Fume composition relationships for alloy-MMA electrodes

By J.C.M. Farrar and A.W. Marshall

Introduction

Welding fume consists of various airborne substances which may constitute a hazard if inhaled or swallowed. Proper assessment of this pollution depends upon information about the fume composition as well as its concentration in the air being breathed. MMA fume originates from the flux-metal system as a whole, and the present discussion concerns the latter ‘alloy fume’ constituents. In order to take account of possible hazards, a guide to various airborne substances is provided by the system known as Occupational Exposure Limits (OEL) and these are listed and defined in the Health and Safety Executive Guidance Note EH40.

When the material being welded is mild steel with approximately matching C-Mn consumables, it is generally sufficient to control the total fume level within the overall current OEL for welding fume of 5mg/m This automatically limits the levels for the individual constituents, e.g. iron oxide, manganese oxide, etc. to below their OEL values. However, it is recognised with alloyed welding consumables that the fume composition is related to the weld metal composition being deposited. Since some of the alloy constituents, e.g. Cr, may have a very low OEL, it may be necessary to apply additional precautions to ensure safe working conditions over and above those necessary to meet the general 5mg/m level for welding fume.

Over the last few years manufacturers of alloyed welding consumables worldwide have generated substantial quantities of fume composition data for electrodes varying from relatively simple low alloy types up to complex high alloy Cr, Ni and Co base alloys. Much of this information is available from responsible manufacturers on request in the form of data sheets, but until now it has not been brought together from these and other published sources in a form which can be readily appreciated and utilised by concerned users of the many different compositions and brands on the market.

Alloy Types

The commonly used alloy constituents which are listed in Guidance Note EH40 are given in Table 1, together with OEL values, the consumable groups in which they occur and the approximate compositional range for that element. From this table it can be seen that elements fall into two clear groups with respect to their short-term OEL values: those having no influence on fume OEL, with a short-term OEL of 5mg/m3 , i.e. Fe and Mo, and those which need to be examined in more detail,

i.e. chromium, particularly as CrIV, nickel, copper, manganese, cobalt, tungsten and vanadium, all of which have short-term OEL values of 1mg/m3 or lower.

This technical note will therefore deal with the relationship between fume composition and alloy weld metal composition for the elements given in Table 1, but will concentrate on those in the low OEL group since it is control of these elements which is most necessary to provide a safe working environment for welders and other personnel.

The Data

Fume data from Metrode’s own investigations together with that published by other manufacturers and various independent studies has been included in this work. A summary of the sources and the scope of the data utilised is given in Table 2 and will be published more fully elsewhere.

The data is presented graphically in all cases on a log/log plot. Averaged values have been used where analysis for more than one electrode size was available for the same product. Note that data boundaries parallel to the 1:1 line also have linear proportionality but with different gradient. The OEL curves are an inverse function of alloy content, with corresponding negative gradient.

Discussion – conclusion

For most elements, sufficient fume analyses are available to provide a coherent view of the influence of alloying on fume composition. The range of fume compositions occurring at various alloying levels enables boundaries to be drawn with reasonable confidence, since there are relatively few outlying points. It should be noted that the 'scatter band' does not represent the range of fume composition produced by any single specific product.

It is clear from the relationships shown in these graphs and their simple interpretation to show fume OEL, that reasonable predictions can be made from a knowledge of weld metal composition. The shape of data envelopes for different elements and their relative positions may also throw further light upon the nature of fume evolution. However, it is felt that their principal value lies in helping to make informed generalisations in the assessment of measures required for fume control.

TABLE 1 (a-Long-term OEL, b=Short-term OEL)
FUME CONSTITUENT mg/m3 CONSUMABLE GROUPS APROX RANGE OF
OEL ELEMENT
a b
Iron: Fe
(oxide Fe2O3) 5 10 All iron base alloys and steels. 40-100%
Residual in nickelbase alloys. 0-10%
Manganese: Mn 1 3 All low alloy stainless steels. 1-7%
Work-hardening Mn steels 10-20%
Chromium:(Cr lI&III) 0.5 - Low alloy steels 1-5%
(CrVI) 0.05 -Hardsurfacing alloys 5-35%
Stainless steels 12-30%
Special alloys Up to 50%
Nickel: Ni Low alloy steels 1-5%
(soluble) 0.1 0.30 Stainless steels 10-30%
(insoluble) 1 3 Ni base & superalloys 30-70%
Electrodes for cast irons 95%+
Molybdenum: Mo Low alloy & tool steels 0.5-5%
(soluble) 5 10 Stainless steels 2-6%
(insoluble) 10 20 Nickel base alloys 5-30%
Copper: Cu Stainless steels 1-3%
(fume) 0.2 - Monels ~30
Cupronickels 70-90%
Cobalt: Co 0. 1 -Pure coppers 95%+
Alloy & tool steels 1-5%
Cobalt alloys 50-60%
Vanadium: V
(pentoxide) 0.05 0.05 Alloy & tool steels Up to 5%
Tungsten: W
(soluble) 1 3 Alloy & tool steels Up to 5%
(insoluble) 5 10 Cobalt alloys 5-15%

