1. Introduction
Welding is an important technological process of joining materials by application of heat and/or pressure. Arc welding is a process used to join metals by using an electric arc to create enough heat to melt metal. Burning the arc between the electrode and metal piece produces a very high temperature, which causes melting of the joint area. Sources of optical radiation in the welding arc include: the arc column, the regions close to the electrode, the liquid metal transported across the welding arc, the molten pool, the heated region of the base material around the molten pool, and the heated end of the electrode wire [
1]. This way, the optical radiation from the whole range of ultraviolet (UV), visible (VIS), and infrared (IR) radiation is a byproduct of the welding process. Directive 2006/25/EC [
2], which specifies minimal requirements related to the health and safety of workers exposed to artificial optical radiation, is an obligatory document for all European Community countries. It lays down the criteria and scope of risk assessment, as well as exposure limit values (ELVs) to avoid adverse effects to eyes and skin due to exposure to artificial optical radiation. The Directive has its scientific basis in the International Commission of Non-Ionizing Radiation Protection (ICNIRP) guidelines published in 2004 [
3]. ELVs for UV radiation stated in the 2004 guidelines have been repeated in the ICNIRP statement on UV protection of workers published in 2010 [
4]. Directive statements had to be implemented into the law of each EC country by 2011. Risk assessment is the scientific approach to identify and quantify health hazards due to optical radiation [
5] and assessment of such hazard is obligatory for employers, which means that each source of optical radiation in the working environment should be identified, and risk assessment should be performed. However, the determination of exposure level is often not a trivial task, especially in the case of technological sources like a welding arc, which is not stable in time.
A welding arc is one of the strongest technological sources of artificial optical radiation, especially in the ultraviolet and visible range [
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15]. It creates a wide range of risks for the skin and eyes of both welders and workers or people around the welding workstation [
13,
14,
16,
17,
18,
19,
20,
21]. Strong UV radiation is emitted by the welding arc into the surrounding environment, and a great number of workers at the workplaces, where welding is performed, are exposed to ultraviolet radiation (UVR). Among those workers are not only professional welders, but also non-professional welders, who perform welding occasionally or are engaged in non-welding tasks [
14]. The measurements in the vicinity of the welding arc showed that actinic radiant exposure (i.e., an 8-h dose) for non-welding workers could reach about 360 J/m
2 (which corresponds to 12×ELV, where ELV = 30 J/m
2), and even about 164 J/m
2 (5.5×ELV) in the non-welding areas. This poses a significant health risk to all workers present in the vicinity of the welding area [
15,
20]. Exposure to optical radiation when striking the arc could also pose hazard for the eyes of everybody present in the vicinity [
7,
9,
14,
22,
23].
The worldwide population of professional welders is estimated at more than three million [
21,
24]. However, it is estimated that more than 110 million workers across the world are exposed to welding-related radiation [
25]. This constitutes a significant number of workers who work in a hazardous welding environment. UV radiation emitted by the welding arc is one of the most important physical agents to consider during risk assessment. Exposure to UVR is a risk factor conducive to the development of: erythema [
17,
26], sunburn/burn [
24,
25], actinic keratosis [
24,
26], skin cancer: basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) [
18,
24,
25], cutaneous malignant melanoma (CMM) [
24,
26], cataract [
21], photokeratoconjunctivitis, often known as “welders flash” [
26,
27], and ocular malignancy [
18]. The International Agency for Research on Cancer (IARC) has identified ocular melanoma in welders as the main adverse health effect due to UV radiation emitted by welding arc [
28]. In everyday practice, acute health effects of UV radiation, especially “welders flash”, occur frequently at workplaces where welding is being performed. It affects both welders and workers in the vicinity of the welding workstation [
14].
