WO1989009183A1

WO1989009183A1 - SOx/NOx POLLUTION CONTROL COMPOSITION AND METHOD FOR ESP BROWN PLUME CONTROL

Info

Publication number: WO1989009183A1
Application number: PCT/US1989/001168
Authority: WO
Inventors: Richard G. Hooper
Original assignee: Natec Mines Ltd
Priority date: 1988-03-21
Filing date: 1989-03-20
Publication date: 1989-10-05
Also published as: CA1334585C; JPH03501944A; EP0364568A1; EP0364568A4; AU3447089A; US4844915A

Abstract

SOx?/NOx? pollution control reagent composition comprising from 0.1 to 25 % urea in intimate admixture with dry, finely divided Nahcolite, a sodium bicarbonate, removing SOx?/NOx? from the flue by dry injection into a flue gas duct upstream of a collection device to provide above about 0.25 seconds of residence time at temperatures in the range of above about 200F followed by collection. The Nahcolite reacts with the SO2? to form sodium sulfate and also removes NOx? in the form of NO. The urea prevents the conversion of NO to NO2? in the exit flue gases below the 50 parts per million visibility threshold and improves the SOx? removal efficiency and utilization of the Nahcolite to near theoretical.

Description

SO/NOx POLLUTION CONTROL COMPOSITION AND METHOD FOR ESP BROWN PLUME CONTROL

FIELD AND BACKGROUND: The invention relates to a so NO,, pollution control reagent composition and method for control of brown plume coloration associated with the injection of sodium reagents into the inlet ductwork to remove S0₂ from the flue gases of industrial and utility plants that fire (burn) fossil fuels or wastes to generate electricity, steam or heat. More particularly, the method of the invention employs a novel composition of sodium bicarbonate (preferably nahcolite) plus urea to suppress the formation of N0₂ causing plume coloration in exit flue gases as a result of removal of S0₂ from the flue gases. The increasing use of sulfur-containing fuels as energy sources, namely coal and oil, has resulted in pollution of the atmosphere throughout the world. The burning of the fuel produces flue gases containing substantial quantities of S0_X, predominantly S0₂ and S0₃. A wide variety of processes have been proposed for the removal of the SO- from flue gases before they are emitted to the atmosphere. Processes using dry solids, liquids or slurries of sorbents injected in a manner to directly contact the flue gases have proven to be effective for control of S0₂ emissions. Of these, the calcium-based wet scrubber systems have experienced the greatest use in industrial and utility application. This involves finely grinding lime or limestone of suitable composition, forming a water slurry or mixture, and providing apparatus and method for slurry contact with the flue gases. The SO- reacts with the calcium reagent to form calcium sulfate/sulfite which is collected and removed in the form of a generally thixotropic sludge. This sludge is usually difficult to dispose because it does not easily dewater, and heavy metals can leach from the waste sludge containment ponds. Further, abrasion to the slurry pumps and other scrubber handling equipment (due to the hardness of the limestone particles) increases maintenance costs. The operating costs are increased because the evaporation of water in the system cools the flue gases nominally by 200°F. The moist flue gases, containing as they do some residual dissolved SO_x- are also acidic. In addition, where the utility or industrial plant is located in an arid region, the water requirements of a wet pollution control process are expensive and drain an already scarce natural resource.

The subject of flue gas desulfurization by the injection of dry sodium-based reagents has been extensively investigated in laboratory, pilot, and full scale applications since the 1960's. These investigations have provided the technical basis to evaluate the economic aspects of the technology and compare its advantage to the current calcium-based technologies. The injection of dry reagents clearly is the least demanding of capital funds for both new and retrofit applications. The use of familiar hardware such as pulverizers, blowers, and silos permit the easy installation and operation of this simple technology.

Accordingly, the injection of dry reagents into flue gases have become of increasing interest. Principal among these is the use of sodium compounds as a reagent or sorbent. Sodium-based reagents that have been used are commercial sodium bicarbonate (NaHC0₃) , light and dense soda ash (Na₂C0₃) , sodium sesquicarbonate (Na₂C0₃»NaHC0₃'' 2H₂0) , trona (a naturally occurring form of sodium sesquicarbonate) , and Nahcolite (a naturally occurring form, of sodium bicarbonate) . The reagents of greatest effectiveness are commercial sodium bicarbonate and Nahcolite. The sodium bicarbonate or Nahcolite is injected into the flue gas duct of a utility or industrial plant, and reacts with S0₂ in the gas stream to produce Na₂SO_A which is collected in a baghouse or on plates of an ESP.

