SIP Call1. To address the long-range transport of ozone, in October 1998, EPA promulgated a rule to limit summer-season NOx emissions in 22 Northeast states and the District of Columbia that the Agency believes are significant contributors to ozone non-attainment in . downwind areas. These states were required to amend their state implementation plans (SIPs) through a procedure established in Section 110 of the CAAA. EPA finalized a summer-season state NOx budget (in tons of NOx) and developed a state implemented and federally enforced NOx trading program to provide for emissions trading by certain electric and industrial stationary sources. The state N0X budget is based on the application of a population-wide 0.15 lb/106 Btu NOx emission rate for large electricity generating units (EGUs) and a 60 percent reduction from uncontrolled emissions for large non-EGUs. The NOx SIP call is projected to reduce summer- season NOx emissions by 1.1 million tons in the affected 22 slates and DC. In response to litigation, on March 3, 2000, the D.C. Circuit Court issued its decision on the NOx SIP call, ruling in favor of EPA on all major issues, including the findings of significant contribution by the 23 states and the emissions reductions that must be achieved. On August 30, 2000, the D.C. Circuit Court extended the deadline for the full implementation of the NOx SIP call from May 1, 2003, to May 31, 2004. Section 126 Petitions10 In addition to promulgating the NOx SIP call, EPA responded to petitions filed by eight Northeastern states under Section 126 of the CAAA. The petitions request that EPA make a finding that NOx emissions from certain major stationary sources significantly contribute to ozone nonattainment problems in the petitioning states. The final Section 126 rule requires upwind states to take action to reduce emissions of NOx that contribute to nonattainment of ozone standards in downwind states. The findings affect large EGUs and both non-EGU boilers and turbines located in 12 northeast states and the District of Columbia. Like the NOx SEP call, EPA has finalized a federal NOx Budget Trading Program based on the application of a population-wide 0.15 lb/106 Btu NOx emission rate for large EGUs and a 60 percent reduction from uncontrolled emissions for large non-EGUs. The final Section 126 actions are projected to reduce summer-season NOx emissions by 510,000 tons in the 12 affected states and D.C. The compliance deadline is May 1, 2003. Title IV NOx Requirements Title IV of the CAAA authorized EPA to establish an Acid Rain Program to reduce the adverse effects of acidic deposition on ecosystems, natural resources, materials, visibility, and public health. Emissions of S02 and NOx from the combustion of fossil fuels are important contributors to acidic deposition from the atmosphere. Title IV includes provisions designed to address NOx emissions from existing power plants. Acid Rain NOx Reduction Program3'4. Under Title IV of the CAAA, the Acid Rain Program uses a two-phased strategy to achieve the required annual reductions in NOx emissions. Effective January 1, 1996, Phase I established regulations for "Group 1" boilers, which include dry-bottom, wall-fired boilers, and tangentially fired (T-fired) boilers. In Phase II, which began 4 ------- on January 1, 2000, lower emissions limits are set for certain Group 1 boilers, and regulations are established for Group 2 boilers, which include cell-burner, cyclone, and wet-bottom, wall-fired coal-fired boilers. The regulations allow for emissions averaging in which the emissions levels established by EPA are applied to an entire group of boilers owned or operated by a single company. By January 2000, the Acid Rain Program required annual average emission rates of 0.46 lb/106 Btu for dry-bottom wall-fired boilers and 0.40 lb/106 Btu for tangentially fired boilers. The limits are 0.68 lb/106 Btu for cell burners, 0.86 lb/106 Btu for cyclones greater than 155 MWC, 0.84 lb/106 Btu for wet-bottom boilers greater than 65 MWe, and 0.80 ib/106 Btu for vertically fired boilers. The Phase I compliance results for 1996 show that, from 1990 to 1996, the overall NOx emission reductions for the affected boilers totaled about 340,000 tons; i.e., a reduction of 33 percent. In Phase II, approximately 1.17 million tons per year of NOx reductions are projected to result from the Acid Rain NOx Program requirements. New Source Performance Standards (NSPS)6. Under the CAAA, new power plants are subject to NSPS that represent maximum allowable emission rates and are based upon the best adequately demonstrated technology. EPA promulgated a revised NSPS for fossil-fuel-fired utilities in 1998. The new standards revise the NOx emission limits for steam generating units and affect only units for which construction, modification, or reconstruction commenced after July 9, 1997. The NOx emission limit in the final rule is 201 nanograms per joule (ng/J) [1.6 lb/megawatt-hour (MWh)] gross energy output regardless of fuel type. For existing sources that become subject to regulation through modification or reconstruction, the NOx emission limit is 0.15 lb/106 Btu heat input. The estimated decrease in baseline nationwide NOx emissions is 25,800 tons per year, which represents about a 42 percent estimated reduction in growth of NOx emissions from new utility and industrial steam generating units subject to NSPS. NOx FORMATION IN COMBUSTION Before examining the control technologies, it would be helpful to review the mechanisms of NOx formation in combustion. These mechanisms form the basis for practical NOx control strategies, particularly those based on modification of the combustion process. NO is formed during most combustion processes by one or more of three chemical mechanisms' '12,13: (1) "thermal" NOx resulting from oxidation of atmospheric molecular nitrogen14, (2) "fuel" NOx resulting from oxidation of chemically bound nitrogen in the fuel, and (3) "prompt" NOx resulting from reaction between atmospheric molecular nitrogen and hydrocarbon radicals. Figure 2 depicts a simplified picture of the major reaction pathways for NOx formation and reduction in combustion. In fuel-lean combustion of nitrogen-free fuels, thermal NOx is the primary component of NOx emissions. Thermal NOx formation is quite sensitive to temperature and can be controlled by appropriately controlling peak temperature in the furnace. In fuel-lean combustion of fuels containing nitrogen (e.g., coal), fuel NOx contributes significantly to total NOx emissions, depending on the weight percent of nitrogen in the fuel.15 Formation of fuel NOx depends on the availability of oxygen to react with the nitrogen during coal devolatilization and the initial stages 5 ------- of combustion. Under fuel-rich conditions, the formation of NOx may compete with the formation of molecular nitrogen (N2) and may result in a reduction of NOx emissions. Prompt NOx contributes a relatively minor fraction of total NOx emissions for both nitrogen-free and nitrogen-containing fuels. The factors identified above that dictate NOx formation (devolatilization of fuel-bound nitrogen, oxygen concentration, and flame temperature) can all be adjusted by controlling the rate at which the fuel and air mix or staging the combustion process such that an initial fuel-rich zone is followed by a burnout zone that is high in oxygen to complete the combustion process, but low enough in temperature to minimize thermal NOx production. Combustion modification NOx controls utilize this combustion staging. NOx CONTROL