WO1999022032A1

WO1999022032A1 - Inclined heat pipe lance or tuyere with controllable heat extraction

Info

Publication number: WO1999022032A1
Application number: PCT/CA1998/000989
Authority: WO
Inventors: Frank Mucciardi; Ning Jin
Original assignee: Mcgill University
Priority date: 1997-10-24
Filing date: 1998-10-22
Publication date: 1999-05-06
Also published as: AU9617198A

Abstract

A heat pipe lance for conveying reagent materials into a metallurgical bath includes an elongated heat pipe vessel (7) defined by a tubular wall and a coolant liquid in the vessel (7) selected from a class of liquids including liquid sodium and liquid potassium. The heat pipe vessel (7) includes an evaporator section near a leading edge of the lance and a condenser section near the trailing end of the lance. The coolant liquid forms a liquid pool (16) in the evaporator section. A separate conduit (9) passes through at least a portion of the heat pipe vessel (7) at the evaporator section and is open to the leading edge of the lance to form a nozzle (14) for passing materials into the molten bath.

Description

"INCLINED HEAT PIPE LANCE OR TUYERE WITH CONTROLLABLE HEAT EXTRACTION"

Technical Field This invention describes an improved design for a heat pipe lance that is used to convey gases, liquids and/or solids into or onto a molten bath commonly encountered in the smelting and refining of metals and materials. This invention also describes an improved design for a heat pipe tuyere that is used to convey gases, liquids and/or solids into a molten bath. Background Art

Current lancing technology for conveying reagents including gases, liquids and/or solids into and onto melts contained in metallurgical vessels and furnaces is classified in either of two categories. One segment of industry utilizes fluid cooled lances with water being the coolant of choice. Such lances are complicated in design as necessitated by the fact that the coolant must be forced at a high enough velocity over the internal area of the nozzle to extract sufficient heat to allow the lance to maintain stable operation. Another segment of industry has adopted a simpler and safer approach whereby the lance is not cooled by an external coolant. In such cases, the lance is simply a tubular section that is consumed with time. The major drawbacks to this approach are that the positioning of the nozzle varies as the lance is consumed and that the cost of replacement pipe to make up for the consumed length can, with time, be substantial. In either case, the lance can be partially immersed in the melt or elevated above the melt while the injection of reagent is carried out .

Recently, a new lance cooling technology as described in U. S. Patent 5,310,166 was advocated as

SUBSTΓRΠΈ SHEET (RULE 26) being superior to the existing techniques described above. This new lance is described as a heat pipe lance essentially comprising two tubular members with the said tubular members defining a closed annular chamber. A working substance (e.g. sodium) occupies a portion of the annular chamber with the remaining space consisting of an inert gas or simply a vacuum. A central inner tube is used to convey reagent onto the melt . While such a lance using heat pipe technology can function m a non-hostile environment, it is the contention of the inventors that this lance ultimately will fail m hostile environments as a result of a mechanism termed 'burn-out¹.

While lancing constitutes a sizeable segment of injection processing, there is another method by which injection is carried out. This other technique is referred to as tuyere injection. A significant portion of injection technology is based on the use of submerged tuyeres for conveying reagents into melts. In its simplest of forms, a tuyere is a p pe embedded through the refractory lining of the containment vessel. The reagent is conveyed through this pipe and into the melt to carry out the smelting and refining of the melt. In hostile environments, the pipe is consumed as the refractory is eroded; thus, the tip of the tuyere is typically flush with the surrounding refractory. Erosion of the refractory can be reduced if the tuyere can be made to extend beyond the face of the refractory and into the melt. At present, there is no tuyere that is non- consumable which can operate extended beyond the refractory face into a hostile melt.

For the purposes of this specification, when the term "lance" is used, it includes a tuyere.

The object of this invention is to provide a significant improvement to the design of a heat pipe lance so as to achieve sufficient cooling of the inner wall of the evaporator and thus avoid the premature failure of the lance especially in those environments where a loss of cooling ability even over a small patch of area can eventually lead to failure by the perforation of the heat pipe outer wall. Another object of the present invention is to describe an improved tuyere design that will allow the tuyere to operate extended into the melt wherein it will be shown that this new tuyere is a deviant of an inclined heat pipe injection lance.

Those familiar in the art of heat pipes will know that a heat pipe contains, within a hermetically sealed chamber, a material known as the working substance. As heat is applied to the evaporator, the working substance contained therein evaporates and the resultant vapor flows to the cooler portions of the heat pipe where it condenses to a liquid and flows back by a combination of gravity and capillary forces, to the evaporator to repeat the cycle. The evaporator portion of the heat pipe lance resides in the hot furnace environment and may also be partially submerged within the melt while the condenser formed by the outer pipe is situated outside the hot furnace environment. The choices of working substance and materials of construction for the tubular members are dependent on the application of the heat pipe as well as the chemical compatibility of the working substance and the pipe materials. For reasons of safety, the working substance that is used should have a boiling point that is higher than the operating temperature of the pipe. In this way, the working substance chamber of the pipe can be operated at negative gauge pressure and thus minimize the safety hazards in the event of lance failure . Another important component in a heat pipe lance is the wick which is used to ensure relatively uniform coverage of the evaporator surface. Typically, the wick comprises several wraps of fine mesh screen made of the same material as the pipes. One of the more popular groupings of materials pairs sodium as the working substance and 304 stainless steel as the wick and pipe materials.

The prior art of heat pipe lances as detailed m U. S. Patent 5,310,166 stipulates an annular configuration m which heat extracted by the vaporization of working substance m the evaporator is transported to the condenser and is released by the condensation of working substance vapor. While this statement has generally been taken to be adequate m describing the process, it nonetheless requires a certain degree of clarification. Condensation of the vaporized working substance, it will be noted, arises on both the condenser surface area (i.e. the inner wall of the outer pipe) and on the outer wall of the central conduit transporting reagent to the melt. The relative percentages of the quantities condensed on the outer pipe and on the inner pipe are a function of a number of parameters including the physical radial dimensions of the pipes, the mass flowrate of reagent through the central conduit, the thermo-physical properties of the reagent, the thermal conductivity of the pipe materials, the effective heat transfer coefficient between the reagent and the inner pipe, and the effective heat transfer coefficient

(including radiation) between the outer surface of the condenser and the environment. While the practice of having two concentric pipes serving as condensation sites may be acceptable for lances where the incoming heat flux is applied principally around the nozzle, the design is not acceptable when:

(1) the evaporator section is relatively long, and

(2) the temperature of the hot furnace environment is such that any portion of the lance evaporator, not cooled for whatever reason, will fail .

In the prior art of annular heat pipe lances, the entire length of reagent conduit pipe acts as a condensing surface for the vaporized working substance. This is so because m the vast majority of cases, the reagent is introduced into the lancing system at a temperature that is relatively low m comparison to the operating temperature of the lance. Working substance condensed on the inner pipe flows down to the lance nozzle and into the liquid pool of working substance accumulated at the nozzle region. This liquid returning via the central conduit is effective m cooling the nozzle and immediate evaporator region. It cannot, however, cool the upper regions of the evaporator as will be demonstrated m the examples to follow.

