Annals of Occupational Hygiene Advance Access originally published online on December 14, 2005
Annals of Occupational Hygiene 2006 50(3):219-229; doi:10.1093/annhyg/mei059
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Original Article |
Workplace Protection of Air-Fed Visors Used in Paint Spraying Operations
1 Health and Safety Laboratory, Harpur Hill, Buxton SK17 9JN, UK and 2 Health and Safety Executive, Bootle L20 3QZ, UK
* Author to whom correspondence should be addressed. Tel: +44 (0)1298 218 329; fax: +44 (0)1294 218 393; e-mail: nick.vaughan{at}hsl.gov.uk
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTIONAir-fed visors are commonly used for protection against exposure to airborne isocyanates during paint spraying. Protection levels for this class of equipment are theoretically adequate, yet isocyanate sensitization in this occupation still occurs. The work reported here set out to establish the level of respiratory protection that is achieved during real paint spraying activities when air-fed visors are used. The work also examined the effects of reduced air supply flow rates on this type of respiratory protection. The workplace study highlighted common problems that occur when attempting to measure protection factors, and process and interpret the collected data. Many of the environments included in this study did not exhibit challenge concentrations high enough to reliably measure the workplace protection factor of this class of device. When detection limits are taken into consideration, the remaining field data suggest that an assigned protection factor in the region of 40 may be appropriate. When well maintained and used in accordance with the manufacturer's instructions, air-fed visors are capable of providing a good level of respiratory protection. The protection given by air-fed visors is strongly dependent on the air flow supplied to them. Laboratory measurements demonstrate that protection falls as the air supply falls. This is a gradual process and does not suddenly occur at any particular air supply flow. Observations made during the field tests indicate that there may be other activities associated with the spraying process that need to be taken into consideration when looking for sources of respiratory sensitization.
Keywords: workplace protection factor • assigned protection factor • air fed visor • paint spraying • isocyanate
|
INTRODUCTION |
|---|
It is estimated that there are
30 000 people exposed to isocyanate-containing paints during spraying activities in motor vehicle repair in the UK (Cowie et al., 2005
). Isocyanates are known respiratory sensitizers. The UK Health and Safety Executive (HSE) guidance on the control of exposure to isocyanate paint sprays promotes ‘best practice’ for the selection and use of respiratory protective equipment, recommending that full facemask airline breathing apparatus be used (HSE, 1999a). However, for reasons of comfort and wearer acceptability, compressed air-fed visors are almost universally used to provide respiratory protection during such operations. The assigned protection factors (APFs) recommended in BS 4275 (BSI, 1997) and HSG53 (HSE, 2005a) are based on professional judgement and are not derived from workplace performance data for this class of equipment.
Previous laboratory studies (Bolsover, 1996) concluded that air-fed visors were capable of providing high levels of protection when used in accordance with the manufacturer's instructions and were well maintained. However, in practice, people are still becoming sensitized to isocyanates, so there was a clear need to gather workplace performance data to establish the effectiveness of air-fed visors.
For this application, there was no readily available instrument for direct ‘real-time’ measurement of isocyanates, so alternative approaches were adopted involving measurement of the volatile organic compounds (VOCs) universally present during spraying.
Various sites were identified including sprayer training schools and vehicle repair centres and a series of measurements made during actual paint spraying operations. By sampling both inside and outside the visor and comparing the results, a workplace protection factor (WPF) can be calculated.
|
AIR-FED VISORS |
|---|
The visors used in the paint spraying industry are generally of a similar construction. They consist of a head harness, a moveable visor and a belt with attached filter/regulator for connection/control of the air supply. The air supply must be from a clean, breathable source and in accordance with the visor manufacturer's specification for pressure and flow. A plastics tube runs from the belt filter/regulator into the visor, where it discharges, usually through a foam diffuser, allowing the air to flow over the inside of the visor. This not only provides air for the wearer but also helps prevent misting, which can obscure vision. Excess supplied air and exhaled air passes through or around the loose fitting faceseal to the ambient atmosphere.