TECHNOLOGIES In general, NOx control technologies may be placed in two general categories: primary control technologies and secondary control technologies. Primary control technologies reduce the amount of NOx produced in the primary combustion zone. In contrast, secondary control technologies reduce the N0X present in the flue gas from the primary combustion zone. Some of the secondary control technologies actually employ a second stage of combustion, such as reburning. Primary Control Technologies In the U.S., popular primary control technologies are low NOx burners (LNB) and overfire air (OFA). These technologies utilize staged combustion techniques to reduce NOx formation in the primary combustion zone. LNB and OFA are described below. Widely Used Primary Control Technologies LNB16,17. A LNB limits NOx formation by controlling the stoichiometric and temperature profiles of the combustion process. This control is achieved by design features that regulate the aerodynamic distribution and mixing of the fuel and air, thereby yielding one or more of the following conditions: (1) reduced oxygen in the primary flame zone, which limits fuel NOx formation; (2) reduced flame temperature, which limits thermal NOx formation; and (3) reduced residence time at peak temperature, which limits thermal NOx formation. In general, LNBs attempt to delay complete mixing of fuel and air mixing as long as possible, within the constraints of furnace design. This is why the flames from LNBs are usually longer that those from conventional burners. The gradual mixing of the combustion air to a fuel-rich flame core is shown schematically in Figure 3. The hardware used to influence the fuel/air mixing varies from manufacturer to manufacturer. LNBs can provide NOx reductions in excess of 50 percent from uncontrolled levels. Overfire Air (OFA)16'17. OFA, also referred to as air staging, is a combustion control technology in which a fraction, 5 to 20 percent, of the total combustion air is diverted from the burners and injected through ports located downstream of the top burner level. OFA is generally 6 ------- used in conjunction with operating the burners at a lower-than-normal air-to-fuel ratio, which reduces NOx formation. The OFA is then added to achieve complete combustion. OFA can be used in conjunction with LNBs. The addition of OFA to LNB may increase the reductions by an additional 10 to 25 percent. Primary control technologies, described above, have been widely implemented on U.S. coal-fired utility boilers to comply with the NOx emissions reduction requirements of Phase I of the Title IV NOx Program. Table 2 provides a summary of primary control applications through 1998.18 These data reflect that primary control technologies have been applied on 177 boilers and have resulted in reductions ranging from 33 to 48 percent, on average, from 1990 emissions levels. In particular, applications of LNB resulted in reductions of greater than 40 percent, on average, from 1990 levels. Recently, advances have been made in primary control technologies aimed at providing greater NOx reduction. These advances are described below. Advances LNB with Multi-level OFA. A concentric firing system with multi-level OFA is now available for tangentially fired boilers. This system incorporates deep air staging to achieve significantly lower NOx emissions. This combustion technology has achieved NOx emissions under 0.15 lb/106 Btu while firing Powder River Basin (PRB) subbituminous coal19, which reflects the potential for achieving low NOx emissions with this technology. Note, however, that for other coals that are lower in fuel volatile content than PRB coals (e.g., eastern bituminous coals) higher NOx emissions may be expected with this technology. Rotating Opposed Fire Air (ROFA)20. The ROFA design injects air into the furnace first to break up the fireball and then to create a cyclonic gas flow to improve combustion. The difference between ROFA and conventional OFA is that ROFA utilizes a booster fan to increase the velocity of air to promote better mixing and to increase the retention time in the furnace. Specific advantages of ROFA include more even distribution of combustion products, less temperature variation across the furnace, and less excess air needed for complete combustion. The technology has been installed on a 175 MWC boiler, and NOx reduction obtained has been over 50 percent. For many coal-fired boilers, it may not be possible to achieve sufficiently low NOx emissions to comply with existing or future NOx regulations by using primary measures alone. These units may require additional primary controls and/or secondary controls for future NOx compliance. Described below are the secondary control technologies applicable to coal-fired electric utility boilers. 7 ------- Secondary Control Technologies In the U.S., popular secondary control technologies are reburning, selective catalytic reduction (SCR), and selective non-catalytic reduction (SNCR). These technologies are described below. Widely Used Secondary Control Technologies Reburning 21'22. In reburning, between 15 and 25 percent of the total fuel heat input is provided by injecting a secondary (or reburning) fuel above the main combustion zone to produce a slightly fuel-rich reburn zone with a stoichiometry of about 90 percent theoretical air. Combustion of reburning fuel at fuel-rich conditions results in hydrocarbon fragments, which react with a portion of incoming NOx to form hydrogen cyanide (HCN), isocyanic acid (HNCO), isocyanate (NCO), and other species. These species then pass through an amine (NHj) + carbon monoxide (CO) step, and the NH, is ultimately reduced to N2. Finally, completion air is added above the reburn zone to complete burnout of reburning fuel. Reburning reaction pathways are shown in Figure 2, and a reburning application is schematically shown in Figure 4. Boiler size, in particular furnace height, can have an impact on one of the key design parameters - gas residence time within the reburn and burnout zones - responsible for a successful reburning application. Sufficient residence time is required to achieve adequate flue gas mixing and to accommodate the NOx reduction kinetics in the reburning zone, and for complete combustion in the bumout zone. Given sufficient time in the reburn zone, reburn zone stoichiometry is another critical parameter that influences NOx reduction. However, reburn zone stoichiometry is directly related to the heat input split between the primary and reburn zones. In general, an increase in the reburn heat input and a commensurate decrease in primary heat input will decrease the stoichiometry in the reburn zone and improve the NOx reduction efficiency. However, this heat splitting is constrained by flame stability considerations in the primary and reburn zones, the potential for unacceptable levels of CO emissions resulting from addition of relatively large amounts of reburn fuel, and the potential for increased boiler tube corrosion within the reburn zone. In addition to the above considerations, the temperature at the point of burnout air addition can place a lower limit on the achievable NOx due to reformation of thermal NOx. The choice of the reburning fuel is determined largely by fuel availability, a balancing of operating costs versus capital costs, and the specifics of the boiler. If natural gas is available on site, it may be used as the rebuming fuel, depending on the price of gas with respect to other fuels such as coal and oil. Compared to natural gas reburning (NGR), coal reburning requires a relatively longer residence time through the reburn zone and a larger upper furnace to achieve adequate bumout. Further, coal reburning requires addition of dynamic pulverizers to produce reburning fuel grade coal. Micronized coal (i.e., coal pulverized to a very high fineness) may be used as a reburning fuel on some boilers that may not have enough volume for normal coal reburning. However, micronized coal reburning requires specialized coal processing equipment, which increases its cost over gas or conventional coal reburning. In addition, for any coal reburning system it is necessary to use recirculated flue gas as the reburning fuel carrier. 