Method and Apparatus for Reducing Nitrogen Dioxide Emissions in a Dry Sodium Scrubbing Process Using Humidification

The invention relates generally to flue gas desulfurization and, more particularly, to a method of reducing nitrogen dioxide (NO.sub.2) emissions from a flue gas duct using a dry sodium scrubbing process without significantly reducing sulfur dioxide (SO.sub.2) and nitrogen oxides (NO.sub.x) removal efficiencies.

To control acid rain the major air pollutants that must be removed from a flue gas effluent produced when burning fossil fuel are sulfur dioxide (SO.sub.2) and nitrogen oxides (NO.sub.x). In air pollution terminology, NO.sub.x refers to two gaseous oxides of nitrogen, namely, nitrogen dioxide (NO.sub.2) and nitric oxide (NO). In the past, dry sodium scrubbing has been one of the methods employed to remove SO.sub.2 and NO.sub.x. In this type of process, a dry sodium compound, such as sodium bicarbonate (NaHCO.sub.3) or sodium sesquicarbonate (Na.sub.2 CO.sub.3 NaHCO.sub.3 2H.sub.2 O), is injected into the flue gas duct that carries flue gas from the combustion chamber of a boiler or the like to the flue gas stack for venting the effluent to atmosphere. Though the sodium injection is generally effective in removing sulfur dioxide (SO.sub.2) and nitrogen oxides (NO.sub.x), during this process a portion of the nitric oxide (NO) is converted to nitrogen dioxide (NO.sub.2), which usually results in a plume emanating from the gas stack having an undesirable brown color. The chemistry of NO.sub.x removal by, for example, sodium bicarbonate is given below.

Because the production of NO.sub.2 in reaction 2 outpaces its removal via reaction 3, the result is a net generation of NO.sub.2 with sodium bicarbonate injection.

Various additives have been used to try to eliminate the brown plume, and some success in controlling nitrogen dioxide emissions has been achieved using ammonia (NH.sub.3) and urea. However, use of these additives can create new air and waste disposal problems, including high NH.sub.3 slip and the presence of soluble NH.sub.3 compounds in the fly ash.

Hence, the invention is directed to providing a dry sodium scrubbing process that reduces nitrogen dioxide emissions to a level below that which causes the creation of a brown plume, without significantly reducing sulfur dioxide and nitrogen oxides removal efficiencies while, at the same time, avoiding the above-mentioned drawbacks and disadvantages.

The invention solves this problem by providing a dry sodium scrubbing process in which NO.sub.2 emissions are reduced without significantly reducing SO.sub.2 and NO.sub.x removal efficiencies by humidifying and cooling the flue gas. More specifically, a sodium additive is injected into a flue gas duct carrying flue gas from a fossil fuel-burning combustor to a particulate collection device to remove sulfur dioxide and nitrogen oxides from the flue gas and the flue gas is humidified by introducing water into the flue gas duct at a point upstream of the particulate collection device.

The step of humidifying the flue gas increases the moisture content of the flue gas to a value that produces an NO.sub.2 concentration downstream of the particulate collection device below the level at which a brown plume becomes visible upon venting the flue gas to atmosphere. A general rule that can be used to determine the brown plume threshold is to reduce the NO.sub.2 emissions in ppm to a value that is less than 200 divided by the gas stack diameter in feet.

The sodium additive is injected into the flue gas duct at a point upstream of the particulate collection device. The flue gas is humidified by introducing water into the duct, most preferably, at a point disposed upstream of the sodium additive injection point. In some cases, it may be desirable to introduce water downstream of the sodium additive injection point and upstream of the particulate collection device.

The sodium additive may be sodium bicarbonate, sodium sesquicarbonate, or a mixture thereof. A further step of injecting a sulfur adsorbing sorbent into the convention bank of the combustor may be optionally employed. An apparatus for reducing nitrogen dioxide (NO.sub.2) emissions without significantly reducing SO.sub.2 and NO.sub.x removal efficiencies in accordance with the invention is also disclosed.