If the condenser is configured m such a way that it can dissipate heat, it will condense working substance. The quantity that it condenses together with that condensed on the inner conduit comprises the total condensed and must at steady state equal the total quantity of working substance that is evaporated. With the annular arrangement stipulated m the prior art, the upper section of the evaporator is fed liquid working substance only originating from condensation on the outer pipe while the lower section of the evaporator is fed liquid working substance originating from condensation on the inner conduit . The nozzle and leading end section of the lance will accumulate a quantity of working substance liquid because of the initial imbalance between the rate at which working substance liquid flows down the central conduit and the rate at which working substance is vaporized from the leading section of the lance. A portion of the evaporator located between the upper and lower sections will under conditions of high heat transfer become deficient m working substance with the result that this intermediate region straddling the upper and lower regions operates m 'dry-out' mode. In effect, there is at steady state an excess of liquid m the lower section and a deficiency of liquid m the intermediate section of the evaporator. The result is that this intermediate section of pipe is not cooled by evaporating working substance and hence heats up to a temperature approaching that of the hot furnace environment . Since cooling of the entire lance evaporator body is a necessity to avoid failure by chemical attack, or by high temperature corrosion or simply by fusion (i.e. 'burnout ' ) , then it stands to reason that uncooled portions of the evaporator ultimately will fail . Whereas the prior art called for the conveying of reagent through the central conduit spanning the entire length of the lance, the present invention calls for the conveying of reagent through a jacket which is applied on the outer surface of a portion of the condenser wall. The jacket ideally is located directly above the evaporator. The lowermost segment of the jacket is connected to an opening m the outer wall of the heat pipe. The hole m the outer wall of the heat pipe is fitted with a conduit which carries the reagent to the tip of the lance. Thus, the reagent enters the jacket, is preheated and subsequently leaves the jacket and continues into the evaporator travelling through a central conduit to the nozzle. The jacket is configured and sized so that most of the preheating of the reagent occurs m the jacket. In this way the reagent is preheated by the energy extracted from the condensation of working substance on the inner wall of the outer pipe with only limited condensation on the outer surface of the new inner conduit. Disclosure of the Invention

While the importance of preheating has not been discussed, it must be noted that preheating of the reagent is, in many metallurgical applications, a desirable feature which can have substantial financial benefits. Moreover, it promotes greater energy utilization efficiencies and lower entropy generation. Thus, the degree of preheating should, under normal circumstances, be optimized. As the incentives for preheating are great, the existing heat pipe technology utilizing an annular configuration (U. S. Patent 5,310,166) capitalizes on the preheating of reagent, but in so doing, .subjects portions of the lance to insufficient cooling and the resultant creation of 'hot Spots ' .

The preheating jacket is an essential feature of the present invention which overcomes the limitation of 'hot spot' formation in high temperature metallurgical applications by removing the central conduit from the condenser section and instead having the reagent flow through an outer jacket. In this way, working substance is forced to condense on the inner surface of the outer wall of the condenser and to flow down this wall and cool the evaporator surface. In the prior art, working substance condensed on the central conduit and flowed down the conduit to the nozzle. While this arrangement cooled the nozzle effectively, it did not cool the evaporator wall sufficiently.

In addition to the preheating jacket, one may incorporate another jacket which resides directly above the preheating jacket. This jacket, to be termed the cooling jacket, can be used to pass a fluid (e.g. air, water, etc.) from the uppermost region to the lowermost region where the cooling fluid is exhausted. To avoid solidification of the working substance in the uppermost section of the chamber, the flow of coolant can be reversed such that it enters via the lowermost section and exits the uppermost section. Regardless of the direction of the flow, one can control the degree of cooling of the upper portion of the condenser by controlling the mass flowrate of cooling fluid; thus, this new heat pipe lance features user-controllable rates of heat extraction.

Each of the jackets contains enlarged inlet and outlet headers. These allow the reagent and coolant to be homogenized and uniformly distributed. The relative length of each jacket depends on the application the lance is to be subjected to. While the desired degree of preheat for the reagent is one parameter that influences the length of the jacket to be used, other factors, such as reagent flowrate, must also be considered in determining the size. In the limit the lance may contain no preheating jacket with a long cooling jacket or it may contain a long preheating jacket with no cooling jacket, with a compromise between these two extremes normally prevailing.

As the jackets will typically reside outside the furnace environment, even though it is equally viable to have the jackets m whole or m part residing within the furnace, they can be produced as half sections which snap together over the outer section of the lance and are sealed by high temperature gaskets. Moreover, the jackets need not be circular in cross-section; they can be of any shape deemed appropriate for the prevailing heat transfer conditions. In the event of failure of the lance, the jackets can be removed and reapplied onto a new lance with relative ease. For applications m which lance life can be considerable, it may be more practical to weld circular jackets onto the outer body of the lance, thereby avoiding the use of sealing gaskets. The operating orientation of the lance is by no means limited to the vertical. With the proper use of a wick m the working substance chamber and the judicious selection of jacket cross-section so as to promote greater heat extraction on the upper axial positions along the pipe as opposed to the lower axial positions, the lance can be operated at any inclination from the vertical down to the horizontal. Near-horizontal operation can be used to describe the working orientation of a submerged tuyere. All aspects of the present invention as described for lances are equally applicable to submerged heat pipe tuyeres extending into a melt . Thus, one should construe the present invention to encompass both lances and tuyeres equally. It is well documented that the liquid pool of working fluid may cause two problems on the heat pipe. One is the significant superheat which may appear m the liquid pool. The superheat m the liquid pool creates a controversy as to the isothermal feature of the heat pipe. The other is the slugging of the working fluid which may occur during the startup of the heat pipe. The slugging in this case refers to a non-stationary state of the working fluid where slugs of liquid, with a characteristic dimension equal to the heat pipe diameter, are periodically ejected from the pool, traverse some distance toward the condenser, and disappear into the liquid film along the heat pipe wall. The slugging is usually caused by the flash vaporization of the working fluid within the pool. These two issues are unsolved problems for gravity-assisted heat pipes. The pool of working fluid has been causing problems for the application of heat pipes m the nuclear industry and in some other industries .

