CA1115060A - Use of argon to prepare low-carbon, low-nitrogen steels in the basic oxygen process - Google Patents

Abstract

ABSTRACT

The nitrogen content of low carbon steel made by the basic oxygen process is minimized by:
(a) introducing nitrogen-free fluid into the vessel before the nitrogen content of the melt has reached its minimum level, (b) adjusting the flow rate of the nitrogen-free fluid to maintain total off-gas flow rate at least equal to that which would have been produced without the nitrogen-free fluid at the time in the refining process when the nitrogen content of the melt reached its minimum level, and (c) the injection of nitrogen-free fluid throughout the remainder of the oxygen blow.
A preferred additional or alternatively separate step involves minimizing the nitrogen content of BOP steel by purging the vessel headspace with a nitrogen-free fluid prior to reblows.

This is a divisional application of applicants' copending patent application, Application No. 288,670, filed October 11, 1977, having a priority date of January 11, 1977.

Description

~ D-10,883 :~:
n3 ~ .

THE USE OF ~RGON TO PREP~RE LOW-CARBON, LOW-NITROGEN STEELS IN THE
BASIC OXYGE~ PROCESS

. . .
This invention relates, ln general, to a process for refining steel, and more specifically, to an improvement in the basic oxygen process wherein molten steel contained in a vessel is refined by top blowing oxygen lnto the melt, i.e. from above the melt surface.
The manufacture of steel by the basic oxygen process, . ;;
also referred to as BOP or BOF process,is well known in ;~ :
the art. When low carbon steel is made by this process, it is often subject to contaminat;on by atmospheric nitrogen. Such contamination tends to cause premature age hardeni~g of the steel, which leads to strain-aging, poor surface properties a~d undesirable appearance of the final product.
The problem of nitrogen pickup during the manufacture ~ :
of low-carbon steels has been addressed by the prior art. Glassman, in U.S. Patent ~o. 3,769,000, describes a method for excluding nitrogen from the melt by placing . .
a hood looqely over the mouth of the refinîng vessel.
Nitrogen from ambient air is excluded by mainta~ning a curtain of carbon dioxide between the hood and the refining vessel. Pihlblad et al, in U.S. Patent No.
3,307,937 disclose a method for excluding atmospheric

-2- :

D~10,883 '''' ~'"

nitrogen from the melt by adjusting the size of the opening through which gas flows out at the top of the vessel, thereby.maintaining positive pressure in the vessel with respect to the ambient atmosphere, even at low carbon levels. Both of these approaches require modification of the BOP vessel which is expensive and cumbersome to utilize, consequently 3 neither has met with signi~icant commerclal success.
In addition to the potential for nitrogen contamination, a second disadvantage of the conven~ional basic oxygen ~ ;
process is the increasing quantity of oxygen that reacts with valuable metal as the carbon content of the melt decreases. Several U.S. patents disclose ways of diluting ~he oxygen with another gas in order to minimize the amount of oxygen that reacts with the metal.
Such patents include Fulton et al U.S. Patent No. -~

3,649,246~and Ramachandran's U.S. Patent Nos. 3,594,155 and 3,666,439. These patents deal only with the problem of increasing the degr~e to which the injected oxygen reacts with carbon rather than the metal, None are concerned with how one might utilize a diluent to minimize nitrogen pickup from the atmosphere during oxygen decarburization in the BOF.

OBJECTS
A~cord~ngly, it is an object of the present invention to prevent contaminat;on of molten errous -3- ~
' '~', :

D-10,883 metal with nitrogen during decarburization by top blowing with oxygen.
It is another obj ect of this invention to produce low-carbon s teels having a low nitrogen content by the basic oxygen process. .
It is still another objeet of this in~ention to :
minimize the amount of nitrogen-free fluid needed to produce low-carbon steel ha~in~ a low nitrogen level.

