The Use of Hydrogen in Shielding Gases for Weld Overlays

Part I: Benefits and Concerns

ARC Specialties has conducted a study looking into the effects of hydrogen additions in GTAW shielding gases for the
welding of high strength steels. The results of this study are given in two reports. Part I, covers the benefits and
concerns of using an argon/hydrogen blend shielding gas. Part II, covers the effects of hydrogen on the weld metal
mechanical properties and weld metal inclusions. Readers are encouraged to go to the ARC Specialties web site
(https://arcspecialties.com/resources/white-papers/) and download Part II of this report. “The Use of Hydrogen in
Shielding Gases - Part II: Effects on Base Metal”

Introduction:
When the suggestion of using a welding shielding gas containing hydrogen (H) is offered, what is the normal reaction?
There are likely several responses, with most of them being negative. We’re not surprised to hear comments like,
“What about the potential for fire or explosion?” or “What about hydrogen-induced cracking (HIC)?”. Because of these
negative responses and the inability to overcome widespread prohibition by the welding academic/technical
community, it was realized that a study of the effects of hydrogen additions to shielding gases was needed to either
confirm or dispel these concerns. While it is understood there could be potential issues regarding the use of hydrogen,
it is also well known there are potential benefits, as discussed below.
Such a study is especially interesting to the technical community involved with the use of the gas tungsten arc welding
process with hot wire filler metal addition (GTAW-HW) for the welding of corrosion-resistant overlays (CROs) on high
strength steels like 4130, 8630 and F22, all of which are common alloys used for components in the upstream sector
(production) of the oil and gas industry. The CRO alloy of choice for many of these “oil patch” applications is INCONEL
625 (625).

Hydrogen will probably always be infamously known for its risk of explosivity and flammability ever since the
Hindenburg disaster shocked the world in 1937. While it is true hydrogen is highly flammable, hydrogen's flammability
limit occurs at 4% concentration, a percentage much greater than the amount of hydrogen used in the shielding gas
mixture that will be suggested in this study (around 2%). It should also be noted that gas mixtures with up to 15%
hydrogen are used in seamers for welding of thin wall stainless steel tanks and tubing for improved penetration and
increased travel speed, without any safety concerns. So the concern for fire or explosion is immediately eliminated.

Hydrogen-induced cracking (HIC) however, is a more legitimate concern among metallurgists familiar with its potentially
catastrophic consequences. Feeding this fear is history found in research papers, presence of cracking in process
components resulting from the presence of hydrogen in conjunction with other contributing factors, and the constant
requirements found in fabrication standards for low hydrogen welding consumables and practices.
One of the more vocal groups arguing against the use of hydrogen in shielding gases are those in the oil and gas
production community, especially those on the Gulf coast. Damaging effects of hydrogen sulfide on carbon and low
alloy steel weldments are well known, and respected, by this group. The risk of hydrogen-induced stress corrosion
cracking is real, which is why the use of corrosion-resistant overlays (CROs) applied using welding on oil patch
components is so important. Without this protective layer, steel components would fail prematurely potentially causing
serious damage, including potential safety and health risks to personnel.
With all of these negative arguments related to the use of hydrogen, why even bother exploring the topic further?
When using the GTAW process, the fact is exposure to an oxidizing environment will degrade the tungsten electrode and
have a negative effect on the welding process and subsequent weld quality. Consequently, inert shielding gases such as
argon (Ar) and helium (He) are required, with argon being the most common choice. When helium is used, more heat is
generated by the welding arc allowing for improvements when welding thick sections or when there is a desire to weld
at higher travel speeds. However, helium costs significantly more than argon, and availability can be an issue.
Consequently, argon is by far the gas of choice for typical CRO welding applications in the oil patch.
Compared to argon and helium, hydrogen is reactive as opposed to inert. This reaction, however, creates a reducing
atmosphere in the weld zone resulting in the reduction of oxides. Additionally, hydrogen has the capability for increased
heat transfer, six (6) times that of argon. Hydrogen readily combines with any oxygen present in the weld zone. Similar
to helium, the addition of hydrogen adds heat to the weld pool, but it also cleans or deoxidizes the weld metal. Perhaps
more importantly, these benefits can be realized with only minor additions of hydrogen to either argon (most common)
or argon/helium mixtures.
With these characteristics, it is easy to see how hydrogen-bearing shielding gases can offer a number of advantages in
terms of both weld quality and productivity. Some of those benefits include:
 Higher heat input allows for the use of faster travel speeds, with better edge-wetting, even at lower welding
currents.
 Produces a cleaner weld surface. This is increasingly more important when making multi-layer welds in small
bores where cleaning between layers is impractical.
 Produces cleaner weld metal, reducing the potential for creating subsurface oxide inclusions which could
subsequently result in surface defects on machined surfaces, requiring repair.
 Removal of oxygen contamination in the shielding gas. This is evidenced by the presence of a clean electrode.
As a result, longer weld times are possible without deterioration of the electrode, and the subsequent need to
interrupt the operation to re-dress the tungsten electrode.

