Hydrogen pick-up after vacuum degassing

J. B. Hurtig and D. Sichen*

The present work aimed at determining the major source for hydrogen pick-up of the steel after
vacuum degassing but before casting. Samples of slag and metal were taken at different stages
during ladle treatment at SSAB. Hydrogen increase after vacuum treatment was observed.
Moisture contents of the industrial slags were analysed and their water capacities were
calculated. It could be seen that the hydrogen increase was correlated to the amount of moisture
in the slag and the water capacity. The study showed that the slag containing most water was also
the heat having the largest hydrogen increase. The slag with most water had the highest water
capacity. It could be concluded that the major source for hydrogen increase after degassing was
due to the slag–metal reaction.
Keywords: Hydrogen, Steel making, Vacuum treatment, Hydroxyl capacity, Ladle, Industrial slag

Introduction respect to hydrogen absorption, the major source for

High end steel grade demands low levels of sulphur, nitrogen,
oxygen and hydrogen as well as inclusion control. Hydrogen
is one of the elements, the content of which is reduced during
the degassing process. A number of publications on the
influence of different factors on the dehydrogenation process
can be found.1–8 Konolov et al.9 showed that a two-stage
neutralisation of the slag reduced the hydrogen content by
12% in the steel. Jauhiainen et al.10 investigated four different
stirring methods concluding that combined gas and down-
ward induction stirring was beneficial for dehydrogenation.10
It has also been investigated whether the hydrogen control
should be started from the BOF (Basic Oxygen Furnace).11
These authors reported that ladle slag played an important
role and it should be beneficial to have a slag with low
basicity11 and high viscosity12 in order to achieve high
dehydrogenation and low resupply of hydrogen from the
slag to the liquid steel. However, for many steel producers a
high basicity slag is of importance to reach low sulphur
concentration and having too high viscosity in a slag creates
practical problems during steel production.
After vacuum treatment it is essential to minimise
hydrogen resupply from the surroundings. In SSAB it was
noticed that the hydrogen levels after vacuum had a large
variation from below 1 ppm up to almost 3 ppm at
departure from the ladle station. This scattered behaviour
has also been reported by other steel producers.12 Further-
more, the hydrogen concentration in the worst heats at the
caster increased with 2 ppm from the value measured at
the ladle station. Therefore, a systematic study is currently
carried out in order to stabilise the hydrogen values from
the ladle and to reduce the increase at the caster.
The possible sources of hydrogen resupply are alloys,
slag, refractory, furnace atmosphere and possibly the
purging gas. To optimise the ladle treatment with
Department of Material of Science and Engineering, Royal Institute of
Technology, SE – 100 44 Stockholm, Sweden
*Corresponding author, email sichen@kth.se
hydrogen increase needs to be identified. The purpose of
the present work is to investigate how much the different
contributors affect hydrogen coming back to the steel
after vacuum degassing. Major focus of this study is to
clarify the slag contribution. For this purpose, sampling
of liquid slag and steel were carried out at different
stages during ladle treatment at SSAB, Oxelösund,
Sweden. These data were analysed along with the data
of moisture level with the help of the water capacity
model developed by Brandberg et al.13
Experimental
Process description
Industrial trials were carried out in the 200 t BOF plant of
SSAB in Oxelösund, Sweden. The main production route is
presented schematically in Fig. 1. During tapping, alloying
is carried out. Thereafter, the steel is transferred to a TN
(Thyssen Niederrhein) station, where it is deoxidised and
desulphurised. After TN treatment the slag is removed and
the ladle is transferred to an integrated LMF-VTD (Ladle
Metallurgical Furnace-Vacuum Tank Degasser) station.
The ladle has two bottom porous plugs for introducing
purging argon gas. After securing the function of the
porous plugs, slag builders are added. These consist of lime
and a mixture of mostly Al2O3 and MgO. The oxides are
kept dry until they are put into the alloying bin. Samples
are taken once a day to ensure low moisture content. At the
LMF, fine adjustments of temperature and composition are
carried out. The liquid steel is transferred to the tank
degasser for vacuum treatment. During degassing, addi-
tional alloying can be performed depending on steel grade.
Vacuum degassing is carried out in order to reduce sulphur,
nitrogen and hydrogen concentrations in the liquid metal
but also for inclusion control. The flowrate of the purging
gas is controlled and also monitored using a camera placed
in the LRF/VD (Ladle Refining Furnace/Vacuum
Degasser) roof. After the degassing process, if necessary,
the ladle returns to the LMF station for trimming