SUMMARY OF THE INV~NTION

The above and other objects, which will readily be apparent to those skilled in the art, are achleved by the present invention, one aspect of which comprises: :
in a process for the production of low-carbon s~eel by blowing oxygen into a ferrous melt contained in a vessel or zone from above the surface of said melt, the ~ .
. . ; .
~mprovement comprising the production of steel having low nitrogen content by: ~ :
(a) introducing nitrogen-free fluid into the vessel `-before the nitrogen content in the melt has reached its ~ :
minimum level, while continuing the blow with oxygen, (b) adjusting the flow rate of said nitrogen-free fluid so that the total off-gas flow rate from the vessel is maintained at least equal to that which would have be~n produced without said nitrogen-free fluid at the time in the refining process when the nitrogen .,,... . . ~. ~ , ' :;' . ;

10,~83 content of the ~elt reached its minimum level, and (c) continuing the injection of said nitrogen-free fluid ~hroughout the remainder of the ox~gen blow.
During practice of the basic oxygen process it is common to interrupt the inj ec tion of oxygen into the melt and then reblow the melt with oxygen. Reblowing ~he melt is often accompanied by a signiicant increase :
in the dissolved nitrogen content of the melt. To prevent this nitrogen pickup when the oxygen flow has been ~ -interrupted the vessel ~hould be p~rged by injection of a nitrogen-free fluid immediately prior to restarting the injection of oxygen. Thereafter the introduction of nitrogen-free fluid into the vessel is resumed before.-~
the nitrogen content in the melt has reached its minimum level, adjusted and continued as above.
The term "nitrogen-free fluid" as used herein is lntended to mean any fluid, other than oxygen, sub-stantially free of nitrogen or nitrogen containing ~ .
compounds. The term includes but is not limited to :~
argon, helium, neon, krypton, xenon, carbon dloxide, carbon monoxide, steam, water, ~ydrogen, gaseous hydro- ~;
carbons such as methane and ethane, liquid hydrocarbons such as kerosene and n-heptane, and mixtures thereof.
The preferred nitrogen-free fluid is argon. ~ :
The terms "low-carbon steel" and "low-nitrogen ~ -steel" as used herein are in~ended to include respectively D-lo~883 . steels having a carbon content no higher than about 0.10 percent, and steels h~ving a nitrogen content no higher than abvut O . 005 percent (50 ppm) .
The term "off-gas" is used to mean the gases which issue from the gas exit port or top opening of the steel refining vessel while oxygen or oxygen and one or more other gases are ~njected into the v~ssel in order to refine the ferrous melt. -The term "reblow" is used to mean a subsequent blowing of oxygen or 02ygen mixed with other gas into a BOP vessel after the initial flow of the oxygen or oxygen-containing mixture has been stopped for any reason. It is possible to have more than one reblow per heat.
The preferred method of injecting the nitrogen-free fluid is to mix it with the oxygen stream; however alternate methods may also be used. The preferred amount of nitrogen~free fluit to use when purging the vessel prior to restarting the injection of oxygen is a volume of gas, measured at 70F and 1 atmosphere pressure, at least equal to 1/2 the vessel head space.
THE DRAWINGS
Figure 1 is a graph illus trating the final nitrogen con~en~ N as a funceion of the final carbon conten~ C
of a series of heats of metal refined by prior art BOP
.

~`~ D-10,883 practices in a typical commercial refining system wlthout using the present invention. This figure illustrates how data obtained without practîcing the in~Tention is used to determine when nitrogen-free fluid injection ;:
should be ~tarted.
Figure 2 is a graphic representation of the change in off-gas flow rate F as a func~ion of carbon content ~:
C for same system for which data i~ shown in Figure 1.
This graph shows how the data, ob~ained wi~hout practicing the invention, is used to determine how much nitrogen~
free fluid is to be inj ected .

DETAILED DESCRIPTION OF THE INVENTION
':
The band formed by curves A and B in Figure 1 shows how the nitrogen content N of the melt varies . 1 with percent carbon C in the melt when the present ~
in~ention is not practiced. Although all BOP systems ~ :.
exhibit curves shaped similarly to Figure 1, the ~ ~ :
~... .
numerical relationship between N and C is specific to .
each BOP system and its manner o~ operation, and must be plotted from data obtained during actual production runs. The reasons for the variations from system to system are: variations in oxygen blowing rate, lance operating position, lance oxygen pressure, lan~e design, melt weigh~ vessel geometry, and so on. It can be seen that as the carbon content C decreases the nitrogen : .
-7- ;