Tests were conducted in accordance with the principles of the invention and excellent results were obtained when the flue gas was humidified to increase the moisture content of the flue gas to a value in the range of 10-20% H.sub.2 O, the sodium additive was injected at a temperature of approximately 240.degree. F.-350.degree. F., and the temperature of the flue gas in the particulate collection device was approximately 180.degree. F.-280.degree. F. However, the preferred value for the sodium additive injection temperature is approximately 250.degree. F., while the preferred moisture content is approximately 15% H.sub.2 O and the preferred temperature of the flue gas in the particulate collection device is approximately 190.degree. F.

FIG. 1 is a flow diagram that schematically illustrates the flue gas desulfurization system of the present invention.

FIG. 2 is a graph that illustrates the effect of humidification prior to sodium bicarbonate injection on SO.sub.2 and NO.sub.x removal and NO.sub.2 emissions in accordance with the principles of the invention.

FIG. 3 is a graph that illustrates the effect of humidification prior to sodium sesquicarbonate injection on SO.sub.2 and NO.sub.x removals and NO.sub.2 emissions in accordance with the principles of the invention.

FIG. 4 is a graph that illustrates the effect of flue gas moisture content on NO and NO.sub.x removals and NO.sub.2 emissions with a baghouse temperature of 260.degree. F.

FIG. 5 is a graph that illustrates the effect of flue gas moisture content on NO and NO.sub.x removals and NO.sub.2 emissions with a baghouse temperature of 190.degree. F.

FIG. 6 is a graph that illustrates the effect of flue gas moisture on the increase in NO.sub.2 emissions.

FIG. 7 is a strip chart graph showing the effect over time of bicarbonate injection with a baghouse temperature of 250.degree. F. and a flue moisture content of 9% on O.sub.2, NO.sub.x, SO.sub.2, and NO.sub.2 levels.

FIG. 8 is a strip chart graph showing the effect over time of bicarbonate injection with a baghouse temperature of 250.degree. F. and a flue moisture content of 15% on O.sub.2, NO.sub.x, SO.sub.2, and NO.sub.2 levels.

FIG. 9 is a strip chart graph showing the effect over time of bicarbonate injection with a baghouse temperature of 200.degree. F. and a flue moisture content of 9% on O.sub.2, NO.sub.x, SO.sub.2, and NO.sub.2 levels.

FIG. 10 is a strip chart graph showing the effect over time of bicarbonate injection with a baghouse temperature of 200.degree. F. and a flue moisture content of 15% on O.sub.2, NO.sub.x, SO.sub.2, and NO.sub.2 levels.

FIG. 1 schematically illustrates in process flow diagram form a fossil fuel-burning facility 10 equipped with flue gas desulfurization apparatus constructed according to the principles of the invention. The furnace or combustor 20 of the facility is fed with a pulverized solid fuel, such as coal in the case of the illustrated coal fuel-fired boiler, and with air to burn the fuel and generate heat in a manner known in the art. From the combustor 20, the gases produced during this process are conducted to a convection section 21 of the boiler where they pass through convective banks 22 and 23 for heat extraction. The first portion 11 of a flue gas duct is connected to the outlet of convection section 21 for conducting the flue gases to a gas stack for venting. A heat exchanger 30 is arranged between the first flue gas duct section 11 and a second flue gas duct section 12 for reducing the temperature of the flue gas to a value, typically 350.degree. F. or less, which is the safe operating temperature of the baghouse. In a utility furnace, this temperature is achieved by use of an air preheater. The outlet of duct section 12 is connected to a baghouse 40. Baghouse 40 contains fabric filters for separating particles made during the combustion process, such as fly ash, from the flue gas before it is conducted to the gas stack for venting. The filtered particles themselves may be recycled or otherwise disposed of as schematically illustrated in FIG. 1. Of course, other types of particulate collection devices, such as electrostatic precipitators or the like, may be used for solids removal.

Apparatus for injecting a sodium additive, such as sodium bicarbonate, sodium sesquicarbonate, or a mixture thereof, is schematically illustrated at 34. Sodium injection apparatus 34 introduces the sodium additive into flue gas duct section 12 via a line 35, which is disposed at a position upstream of the baghouse inlet and downstream of the heat exchanger outlet. Any type of injection apparatus known in the art may be used to introduce the sodium additive into the gas duct. For example, injection can be accomplished directly by a compressed air-driven eductor.