As was mentioned before, the Directive [
2] refers to the risks to the health and safety of workers due to adverse effects caused by exposure to artificial optical radiation to the eyes and skin. The requirements included in the Directive apply to workers exposed to optical radiation emitted by the welding arc, both welders and workers who perform their tasks near the welding station. The employer is obligated to perform risk assessment of optical radiation for all employees, not only welders, who are exposed to it. An appropriate exposure assessment must be established; in the case of welding, it is usually determined on the basis of the results of measurements made at workstations. The exposure level has to be evaluated in relation to the exposure limit values (ELVs) [
2].
The measurement of optical radiation emitted by the welding arc is not a trivial task, largely because of arc emission, which is unstable in time, especially at the beginning, during arc ignition. The arc ignition process affects the subsequent arc welding process [
29] and subsequently a stable welding process, which is essential for good joint quality [
1]. For this reason, there are several methods of arc stability determination based on measurements and analysis of welding current intensity and voltage [
30], noise emitted, arc light emission [
1], or acoustic emission from the material. According to Suban and Tusek, an ideal welding arc could be characterized by at least the following properties: a uniform material transfer, constant arc length, and no spatters. It was proven that some welding parameters could influence the arc stability. For example the analysis of MIG/MAG (metal inert gas/metal active gas) welding process stability showed that the type of shielding gas influences the stability and should be selected according to the material transfer mode [
30]. Many studies were carried out to measure optical radiation from the welding arc for the purpose of hazard evaluation, and they proved the strong influence of welding parameters on UV radiation emission [
8,
10,
14,
31]. The relative spectrum of arc emission depends mainly on the chemical composition of the welded material, but the absolute intensity of the radiation depends on the wire/electrode diameter, the chemical composition of material from which the electrode is made, current intensity, and arc voltage [
5,
8,
10,
13]. For example, the study of Nakashima et al. proved that the UV emission is more hazardous when the welding materials include magnesium and when the electrodes contain oxides [
14]. Another study by Garcia-Guinea et al. drew attention to the relationship between ultraviolet emission and consumable electrode coatings’ mineralogy [
31]. A strong effective irradiance dependence on the direction of emission from the arc was also proven [
10,
14]. The study of Okuno et al. for MAG welding with CO
2 shielding gas showed that the maximum emission was at 50°–60° from the surface of the base metal. Another study by Nakashima et al. carried out for TIG welding showed the highest effective irradiance at 40° from the base metal, but at the same time, no dependence of effective irradiance versus the angle of inclination with respect to the welding direction was found [
14]. The differences between the angles of maximum emission could be the result of different characteristic of the radiation reflected from the welded materials, because at those angles (40°–60°), both direct and indirect UV radiation (reflected from the welding pool, etc.) were measured.
The changes of UV emission during arc ignition are mentioned in many articles [
7,
8,
9,
10,
12,
14,
15]. However, a literature overview revealed that only one publication, published in 1985 [
7], was focused on the study of optical spectra for arc ignition. At that time, Eriksen made measurements using a diode array rapid scan spectrometer (DARSS), which allowed him to measure the complete spectrum in the UV and visible range in 5 ms [
7]. Eriksen’s study was carried out for MIG welding and proved that after only one arc ignition (lasting less than 50 ms, but of emission exceeding more than 10-times the steady state of arc burning), the ELVs were much exceeded [
7]. However, in the article from 1991, he wrote that the ignition phase of MIG/MAG welding arcs lasts approximately 25 ms [
6]. It seems strange that in the period of about 30, except for one conference paper delivered in 2018 [
23], there was no interest in this interesting phenomenon in relation to UV or blue light hazard evaluation (conclusion drawn from the overview of published studies and research reports). A few studies on optical hazard evaluation from arc welding report a high increase of optical radiation emission during arc ignition [
8,
9,
10,
12,
14,
15], but all of them relate to the results of Eriksen’s study [
7]. Other studies related to UVR exposure assessment exclude arc ignition emission; measurements (or analysis) were started after the arc ignition [
8,
9,
10,
13,
14]. Depending on the study and measuring equipment (polysulfone film, radiometer, CCD spectroradiometer), a single (or continuous) measurement duration of arc radiation (after arc ignition) was from 5 s–50 s, for example: 6 s [
9], 50 s [
12], 10 s [
5], 5 s [
14], and 15 s [
13]. It could be concluded that studies conducted in the past confirmed that:
UVR emitted by the welding arc is highly hazardous not only for welders, but also for non-welders performing the work in the vicinity, which is reflected in UV exposure-related skin and eye diseases found in this population group,
ELVs to arc ignition could result in exceeding allowable daily exposure.
