SONIC INJECTION FURNACE

Abstract

A low-NOx end-fired furnace for melting glass equipped with an overhead burner includes an inlet duct for oxidizer, including 15% to 30% of oxygen, in its upstream wall, a duct for receiving the combustion flue gases in its upstream wall, and a sonic injection system including at least one injector for injecting a jet of a gas at a speed at least equal to 80% of the speed of sound, referred to as a sonic injector, opening into the upstream wall or opening into the duct for receiving the combustion flue gases, the sonic injector injecting its gas counter-current to the stream of the combustion flue gases that are heading toward the duct for receiving the combustion flue gases.

Claims

1. An end-fired furnace for melting glass comprising: an overhead burner; an oxidizer inlet duct in an upstream wall of the end-fired furnace, the inlet duct being in fluid communication with a source of oxidizer comprising 15% to 30% of oxygen; a combustion flue gas duct constructed and arranged to, in operation, receive combustion flue gases, the duct opening into the upstream wall of the end-fired furnace; and a sonic injection system comprising at least one sonic injector opening into the upstream wall or opening into the combustion flue gas duct and constructed and arranged to inject a jet of a gas at a speed at least equal to 80% of the speed of sound, said sonic injector being further configured and arranged to, in operation, inject its gas counter-current to a stream of the combustion flue gases that are heading toward the combustion flue gas duct.

2. The furnace as claimed in claim 1, wherein the sonic injector is configured and arranged to inject its gas at a speed that is at least 95% of the speed of sound.

3. The furnace as claimed in claim 2, wherein an impulse component of the sonic injection system perpendicular to the wall comprising the combustion flue gas duct is greater than 5 newtons.

4. The furnace as claimed in claim 2, wherein an impulse component of the sonic injection system perpendicular to the wall comprising the combustion flue gas duct is greater than 10 newtons.

5. The furnace as claimed in claim 1, wherein a cross-sectional area of the oxidizer inlet duct in the upstream wall is within the range extending from 0.5 to 3 square meters and wherein a cross-sectional area of the combustion flue gas duct is within the range extending from 0.5 to 3 square meters.

6. The furnace as claimed in claim 1, wherein a discharge area of the sonic injection system is within the range extending from 0.2 to 4 square centimeters.

7. The furnace as claimed in claim 1, wherein an impulse component of the sonic injection system perpendicular to the wall comprising the combustion flue gas duct is greater than 5 newtons.

8. The furnace as claimed in claim 1, wherein an impulse component of the sonic injection system perpendicular to the wall comprising the combustion flue gas duct is greater than 10 newtons.

9. The furnace as claimed in claim 1, wherein each sonic injector of the sonic injection system opens into the combustion flue gas duct or into the upstream wall at a point closer to the combustion flue gas duct than to the oxidizer inlet duct.

10. The furnace as claimed in claim 1, wherein each sonic injector of the sonic injection system opens at a position that is less than 1 meter from an edge of the combustion flue gas duct.

11. The furnace as claimed in claim 10, wherein each sonic injector opens at a position that is less than 0.5 meter from the edge of the combustion flue gas duct.

12. The furnace as claimed in claim 1, wherein the jet of gas comprises air.

13. The furnace as claimed in claim 1, wherein the sonic injection system is in fluid communication with a source of air, and wherein the jet of the gas comprises a jet of air.

14. The furnace as claimed in claim 1, wherein the sonic injection system comprises a plurality of sonic injectors.

15. The furnace as claimed in claim 1, wherein the sonic injection system is controllable to deliver the jet of gas of from 0.2% to 5% of a volume of oxidizer introduced by the oxidizer inlet duct.

Description

[0032] The figures are not to scale.

[0033] FIG. 1 represents an end-fired furnace according to the invention seen from the side in cross section.

[0034] FIGS. 2a-e show various sonic injection systems that can be combined with a duct for receiving the flue gases in a furnace according to the invention.

[0035] FIG. 3 represents, as a cross-sectional side view, a duct for receiving the flue gases equipped with a sonic injector opening into the same upstream wall of the furnace according to the invention.

[0036] FIG. 4 is a representation of the gas flows in an end-fired furnace according to various configurations.

[0037] FIG. 1 represents an end-fired furnace 1 seen from the side in cross section, the molten glass 2 flowing from left to right. The cross section is made through the duct 3 for receiving the flue gases, which are sent into the regenerator 4 containing refractory elements 5 that the flue gases will reheat. After phase inversion, this opening will be used as burner oxidizer inlet duct for a flame, said oxidizer then being reheated by the refractory elements 5. As represented, the duct is in the flue gas receiving phase. A compressed gas injector 6 opens into the upstream wall 33 which is also the wall into which the duct for receiving the flue gases opens. The jet 8 of compressed gas is sent toward the stream of flue gases (depicted by the dotted lines 9), heading toward the duct for receiving the flue gases, while being counter-current thereto.

