UC Berkeley

Linear mixed models were used to analyze data compiled from international sources to simultaneously estimate the fixed effects, associated with process characteristics and sampling regimen, and the variance components, associated with the random effects. The fixed effects explained 55% and 49% of variation in TP and Mn exposures, respectively. The country, industry/trade, ventilation condition, type of work/welding process, and material employed appeared to be the major factors affecting exposures to TP and Mn. Measurements in the U.S. were generally higher than those in other countries. Exposure to TP was 64% higher in enclosed spaces and 42% lower with local exhaust ventilation, was higher among boiler makers, and was higher when a mild-steel base metal was used. Exposure to Mn was 318% higher in enclosed spaces and 67% lower when local exhaust ventilation was present. The measured Mn air concentration was significantly related to the composition of the consumables, but not to the base metal. Resistance welding produced significantly lower TP and Mn exposures compared to other welding processes. After controlling for fixed effects, variance components between groups and between individual workers within a group were reduced by 89% and 57% for TP, and 75% and 63% for Mn, respectively. The within-worker variance component in Mn exposure was three times higher than that of TP, indicating that day-to-day and within-day variations in TP and Mn exposures were influenced by different factors that were not captured equally well by the mixed models for these two contaminants. Interestingly, exposures to TP and Mn had not changed over the 40 years of observation. The estimated probabilities of exceeding occupational exposure limits were very high (generally much greater than 10%) for both agents. Welding exposures to TP and Mn vary considerably across the world and across occupational groups. Exposures to both contaminants have been and continue to be unacceptably high in most sectors of industry. Because exposures to the two agents have different determinants, separate control strategies should be used for reducing welders' exposures to TP and Mn.

A respiratory health survey conducted in an automobile assembly plant in 2000-2001 found that welders had elevated rates of self-reported respiratory symptoms compared to painters and assembly workers. Subsequently, the ventilation system was improved at the body weld department. In a follow-up study, particle spatial distributions were analyzed, following a mapping protocol developed specifically for this work place, to evaluate the effectiveness of the changes. Significant temporal and spatial variations were observed. Temporal variation during a shift was monitored with over-shift stationary sampling at fixed locations. Spatial variation was evaluated with 1-minute time-weighted average particle concentrations measured throughout the process areas (212 locations). The arithmetic spatial mean across 212 locations for the respirable particles varied from 305μg/m3 to 501μg/m3 on six sampled days, with a standard deviation of 71μg/m3, indicating that the difference between before and after countermeasures must be at least 191μg/m3 in order to be considered statistically significant at the given sample sizes. The available data were not sufficient to evaluate the reduction of the particle concentrations after the countermeasures. The map of particle mass concentration revealed several high concentration areas, requiring further investigation and potentially higher level of controls. Resistance welding needed to be effectively controlled as it could be the major particle emitting source in the facility. The map of submicrometer (0.014μm to 1.0μm) particle count concentration presented different patterns from that of respirable particle mass concentration, indicating that the submicrometer particles tended to be more evenly distributed over the process areas. Workers not in close proximity to intensive welding operations might be exposed to fine particles at levels higher than had traditionally been thought. Mapping was demonstrated to be an effective method to assess particle spatial distributions. A well-designed sampling protocol is critical in order to achieve the specific aims of a mapping study.

A pilot study was conducted in three Chinese manufacturing facilities to characterize welders' exposure to particulate matter (PM) and airborne manganese (Mn) from common welding processes, with emphasis on Mn distribution in submicrometer particles. Particle air concentration was measured as 8-hour time-weighted averages (TWAs) for total and respirable particles. Mn air concentration (8hr TWA) was measured as Mn in total and respirable particles. Mn size distribution was assessed using multi-stage impactors with cut-points of 0.25μm, 0.5μm, 1.0μm and 2.5μm. The welding processes investigated were shielded metal arc welding, gas metal arc welding, submerged arc welding and plasma arc welding. Overall arithmetic means (AMs) across processes and factories were 2.58 mg/m3 (range: 0.338 mg/m3 - 27.8 mg/m3, GM: 1.28 mg/m3, GSD: 3.27) and 1.46 mg/m3 (range: 0.011 mg/m3 - 14.7 mg/m3, GM: 0.698 mg/m3, GSD: 3.37) for total and respirable particles (8hr TWAs), respectively. Overall AMs for Mn air concentrations were 0.122 mg/m3 (range: 0.001 mg/m3 - 1.30 mg/m3, GM: 0.058 mg/m3, GSD: 3.40) and 0.073 mg/m3 (range: 0.001 mg/m3 - 0.650 mg/m3, GM: 0.036 mg/m3, GSD: 3.33) for Mn in total and respirable particles, respectively. Particle and Mn concentrations varied over 4-fold by process. Shielded metal arc welding produced higher air concentrations for both agents compared to gas metal arc welding and submerged arc welding. Plasma arc welding resulted in the lowest concentrations. Manganese was found to be more concentrated in respirable particles than in total particles. Four percent of the particle mass of total particles was composed of Mn, while it was 5% for respirable particles. Data from the multi-stage impactor further revealed that majority of Mn mass, 97% for plasma arc welding and over 85% for shielded metal arc welding and gas metal arc welding, was distributed in particles smaller than 0.5μm. Percentage of particle mass made of by Mn increased three to twenty times as particle size decreased from 2.5μm - 10μm to <0.25μm. These findings are of great significance in that Mn primarily targets the central nervous system and Mn in small particles in the nano-size range has higher potential to reach the brain than Mn in larger particles. Therefore, welders' risk of developing neurological effects due to exposures to Mn may be higher than it had been traditionally thought. It was also observed that Mn size distribution varied by processes. Plasma arc welding and gas metal arc welding could be more hazardous than submerged arc welding when particle and Mn air concentrations are comparable. Shielded metal arc welding should be evaluated and controlled with high priority.