~`

.

~~~~ D-10,883 content N also decreases until a minimum is reached, a~
which point the nitrogen content begins to rise again.
The nitrog n content of the melt~is used to determine when injection of the non-nitrogen fluid should begin in accordance with the present invention. However, since the nitrogen content is not often regularly measuredg as is carbon content, and since nl~rogen content is a function of carbon content for a given BOP vessel, as shown in Figure 1, the carbon content can be used to ;`
determine the nitrogen content.
From Figure 1 it can be seen that the nitrogen content o this particular system is at a minimum when the carbon content of the melt is approximately 0.08 percent.
Figure 2 shows how the off-gas flow rate F varies wi~h carbon content C for the given BOP refining system at a givan oxygen blowing rate without using the method of the present i~vention. Approximate off-gas flow rates can be determined wi~hout a flow meter by preparing a graph of carbon content versus time, determining the rate at which carbon is removed by the slope of the plot, and calculating the of-gas rate by assuming that the carbon removed is converted to carbon monoxide and that this carbon monoxide constitutes all of the off-gas. As with Figure 1, each BOP system will have , 10,883 . its own curve for ~his relationship depending upon system characteristics and manner of operatlon.
Wh~le we do not wish to be tied to any particu~ar theory, it i~ a hypothesis of this inventlon that nitrogen contamination in the basic oxygen process, occuring mainly during the lat~er s~ages of decarburization when the carbon content of the steel is low, is caused as follows. At high carbon levels the rate of carbon mono~ide generation during the oxygen blow or decarbur- ~ :
ization period produces off-gas ra~es sufficient to prevent significant infiltration of the surrounding atmosphere into the vessel. In addition, at high carbon levels, the carbon monoxide boil is sufficient to sparge some of the nitrogen that may be dissolved in the steel. ;
During the initial stages of decarburization therefore, the nitrogen level in the steel decreases, as shown in Figure 1. Beyond a eertain carbon level however, as the carbon con~ent drops, the nitrogen content of the melt increases. It is believed that the reason for such ~ ;:
increase is that as the carbon level drops, the rate of C0 formation by ~he ~ecarburization reaction and consequent ~:
off-gas evolution drops, making it possible for abmospheric ~ ;
nitrogen to enter the head space of the vessel and be absorbed by t~e melt. The oxygen jet helps carry the nitrogen down into the melt. Hence, as off-gas flow rate decreases, as shown in Figure 2, infiltration of :~
`. `':' _9_ 10,883 $~

. atmospheric nitrogen into ~he vessel is increased, and eventually a point is reached in which the nitrogen infiltrates at a rate sufficient to cause a net increase in the nitrogen content of the steel produced, Prsctice of the present invention will now be described with reference to Figures 1 and 2. From actual operating data one obtains N*, the minimum nitrogen con~ent a~tained during an oxygen blow for the particular system on which the invention is to be practiced.
In Figure 1 N* is about 19 to 25 parts per million. One : ;
then reads C*, the carbon content corresponding to N*.
From Figure 1 it can be seen that C* is 0.08%. Injection of the nitrogen-free fluid must be started no later than when the carbon content is C*. To de~ermine the :
rate of injection of nitrogen-free f~uid, one takes the carbon content at C* and reads on Figure 2 the off-gas flow rate, F* corresponding to C*. F* is the value ;.
below which the off-gas flow rate must not be allowed to fall during the refin~ng process. In accordance with : .
this invention, the off-gas rate is maintained above ;-.
this minimum value by maintaining the rate of injection of nitrogen-free fluid sufficien~ ~o maintain the total off-gas flow rate above F*. ~ :
In summary, from Figure 1 one obtains the latest ~.
point in time at which to begin injecting the nitrogen-fr~e fluid while from Figure 2 one obtains the minim~n : , .
- 1 0 - ' ' "

. .

- , ~ -. .: .

D-10,883 amount of nitrogen-free fluid that needs to be added in accordance with the present invention in order to prevent contamlnation of the melt with atmospheric nitrogen.