ß 2015 Institute of Materials, Minerals and Mining

Published by Maney on behalf of the Institute
Received 26 November 2013; accepted 22 April 2014
DOI 10.1179/1743281214Y.0000000199 Ironmaking and Steelmaking 2015 VOL 42 NO 1 49
Hurtig and Sichen Hydrogen pick-up after vacuum degassing
1 Schematic illustration of production route at SSAB in Oxelösund
temperature and composition before departure to the
continuous casting machine.
Sampling
The ladle treatment of the industrial trials can be divided
into three process steps, namely before vacuum (BV)
treatment, vacuum treatment and after vacuum (AV)
treatment. The last possibility for sampling at the
secondary metallurgy station was before departure
(BD) of the steel to the casting. Steel and slag samples
were taken along with hydrogen measurements at
different stages during the ladle treatment according to
Table 1. The hydrogen content was measured manually
with Hydris. Seven heats were studied with extra
sampling after vacuum treatment, all from the same
steel grade. Many additional heats were also studied,
where hydrogen analyses were taken after vacuum
degassing and before departure to statistically examine
the effects of different factors on hydrogen absorption.
In order to clarify if the hydrogen pick-up was from
the slag, heating or additives, the hydrogen content was
measured before vacuum degassing, [H]BV, as soon as
possible after vacuum treatment, [H](AV–1) and again
7 min after the vacuum treatment but before any heating
or additions had been made [H](AV–2). It was ensured
that the stirring intensity was low without any open eye
formation during this period. In this way, any other
source of hydrogen except slag was minimised. These
samples could also indicate how fast the transfer of
hydrogen from the slag back into the metal took place
after the degassing process. Hydrogen content was also
measured before departure, [H]BD, of the steel after final
adjustments of temperature and composition.
2 Hydrogen concentration as a function of time after
vacuum treatment
50 Ironmaking and Steelmaking 2015 VOL 42 NO
The steel samples and temperature measurements
were taken with an automatic sampler and the slag
samples with a sampling scope. The slag was rapidly
cooled and sealed immediately in glass bottle. The
composition of steel and slag as well as the moisture
content was evaluated by the chemical laboratory at
SSAB. The moisture content in the slag was determined
by a thermogravimetric method. The slag sample was
heated to 873 K and held at this temperature for 2 h in
an argon atmosphere. The weight loss was assumed to
correspond to the moisture content in the slag.
Results
Figure 2 presents the hydrogen concentration as a
function of time after degassing for heat 41, 44–45 and
51–54. These heats are denoted as A–G to help later
discussion. From the figure it can be seen that the
hydrogen increase was faster just after vacuum degas-
sing. However, it must be pointed out that the
measurements were not made continuously.
Figure 2 indicates that the hydrogen concentration
after dehydrogenation varies from heat to heat. Jha
et al.11 reported importance of hydrogen control starting
from the BOF. According to these authors, the initial
hydrogen concentration before degassing is significantly
important for the final concentration. Therefore, Fig. 3
presents the hydrogen concentration directly after
vacuum treatment, [H](AV–1) as a function of the
hydrogen concentration before vacuum treatment
[H]BV. In contrast to Jha’s suggestion, there is no clear
relationship between the start concentration and the
concentration directly after vacuum treatment in the
heats studied.
3 Hydrogen concentration [H]AV–1 as a function of hydro-
gen concentration before vacuum treatment
1
 Hurtig and Sichen Hydrogen pick-up after vacuum degassing

