Shielding gas formulation - its effect on arc welding fume emissions

TWI Bulletin, January - February 1998

Graham Carter joined TWI in 1965, and for the past 14 years has been actively involved in work related to welding fume, its generation and control. He is the UK representative on CEN/TC121/SC9, WG1 and WG2.

The effect of 18 shielding gas formulations on fume emission rate and composition was examined during GTAW of stainless steel, GMAW of stainless steel, carbon steel and aluminium, FCAW of carbon steel. The additional effects of different metal transfer conditions and the use of pulsed current were also examined. The emission rate of carbon monoxide was determined when using shielding gases containing carbon dioxide. Graham Carter and Richard Wiktorowicz report.

The control of substances hazardous to health regulations (COSHH) 1994 ^[1] require that exposure to welding fume does not exceed certain concentrations known as occupational exposure limits (OELs). The associated approved code of practice (ACOP) ^[2] describes an hierarchy of events that should take place to achieve compliance. The first priority must be prevention of exposure and consideration must be given to changes in work methods, substitution of hazardous substances and process modification to eliminate production of hazardous by-products. If control cannot be achieved by these routes, then engineering controls will be required or it may be necessary to resort additionally to respiratory protection equipment.

Given that, at least for metals, there is generally no viable joining alternative to welding, prevention of exposure to fume becomes difficult. However, it is possible that modification of the welding process could be used to reduce exposure by reducing the amount of fume produced at source. In this respect, economic and quality considerations generally require that welding parameters are predefined within certain narrow limits thus limiting scope within this area. However, a possible variable in gas shielded welding processes is the formulation of the shielding gas.

It is possible to reduce the control requirements for welding fume, whether it be particulate or gaseous fume, either by reducing the amount of fume generated or by effecting changes in the composition of the particulate fume to make it less harmful. This article reports work carried out to examine the effect of shielding gas formulation on emissions from a number of welding process and material combinations, in particular the emission rate and composition of particulate fume and the emission rate of carbon monoxide. Particulate fume emission rate and fume composition were measured during steady current GMAW of stainless steel, carbon steel and aluminium, pulsed current GMAW of carbon steel, GTAW of stainless steel and FCAW of carbon steel. Carbon monoxide emissions were measured from carbon steel and stainless steel GMAW and from carbon steel welding using the FCA process.

Work programme

The work programme was carried out in three parts. Part 1 evaluated the emission rate of particulate fume. Part 2 established the emission rate of carbon monoxide and Part 3 determined particulate fume composition.

The wires for GMAW and FCAW were 1.2mm diameter whilst those for GTAW were 2.4mm. The specifications are given in Table 1.

Table 1: Wire specifications

It is well recognised that fume emission rate varies with welding current and arc voltage. To isolate the effects of the shielding gas, tests were performed at a fixed current. The chosen currents for steady current GMAW and FCAW were 160A for dip transfer and 270A for spray transfer with voltage adjusted to provide optimum welding conditions. GTAW was carried out at a nominal current of 225A. GTAW and FCAW were performed using dc - polarity whilst GMAW used dc +. Pulsed GMAW was performed, following procedure development to optimise joint completion rate, with visual acceptability, of a fillet weld of 6mm leg length, made in the flat position. The conditions are summarised in Table 2.

Table 2: Conditions for pulsed GMAW

Eighteen shielding gas formulations were tested. For the welding of steels, the shielding gas was either carbon dioxide or mixtures of argon/carbon dioxide, argon/carbon dioxide/ oxygen and argon/helium/carbon dioxide. Argon or argon/helium or argon/hydrogen formulations were used with GTAW of stainless steel.

A fume box technique was used to determine fume emission rate. The procedure conformed to BS7384:1991 ^[3] and to a draft European/ISO standard ^[4] . Welding was performed, manually, bead on plate on matching test pieces of stainless steel, carbon steel or aluminium alloy. Five tests were performed to evaluate each particulate fume emission rate and three tests to evaluate carbon monoxide emissions.

Fume for analysis, from GMA and FCAW, was collected on paper filters by repeated welding in the fume box. Analysis was performed using X-ray fluorescence spectrometry using a fusion technique.

Fume from GTAW was collected using equipment normally used for breathing zone sampling according to BS6691 ^[5] . Fume was collected on a cellulose ester filter by placing the sampling head above the arc during welding on a test piece, placed on a table in a welding bay. Analysis was performed using ICP-AES spectrometry, following dissolution in a nitric/hydrochloric/hydrofluoric acid mixture, using a high pressure microwave digestion system.

Particulate fume emission rates

Fume emission rate data are presented graphically in Figs 1-7. From the results, it may be seen that both the welding process and the material welded play an important role in determining particulate fume emission rate. GTAW provided the lowest emission rates regardless of the shielding gases used and GMAW of aluminium generated substantially more fume than all the other GMAW. For ferrous welding, dip transfer used with stainless and carbon steel provided broadly similar emission rates, but spray transfer used with carbon steel generated more fume than either. FCAW and pulsed GMAW resulted in similar emission rates to steady current GMA welding using spray transfer.