` TABLE 2 : Principal sources of data

1. General

Welding Fume, sources, characteristics, control. 3v. TWI 1981-83. Guidance Note EH40, Health and Safety Executive, London HMSO, pub. Annually

2. Consumable manufacturers

Fume classes of covered welding electrodes, Avesta Jernverks, Avesta (1981). Skyddsblad. Elgasvets AB, Lerum (1980). Skyddsblad. Vid svetsning bildas rök och gaser. Esab AB, Goteborg (1981). Røgklassificering. Esab, København-Valby (1980). Measurement of fume emission rates and composition of a range of MMA electrodes, LD Report

26584. (Metrode) TWI 1980.

Determination of fume emission rate and composition for MMA electrodes, Project Report 23791/1/85 (Metrode) TWI 1985. Røgklassificering of Philips Elektroder. V. Løwener, Kjøbenhaven (1979-81). Skyddsblad. Philips Svets Nordika AB (1981).

3. Specific studies/reviews

Chromium in welding fume. Safety Information Sheet No.11, issue 2, TWI, 1986. Tandon, R.K. et al. Investigation of hardfacing and HSLA steel electrodes. Weld. and Met. Fab. March 1983 pp.43-47. Kimura, S. et al. Investigations on Chromium in Stainless Steel Welding Fumes, Weld.J. 1979,58,7,195-204s. Moreton, J. et al. Ann.Occup.Hyg.1983,27,2,137-156. Moreton, J. et al. Met.Constr.1985,17,12, 794-798.

Fig.l: Relationship between Fe content of weld metal and fume. OEL values for Fe (as Fe203) and total fume are both 5mg/m3, thus the control of fume to this level will automatically ensure that fume Fe is below its OEL. In practice the fume Fe content is unlikely to exceed 40% for mild steel or low alloy electrodes.

Fig.2: Relationship between Mn content of weld metal and fume. The OEL for Mn is lmg/m3 or one fifth that for total fume, and this value is exceeded when the fume Mn content exceeds 20%, corresponding to a weld metal Mn content of at least 6%. In practice, the highest manganese weld metals are 'Hadfield Steel' types with 10-15%Mn, for which the fume OEL would be around 3mg/m3 . Note the relatively high volatility of this element, above the 1:1 line. .

Fig.3: Relationship between Cr content of weld metal and total Cr in fume. Normally 60-100% of total Cr is the hexavalent form (Cr6) with an OEL of 0.05mg/m3 . Fume may contain up to 1% Cr6 for a total fume OEL of 5mg/m3. In the most simple case, total Cr can be read as Cr6, indicating a reduction in fume OEL for weld metal Cr contents exceeding 2-5%. The median line for fume Cr lies at roughly 60% of the upper boundary level and indicates a typical OEL of 0.8mg/m3 for a weld Cr of 20%.

Fig.4: Relationship between Cr content of weld metal and Cr6 in fume. Fewer reliable data points are available than for total Cr. Comparison shows that the upper Cr6 boundary in this graph is equivalent to the medium line in Fig.3 and is evidently a more realistic indication of the ‘worst-case' fume OEL determined by weld metal Cr content. In this figure, fume OEL is influenced at weld metal Cr contents above 2.5-12%, the median predicting an OEL of 1.4mg/m3 at 20%Cr.

Fig.5: Relationship between Ni content of weld metal and Ni in fume. The OEL for soluble Ni is 0.l mg/m3 or 2% of that for 5mg/m3 . The upper boundary of the scatter band shows that weld metal above 12%Ni content could dictate overall fume OEL. However, this ignores the relative proportion of insoluble nickel with OEL lmg/m3. Note also the similarity of the data envelope to that of iron, Fig.l.

Fig.6 : Relationship between Cu content of weld metal and fume. The envelope for data in this case is more tentative as fewer analyses were available. The DEL for Cu is 0.2mg/m3 or 4% of total fume, and the figure predicts that this level will be exceeded in weld metals containing between 7% and 35% Cu

Fig.7 : Relationship between Co content of weld metal and fume. The data for this relationship is rather sparse and the boundaries therefore tentative. The OEL for Co is 0.lmg/m3 or 2% of total fume. The figure predicts that cobalt will control overall fume OEL for weld metals above 5-15%Co.

Other alloying elements which may be present in fume and for which DEL values have been assigned are as follows.

a) Molybdenum. The OEL is 5mg/m and therefore has no effect on overall fume OEL. Nickel base alloys may contain up to 30%Mo, and this is likely to give a fume Mo content of less than 15%.

b) Tungsten. The OEL is Img/m , but is unlikely to have a controlling effect on fume OEL. Some cobalt alloys have up to 15%W, but the fume OEL is dominated by Co.

c) Vanadium. The OEL is 0.05mg/m , the same values as hexavalent chromium, but alloying levels are relatively low at up to 5% in some tool steels. The sparse data indicates fractional levels in welding fume.