Ann. occup. Hyg., Vol. 46, No. 1, pp. 33-42, 2002
© 2002 British Occupational Hygiene Society
Published by Oxford University Press
Article |
Control of Occupational Exposure to Hexavalent Chromium and Ozone in Tubular Wire Arc-welding Processes by Replacement of Potassium by Lithium or by Addition of Zinc
1Department of Environmental Science, University of Bradford, Bradford BD7 1DP, UK; 2University of Tarbiat Moddarres, Tehran, Iran
Received 4 September 2000; in final form 31 March 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Hexavalent chromium [Cr(VI)] and ozone are produced in many arc-welding processes. Cr(VI) is formed when welding with chromium-containing alloys and is a suspected carcinogen. Ozone is formed by the action of ultraviolet light from the arc on oxygen and can cause severe irritation to the eyes and mucous membranes. Previous work has demonstrated that reduction of sodium and potassium in manual metal arc-welding electrodes leads to substantial reductions in Cr(VI) concentrations in the fume as well as a reduction in the fume formation rate. In this paper replacement of potassium by lithium in a tubular wire welding electrode (self-shielding flux-cored) is shown to give reductions in Cr(VI) concentrations and fume formation rates. Previous work has also demonstrated that use of a tubular wire (metal cored) containing 1% zinc can, under certain conditions, result in a reduction in Cr(VI) formation rate and in ozone concentration near the arc but with a rise in the total fume formation rate. The effects of different shield gases and different levels of zinc are examined. An experimental chromium-containing tubular wire with 1% zinc was used with the following shield gases: argon, Argoshield 5, Argoshield 20, Helishield 101, Ar + 2%CO2, Ar + 5% CO2, Ar + 1% O2 and Ar + 2% O2. The wire gave >98% reduction in Cr(VI) formation rate compared to the control wire provided the shield gas contained no oxygen. When the shield gas did contain oxygen, 1% zinc enhanced Cr(VI) formation rate, resulting in more than double the rates measured when welding with the control wire. Experiments with zinc concentrations, from 0.018 to 0.9% using Helishield 101, gave results indicating that there is an optimum zinc concentration from the point of view of Cr(VI) reduction. Implications of the use of lithium or zinc on the overall exposure risk are discussed.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Electric arc welding is a widely used industrial process for joining of metal components. The necessity and enormous economic value of such processes are well recognized. Even though welding is a critically important fabrication process underpinning industry in the UK and elsewhere, occupational exposure to welding fumes and gases continues to be an unresolved issue which has been of greater significance since the widespread introduction of welding in the 1950s. In recent years the use of local exhaust ventilation (LEV) throughout larger UK industry has somewhat reduced occupational exposure to welding fumes. However, in many small to medium-sized enterprises (SMEs), it is common to find little or no LEV control of welding fumes. This situation is mirrored in other developed countries, while the complete absence of LEV is the norm in third world and developing countries. LEV must therefore be viewed as an intermediate control measure while more effective engineering controls are sought.
This paper examines the replacement of potassium by lithium as a means of controlling hexavalent chromium [Cr(VI)] in a self-shielding flux-cored wire. The paper also examines the use of zinc in an experimental chromium-containing metal cored wire for the control of Cr(VI) and ozone. A 1% zinc wire with different shield gases was investigated. The best shield gas from the point of view of reducing Cr(VI), under the welding conditions used, was found to be Helishield 101 (BOC Gases). A range of zinc concentrations from 0.018 to 0.9% with Helishield 101 was investigated.
Zinc addition could be used in flux-cored wires but replacement of potassium with lithium is not applicable to metal-cored wires since they do not normally contain potassium.
When using chromium-containing electrodes, the reactive metal species Cr(VI) is formed in the fume. Kazantzis (1972) describes excess lung cancers in chromite ore processors, suggesting an association with Cr(VI). The evidence relating to Cr(VI) exposure in welding fume has not yet demonstrated that it causes cancer in welders, but it is sensible to limit exposure during welding. Concern regarding exposure to Cr(VI) is expressed in the UK by the Health and Safety Commission specifying a maximum exposure limit (MEL) (long-term limit) of 0.05 mg/m3. Ozone is formed in arc welding by the action of ultraviolet light on oxygen. Ozone causes severe irritation to the eyes and mucous membranes and in severe cases pulmonary oedema. Ozone is also known to exacerbate existing asthma. Other fume components and gases have been allocated occupational exposure limits (OELs) either as MEL values or occupational exposure standard (OES) values. The total welding fume OES is 5 mg/m3 and is exceeded in some welding operations.