Two welds made with the same “Dirty” wire at the same welding parameters, both have two (2) layers of 625 wire.

Top weld - was made using argon shielding gas.

Note the irregular edges and oxide spots on
surface of weld.

Bottom weld - was made with 98% argon/2%

hydrogen shielding gas. Note the clean straight
edges on weld beads and the much cleaner
weld surface.

Electrode contamination after Electrode after 67 minutes of

67 minutes of arc time using arc time using argon/hydrogen
argon shielding gas, top weld shielding gas, bottom weld

History of hydrogen in welding:

The use of a hydrogen component in a shielding gas for welding is certainly not a new concept. In fact, hydrogen was
the first gas used for shielding of an arc welding process. Introduced in the 1930s, the atomic hydrogen welding (AHW)
process used hydrogen for a shielding gas. It produced higher welding temperatures than any of the welding processes
available at the time. Not only does it have a high temperature arc, but it has a very high thermal conductivity. The heat
of the arc ionizes the hydrogen, separating the atoms of the hydrogen molecule. When the ionized gas cools in the weld
zone, the atoms recombine, releasing the absorbed energy. Another beneficial characteristic is the reducing potential of
hydrogen. Oxygen has a greater affinity for hydrogen than molten steel, so if oxygen-bearing contaminants are present
in the arc atmosphere, they will combine with the hydrogen and evolve instead of becoming surface oxides or
entrapped oxide inclusions. Atomic hydrogen welding has been used for welding tool steels and hard facing of high
strength steel rock bits for decades with no detrimental effects. Through the years, hydrogen has been added to
shielding gases for GTAW, GMAW, and PAW to take advantage of its high heat transfer and reducing atmosphere. When
one observes a weld made with hydrogen additions, it is clearly evident the surface is cleaner. Figure 1 is an example of
an INCONEL 625 CRO welded with argon/hydrogen shielding in its as-welded condition.

Figure 1. Two-layer, GTAW-HW ERNiCrMo-3 Corrosion Resistant Overlay welded using 98% Ar/2% H2 shielding gas
To better understand the concerns related to the presence of hydrogen during welding, it is appropriate to briefly
discuss the conditions leading to hydrogen-induced cracking (HIC). While the presence of hydrogen is a necessary factor
for this damage, there are other contributing elements. For HIC to occur in steel weldments, three (3) conditions must
exist: 1) the presence of a crack-susceptible microstructure (HAZ hardness greater than 20 HRC); 2) the presence of
applied stress (residual stresses from welding are sufficient); and 3) the presence of hydrogen. For the applications cited
here where CROs are welded to hardenable steels such as 4130, 8630 and F22, the presence of a high HAZ hardness and
residual stress are expected. One might ask with those two (2) factors present, how can HIC be avoided when hydrogen
is present as a component of the shielding gas? To answer that, we need to understand what happens to the hydrogen
during the welding operation.
As mentioned above, some of the available hydrogen is converted to monatomic hydrogen which is highly soluble in
molten steel. When welding high strength steel, this diffusible hydrogen (monatomic hydrogen dissolved when the steel
is molten) must be allowed to evolve from the steel so it does not become trapped in the solidified steel, or it could
potentially lead to HIC. Hydrogen is not a concern for stainless steels and nickel alloys because the atomic structure of
these metals provides for the hydrogen to remain mobile. The higher the temperature, the faster the hydrogen will
diffuse from the steel and into the CRO. Most of the hydrogen, however, will be consumed in the reduction of oxygen in
the weld zone. Therefore, the increased hydrogen mobility at elevated welding temperatures, and the elimination of
hydrogen as a result of the reducing reaction are thought to explain how the hydrogen is eliminated and not available to
contribute to the occurrence of HIC. While this argument is technically sound, the purpose of this study was to perform
various analyses to confirm that hydrogen does not present a concern when used as a component of the welding
shielding gas. The testing methods and results are found below.