However, I have observed that the use of sodium can result in production of a reddish-brown plume coloration in stack gases downstream of the particulate control device. The use of the sodium bicarbonate not only removes S0₂, but also removes some NO_x (NO and N0₂) . While the precise mechanism is not known at this time, it is presently thought that some step within the overall sulfation reaction (reaction of sodium reagent with SO_x) initiates the oxidation of NO to N0₂. It is the presence of the N0₂ in the exiting flue gases which is the source of the plume coloration.

Accordingly, this promising sodium reagent S0_X control process has a serious disadvantage, in that its use creates an N0₂ plume in'the process of its removal of S0_X pollution from flue gases. Accordingly, there is a great need for a method by which the N0₂ produced by the use of sodium reagents can also be removed or suppressed, while not sacrificing S0_X removal. Nearly all previous investigations have linked the injection of sodium based reagents with the use of a fabric filter for particulate control. However, as this invention shows, the injection of dry sodium reagents is also effective even when an electrostatic precipitator (ESP) is used for particulate and S0_x/N0_x control.

SUMMARY OF THE INVENTION: The method of this invention comprises the introduction of a novel S0_x/N0_x reagent composition, preferably in a finely divided dry form (but also in solution or slurry form) , into the flue gas of a utility or industrial plant in the range of from about .1 to 1.5 the Normalized Stoichiometric Ratio of the SO_x to be removed. The novel SO„/NO_x reagent composition of this invention comprises a dry finely divided mixture of a sodium reagent, preferably sodium bicarbonate, or Nahcolite (either or both herein termed "bicarb"), with urea in the range of from about .1 to 25% by weight urea in the total composition, preferably about 5 to 15%. This bicarb/urea composition is preferably finely ground uniform admixture of urea in bicarb, or a uniform admixture of unground urea and finely ground bicarb which is injected evenly throughout the gas stream ahead of the ESP. The reaction products collected by the ESP can be disposed in the same manner as sodium reagents used without the use of urea.

Surprisingly, I have discovered that sodium reagents (preferably bicarb) not only remove S0₂ but also remove some N0_X (NO and N0₂) in direct proportion to the amount of S0₂ removed. The ratio of

removed ranges from about 1:6 to 1:3 with 1:4.5 about average. In the process there is also some conversion of nitrogen oxide (NO) to nitrogen dioxide (N0₂) .

In the process, there is some conversion of nitrogen oxide (NO) to nitrogen dioxide (N0₂) . Some, but not all of the N0₂ reacts with the carbonates and/or sulfates produced by the injection of the reagent. The remaining unreacted N0₂ exits the particulate collection device and is responsible for producing a reddish-brown plume coloration. Surprisingly, the addition of urea to the reagent reduces or eliminates the unreacted N0₂ from the exit gas stream and thereby prevents the unwanted plume coloration.

I have also discovered, quite unexpectedly, that not only does urea reduce NO- in the exit gas stream, but also the efficiency of removal of S0_X by the sodium reagent increases by the use of urea. The percent utilization of bicarb can be increased by 10% or more, and on the average the increase is 25%. Conversely, to remove a given percentage of S0_X, the Normalized Stoichiometric Ratio (NSR) can be reduced to near theoretical, while the N0₂ removal is increased some 50%. Thus, for example, for 65% S0_X removal, the NSR drops from 1.2 to about .7 while N0₂ removal increases from 16% to 32% without exceeding the visible plume threshold.