For a heat pipe, a relatively large superheat m the liquid pool obviously reduces the heat transport capacity through the liquid pool. For a heat pipe cooled top-blowing injection lance, the nozzle is covered by the liquid pool. This region typically withstands the biggest heat fluxes and is exposed to the highest oxidation rates It is thus the place which requires the most cooling. It is an area where every additional degree of temperature can be of concern. Therefore, the big superheat m the liquid pool may be too costly for the life of the heat pipe cooled injection lance. Brief Description of the Drawinσs

The invention will now be disclosed with reference to the accompanying drawings m which:

Fig. 1 is a diagram depicting the longitudinal cross-section of the heat pipe lance; Fig. 2 is a view of the radial cross-section of the heat pipe lance at section F;

Fig. 3 is a view of the radial cross-section of the heat pipe lance at section G;

Fig. 4 shows the longitudinal cross-section of an annular heat pipe lance as described m the prior art (U. S. Patent 5,310,166);

Fig. 5 is a diagram depicting the longitudinal cross-section of the heat pipe lance;

Fig. 6 is a view of the radial cross-section of the heat pipe lance at section J;

Fig. 7 is a view of the radial cross-section of the heat pipe lance at section K;

Fig. 8 is a diagram showing the temperature curves plotted against a different distance from the tip or leading end of a lance;

Fig. 9 is a schematic diagram of a heat pipe showing the working fluid;

Fig. 10 is a diagram showing a temperature profile of a sodium heat pipe; Fig. 11 is a diagram showing a temperature profile of a potassium heat pipe;

Fig. 12 is a diagram showing a temperature profile of a water heat pipe; Fig. 12a is a diagram temperature profile of superheat for nucleate pool boiling of sodium;

Fig. 13a is a top plan view of a temperature stabilizer for use m the heat pipe lance;

Fig. 13b is an axial cross-section taken through the stabilizer shown m Fig. 13a;

Fig. 14a is a schematic view of a heat pipe lance according to the prior art before its modification for reducing the superheat m the liquid pool; Fig. 14b is a schematic diagram, similar to

Fig. 14a, but showing the heat pipe lance with the stabilizer installed for reducing the superheat in the liquid pool;

Fig. 15 is a diagram showing the temperature profile of a sodium heat pipe lance with a temperature stabilizer;

Fig. 16 is a diagram showing the temperature profile of a potassium heat pipe lance with a temperature stabilizer; and Fig. 17 is a diagram showing the temperature profile of a water heat p pe lance with a temperature stabilizer. DESCRIPTION OF THE PREFERRED EMBODIMENTS

The longitudinal cross-section of the heat pipe lance shown m Fig. 1 comprises a lance body 7, on whose inner surface is applied a wick 8. The working substance, predominantly m the vapor state, fills the heat pipe chamber 15. At the leading end of the heat pipe one may find a small pool of liquid working substance 16. The upper section of the lance is fitted with a cooling jacket 1, into which coolant is pumped through inlet 10, v a feed pipe 4, and is distributed in header 21. The coolant then travels through the annular heat transfer gap 5, to the outlet distribution header 20, and out through the exhaust port 12.

Below the cooling jacket 1 is the reagent preheating jacket 2, into which reagent is introduced via inlet 11, through reagent feed pipe 3. The reagent is distributed by the inlet header 18 and flows through the annular preheating gap 6 to the outlet distribution header 19. The reagent then enters through conduit inlet 13, which penetrates through the wick 8, and the wall of the heat pipe 7. The reagent is forced through the central conduit 9 to the lance nozzle 14. Both the coolant and reagent jackets depicted m Fig. 1 are equipped with dual intake pipes and inlets 4 and 10 for the coolant circuit and 3 and 11 for the reagent circuit to provide better homogenization. Also shown m Fig. 1 is nipple 17, which is used to permanently seal the working substance chamber 15 prior to the application of the coolant jacket 1.

Fig. 2 shows a view of the radial cross-section of the heat pipe lance at location F. Reagent is contained by the outlet distribution header 19 and the lance wall 7 as it flows through the preheating gap 6. Reagent enters the central conduit inlet 13 and travels through conduit 9 on its way to the nozzle. The working substance chamber 15 is bounded by the wick 8.

Fig. 3 shows a view of the radial cross-section of the heat pipe lance at location G. Coolant is contained by the coolant jacket 1 and the lance wall 7 as it flows through the annular heat transfer gap 5. The working substance chamber 15 is bounded radially by the wick 8. Fig. 4 shows a longitudinal cross-section of a typical, annular heat pipe lance as detailed m the prior art (U. S. Patent 5,310,166). The working substance chamber 15 is bounded by the inner surface of the outer heat pipe wall 7 and the outer wall of the inner conduit 22. A wick 8 is applied on the inner surface of the heat pipe wall. Reagent enters the inner pipe 22, via the inlet 23, flows through the passage 6, and is discharged into the melt by the nozzle 14. A pool of working substance 16 resides on the base of the lance.

Also indicated in Fig. 4 are six surface areas as denoted by the labelling a through f. Surface area a refers to the evaporator surface area formed by the base of the leading end of the heat pipe. Surface area b indicates the vertical surface area of the outer pipe m direct contact with the pool of working substance. Surface area d refers to the intermediate vertical surface area of the evaporator while surface area e refers to the top vertical surface area of the evaporator. The two remaining surfaces, c and f, represent surface areas available for condensation. Surface area c is delineated by the outer surface of the central inner conduit spanning the entire length of the conduit . Surface area f represents the inner vertical surface of the outer pipe extending from the top of the evaporator to the top of the heat pipe .

Removal of a large part, if not all, of the inner conduit that constitutes a major feature of the annular heat pipe lance as proposed in the prior art is a prerequisite to the execution of the present invention. Since the inner conduit carrying the reagent to the melt acts as a condenser for the working substance, it stands to reason that whatever is condensed on the inner conduit flows down the walls of the conduit to the nozzle region. Thus, this condensate bypasses the evaporator walls. Under high heat flux and hard blowing conditions, a peculiar set of circumstances may culminate m the drying up of the evaporator walls starting from somewhere above the liquid pool and extending upwards, until m the limit, the vast majority of the evaporator may be dry.

The drying up of the evaporator occurs when the rate of condensation of working substance on the inner conduit exceeds the rate of evaporation from the liquid pool. With reference to Fig. 4 depicting the annular heat pipe lance detailed m U. S. Patent 5,310,166, it can be shown that when the rate of condensation on surface c exceeds the combined rates of evaporation from surfaces a and b, then some working substance vaporized on the top and intermediate evaporator surface is diverted and condensed on the inner conduit . The continuance of this situation can lead to a virtual drying up of the intermediate evaporator. Moreover, as the intermediate and then the top evaporator area is dried up, the outer condenser is also dried up, with the result that the bulk of the evaporator walls become excessively hot and subject to the usual failures encountered m such processes. However, it will be noted that such failures typically will occur at locations above the liquid pool that collects m the nozzle region. Thus, the rate of condensation on surface c must be less than the combined rates of evaporation from surfaces a and b. Transforming this statement into one of mass flows, it is clear to someone knowledgeable in the art that the mass flow of condensate (i.e. C) produced on surface c must be less than the combined mass flows of evaporation from surfaces a (i.e., A) and b (i.e. , B) . Thus,

C < A + B where the units of C, A and B are m mass of working substance per unit of time. As the rate of condensation on surface c is directly linked to the heat extraction capability of the reagent, it is obvious, for example, that for a given reagent, heat extraction increases with increasing reagent velocity. Since the rate of heat extraction is directly linked to the rate of condensation by the thermodynamic property called the heat of vaporization, it then stands to reason that C will increase with increasing reagent flow velocity. If conditions are such that C exceeds the sum of A and B, the creation of 'hot spots' on the evaporator wall will be the outcome. Given adequate wick ng of the evaporator walls, the hot spot will manifest itself m the shape of a ring that envelopes the entire circumference and covers the intermediate vertical surface area of the evaporator. The width of the hot ring is dictated by the degree of imbalance between mass flow C and the sum of mass flows A and B. Once equality is established between these streams, the width of the hot ring stabilizes and steady state is said to be attained. To force the elimination of the hot ring (i.e., intermediate surface area d) requires that C be maintained at a level which is less than the sum of A and B. Under this operating constraint, the hot ring (i.e., intermediate surface area d) will disappear and isothermal processing will be restored.