8 ------- The first application of reburning technology to a wet-bottom, wall-fired unit was to the 300 MWC Ladyzhin plant in Ukraine. This application demonstrated that 50 percent NOx reduction was achievable with natural-gas-based reburning technology on wet-bottom, wall-fired boilers. In general, reburn technology has demonstrated greater than 50 percent NOx reduction on several coal-fired boiler types such as cyclone-fired, dry-bottom wall-fired, and wet-bottom wall- fired.23,24 Table 3 presents the reburning applications at coal-fired utility boilers. As seen in this table, 16 utility boilers worldwide and 14 in the U.S. have either used, or will use, reburning as their secondary NOx control technology. The NOx reductions at these boilers either achieved or expected to be achieved range from 39 to 67 percent. The largest application of reburning is at Southern Company's 818 MWe Shearer Unit 1, which utilizes coal rebuming. This application started in 2001.25 The largest application of NGR is at Scottish Power's -600 MW,. Longannet Unit 2.26 EPA is currently participating in a demonstration of a multifuel reburn system (gas, oil, and/or coal) on a 300 MWe wet-bottom wall-fired boiler in Ukraine. *70 "77 09, SNCR . SNCR is a postcombustion technology in which a reagent (ammonia or urea) is injected into the furnace above the combustion zone, where it reacts with NOx to reduce it to N2 and water. In general, SNCR reactions are effective in the range of 980 to 1150 °C. The high temperature necessary for the reaction to proceed requires that the reagent normally be injected into the boiler's upper furnace region, as shown in Figure 5, Although the actual reactions are more complex, the overall stoichiometric reactions for urea and ammonia SNCR are: (NH2)2CO + 2NO + 1/2 02 2H?0 + C02 + 2N2 (1) 2NH3 + 2NO +l/2 02-> 2N2 + 3H20 (2) After being injected into hot flue gas, urea decomposes into ammonia, which participates in SNCR chemistry. In general, ammonia may reduce NOx, oxidize to form NOx, or remain unreacted and pass through the stack. This unreacted portion is referred to as ammonia slip. Inadequate flue gas temperature and/or reaction time for SNCR kinetics, and mixing of the reagent with flue gas can contribute to an increase in ammonia slip. Relatively high concentrations of ammonia slip can react with S02 and sulfur trioxide (SO3) in the flue gas and form ammonium sulfates and bisulfates which, in turn, can cause plugging of the air preheater passages. Furthermore, ammonia slip can also reduce the marketability of flyash by making it odorous. For these reasons, ammonia slip is normally well controlled through proper specification, design, and operation of an SNCR system. Although the dominant reactions in the SNCR process result in reduction of NO to nitrogen, a significant competing reaction is the oxidation of the SNCR reagent to form NOx. This oxidation reaction becomes more significant as temperature is increased. Because of this competing oxidation reaction, there is not a one-to-one relationship between reagent injected and 9 ------- N0X reduced. It typically requires more urea or ammonia to reduce NOx than is suggested by the stoichiometric equations above. A common misunderstanding of the SNCR process is that the reagent must be injected into the flue gas where the gas is between 980 and 1150 °C. Because of this misconception, in the past it was believed that SNCR could not be used on cyclone- of wall-fired wet-bottom boilers. However, commercial SNCR systems on these boiler types exist today.22 In fact, most electric utility SNCR systems operate effectively with reagent injection where gas temperatures are above 1150 °C. In such cases, sufficiently high initial NOx concentrations cause the reduction kinetics to still dominate over the oxidation kinetics. Also, the reactions normally occur downstream of the injection location, after some cooling of the flue gas. At the low end of the SNCR temperature range, below 980 °C, the SNCR kinetics becomes slow. Nevertheless, there are applications on fluidized bed combustors where sufficient mixing time is available at these temperatures for the SNCR reactions to reach completion, and result in high levels of NOx reduction with low levels of ammonia slip. Urea and ammonia reagents require different injection approaches. Urea reacts over a somewhat broader and slightly higher temperature range than ammonia (about 55-83 °C higher), making it somewhat more compatible with the temperatures found in most utility furnaces. Further, reagent droplets containing urea penetrate well into flue gas before complete vaporization. In general, this results in improved mixing of the reagent in large flue gas volumes compared to that possible with ammonia injection. Finally, safety associated handling of urea handling may be more acceptable than that associated with handling of ammonia. For these reasons, urea is the preferred reagent in current SNCR applications on utility boilers. Boiler load changes can impact the performance of an SNCR application due to changes in two key parameters: temperature of flue gas at the furnace exit and gas residence time in the furnace. Two measures are usually included in the design of SNCR systems, namely :(1) a multi- level injection system, and (2) an ability to change the amount of reagent to be injected. It is not unusual for an SNCR system on a load-following boiler to have three injection zones, each with several in jectors. Since each of these injection zones is switched on and off automatically by the plant's control system, this does not add complexity to plant operation. SNCR has been applied in the U.S. on a wide variety of boilers firing a range of fuels. Table 4 shows SNCR applications on electric utility boilers in the U.S.29 As boiler size increases, the capability to uniformly distribute a chemical reagent, urea or ammonia, throughout the furnace volume may diminish, and, therefore, may negatively impact NOx removal efficiency. Data in Table 4 show that, while smaller boilers (e.g., Salem Harbor) may be able to achieve greater than 40 percent NOx reduction, larger boilers (e.g., Cardinal) may be capable of achieving reductions of only 30 percent or less. SCR22,30. SCR is a postcombustion NOx reduction technology in which ammonia is added to the flue gas, which then passes through layers of a catalyst. The ammonia and NOx react on the surface of the catalyst, forming N2 and water. SCR reactions are generally effective in a temperature range of 350 to 400°C. In general, SCR is capable of providing high levels of NOx 10 ------- reduction, ranging from 80 to greater than 90 percent. The overall stoichiometric SCR reactions are: 2NH3 + 2NO + 1/2 02-> 2N2 + 3H20 (3) 4NH3 + 2N02 + 02 -> 3N2 + 6H20 (4) In most utility boiler applications, the catalyst is installed in a separate reactor positioned downstream of the