Testing Program:
As stated above, the primary goal of this study was to prove that the addition of a small amount of hydrogen to the
shielding gas does not result in any degradation of mechanical properties of the weld metal zone (WMZ), heat-affected
zone (HAZ), or base metal zone (BMZ). A related concern is the potential for hydrogen-induced cracking (HIC) due to the
introduction of hydrogen during the welding operation.
To review, for HIC to occur, three (3) coincidental conditions are necessary: the presence of a crack-susceptible
microstructure, the presence of stress, and the presence of hydrogen. If one or more of these conditions is absent, the
threat is eliminated. All three (3) materials previously mentioned are considered to be hardenable, so the potential for a
HAZ microstructure with a hardness above the 20 HRC threshold is virtually guaranteed. While the use of preheating and
postweld heat treatment (PWHT) will tend to reduce hardnesses and residual stresses, the potential still exists that even
with these measures, one (1) or two (2) of the contributing conditions could exist. These thermal treatments also play a
key role in removal of diffusible hydrogen from the weld zone, with both temperature, and time at temperature, being
key factors.
The risk of HIC is minimal as long as proven procedures are followed and the welding operation is properly controlled.
The testing program presented here is an attempt to show the use of hydrogen-bearing shielding gases does not
introduce hydrogen to the weld or base metals.
The testing program was separated into three (3) specific parts: determination of hydrogen content, resulting weld
mechanical properties, and inclusion content. For each of these, the welding parameters were held constant to
minimize variability in the resulting welds. All welding was done using pulsed gas tungsten arc welding (GTAW-P) with
hot wire (AC-energized) filler metal addition. Specifics of the welding and test results for each part are found below.

Hydrogen Content Testing:

The purpose of these tests was to measure the amount of diffusible and residual hydrogen remaining in the weld when
using a hydrogen-bearing shielding gas. In addition to those welds made with the argon/hydrogen shielding gas, control
welds were made (using the same welding parameters for tests #1 and #2) with argon only. Test welds #1 - #4 were
made using the following conditions:
Test Conditions for Hydrogen Testing
Wire Feed Hot Wire
Shielding Peak Arc Average Arc Arc Travel Speed, Hot Wire Heat Input
Test # Condition Speed, in/min Voltage,
Gas * Current, A Current, A Voltage, V in/min [mm/min] Current, A Rate, J/in
[mm/min] ACV
1 Argon PWHT 325 285 13.0 10.0 [255] 80 [2,030] 30 2.0 22,541
2 98Ar/2H2 PWHT 325 285 13.0 10.0 [255] 80 [2,030] 30 2.0 22,541
3 Wire test only ** --- --- --- --- --- --- --- ---
4 98Ar/2H2 ISO 3690 260 195 12.8 8.25 [210] 80 [2,030] 59 2.5 19,354

NOTES: Base metal used for test coupons was A-36

* Argon – A5.32, SG-A; 98Ar/2H2 – A5.32, SG-AH-2
** Wire – ERNiCrMo-3, Diameter – 0.045 in [1.1 mm]

Knowing preheat and PWHT are always requirements of the welding procedure, the first series of tests was designed to
simulate normal production welding. The weld coupons were preheated to 500° F [260° C], welded, and then subjected
to PWHT for four (4) hours at 1,175° F [635° C]. Welds for Test #1 were made using 100% Ar shielding and those for Test
#2 were made with the 98% Ar/2% H2 shielding gas. Test #3 was conducted to measure the residual hydrogen in the
welding wire in its as-received condition. Test #4 was conducted using the standard AWS A4.3 / ISO 3690 procedure for
capturing and measuring diffusible hydrogen, which is considered to be a worst-case scenario. To accomplish this, the
coupon is welded, followed by immediate quenching in ice water, blown dry and placed in liquid nitrogen. According to
the testing standards, the welded coupon must be subjected to this routine from welding to liquid nitrogen in less than
35 seconds. Because of the measures taken to assure the hydrogen is completely contained in the weld metal, this test
is considered to be the most stringent measure of entrapped hydrogen in weld metal and is the standard method for
evaluation of hydrogen content in welding filler metals. These same containment methods were not used for Test #1
and Test #2.