Water injection apparatus schematically illustrated at 31 humidifies and cools the flue gas by injecting water into flue gas duct section 12 via a line 33, which is disposed at a position upstream of the baghouse inlet and, preferably, upstream of the sodium additive injection point as well. Optionally, depending upon the degree of humidification desired, the water injection apparatus 31 may conduct some or all of the water to be injected via suitable valving to a line 32, which is connected to duct section 11 at a position upstream of the heat exchanger inlet. In either event, the water injection apparatus includes appropriate valves, spray nozzles, and the like for conducting the water into the flue gas duct.

A sorbent, such as limestone, may be optionally injected via one or more sorbent injection apparatuses 24a-24e directly into the convection section 21 of the boiler to adsorb sulfur dioxide. In general, the efficiency of the sorbent injection at this point decreases when low SO.sub.2 concentrations are reached. Thus, especially when high SO.sub.2 removal efficiencies are desired, this sorbent injection would not obviate the sodium additive step carried out by injection apparatus 34.

Gas analysis apparatus 50 may be provided throughout the system, as illustrated at points upstream and downstream of the heat exchanger and in the gas outlet section of the baghouse. Additionally, if one or more of the sorbent injection apparatuses 24 are not provided, gas sampling apparatuses may be provided at these points as well.

The invention will now be illustrated by the examples shown on the following tables, which contain test data collected from a pilot plant facility constructed according to the principles of the invention. The pilot scale facility was generally configured as illustrated in FIG. 1 and used coal as fuel, which was burned in the combustor at a rate of 0.25.times.10.sup.6 Btu/h. No sorbent was injected into the boiler during any of these tests. The width a of the combustor 20 was 18 inches and the width b of the convection section 21 was 6 inches, as was the width c of the duct 11. The temperature of the flue gas at the locations of sorbent injection/gas analysis apparatuses 24a-24e was approximately 1200.degree. F., 1100.degree. F., 1000.degree. F., 900.degree. F. and 800.degree. F., respectively. The heat exchanger 30 cooled the flue gas to a temperature of 350.degree. F. at the outlet of the heat exchanger.

During the tests discussed below, the characteristics of the flue gases were measured at the three positions shown at 50 in FIG. 1, i.e., at the outlet of the convective section 21, at the baghouse inlet, and the baghouse outlet. The amounts of SO.sub.2 and NO.sub.x measured were based on the respective differences between the SO.sub.2, NO, NO.sub.2 and NO.sub.x levels at the convective section outlet without any injection and those levels measured at the baghouse outlet after sodium injection. All measurements shown in the data tables were normalized to 3% O.sub.2. Actual O.sub.2 values were approximately 4% at the convective section outlet, 4% at the baghouse inlet and 8% at the baghouse outlet.

The flue gas flow from the furnace was approximately 4.2 pound/min. and the residence time between the sodium injection point and the baghouse inlet was about 1.0 seconds during which time the temperature remained steady. The average temperature inside the baghouse was measured by inserting a thermocouple through the wall of the baghouse between bags and should have accurately represented the temperature of the sodium additive deposited on the bags. The gas residence times and cooling rates were comparable to those expected for full-scale boilers.

The sodium compounds injected were sodium bicarbonate, NaHCO.sub.3 and sodium sesquicarbonate, Na.sub.2 CO.sub.3 NaHCO.sub.3 2H.sub.2 O. The mass mean diameter particle sizes were 12.9 microns for the bicarbonate and 12.2 microns for the sesquicarbonate. For all tests, the initial SO.sub.2 concentration was 2600 ppm, the initial NO concentration was 350 ppm, and the initial moisture content was 7-8% H.sub.2 O. The baghouse temperature was lower than the sodium injection temperature, due to heat loss through the walls and air leakage. The injected sodium compounds were entrained in flue gas at the injection temperature for about one second, after which they entered the baghouse, where they remained until removed from the bags. The reactions between the sodium compounds and SO.sub.2 and NO.sub.x therefore took place initially at the injection temperature (one second) and subsequently at the baghouse temperature (minutes).