By virtue of the present elaboration, one can surmise that C (the mass flow of condensate on the inner conduit) can be reduced by decreasing c (the surface area of the inner conduit active m the condensation of working substance) , or alternatively by insulating area c, thus reducing its effectiveness m condensing working substance. The use of insulation, however, entails limited heat extraction by the reagent with the result that preheating of the reagent and energy recycling are both reduced. An energy efficient solution as detailed m the present invention calls for the preheating of the reagent m an outer jacket. In this way, energy is recycled to the reagent and condensation is forced to the inner surface of the outer wall of the heat pipe lance. As has already been stated, the heat pipe lance can be operated m positions other than the vertical. In such cases, it is necessary to ensure the proper refluxing of the entire evaporator walls with liquid working substance. To achieve this, it may be necessary, in addition to the use of a wick, to compartmentalize the coolant and reagent jackets to ensure that sufficient condensation occurs on the upper sections of the jackets. To illustrate this feature of multiple (i.e., split) jackets, one can consider a heat pipe tuyere configuration for which a portion of the tuyere extends through the vessel lining and into the melt. The tuyere is angled such that it is near horizontal although a slight inclination towards the melt is preferable (i.e., angle of inclination from the vertical <90 ) . Shown m Fig. 5 is a longitudinal cross-section of the heat pipe tuyere that forms part of the embodiments of the present invention. Assuming a vertical vessel lining, the tuyere is installed such that it is at angle s from the vertical where s is less than 90 . For the sake of simplicity, the coolant and reagent jackets have each been split into two segments, the upper and lower segments, as detailed in Fig. 5, although one can clearly envisage the use of a greater number of compartments if deemed necessary. The heat pipe tuyere, shown m longitudinal cross-section m Fig. 5, comprises the upper coolant jacket 51 and the lower coolant jacket 52. Coolant to the upper coolant jacket 51 is fed via entry port 59, through intake pipe 57, and distributed by header 79 into the heat transfer passage 53. The coolant proceeds to the exnaust header 81 and exits via the outlet 76. The lower coolant jacket 52 is fed coolant via entry port 60, through intake pipe 58, and distributed by header 78 into the heat transfer passage 5 . The coolant proceeds to the exnaust header 80 and exits through the common outlet 76. The upper and lower coolant jackets are physically separated by the divider 75 (shown m Figs. 6 and 7) which allows for individualized control of coolant flow rate through each of the jackets. The reagent preheating portion of the heat pipe tuyere also comprises an upper preheat jacket and a lower preheat jacket. Reagent enters the upper preheat jacket 65 via entry port 63 , and travels through intake pipe 61 into the intake neader 83, and through the preheating passage 67. Reagent flows into the exhaust header 85, enters the central conduit through port 70, and flows through the central conduit 72 to the tuyere nozzle 73. Similarly, reagent enters the lower preheat jacket 66, via entry port 64, and travels through intake pipe 62 into the intake header 82, and through the preheating passage 68. Reagent from the lower preheat jacket then flows into the exhaust header 84, enters the central conduit through port 71, and merges with the reagent flowing from the upper preheat jacket. The combined reagent flow moves through central conduit 72 and on to the tuyere nozzle 73.

The tuyere body 55 is fitted on the inner surface with a wick 56. The working substance chamber 69 contains the vapor/liquid working substance which is sealed inside the chamber via nipple 74. The central conduit 72 and the leading face 78 of the tuyere perpendicular to the longitudinal axis of the tuyere, are also fitted with a wick 77 to enhance even distribution of condensed working substance.

SUBSTTTUTE SHEET (RULE 26) Fig. 6 shows a view of the radial cross-section of the heat pipe tuyere at location J. The schematic shows a section through both the upper and lower reagent preheat segments of the tuyere assembly. The upper preheat segment is located on the right of the line of symmetry while the lower preheat segment is located on the left. The upper and lower portions of the preheat jacket are separated by divider 75. Reagent m the upper preheat jacket is contained by the outlet distribution header 85 and the lance wall 55 as it flows through the upper preheating gap 67 into the central conduit inlet 70 and down the central conduit 72 to the nozzle. Similarly, reagent m the lower preheat jacket is contained by the outlet distribution header 84 and the lance wall 55 as it flows through the lower preheating gap 68 into the central conduit inlet 71 and down the central conduit 72. The working substance chamber 69 is bounded on its perimeter by the wick 56.

Fig. 7 shows a view of the radial cross-section of the heat pipe tuyere at location K. The schematic shows a section through both the upper and lower coolant jackets which are separated by divider 75. The upper coolant jacket 51 and the lance wall 55 constrain coolant as it flows through the upper heat transfer gap 53. Inside the working substance chamber 69, a wick 56 is affixed to the inner wall of the heat pipe. Similarly, the lower coolant jacket 52 and the lance wall 55 constrain coolant as it flows through the lower heat transfer gap 54. The methods of the present invention are now illustrated by way of the following examples. Example 1 :

A heat pipe lance as described m the prior art (U. S. Patent 5,310,166) was tested m a radiation furnace The lance measured 50 cm m length and 2.5 cm m outer diameter with a wall thickness of 1.6 mm. A central inner conduit: spanning the entire length of the heat pipe measured 6 mm m outer diameter with a wall thickness of 1 mm. The lance nozzle was 2 mm in internal diameter. Two wraps of 100 mesh screen were embedded on the inner surface of the outer wall of the pipe. All construction materials were 304 stainless steel. The pipe was charged with 17 g of sodium, evacuated and sealed. A test was run to determine the effect of injecting reagent through the central conduit on a number of outer wall operating temperatures of the heat pipe. The furnace was set at 843 C for the entire test. Four distinct blowing conditions wherein air was used as the reagent were studied. In the first trial, no reagent was blown through the lance of which 30 cm were positioned within the hot furnace environment . Eleven thermocouples were strapped onto the outer surface of the pipe at various vertical positions. Once steady state was attained, the readings from the thermocouples were noted and the flow rate changed to the next level of 0.15 Nv£/s . Once again, the new temperature readings were recorded once steady state was attained. Two other reagent gas flowrates of 0.30 and 0.40 Ni/s were also tested. The results from all four tests are shown m Fig. 8. It will be noted that with no flow through the central conduit, the evaporator section which spans the first 30 cm was very uniform m temperature (approx. 735 C) . Beyond the evaporator and extending to the top of the pipe, the temperature dropped to 630 C m an ambient environment of 20°C.

