Most major industrialized urban areas in the U.S. are unable to meet the National Ambient Air Quality Standards (NAAQS) for ozone. Atmospheric studies have shown that ozone formation is the result of a complex set of chemical reactions involving volatile organic compounds (VOCs) and nitrogen oxides (NOx). Those studies indicate that many urban areas with VOC/NOx ratios greater tan 15:1 can reduce ambient ozone levels only by reducing NOx emissions. Many states, therefore, are implementing NOx control regulations for combustion devices in order to achieve compliance with the NAAQS ozone standard.
This article discusses the characterization of NOx emissions from industrial combustion devices. It then provides guidance on how to evaluate the applicable NOx control technologies and select an appropriate control method.
Characterizing Emissions
Most industrial combustion devices have not been tested to establish their baseline NOx emission levels. Rather, the NOx emissions from these units have been simply estimated using various factors. In light of recent regulations, however, it is mandatory that the NOx emissions from affected units now be known with certainty. This will establish each unit’s present compliance status and allow definition of fee applicable control technologies for those units that will require modification to achieve compliance.
It is, therefore, important to test each combustion device to verify its NOx emissions characteristics. The testing process should be streamlined to provide timely and necessary information for making decisions regarding the applicability of NOx control technologies.
The basic approach is to select one device from a class of units (that is, of same design and size) for characterization testing (NOx, CO2, and 02). Testing is conducted at three load points that represent the normal operating range of the unit, with excess oxygen variation testing conducted at each load point. Figure 1 illustrates the typical characterization test results. The remaining units in the class are tested at only one load point, at or near full load.
The operational data obtained during testing, in conjunction with the NOx and CO data, are used to define the compliance status of each unit, as well as the applicable NOx control technologies for those devices that must be modified. In most instances, this approach will allow multiple units to be tested in one day and provide the necessary operational data the engineer needs to properly evaluate the potential NOx control technologies.
Fundamental Concepts
Reasonably available control technology (RACT) standards for NOx emissions are defined in terms of an emission limit, such as 0.2 lb NOx/MMBtu, rather than mandating Specific NOx control technologies. Depending on the fuel fired and the design of the combustion device, a myriad of control technologies may be viable options. Before selecting RACT for a particular combustion device, it is necessary to understand how NOx emissions are formed so that the appropriate control strategy may be formulated.
NOx emissions formed during the combustion process are a function of the fuel composition, the operating mode, and the basic design of the boiler and combustion equipment. Each of these parameters can play a significant role in the final level of NOx emissions.
NOx formation is attributed to three distinct mechanisms:
1. Thermal NOx Formation;
2. Prompt (i.e.. rapidly forming) NO formation; and
3. Fuel NOx formation.
Each of these mechanisms is driven by three basic parameters – temperature of combustion, time above threshold temperatures in an oxidizing or reducing atmosphere, and turbulence during initial combustion.
Thermal NOx formation in gas-, oil-. and coal-fired devices results from thermal fixation of atmospheric nitrogen in the combustion air. Early investigations of NOx formation were based upon kinetic analyses for gaseous fuel combustion. These analyses by Zeldovich yielded an Arrhenius-type equation showing the relative importance of time, temperature, and oxygen and nitrogen concentrations on NOx formation in a pre-mixed flame (that is, the reactants are thoroughly mixed before combustion).
While thermal NOx formation in combustion devices cannot actually be determined using the Zeldovich relationship, it does illustrate the importance of the major factors that Influence thermal NOx formation, and that NOx formation increases exponentially with combustion temperatures above 2.800°F.
Experimentally measured NOx formation rates near the flame zone are higher than those predicted by the Zeldovich relationship. This rapidly forming NO is referred to as prompt NO. The discrepancy between the predicted and measured thermal NOx values is attributed to the simplifying assumptions used in the derivation of the Zeldovich equation, such as the equilibrium assumption that O = ½ 02. Near the hydrocarbon-air flame zone, the concentration of the formed radicals, such as O and OH, can exceed the equilibrium values, which enhances the rate of NOx formation. However, the importance of prompt NO in NOx emissions is negligible in comparison to thermal and fuel NOx.