The main physical differences between different models of visor relate to the method of sealing against the face, with several types being available. These can be foam pads shaped to fit the contours of the face, or cloth/paper/plastic material with an elasticated edge to hold it against the face. Both can be equally effective in use, the choice lying with the user for personal comfort.
European performance standards exist which define classes and levels of protection for this type of RPE. In this study, all the visors used were classified as providing equivalent protection to that required by standards EN 270 (CEN, 1994) or EN 1835 ‘LDH3’ (CEN, 1999), i.e. a maximum inward leakage of 0.5%, giving a nominal protection factor of 200. This type of equipment is currently considered to have an APF of 40 (BSI, 1997; HSE, 2005a).
During the survey, the air supply flow to the visors was not adjusted at any time but left at the setting normally used by the wearer. The manufacturers stated minimum design flow rates for the devices were of the order of 180 l min–1. Measurements taken on-site showed that none of those tested were below this level, and some flows were in excess of 220 l min–1.
Where ‘pumped’ samples were involved in the measurement of WPF, each visor had to be fitted with a sample port and ball probe of the type specified in current European Standards (e.g. EN 1835). The ball probe was positioned so as to be close to the wearer's lips when the visor was worn. This allowed a continuous sample to be taken from the wearer's breathing zone. For comparison, another sample line was positioned outside on top of the visor browguard, and fitted with a particle filter to prevent spray droplets passing into the line. In this position (visible in Fig. 3), the ambient sample was within the immediate working environment of the sprayer but clear of any clean air wash from the visor.
|
TEST SITES |
|---|
Measurements were made at a number of field trial sites, covering a range of differing types of engineering control. These included commercial repair shops, paint manufacturers, spray equipment manufacturers, spray training establishments and Ministry of Defence (MOD) repair shops. The type of spray booth used in these establishments covered the range of ventilation systems available (cross flow, down flow and wet back), but the level of ventilation provided at each site was not measured as it was outside the scope of this work. Activities undertaken in these booths varied from component painting and small-area accident damage respray to complete repainting of large military vehicles.
Figures 1 and 2 show photographs taken during the trials on-site.
|
|
|
SAMPLING AND MEASUREMENT METHODS |
|---|
Isocyanate measurement
The methods available for measurement of isocyanates inside the visor either lacked sensitivity or the ability to produce results in real time. The situation was made more difficult by relatively short spraying times and efficient ventilation systems. The accepted method for isocyanate measurement is that described in Methods for the Determination of Hazardous Substances (MDHS) 25—organic isocyanates in air (HSE, 1999b
). This involves the collection of a sample on an open-faced filter for later laboratory analysis, and although this was used to make some measurements of isocyanate exposure levels during spraying, it was not considered suitable for taking samples from within the air-fed visor, where samples would be prone to enhanced denaturing owing to the moisture in exhaled breath.
From open-faced filter samples (MDHS 25 method) taken at the sprayer's chest, detectable levels of isocyanates were measured at four out of five spray booths visited, and results are shown in Table 1. The calculated minimum protection factor required to keep the exposure level at or below the maximum exposure level (MEL—now known as WEL—workplace exposure level; see HSE, 2005b) of 0.02 mg m–3 (8 h TWA) is also shown, assuming the upper limit of the measured ranges at each site.
|
BSI and HSE guidance (BSI, 1997
; HSE, 2005a) allocates EN 270 or EN 1835 LDH3 type air-fed visors an APF of 40.
WPF measurement
Ideally, the standard laboratory method for measuring inward leakage would have been used, i.e. using sulphur hexafluoride (SF6) as a tracer gas, as this relies on well-established technology with known performance and detection limits. However, this would have involved flooding the work area with substantial quantities of the gas, or enclosing the sprayer's head in an atmosphere containing SF6, and was considered impractical.
Several different WPF measurement methods were identified, which did not involve the direct assessment of isocyanates but were considered capable of giving a representative indication of the air-fed visor performance in actual use. These methods relied on the measurement of one or more VOCs that are invariably generated by the spraying activity simultaneously with isocyanates. By sampling the external atmosphere for VOCs and comparing it with a simultaneous VOC sample taken from within the air-fed visor while the wearer is spraying, a WPF can be calculated. Each of the methods tried is described briefly below, with an outline of problems encountered and an assessment of the reliability of the results obtained.