With a reagent flowrate of 0.15 Nvfc/s, the temperature of the evaporator at the bottom decreased to about 721 C at steady state. A hotter intermediate zone measuring about 7 cm m height was detected as can be

SUBSTΓΓUTE SHEET (RULE 26) determined from the data presented m Fig. 8 At this flowrate, the condition of C being initially greater than the sum of A and B must have existed. In other words, the rate of condensation of working substance on the central conduit exceeded the rate of evaporation from the base and lower area of the evaporator. This then led to the formation of the intermediate area d on the evaporator from which vaporization of working substance was impeded with the consequence that surface area b of the liquid pool increased and the equality of C with the sum of A and B was established. At this point in time, steady state was said to be achieved and stable operation of the heat pipe lance could be maintained.

When the flowrate was increased to 0.30 N-β/s, the temperature of the evaporator at the bottom decreased to 712 C at steady state. The hot intermediate zone in this case increased to about 21 cm m length. While the condenser temperature dropped to about 517 C, the evaporator surface temperature peaked at about 824 C - only 19 C below the furnace temperature.

An increase in the flow rate to 0.40 Nj£/s showed a similar trend, the bottom evaporator temperature decreased to about 678 C while the condenser temperature dropped to 411 C. However, the intermediate hot zone grew m length to about 27 cm and attained a peak temperature of about 817 C. This clearly illustrates that by forcing condensation of working substance on the inner conduit, one can greatly distort the operating characteristics of the heat pipe lance. A lance operated in this fashion can readily fail m high temperature applications. As a final point, it is to be noted that the hot ring was clearly visible during those tests with a reagent flowrate of 0.15 N L /s or greater, thus confirming the temperature data. No hot ring was present with zero flow as expected. A similar test to the above was carried out with a lance designed with the embodiments of the present invention. The reagent air entered a preheating jacket which was 20 cm m length, was preheated and then allowed to flow into a central conduit leading to the nozzle. As before, flowrates of 0, 0.15, 0.30, and 0.40 Njt/s were tested. In all four cases, the temperature of the outer heat pipe evaporator surface was maintained at a relatively uniform level . There was no evidence of the onset of a hot spot. While the steady state temperature of the evaporator decreased with increasing flow rate, as was expected, the temperature of the entire length of evaporator was uniform for each flow rate. Example 2 : A heat pipe lance designed according to the embodiments of the present invention was used to make steel from pig iron initially containing 4.1%C (by weight) . The lance was 50 cm in total length with 20 cm of uncovered evaporator length, a 10 cm long preheating jacket and a 20 cm long cooling jacket. The heat pipe body was 2.5 cm in outer diameter with a wall thickness of 1.6 mm. Both jackets were affixed to the outer heat pipe wall to create a gap of 1.5 mm in width. The central conduit connecting the preheating jacket and the nozzle was 6 mm in outer diameter with a wall thickness of 1 mm. Two wraps of 100 mesh screen were embedded on the inner surface of the outer wall of the pipe. The pipe was charged with 17 g of sodium, evacuated and sealed. All materials of construction were 304 stainless steel with the exception of the leading face of the lance which was made of nickel .

A charge of 12 kg of pig iron was melted in a 14 cm diameter alumina crucible in an induction furnace. Once at the desired temperature of 1250 C, the crucible was tipped to 30 from the vertical. The lance was positioned, also at an angle of 30 , such that the nozzle was 7 cm from the melt surface. Oxygen reagent was blown at a rate of 0.30 Nj-t/s into the melt through the 2 mm nozzle. Thermocouples strapped onto the body of the lance and on the base monitored the temperatures. Blowing of reagent was continued for a total of 30 minutes after which time the melt carbon concentration had been reduced to 0.06% C.

During the test the lance body was found to be uniform in temperature but dependent on heat flux with fluctuations occurring only in time. A maximum temperature of 850 C was recorded at the tip of the lance where incipient heat fluxes are highest . The lance operating temperature was readily adjusted by varying the flow rate of cooling air through the coolant jacket. For example, a flow rate of 1 N%£. /s of air reduced the temperature by slightly more than 50 C. At the end of the blow which lasted about 25 minutes, the melt temperature was about 1600 C. As the lance operated at a relatively low temperature with the working substance averaging about 560 C, an accretion formed on the leading face of the lance and on the lower body of the evaporator. Once the accretion was removed at the end of the test, the nozzle and leading face were found to be in excellent condition with no signs of wear or attack. The lance was reused on a number of occasions with identical results . Example 3 :

The heat pipe lance as described m Example 2 was also used to convert copper matte (predominantly CU2S) containing 70% Cu (by weight) to blister copper. The equipment used m the test was identical to that described m Example 2 The charge consisted of 5 kg of matte and the lance was positioned at an angle of 30 from the vertical to within 2 cm of the melt surface at

SUBSTTTUTE SHEET (RULE 26) the start of the test . Oxygen was blown at a rate of 0.30 N jL/s for a period of 35 minutes. It was not necessary to use cooling air in this test. At the end of the blow virtually all the sulphur had been oxidized to SO2 gas. Blowing was initiated at a temperature of 1280 C while the final melt temperature was about 1500 C. As with the steelmakmg test, an accretion had formed on the lance during the blow. There were no signs of wear or chemical attack of the lance or nozzle. The lance was retested a number of times at varying angles of inclination with identical results.

While a heat pipe lance or tuyere as described above can, in many circumstances, be sufficient, there are other cases when it may be necessary to incorporate a device that can homogenize the temperature of the liquid pool m the vicinity of the tip to reduce the overall temperature at the tip. The device that has been discovered and adopted has been termed a "temperature stabilizer" . It is especially effective when the working substance is conductive (e.g., potassium, sodium - alkali metals) . Its use provides two mam advantages: (1) homogenizes temperature gradients within the liquid pool; and (2) promotes boiling and sprays liquid working substance the evaporator section. For optimum use a heat pipe lance or tuyere should incorporate a temperature stabilizer as described the following embodiment .

Fig. 9 shows the schematic diagram of a vertical gravity-assisted heat pipe with a liquid pool at the bottom end of its evaporator section. Apparently, the working fluid is one of three essential components which make up the heat pipe. Because of this, the difficulty of calculating the right amount of working fluid for the heat pipe, and the dynamic features of the heat pipe lead to an over charging of a heat pipe. Since all gravity-assisted heat pipes have to be charged with some amount of excessive working fluid to ensure that they can operate under various operating conditions, the presence of a liquid pool is inevitable. In addition to that, working fluid over and above that required to simply maintain a vertical gravity-assisted heat pipe functional is needed to ensure that the heat pipe does not encounter any difficulty at startup from a frozen state. Usually, the amount of working fluid in the liquid pool is decided by the heat extraction rate, the heating speed at the startup stage, the wick structure in the heat pipe, and the thermal mass of the heat pipe. However, excessive working fluid may cause the heat transport capacity of the heat pipe to decrease. This decrease in heat transport capacity relates to the heat transfer mode in the liquid pool.