boiler economizer and immediately upstream of the air preheater (see Figure 6). Under low-load conditions, an economizer bypass is sometimes utilized to ensure proper flue gas temperature at the SCR reactor inlet. The installation shown in Figure 6 is called a "hot-side" and/or "high-dust" SCR installation. A "low-dust" application may be used at facilities with hot- side ESPs. In this case, the SCR reactor would be installed immediately downstream of the ESP and prior to the air preheater. An alternative is a "cold-side" or "tail-end" SCR installation in which the SCR reactor is located after the FGD system. This installation requires reheating of flue gas with auxiliary fuel or other means prior to entry into the SCR reactor. Because of cost considerations, the majority of SCR installations on utility units are of the "high-dust" type. An ammonia injection grid is located upstream of a titania-vanadia (T1O2-V2O5) catalyst. Flue gas with ammonia passes through the catalyst, which may be of the ceramic honeycomb or coated parallel metal plates type. The catalyst provides the active sites where the ammonia and NOx reduction reactions take place. Because of the temperature at which this reaction occurs, the competing ammonia oxidation reaction is not significant. This results in two advantages over SNCR: (1) much lower outlet NOx concentrations are possible, and (2) the reagent is more efficiently utilized. Since most of the NOx is in the form of NO in boilers, the ratio of ammonia added to NOx reduced is typically close to 1:1 [see Eq. (3)]. The local molar ratio of ammonia to NOx in the flue gas has a great impact on SCR performance. This process parameter becomes more critical for SCR systems designed for high reduction efficiencies. If appropriate distribution of this parameter is not possible, then additional catalyst is needed to ensure adequate performance. Gas velocity and spatial temperature distribution at the catalyst face are somewhat less critical, but are still important. Achieving proper ammonia/ NOx ratio throughout the flow field and proper gas flow and temperature distribution are addressed in the design stage through flow modeling and through optimization of the ammonia injection grid at SCR system start up and periodically thereafter. Catalyst deactivation occurs as a result of impurities in the gas stream that can produce poisoning of the catalyst material or blinding deposits. However, for coal-fired boilers, catalysts have been developed that can tolerate coal impurities and provide reasonable catalyst lifetimes, typically, in the range of about 14,000-32,000 operating hours before replacing a portion - about 25 or 35 percent ~ of the total catalyst loading.22 The cost of catalyst replacement has been addressed by developing catalyst management plans, which minimize the cost of catalyst replacement over the projected lifetime of the SCR system's operation. In such plans, the catalyst is usually installed in layers to permit periodic 11 ------- replacement of portions of the total catalyst loading. In this manner, the activity of the catalyst is more fully utilized over the life of the installation. The time between catalyst layer replacements is measured in years. At the dry-bottom coal-fired U.S. boilers equipped with full SCR, the planned time between catalyst changes is typically about 24,000 operating hours and at one boiler it is 32,000 hours, or 3 or more years of operation on a typical unit. The catalyst replacement frequency for boilers that reinject fly ash may be on the order of 14,000 hours, or nearly 2 or more years of operation. SCR operators and catalyst manufacturers periodically monitor the activity of portions of the catalyst to ensure that the catalyst is maintaining the planned amount of activity. With this testing, it is possible to predict future catalyst replacement schedules. As mentioned before, difficulties with arsenic poisoning of SCR catalyst on wet-bottom boilers that reinject their fly ash are well known. Most operators of these facilities add limestone to their coal as a slag fluxing agent and also have an accelerated catalyst management plan. In general, a high arsenic concentration in the flue gas in contact with the catalyst will increase catalyst replacement frequency and/or increase catalyst loading compared to low arsenic applications. For example, at Merrimack 2, a cyclone boiler with 100 percent fly ash reinjection, the expected time between replacement of layers is 14,000 operating hours. However, this is less than the typical time between replacement of catalyst layers for SCR systems on dry-bottom pulverized-coal boilers (-24,000 operating hours).22,31,32 In the U.S., low-sulfur, high-calcium subbituminous coals from the Powder River Basin (PRB) are used at many power plants to comply with S02 regulations. Since these coals are unique to the U.S., until recently, there was no experience with SCR on facilities firing this coal. There is some concern that firing of such coals may lead to deposits on SCR catalysts and accelerated deactivation of these catalysts. According to one supplier, the rate of catalyst deactivation is expected to be within an acceptable range for commercial use.33 Commercial experience with PRB coal-fired electric utility boilers that have recently commenced operation with SCR will shed more light on this issue.34 Another issue that has recently surfaced is catalyst arsenic poisoning on dry-bottom boilers firing coals from western Pennsylvania and West Virginia. These coals do not have unusually high arsenic contents (although higher than for European coals). However, these coals sometimes have unusually low content of free calcium oxide (CaO) in the fly ash. CaO acts to scavenge gaseous forms of arsenic to form calcium arsenide. If free CaO is too low (below about 2.5 percent by weight of the fly ash), it is possible that arsenic will not be scavenged and will lead to poisoning of the catalyst. Note that arsenic oxide chemically bonds to the catalyst surface so that the catalyst cannot be cleaned. As a result, the catalyst is permanently poisoned. To address this issue, some facilities have accelerated their catalyst management plans, and others are adding small amounts of pulverized limestone to their coal.35 SCR has been extensively used to control NOx from hundreds of utility and industrial boilers in Japan and Germany, and several coal- and gas-fired utility boilers in the U.S. Table 5 shows a list of U.S. utility facilities where SCR is currently in operation, under construction, or under agreement for installation.29 Many of these applications are designed to provide NOx reductions of 80 percent or greater.29 This table also indicates those facilities at which SCR was 12 ------- in operation in 1998, Based on the data in Table 5, SCR has become a popular NOx control technology in the U.S. in the last 3 years. Recently, advances have been made in secondary control technologies aimed at providing large NOx reductions, more efficient use of reagents, and eliminating any public concerns with the transport and handling of the ammonia reagent used in SCR applications. These advances are described below. Advances Advanced Gas Rebuming (AGR)12. With AGR, a nitrogen compound (typically urea or ammonia) is added downstream of the rebuming zone. The reburning system is adjusted for somewhat lower NOx reduction to produce free radicals that enhance SNCR NOx reduction. AGR systems can be designed in two ways: (1) non-synergistic, which is essentially the sequential application of