Adsorption tubes
Chromosorb 106 (Advanced Minerals, Goleta, CA) chemical adsorption tubes were used in pairs to obtain both external and internal samples for each test. Samples were quantitatively analysed for a range of different compounds by thermal desorption and gas chromatography—flame ionization detector. The tubes can be used either diffusively or pumped, depending upon the expected concentration of the chemical(s) to be detected. We were expecting relatively short 10–15 min tests, so initially we used pumped samples to ensure sufficient collection of VOCs for more reliable analysis. Diffusive samples were also taken for comparative purposes. Adsorption tube data can only generate a time-weighted-average measurement of exposure.
At each site, a range of WPF results was obtained from pumped adsorption tube samples, but these were found to be highly inconsistent within each individual test. When the four or five highest analyte concentrations recorded for each pair of test samplers were used to calculate protection factors, the results were very different (ranging from a calculated WPF of <1 to over 700), when broadly similar results would have been expected. In addition, the few diffusive samples that were taken indicated significantly higher peak levels of analyte, suggesting that there was a problem with breakthrough or incomplete collection in the pumped tube samples.
Further measurements were made at one site, where work patterns offered the best opportunity of collecting significant quantities of material, using non-pumped tubes. However, in-facepiece samples from this environment were again usually below the limit of detection of the analytical method, and consequently no reliance could be placed on calculated WPF values. These values showed similar range and variability to those derived from the pumped tube samples.
We decided to reject information from the adsorption tube samples as unreliable, and until further investigations can be carried out into the source of these inconsistencies, we consider that this approach to the measurement of WPFs in similar situations should not be relied upon.
Photoionization detector
Portable photoionization detectors (PIDs) were used to sample total VOCs and log the measured concentrations for downloading onto a PC, allowing the results to be examined and analysed later. Two devices were used simultaneously, one to sample the external atmosphere that the wearer was exposed to during spraying and a second unit to sample from inside the air-fed visor. (These samples were collected simultaneously with one or other of the adsorption tube samples described above. We had intended to compare the WPFs derived by the different methods as a means of corroboration.) At the end of each test run, the logger memory was downloaded onto a laptop computer and the PIDs reset ready for the next run. The data were in the form of continuous records of instantaneous VOC concentrations inside and outside the visor.
The two PIDs and their sampling pumps were worn by the sprayer during the tests to keep sample lines as short as possible, using a carrying jacket with adjustable pocket locations. This jacket is shown in Fig. 3, and both PID and adsorption tube samplers are also visible. For general spraying activities, the jacket was worn with the sampling devices on the back, but for those times when under vehicle spraying was being done, the sampling devices were worn on the front. Sampling tubes were routed such that they did not interfere with the spraying operations.
|
The PID samples produced apparently useable results, but in a number of cases, the calculated WPF values were unquantifiable as the measured levels of VOCs inside the facepiece, or outside, or both, were below the detection limit of the PID. One site in particular had very low levels of VOCs present inside the booth when spraying, attributed to the high level of ventilation in the booth removing the overspray. This low challenge concentration artificially reduces calculated WPFs for that site.
In order to make use of such unquantifiable results, a commonly applied data treatment is to substitute a minimum value for any ‘zero’ result, to produce a worst case WPF calculation. The minimum value chosen in our case (0.05 p.p.m.) was half the resolution limit of the PID for VOCs, because any measured values of zero p.p.m. could in fact be up to this value. Actual WPFs cannot be less than those calculated after such substitutions and may, in fact, be substantially higher. A typical result is shown in Fig. 4, for data collected at Site 4. For this sample, in-visor measurements were zero throughout and have been substituted by 0.05 p.p.m. Consequently, the plotted values of protection factor are an entirely artificial function of the externally measured VOC concentration, and are inevitably lower than the real level of protection which was being achieved at this time.