Table 1 below shows the results of sodium additive tests conducted with sodium bicarbonate as the additive both with and without humidification. The NO.sub.x removals and NO.sub.2 emissions have been normalized by the change in SO.sub.2 concentration because the NO.sub.x removal and NO.sub.2 formation appear to be driven by the reaction of SO.sub.2 with the sodium additive. This method of normalizing the data allows for easy estimation of NO.sub.x reductions and NO.sub.2 emissions for a given SO.sub.2 reduction.

As shown below, Table 1 sets out for each test, the Na.sub.2 /S ratio, which is the amount of sodium in the additive Compared to the amount of sulfur in the flue gas, as well as amounts of sulfur dioxide removed in ppm, the percentage of Na.sub.2 utilized, the amount of nitric oxide (NO) removed in ppm, the amount of nitrogen dioxide (NO.sub.2) generated in ppm, and the amount of nitrogen oxides (NO.sub.x) removed in ppm. All of these values were measured in the manner discussed above. In addition, the normalized ratios of the respective changes in nitric oxide (NO), nitrogen oxides (NO.sub.x), and nitrogen dioxide (NO.sub.2) relative to the change in sulfur dioxide (SO.sub.2) are set out. Na Inj. temp. refers to the injection temperature in degrees Fahrenheit of the sodium additive, which also is the temperature of the flue gas at the baghouse inlet. The average temperature at the center of the baghouse is set forth in the last column.

The test results from Table 1 clearly show that humidifying the flue gas stream prior to sodium injection significantly reduces the amount of NO converted to NO.sub.2. Comparison of the NO.sub.2 emissions in Test Nos. B1-2, B-4, and B-15 with the NO.sub.2 emissions in Tests Nos. B-14 and B-9 shows a decrease in NO.sub.2 emissions to values of 18 and 13 ppm, respectively, from values of 41, 43 and 58, respectively. The amount of NO.sub.2 remaining after the humidification was less than 20 parts per million, which generally is below the NO.sub.2 level at which a brown plume is created.

FIG. 2 graphically illustrates the effect of humidification prior to sodium bicarbonate injection on SO.sub.2 and NO.sub.x removals and NO.sub.2 emissions from the data obtained from Table 1. The vertical axis is labelled .vertline..DELTA..vertline. ppm, which is the absolute value of the amount of the particular gas species present before the sodium additive injection minus the amount present after injection, in parts per million. In the case of SO.sub.2 and NO.sub.x, this is the amount of particulate matter removed while in the case of NO.sub.2, this is the amount of NO.sub.2 emissions generated. The average temperature in the baghouse is plotted on the lower horizontal axis. The upper horizontal axis roughly indicates the relationship between temperature and the moisture content of the flue gas as the gas leaving the heat exchanger 30 at approximately 350.degree. F. is cooled by the humidification step. Baghouse temperatures above about 250.F are considered to result in non-humidified flue gas while those below about 200.degree. F. are considered humidified. The graph of FIG. 2 also illustrates that NO.sub.x efficiency is fairly constant regardless of humidification and that the humidification step actually improved the SO.sub.2 removal efficiency at baghouse temperatures below 250.degree. F.

Table 2 below and FIG. 3 illustrate the results of similar tests conducted using sodium sesquicarbonate as the sodium additive. In these tests, humidification decreased NO.sub.2 emissions from values ranging between 54-84 ppm to 14-17 ppm, while NO.sub.x removal efficiency remained approximately constant, and SO.sub.2 removal efficiency decreased somewhat.

From these two series of initial tests, the most optimal results occurred when the sodium injection temperature was about 250.degree. F. and the flue gas was humidified and cooled to approximately 190.degree. F. Subsequent tests were conducted to more precisely determine the effect of the flue gas moisture concentration on the NO.sub.x /Sodium chemistry. These tests, discussed below, revealed that excellent results are obtained with a moisture flue gas content in the range of 10-20%, with about 15% being preferred, a sodium additive injection temperature of 240.degree. F.-350.degree. F., with about 250.degree. F. being preferred, and a baghouse temperature of 180.degree. F-280.degree. F., with about 190.degree. F. being preferred.