In the heat transfer literature, when a heating surface is submerged in an otherwise quiescent pool of liquid, and heat is transferred to the liquid by free convection and bubble agitation, the process is termed pool boiling. The liquid pool formed by the working fluid in a vertical gravity-assisted heat pipe undergoes the process of pool boiling.

The pool of working fluid in a vertical gravity-assisted heat pipe can undergo a typical pool boiling process. To achieve a high heat transport rate through the heating surface and the liquid pool in a heat pipe, nucleate boiling is the stage for the liquid pool of the working fluid to operate at . The superheat requirements for nucleate boiling in the liquid pools were reported when potassium and sodium were chosen as the liquids (J. C. Chen, "Incipient Boiling Superheats in Liquid Metals", Trans. Am. Soc. Mech. Engrs, J. Heat Transfer C, Vol. 90, pp. 303-312, 1968; and R. E. Holtz, and R. M. Singer, "On the Initiation of Pool Boiling in

SUBSTTTUTE SHEET (RULE 25) Sodium" , Paper presented at the Tenth National Heat Transfer Conference, Philadelphia, Pa., August, 1968.) These were also confirmed m this study.

Fig. 10 shows the superheat requirement for nucleate boiling the liquid pool m a vertical gravity-assisted heat pipe with sodium as the working fluid. In Fig. 10, as determined m the present study, curve #1 is the pool temperature, while curve #2 is the saturation temperature. The superheat requirement for nucleate boiling m the liquid pool varies between 60°C and 80°C for a saturation temperature ranging between 540°C and 570°C.

In addition to the measurements on the sodium system, the potassium and water systems were also studied. Fig. 11 shows the superheat requirement for nucleate boiling m the liquid pool in a vertical gravity-assisted heat pipe with potassium as the working fluid. In Fig. 11, curve #1 is the pool temperature, while curve #2 is the saturation temperature. The superheat requirement for nucleate boiling in the liquid pool varies between 60°C and 80°C for a saturation temperature between 330°C and 360°C. Fig. 12 shows the superheat requirement for nucleate boiling m the liquid pool m a vertical gravity-assisted heat pipe with water as the working fluid. In Fig. 12, curve #1 is the pool temperature, while curve #2 is the saturation temperature . The superheat requirement for nucleate boiling in the liquid pool varies between 5°C and 20°C.

The significant superheat the liquid pool of the working fluid is a distinct feature which the vertical gravity-assisted heat pipe has.

Basically, there are two parts which contribute to the temperature difference between the heating surface of the bottom cap and the free surface of the liquid pool. One is the temperature difference between the

SUBSTITUTE SHEET (RULE 25) heating surface of the container and the adjacent liquid. The other is the superheat m the liquid pool.

It has been well documented that partial film boiling and film boiling have to be prevented because of a huge temperature difference between the heating surface and the bulk of the liquid. On the other hand, free convection boiling cannot be maintained for high heat transport rates. Therefore, to achieve the best heat transport results from pool boiling, the most useful regime is nucleate boiling.

As described in the last section, the most important regions for achieving high heat transfer coefficients are the free convection boiling region and nucleate boiling region. As heat flux increases, the heat transport mode in the liquid pool switches from free convection boiling to nucleate boiling. However, one parameter that should be noticed is the depth of the pool. It is another parameter, which also determines when the heat transfer mode changes from the free convection boiling to the nucleate boiling as heat flux increases.

Suppose a pool of sodium undergoes free convection boiling. If heat is applied to the bottom and the side wall of the container, and the heat fluxes on the bottom and on the side wall are the same, then the temperature difference through the pool is a function of the heat flux applied on the container wall and the depth of the pool. If the pool is stationary (i.e., motion is negligible) , the relationship between the temperature difference, the heat flux, and the depth can be determined by the basic law governing heat conduction:

.7' = * Eσ. 1

Δx For example, the evaporator of a 600 mm long heat pipe with 25.4 mm O.D. and 1.65 mm wall thickness is exposed to 875 °C hot furnace environment and the heat pipe operates at 600°C. If the emissivity of the evaporator is 0.8 and ambient temperature is 27°C, then, the heat flux on the free surface of the liquid pool of sodium is given m Table 1. Assuming the heat flux applied to the evaporator is the same, the smaller the depth of the liquid pool is, the smaller the total surface area of the evaporator immersed in the liquid pool is. Therefore, the heat flux on the top free surface of the liquid pool is proportional to the depth of the liquid pool. According to Fig. 12a, 81°C of superheat the liquid pool is required to have the heat transfer mode change from free convection boiling to nucleate boiling while the saturation temperature is 600 °C. Then, the thickness of the free convection boiling layer can be calculated by Eq. 1, and is shown Table 1, assuming the thermal conductivity of sodium at 600°C is 64.6 W/m-°C. Therefore, for the conditions given in the above example, nucleate boiling will not occur when the depth of the liquid pool is less than 20 mm, because the thickness of the free convection boiling layer is bigger than the actual depth of the liquid pool. The temperature difference through the layer of free convection boiling decreases as the depth of the liquid pool decreases from 20 mm. The temperature difference through the layer of free convection boiling can be significantly smaller than the superheat required by nucleate boiling when the depth of the liquid pool is small enough. For instance, the temperature difference for nucleate boiling is 81°C for this case, but the temperature difference through the liquid pool is only 7.7°C when the depth of the liquid pool is 5 mm, and 22.8°C for 10 mm. However, nucleate boiling will occur when the depth of the liquid pool is slightly bigger than 20 mm.

Based on the analytical results presented above, 5 to 10 mm depth of liquid pool is ideal for the heat pipe, because the temperature differences are much smaller than the superheat required for nucleate boiling. Unfortunately, 5 mm depth of liquid pool with 22.098 mm internal diameter translates to 1.5 grams in weight for sodium at 600°C. Thus, in practice, it is not practical to fix the depth of liquid pool to 5 or 10 mm because of the dynamic features of the heat pipe and because of the amount of the working fluid required at the startup stage. This leaves a big question mark on how to find a practical way to control the temperature difference through the liquid pool to within a much smaller range ⁽say 10 to 15°C) than the superheat required by nucleate boiling, when the liquid in the pool has to transfer heat by the nucleate boiling mode.

Table 1 The heat βux on the evaporator

SUBSTITUTE SHEET (RULE 25) For a vertical gravity-assisted heat pipe with an alkali metal as working fluid, the depth of liquid working fluid pool is usually more than 20 mm depending on the size of the evaporator and total length of the heat pipe. For most of the high temperature applications of heat pipes, the heat transfer mode liquid pools is nucleate boiling.