|
In Fig. 4, the bar shows the wearer's activity—spraying, not spraying or out of the booth. As the external concentration increases so does the WPF, which appears to range during spraying activity from not <10 up to a peak of 330. Other sets of results where this substitution method has been used, and again where only zero was measured inside the visor, gave a range of WPFs of 40–135. In all these cases, the resulting WPF value is purely a function of the prevailing external concentration and the fairly arbitrary substitution of the 0.05 p.p.m. for all ‘zero’ values. No significance can be attached to these calculated WPFs, other than them being an absolute worst case measurement.
|
USE OF INFRARED DETECTORS |
|---|
A further series of field tests was conducted using MIRAN infrared absorption spectrophotometers (Quantitech, Milton Keynes, UK) to measure one specific VOC, n-butyl acetate, which had been detected in significant quantities in all the work environments. This measurement technique potentially had a comparable detection limit to the PIDs and could similarly be used to give real-time concentration measurements.
However, there were some drawbacks to this method. The instruments could not be sited close to the wearer inside the spray booth as they are neither portable nor intrinsically safe, and they are powered by mains electricity. Additionally, while they are capable of detecting a range of components, they could only, in this application, be used to quantify a single component. The units had to be sited some distance from the actual spraying operation, and required a sufficient length of sampling line, to allow the sprayer to carry out his work unimpeded.
Two MIRAN infrared spectrophotometers were set up to monitor n-butyl acetate. One unit was used for the internal sample (i.e. from inside the visor) and the second for the external sample. The outputs from both were logged on a small portable data logging unit (Eltek, Cambridge, UK), which was subsequently downloaded onto a PC for analysis. The detectors required 10 l min–1 sample flow, so suitable pumps were required to pull the sample down the long, narrow-bore sample lines.
In use, the sample lines had to be robust and securely attached to the wearer. This was achieved by fixing them to the compressed air supply line for the visor (which the wearer already had to drag behind him as he moved around the job), to the waist belt and then running them up over the wearer's shoulder to the visor. The external sample line was fixed to the top of the browguard so that it extended just over the brow, while the internal line was attached to a smaller bore tube that could be fitted inside the visor, and terminated in a ball probe close to the wearer's nose and mouth. Again a particle filter was fitted to the end of the external sample line to prevent solids/droplets from entering and contaminating the sample.
Two field sites were selected, one of which had been previously visited. Both were MOD sites involved with large-scale vehicle repair, which meant that longer spraying times could be expected. This would possibly be helpful in obtaining more meaningful results.
|
MIRAN DATA ANALYSIS |
|---|
Simple processing of raw data from the IR spectrophotometers
The Miran used for the in-visor sampling was set to a sensitive range and appeared to be prone to zero drift (tending to shift negative with time), so the results from this unit had to be corrected to take this into account. Examination of the detailed recordings of in-visor data allowed stable minimum values to be identified during periods of no exposure; recorded data were then adjusted such that this minimum corresponded to zero. The range of results obtained in several sets of measurements from two sites are summarized in Tables 2 and 3.
|
|
The operations at Site 5 (Table 2) occasionally involved spraying the underside of a large truck. This required the sprayer to spend some time on his back in awkward conditions. The spraying action was in an upwards direction and it was under these conditions that the highest external levels of solvent were encountered. This was partly due to the sprayer's close proximity to the component being sprayed and partly to the reduced ventilation under the vehicle. As a result of these high external levels, the calculated protection factors are also relatively high at this time.
When the spraying operation moved to the sides of the truck, the external concentrations reduced accordingly, although some measurements in excess of 100 p.p.m. were still recorded. This resulted in the calculated protection factors also being reduced overall. However, relatively high protection factors in excess of 2000 were apparently still achieved at times when the internal sample was low (0.1 p.p.m.).
At Site 6 (Table 3), external levels were generally lower than those of Site 5 owing to the different types of components being sprayed. Most of these were of a smaller scale or smaller surface area compared with the large truck encountered at Site 5. As a result, the ventilation system (in this case a wet back type) was able to clear the overspray more efficiently. The maximum recorded external level was just 76 p.p.m., with a low of just 2 p.p.m. Calculated protection factors ranging from 3 to 319 resulted from these samples.