In some cases, precise instantaneous measurement of neither ~he carbon con~ent, nor the ni~rogen content of the melt is available during decarburization. It is therefore more convenlent to practiee the invention by start~ng injection of the nitrogen-free fluid some--what in advance of the time when the nitrogen content is squal to N* and the carbon content is C*. If a BOP
system has no means for constantly monitoring the off-gas flow rate or means for controlling the off-gas rate by varying the amount of nitrogen-free fluid that is injected into the vessel, the invention can still be practiced by introducing the nitrogen-ree fluid at a constant rate sufficient to mainta~n the total off-gas rate at least equal to F*.
It i~ not uncommon during practice of the basic ~. :
oxygen process to interrupt the injection o~ oxygen into .~:
the melt prior to achieving the final des~ red degree of decarburization. When this occurs it is necessary to ~`
reblow the melt. Similarly, it is also often necessary to rehl~w the melt even ~hough the inal desired carbon level has been reached, either because the temperature of the molten steel is too low, or because so~e other element or impurity is not at the desired level.

' ' -D-10,883 ~atever the reason, reblowing of the molten st~el ~ s not at all uncommon. When a ~elt is reblown during conventional practice of the basic oxygen process it is often accompanied by a sign~ficant increase in dissolved nitrogen content. The amount of this increase will vary.
Typical nitrogen pickup during con~entional reblowing is in the range of 2 to 10 ppm, with increases of up to 15 or 20 ppm not uncommon. Further, if se~Teral reblows in succession are requir~d, the final nitrogen l~vel may be as much as 80 to lO0 ppm higher than N* and 40 to 60 ppm higher than the maximum acceptable level for some grades of low-carbon, low-nitrogen steel.
It is believed that the reason for such high nitrogen pickup is that while refining is temporarily stopped, ~;
atmospheric nitrogen diffuses into the vapor or head-space of the vessel and is absorbed by the melt during the ~ubse~uent reblow. In accordance with this invention, nitrogen is removed from the vessel by purging the vessel with a nitrogen-free fluid, just prior to starting the reblow and by maintaining the off-gas flow rate no lower than F* during the reblow. While any amount of purging will be helpful it has been found that purging :
with a volume of gas (measured at 70F and atmospheric pressure) approximately equal to half the ~otal volume of the headspace of the vessel is sufficient to minimize ;
the nitrogen pickup by the steel during the reblow.
Purging with less inert gas is likely to be insufficient, ;~

D-10, 883 while purging with more is technically acceptable but uneconomical. It should be noted that if multiple reblows are required, the vessel must be purged prior to each reblow.
Argon is the preferred nitrogen~free fluid for use . -in the present invention. This gas has the advantages of being inert chemically, of being the laast expensive and most abundant of the chemically inert gases, vf being the leas~ disruptive to the thermal balance in the vessel, and also of favorably af~ecting the reaction of oxygen with carbon by diluting the effluent carbon monoxide. Other nitrogen-free gases can also be used, ~s well as liquids which vaporize readily a~ steel refining temperatures. Exampl~s of other nitrogen-free fluids include, but are not limited to: helium, neon, krypton, xenon, carbon dioxide, carbon monoxide, steam, wa~er, hydrogen, methane, liquid hydrocarbon~, gaseous hydrocarbons,or mixtures thereof, including mixtures with argon.
When using a flammable gas such as methane or hydrogen, special precautions should be taken to avoid forming an explosive mixture prior to ~njec tion into ~he refinin~ vessel. The flammable gas will, of course, rPact with oxygen i~ the vessel. This reaction ~ust be taken into account when calculating the amount of off-gas that will be produced for each quantity of D-10, 883 . flammable gas added.
In order to best attain the further benefits of minimizing the amount of metal oxidized, and of reducing the amount of oxygen dissolved in the melt, the preerred means for injecting the nitrogen-~ree fluld into the vessel is to mix it with the oxygen, i that can be accomplished without forming an explosive mixture. By using argon the possibility of creating an explosive mixture is entirely elimina~ed. By injecting the ~ :
nitrogen-free fluld admixed with oxygen, the invention may be practiced on existing BOP systems with very little in~estment since there is no need to add new injection equipment. I~ is po~sible simply to meter the nitrogen-free fluid into the oxygen line at some point upstream of the oxygen lance. However, ~ :
