Determinants of Exposure to Respirable Quartz Dust in the Construction Industry

ABSTRACT

Because most masonry building materials contain quartz and because these materials are subjected to a variety of treatments during the building process, quartz is encountered everywhere in building operations. The level of exposure to respirable quartz has been measured for some highly exposed groups of employees. At 30 construction sites personal air sampling (PAS) measurements of respirable dust and quartz have been performed and 171 samples have been taken. Both respirable dust and quartz levels were high. Respirable quartz exposures of more than ten times the Dutch limit value of 0.075 mg/m3 TWA were common, but exposures up to 200 times the Dutch limit value were also found. The measurements were task oriented.

By statistical analysis the contribution of the different determinants to the total exposure has been identified. With this approach, directions for an effective control measures programme can be given.

MATERIALS AND METHODS

Selection of construction sites

Construction sites where the research was carried out were acquired in two ways, by contacting companies listed in the telephone Yellow Pages and via the health and safety co-ordinators of several big construction companies. The selection was aimed not only at the occupations, but also at a variation in the conditions in which the jobs were carried out, e.g. variation in type of building material and level of control measures.

Measurements: task based PAS and source oriented measurements

A total of 30 construction sites were selected. Some sites were studied on more than one day. At each site employees were asked to co-operate by wearing personal air sampling equipment for some hours. The large majority of them were willing to participate in the study. Only two of them refused because they considered it too much of a fuss. Measurements were task-oriented. Because of the nature of the jobs (at different sites, ready-go-home system), the sampling time was mostly between 3 and 4 h.

In addition to the personal air-sampling, source-oriented ambient measurements were taken. A direct reading dust monitor was used both near the working environment of the worker and at the background to quantify the (relative) contributions of the different sources mentioned above.

Description of construction sites

A questionnaire and a checklist developed for this study were used to enable a qualitative description of the different workplaces where the measurements were conducted. In the questionnaire the workers at the construction sites are asked about their normal working habits, working hours, use of personal protection equipment, use of other control measures and possible improvement of the working conditions at the construction sites. The checklist was used to systematically report on factors at the workplace which might have an influence on the total exposure of the workers. Information was gathered on: type (and if available make) of building material, type and make of equipment, stage of the building process (layout, with or without glass in the windows), presence of other workers, use of LEV, good house keeping. Observations were made according to a set protocol. Reports of the weather conditions during the days of the measurements came from the nearest weather station.

Respiratory protection equipment

Respiratory protection equipment was used at 30% of the construction sites. The protection used consisted mainly of very simple disposable paper masks. In only one case was the worker wearing a filtering facepiece of P2 quality.

Sampling and analysis

Samples were collected on Millipore mixed ester filters (type RA 1.2 μm, 25 mm) using Casella cyclones as sampling heads in combination with Dupont P-2500 or Gilian Gilair pumps with a flow of 1.9 l./min. Casella cyclones collect the respirable dust fraction, which is relevant in determining exposure to respirable quartz. Filters were weighed before and after sampling with a Mettler balance (type AT 261 DeltaRange, Switzerland).

A selection of 61 of the 181 filters was sent to an external laboratory (Ascor Analyse, The Netherlands) to determine the content of respirable crystalline silica of the respirable dust. The analysis was performed by X-ray diffraction (XRD) according to NIOSH method 7500. The detection limit of this method is 10 μg/filter. The selection was done in such way that for each construction site at least two filters were analysed, representing the tasks that were performed and the building materials processed. The quartz content of these filters is considered to be representative of the filters taken under the same circumstances, and was used to calculate the quartz content of the samples taking at the same building sites working with the same building materials.

As a direct-reading aerosol monitor the MiniRAM (model PDM-3, USA) was used. This is a light scattering aerosol monitor that responds to particles in the range 0.1-10 μm. It is calibrated on Arizona Road Dust and not on 'construction' dust, so results are comparative only. The MiniRAM was connected to a data logger (Metrosonics, USA). The logged data were read into a personal computer using Metrosoft software. The results were plotted and interpreted by comparing the variation in exposure to the results of the observations made synchronously at the workplace.