When nitrogen is introduced with the fuel, completely different characteristics are observed. The NOx formed from the reaction of the fuel nitrogen with oxygen is termed fuel NOx. The most common form of fuel nitrogen is organically bound nitrogen present in liquid or solid fuels where individual nitrogen atoms are bonded to carbon or other atoms. These bonds break more easily than the diatomic N2 bonds so that fuel NOx formation rates can be much higher than those of thermal NOx. In addition, any nitrogen compounds (e.g., ammonia) introduced into the furnace react in much the same way.
Fuel NOx is much more sensitive to stoichiometry than to thermal conditions. For this reason, traditional thermal treatments, such as flue gas recirculation and water injection, do not effectively reduce NOx emissions from liquid and solid fuel combustion.
NOx emissions can be controlled either during the combustion process or after combustion is complete. Combustion control technologies rely on air or fuel staging techniques to take advantage of the kinetics of NOx formation or introducing inerts that inhibit the formation of NOx during combustion, or both. Post-combustion control technologies rely on introducing reactants in specified temperature regimes that destroy NOx either with or without the use of catalyst to promote the destruction.
Conbustion Control
The simplest of the combustion control technologies is low-excess-air operation–that is, reducing the excess air level to the point of some constraint, such as carbon monoxide formation, flame length, flame stability, and so on. Unfortunately, low-excess-air operation has proven to yield only moderate NOx reductions, if any.
Three technologies that have demonstrated their effectiveness in controlling NOx emissions are off-stoichiometric combustion. low-NOx burners, and combustion temperature reduction. The first two are applicable to all fuels, while the third is applicable only to natural gas and low-nitro-gen-content fuel oils.
Off-stoichiometric, or staged, combustion is achieved by modifying the primary combustion zone stoichiometry – that is, the air/fuel ratio. This may be accomplished operationally or by equipment modifications.
An operational technique known us burners-out-of-service (BOOS) involves terminating the fuel flow to selected burners while leaving the air registers open. The remaining burners operate fuel-rich, thereby limiting oxygen availability, lowering peak flame temperatures, and reducing NOx formation. The unreacted products combine with the air from the terminated-fuel burners to complete burnout before exiting the furnace. Figure 2 illustrates the effectiveness of this technique applied to electric utility boilers. Staged combustion can also be achieved by installing air-only ports, referred to as overfire air (OFA) ports, above the burner zone. redirecting a portion of the air from the burners to the OFA ports. A variation of this concept, lance air, consists of installing air tubes around the periphery of each burner to supply staged air.
BOOS, overfire air, and lance air achieve similar results. These techniques are generally applicable only to larger, multiple-burner, combustion devices.
Low-NOx burners are designed to achieve the staging effect internally. The air and fuel flow fields are partitioned and controlled to achieve the desired air/fuel ratio, which reduces NOx formation and results in complete burnout within the furnace. Low-NOx burners are applicable lo practically all combustion devices with circular burner designs.
Combustion temperature reduction is effective at reducing thermal N0x but not fuel NOx. One way to reduce the combustion temperature is to introduce a diluent. Flue gas recirculation (FGR) is one such technique.
FGR recirculates a portion of the flue gas leaving the combustion process back into the windbox. The recirculated flue gas, usually on the order of 10-20% of the combustion air provides sufficient dilution to decrease NOx emission. Figure 3 correlates the degree of emission reduction with the amount of flue gas recirculated.
On gas-fired units, emissions arc reduced well beyond the levels normally achievable with staged combustion control. In fact, FGR is probably the most effective and least troublesome system for NOx reduction for gas-fired combustors.
An advantage of FGR is that it can be used with most other combustion control methods. Many industrial low-NOx burner systems on the market today incorporate induced FGR. In these designs, a duct is installed between the stack and forced-draft inlet (suction). Flue gas products are recirculated through the forced-draft fan, thus eliminating the need for a separate fan.