it is also possible to practice the invention by injec~ing the nitrogen-free fluid by a separate injecting lance, tuyere, or other injecting means located any place in the vessel, be it in the headspace, below the surface of the melt, or as a separate conduit within the oxygen lance.

-14- :

, .. , .. . - , . . .. .

--- D~10,883 The following examples will serve to illustra~e the practice of the present invention.

~PL~S
5everal steel heats werP reff ned by top blowing : :
in a BOP refining system havi~g the following ckaracteristics: .
Vessel volume 5000 ft3 Vessel mouth area 95 ft2 Total charge (pig iron and 235 tons scrap metal) . ;-Average amount of pig iron in charge 162 tons Average pig iron composition 4.5% c rbon :~
l~0~/o silicon ~ :
0~8% man8anese Nitrogen-~ree fluid ~rgon gas Oxygen blowing rate Withou~ 20,00Q ft3/min argon: (at 70F and 1 a~m) ;`
with 16,500 ft3/min argon: (70F and 1 atm) Off-gas tempera;ture 2900F :
The size of the lance limited the total flow rate of injected gas such that the oxygen blowing rate had ~o be reduced while argon was being injec:ted. The ln~ention is preferably practiced by maintaining a constant oxygen blowing rate throughout the entire heat.

... . , : . , , ,., , ~. .; . .. ~ ., ~~` D-10,8~3 The graphs relating nitrogen content and off-gas flow rate for this vessel with carbon content of the melt are chown in Figures 1 and 2. From the graphs i'c can be seen that the minimum nitrogen level, N*, occurs at a carbon content of approximately 0.08% and an of-gas rate of 15,000 ft3/min (measured at 2900F and :
1 atmosphere or pressure). Thus, in order to properly practica this invention, the latest point in t~.me for -~ .
introduction o~ nitrogen-free fluid into the vessel, is : :
at a ni~rogen conten~ of about 19 to 25 parts per million -or a carbon content of 0~08~/o~ The argon must be ~njected at a rate sufficient to maintain the off-gas rate at 15,000 ft3/min measured at 2900F and 1 a~mosphere, or ~ ~.
abou~ 2300 ft3/min measured at 70F and 1 atmo~phere.
Argon was introdueed into the BOP vessel via the oxygen lance by metering argon into the cxygen supply line upstream of the lance. Since a precise means to :: .
continuously measure the nitrogen or carbon content of :~
the melt during the refining process was not available, the argon flow was begun when the carbon content was :~
estimated to be between 0.10% and 0.15%. To maintain an off-gas rate of 15,000 ft3/min at 2900F, 3000 ft3/min of argon measured at 70F, or 19,000 ft3/min at 2900F, was injected. The extra gas was added to provide a s~fety factor in case all the argon was not heated to 2900F. Some runs were performed with argon:added at ;

. . ' : -. ' ' : :

D-10, 883 ~' ';, ', .
constant rate as low as 2000 ft3/min (at 70F and 1 ; ~
atm). These runs also gave satisfactory results. ::
Table 1 shows the results ob~cained UpOIl the f~rst stoppage of oxygen or first turn do~n, for heats ~-in which reblow~ng was not required prior ~o ~he tlme that argon was added to maintain the off-gas flow rate. ~ :
T ~
Heat No: 1 2 3 Argon rate (f t3 /min at 70F and 1 atm) 0 2000 3000 Duration of total ~:
oxygen blow ~minu'ces) 17 17 16 ~ :

Duration of argon injection (minutes)0 4.25 2.00 : ~
Temperature (~F)2880 2935 2890 ~:

Carbon content at ~ :
first turndown ~/O)0.03 0.03 0.03 Nitrogen content at ~
first turndown (parts ; ~ :
per mi1lion) 33 20 2~ :~
The results in Table 1 show the lower nitrogen ~ :
content obtained while practicing the inven~on in Heats No. 2 and 3 as compared with Heat No. 1, during which the invention was not practiced.
Table 2 illustrates the effect of purging the vessel prior to a reblow. In these heats argon was not introduced into the vessel prior to the first turn down. ~:
It was used to purge the ve~sel prior to the reblow and also added to the oxygen during each reblow. It 10,883 is evident that purging the head space followed by addition of argon to the oxygen during the reblow essentially eliminates pickup of nitrogen even when the carbon content is as low as 0 . 03% . Consider, for e~arnple, :
Heat No. 1 where the purpose of the reblow was to raise the melt temperature. The carbon content was 0,03V/~ both before and after the reblow- i.e. there was little or no carbon removal and hence there would, in the absence of argon, be little or no off-gas. Because the vessel was first purged with argon and then reblown with oxygen plus argon the total nitrogen pickup during the reblow was minus 1 ppm, i.e. the nitrogen level actually decreased. At this low carbon level one would anticipate a nitrogen pickup of at least 5 ppm if argon purging and argon addition during the reblow had not been practiced.