I believe that the reagent (sorbent) composition of this invention could be applied to a wide variety of dry, wet or slurry SO^NO,, pollution control processes employing various types of injection and collection techniques, such as ESPs, baghouses, spray dryers, cyclones, fluidized bed injection or collection systems, scrubbers, counter-current^*or co-current contactors and the like. I also expect the reagent composition of this invention to give equivalent results using any of those devices to treat a wide range of flue gases types, compositions and conditions, e.g., incinerators, paper plants, boilers, furnaces, coke ovens, refineries, smelters, steam plants, utilities and all variety of industrial plants. While I prefer the dry powdered urea to be introduced simultaneously with the sodium reagent as a substantially uniform mixture, it can be separately introduced into the flue gas stream at or near the location of the sodium-based reagent. For example, a solution of urea in water can be sprayed into the flue gases ahead of, coordinate with (in staggered array) , or just downstream of the dry sodium reagent injection point. The urea solution may be sprayed onto the bicarb crystals and let dry, or the thus-coated bicarb crystals can be injected damp into the flue gases. DRAWINGS: The invention is further described in connection with the drawings in which:

Fig. 1 is a graph showing percent S0₂ removal vs NSR for sodium bicarbonate, with the straight line being theoretical for the reaction with S0₂ and the curve being actual results from injection of sodium bicarbonate ahead of an ESP.

Fig. 2 is a graph of average reagent utilization vs NSR based on actual results of the tests shown in Fig. 1 (sodium bicarbonate in an ESP) .

Fig. 3 is a graph of NO removal vs. S0₂ removal showing the range (cone) of linear relationships within which NO removal increases as S0₂ removal increases by use of sodium bicarbonate as a dry reagent injected ahead of an ESP.

Fig. 4 is a graph of the N0₂ emissions at the ESP outlet vs. the S0₂ removed due to the sodium reagent injection. The circles are the results obtained with the addition of urea, the triangles are the results obtained without urea addition. In this example, urea (in combination with sodium bicarbonate) was injected at about 8% urea, i.e., about .085 lbs. urea/lb of bicarb. There was no plume coloration due to the presence of N0₂ when the sodium bicarbonate/urea composition of this invention was injected.

Fig. 5 is a graph showing the removal efficiency of S0₂ and NO_x using urea in accordance with this invention expressed in terms of percent removal vs NSR. Included in the figure is an illustration of the results from a full scale application of the injection of the Nahcolite form of sodium bicarbonate without the addition of urea as compared to the increase in S0₂ and N0_X removal with the use of the composition of this invention, sodium bicarbonate (Nahcolite) plus urea. DETAILED DESCRIPTION OF THE BEST MODE: While the terms sodium bicarbonate and Nahcolite are used interchangeably throughout in case of generic descriptive context, the terms are used specifically in the Examples, and the preferred reagent (or sorbent) is sodium bicarbonate. The term "reagent" means the sodium reagent alone, while "reagent composition" means reagent plus urea (also referred to as bicarb/urea) .

DRY INJECTION PROCESS DESCRIPTION: While I do not wish to be bound by theory, I presently believe the reaction of the oxides of sulfur (S0₂, S0₃) in the flue gas occurs with sodium carbonate (Na₂C0₃) , the product of the thermal decomposition of Nahcolite or sodium bicarbonate (NaHC0₃) . Therefore, when sodium bicarbonate/Nahcolite (herein bicarb) is injected, the first step necessary for SO- removal is the decomposition into sodium carbonate:

(1) 2NaHC0₃ •► Na₂C0₃ + H₂0 + C0₂ The rate of the decomposition of the bicarb controls the distance downstream of the injection location where the SO- reaction begins, and, generally speaking, the temperature of the particular flue gas at the injection location determines the rate of decomposition. Other factors influencing decomposition rate include coal type, fly ash composition, and C0₂, NO and humidity in the flue gas. Under the particular conditions of the tests reported here, injection of bicarb into flue gas at temperatures greater than 275°F will assure that some sodium carbonate will be available for reaction within

0.4 seconds; this amounts to about 20 feet at typical gas velocity in the ductwork. The - time/temperature relationship that the reagent experiences is essential to the efficiency of the S0₂ removal and the utilization of the reagent. Injection into higher flue gas temperatures (275°-700°F) improves S0₂ removal and reagent utilization; injection into flue gas below 250°F can significantly reduce reagent utilization and slow thermal decomposition. Once the decomposition has proceeded sufficiently to provide sites of sodium carbonate, the S0₂ then rapidly reacts according to the following overall reaction:

(2) Na₂C0₃ + S0₂ + 1/2 0₂ >^► Na₂S0₄ + C0₂

Previous efforts relating to this technology were almost entirely focused on the injection of sodium reagents upstream of a baghouse. This invention demonstrates the full-scale (over 100 MW) application of the injection of bicarb upstream of a ESP, and documents the unexpected improvement in the reduction of SO„/NO_x emissions by use of urea in combination with bicarb, preferably a composition of bicarb/urea.