The superheat required for nucleate boiling is a function of the radius of bubbles in the liquid pool, and the radius of the bubbles is decided by the size of nucleation sites. The smaller the size of the nucleation site is, the higher the superheat in the liquid pool that is required. To achieve smaller superheat, the size of the nucleation sites snould be increased. However, the size of nucleation sites is a function of the quality of the internal surface of container wall, and not much difference can be made to physically increase the size of nucleation sites on the internal surface of the container wall. Typically, the radii of the bubbles are between 1 to 10 micrometers.

It has been shown that a way for reducing the superheat m nucleate boiling is to increase the radius of bubbles, and it needs to be done by some method other than physically increasing the size of the nucleation sites on the internal surface of the container wall.

Much attention was paid to this aspect this study. Six vertical gravity-assisted heat pipes were made to investigate the superheat phenomenon of nucleate boiling m the liquid pool. One possible solution which was found during this study is to separate the liquid pool into two parts at the place of 5 to 10 mm above the bottom by using a plate with a few holes on it. In the pool below the plate, the heat will be transported by free convection boiling. In the pool above the plate, the heat will be transported by nucleate boiling. Using

SUBSTTTUTE SHEET (RULE 26) this method, the superheat requirement for the nucleate boiling the liquid pool above the plate may be controlled to within 15°C, because it is no longer the function of the amount of the liquid m the pool. Fig. 13 shows a drawing of a piece of plate which needs to be embedded the liquid pool to initiate the much bigger bubbles than normal nubbles with the radius between 1 and 10 micrometers. Practically, it is an easy solution. For implementation of the new method, the plate made from the same material as that of the container is fixed the liquid pool parallel to the upper surface of the bottom cap with a 5-10 millimeter gap m between. There are a number of holes, usually 3 to 5 holes, normal to the surface of the plate. The holes are usually about 3 mm diameter through the plate. This plate is referred to as a temperature stabilizer this study. The purpose of the temperature stabilizer is to reduce the difference between the liquid temperature and the saturation temperature to the minimum. Fig. 14 shows two diagrams of before and after the modification at this stage. The following is an explanation of how the temperature stabilizer works to reduce the superheat the liquid pool.

In a vertical gravity-assisted heat pipe with the proposed temperature stabilizer the liquid pool, the liquid pool is separated by the temperature stabilizer into two sections. The heat transfer modes are different m these two sections.

In the bottom section, it usually undergoes free convection boiling. Because the depth of the bottom section is only about 5 mm, the superheat can only be 7.7°C if the operating temperature of the heat pipe is 600°C and the heat source is about 875°C. From the example given, it has been shown clearly that under the same heat fluxes m a liquid pool of about 5 mm depth,

SUBSTTTUTE SHEET (RULE 26) the temperature difference can be an order of magnitude less than the superheat required for nucleate boiling in a pool of several times the depth of 5 mm. Based on the observation in this study, there is no perfect contact between the liquid and the bottom surface of the temperature stabilizer, which means there is a free liquid surface in the bottom section. Vapor generated on the free liquid surface is forced to go through those holes on the temperature stabilizer. Generally, at relatively low heat load, the bottom section works as a heat pipe while the bottom surface of the temperature stabilizer works as the condenser. As heat load increases, at some point the evaporation will exceed the condensation in the bottom section, which leads to some vapor generated in the bottom section to go into the top section through the holes on the temperature stabilizer.

In the top section, when there is no vapor getting into it from the bottom section, it performs as a regular liquid pool, low superheat for free convection boiling and high superheat for nucleate boiling. When there is vapor getting into it from the bottom section, the size of the bubbles is no longer 1 to 10 micrometers as generated by the nucleation sites. If it is assumed that vapor forms spherical bubbles when it gets into the top section and the radius of the spherical bubbles equals the radius of those holes on the temperature stabilizer, it is clear that the radii of the new bubbles are orders of magnitude bigger than that of the initial bubbles generated in the nucleation sites on the heating wall .

Since the pressure difference will decrease as the radius of the bubbles increases, the superheat in the top section will be significantly reduced as the bubble size increases from 1 to 10 micrometers to more than 1000 micrometers. Based on the better understanding of the boiling process liquid pool, and the development of the temperature stabilizer, a number of heat pipes and heat pipe cooled injection lances which incorporated the temperature stabilizer were designed, fabricated and tested. As to be shown, they all worked as expected and reduced the superheats substantially.

Fig. 15 shows the experimental result of a vertical gravity-assisted heat pipe with sodium as working fluid. In Fig. 15, curve 1 represents the temperature in liquid pool, and curve 2 represents the saturation temperature the evaporator. The difference between curve 1 and curve 2 represents the superheat of the liquid sodium. In this sodium heat pipe, thermocouple #1 was positioned the bottom section of the liquid pool. In Fig. 15 temperature curves can be divided into three regions on time axis. The first region is the one from 0 to 180 seconds, the second region is the one from 180 to 2400 seconds, and the third region is the one after 2400 seconds. In the first region, the temperature m the liquid pool was lower than the temperature the rest of the evaporator because of different thermal mass m these two sections. After the heat pipe was heated for 180 seconds, the temperature of the liquid sodium caught up, and then passed over the saturation temperature in the rest of the evaporator. In the second region, the superheat of the liquid increased to as big as 100°C as if the temperature stabilizer was not the liquid pool because the temperature stabilizer was not functioning. At 2400 seconds, the temperature stabilizer started to function as shown on curve 1 and curve 2. When the temperature stabilizer started to work, the temperature of the liquid decreased as much bigger bubbles passed through the liquid pool, and the saturation temperature m the rest of the evaporator

SUBSTTTUTE SHEET (RULE 26) increased slightly at the same time, which indicates the heat pipe worked more efficiently. The superheat of the liquid sodium was reduced to about 25 °C, which is substantially smaller than the superheats on those heat pipes and heat pipe cooled injection lances prior to the modification.

Fig. 16 shows the experimental result of a vertical gravity-assisted heat pipe cooled injection lance with potassium as working fluid. In Fig. 16, curve 1 represents the liquid temperature, and curve 2 represents the saturation temperature. In this potassium heat pipe cooled injection lance, thermocouple #1 was positioned right above the temperature stabilizer the top section of the liquid pool. In Fig. 16, there are two peaks on curve 1. At time 300 seconds, curve 1 peaked out. It indicates that the temperature stabilizer started to work. Between 350 and 950 seconds, the temperature stabilizer was working, and the superheat of the liquid was a function of the saturation temperature of the heat pipe. It is shown in Fig. 16 that the superheat decreased as the saturation temperature increased. Between 950 and 1050 seconds, the curve 1 peaked out again. This peak indicates how big the superheat of the liquid can be if the temperature stabilizer was not working After the second peak, the superheat of the liquid was about 5°C, which is substantially smaller than the superheats on those heat pipes and heat pipe cooled injection lances prior to the modification. Fig. 17 shows the experimental result of a vertical gravity-assisted heat pipe cooled injection lance with water as working fluid. In Fig. 17, curve 1 represents the temperature of water, and curve 2 represents the temperature of steam above the liquid pool. As the potassium heat pipe cooled injection

SUBSTTTUTE SHEET (RULE 25) lance, thermocouple #1 was positioned right above the temperature stabilizer in the top section of the liquid pool. In Fig. 17, both curve 1 and curve 2 indicated that the temperature stabilizer made three attempts to become functional, and finally, it stayed functional. During the period of the temperature stabilizer staying functional, the superheat of the liquid was within 2°C, which is substantially smaller than the superheats on those heat pipes and heat pipe cooled injection lances prior to the modification also.