Hydrogen testing:
For this study, measurement of the amount of hydrogen present in the various samples was accomplished using the
testing methods specified in AWS D4.3, Standard Methods for Determination of the Diffusible Hydrogen Content of
Martensitic, Bainitic, and Ferritic Steel Weld Metal Produced by Arc Welding and ISO 3690, Welding and allied processes -
Determination of hydrogen content in arc weld metal. While other techniques are referenced in the standards, the gas
chromatograph method was chosen for measuring hydrogen levels in this study, since it is generally considered to be
the most accurate method.
To better understand the test results reported here, it is important to realize there are two different types of hydrogen
in weld metal. First, there is diffusible hydrogen, which is the hydrogen released from the weld metal over time. The
rate of release is increased at higher temperatures; hence the specification of postheating, sometimes is referred to as a
hydrogen bakeout, used in cases where there is a concern hydrogen has been absorbed during the welding operation.
The other form of hydrogen is the non-diffusible, or residual hydrogen. This form of hydrogen is not as mobile as
diffusible hydrogen, so it becomes trapped, remaining in the solid weld metal, unable to escape.
For the diffusible hydrogen tests, five (5) sample welds were made on A36 carbon steel coupons which had been
specially prepared and thermally treated per the requirements of the testing standards. Four (4) samples were needed
for the test and the fifth sample was a backup in the event an additional test is needed. Each weld sample was heated
to 750° F [400° C] for 45 minutes in the gas chromatograph. This is twice the ISO requirement of 21 minutes. This was
done to assure all the diffusible hydrogen would be detected. The values of the four tests were averaged together for
the final test value.
To measure residual hydrogen, a small sample is removed from the weld bead after the completion of the diffusible
hydrogen test. This weld metal sample, weighing from 0.1 – 1.0 g, typically in the range of 0.4 – 0.6 g, is placed in a
carbon crucible and heated to 2,730° F [1,500° C] in a small induction furnace. This heats the sample above the melting
point, allowing any entrapped hydrogen in the nickel weld metal to be released. The result reported here is the average
of three (3) tests.

Hydrogen Test Results:

When reviewing these results, it is important to note, as a point of reference, that the lowest (most stringent) hydrogen
classification used by AWS to indicate the permissible amount of hydrogen is “H4”, which designates the amount of
hydrogen as 4 ml/100 g. As described above, the first two (2) tests were welded using the same (production) welding
conditions with the only difference between the two (2) being the shielding gas. Samples #1 were welded with argon,
and Samples #2 were welded with the argon/hydrogen mixture. The table below summarizes the averages from these
two (2) welded samples.
Hydrogen Determination
Test Number/Gas Diffusible Hydrogen Residual Hydrogen
#1 - Preheat & PWHT – Ar 0.58 ml/100g 3.35 ppm
#2 - Preheat & PWHT – Ar/H2 0.65 ml/100g 3.37 ppm

These results are quite revealing. The first observation is the amount of diffusible and residual hydrogen is virtually the
same when comparing the welds made with argon shielding and those made with argon/hydrogen shielding. Of even
greater importance is the magnitude of the diffusible hydrogen results. In both cases, the amount of diffusible hydrogen
present is roughly six (6) times less than the 4 ml/100 g limit of the H4 designation.
The fact that the amount of hydrogen present in these samples is virtually the same leads one to ask about the source of
the hydrogen. While the diffusible hydrogen amounts are close, the residual hydrogen amounts are virtually identical.
It is suspected the source is the wire itself. In the manufacture of stainless steel and nickel alloy wires, the wire must be
annealed as it moves through the sequential reducing dies since these materials tend to harden due to the cold working.
In most filler metal manufacturing operations, this annealing is accomplished in a hydrogen atmosphere since the
hydrogen inherently cleans the surface of the wire by removing oxides. As a result, there is a tendency for some
hydrogen to be absorbed in the wire as a natural result of the annealing operations applied during the drawing process.
To confirm this theory, a third series of tests (Test #3) was conducted. In this series of tests, samples of the filler metal
wire were melted to determine the amount of residual hydrogen present. As shown in the table below as Test #3, the
average of the tests performed revealed the amount of residual hydrogen present in the wire was 8.57 ppm, which is
roughly 2.5 times the amount of hydrogen present in the weld deposit. So during the welding process, when the wire
was taken to a full melt in a protected atmosphere, some of the hydrogen present in the wire is actually removed. It
must be noted that this was the hydrogen level in the heat of wire used for our testing. Amounts of hydrogen present in
wires from other manufacturers may vary, but it is suspected those levels will also be higher than the resulting amounts
of hydrogen in welds made with these wires. This test confirms the amount of hydrogen in the filler metal wire is
greater than the amount that ends up in the weld deposit.
The fourth hydrogen test (Test #4) was the most severe test in which the weld metal was tested using the standard
procedure for measurement of diffusible hydrogen. This procedure is defined in AWS A4.3 and ISO 3690. This test was
only applied to test welds made using argon/hydrogen shielding gas. As was the case for the first two (2) tests, the test
result is an average of four (4) test samples. The weld metal residual hydrogen level was also measured. The amounts
of diffusible and residual hydrogen are reported in the table below as Test #4.