Table 1 Sampling during ladle treatment

Sample BV AV-1 AV-2 BD

Steel Heat 1–54 Heat 1–54 Heat 41, 44–45, 51–54 Heat 1–54
Slag 0 heats 0 heats Heat 41, 44–45, 51–54 0 heats
Hydrogen Heat 1–16 Heat 1–54 Heat 41, 44–45, 51–54 Heat 1–54

It could be expected that the absorption of hydrogen
should be higher when the concentration after degassing
is low. Therefore, hydrogen resupply after vacuum
degassing is plotted as a function of the hydrogen
content measured directly after vacuum treatment in
Fig. 4.
The hydrogen resupply, denoted as DH1, is defined as
the difference between the second and the first hydrogen
sample taken after vacuum treatment according to
equation (1)
DH1 ~½H AV{2 {½H AV{1 (1)
Figure 4 shows no clear relationship between the
hydrogen resupply and the hydrogen concentration after
vacuum treatment. It had been expected that the
hydrogen increase would be higher when ½H AV{1 is
low, since the driving force for the liquid steel to absorb
hydrogen is larger. However, this expectation is ruled
out by the results shown in Fig. 4. The hydrogen supply
seems to have a superior relevance.
The hydrogen resupply between the second sampling
after degassing and steel departure is defined as
DH2 ~½H BD {½H AV{1 (2)
of the slag. The slags containing most water are from the
same heats that have the highest hydrogen increase.
Discussion
During vacuum degassing, the hydrogen content in the
steel is decreased due to high argon flow combined with
low pressure. The good contact between the rising argon
bubbles and the steel greatly enhances the removal of
hydrogen from the liquid steel to the gas bubbles. In this
period, an open-eye is formed. Since most of the gas
escapes from the bath to the gas phase through the
open-eye, the contact between gas and slag is limited.
While hydrogen would be transferred from the slag to
the liquid metal due to the low hydrogen content in
liquid steel, the efficient removal of hydrogen by gas
bubbles would still reduce the hydrogen content in the
metal.
After vacuum treatment, the total pressure in the ladle
is increased and the argon flow is considerably
decreased. In this period, there is no longer a driving
force for dehydrogenation. Instead, as long as OH2 ions
are present in the ladle slag, hydrogen is transferred
from the slag back into the steel bath.14 The resupply of
hydrogen of the steel from the slag can be described by
reaction (4)

In Fig. 5a–d, DH2 is plotted as functions of ladle life, 2HzOzO2{ '2OH{ (4)
heating time, alloy additions and time to departure,
respectively.
The underline in the above equation indicates that the
Note that the number of data points for alloying with
element is dissolved in the liquid steel. Note that the
more than 300 kg is limited. There is no clear trend in
reaction takes place in the reverse direction, from right
any of the figures. This has also been reported
to left.
previously.8 Since the alloys are added as one batch, it
The equilibrium constant of reaction (4) is expressed
is impossible to separate the contribution of each alloy.
as
Nevertherless, it can be concluded that in total there is
no significant correlation of hydrogen resupply because a2OH{ ðwt-%OH{ Þ2 :fOH2
{
of alloy addition. K4 ~ ~ (5)
aH :aO :aO2{
2 2: 2 :
H
:
½wt-%H f aO a 2{ O
The moisture contents of the slags taken from the
seven heats are listed in Table 2 along with the slag
compositions. It can be seen that the moisture content in
the slag varies considerably.
In order to investigate the correlation between the
moisture content in the slag and the hydrogen absorp-
tion in the liquid steel the relative hydrogen resupply
after vacuum treatment is plotted as a function of the
moisture content in the slag in Fig. 6.
The relative hydrogen increase, DH3R, is defined as
½H AV{2 {½H AV{1
DH3R ~ |100 (3)
½H AV{1

A strong correlation between the hydrogen increase

and the amount of moisture in the slag is brought out by
the experimental observation shown in Fig. 6. The
hydrogen absorption increases with the moisture content 4 Hydrogen pick-up after vacuum treatment as a function
of hydrogen content just after vacuum treatment