The working mode of the temperature stabilizer not only can be identified from the temperature difference but also can be identified by the sound caused by big bubbles breaking the surface of the pool . Our tests with sodium, potassium, and water were consistent and in all cases a pinging sound was initiated as the temperature stabilizer began to function. The sound persisted as long as the stabilizer functioned. If the stabilizer stopped working, the sound ceased. The optimum lance or tuyere is one which incorporates both embodiments of the present invention. The unit should provide for complete coverage of the evaporator by the working substance. This can be achieved by replacing the central reagent feed pipe with a reagent preheating jacket. The other embodiment can be adopted to minimize the temperature gradient in the liquid pool . It requires that the leading end of the lance be fitted with a perforated plate. This plate promotes boiling at a temperature that is comparable to the saturation temperature.

Claims

CLAIMS ;

1. A heat pipe lance for conveying reagent materials into a metallurgical bath, wherein the lance includes an elongated heat pipe vessel defined by a tubular wall, a coolant liquid in the vessel selected from a class of liquids including liquid sodium and liquid potassium, the heat pipe vessel including an evaporator section near a leading edge of the lance and a condenser section near the trailing end of the lance, the coolant liquid forming a liquid pool in the evaporator section, a separate conduit passing through at least a portion of the heat pipe vessel at the evaporator section and open to the leading edge of the lance to form a nozzle for passing materials into the molten bath, and a barrier member within the heat pipe vessel extending laterally of a longitudinal axis of the heat pipe within the evaporator section to separate the evaporator section into two portions, and liquid communicating openings provided in the barrier to allow liquid communication between the two portions of the evaporator section.

2. The heat pipe lance in accordance with claim 1, wherein the tubular wall forming the heat pipe is provided with a wick extending on an interior surface of the tubular wall in order to allow for the uniform distribution of the liquid condensing in the condenser section back to the evaporator section.

3. The heat pipe lance as defined in claim 1, wherein the barrier is in the form of a plate extending laterally of the evaporator section within the cooling liquid in order to reduce the temperature gradient along the longitudinal axis of the evaporator section such that evaporation takes place in both portions of the evaporator section.

SUBSTTTUTE SHEET (RULE 25)

4. The heat pipe lance as defined claim 1, wherein the conduit extends from the leading edge of the lance centrally of the evaporator section and out of the heat pipe vessel at a point remote from the evaporator section such that condensation will be avoided on the interior wall of the conduit .

5. The heat pipe lance as defined m claim 1, wherein the conduit extends from the leading edge of the lance to the trailing edge of the lance concentric with the heat pipe, and insulation means are provided on the conduit to prevent condensation on the walls of the conduit .

6. The heat pipe lance as defined claim 1, wherein C < A + B when C is the mass flow expressed in time, of the condensation of the vaporized cooling liquid on the surface of the conduit and A and B are the mass flows expressed in time, of the vaporization from the liquid pool in the vaporizer section.

7. The heat pipe lance as defined m claim 3, wherein the heat pipe has an inner diameter of 1 inch and the lateral wall extending through the vaporizer section has between three to five openings of 1/8 inch diameter.

8. The heat pipe lance as defined in claim 1 is embedded in the wall of the vessel and angled to function as a tuyere .

9. A heat pipe lance comprises a first elongated tubular vessel forming the heat pipe and including a leading end and a trailing end, the tubular vessel having a vaporizer section near the leading end and a condenser section near the trailing end, a cooling liquid within the tubular vessel forming a liquid pool the vaporizer

SUBSTITUTE SHEET (RULE 25) section, a reagent conduit forming a nozzle at the leading edge of the vessel and passing through the vessel and concentric with the vessel at least m the area of the vaporizer section, the conduit extending through a tubular wall of the vessel and being sealed from the heat pipe, a first annular fluid conducting jacket surrounding the tubular wall of the heat pipe and communicating with the conduit, and inlet means connected to the jacket remote from the vaporizer section and extending at least over a portion of the condenser section of the heat pipe tubular wall, the inlet means communicating with the annular jacket and the conduit communicating with the jacket, and means for feeding the reagent through the annular jacket for preheating of the reagent before entering the conduit by heat exchange relationship with the tubular wall of the heat pipe vessel .

10. The heat pipe lance as defined in claim 9, wherein a second annular jacket surrounds the heat pipe tubular wall forming the vessel at the condenser section of the vessel and longitudinally adjacent the first jacket, the second jacket surrounding the condenser section of the heat pipe vessel tubular wall and including means for circulating a cooling fluid through the second jacket for the purposes of controlling the rate of condensation within the condenser section of the heat pipe vessel.

11. The heat pipe lance as defined claim 10, wherein a wick extends on the inner surface of the tubular wall forming the vessel to allow for the uniform distribution of condensed liquid from the condenser section to the evaporator section of the heat pipe.

12. The heat pipe lance as defined in claim 11, wherein the wick is a wire mesh.

SUBSTTTUTE SHEET (RULE 26)

13. The heat pipe lance as defined in claim 9, wherein the rate of condensation of the cooling liquid on the surface of the conduit is maintained at a value that is less than the rate of evaporation of the coolant liquid from the liquid pool in the evaporator section.

14. The heat pipe lance as defined in claim 10, wherein the rate of circulation of the coolant fluid in the second jacket is controlled in order to adjust the rate of heat extraction of the heat pipe.

15. A lance for a metallurgical bath comprising an elongated tubular cooling chamber having a forward portion and a rearward axial portion, a conduit for passing a reagent extending concentrically within the forward portion of the cooling chamber and terminating at a tip formed at the forward end of the cooling chamber for passing a reagent into the melt, a preheating chamber concentric with the exterior of the rear portion of the cooling chamber, the reagent conduit communicating with the preheating chamber and means for feeding reagent through the preheating chamber, and a cooling liquid provided in the cooling chamber.

16. The heat pipe lance as defined in claim 9, wherein the coolant liquid within the heat pipe is selected from a class of liquids which includes sodium and potassium.

17. The heat pipe lance as defined in claim 9 is embedded in the wall of the vessel and angled to function as a tuyere .