In all, in excess of 8000 data pairs were generated using the MIRAN analysers, sampling at a rate of 1 s–1. Typical concentration measurements are presented in graphical form in Figs 5 and 6, which show the internal and external concentrations plotted against time for two of the sampling runs.
|
|
From these plots of internal and external sample concentrations, it can be seen that the level of internal sample is relatively very low and does not change significantly with changes in external concentration; there is no evidence of internal concentrations tracking external concentrations. On this basis, it would, therefore, be reasonable to conclude that the visors are consistently providing a good level of protection regardless of the task being performed. However, this conclusion disagrees with some of the low numerical values of protection factor in Tables 2 and 3. This indicates that the situation is considerably more complex than a simple intuitive analysis would suggest.
Detection limit of field measurements
The detection limit of an analytical system is commonly defined as three times the standard deviation of a blank measurement (Royal Society of Chemistry, 1987). A complicating factor in the field measurements was that the ‘blank’ level could not be determined with certainty and that there was unexpected interference at the IR wavelength used to detect n-butyl acetate (8.2 µm) between this substance and airborne moisture. This was most apparent in the in-facepiece samples because of the moisture in the wearer's exhaled breath. The size of this effect would vary from one set of measurements to the next, depending on the condition of the air when the MIRAN was zeroed at the start of a test, but maximum and minimum possible values were established in the laboratory.
If there was no such moisture interference operating, the detection limit of the MIRAN system in this application is likely to be in the region of 0.1 p.p.m. of n-butyl acetate. In the worst case interference situation, the additional variability introduced by changes in moisture in the sample would be in the region of 1.3 p.p.m. n-butyl acetate. A probable representative value that can be applied to our field test results is 0.5 p.p.m. n-butyl acetate. These values are similar in range and magnitude to the in-facepiece concentration measurements observed during our workplace studies (0.01–1.14 p.p.m.). It is perhaps questionable whether significant levels of n-butyl acetate have been detected inside the facepieces at any time during our studies.
Data treatment and interpretation
In spite of the doubtful significance of much of our in-facepiece data, it was collected in real working situations, and at the very least WPFs calculated from it could represent a worst case performance for this type of RPE, operating in these environments at the flow rates observed. As with the PID data, the initial step in the treatment of the collected data was to substitute all zero-corrected measurements below the assumed detection limit by the assumed detection limit itself. Again, as with the PID data, the resulting values of WPF calculated from MIRAN data pairs where this substitution has been made will be lower than the real WPF that was achieved at the time. The data treatment method described by Vaughan and Rajan (2005) was then applied. This data treatment method rejects unreliable data, using criteria based on the detection limits of the measurement system being used. Applying an iterative approach to assumed PF, it arrives at a supportable estimate of given percentiles of the remaining distribution of measured WPFs (such as the 5th percentile, conventionally taken to be the APF).
Data pairs are only rejected if both the ‘in-facepiece’ concentration is at the detection limit and the challenge concentration is less than the product of the assumed PF and the detection limit. All data pairs where ‘in facepiece’ concentration is above the detection limit, or challenge concentration exceeds the detection limit-assumed PF product, are accepted.
Table 4 shows the outcome of applying this data treatment system to the MIRAN field data. Three different detection limits for n-butyl acetate have been applied to the data:
- 0.1 p.p.m., assuming there was no effect of moisture in the sampled air. This is unlikely to have been the case in our measurements;
- 0.5 p.p.m., assuming the most probable magnitude of moisture interference effect in our measurements;
- 1.0 p.p.m., assuming the highest likely magnitude of moisture effect in our measurements.
|
|
The assumed detection limit of 0.5 p.p.m. is probably most closely representative of the circumstances of these field trials. This provides an estimate of the APF for this class of device, based on field data from seven sets of measurements at two sites, incorporating more than 8000 data points, if the assumption of 0.5 p.p.m. detection limit is reasonable. By definition, 95% of the calculated WPF values exceed this protection factor.
Vaughan and Rajan (2005) also set out criteria for the design of WPF studies to minimize the complications of the detection limit of field measurements and inherent bias in derived PF values. The data from this study of air-fed visors can be assessed retrospectively against these criteria.