Hydrogen Determination
Test Number Diffusible Hydrogen Residual Hydrogen
#1 - Preheat & PWHT - Ar 0.58 ml/100g 3.35 ppm
#2 - Preheat & PWHT – Ar/H2 0.65 ml/100g 3.37 ppm
#3 - NiCrMo-3 wire * 8.57 ppm
#4 - ISO 3690 – Ar/H2 0.90 ml/100g 3.82 ppm
* Diffusible hydrogen test was not run on the welding wire

Since Test #4 was conducted using standardized procedures typically used for measurement of diffusible hydrogen in
weld filler metals, the reported amount of diffusible hydrogen can be directly compared to other filler metals. As
discussed above, the lowest classification of hydrogen content for filler metals is H4, which means there is less than 4
ml/100 g of diffusible hydrogen. Therefore, the results of Test #4 show the amount of diffusible hydrogen present was
less than 25% of this classification limit. Compared to Test #2, the amounts of diffusible and residual hydrogen present
in this test were about 40% and 15% higher, respectively. This clearly shows the methods used for determination of
hydrogen content are more stringent because all of the hydrogen is retained in the sample. Without these measures,
the results from Tests #1 and #2 show that some of the hydrogen will naturally evolve.
With it now proven that the presence of hydrogen in the shielding gas does not significantly increase the amount of
diffusible or residual hydrogen in the weld deposit, another question has been raised regarding the potential for
hydrogen remaining in the weld metal to diffuse into the carbon steel. To address this concern, one must consider the
potential for hydrogen to migrate through the solid metal. It should be noted that nickel has a high affinity for hydrogen
and the equilibrium path is easier for the hydrogen to diffuse thru the nickel alloy rather than into the HAZ of the high
strength low allow steel. The mobility of hydrogen is based upon its solubility, which is dramatically reduced as the
temperature of the steel decreases. At an elevated temperature 1,300° F [705° C] the solubility of hydrogen in carbon
steel is about 8 ppm compared to less than 1 ppm at room temperature. As the temperature of the steel decreases, the
pressure of the hydrogen in the steel increases due to the reduced solubility. To avoid HIC damage, any hydrogen
introduced into the steel must be removed. This is where the time and temperature become important. Hydrogen will
migrate out of the steel 250 to 400 times faster at 400° F [205° C] than at room temperature. Therefore, the use of
preheat and allowing the weld zone to cool slowly gives the hydrogen more mobility and time to move out of the carbon
steel and into the more soluble CRO weld metal. To reverse this process, significant amounts of time, temperature and
pressure would be required.

Summary and Conclusions:

This study has primarily been conducted to dispel impressions that the use of a hydrogen-bearing shielding gas will
cause hydrogen pickup in the weld and base metals. In the process of dispelling these fears, it shows there are
numerous benefits associated with the addition of 2% hydrogen to the shielding gas.
Below are some of the points confirmed in this study:
 In a worst-case scenario with weld metal in the as-welded condition, diffusible hydrogen levels were one fourth
the AWS H4 classification for filler metal hydrogen content.
 When using argon/hydrogen shielding gas for joining high strength steels, there is no evidence of any
degradation of mechanical properties in the weld metal and heat-affected zones.
 The study shows argon/hydrogen shielding significantly reduces the inclusion content of the weld deposit.
The benefits of using hydrogen as a component of welding shielding gases have been well known for decades; however,
these potential benefits have been quickly ignored due to the overriding concern for potential damage or even failures.
It was with this in mind that this study was conducted. Years of experience had prompted the need to show that using
argon/hydrogen shielding gases for applying CROs to high strength steels was indeed safe. When using argon/hydrogen
shielding gas and normal low hydrogen practices, top quality welds can be produced with high productivity resulting in
desirable mechanical properties and improved cleanliness.
This article only covered the test results showing the hydrogen content of welds made with an argon/hydrogen blend
shielding gas. The study also included the effects of hydrogen on the mechanical properties and the effects on weld
metal inclusions. For the full report covering all the details on test weld parameters, comparison of mechanical
properties and weld metal inclusion comparisons, readers are encouraged go to the ARC Specialties web site
(https://arcspecialties.com/resources/white-papers/) and download the complete report.

Contact information:

Dave Hebble dave@arcspecialties.com

Richard Holdren richard@arcspecialties.com

ARC Specialties
1730 Stebbins Dr.
Houston TX 77046
713-631-7575
www.arcspecialties.com