TESTWORK: The reagent composition and methods of this invention were tested on a coal-fired power plant having an output in excess of 100 MW. For comparison tests, the boiler flue gases were divided into two equal streams so that simultaneous comparison of treated vs. untreated flue gases could be made. The flue gas streams were then recombined and emitted up a common stack.

In the tests of this invention, dry sodium sorbent (sodium bicarbonate as the reagent) injection was employed for S0₂ control. The dry sodium bicarbonate reagent was pulverized and then pneumatically conveyed to the flue gas ductwork between the air heater and particulate control device (ESP) . The reagents were introduced into the ductwork through an injection system comprising simple injection nozzles (pipes) arranged in a manner to distribute the dry reagent particles evenly throughout the gas stream. The reaction products are then collected for disposal along with the fly ash.

The sodium bicarbonate reagent was pulverized to the desired particle size, and supplied into the feeder hopper. The feed screw from an "Acrison" feeder delivered the sodium bicarbonate reagent into an air lock feed system that was connected to a positive displacement blower. Reagent was conveyed pneumatically to the ductwork through 3-inch piping to 2-inch nozzles for injection into the flue gas. The injection nozzles were oriented pointing upstream to insure good mixing through gas flow turbulence.

The reagent composition injection system was capable of variable feed rate up to rates permitting injecting the reagent composition of this invention at a rate necessary to- remove up to 80-100% of the S0₂.

Flue gas temperature and volume of gas flow were fixed by the plant operations. Flue gas temperature at the injection location ranged from 300-325°F. The location of the reagent injection and the flue gas velocity determine the distance and time within which the reagent must decompose and react with S0₂ before it is removed from the gas stream and collected by the ESP. It is estimated that the residence time of the reagent in the hot flue gas stream before collection by the ESP was about 1.5 seconds.

The coal used during the tests was a lignite having from about 6500-8900 BTU/lb. Coal was sampled from the burner feed pipes in order to follow the trends of fluctuations in ash, sulfur, and moisture content of the fuel. The ash content ranged from nominally 23 percent to greater than 30 percent; averaging about 27%. Particulate mass measurements indicate that about 20 percent of the total ash in the fuel was carried to the ESPs. The percent sulfur (dry basis) in the fuel varied from 1.3% to 1.5% without a particular pattern or trend 10 throughout the test period. The as-fired fuel moisture content ranged from 1.4% to 3.0% without evidence of a trend. The spent reagent was predominantly Na₂S0₄, plus in appropriate cases, some unknown urea reaction produc .

REAGENTS: Chemical analysis of the Nahcolite reagent provided the basis to calculate the stoichiometric ratio (NSR) and percent utilization. The sodium bicarbonate was provided by Industrial Resources, Inc. of Lakewood, Colorado from its deposit of Nahcolite in the Picance Creek Basin Deposit in Northwestern Colorado. The sodium bicarbonate was in the form of a crystalline fine powder having a mean mass diameter of 18 um, and a purity in of nearly 98% NaHC0₃. The urea was commercially available agricultural grade urea in coarse crystal flakes. The sodium bicarbonate may also be any other commercial grade of sodium bicarbonate, such as Church and Dwight 3DF (a fine powder) . The sodium bicarbonate was placed in a large hopper, fed by a screw feeder into a rotary air lock. The urea was placed in a small hopper, and fed at the rate of .085 lb. urea/lb. of the sodium bicarbonate by an Acrison screw feeder into the same rotary air lock into which the sodium bicarbonate was being fed. The two components were mixed together in the airlock, and then the resulting reagent composition (sodium sorbent, in this case sodium bicarbonate, plus urea) was fed into the feed pipe and conveyed pneumatically by fresh ambient air supplied by the air blower to the feed nozzles in the flue gas duct.