For fumes containing chromium the MEL for Cr(VI) can be exceeded before a total fume concentration of 5 mg/m3 is reached. In certain situations Cr(VI) becomes the key component for which the MEL is exceeded first. This can be the case in manual metal arc (MMA) welding with chromium-containing electrodes, when typical Cr(VI) concentrations are ~5% of the total fume. This requires total fume control to <1 mg/m3 if the Cr(VI) MEL is to be complied with. Relatively high concentrations of Cr(VI) are also encountered in many solid and tubular wire welding operations involving chromium-containing alloys.
The concept of a key component ignores the consideration of the additive effects of fume components. For example, nickel is usually present in alloys containing chromium and certain species have also been associated with cancer. It would be precautionary to consider the effects additive although there are differing views on the consideration of additive effects. Meeting the limit for Cr(VI) would imply exceeding the additive limit if nickel compounds were present and further reduction of total fume concentration, below that necessary to meet the Cr(VI) MEL, would be necessary. It is important to add that meeting the MEL is a minimum requirement and it is expected that exposure should be reduced so far as is reasonably practicable.
Studies in the 1970s by Japanese workers lead to the development of low fume, low Cr(VI), MMA electrodes. Kimura and co-workers found that the Cr(VI) in fume from MMA electrodes was in the form of sodium and potassium chromate or dichromate (Kimura et al., 1979). Sodium and potassium are used in the form of silicates to aid extrusion and binding of electrode coverings. They are also the prime source of alkali ions in the arc plasma and are important aids to arc formation and stability. Reduction of the (Na2O + K2O) level to <0.5% resulted in a dramatic reduction in the Cr(VI) content of the fume. Fume Cr(VI) concentrations of >4% were reduced to <1%. There was also a significant reduction in fume formation rate (FFR). Use of lithium in place of the sodium and potassium was found to increase Cr(VI) but not above 1% in the fume when using up to 3.75% Li2O in the covering. Lithium can be used as a binder in the form of a silicate and as an aid to arc formation and stabilization in place of sodium or potassium. Table 1 shows results for FFR and Cr(VI) for the standard and improved electrodes. The improved electrodes gave weld metal that fully satisfied the composition, corrosion resistance and mechanical properties of the respective American Welding Society (AWS) specifications. Subsequently the welding manufacturers ESAB developed a low-fume, low-Cr(VI) MMA electrode based on reduction of sodium and/or potassium. Griffiths (1991) also reports development of low-fume, low-Cr(VI) electrodes for stainless steel MMA welding, based on replacement of sodium and potassium by a form of lithium silicate. Careful balancing of the formulation was necessary to achieve acceptable welding characteristics. Use of the low-fume electrodes required no additional fume control above that required for carbon manganese steel electrodes, and they have been used in commercial situations including construction of the Sizewell pressurised water reactor and offshore pipework manufacture.
|
Hirst (1990) applied the replacement of potassium by lithium to a self-shielding, flux-cored wire and found reduction of Cr(VI). Further study of the wire, produced for Hirst by ESAB, is reported in this paper.
Ozone is a major exposure risk in metal inert gas (MIG) and tungsten inert gas (TIG) welding of aluminium and sometimes in MIG welding of mild steel. The wires used in this research are tubular (metal-cored) wires. Although ozone may not be the major exposure risk with these wires, it is still desirable to reduce ozone exposure.