[0038] FIGS. 2a-e show various sonic injection systems that may be combined with a duct for receiving the flue gases 20. These various configurations are illustrated by the examples. The sonic injection system may comprise 3 injectors 21 and may be located just above the duct for receiving the flue gases 20 (FIG. 2a) or at a certain distance above the flue gases (FIG. 2b). FIG. 2b schematically shows a sonic injection system 62 that includes three injectors 22. The injectors of the sonic injection system may be located beneath the duct for receiving the flue gases (FIG. 2d). In FIG. 2e, a single injector 26 is used and is located beneath the duct for receiving the flue gases 20. Here the injector is slightly to the left, to the side of the longitudinal axis of the furnace. In FIG. 2c), sonic injectors 23,24 have been combined in the sonic injection system above 23 and below 24 the duct for receiving the flue gases. The term d denotes what is understood by the distance between the injector and the edge of the duct. Considering the small diameter of the sonic injector, the distance d is taken starting from its axis.

[0039] FIG. 3 represents a duct 30 for receiving the flue gases, seen from the side, equipped with a sonic injector (31) opening into the same upstream wall 33 of the furnace. The sonic injector is at the distance d from the edge of the duct. The sonic injector delivers its gas with an impulse represented by the vector 32, which may be broken down into a component 34 perpendicular to the upstream wall 33 and another component 35 parallel to the wall.

[0040] FIG. 4 is a representation of the gas flows in a reference end-fired furnace without additional injection of gas through a wall (configuration a), in an end-fired furnace with sonic injection according to the invention (configuration b), and in an end-fired furnace with sonic injection in a side wall (configuration c). The configuration b) of FIG. 4 shows the overhead burner 50 and the inlet duct for oxidizer 51.

Examples 1 to 10

[0041] The tests were carried out with an end-fired furnace comprising two burners operating in inversion, having a power of 13.3 megawatts and the oxidizer of which was air. Each air inlet duct had an area of 1.55 m.sup.2 (2200 mm wide and 800 mm high). The furnace was supplied with soda-lime type batch material, including 95% by weight of cullet. It operated with an output of 330 tonnes per day. The furnace had a surface area of 94 m.sup.2. The temperature of the glass at the outlet of the furnace was around 1300 C. The crown was at a temperature of around 1600 C.

[0042] One or more sonic compressed air injectors were placed in the vicinity of the duct for receiving the combustion flue gases in order to form sonic injection systems. These injectors had a convergent end. The gas injected was at 25 C. Table 1 gives the various operating conditions and also the results in terms of content of NOx in the flue gases. Four possible sonic injector positions were tested:

[0043] just above the flue gas receiver: the injector is exactly at the edge of the duct for receiving the flue gases with a downward angle of 20 with respect to the horizontal;

[0044] slightly above the duct for receiving the flue gases: the injector is 400 mm above the upper edge of the duct for receiving the flue gases with a downward angle of 20 with respect to the horizontal;

[0045] below the duct for receiving the flue gases: the injector is 250 mm below the duct for receiving the flue gases with an upward angle of 5 with respect to the horizontal.

[0046] All the sonic compressed air injectors above the duct for receiving the flue gases had an internal diameter of 5 mm. The sonic compressed air injectors underneath the duct for receiving the flue gases had an internal diameter of 6 mm.

[0047] Indicated in table 1 are:

[0048] the relative pressure: this is the pressure of the reservoir supplying the sonic injector;

[0049] the flow rate: this is the total flow rate of compressed air (sum of the flow rates of all the sonic injectors);

[0050] the injection speed: this is the speed of the air at the outlet of the sonic compressed air injector;

[0051] sonic impulse: this is the sum of the impulses of the sonic injectors (in the case of example 2, no injection is sonic but the impulse of the gas has nevertheless been depicted at 50% of the speed of sound in the sonic impulse column for the sake of simplification);

[0052] NOx: this is the concentration in mg/Nm.sup.3 standardized at 8% oxygen in dry flue gas;

[0053] variation: this is the variation in NOx relative to a reference (without sonic injection of compressed air).

[0054] The variation results from examples 2 to 7 are given relative to example 1. The variation results of examples 9 to 10 are given relative to example 8.