The reagent composition can be prepared, for example, by intermixing the sodium bicarbonate and urea in the appropriate proportions to produce for example, standard grades of 1%, 5%, 10%, 15% and above, or produce custom percentage mixtures, then bagged and shipped to the point of utility or industrial plant use. As an alternative to intermixing the crystalline urea with the sodium bicarbonate, a water solution of the urea can be sprayed onto the crystalline sodium bicarbonate on a conveyor belt, dried, and stored in a bin at the use site, or may be dried and bagged at the sodium bicarbonate production facility for shipment to the users. Still other modes of mixing the composition components and their simultaneous or sequential injection to the flue gas is within the scope of this invention.

Particle size analysis of the sodium bicarbonate was performed on a selected sample collected during the demonstration. The sample was diluted in an organic liquid and particle size distribution was measured between 3 um and 175 um (micro meter) . The mass median diameter (MMD) of the reagent agglomerates (as received) was «18 um. A sample was then ultrasonically dispersed and the single particle mass median diameter (MMD) was ~7 um.

Feed rate of the reagent was controlled both automatically and manually by the controls of the loss- in-weight (Acrison) feeders. Blowers provided a constant supply of ambient air that conveyed the reagent to the ductwork. S0₂ removal as a function of the amount of reagent or reagent composition injected (presented as Normalized Stoichiometric Ratio, NSR) is a measure of the effectiveness of the reagent and its distribution in the gas stream. The amount of sodium bicarbonate injected was controlled to produce several levels of S0₂ removal that provide the basis to evaluate this specific installation. Typical injection rate was .085 lbs. urea/lb sodium bicarbonate (in the 5-10% range) , with the reagent composition mixture being fed at the rate of 2800-2900 lb/hr. Emissions were monitored at the air heater inlet and the ESP outlet to assess the impact of the injection of the dry sodium bicarbonate on the collection efficiency of the ESP.

GAS MEASUREMENTS: Sulfur dioxide (S0₂) , nitrogen oxide

(NO) , total oxides of nitrogen (N0_X) , Oxygen (0₂) , carbon dioxide (C0₂) , and nitrogen dioxide (N0₂) were continuously sampled, monitored and recorded on analog strip charts. Calibration was conducted several times daily with standard bottled calibration gases. S0₂, NO or NO_x, and O_z were measured for selected periods at one location and then switched to the other location for selected periods. The continuous monitoring system used the analytical instruments as described below:

Flue Gas Component Measurement Method S0₂ and N0₂ Ultraviolet, photometric

Nitrogen Oxides Chemiluminescence

Oxygen Electrochemical

Carbon Dioxide NDIR, photometric

PARTICULATE MEASUREMENTS: Mass and particle size measurements were taken throughout the project to establish emission characteristics entering and exiting the ESP handling flue gases during treatment. Mass measurements were also taken at the inlet and outlet of the ESP during non-treatment for comparison.

Sampling methods for velocity, moisture and particle mass concentration followed federal procedures of Methods 1,2,3,4, and 17. Particle size distributions were obtained using standard Anderson cascade impactors. GASEOUS EMISSIONS: Using the monitors described above, samples of flue gas were withdrawn through heated sample lines and analyzed for concentrations of S0₂, NO, NO_x, N0₂, C0₂, and 0₂. A manifold of three stainless steel, sintered filters was located across the ductwork at both the air heater inlet and ESP outlet. The three sample locations in each of the ducts provided some averaging of possible stratification of gas constituents. A comparison of the concentrations of SO- and NO from each of the three locations in the inlet and outlet ducts indicated no stratification from the boiler. Further, the manifolded sample provided much the same results as a single sample taken only from the middle of the ductwork. This was true for both the inlet and outlet locations.

EXAMPLES 1-4; UTILIZATION OF NaHCOs ALONE FOR S0₂ REMOVAL: The injection of sodium bicarbonate to remove S0₂ in flue gas from fossil fuel fired boilers is a well documented technology. However, this test example is unique because the injection of the Nahcolite reagent is upstream of an ESP, not a fabric filter. In the prior art it was. perceived that the time necessary for the decomposition of the reagent and subsequent reaction with S0₂ was too great for significant S0₂ removal to occur in the ductwork preceding a particulate control device. Therefore, it was perceived that the S0₂ primarily reacted with the decomposed reagent as it passed through the cake on the fabric filter surface. Measurement of S0₂ removal prior to the particulate control device often substantiated this belief. This, then, curtailed any effort to employ dry sorbent injection technology with an ESP.