We must first assume a value for the detection limit of in-facepiece measurements (Ui); based on the information above, a value of 0.5 p.p.m. is assumed. Second, we must choose a value for the WPF we are trying to measure with confidence (expected PF). In what follows, we assume a choice of two values; 200 is the nominal PF for this class of RPE, derived from the permitted maximum inward leakage in standard tests; 40 is the currently accepted APF. Based on these values, a limiting lower external concentration (Cu) can be calculated, below which data pairs should be rejected.
Table 5 summarizes the outcome of applying these parameters to the complete set of MIRAN air-fed visor data, rejecting all data where the external concentration was below Cu.
To have confidence in measurement of a PF of 40 when using this detection system, external concentrations must exceed 200 p.p.m. However, when data with challenge concentrations <200 p.p.m. are eliminated from the dataset, only 11% of data remain and the resulting value of the APF (5th percentile) for this residual data is 312. This is far higher than the expected PF value of 40. The only inference we can make from this result is that the real APF value is not <40, and is probably higher.
Adopting the higher value of 200 for expected PF, external concentration must exceed 1000 p.p.m. for confident measurement of PF. For the data collected in this study, the maximum concentration observed was 663 p.p.m.; there are no residual data from which to calculate an APF value.
|
OVERALL WPF DISCUSSION |
|---|
As there was no suitable direct measurement method for isocyanates, other components generated during the spraying process have been used as surrogates. All this work is based on the assumption that these surrogate compounds are reliable indicators of the performance that would be achieved by the RPE against isocyanates. This assumption is not unreasonable, as both are vapours.
As with all studies of this kind, it cannot be guaranteed that the work methods and techniques adopted by the test subjects during tests were representative of their normal practice. However, no unusually careful modes of operation or techniques were observed during these tests. Indeed, some examples of poor practice were seen, such as lifting or removing the visor while still in the contaminated area.
In common with virtually all analytical and sampling techniques, those used in this study have an inherent detection limit. Where one or more of the quantities being measured are of a similar order of magnitude to the detection limit, derived data will be unreliable and should not be included in the population of valid data for WPF assessment. This effect has been a significant complication in this study, for adsorption tubes, PIDs and MIRAN infrared absorption spectrophotometers. The effects of moisture interference on the MIRAN readings have also served to greatly inflate the detection limit of that system.
Qualitative assessment of the external and in-visor concentrations implies that a good level of protection is being achieved, with no evidence of any consistent relationship between in-visor and external concentrations. This is contradicted by ‘face-value’ quantitative data analysis, which infers that an extremely low APF (in the region of 12) applies. However, when the detection limit is properly included in the calculations (being used to reject inherently unreliable and biased data), the likely APF increases to 40. This is the same as the currently accepted APF for these devices but considerably lower than permitted during certification tests. It is worth remembering here that the values of PF derived in this work should be regarded as worst case minima. The substitution of the detection limit value for lower in-facepiece data will artificially reduce calculated values. In comparison with the ‘required’ PF levels in Table 1, these devices appear to provide adequate levels of protection for the great majority of the time they are used. As a worst case, the data suggest that a PF of less than the required 50 applies for only 9% of the time.
Calculated protection factors where the measured internal concentration is comparable with the detection limit, and the external concentration is also relatively low, are meaningless, will always be inherently biased low, and should be excluded from calculations of WPF. For reliable measurement, the in-facepiece concentration should be significantly above the detection limit, and the externally measured concentration should exceed a value that can be calculated using the procedure of Vaughan and Rajan (2005). This constraint renders considerable quantities of our data invalid, but what remains again infers that the APF for this class of RPE is not <40. This also highlights the importance of the environment in which measurements are made; the most reliable data will come from high concentration challenges. Many of the environments used in this study were inadequate in this respect.
All of this emphasizes the fact that in many workplace situations it is difficult to reliably ascertain a WPF, and also underlines the possibility that a measured WPF, taken at face value, is not necessarily the best indicator of the performance of a device in the workplace.
|
OBSERVATIONS OF SPRAYING PRACTICE |
|---|
While the main intention of this study was, if possible, to determine WPFs for air-fed visors, observations made during the study are also relevant to the subject of sensitization. A number of factors that might add to, or indeed be the major factor in, sensitization taking place were identified.