Examples 11 and 12

[0055] The tests were carried out with an end-fired furnace comprising two burners operating in inversion, having a power of 11 megawatts and the oxidizer of which was air. Each burner air inlet duct (or duct for receiving the flue gases, depending on the inversion phase) had an area of 2 m.sup.2 (2300 mm wide and 960 mm high). The furnace was supplied with soda-lime type batch material, including 60% by weight of cullet. It operated with an output of 250 tonnes per day. The furnace had a surface area of 85 m.sup.2. The temperature of the glass at the outlet of the furnace was around 1300 C. The crown was at a temperature of around 1600 C. The sonic injection system comprised only a single sonic compressed air injector with a convergent end. The latter was placed 300 mm below the duct for receiving the flue gases and at 650 mm from the lower corner of the duct for receiving the flue gases closest to the longitudinal axis of the furnace. The sonic injector injected its gas with an upward angle of 5 with respect to the horizontal and had an internal diameter of 8 mm in diameter. The gas injected was at 25 C.

[0056] Table 1 gives the various operating conditions and also the results in terms of content of NOx in the flue gases.

TABLE-US-00001 TABLE 1 Number of Number of Number of % NOx injectors injectors injectors Relative Flow Speed Injection Sonic [mg/Nm.sup.3 just 400 mm 250 mm pressure rate of speed impulse @ 8% Ex. No. above above below [bar] [Nm.sup.3/h] sound [m/s] [N] O.sub.2] Variation 1 601 (comp) 2 3 0.2 35 50 174 2 607 1% (comp) 3 3 0.7 65 84 289 7 559 7% 4 3 1.7 104 100 346 14 546 9% 5 3 4.2 200 100 346 33 405 33% 6 0 3 4.2 200 100 346 33 390 35% 7 2 4.2 200 100 346 33 397 34% 8 778 (comp) 9 3 4.2 200 100 346 33 494 37% 10 3 2 4.2 400 100 346 65 354 54% 11 1 3 132 100 346 21 457 12 1 3.5 148 100 346 24 381

Examples 13 to 15

[0057] Numerical simulations were carried out of the flow of the combustion flue gases in an end-fired furnace in operation, in the following configurations:

[0058] a) reference: no additional injection of gas (see FIG. 4a));

[0059] b) additional injection of 181 Nm.sup.3/h of sonic gas in the upstream wall in accordance with the present invention;

[0060] c) additional injection of 150 Nm.sup.3/h of sonic gas in the side wall (see FIG. 4b, the additional injection being symbolized by a bold arrow); the injector formed an angle of 60 with the side wall of the upstream wall side and was at a distance from the upstream wall equal to 23% of the total length of the side wall; this example is given by way of comparison and is not according to the present invention;

[0061] d) additional injection of 150 Nm.sup.3/h of sonic gas in the upstream wall (see FIG. 4c, the additional injection being symbolized by a bold arrow) in accordance with the present invention.

[0062] The end-fired furnace comprises two burners operating in inversion, having a power of 11 megawatts, and the oxidizer of which is air. Each burner air inlet duct (or duct for receiving the flue gases, depending on the inversion phase) has an area of 1.55 m.sup.2 (2200 mm wide and 800 mm high). The furnace has a surface area of 94 m.sup.2. The temperature of the glass at the outlet of the furnace was around 1300 C. The sonic gas injected was at 25 C.

[0063] FIG. 4 shows the effect of these injections on the gas flows in the laboratory volume of the furnace. In order to visualize these flows, the speed vectors have been shown in a vertical plane passing through the middle of the air inlet duct. This representation makes it possible to visualize the vertical recirculation of the gases. Large differences are observed depending on the configurations.

[0064] It is seen that the sonic injection according to the invention leads to the broadest recirculation in the laboratory volume. Furthermore, the results on the NOx expressed relative to a reference without sonic injection demonstrate the superiority of the injection according to invention, as shown by table 2. The variation column gives the concentration of NOx relative to the reference configuration without additional gas injection (FIG. 4a). The injection according to invention leads to a reduction, respectively, of 15% and 20% in NOx. The injection of sonic gas in the side wall does not produce a substantial reduction in NOx.

TABLE-US-00002 TABLE 2 Number of injectors Number of 250 mm Flow Injection Sonic injectors in below the duct for rate % Speed speed impulse Variation Ex. No. side wall receiving the flue gases [Nm.sup.3/h] of sound [m/s] [N] (%) 13 2 181 100 346 27 19% (FIG. 4b) 14 (comp) 1 150 94 326 17 1% (FIG. 4c) 15 1 150 94 326 17 15%