Ironmaking and Steelmaking 2015 VOL 42 NO 1 51

Hurtig and Sichen Hydrogen pick-up after vacuum degassing

5 Hydrogen increase as a function of a ladle life, b heating time after vacuum treatment, c amount of alloy addition after
vacuum treatment and d time from end of vacuum treatment to departure

Equation (5) can be rearranged as Hydroxyl capacity is expressed as

ðwt-%OH Þ fH{
1=2
1=2
~ K4 :aO :aO2{ (6) (wt-%OH) K8 :aO2{
½wt-%H fOH COH ~ ~ (9)
ðPH2 O Þ1=2 fOH{

The partition coefficient for hydrogen is defined by

Combination of equations (5) and (9) leads to
(wt-%OH)  : 1=2
L~ (7) (wt-%OH) K7 a O
½wt-%H L~ ~ :COH :fH (10)
½wt-%H K11

The hydrogen exchange between slag and metal is

controlled by the partition coefficient (L) and the mixing From the industrial perspective, it is more common to
fraction between slag and metal. Since the stirring use the term water capacity instead of hydroxyl capacity.
intensity was kept minimal during the industrial trials, Water capacity, CH2 O , is defined as
this mixing fraction can be assumed to be approximately
the same in different heats and therefore it is the
partition coefficient that differs between the melts.
The hydroxyl capacity is defined based on reaction (8)
H2 OzO2{ '2OH{ (8)

Table 2 Slag compositions and moisture data of sampled

heats/mass-%

Heat Al2O3 CaO MgO SiO2 H2O in slag

A 40.9 49.8 6.9 2.4 0.28

B 36.3 54.9 6.3 2.5 0.18
C 37.5 53.7 6.5 2.3 0.05
D 32.3 41.1 17.0 9.5 0.42
E 33.5 47.6 9.2 9.7 0.05
F 39.9 50.9 6.7 2.5 0.1 6 Relative hydrogen pick-up of steel after vacuum treat-
G 39.5 50.6 7.2 2.6 0.34
ment as a function of moisture content of slag