NGR and SNCR [i.e., the nitrogen agent (urea or ammonia) is injected downstream of the burnout air]; and (2) synergistic, in which the nitrogen agent is injected either with or before the burnout air. To obtain maximum NOx reduction and minimum ammonia slip in non-synergistic systems, the nitrogen compound must be injected so that it is available for reaction with the furnace gases within a temperature zone around 1000 °C. A synergistic AGR system was demonstrated on the 104 MWe Greenidge Unit 6 in New York to reduce NOx emissions by 68-76 percent. However, it could not reduce ammonia slip to less than 10 ppm.35 Fuel-Lean Gas Reburning (FLGR)22. FLGR, also known as controlled gas injection, is a process in which careful injection and controlled mixing of natural gas into the furnace exit region reduces NOx. The gas is normally injected into a lower temperature zone than that in NGR. Whereas NGR requires 15-20 percent of furnace heat input from gas and requires bumout air, the FLGR technology achieves NQx control using less than 10 percent gas heat input and no bumout air. Lower NOx reductions are achieved with FLGR when compared with NGR. FLGR has been demonstrated to reduce NOx emissions by roughly 33-45 percent at full load, with less than 7 percent of the heat input attributed to the re burn fuel. Table 6 lists FLGR applications.35 Amine-Enhanced Fuel-Lean Gas Reburn (AEFLGR)22. AEFLGR is similar to AGR, except that bumout air is not used, and the SNCR reagent and rebum fuel are injected to create local, fuel-rich NOx reduction zones in an overal 1 fuel-I ean furnace. The fuel-rich zone exists in local eddies, as in FLGR, with the overall furnace in an oxidizing condition. However, the SNCR reagent participates with natural gas (or other hydrocarbon fuel) in a NOx reduction reaction, which is believed to be different than the reaction that occurs when ammonia or urea is used in SNCR. In SNCR the NOx reduction occurs in an oxidizing environment, while in AEFLGR the ammonia or urea is injected into the reducing zone. High reductions are possible because, with the local low oxygen environment, the AEFLGR NOx reduction reaction does not have to compete as much with the Zeldovich thermal NOx reaction that limits SNCR performance. Preliminary results at a demonstration plant show approximately 73 percent reduction at about 40 percent load, 60 percent reduction at 60 percent load, and 30-40 percent reduction at full load. 13 ------- AEFLGR is in commercial operation at PSE&G's Mercer station units 1 & 2. PSE&G is also planning a commercial AEFLGR system at Hudson, unit 2. AEFLGR was also commercially demonstrated at Wisconsin Electric Power Company's Pleasant Prairie 1. Table 6 lists AEFLGR applications.35 Hybrid Selective Reduction (HSR)22. HSR is a combination of SNCR and SCR that is designed to provide the performance of full SCR with significantly lower costs. In HSR, an SNCR system is used to achieve some NOx reduction and to produce a controlled amount of ammonia slip that is used in a downstream in-duct SCR reactor for additional NOx reduction. The HSR technology has lower capital costs than SCR, allows installation to be phased in, provides better reagent utilization than SNCR, and has very good ammonia slip control. A test done with this hybrid system showed 95 percent NOx reduction with less than 5 ppm ammonia slip and 55 percent reagent utilization. In-duct SCR, There has been one full-scale demonstration installation of In-duct SCR on a coal-fired unit at PSE&G's Mercer Generating Station. The approach entailed installing the catalyst in an expanded duct rather than a separate reactor and was selected because the site did not have the necessary room for a full SCR and the associated ductwork; utilizing a full SCR would have required substantial, costly modifications to the entire facility. At Mercer, the In-duct SCR handles 25 percent of the total flow from a 320 MWe boiler and follows a commercially operating SNCR system that treats all of the boiler gases. This unit demonstrated between 85 and 90 percent NQX reduction with under 10 ppm ammonia slip at the air heater inlet. At the outlet of the catalytic air heater, 90 percent reduction was achieved with no measurable ammonia slip.36 To enhance the performance of In-duct SCR, some vendors offer catalyst-coated air heater baskets for Lungstrom type air heaters. In these situations, the plant's existing hot-end baskets are removed from the Lungstrom air preheater, and they are replaced with baskets that are coated with catalyst. These catalyst-coated air heater baskets may be used separately, or in addition to In-duct SCR. Testing at Mercer Generating Station and experience in Europe with this catalyst in coal-fired systems demonstrate that this approach is capable of providing some additional NOx reduction and, more importantly, can be very effective in eliminating ammonia slip. For example, the catalytic air heater baskets installed at Mercer station provided sufficient ammonia slip control so that an additional 20 percent reduction was achieved while maintaining the ammonia slip limit.36 However, catalytic air heater technology generally will not provide 37 sufficient reduction of NOx to be a stand-alone technology. Ammonia on Demand (AOD)n. In AOD, ammonia reagent for SCR application is produced from urea through on-site hydrolysis. In this conversion process, urea pellets are mixed with small amounts of deionized water. This mixture is then pumped to a pressure vessel (the hydrolizer) where heat and pressure are applied and urea sublimates to ammonia and carbon dioxide (CO2). Steam is then used to strip out the ammonia gas and the steam/ammonia mixture is then supplied to the SCR system. AOD eliminates any public concerns with transport and handling of ammonia and associated Occupational Safety & Health Administration (OSHA) requirements. However, the incorporation of AOD into an SCR installation would add a small 14 ------- chemical manufacturing plant and associated complexities. To date it has only been demonstrated at a large 565 MWe oil-fired utility boiler that slowly changes load. SUMMARY Recently, a number of regulatory actions have been taken in the U.S. focused on reducing NOx emissions from stationary combustion sources, particularly electric utility boilers. As a result of these regulations, state-of-the-art NOx control technologies have been applied to a large number of coal-fired U.S. utility boilers. This paper reviews these technologies and their applications. In general, NOx control technologies may be placed in two general categories: primary control technologies and secondary control technologies. Primary control technologies reduce the amount of NOx produced in the primary combustion zone. In contrast, secondary control technologies reduce the NOx present in the flue gas from the primary combustion zone. The popular primary control technologies in use in the U.S. are LNB and OFA. These technologies utilize staged combustion techniques to reduce NOx formation in the primary combustion zone. Primary control technologies have been widely implemented on U.S. coal-fired utility boilers to comply with the NOx emissions reduction requirements of Phase I of the Title IV NOx Program. Data reflect that primary control technologies have been applied on 177 boilers and have resulted in reductions ranging from 33 to 48 percent, on average, from 1990 emissions levels. In particular, applications of LNB resulted in reductions of greater than 40 percent, on average, from 1990 levels. Advances