Goralski (1991) found that welding operations involving zinc produce very little ozone compared to mild steel welding. Ozone is formed by the action of ultraviolet light generated in the arc. Since there was little change in ultraviolet emission between mild steel and zinc-plated welding operations, it was concluded that most of the ozone must rapidly react with zinc vapour close to the arc. Redding (1995) examined the effects of the addition of the volatile metals zinc, magnesium and aluminium, at a concentration of 1%, to tubular wires, on Cr(VI) formation, over a range of voltages, using one shield gas. Dennis et al. (1996) reported on this zinc, aluminium and magnesium work. Mortazavi (1995) studied the same wires using one voltage setting and one shield gas. He measured fume formation rate, ultraviolet light, ozone and Cr(VI), and found the 1% zinc wire to be the most promising in relation to control of Cr(VI) and ozone. Mortazavi went on to study the effects of using the 1% zinc wire with different shield gases on fume formation rate, ultraviolet light, ozone and Cr(VI). This latter work, on 1% zinc wire with different shield gases, is reported here. The effects of different concentrations of zinc on ozone and Cr(VI) are also reported in this paper and compared with results of experiments on the same wires carried out by Moore (1997). Work on different concentrations of zinc formed part of a paper by Mortazavi (1997).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Welding and fume collection
Welding took place onto a variable-speed, rotating workpiece above which a welding torch (BOC MG5/MXA503) was fixed so that a uniform weld could be achieved. The workpiece composition was mild steel. The height of the torch above the workpiece could be adjusted by means of a rotating ball screw to which the torch holder was attached. The welding set was a constant-voltage, direct current power source (BOC SMR500) with a wire feed unit (BOC TF2.0S). Shield gas was supplied to the torch via a flowmeter. The welding torch and workpiece were covered by a large conical fume box of similar construction to that recommended by the American Welding Society (1979). An extraction fan drew fumes to the top of the chamber where they were collected on Whatman GF/A glassfibre filters. Welding was carried out for a controlled period of 10, 15 or 20 s and the extraction fan was left operating for a further 90 s to ensure collection of fume in the chamber. An electronic sequence controller was used to control welding and extraction times. Direct weighing of the filter before and after welding allowed calculation of FFR expressed as mass per unit time. Fume was collected for analysis by carefully scraping it off the filter into a glass vial.
Wires studied
For the study of replacement of potassium by lithium, a commercial self-shielded, flux-cored wire, Murair 20/9/3 (Murex Welding, part of the ESAB Group), was used. Self-shielded wires are designed to produce shielding gases and vapours from the core materials and do not require a separate shielding gas supply. A lithium-containing equivalent (CXL71) was made. The compositions of the wires are given in Table 2. Murair 20/9/3 is produced with a diameter of 1.6 mm and deposits a nominal 20% chromium, 9% nickel, 3% molybdenum weld metal. Welding was DC electrode positive with a 20 mm stand-off (contact tip of welding torch to workpiece distance) and no shield gas. A range of voltages was studied.
|
For the study of the effect of zinc a series of experimental tubular wires (metal cored) was prepared by ESAB(UK). The wires had a mild steel sheath with a core containing chromium and different amounts of zinc. For the experiments with different shield gases the 1% zinc and control wire compositions are given in Table 3 and the welding conditions in Table 5. For the 0.018–0.9% zinc experiments with one shield gas the wire compositions are given in Table 4 and the welding conditions in Table 6.
|
|
|
|
Shield gases
The shield gas compositions and estimated flowrates at the torch are given in Table 7. Some of the gases were not available commercially and were specially prepared by BOC Gases. Gas flowrates were measured using a variable-area glass flow tube and float (PLATON UK), with corrections for differences in gas densities to give the same volumetric flowrate for each gas. It was originally assumed that the pressure in the flowtube was close to atmospheric. Later it was found that there was a large pressure drop across the system downstream of the flowtube. Pressures at the flowtube were measured to allow estimation of the volumetric flowrate at the welding torch. The estimated flowrates were similar except for the Helishield H101, which was ~30% lower.
|
Measurement of ultraviolet light
Ultraviolet light was measured indirectly through its ozone-forming capabilities (UV–ozone). A quartz-windowed glass cell was positioned with the window facing the welding arc at a distance of 0.5 m and an angle of 30° from the workpiece. Clean dry air was drawn through the cell at a constant flowrate of 0.9 l/min and the ozone concentration in the exit air measured using the chemiluminescent ozone analyser described below. The quartz window was necessary as it transmits ultraviolet light. The window was protected from weld spatter damage by a device that directed a jet of air across the face of the window.