52 Ironmaking and Steelmaking 2015 VOL 42 NO 1


Table 3 Water capacities (ppm H2O atm21/2)
sampled heats
Heat CH2 O
A 1761
B 1631
C 1626
D 1828
E 1404
F 1648
G 1687
ppmH2 O 104 : MH2 O
CH2 O ~ ~C {:
OH
ðpH2 O Þ1=2 2 MOH{
where ppm H2O is the water content in the slag in ppm
unit and PH2 O is the partial pressure of water.
Since the hydrogen concentration in the liquid metal,
[wt-%H] is reduced by vacuum degassing and the
degassing would not reduce the OH2 concentration in
the slag considerably, hydrogen is supplied back from
slag to the steel to meet the thermodynamic constraint,
namely equation (4).
It is often believed in the industry that the lower the
hydrogen level is after vacuum treatment the higher the
increase would be after vacuum. However, the industrial
trials reveal that the hydrogen increase after vacuum
correlates to the moisture content in the slag rather than
to the absolute value after vacuum treatment. This
implies that the slag composition is a factor that has to
be considered.
It is possible to calculate the hydroxyl capacity (or
water capacity) of the slag based on its composition
using a model developed for the quaternary slag system
Al2O3–CaO–MgO–SiO2 by a previous study.13 The
water capacity model considers a silicate melt that
consists of cation grouping and oxygen ions. The water
capacity is assumed to depend on the binary interactions
between the cations in the presence of oxygen ions. It
has been shown that the water capacity increases with
the increasing CaO content keeping MgO between 5 and
10 wt-%. Within this range the increase is irrespective to
the MgO concentration. Furthermore, keeping CaO and
MgO constant at 50 and 10 wt-% respectively the water
capacity increases by replacing SiO2 by Al2O3. The
calculated water capacities for the sampled heats are
presented in Table 3.
7 Relative hydrogen pick-up of steel after vacuum treat-
ment as a function of calculated water capacity for
sampled heats
Hurtig and Sichen Hydrogen pick-up after vacuum degassing
of the
(11)
8 Relative hydrogen pick-up of steel after vacuum treat-
ment as a function of basicity
In Fig. 7, DH3R is plotted as a function of the
calculated water capacity. It is seen that the hydrogen
resupply to the steel after vacuum treatment increases
with the water capacity of the slag. The slag containing
most water also is the slag with the highest water
capacity. The heat with the highest increase in hydrogen
concentration has the slag with the highest content of
moisture and the highest water capacity. The heat that
does not have any hydrogen increase having the slag
with the lowest content of moisture and the lowest water
capacity. The amount of water adsorbed by the raw
materials is dependent on the humidity of the surround-
ing atmosphere as well as on the handling of the raw
materials in the steel plant. Jha et al.11 indicated that the
slag basicity should be kept low in order to decrease
hydrogen increase. To examine this aspect, the DH3R is
plotted as a function of basicity in Fig. 8.
For the slag compositions studied in the trials, there is
a weak correlation between the hydrogen resupply and
basicity in the composition range study. While the
variation of the basicity is small, the variation in water
capacity of the slag significantly influences hydrogen
supply. This implies that good sulphur refining slag that
has high basicity does not necessarily have strong impact
on hydrogen pick-up. In this study, the slag with the
highest water capacity could contribute with 2?5 ppm
hydrogen in the steel while the slag with the lowest water
capacity holds a potential of 0?3 ppm increase.
Dor et al.12 have reported that the water content in
the slag after vacuum treatment could be 200 times
higher than the predicted value from the metal–slag
equilibrium.12 The reason for this is that the dehydro-
genation kinetics in steel is much faster than it is in
slag.14 This study supports their conclusions. Hence, the
concentration of hydroxyl ions in the ladle slag can be
the major source for hydrogen transferring back into the
liquid steel after vacuum degassing.
According to Dor et al.12 high viscosity slag results in
better dehydrogenation and low resupply of hydrogen
from the slag to the liquid steel. Their observation is in
accordance with the present finding. Slag with high
viscosity would lead to lower mass transfer rate between
steel and slag and therefor slower hydrogen resupply.
Previous studies have shown that the slag phase does
not affect the hydrogen removal kinetics during vacuum
degassing treatment. Since both steel and slag are
protected from the surrounding atmosphere during
vacuum degassing, there is no water transferring from
Ironmaking and Steelmaking 2015 VOL 42 NO 1 53
Hurtig and Sichen Hydrogen pick-up after vacuum degassing
the atmosphere to the slag.15 However, after vacuum
degassing the slag is not protected from the surrounding
atmosphere and there is a constant supply of hydrogen
from the moisture in the air to the steel through the slag