Recently, pilot plant efforts to combine dry injection of sodium reagent with an ESP met with some but erratic and inconclusive results in terms of S0₂ removal and reagent utilization. In contrast, the examples of this invention expand these efforts to a full scale demonstration, and broaden the horizon for the technology by evidence of the combination of S0₂ and N0_X control that is achieved without production of brown lume.

Fig. 1 shows the results in graph form of four tests (Examples 1-4) of Nahcolite injected upstream of an ESP in terms of S0₂ Removal vs. Normalized Stoichiometric Ratio (NSR) . The NSR is a measure of the amount of reagent injected relative to the amount theoretically required. For example, an NSR of 1.0 would mean that enough material was injected to theoretically yield 100 percent removal of the S0₂ in the inlet flue gas; an NSR of 0.5 would theoretically yield 50 percent S0₂ removal. The solid straight line indicates the theoretical use of the sodium bicarbonate. The curved line drawn through the data points indicates the actual S0₂ removal obtained in Examples 1-4, each example being one of the four data points in each curve.

These results are the first of their kind for a power plant ESP at full scale (over 100 MW) , and are extremely encouraging for the future of this process and sodium bicarbonate reagent. At the lower injection rates, the S0₂ removal is near the theoretical limit. However, this falls off dramatically as more reagent is injected.

Percent utilization is a measure of the effective use of the reagent and is determined by dividing the percent SO- removal by the NSR. The percent utilization curve obtained in Examples 1-4 is shown in Fig. 2. The low utilization of reagent at the higher NSR values is believed to result from a somewhat uneven distribution of the reagent in the gas stream. This can be visualized as putting too much reagent in part of the gas stream (removing all the S0₂ in that space) and too little in the remaining space. There is too little time (too short ductwork) for the reagent to distribute evenly throughout the gas stream and react before it is collected by the ESP. Although utilization is expected to decrease with increasing NSR in all practical applications, the utilization as illustrated in Fig. 2 could be improved with a redesign of the injection nozzles.

EXAMPLES 5-44; REMOVAL OF THE OXIDES OF NITROGEN (NO and N0₂) ; PLUME FORMATION: I have discovered that nitrogen oxide (NO) was also removed upon injection of sodium bicarbonate. While simultaneous removal of small percentage of NO has been occasionally observed (to a limited degree) by several investigators of this technology, the data from the tests here demonstrate a correlation that leads me to conclude that NO removal is a direct function of S0₂ removal. Examples 5-44, shown graphically as the triangular data points in Fig. 3, illustrate this correlation. Although some data scatter is evident, it is clear that the change in NO concentration is directly a function of the amount of S0₂ removed. In Fig. 4 "Without Additive" means Nahcolite alone, while "With Additive" means urea plus Nahcolite injected as a composition mixture.

I also observed that the effect of NO removal with S0₂ removal is associated with a coloration of the plume emitting from the stack. This plume coloration is often described as a "brown plume"; N0₂ is a reddish-brown gas easily visible to the human eye. Most gases emitting from the stack are not visible to the human eye.

The N0₂ concentrations in the flue gas without the injection of reagent were measured to be about 20 ppm at stack conditions. As the reagent injection resulted in S0₂ removal, N0₂ concentrations at the ESP outlet increased as shown in the test results plotted in Fig. 4. All concentrations on this graph are corrected to zero percent (0%) 0₂ levels, putting them on an equal basis. The horizontal line estimates the minimum N0₂ concentration visible to the human eye at about 45-50 ppm (corrected to 0% 0₂) .

Usually flue gases from utility plants average 20- 25 ppm N0₂ and 400-800 ppm NO. Fig. 4 shows the increase in N0₂ in the outlet (conversion of NO to N0₂) during the SO_x removal with Nahcolite via the sulfation reaction (2) above. Thus, for example, at 1000 ppm S0₂ removed by Nahcolite, approximately 225 ppm NO (see Fig. 3) , is converted to N0₂. That plus the baseline value of 25 ppm totals 250 ppm N0₂ in the flue gas. Fig. 4 shows the outlet N0₂ is about 110 ppm. Thus a net 140 ppm N0₂ reacts with the sodium reagent (or possibly some other reaction product) . The level of N0₂ visible to the human eye is dependent upon several site specific parameters. These parameters include stack dimensions, meteorology factors, position of the viewer, and possibly the particulate concentration and size distribution exiting the stack. Plume coloration is not the same as stack opacity as measured by plant monitors. The relationship between visible N0₂ concentrations in the flue gas and stack opacity measured by conventional methods has not been established. As shown in Fig. 4, without injection of reagent

(i.e. zero percent S0₂ removal at the left end of the graph) , outlet N0₂ concentration is just under the minimum necessary to cause plume coloration. However, with reagent injection and subsequent S0₂ removal of above about 300 ppm S0₂, N0₂ concentration rises significantly and causes plume coloration.