Measurement of ozone concentrations
Ozone concentrations were measured at a fixed sampling position relative to the welding arc at an angle of 60° from the horizontal and at a horizontal distance from the weld pool of 75 mm. The sampling probe was attached to the welding gun and connected via a 2 m PTFE tube to a precalibrated chemiluminescent ozone analyser (Analytical Instrument Development Inc. model 650). The analyser was capable of measuring ozone concentrations in the range 0.001–100 p.p.m. Two 37 mm diameter, 0.2 µm PTFE filters were inserted in the connecting tube to prevent welding fume from entering the analyser. The first filter (nearest the sampling tip) was renewed after every test and the second filter was renewed after nine tests. Before undertaking any measurements on welding emissions the sampling probe, PTFE tubing and filters were conditioned to prevent subsequent adsorption of ozone. Conditioning was by passage of 10 p.p.m. ozone from a Penwalt ozone generator (Wallace and Tiernan) for 10 min. Ozone measurements were averaged over a 60 s period after first welding for 60 s to allow for establishment of the welding regime. All experiments were performed in triplicate.
Cr(VI) analysis
Cr(VI) analysis involved the preliminary step of extraction using 2% NaOH/3% Na2CO3 solution. Accurately weighed samples of ~25 mg were placed in 150 ml conical flasks to which 40 ml of 2% NaOH/ 3% Na2CO3 solution was added, together with 10 ml deionized water. The solution was simmered for 20 min in covered flasks. Filtration, cooling and making up to 50 or 100 ml in volumetric flasks followed and the Cr(VI) was determined on an atomic absorption spectrometer (Perkin Elmer 1100).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Control of Cr(VI) by replacement of potassium by lithium in Murair 20/9/3 flux-cored wire
Table 8 gives the experimental results for Cr(VI) measurements and Table 9 gives experimental results for FFR measurements. Figure 1 shows how Cr(VI) concentration increases with voltage for the Murair 20/9/3, whereas the CXL71 wire produced not only a slight fall in Cr(VI) concentration with increasing voltage but also very much lower concentrations. At all voltages the concentrations were reduced well below the 1% level, thereby removing the possibility of Cr(VI) being the key component in determining allowable fume concentrations (but note discussion of key component concept in the Introduction). Figure 2 shows how FFR increases with voltage. FFR for the Murair is approximately twice that for the CXL71 at a given voltage. The combination of lower Cr(VI) concentrations and lower FFR for the CXL71 wire produces a large reduction in Cr(VI) exposure risk. There is, however, the possible increased risk associated with other hazardous components of the fume, which may have increased.
|
|
|
|
A comparison of the composition of the fumes produced by the standard and improved MMA rods described by Kimura et al. (1979) is given in Table 10. Since the FFR was lower in the improved electrodes the concentrations of the various components have been reduced proportionately in the ‘corrected improved’ column of the table to give fairer comparison of exposure risks between the two electrodes. It can be seen that there are increases in fume concentrations of some components. Of particular concern are the increases in nickel and lithium. Stern (1981) examined whether the improved electrodes do reduce overall exposure risk. Some forms of nickel have already been described as being associated with cancer. Stern (1981) describes lithium as an extremely biologically active ion. He derives an exposure limit based on maximum allowable blood serum levels of lithium when it is used as a therapeutic drug combined with a safety factor of 100 and obtains a value of 0.05 mg/m3 which is the same as for Cr(VI). A more realistic estimate may be much higher than this, but the point being made is that there must be an improvement in absolute risk. The assessment of exposure risk for the improved electrodes is open to debate, but it is reasonable to consider exposure risk implications of changes of concentrations of components other than Cr(VI) before commercial exploitation of such electrodes.
|
The replacement of potassium with lithium could be applied to gas-shielded, flux-cored wires but the effect of the presence of shield gas and of different flux formulations would require investigation.