layer. If slag having high water capacity is employed, a
greater gradient of OH2 through the slag would be
expected. The big gradient of OH2 associated with high
water capacity slag would in turn enhance the hydrogen
transfer from the surrounding to the steel bath, thereby
increasing the hydrogen concentration of the steel.
Furthermore, it can be assumed that the hydrogen
increase directly from atmosphere is limited but the
basic slag layer acting as transferring agent.
The result from this investigation shows that the
supply of hydrogen from the slag is superior to the other
factors, namely refractory, heating time after degassing
as well as time from degassing to casting since the major
increase comes within 5–10 min. The strongest influence
on the hydrogen increase is the slag. It seems that the
increase is not substantial when CH2 O ,1650. Above this
water capacity, the increase becomes dramatic. Since the
hydrogen increase after vacuum treatment indicates low
correlation to slag basicity, it is possible to have slag
beneficial for sulphur refining and still minimise hydro-
gen increase. In this study, a slag contains 30 wt-
%Al2O3, 50 wt-%CaO, 10 wt-%MgO and 10 wt-%SiO2
had the lowest water capacity values and one of the
lowest moisture concentration in the slag. When
designing the slag composition water capacity should
also be considered.
Conclusions
Steel and slag samples were taken at different stages of
ladle treatment. From these samples, it was found that
the increase in hydrogen in liquid steel was dependent on
the moisture content in the slag phase. Water capacities
of the sampled slags were calculated in order to examine
whether ladle slag was the major source for increasing
hydrogen content after vacuum degassing. Furthermore,
the water capacity was correlated to moisture content of
ladle slag. The slag with the highest concentration of
water also had the highest water capacity and associated
with the highest hydrogen increase in the steel after
vacuum treatment.
After vacuum degassing there is a constant supply of
water to the slag layer. If a slag with high water capacity
is used, it could raise the rate of hydrogen transferring
54 Ironmaking and Steelmaking 2015 VOL 42 NO
back into the liquid steel. It is possible to design a slag
with high basicity and low hydrogen capacity.
Acknowledgement
The authors are grateful to SSAB EMEA, Oxelösund
for supporting the industrial sampling and analysing.
References
1. N. Bannenberg: ‘Demands on refractory material for clean steel
production’, Unitec ’95 congress proceedings: Global development
of refractories. UNITECR Kyouto, Japan, 36–59, 1995.
2. S. Misra, R. P. Stone, M. Kan and R. J. Fruehan: ‘Hydrogen and
nitrogen pickup from ladle additions’, Iron Steel Technol., 2006, 3,
(3), 236–245.
3. D. Nolle, U. Euelenberg, A. Jahns and H. Miska: ‘Application of
ladle stand degassing in the compound LD Steelmaking and
continuous casting process’, Stahl. Eisen., 1989, 109, (19), 107–110.
4. E. F. Mazurov, V. V. Shakhnovich, A. F. Kablukovskiy and S. A.
Kiseleva: ‘Refining of steel in a 150 ton ASEA-SKF ladle furnace
system’, Sov. Mater. Sci. Rev., 1987, 1, (2), 187–191.
5. F. Steiner: ‘Ladle furnace Cum vacuum treatment process for
production of high quality steel–some lessons’, Tool Alloy Steel,
1985, 19, (1), 31–37, 26, 12.
6. S. Misra and R. J. Fruehan: ‘Hydrogen and nitrogen control in
ladle and casting operations’, Iron&Steel Society international
technology conference and exposition 2003; Indianapolis; USA;
27–30 A pr. 2003, pp. 873–882.
7. M. Hallberg, T. L. I. Jonsson and P. G. Jönsson: ‘A new approach
to using modeling for on-line prediction of sulphur and hydrogen
removal during ladle refining’, ISIJ Int., 2004, 44, (8), 1318–1327.
8. N. P. Kojola and J. A. E. Hurtig: ‘Nitrogen and hydrogen refining
during vacuum treatment of liquid steel’, Proc. Int. Symp. on ‘The
recent developments in plate steels’; 2011, Warrendale, PA, AIST.
9. V. S. Konolov, Y. U. B. Kushnir, Y. U. G. Korinovskii, S. I.
Lavrenko and L. M. Bagaeva: ‘A process for the deoxidation of a
wheel steel involving two-stage neutralization of the slag in
Furnace and in Ladle’, Metall. Gornorudn. Prom-st., 1982, 3, 15.
10. A. Jauhiainen, L. Jonsson, P. Jönsson and S. Eriksson: ‘The
influence of stirring method on hydrogen removal during ladle
treatment’, Steel Res., 2002, 73, (3), 82–90.
11. K. N. Jha, M. K. Sardar, N. N Jha and S. Chakraborty: ‘Hydrogen
control during steel making for medium carbon wheels’, Scand. J.
Metall., 2003, 32, 296–300.
12. P. Dor, B. Carrier, M. Nadif and C. Gatellier: ‘Influence of slag on
steel dehydrogenation under vacuum, Rev. Metall./Cah. Inf. Tech.,
1988, 85, (4), 307–316.
13. J. Brandberg, L. Yu and D. Sichen: ‘Water capacity model of
Al2O3-CaO-MgO-SiO2 quaternary slag system’, Steel Res., 2007,
78, 460–464.
14. T. Rosenquist: ‘Principles of extractive metallurgy’, 2nd edn; 1983,
London, McGraw-Hill Book Co.
15. J. Brandberg: ‘Process model for dehydrogenation during vacuum
degassing’, 2006.
1