The entire process of preparing a vehicle/component for spraying needs to be taken as a whole rather than concentrating on just the actual spraying. During spraying, a worker is generally well protected by wearing coveralls, gloves and RPE incorporating eye and face protection. Provided these items are suitable, correctly worn, well maintained and changed frequently, the exposure to contaminants during this phase of the work will be well below the recognized exposure limits, and probably far less than when carrying out some other part of the process, such as preparing the vehicle or component, or mixing the paint.
Another area for consideration is that of donning and removing the protective equipment, including the RPE. If the equipment has been used previously, it may well be contaminated and must be handled appropriately. Procedures need to be in place to ensure that contaminated equipment is not mixed with clean items. This particularly applies to the visors themselves. Observations showed that on a number of occasions, the visors were left in the spray booth while the worker took a break. Sometimes the visors were lifted briefly during use to look at the workpiece; this was done with a contaminated gloved hand that would at some point have to make contact with an inner surface of the visor. All of these things could contribute to the problem of sensitization and it may well prove to be the case that they are a more significant source of exposure than the actual spraying process.
|
INWARD LEAKAGE AS A FUNCTION OF AIR SUPPLY |
|---|
This was most easily studied by measuring the inward leakage at a range of differing air supplies in the laboratory. Three visors of similar construction but with different types of face seal (two shaped foam and one elasticated non-woven fabric type; see Figs 7
–9) were selected and fitted with a ball sample probe for taking a sample from the wearer's breathing zone. Three test subjects were then asked to wear the visors while exercising on a treadmill in a test chamber. The atmosphere in the test chamber was controlled to provide a stable level of SF6 tracer gas. By comparing the level of tracer inside the visor with that in the chamber, a measure of the inward leakage could be obtained.
|
|
|
While the subject was exercising, the air supply was set at three different levels to determine the effect this had on performance. A standardized set of exercises was performed at each setting of the air supply to obtain comparable results. These consisted of walking, looking from side to side, looking up and down, reciting a passage out loud and finally another period of walking.
The air supplies used were 160 l min–1 (test subject 3, 150 l min–1), 120 l min–1 and 100 l min–1. These flows are all substantially lower than the nominal manufacturer's recommended flow for these devices of 180 l min–1, and even further below the flows observed in the field studies.
The effects are clearly evident in Fig. 10, which shows the results for the different combinations of device and test subject, during the talking phase of the exercise. This phase is generally the worst case in that it produces the highest leakage figures due to the abnormal breathing pattern, or movements of the face and/or seal during this exercise. During talking, breathing tends to change from a fairly regular sinusoidal form to an irregular triangular pattern resulting in high instantaneous peak demands on the air supply.
|
It can be clearly seen that regardless of the actual level of leakage, the inward leakage increases as the air supply flow decreases. There does not appear to be any particular cut-off point where the level of protection suddenly falls, rather a gradual increase in leakage as the air supply decreases.
Of the three designs tested, Visor A appears to be less critically affected by reduced flow than the other two. In common with Visor B it has a foam face seal, while Visor C has an elasticated non-woven fabric seal. It appears that the seal material alone is not the determining factor here.
EN 1835 and EN 270 devices must not exceed the maximum permitted level of inward leakage at the manufacturer's minimum stated flow. For the EN 1835 LDH3 and EN 270 devices considered here, the maximum inward leakage is 0.5%. (For LDH1 and LDH2 devices, the maximum inward leakage is 10 and 2%, respectively.) The degradation in performance that we have observed for EN 270 and LDH3 visors is likely to be similar in form for LDH1 and LDH2 devices.
Depending on the design of the individual device, moderate reductions in air flow below the specified minimum value (180 l min–1 in this case) may degrade the visor performance to the extent that the device would fail a standard inward leakage test. This is a clear indication of the importance of setting and maintaining the air supply to at least the manufacturer's minimum design condition (either specified as a flow rate or a supply pressure) and checking it as a matter of routine. All such RPE must be supplied with a means of checking that the air supply is adequate, and this should be routinely done immediately prior to every use. Many also incorporate in-use warning systems that alert the user to an inadequate supply.