EXAMPLES 45-46 - EFFECT OF UREA PLUS SODIUM BICARBONATE:

Results from testing a mixture of urea and sodium bicarbonate as the reagent are illustrated in Fig. 4 in the darkened circles lying just above 1200 ppm S0₂ and the open baseline circle. Two effects should be noted: (1) , the N0₂ concentration at the ESP outlet was considerably below the visible threshold; and (2) , S0₂ removal increased with the additive. This is shown by comparing the position of the open circle referred to by the legend "Baseline Before Additive" (a prior example) , to the position of the two darkened circles (Examples 45 and 46). When using .085 lb. urea/lb. sodium bicarbonate, for example, the same 0.5 SR increased removal 50% (from 800 to 1200 ppm S0₂) , while the N0₂ dropped from 90 ppm to 20 ppm. The most unexpected result is that not only was no plume coloration visible, but the S0₂ removed efficiency/utilization improved significantly.

The result of including the urea additive with the Nahcolite is shown in Fig. 5. Again, the straight line is the theoretical maximum S0₂ removal and the curved lines are the results obtained without the urea additive to the sodium sorbent. The labeled triangle is the S0₂ removal obtained while injecting the reagent composition of this invention, and the labeled circle is the NO_x removal. These examples show unexpectedly improved S0₂ removal with simultaneous N0_X removal, and without the undesirable "brown plume".

PARTICULATE EMISSIONS: Once the dry reagent has been injected into the flue gas it must be captured by the particulate emission control device. This aspect of the technology is certainly no less important that its FGD capability because particulate emission regulations also must be obeyed. Injecting more material (reagent and/or reagent composition) into the ESP would be expected to result in some loss of efficiency or increase in total particulates emitted.

The Percent Efficiency of the ESP can be calculated from mass measurements using EPA Method 17. The amount of fly ash measured at the air heater inlet ranged from 1624 pounds/hr to 8310 pounds/hr. (average = 3634 pounds/hr.) Outlet mass emissions from the ESP immediately increased following the initial reagent/reagent composition injection period but did not increase with greater reagent injection rates throughout the test period. During periods of reagent injection, the ESP emissions were 64-100 percent greater than emissions from the North ESP. The reagent injection rates during these periods were 1450-6500 pounds/hr; this is 50-100 percent of the amount of fly ash at the injection location. Outlet particulate size distributions with and without injection are not significantly different. Particles slightly less than 1 um (Log Dp = 0) in diameter were some of the least efficiently collected. This is very typical of ESP performance. Particles in the size range of ∞7 um were also less efficiently collected. This effect is likely a result of the large increase of particles at the ESP inlet due to reagent injection (MMD «7 um) . However, opacity excursions above permitted limits were not exceeded, indicating the process of this invention does not pose problems with regard to particulates.

During periods of reagent injection (without urea additive) when S0₂ removal was greater than 25 percent,

N0₂ concentrations developed in the flue gas caused a visible plume coloration. The group of test examples using the reagent composition of this invention (the same Na bicarbonate reagent with urea additive) , resulted in simultaneous removal of S0₂ and N0_X from the flue gas without the visual plume coloration. The continuous monitoring instruments substantiated that N0₂ concentration in the outlet flue gas was below visual levels.

It should be understood that various modifications within the scope of this invention can be made by one of ordinary skill in the art without departing from the spirit thereof. For example, the detailed description above illustrates the invention by way of example, not by way of limitation of the principles of the invention. That description will clearly enable one skilled in the art to make and use the invention, and describes several embodiments, adaptations, variations, alternatives and uses of the invention, including what I presently believe is the best mode of carrying out the invention. I therefore wish my invention to be defined by the scope of the appended clams as broadly as the prior art will permit, and in view of the specification if need be.