In view of the development of measurement techniques, it is now possible to try to determine the actinic radiant exposure from arc ignition and evaluate the actinic UV hazard and compare the results with those reported in the previous study of Eriksen [
7]. Previous studies carried out by the present authors [
23] indicated that the duration of UV radiation peak of arc ignition could be in the range from 20 ms–220 ms depending on the welding technique, and these results were consistent with the results obtained by Eriksen [
7]. The estimated actinic irradiance during the ignition phase was 1.2–5-times bigger than the average actinic irradiance after ignition (maximum 26-times bigger for the worst case). It corresponded to actinic radiant exposure of 30 J/m
2 reached from 0.5 s–37.5 s under average conditions and 0.1 s for the worst case [
23].
The article presents a new measurement method of UV radiation emitted during the arc ignition phase based both on CCD spectroradiometric measurements of irradiance spectra and on a quick photometer/flicker meter (based on photodiode) for measurements of photocurrent distribution over time. Based on a set of measurements carried out for the MAG-P process for different welding currents, hazard assessment during the welding arc ignition phase was performed using two approaches of actinic radiant exposure determination. Particular attention has been paid to the actinic hazard to the eyes during the aversion response at unintentional exposure of the unprotected eye, which may adversely affect the health of workers in the vicinity of the welding station. The aim of the article is to present a new evaluation method of UV hazard arising from the welding arc ignition phase using a CCD spectroradiometer combined with a photometer/flicker meter.
3. Results
An ignition peak can be identified for each registered distribution of photocurrent over time, although higher actinic irradiances were observed for lower welding currents. The time of ignition peak
ranged from 0.009 s–0.015 s. The total actinic radiant exposure during the adopted 0.25 s (equal to eye aversion response time) after the start of welding arc ignition was calculated by applying both approaches described above. The total actinic radiant exposures calculated for the time longer than the adopted eye aversion time are presented in the
Table A2 and
Table A3 (
Appendix A.3).
The results of the first approach of UV evaluation are presented in
Table 4. The data obtained in the first approach can be characterized by large data scatter. The maximum peak actinic irradiance tends to be biggest for the smallest welding arc current, i.e., 70 A, then it decreases at 100 A, and increases at 120 A, and finally again decreases at 180 A, where it reaches its minimum. At first glance, this was surprising for us, so we repeated the measurement series for those currents a few times. The results revealed the same tendency.
The average actinic irradiance after the ignition phase increased according to the welding current; however, the increase was not simply proportional; in the range between 120 A and 240 A, it seemed to be approximately constant. Therefore, it could mean that it was probably related to the change in the mode of metal transfer in the arc during ignition and increasing amount of welding fumes close to the arc. The decrease of irradiance in the range of welding arc currents between 240 A and 400 A was also observed by Okuno et al. [
10] during the measurement of effective irradiance during MAG welding with solid wire. The authors [
10] attributed this decrease to the change in the mode of metal transfer in the arc from dip mode through globular mode to spray mode.