Water injection is another method that works on the principle of combustion dilution, very similar to FGR. In addition to dilution, it reduces the combustion air temperature by absorbing the latent heat of vaporization of the water before the combustion air reaches the primary combustion zone.
Few full-scale retrofit or test trials of water injection have been performed. Until recently, water injection has not been used as a primary NOx control method on any combustion devices other than gas turbines because of the efficiency penalty resulting from the absorption of usable energy to evaporate the water. In some cases, water injection represents a viable option to consider when moderate NOx reductions are required to achieve compliance.
Reduction of the air preheat temperature is another viable technique for culling NOx emissions. This lowers peak flame temperatures, thereby reducing NOx formation. The efficiency penalty, however, may be substantial. A rule of thumb is a 1% efficiency loss for each 40º F reduction in preheat. In some cases this may be offset by adding or enlarging the existing economizer.
Post-Combustion Control
There are two technologies for controlling NOx emissions after formation in the combustion process – selective catalytic reduction (SCR) and selective noncatalytic reduction (SNCR). Both of these processes have seen very limited application in the U.S. for external combustion devices. In selective catalytic reduction, a gas mixture of ammonia with a carrier gas (typically compressed air) is injected upstream of a catalytic reactor operating at temperatures between 450º F and 750º F. NOx control efficiencies are typically in the 70-90% percent range, depending on the type of catalyst, the amount of ammonia injected, the initial NOx level, and the age of the catalyst.
The retrofit of SCR on existing combustion devices can be complex and costly. Apart from the ammonia storage, preparation, and control monitoring requirements, significant modifications to the convective pass ducts may be necessary.
In selective noncatalytic reduction, ammonia- or urea-based reagents are injected into the furnace exit region, where the flue gas is in the range of 1,700-2,000º F. The efficiency of this process depends on the temperature of the gas, the reagent mixing with the gas, the residence time within the temperature window, and the amount of reagent injected relative to the concentration of NOx present. The optimum gas temperature for die reaction is about 1,750°F; deviations from this temperature result in a lower NOx reduction efficiency. Application of SNCR, therefore, must be carefully assessed, as its effectiveness is very dependent on combustion device design and operation.
Technology Selection
As noted previously, selection of applicable NOx control technologies depends on a number of fuel, design, and operational factors. After identifying the applicable control technologies, an economic evaluation must be conducted to rank the technologies according to their cost effectiveness. Management can then select the optimum NOx control technology for the specific unit.
It should be noted that the efficiencies of NOx control technologies are not additive, but rather multiplicative. Efficiencies for existing combustion devices have been demonstrated in terms of percent reduction from baseline emissions level. This must be taken into account when considering combinations of technology.
Consider, for example, the following hypothetical case. Assume a baseline NOx emissions level of 100 ppmv and control technology efficiencies as follows: low-excess-air operation (LEA), 10%; low-NOx burners (LNB), 40%; and flue gas recirculation (FGR). 60%. The three controls are installed in the progressive order of LEA-LNB-FGR.
It should also he noted that combining same-principle technologies (for example, two types of staged combustion) would not provide a further significant NOx reduction than either of the combination, since they operate on the same principle.
It must be emphasized that virtually all of the available control technologies have the potential for adversely affecting the performance and/or operation of the unit. The operation data obtained during the NOx characterization testing, therefore, must be carefully evaluated in light of such potential impacts before selecting applicable control technologies. Operational limitations such as flame envelope, furnace pressure, forced-draft fan capacity, and the like must he identified for each potential technology and their corresponding impacts quantified. (Reference (4), for example, discusses these items, in detail.)
As anyone familiar with combustion processes knows, one technology does not fit all. Careful consideration must he used to select the appropriate, compatible control technology or technologies to ensure compliance at least cost with minimal impact on performance, operation, and capacity.