in primary control technologies include LNB with multilevel OFA and ROFA systems. The former system has been able to provide relatively low NOx emissions of 0.15 lb/106 Btu at a subbituminous coal-fired boiler. The latter system has been used to achieve NOx reductions in excess of 50 percent at a utility boiler. The secondary NOx control technologies being used on U.S. coal-fired utility boilers include reburning, SNCR, and SCR. Of these boilers, 14 have used, or will use, reburning as their secondary NOx control technology. The NOx reductions at these boilers, either achieved or expected to be achieved, range from 39 to 67 percent. Twenty-two U.S. coal-fired utility boilers either have used SNCR, or will use it in the future. NOx reductions achieved, or projected, at these applications range from 20 to 62 percent. In the last 3 years, SCR has been the preferred secondary technology at numerous U.S. coal-fired utility boilers. Current data indicate that 79 boilers either use, or will use, SCR for NOx control. Many of these applications are designed to provide reductions in excess of 80 percent. Advances in secondary control technologies include AGR, FLGR, AEFLGR, HSR, In- duct SCR, and AOD. FLGR is a variation of NGR, focused on reducing gas usage. AGR and AEFLGR are combinations of NGR and SNCR, focused on obtaining large NOx reductions. AGR has been demonstrated on a 105 MWC utility boiler to reduce NOx emissions by 68-76 percent; however, it could not reduce ammonia slip to less than 10 ppm. FLGR has been applied 15 ------- at five coal-fired U.S. boilers to achieve NOx reductions ranging between 27 and -40 percent. Also, four coal-fired U.S. boilers have used AEFLGR to achieve NOx reductions ranging between 50 and 70 percent. HSR is a combination of SNCR and SCR that is designed to provide the performance of full SCR with significantly lower costs. A test with HSR showed 95 percent NOx reduction with less than 5 ppm ammonia slip and 55 percent reagent utilization. In-duct SCR entails installing the catalyst in an expanded duct rather than a' separate reactor. This may be an attractive option for plants with constrained footprints. An application of In-duct SCR, in conjunction with SNCR, has demonstrated between 85 and 90 percent NQX reduction with under 10 ppm ammonia slip at the air heater inlet. In AOD, the ammonia reagent for SCR application is produced from urea through on-site hydrolysis. AOD eliminates any public concerns with transport and handling of ammonia and associated OSHA requirements. To date it has been demonstrated only at a large 565 MWe oil-fired utility boiler that slowly changes load. ACKNOWLEDGEMENTS/DISCLAIMER The research described in this article has been reviewed by the U.S. Environmental Protection Agency and approved for publication. The contents of this article should not be construed to represent U.S. government policy nor does mention of trade names or commercial products constitute endorsement or recommendation for use. 16 ------- REFERENCES Price, D.; Bimbaum, R.; Batiuk, R.; McCullough, M.; Smith, R. Nitrogen Oxides: Impacts on Public Health and the Environment; U.S. Environmental Protection Agency, Office of Air and Radiation, Washington, DC, August 1997; EPA-452/R-97-002 (NTIS PH98-104631). Bimbaum, R. National Air Quality and Emissions Trends Report, 1997; U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, NC, 1998; EPA- 454/R-98-016. Acid Rain Program: Nitrogen Oxides Emission Reduction Program: Direct Final Rule, Federal Register, Volume 60, No. 71, April 13, 1995. Acid Rain Program: Nitrogen Oxides Emission Reduction Program: Final Rule, Federal Register, Volume 61, No. 245, December 19, 1996. Ozone Transport Commission (OTC) NOx Budget Program, U.S. Environmental Protection Agency, Washington, DC, June 2001. Available at the web site http://www.epa.oov/airmarkets/otc/index.html. Revisions of Standards of Performance for Nitrogen Oxide Emissions From New Fossil- Fuel Fired Steam Generating Units; Revisions to Reporting Requirements for Standards of Performance for New Fossil-Fuel Fired Steam Generating Units, Federal Register, Volume 63, No. 179, September 16, 1998. Regional Transport of Ozone, U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, NC, Federal Register, Volume 63, No. 207, October 27, 1998. Clean Air Act, 1990 Amendments, U.S. Environmental Protection Agency, Office of Air and Radiation, Washington, DC. Available at the web site http://www.epa.gov/oar/caa/contents,html. Memorandum of Understanding Among the States of the Ozone Transport Commission on Development of a Regional Strategy Concerning the Control of Stationary Source Nitrogen Oxide Emissions, September 27, 1994, Ozone Transport Commission, Washington, DC, Available at the web site http://www.sso.org/otc/Formal%20Actions/att2.HTM. Section 126 Petitions - Findings of Significant Contribution and Rulemakings, U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, NC. Available at the web site http://www.epa.gov/ttn/rto/126/index.html. 17 ------- 11 12 13 14 15 16 17 18 19 20 21 22 23 Miller, J. A.; Bowman, C.T. Mechanism and modeling of nitrogen chemistry in combustion, Prog. Energy Combust. Sci. 1989,15, 287-338. Wendt, J.O.L. Fundamental coal combustion mechanisms and pollutant formation in furnaces, Prog. Energy Combust. Sci. 1980, 6, 201-222. Bowman, C.T. In Fossil Fuel Combustion: A Source Book; Bartok, E. and Sarofim, A.F., eds.; John Wiley & Sons, Inc., New York, NY, 1991; pp. 228-252. Zeldovich, Y.B. The oxidization of nitrogen in combustion explosions, Acta Physiochimica, U.S.S.R. 1946, 21, 577-628. Pohl, J.H.; Sarofim, A.F. Fate of coal nitrogen during pyrolysis and oxidation, Proc. of the Stationary Source Combust. Symp., Volume I, pp. 1-125 through 1-152, U.S. Environmental Protection Agency, Industrial Environmental Research Laboratory, Research Triangle Park, NC, June 1976; EPA-600/2-76-152a (NTIS PB-256320). Stamey-Hall, S. Alternative Control Techniques Document - NOx Emissions from Utility Boilers', U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, NC, March 1994; EPA^53/R-94-023 (NTIS PB94- 184165). Analyzing Electric Power Generation Under CAM; U.S. Environmental Protection Agency, Office of Air and Radiation, Washington, DC, March 1998. Available at the web site http://www.epa.gov/capi/ipm/update.htm. Krolewski, M.J.; Migst, A.S. Recent NOx reduction efforts: an overview, ICAC Forum 2000, Cutting NOx Emissions, Roslyn, VA, March 2000. Hoh, R.H.; Jennings, P. Reliant Energy, W.A. Parish 7 maintaining 0.15 LB NOx/MMBtu, ICAC Forum 2000, Cutting NOx Emissions, Roslyn, VA, March 2000. Ralston, J.; Fischer, E. Application of Mobotec ROFA technology on a 150 MW coal- fired CE boiler, Electric Utilities Environmental Conference, Tucson, AZ, January 2001. Wendt, J.O.L.; Sternling, C.V.; Matovich, M.A. Reduction of sulfur trioxide and nitrogen oxides by secondary fuel injection, Proc. 14th Symp. (Int.) on Combustion, The Combustion Institute, Pittsburgh, PA, 1973; p. 881. Staudt, J. Status Report on NOx Control Technologies and Cost Effectiveness for Utility Boilers, Northeast States for Coordinated Air Use Management, Boston, MA, 1998. Farzan, H.; Maringo, G.J.; Johnson, D.W.; Wong, D.K.; Beard, C.T.; Brewster, S.E. B&W's advances on cyclone NOx control via fuel and air staging technologies, EPRI- DOE-EPA Combined Utility Air Pollution Control Symposium: The MEGA Symposium, Atlanta, GA, August 16-20, 1999; EPRITR-113187-V2. 