All statistical analyses were performed with SAS software (SAS Institute Inc, 1994).

Determinants of exposure

One of the objectives of this study was to set priorities for control measures. In order to achieve this objective a statistical approach was used.

Generally, personal exposure is composed of the contributions of different determinants. Determinants were divided into four different categories: agents, processes or appliances, work practices and working environment and their contributions to exposure were assessed by statistical analysis. This approach is called the 'multiple source model' and is described by Buringh et al. (1992). Analyses of the statistical model were performed by analyses of variance within the framework of general linear models (GLM) (Draper and Smith, 1981), giving parameter estimates that indicate the contribution of the four determinants to the quartz exposure. Based on these results an estimate of exposure can be made of a work situation for which the four determinants are well described.

The determinants in their turn consist of variables on which information was gathered during the field studies. These variables were chosen because either information from earlier publications or our own workplace observations indicated their influence on the eventual exposure of the construction workers. On some of these variables quantitative information could be obtained; others had to be assessed by workplace observations. For determinants consisting of more variables, sum scores of the separate variables were made. In the regression analysis quartz exposure was the dependent variable and the determinants were the independent variables.

The four distinctive determinants are now described in further detail.

Agents: type of building material used. In the analysis a ranking of the building material is used according to its quartz content. Table 1 gives an overview of the building materials applied at the 30 construction sites under study, and their quartz content as known from literature with increasing percentage of quartz.

Processes and appliances: type of process applied, without or without use of local exhaust ventilation, make of equipment used; The influence of these variables together forms the determinant 'process/appliance'. Table 2 shows the different types of processes used with the three occupations.

Working environment: 30 construction sites were studied, varying from small apartments to large demolition sites. To standardise the influence of working environment, the layout of the building under construction is described by a number of descriptors: glass already present in the windows (0/1), presence of other workers causing extra exposure (0/1), weather conditions (rain (0/1), wind (0/1)); A sum score composed of dummy variables of these descriptors is used to assess the influence of the determinant 'working environment'. The sum score, i.e. the contribution from working environment, is lowest when no glass is present in windows, no other workers are present, when it is raining and windy.

Work practices: use of personal protective equipment (ppe) and the general impression of neatness when working. Information on whether or not ppe is used and scoring of 'good housekeeping' practices on a scale of 1-3 are combined to describe the determinant 'work practice'. Adverse work practices i.e. leading to higher exposures consist of 'not wearing ppe' and 'bad housekeeping'.

RESULTS

In Tables 3 and 4 the results of the respirable dust and quartz dust determinations are summarized.

The results of the measurements of respirable dust and quartz dust for the total population and those for the three separate populations are lognormally distributed. Therefore the geometric means and standard deviations are reported. The highest respirable dust and quartz levels were found in the same sample, which was taken during demolition work. Tiles were removed from the walls with jackhammers in a confined space without natural ventilation. In recess milling the highest respirable quartz sample occurred when milling in sand lime stone with a quartz content of over 40%.

Figures 4-6 illustrate examples of dust measurements as determined by means of the MiniRAM connected to the data logger. In Fig. 4 recesses were milled in a room with glass in the windows. At 15:31 a new room is entered, and concentration drops considerably.

Figure 5 is made during the disposal of wet demolition waste. Only the surface of the waste is humid; at 15:35 the concentration rises due to the fact that the underlying dry and dusty material is scooped up. Figure 6 shows the influence of the equipment used in inner walls construction. In Fig. 6(a) a special type of clipper is used to bring the elements to the correct size. In Fig. 6(b) this is done by applying a circular saw.

Table 5 shows the results of the statistical analysis aimed at determining the influence and contribution of the four determinants to total exposure. The results of this modelling are shown to be significant for all groups of construction workers. Between groups the influence of the four determinants however differs. In principle, all determinants would be expected to contribute positively to exposure. A higher quartz content of agent used, dusty working methods, and little use of LEV, in a working environment with hardly any natural ventilation and more workers present at the workplace, and a low level of housekeeping and limited use of PPE, were expected to contribute significantly to higher quartz exposure levels. Exceptions to these results will be discussed.