As the product of maximum peak actinic irradiance and peak duration, the maximum peak actinic radiant exposure for 70 A was about 5.3-times bigger than that for 180 A. However, when we compare the total actinic radiant exposure for different welding currents, the biggest one is for 120 A, and it is 1.9-times bigger than for 180 A (the smallest one). However, the big difference in peak actinic radiant exposure between 70 A and 180 A has been compensated during the arc emission after ignition (over 0.25 s), so the ratio of total actinic radiant exposures decreased to about 1.4. The calculated peak actinic radiant exposure and total actinic radiant exposure
of the welding arc determined with the second approach are presented in
Table 5. Additionally, the last column presents the ratio between total radiant exposures determined using both approaches. The calculations of radiant exposure were based on the same average actinic irradiances
and recorded data as in the first approach. The peak actinic radiant exposure was the biggest for 120 A and approximately the same for 70-A welding current. The smallest one was for 180 A. The ratio between the maximum and minimum of the above-mentioned values was 2.83. Still, when we look at the total actinic radiant exposure, we can observe its increase with welding current increasing, with one exception for 180 A. The biggest one is for 240 A and the smallest for 100 A, giving the biggest/smallest ratio of 2.81.
The last column in
Table 5 presents the ratio of total actinic radiant exposures obtained using both approaches. This way, we could see the differences of the determined radiant exposures for actinic hazard evaluation. It could be noticed that with the increase of the welding current, the ratio decreased from about three (at 70 A) to about one (at 240 A), which proves that radiant exposure using the first approach, especially for smaller welding currents, was overestimated. However, for a longer time of exposure like 3 s, the ratio of actinic radiant exposures was approximately 1.0, which suggests that the approach used for calculation was less significant and the contribution of peak radiant exposure decreased (see
Appendix A.3).
Figure 6 shows the distribution of actinic radiant exposure over time for different welding currents. The slopes of plots present the radiant exposure growth rate for particular welding currents. It could be concluded that for smaller welding arcs, the radiant exposure grew rapidly to higher values during ignition and then slowly increased. Conversely, in the case of bigger welding arcs, radiant exposure grew less rapidly and to smaller values during the ignition time and then increased evenly and almost linearly in the remaining time. The higher the current, the bigger the growth rate. This also indicates that the contribution of peak radiant exposure to total radiant exposure was bigger for smaller currents than for higher currents.
The contribution of peak radiant exposure to total radiant exposure obtained using both approaches is presented in
Figure 7. The first approach of actinic radiant exposure determination for hazard evaluation resulted in a much bigger contribution of peak actinic radiant exposure to total radiant exposure than in the second approach. The maximum difference was about 20% (for 70 A and 120 A) and the smallest about 4% (for 180 A). The decrease of that contribution with increasing welding current is also visible in this figure.
4. Discussion
Due to the adopted measurement geometry, the obtained results of actinic irradiance and total actinic radiant exposure were lower than the maximum results recorded for the MAG-P or MAG processes in similar conditions reported in other articles [
6,
10,
23]. Determination of maximum possible actinic irradiances at the accepted distance from welding and the height of probe above the floor arc was not a goal of this study. However, the relation between the increase of actinic irradiance of the ignition phase and average actinic irradiance of welding action should be approximately the same in all directions of arc emission, which we assumed in our analysis.
The improved method of recording the variability of welding arc ignition allowed for a more accurate mapping of the peak output shape of arc ignition and thus a more precise determination of its duration and the maximum value. The ratio of maximum peak irradiance and average irradiance (during the “stable” phase of arc emission) was in the range of about 6 and 47, which showed that the irradiance peak during the ignition phase could be much bigger than 10, as was stated by Eriksen [
6]. This discovery was possible with the new method of registration. The range of ignition peak duration identified using the new photometer/flicker meter showed that the peak duration was narrower than the ranges determined in previous studies, i.e., 25 ms and 50 ms [
6,
7,
23]. This suggests that actinic radiant exposure of UV during arc ignition evaluated on the basis of the previous results can be overestimated, and the subsequent hazard evaluation can be distorted. A higher accuracy of measurement obtained thanks to the proposed method of measurement can be expected to give a lower measurement error, which is crucial for the determination of expanded uncertainty, which should not be bigger than 30% according to EN 14255-1 [
32]. Taking into account the wide range of uncertainty components related to arc welding (fumes, instability of arc emission over time, arc movement, reflections from welding area and surroundings, etc.), efforts should be focused on reducing any measurement errors related to measurement geometry, measurement processing procedure, and measurement equipment quality.