18 ------- Engelhardt, D.; Folsom, B.; Latham, C.; Moyeda, D.; Payne, R.; Sommer, T.; Brocato, H.; Maziarz, M. Updated experience using reburn technology for utility boiler NOx emissions reduction, EPRI-DOE-EPA Combined Utility Air Pollution Control Symposium: The MEGA Symposium, Atlanta, GA, August 16-20, 1999; EPR1TR- 113187-V2. Personal communication between R.E, Hall of EPA (NRMRL-RTP) and R. Payne of GE- Energy and Environmental Research Corp., Irvine, CA, July 2001. Ghiribelli, L.; Pasini, S.; Benedetto, D,; De Santis, R.; Mainini, G. Results from the application of a gas reburning technology on a 600 MW coal fired boiler, EPRI-DOE- EPA Combined Utility Air Pollution Control Symposium: The MEGA Symposium, Atlanta, GA, August 16-20, 1999; EPRITR-113187-V2. Lyon, R.K. Method for the reduction of the concentration of NO in combustion effluents using ammonia. U.S. Patent No. 3,900,554 (1975). White Paper: Selective non-catalytic reduction (SNCR) for controlling NOx emissions; SNCR Committee, Institute of Clean Air Companies, Inc., Washington, DC, 1997. Personal communication between R.K. Srivastava of EPA (NRMRL-RTP) and D. Foerter of Institute of Clean Air Companies, Inc., Washington, DC, June 2001. White Paper: Selective catalytic reduction (SCR) control of NOx emissions, SCR Committee; Institute of Clean Air Companies, Inc., Washington, DC, 1997. Lauber, J.; Cohen, M.; Donais, R. The integration of low NOx control technologies at the Southern Energy, Inc. Birchwood Power facility, EPRI/DOE/EPA 1997 Combined Utility Air Pollution Control Symposium, Washington, DC, August 25-29, 1997; EPRI TR-108683-V1. Cochran, J.R.; Scarlett, D.; Johnson, R. SCR for a 460 MW coal fueled unit: Stanton Unit 2 design, startup, and operation, EPRI/DOE/EPA 1997 Combined Utility Air Pollution Control Symposium, Washington, DC, August 25-29, 1997; EPRI TR-108683-V1. Pritchard, S.; Hellard, D.; Cochran, J. Catalyst design experience for 640 MW cyclone boiler fired with 100% PRB fuel, EPRI-DOE-EPA Combined Utility Air Pollution Control Symposium: The MEGA Symposium, Atlanta, GA, August 16-20, 1999; EPRI TR-113187-Y2. Cochran, J.; Hellard, D.; Rummenhohl, V. Design and initial startup results from the New Madrid SCR retrofit project, ICAC Forum 2000, Cutting NOx Emissions, Roslyn, VA, March 23-24, 2000. ------- Personal communication between R.K. Srivastava of F.PA (NRMRL-RTP) and. J. Staudt of Andover Technology Management, Andover, MA, June 2001. Huhmann. A. Evaluation of retrofitted combustion NOx control technology on a wet bottom, coal-fired utility boiler at Mercer Generating Station, First Annual DOE Conference on SCR and SNCR, Pittsburgh, PA, May 21-23, 1997. Huttenhofer, K.; Beer, J.K.; Smeets, H.; van der Kooij, J. The DeNOx air preheater downstream of a coal-fired boiler, EPRI/EPA 1993 Joint Symposium on Stationary Combustion NOx Control, May 24-27. 1993. Slocomb, S.H.; Raczynski, D.T. Comparison of the cost effectiveness of new NOx control technologies with conventional selective catalytic reduction for combined cycle combustion turbine power plants, Air & Waste Management Association's 94th Annual Conference & Exhibition, Orlando, FL, June 24-28, 2001. 20 ------- Table 1. NOx reduction regulations under Titles I and IV of the CAAA. Regulatory action Affected regions Compliance date Control period NOx reductions Title I OTC1 NO, Budget Program 12 States & DC: CT, DE, ME, MD, MA, NH, NJ, NY, PA, RI, VT, VA Phase II: May I, 1999 Phase III: May 1, 2003 ozone seasond 246,000 tons' in 1999, 322,000 tons in 2003 NOx SIPl> Call 22 States & DC: AL, CT, DE, GA, IL, IN, KY, MD, MA, MI, MO, NJ, NY, NC, OH, PA, RI, SC, TN, VA, WI, and WV May 31,2004 ozone season 1.1 million tons in 2007 Section 126 Rule 12 States & DC: DE, IN, KY, MD, MI, NJ, NY, NC, OH, PA, VA, and WV May 1,2003 ozone season 510,000 tons in 2007 Title IV Acid Rain Program nationwide Phase I: January 1, 1996 Phase II: January 1, 2000 annual 340,000 tons per year 2.06 million tons/yr NSPS' nationwide July 9, 1997 annual 25,800 tons/yr a OTC = Ozone Transport Commission. b SIP = State Implementation Plan. c NSPS = New Source Performance Standards. 11 Ozone season = time period May 1 through September 30. c Kg = tons *907.18. 21 ------- Table 2. Primary NOx control technology applications on boilers affected under Phase I of Title IV NOx Reduction Program, • Boiler type Technology Number of 1998 average. NOx reduction boilers emission rate, from 1990 levels, lb/106 Btub percent Dry-Bottom LNB 66 0,45 44 Wall-Fired LNB with OFA 21 0.47 48 LNB 44 0.36 43 Tangential SOFA3 23 0.37 33 LNB with SOFA 23 0.36 45 * Separated ovcrfire air. hng/J = lb/106 Biu *431.0017. 22 ------- Table 3. Reburning applications on coal-fired utility boilers (demonstration systems not currently in operation shown in italics). Boiler Rating, MWC Furnace Reburn fuel, heat input percent FGR'1 Baseline NOx, lb/106 Btub or (ppm) Reduction, percent Chalk Point 1 350 Wall Gas, c No 0.80 c Chalk Point 2 350 Wall Gas, c No 0.80 c Edge Moor 4 160 Tangential Gas, 23.5 No 0.31 39 Crane 1 200 Cyclone Gas, 18 No 1.5 60 Crane 2 200 Cyclone Gas, 18 No 1.5 60 Allen 1 330 Cyclone Gas, c No 1.20 65 Hatfield 2 600 Wall Gas, c No 0.60 c Hennepin 1 71 Tangential Gas, 18 Yes 0.75 67 Grecnidge 6 104 Tangential Gas, 15 No 0.62 52 Longannet 2, Scotland 600 Wall Gas, -20 Yes (-320 pprn) 50 Niles I 108 Cyclone Gas No (650 ppm) 53 Nelson Dewey 2 100 Cyclone Coal, 30 Yes 0.75 56 Lakeside 7 33 Cyclone Gas, 26 Yes 0.95 66 Ladyzhin 6, Ukraine 300 Wall, wet Gas, 12 Yes 0.82 -50 Cherokee 3 158 Wall Gas, 22 Yes 0.75 64 Shearer 1 818 Tangential Coal, c Yes c c a FOR = flue gas recirculation. b ng/I = lb/10 Btu * 431.0017. c Currently, data are not available. 23 ------- Table 4, SNCR applications on U.S. coal-fired utility boilers, including commercial demonstrations, systems in operation, or systems under contract.3 Plant Rating, MWe Furnace Baseline NOx, Ib/IO6 Btud Reduction, percent NH3 slip, ppm Salem Harbor 1 85 Wall 0.42 30 <10 2 85 Wall 0.60 50 <10 3 150 Wall 0.43 30 <10 Somerset 8b 112 Tangential 0.5-0.92 24-62 <5 Mercer 1 320 Wall, wet 1.6-1.8 25-33c <5 2 320 Wall, wet 1.6-1.8 25-33" f Hudson 2 620 Wall 0.65 25e f Edge Moor 3 84 Tangential 0.70 45 <5 Indian River 3 178 Wall 0.46 26 f 4 440 Turbo 0.44 34 Seward with catalystc 150 Tangential 0.75 0.75 42 -55 <2 Cromby 1 160 Wall 0.35 25 <10 B, L. England 1 130 Cyclone 1.31 31 <5 2 163 Cyclone 1.40 36 <5 Schiller 4 50 Wall 0.4-0.45 43-55 <10 5 50 Wall 0.4-0.45 43-55 <10 6 50 Wall 0.4-0.45 43-55 <10 Miami Fort 6 150 Tangential 0.55 35 f Cardinal 2 620 Wall 0.57-0.75 30 <5 East Lake 3 123 Tangential 0.40 20-32.5 f Sammis 2 185 Wall 0.45 25-30 f Ashe vi lie 1 207 Wall 0.58 25e f a All SNCR applications use urea as the SNCR reagent, b Somerset controls to 0.25 lb/106 Btu during the ozone season and to 0.35 the rest of the year. C Seward has added a layer of catalyst and operates at levels below 0.40 lb/106 Btu. dng/J = lb/10A Btu *431.0017. 'These units are also being equipped with AEFLGR. Percent reduction shown here is that achieved with SNCR only. fCurrently. data are not available. 