The CXL71 wire was compared with the standard Murair 20/9/3 wire for ‘weldability’ and a qualitative assessment report is given in Table 11. In addition to assessment of the overall improvement in risk there are other considerations, e.g. weldability and mechanical strength of welds, that have to be considered before a wire such as CXL71 is considered commercially viable.
|
Using the 1% zinc wire with commercial shield gases containing argon with different levels of oxygen and carbon dioxide in comparison with argon-only shield
A series of experiments was performed using argon, Argoshield 5 and Argoshield 20. The latter two gases are argon based, each with 2% oxygen together with 5% and 20% carbon dioxide respectively. FFR and Cr(VI) concentration in the fume were measured and from this Cr(VI) formation rates [Cr(VI)FR] calculated. Figure 3 shows FFR, which increased as CO2 in the shield gas increased both for the control and for the 1% zinc wire. The FFR for the 1% zinc wire was about three times the FFR for the corresponding control. Higher rates would be expected as zinc is a volatile metal compared to the other main metals in the wire and will evaporate and combine with oxygen to form ZnO fume. It can be calculated that if all the zinc were to form ZnO fume, this would result in a ZnO fume formation rate of 0.685 g/min. Comparison of FFR values for the control and the 1% zinc wires indicates that most of the zinc is forming ZnO fume, although analysis of the fume would be necessary to verify this.
|
The Cr(VI) concentration in the fume, Figure 4, was very low for the 1% zinc wire with argon shield when compared with the value for the control. For the 1% zinc wire, the Cr(VI) concentration greatly increases when the shield gases containing O2 and CO2 are used. For the control wire, and relative to argon shield, the Cr(VI) concentration increased for the 5% CO2 gas and decreased for the 20% CO2 gas.
|
Figure 5 shows Cr(VI)FR with different shield gases. Comparison of different Cr(VI)FR values gives a measure of relative exposure risk. Using argon-only shield there was great reduction in Cr(VI)FR using the 1% zinc wire compared to the control but when using the other two shield gases, containing O2 + CO2, the 1% zinc wire produced substantially higher Cr(VI)FR compared to the control.
|
In order to differentiate between the effects of the O2 and the CO2 in the shield, further experiments were done with argon-based gases containing either O2 or CO2. In addition the effect of using a gas containing 38% helium was studied. An arc shielded with helium produces less ultraviolet light at the ozone-producing wavelengths than a comparable argon-shielded arc, although Smårs (1980) concludes that in the presence of metal vapour the metal vapour spectrum dominates, resulting in similar ozone levels for argon- and helium-shielded arcs.
Using the 1% zinc wire with shield gases containing argon and different levels of either oxygen or carbon dioxide in comparison with argon-only shield; effect of using Helishield 101 (H101)
Figure 6 shows UV–ozone results for the different shield gases. For both wires: H101, Ar + 2% CO2 and Ar + 5% CO2 gases gave less UV–ozone than Ar + 1% O2 and Ar + 2% O2 gases. H101 gave the lowest value, which seems to refute Smårs’s conclusion to some extent. The 1% zinc wire gave much less UV–ozone compared to the control, and this is probably in large part due to the higher FFR for the 1% zinc wire The fume blocks the ultraviolet light from reaching the UV–ozone cell which is located 0.5 m from the arc. Figure 7 gives ozone concentrations near the arc. For both wires: H101, Ar + 2% CO2 and Ar + 5% CO2 gases gave less ozone than Ar, Ar + 1% O2 and Ar + 2% O2 gases. The 1% zinc wire in all cases gave >95% reduction in ozone compared to the control. Suggested modes of action of the zinc are:
|
|
1. Increased fume levels block ultraviolet light responsible for the formation of ozone.
2. Ozone reacts with zinc to form ZnO and O2.
3. ZnO catalytically decomposes ozone.
Figure 8 shows FFR results. Using H101 shield resulted in least fume. The 1% zinc wire produced roughly three times the FFR of the control wire. FFR varied little with shield gas. Figure 9 shows Cr(VI) concentration in the fume. For both wires, H101 shield gave the lowest Cr(VI) concentrations in the fume. Ar + 1% O2 and Ar + 2% O2 shield gases gave the highest Cr(VI) concentrations. For shield gases without O2 the 1% zinc wire resulted in >98% reduction in Cr(VI) concentration, but when O2 was present the reduction was <50%. Combining FFR and Cr(VI) concentration to give Cr(VI)FR resulted in Figure 10 and here the difference between the O2 containing gases and the other gases was more striking. The 1% zinc wire gave 70–100% reduction in Cr(VI)FR when no O2 was present in the shield and >100% increase in Cr(VI)FR when O2 was present. Suggested modes of action of the zinc are:
|
|
|
1. Increased fume levels block ultraviolet light responsible for the formation of ozone, leading to lowered oxidizing potential in the vicinity of the arc and hence less Cr(VI) formation.
2. Reaction of ozone with zinc to form ZnO, leading to lowered oxidizing potential in the vicinity of the arc and hence less Cr(VI) formation.
3. Reaction of zinc with oxygen to form ZnO, leading to lowered oxidizing potential in the vicinity of the arc and hence less Cr(VI) formation.
4. Reduction of Cr(VI) directly by zinc.
However, this would not account for the finding that the 1% zinc wire enhances Cr(VI)FR with oxygen-containing shield gases compared to the control wire, even though measured ozone levels with the control are much higher. More detailed consideration of the chemical processes occurring in the arc and shield gas zone may explain this anomaly.
In conclusion, the use of 1% zinc in the tubular wire described and under the conditions described (one voltage setting) gave >95% reduction in ozone in all cases. The 1% zinc wire gave >70% reduction in Cr(VI)FR compared to the control wire, provided the shield gas contained no oxygen. When the shield gas did contain oxygen, 1% zinc enhanced Cr(VI)FR, resulting in more than double the amounts produced when welding with the control wire. Helishield H101 was found to be the most effective gas in terms of reducing occupational exposure to fume, ozone and Cr(VI). Commercially, Ar/He/CO2 mixtures are one of the recommended shield gas mixtures for stainless steel welding but others can contain oxygen.
Effects on Cr(VI) and ozone of different levels of zinc in combination with Helishield 101
FFR and Cr(VI) concentration in fume was evaluated for the five experimental wires. Ozone concentration was also determined. Moore (1997) repeated the experiments, except ozone was not determined. A summary of the results is given in Table 12. Figure 11 shows variation of Cr(VI) concentration in fume with zinc content of wire and Fig. 12 shows variation of FFR with zinc content of wire. Cr(VI) concentration in fume falls with increasing zinc in wire, although there are considerable differences in the measured effects between our results and Moore’s. FFR rises with increasing zinc content and this is attributable to the vaporization of the volatile zinc, which oxidizes and contributes to the fume. Again there are considerable differences between our results and Moore’s. Moore observed that the fume from the 0.018% zinc wire was a brown colour, indicating iron oxide, and the fume from the 0.9% zinc wire was a pale grey, indicating a high concentration of zinc oxide (ZnO).
|
|
|
Figure 13 shows variation of Cr(VI)FR with zinc content of wire. Our results indicate a continuous reduction in Cr(VI)FR as zinc concentration in the wire increases. Moore’s results indicate that there is an optimum value for zinc concentration in the wire intermediate between the two extremes. Obviously further work is desirable to confirm which results are more accurate. Cr(VI) is formed from Cr and Cr(III) by oxidation. Oxygen and ozone can both bring about this oxidation. Modes of action for the zinc, a strong reducing agent, in achieving lower Cr(VI) concentrations in the fume have already been suggested above.
|
As discussed in the section concerning the low-fume, Murair-type wire, there is a need to assess more fully the overall exposure risk associated with changes in wire compositions. Use of zinc-containing wires can give rise to fume containing high zinc concentrations in the fume compared to non-zinc wires. The OEL for ZnO is 5 mg/m3 so sufficient control of fume exposure should result in adequate control of zinc exposure when using metal-cored wires with zinc additions. Excessive exposure to ZnO fumes can result in the acute health effect known as ‘metal fume fever’, a temporary condition with irritation of the respiratory tract and influenza-like symptoms including fever. Recovery usually occurs within 24 h. There may be changes in the concentrations of other fume components that have a bearing on exposure risk. In addition to Cr(VI) reduction, use of the zinc-containing wires containing 0.09% or more zinc resulted in very large reductions in ozone compared to the 0.018% zinc wire, giving additional benefit in terms of exposure risk reduction.