For a given process and transfer conditions, examination of the effects of the shielding gases showed a tendency, with argon based gases, for emission rate to increase with higher helium concentrations or the amount of active gas present (carbon dioxide plus oxygen). This is in broad agreement with the work of Hilton and Plumridge ^[6] although the correlation here between these factors and emission rate was less marked. The observed differences in emission rate were small and in absolute terms amounted to just a few mg/s although the relative differences could sometimes be large.

Carbon monoxide emission rates

It would be expected that the emission rate of carbon monoxide would increase with the concentration of carbon dioxide in the shielding gas. This section examines the emission rate of carbon monoxide in those terms.

With the exception of pulsed GMAW with a carbon dioxide shielding gas, FCAW provided the highest emission rates of carbon monoxide, Figs 8-12. Carbon monoxide emission rates with steady current GMAW and dip transfer were rather higher for carbon steel welding than with stainless steel. Spray transfer again generated more fume than dip transfer, as for particulate fume. The results with steady current spray transfer were very similar to the result with pulsed current, Figs 11 and 12.

Although the carbon monoxide emission rate generally increased in line with the carbon dioxide content of the shielding gas, the rate of increase appeared to be dependent on whether the shielding gas was a two or three component mixture as shown in Figs 13 and 14. Generally, helium appeared to enhance carbon monoxide emissions as shown by the fact that helium/argon mixtures containing around 2%CO ₂ generated slightly more carbon monoxide than an argon/2%CO ₂ mixture, Fig 9.

Fume composition

Guidance Note EH54 from the Health and Safety Executive ^[7] describes how to modify the welding fume exposure standard to take account of toxic substances present in the fume using the following equation to calculate a new control limit:

Fume control level =

mg/m³

This modified fume exposure limit has been calculated for the fume composition analysed here and is reported for stainless steel welding Table 3 together with the fume composition data. The fume composition data for carbon steel and aluminium welding have not been reported because the changes in composition were not sufficient to alter the welding fume exposure limit of 5mg/m ³.

Table 3: Fume composition data for stainless steel welding

For GMAW of stainless steel, the modified fume control level is dictated by the concentration of chromium present. The fume control levels are similar because the levels of chromium are similar, showing that the effect of the shielding gases tested on fume composition was small.

The fume compositions reported for GTAW of stainless steel must be regarded as semi-quantitative because the small amounts of fume collected presented analytical problems. Even so, the compositions are surprising with up to 60% manganese present in the fume when only 2% was present in the wire and plate. Even with 60% manganese present in the fume it is considered unlikely that exposure problems will occur because of the very low fume emission rate. However, further work is considered necessary to evaluate the situation.

Discussion

The results show that the welding process and welding parameters have a major role in determining particulate fume emission rate while the role of the shielding gas is relatively small. However, the shielding gas is extremely important in dictating the extent of carbon monoxide emissions. For example, GTAW always gave less fume than the other processes, whilst dip transfer always generated less fume than spray. Shielding gases containing more carbon dioxide gave rise to higher emissions of carbon monoxide.

For a given set of welding conditions, larger concentrations of active gases (CO₂ + O₂) or helium in the shielding gas tended to promote fume formation but the differences in emission rate were usually small and the correlation sometimes poor. Consequently, there did not appear to be a major advantage in formulating gases specifically to reduce particulate emissions. That said, variations in emission rate of around 2 mg/s at the 5mg/s level occurred, and although small in absolute terms, could be viewed as a significant (60%) reduction. However, it is not expected that differences of this magnitude will cause major changes in fume control strategy in most work situations. Similarly, with the possible exception of carbon dioxide shielding gas, it is not expected that carbon monoxide emissions will give rise to exposure problems.

The effect of gas formulation on fume composition was small and, for the carbon steel and aluminium welding, insignificant in terms of warranting changes in fume control requirements. The fume control levels calculated from changes in fume composition during stainless steel welding ranged from 3.1 to 3.8mg/s. As such, they were similar and little benefit would result from using one gas rather than another. Although there were considerable differences between the fume compositions when welding stainless steel using argon/helium or argon/hydrogen gas mixtures with the GTA process, the differences are not considered important because of the associated, low, emission rates.

Conclusions

1. Shielding gas formulation did not have a major influence on particulate fume emission rate. Although greater concentrations of active gases in the gas tended to raise emission rates the differences were usually small.
   
   
2. Shielding gases containing higher concentrations of carbon dioxide generated more carbon monoxide. The effect was modified by the presence of helium or oxygen in argon based gases. Helium tended to promote the formation of carbon monoxide whilst oxygen moderated it.
   
   
3. For the process/material/gas combinations tested, shielding gas formulation did not have a major effect on fume composition as judged from calculated fume control levels.

References