The test conditions for these humidification experiments are shown in Table 3 below. These tests were conducted at two injection temperatures of 250.degree. F. and 350.degree. F. corresponding to baghouse temperatures of approximately 190.degree. F. and 260.degree. F., respectively, using both bicarbonate (test nos. marked "B") and sesquicarbonate (test nos. marked "S"). Tests were conducted under three different moisture conditions as indicated in the % H.sub.2 O column in Table 3:

(i) when burning coal at a 9% flue gas moisture level;

(ii) when burning coal and using pre-sodium injection humidification to produce either a 15% or 13.4% moisture content; and

(iii) when burning natural gas to produce a 17% H.sub.2 O flue gas concentration. Also as indicated in Table 3, the temperatures of the flue gas were measured at the inlet to the baghouse (T.degree.F Inlet), in the center of the baghouse (T.degree.F Baghouse), and at the outlet of the baghouse (T.degree.F Baghouse). The percent of oxygen in the flue gas in the baghouse was also measured (% O.sub.2 Baghouse). Table 3 also gives the baghouse inlet and baghouse outlet SO.sub.2 and NO.sub.x concentrations normalized to 3% O.sub.2.

Table 4 below summarizes the results of the humidification tests set forth in Table 3. Table 4 gives the percentage SO.sub.2, NO.sub.x, and NO reductions and the change in NO, NO.sub.x, and NO.sub.2 concentrations from the injection point to the baghouse exit, which have been normalized by the change in SO.sub.2 for the reasons previously stated.

The results of the tests summarized in Tables 3 and 4 are graphically shown in FIGS. 4-10. FIGS. 4 and 5 illustrate the effect of changes in flue gas moisture content on NO and NO.sub.2 removals and NO.sub.2 increase (normalized by the change in SO.sub.2 concentration) for 260.degree. F. and 190.degree. F. baghouse temperatures, respectively. The normalized change in concentration of the gas species is plotted on the vertical axes, while the moisture content of the gas is plotted on the horizontal axes. These graphs show that the conversion of NO to NO.sub.2 and NO.sub.x removal are both affected by the flue gas moisture level. As shown in FIG. 5, with a 190.degree. F. baghouse temperature and a 15% moisture concentration, the increase in NO.sub.2 emissions was only 3 ppm per 1000 ppm SO.sub.2 removed.

The NO.sub.2 emission results for these tests are shown in FIG. 6. The amount of increase of NO.sub.2 emissions in ppm is shown on the vertical axis, while the moisture content of the flue gas is shown on the horizontal axis. The tests results show a definite relationship between both the baghouse temperature and the flue gas moisture level. As shown in FIG. 6, with use of either sodium bicarbonate or sodium sesquicarbonate as the additive, less than a 10 ppm NO.sub.2 increase was measured at a .DELTA.NO.sub.2 /.DELTA.SO.sub.2 =1.0 when the baghouse temperature was about 200.degree. F. and the moisture content was 15% or greater.

FIGS. 7 through 10 show the results of the humidification tests that used bicarbonate injection as strip chart records for average baghouse temperatures of 200.degree. F. and 250.degree. F. and flue gas moisture levels of 9% and 15%. A strip chart record shows the change in concentration of the particular gas species over time, as the vertical and horizontal axes of these figures indicate. FIG. 7 shows the rate of removal of SO.sub.2 and NO.sub.x for a baghouse temperature of 250.degree. F. and a 9% moisture content. The SO.sub.2 reached an equilibrium concentration approximately 6 minutes after the sodium injection is begun, while the NO.sub.2 remained low until about 4 minutes after sodium injection began.

FIG. 8 shows the results for a baghouse temperature of 250.degree. F. and a H.sub.2 O content of 15%. At t=30 min, the additional humidification step was stopped thereby reducing the moisture content to the original 9% and providing a rapid rise in NO.sub.2 emissions. Thus, the effect of the additional humidification step appears to be to maintain the NO.sub.2 concentration lower for a longer time after bicarbonate injection and to reduce the amount of NO.sub.2 formed.

FIGS. 9 and 10 show similar results to those depicted in FIGS. 7 and 8 but for a 200.degree. F. baghouse temperature and moisture contents of 9% and 15%, respectively. As evident from these figures, the effect of H.sub.2 O addition is similar to that found at the higher temperature.