These measurements were taken in moderately still air in a test chamber. It is known that strong external air currents, such as may occur close to ventilation equipment, can also detrimentally affect the performance of loose-fitting RPE by increasing inward leakage (Bolsover and Hughes, 1998). Such effects are to some extent taken into consideration, when laboratory testing for certification (by exposing the device to 2 m s–1 windspeeds during the inward leakage test), but may further degrade the protection provided by this type of equipment if air supply rates are lower than specified.
|
CONCLUSIONS |
|---|
The environments included in this study did not exhibit challenge concentrations high enough to confidently measure the WPF of this class of device. When inherently biased data are eliminated and detection limits are taken into consideration, the field data suggest that the APF for this class of air-fed visor (with protective performance as defined in EN 270 or EN 1835 LDH3) is not <40. Available evidence suggests that the true APF for this class of equipment, when operated correctly, may in fact be higher, but no value can be confidently assigned.
It is clear that when well maintained and used in accordance with the manufacturer's instructions, air-fed visors are capable of providing a good level of respiratory protection. This has been shown previously in laboratory tests and is borne out by the results of this study.
The relatively low external challenge concentration levels at some sites also show that a well-ventilated spray booth can significantly reduce the exposure to isocyanates and solvents. Engineering the problem out should always be a higher priority than providing PPE when dealing with any hazard, and it is clearly possible to reduce the hazard in the case of spray painting by means of good ventilation.
The level of protection given by air-fed visors is strongly dependent on the air flow supplied to the device. The experiments reported here show that the protection falls as the air supply falls. This is a gradual process and does not suddenly occur at any particular air supply flow. The inference here is that the level of protection we have seen in our field trials may not be achieved if the same devices were operated at flow rates below those we observed.
Observations made during the field tests indicate that there may be other aspects of the spraying process that need to be taken into consideration when looking for sources of sensitization. General preparation of the parts to be sprayed, preparation/mixing of the paints, donning and removing protective equipment and dealing with contaminated equipment all need to be studied further.
Received June 3, 2005; in final form September 12, 2005
|
REFERENCES |
|---|
Bolsover JA. (1996) Effectiveness of air fed visors for paint spraying. HSE Internal Report IR/L/PE/96/6.
Bolsover JA, Hughes AS. (1998) RPE: Effects of wind speed and direction on the effectiveness of two powered helmets. HSL Internal Report IR/L/PE/98/04.
BSI. (1997) BS 4275:1997 Guide to implementing an effective respiratory protective device programme. London: BSI.
CEN. (1994) EN 270:1994 Respiratory protective devices—compressed airline breathing apparatus incorporating a hood—requirements, testing, marking. Brussels: CEN.
CEN. (1999) EN 1835:1999 Respiratory protective devices—light duty construction compressed airline breathing apparatus incorporating a helmet or a hood—requirements, testing, marking. Brussels: CEN.
Cowie H, Hughson G, Creely K et al. (2005) An occupational hygiene assessment of the use and control of isocyanates in the UK. HSE Research Report 311.
HSE. (1999a) Isocyanates: Health hazards and precautionary measures. Guidance note EH 16. HSE Books, Sudbury, Suffolk, UK.
HSE. (1999b) Methods for the Determination of Hazardous Substances (MDHS) 25/3, organic isocyanates in air: laboratory method using sampling either onto 1-(2 ethoxyphenyl) piperazine coated glass fibre filters followed by solvent desorption or into impingers and analysis using high performance liquid chromatography. HSE Books.
HSE. (2005a) HSG53—respiratory protective equipment at work—a practical guide. HSE Books, Sudbury, Suffolk, UK.
HSE. (2005b) EH40/2005—Workplace exposure limits. HSE Books, Sudbury, Suffolk, UK.
Royal Society of Chemistry. (1987) Analytical Methods Committee. Analyst; 112: 199–204.[CrossRef]
Vaughan NP, Rajan-Sithamparanadarajah B. (2005) Meaningful workplace protection factor measurement: experimental protocols and data treatment. Ann Occup Hyg; advanced access published May 6, 2005.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||