Even if zinc additions did give substantial reduction in overall occupational exposure risk there are other considerations in relation to such wires, which would need to be evaluated before commercial exploitation. The Helishield H101 (BOC Gases) shield gas used here was found to be the best at achieving reduction of Cr(VI) but it is expensive compared to Ar/CO2 shield gases. Alternative shield gases may be preferable. It is also desirable to evaluate the wires under a range of realistic welding conditions. Cr(VI) can be increased by the addition of zinc to wires in some circumstances. The weldability and mechanical integrity of zinc-containing wires and cost implications of using them would need to be examined before such wires could be used commercially.
Research into the effects of zinc additions to stainless steel flux-cored wires as opposed to the metal-cored wires may also demonstrate reductions in Cr(VI). Oxides in the flux may react with the zinc, preventing it from reducing Cr(VI) or ozone. The presence of chlorides in the flux could lead to formation of ZnCl, which has a lower OES than ZnO. Consideration of this and other fume composition changes resulting from the use of zinc would have to be taken into account in an overall risk comparison between wires with and wires without zinc additions.
Stainless steel welding using tubular wires has become more commonly used in recent years and Cr(VI) concentration in fume from such wires can be several per cent of the fume. There is much concern over control of exposure to Cr(VI). There is considerable scope for the application of the Cr(VI) control measures examined above provided the concerns expressed in the discussion are adequately addressed.
Acknowledgements—This study was carried out as part of work supported by the Health Ministry of Iran and EPSRC Grant GR/H93040. Welding equipment and consumables were supplied by ESAB(UK).
|
FOOTNOTES |
|---|
* Author to whom correspondence should be addressed. e-mail: j.h.dennis@bradford.ac.uk
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
American Welding Society. (1979) Laboratory methods for measuring fume generation rates and total fume emissions of welding and allied processes (F1. 2–79). Miami, FL: American Welding Society.
Dennis JH, French MJ, Hewitt PJ, Mortazavi SB, Redding CAJ. (1996) Reduction of hexavalent chromium concentration in fumes from metal cored arc welding by addition of reactive metals. Ann Occup Hyg; 40: 339–344.
Goralski KA. (1991) The interaction of ozone and zinc in MIG welding. Geelong, Australia: Deakin University.
Griffiths T. (1991) Development of stainless steel welding electrodes having a low level of toxic chromium in the fume. Strasbourg Seminar on Welding Fume: Effects, Control and Protection, paper 6. Abington, UK: The Welding Institute.
Hirst AA. (1990) Fume formation in flux cored metal inert gas welding. Ph.D. thesis, University of Bradford, UK.
Kazantzis G. (1972) Chromium and nickel. Ann Occup Hyg; 15: 25–29.
Kimura S, Kobayashi M, Godai T, Minato S. (1979) Investigations on chromium in stainless steel welding fumes. Weld J; July:195S–203S.
Moore T. (1997) Control of occupational exposure through process modification of metal inert gas (MIG) welding. Department of Environmental Science Student Report, University of Bradford.
Mortazavi SB. (1995) Interelationship between ultra-violet, ozone and hexavalent chromium in metal inert gas (MIG) welding. Ph.D. thesis, University of Bradford.
Mortazavi SB. (1997) Engineering control of occupational exposure to welding fume by process modification. Weld World; 39: 297–303.
Redding CAJ. (1995) Ozone and hexavalent chromium formation in continuous arc welding. M.Phil. Thesis, University of Bradford.
Smårs E. (1980) The problem of ozone in the MIG welding of aluminium. Research report GIF-0500, AGA AB Innovation, Sweden
Stern RM. (1981) Occupational health risk assessment. Report 81.31, Danish Welding Institute, Denmark.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. S. YOON, N. W. PAIK, and J. H. KIM Fume Generation and Content of Total Chromium and Hexavalent Chromium in Flux-cored Arc Welding Ann. Hyg., November 1, 2003; 47(8): 671 - 680. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||