24 ------- Table 5. SCR applications on U.S. coal-fired utility boilers. The applications include systems in operation and systems that are planned. (Applications in italics were the only ones in operation in 1998.) Boiler Rating, MWC Furnace New or Retrofit Gorgas 10 700 Tangential Retrofit Widows Creek 7 575 Tangential Retrofit Widows Creek 8 550 Tangential Retrofit Indiantown 330 Wall, dry New Stanton 2 460 Wall, dry New Bowen 1 805 Tangential Retrofit Bowen 2 788 Tangential Retrofit Bowen 3 952 Tangential Retrofit Bowen 4 952 Tangential Retrofit Hammond 4 578 Wall, dry Retrofit Wansley 1 950 Tangential Retrofit Wansley 2 950 Tangential Retrofit Baldwin 1 560 Cyclone Retrofit Baldwin 2 560 Cyclone Retrofit Baldwin 3 635 Tangential Retrofit Coffeen 1 390 Cyclone Retrofit Coffeen 2 617 Cyclone Retrofit E.D. Edwards 3 365 Wall, dry Retrofit Gibson 2 668 Wall, dry Retrofit Merom 1 490 Op/Turbo Retrofit Merom 2 490 Op/Turbo Retrofit Michigan City 12 520.9 Cyclone Retrofit R.M. Schahfer 14 511 Cyclone Retrofit East Bend 2 669 Wall, dry Retrofit Paradise 1 704 Cyclone Retrofit Paradise 2 704 Cyclone Retrofit Paradise 3 1150 Cyclone Retrofit Brandon Shores 1 620 Wall, dry Retrofit Brandon Shores 2 685 Wall, dry Retrofit Morgantown 1 626 Tangential Retrofit Morgantown 2 626 Tangential Retrofit Wagner 3 359 Wall, dry Retrofit Hawthorn 5 540 Tangential Retrofit latan 725 Wall, dry Retrofit New Madrid 1 600 Cyclone Retrofit New Madrid 2 600 Cyclone Retrofit Sioux 1 549.8 Cyclone Retrofit Sioux 2 549.8 Cyclone Retrofit Thomas Hill 2 285 Cyclone (Continued) 25 ------- Table 5. (Cont.) Boiler Rating, MWC Furnace New or Retrofit Roxboro 3A 745 Wall, dry Retrofit Merrimack 1 122 Cyclone Retrofit Merrimack 2 330 Cyclone Retrofit Logan 218 Wall, dry New Carney's Point 1 187 Wall, diy New 2 187 Wall, dry New Somerset 1 690 Wall, dry Retrofit Eastlake 5 680 Wall, dry Retrofit J. M. Stuart 1 610 Wall, dry Retrofit J. M. Stuart 2 610 Wall, dry Retrofit J. M. Stuart 3 610 Wall, dry Retrofit J. M. Stuart 4 610 Wall, dry Retrofit Killcn Station 2 612.5 Wall, dry Retrofit Miami Fort 7 557 Wall, dry Retrofit Miami Fort 8 557 Wall, dry Retrofit Zimmer 1 1300 Opposed Retrofit Beckjord 5 244 Tangential Retrofit Beckjord 6 460 Tangential Retrofit Bayshore 4 220 Wall, dry Retrofit Bruce Mansfield 1 913.8 Wall, dry Retrofit Bruce Mansfield 2 913.8 Wall, dry Retrofit Bruce Mansfield 3 913.8 Wall, dry Retrofit Homer City 1 660 Wall, dry Retrofit Homer City 2 660 Wall, dry Retrofit Homer City 3 692 Wall, dry Retrofit Montour 1 806 Tangential Retrofit Montour 2 819 Tangential Retrofit Allen 1 330 Cyclone Retrofit Allen 2 330 Cyclone Retrofit Allen 3 330 Cyclone Retrofit Bull Run 1 950 Tangential Retrofit Cumberland 1 1300 Wall, dry Retrofit Cumberland 2 1300 Wall, dry Retrofit Parish 5 734 Wall, dry Retrofit Parish 6 734 Wall, dry Retrofit Birchwood 240 Tangential New Chesterfield 6 694 Tangential Retrofit Harrison 1 684 Wall, dry Retrofit Harrison 2 684 Wall, dry Retrofit Harrison 3 684 Wall, dry Retrofit 26 ------- Table 6. FLGR and AEFLGR applications on U.S. coal-fired utility boilers. Demonstration systems not currently in operation shown in italics. Rebum nh3 slip, Ppm Initial Boiler Rating, MWC FLGR or AEFLGR Furnace fuel, heat input NOx, lb/106 Reduction, percent percent Btuc Joliet 340 FLGR Cyclone Gas, 6 b 1.106 38 Elrama 1 112 FLGR Roof Gas, 5 b 0.59 30-35 Elrama 2 112 FLGR Roof Gas, 5 b 0.59 30-35 Elrama 3 112 FLGR Roof Gas, 5 b 0.59 30-35 Riverbend 140 FLGR Tangenti al Gas, ~5 b 0.45 -40 Mercer 1 320 AEFLGR Wall Gas, 6-7 < 5 ppm 1.5 50-70 Mercer 2 320 AEFLGR Wall Gas, 6-7 < 5 ppm 1.5 50-70 Hudson3 660 AEFLGR Wall Gas, b <5 ppm b b Asheville 1 207 AEFLGR Wall Gas, 5 b b 50 Pleasant Prairie 600 FLGR AEFLGR Turbo Gas, ~5 b 0.46 0.46 27 57 * Hudson is already equipped with an SNCR system. b Currently, data are not available. c ng/'J = lb/106 Btu *431.0017. 27 ------- Transportation 11,595,000 tons 50% Biogenic and miscellaneous 346,000 tons Industrial processes 917,000 tons 4% Electric utilities 6,178,000 tons Industrial and other combustion 4,546,000 tons 19% Figure 1. Sources of NOx emissions in the U.S. in 1997. To convert emissions in tons to kg, multiply by 907.18. 28 ------- Hydrocarbon fuel HCN O, OH i k OH, O, H HNCO NCO CH, fixation NH, CH, NH, O, OH NO Figure 2. NOx formation and destruction pathways. 29 ------- Swirling flow Combustion product recirculation 2one Secondary air Fuel and primary air (fuel-rich) Fuel-rich axial flame core Gradual mixing of partiaJly burned products and secondary air Figure 3. Schematic of a low NOx burner. 30 ------- Burnout Burnout zone Rebum Reburn fuel zone Pnmary combustion zone Economizer Airpreheater Flue gas to Main fuel pnmary air Secondary Figure 4. Schematic of a reburning application. 31 ------- Reagent Injection Reagent metering and control Boiler Economizer Air preheater Flue gas to \ stack Main fuel primary air Secondary air From urea or ammonia storage Figure 5. Schematic of an SNCR application. 32 ------- Ammonia injection SCR reactor §" O a Boiler Economizer Air p reheat er Flue gas to stack Main fuel and primary air Secondary air Ammonia vaporizer Air blower Liquid ammonia storage tank Figure 6. Schematic of an SCR application. 33 ------- TECHNICAL REPORT DATA N RMRL~ RTP~ P~ 622 (Please read Instructions on the reverse before completii 1- REPORT NO. 2. EPA/600/A—01/115 3. RECIP 4 TITLE AND SUBTITLE Control of NOx Emissions from U. S. Coal-fired Electric Utility Boilers 5. REPORT DATE 6. PERFORMING ORGANIZATION CODE 7. AUTHORS R. K. Srivastava and R. E. Hall 8. PERFORMING ORGANIZATION REPORT NO 9. PERFORMING ORGANIZATION NAME AND ADDRESS See Block 12 10. PROGRAM ELEMENT NO. 11. CONTRACT/GRANT NO. NA (Inhouse) 12. SPONSORING AGENCY NAME AND ADDRESS U. S. EPA, Office of Research and Development Air Pollution Prevention and Control Division Research Triangle Park, North Carolina 27711 13. TYPE OF REPORT AND PERIOD COVERED Published paper; 6-7/01 14. SPONSORING AGENCY CODE EPA/600/13 15 supplementary notes ^ppcd project officer is Ravi h. Srivastava, Mail Drop 65, 919/ 541-3444. For presentation at the All-Russian Thermal Engineering Institute (VT1) 80th Anniversary Conference, Moscow, Russia, October 9-10, 2001. 16.abstract paper discusses the control of nitrogen oxide (NOx) emissions from U.S. coal-fired electric utility boilers. (NOTE: In general. NOx control technologies are categorized as being either primary or secondary control technologies. Primary technologies reduce the amount of NOx produced in the primary combustion zone.. Secondary technologies reduce the NOx present in the flue gas from the primary combustion zone.) Primary technologies in use in the U. S. are low NOx burner (LNB) and overfire air (CFA). They utilize staged combustion to reduce NOx forma- tion in the primary combustion zone. Data reflect that primary technologies, applied on 177 boilers, have resulted in reductions of 33~48%, on average, from 1990 emis- sions levels. In particular, applications of LNB resulted in reductions of > 40%, on average, from 1990 levels. Secondary technologies used on U.S. coal-fired utility boilers include reburning, selective noncatalytic reduction (SNCR), and selective catalytic reduction (SCR). Of these boilers, 14 have used, or will use, reburning as their NOx control technology. The NCx reductions achieved, or expected to be achieved, at these boilers range from 39 to 67%. Of the U.S. coal-fired utility boil- ers, 22 have used, or will use, SNCR. NOx reductions achieved, or projected, at these boilers range from 20 to 62%. Data indicate that 79 boilers will use SCR. 17. KEY WORDS AND DOCUMENT ANALYSIS a. DESCRIPTORS b. IDENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group Pollution Boilers Nitrogen Oxides Flue Gases Emission Coal Combustion Electric Utilities Pollution Control Stationary Sources 13 B 13A 07b 14G 2 ID 21B 18. DISTRIBUTION STATEMENT 19, SECURITY CLASS (This Report) 21. NO. OF PAGES 20. SECURITY CLASS (This Page) 22. PRICE EPA Form 2220-1 (Rev 4-7? ) PREVIOUS EDITION IS OBSOLETE forms/admin/lechrpt.frm 7/8/99 pad ------- |