Heat No. 4 is an example of a heat where multiple reblows were required. Argon purging was used prior to ;
each reblow and argon was added to the oxygen during ~ ;~
each reblow. Again it is evident from the results shown in Table 2 that the addition of argon in accordance witb this invention resulted in a cumulative nitrogen pickup of minus 3 ppm (i.e. a nitrogen decrease3 after four consecutive rebl~ws. Normally, at these low carbon ;~
levels in the absence of argon additlon, one would anticipate a minimum cumulative nitrogen pickup of about 20 ppm after 4 reblows, and a total pickup of 40 to 60 ~ ;
ppm would not be u~usual.

.

, . , -~ ~ ~ In o ~ o o o u~
c~
E~
~ -c`~ o æ c ~:4 ~ c~
c~
~ ~ ~ ^
æ ~ E~
~ ~ ~ ~ ~ ~ ~ ~ ~1 ':

~ _~ ~ ~ ~0 ~ O ~ O
O 1 C~ G ~ I O
_l O

:
.~ ~ O ~
~3 ~ ~ ~ ~ o O O O :
:~
~ ~- ~::
e 3 J~ e ~ O O O O O O ~:

I . ;~.

3 P~3 1 o o o O o O O
~ ~oO ~ O ~ ., ~ ¢~ 8 : :
:q a _ ~ lo g o o o g g ~ ~
o _ o o ~ o o o o oo o _~ ~ U~ O
C~l ~
~3 :

.
a~ .
6q :
o~
æ -19 Table 3 illustrates the results of practicing the inven~ion when it is necessary to reblow a heat af~er argon addition to maintain the minimum o~-gas flaw rate prior to first turn down. In Heat No. 6, argon flow was initia~ed at a rate of 2000 SCFM 390 seronds prior to the first turn down. A~ turn down the temperature was 2950F, carbon 0.13% and nitrogen 16 ppm. The vessel w~s then purged with 2500 SC~ of argon and reblown for 60 ~econds with 16,500 SCFM oxygen and 3000 SCFM argon. After 60 seconds the temperature was 2860F, ~:
carbon was 0.07% and nitrogen was l9 ppm. The vessel was again purged with 2500 SCF argon and again reblown for 60 seconds with 3000 SCFM argon and 16,500 SCFM
oxygen, and at turn down the temperature was 2910F, carbon was 0.04% and nitrogen~ l~ ppm. Total nitrogen pickup during the two reblows was 2 ppm. The heat was then tapped.
Heat No. 7 i5 similar to Heat No. 6 excep~ that only one reblow was required, and the nitrogen pickup was minus 2 ppm, i.e. the n~trogen le~el decreased. :
;' ~'''"' -20- ~

a~
-~ ~
~ --~
~ Q) ~ o ~ o o o a\ 3 p, o u~
v o E;--~ QC~
~1 ~ N
¢ P~
~: _ Z g ¢

i~ ~ ~D a~ I ~ .
æ ~ P~
~ O ~_ Iq ~7 .

p ~ I~ ~ ~ ~
~ ~_ ~ o o o o J~ ~
~q ~ ~ ~ o o o 3 ~1 0 : .
O ¢ ~
.CI

O ~ ~ ,. l _~ O O
~4 O ~ ~
O O
a~ o ~ ~ C~
~0 ~ . .,~
O o o U~ C~
:~+ O ~ ~ ~ ~D C`l .D
E~bO ~ a~ ~ e~
~ O ¢ p,_ )-I ' ~ . : .
Q~
~ ~C`I
~a o ., en ~ o o o o o Q~ ~0~ O O o O o ~ ~ ~ o ~ e~

= 30 1 ~ ~
~-~ 3 a~ ~ o _ I o o I o00 0--I ~ O ~ O

O
I _I
a) o ~
:r: æ -21-... .

Claims (6)

WHAT IS CLAIMED IS:

1. In a process for the production of low carbon steel by blowing oxygen into a melt contained in a vessel from above the surface of said melt wherein the oxygen blow has been interrupted before the nitrogen content of the melt has achieved its minimum level, the improvement comprising:
(a) purging said vessel headspace by injec-tion of a nitrogen-free fluid immediately prior to restarting the injection of oxygen, and after restarting the injection of oxygen, (b) introducing nitrogen-free fluid before the nitrogen content in the melt has substantially achieved its minimum, (c) adjusting the flow of said nitrogen-free fluid so that the total off-gas flow rate from the vessel is maintained at least equal to that which would have been produced without said nitrogen-free fluid, at the time in the refining process when the nitrogen content of the melt reached its minimum level, and (d) continuing the injection of said nitrogen-free fluid substantially throughout the remainder of the oxygen blow.

2. The process of claim 1 wherein the nitrogen-free fluid is argon.

3. The process of claim 1 wherein the mitrogen-free fluid is injected admixed with the oxygen.

4. The process of claim 1 wherein the purge con-tains a volume of gas measured at 70°F and 1 atmosphere substantially equal to half the total vessel headspace.

5. The process of claim 1 wherein the nitrogen-free fluid is injected during the oxygen reblowing at a constant rate at least equal to the off-gas rate obtained at the time in the process when the nitrogen content of the melt has achieved its minimum.

6. The process of claim 1 wherein the steel produced has a carbon content below 0.10% and a nitrogen content below 50 ppm.