Another important aspect that strongly influences the determined parameters for UV hazard evaluation is the way in which parameters based on the obtained measurement results were calculated. As was presented in the article, consideration or not of the phase of arc ignition for short duration exposures (like eye aversion response) in actinic hazard evaluation of eye could have critical consequences. For the MAG-P welding process under investigation, omission of peak radiant exposure especially for low welding currents (between 70 A and 120 A) could result in a 44–66% decrease of the calculated radiant exposure. Furthermore, the determined maximum allowed time of exposure will be longer than it should be, and as a consequence, it will lead to increased hazard for workers related to overexposure to UV radiation.
The next important aspect is the determination of actinic radiant exposure used for photobiological hazard evaluation. Integration of actinic irradiances used to determine total actinic radiant exposure in the second approach leads to the more precise determination of the actinic hazard, especially for short exposure time including arc ignition (with special attention paid to aversion response time). For MAG-P welding, it is significant especially for smaller welding currents (70 A–120 A), while it does not seem to be so important for currents above 200 A (see
Table A4 in
Appendix A.3).
The same tendency may be observed for both approaches. As the welding current increases, the significance of ignition peak actinic radiant exposure for total actinic radiant exposure decreases (
Figure 7). The ratio
decreases with increasing welding current, which corresponds to lower contributions of peak actinic radiant exposure to total radiant exposure presented in
Figure 6 and
Figure 7.
However, the results of total actinic radiant exposure during eye aversion time were much lower than exposure limit values 30 J/m
2, which means that a single glance of unprotected eye at arc ignition does not pose a high risk of adverse effect. However, multiple glances could lead to a high risk of acute effects like photokeratoconjunctivitis (“welding flash”), which was stated in other publications [
7,
10,
14]. Unprotected skin, especially the skin of the head or neck, can be exposed to welding arc radiation for a longer time because of the lack of aversion response time. The importance of this fact both for welders and non-welders was emphasized by Tenkate [
15,
20]. Therefore, for this reason, the maximum permissible exposure time was determined in our study (
Appendix A.3). It was in the range of 7 s–13 s (first approach) and 10 s–29 s (second approach). However, as reported in our study and in other published studies [
10,
11,
14,
20], the total daily exposure time of unprotected parts of skin to UV arc radiation could be easily much longer than permissible exposure times.
5. Conclusions
While the exposure is brief for each arc ignition, if it occurs many times in a workday, the total actinic radiant exposure may become sufficient to cause acute adverse effects like photokeratoconjunctivitis, erythema, and burn. If it is repeated for many days, chronic adverse effects of prolonged exposure to UV like ocular malignancy, actinic keratosis, or different kinds of skin cancer could appear. In fact, the total exposure time can easily exceed the permissible exposure time obtained in our study, i.e., 7 s–29 s at the distance of 1 m from the arc. This means that UV radiation from MAG-P arc welding is actually hazardous for the eyes and skin not only for welders, but also for workers in the vicinity of the welding arc.
The presented measurement method using a CCD spectroradiometer and photometer/flicker meter enables determination of radiant exposure of arc ignition radiation. This method is especially recommended for evaluation of UV hazard to eyes during the aversion response time exposure. Additionally, it reduces the uncertainty of welding arc emission measurements for the purpose of determining parameters for photobiological hazard evaluation through the limiting of errors related to the measurement method. The proposed method could be easily expanded to other hazards related to optical radiation emitted by welding arc, especially to blue light hazard evaluation.
The way of determination of UV radiant exposure during the arc ignition time, based on the obtained measurements results, has significant influence on hazard evaluation. For the MAG-P welding process and for the welding currents used in the study, it was especially significant for exposure duration below 3 s. The measurements using the same method should be continued for other welding techniques and welding parameters to determine characteristic dependencies occurring during arc ignition.
It is important to promote knowledge about the new methods of photobiological hazard evaluation among health safety officers, employees, and other experts who perform risk assessment, as this will contribute to the quality of hazard evaluation and will help to improve occupational safety.