Claims

CLAIMS : 1. An SO_/NO_x pollution control reagent composition comprising from .1 to 25% urea in admixture with a sodium-based sorbent selected from the group consisting essentially of sodium bicarbonate, Nahcolite, soda ash, trona, sodium sesquicarbonate, and combinations thereof. 2. An SO_/NO_x pollution control reagent composition as in Claim 1 wherein said reagent is in the form of a solution. 3. An SO_ NO_x pollution control reagent composition as in Claim 1 wherein said composition is in a finely divided state, and said urea is in a dry, crystalline form dispersed substantially uniformly throughout said composition. 4. An SO_x/NO_x pollution control reagent composition as in Claim 1 wherein said urea is in the form of a solution which is applied to said sodium sorbent to form said admixture, and thereafter said admixture is dried to form said reagent composition. 5. An SO„/NO_x pollution control reagent composition as in Claim 1 wherein said reagent is in the form of a slurry. 6. An SO_x/NO_x pollution control reagent composition as in Claims 2, 3, 4 or 5 wherein said reagent is sodium bicarbonate or Nahcolite. 7. An SO_x/NOχ air pollution control process comprising the steps of: a) introducing a sodium-based sorbent into a stream of flue gas containing SO_x and NO_x, said flue gas having a temperature above about 200°F; b) introducing urea into said flue gases in an amount ranging from about .1 to about 25% by weight of said sodium-based sorbent which is selected from the group consisting essentially of sodium bicarbonate, Nahcolite, soda ash, trona, sodium sesquicarbonate, and combinations thereof; c) maintaining said sorbent and said urea in contact with said flue gas for a time sufficient to react said sorbent with some of said S0_X and N0_X thereby reducing the concentration of said S0_X and N0_X in said flue gas; and d) said urea, as compared to use of said sorbent without said urea, increasing both the percent S0_X removal and the percent utilization, expressed as percent S0_X removal divided by the normalized stoichiometric ratio, and increasing the percent N0_X removal by said sorbent while suppressing conversion of NO to N0₂ by said sorbent to below about the 50 ppm visible threshold to prevent brown plume formation. 8. An SO_x/NO_x air pollution control process as in Claim 7 which includes the added step of collecting said reacted sorbent in an ESP. 9. An SO„/NO_x air pollution control process as in Claim 8 wherein said reagent is sodium bicarbonate or Nahcolite. 10. An SO^/NO_j, air pollution control process as in Claim 9 wherein said urea is injected simultaneously with said reagent. 11. An SO_x/NO_x air pollution control process as in Claim 9 wherein: a) said reagent introduction step includes feeding said reagent in a dry finely divided condition from a first hopper at a metered rate in relation to a predetermined amount of SO_x reduction desired; b) said urea introduction step includes feeding said urea in a dry crystalline form from a second hopper at a rate within said .1 to 25% by weight range in relation to a predetermined S0_X removal percent utilization increase desired; and c) said urea and reagent are mixed together to form a reagent composition admixture just prior to introduction of said admixture into said flue, gas duct. 12. An SO_/NO_x air pollution, control process as in Claim 11 wherein said mixture is pneumatically conveyed by air at substantially ambient temperature to said flue gas duct and injected therein. 13. An SO_x/NO_jj air pollution control process as in Claim 8 wherein: a) said reagent introduction step includes feeding said reagent in a dry finely divided condition from a first hopper at a metered rate in relation to a predetermined amount of SO_x reduction desired; b) said urea introduction step includes feeding said urea in a dry crystalline form from a second hopper at a rate within said .1 to 25% by weight range in relation to a predetermined SO_x removal percent utilization increase desired; and c) said urea and reagent are mixed together to form a reagent composition admixture just prior to introduction of said admixture into said flue gas duct. 14. An SO_x/NO_x air pollution control process as in Claim 13 wherein said mixture is pneumatically conveyed by air at substantially ambient temperature to said flue gas duct and injected therein. 15. An SO„/NO_x air pollution control process as in Claims 7 or 9 wherein said urea and said sorbent are provided as a premixed reagent composition of said sodium-based sorbent with said urea, and said sorbent and urea injection steps are combined into a single step of injecting said premixed reagent composition into said flue gas.