Predicting exposure based on fundamental variables that are easily accessible to the occupational hygienist requires the development of a model. A physical model is chosen for this purpose. It comprises two compartments, or working zones: a zone close to the source (Near Field) and a zone far from the pollutant source (Far Field). In each of these two zones, the pollutant is assumed to be well mixed (homogeneous concentrations).

The following parameters are necessary for the application of the model:

VME Valeur Moyenne d’Exposition [mg/m3] – Reference exposure concentration

G total emission [mg/min], without considering the exitence of local exhaust ventilation (LEV); mean emission over the working period in question; variable emission can also be represented by graphic input

T pollutant emission time [min]

Q general workplace air supply [/h]; if there is no mechanical ventilation, the default entry is 1 air exchange/hour

rLEV efficiency of local exhaust ventilation over the source [%],

• 0% if there is no LEV
• 30% if LEV is poor (capture velocity < 0.25 m/s)
• 80% if LEV is mediocre (capture velocity 0.25-0.5 m/s fora suction hood)
• 90% if LEV is good (capture velocity > 0.5 m/s for a suction hood)
• 98% if LEV is excellent (e.g. capture velocity> 0.5 m/s)
• β exchange rate between NF and FF compartments due to air currents
• low (3-6 m3/min)
• moderate (6-11 m3/min)
• high (11-22 m3/min)

d distance from source operator or radius of the NF zone [m]

V workplace volume (m3)

A schematic diagram of the model is presented below:

Among the required parameters, rLEV but especially G are difficult to estimate. Estimation of the latter will obviously differ, depending on whether the source is a gas, a volatile product, dust or fumes. Methods ought to be established for estimating G classes, depending on what is observed in the field and the processes involved. This parameter is probably much easier to estimate than exposure directly. The people in charge of the processes probably have useful knowledge in this context. At present there is no general tool available for estimating G. However, it is one of subjects in which IST is interested.

The model equations are relatively simple and pre-programmed below so that a direct calculation can be done. All the above-mentioned parameters must be estimated and input into the model. With regard to emission, the model allows for variable emissions over time to be programmed, so that intermittent activities can be simulated.

The results are presented graphically in the form of the ratio to VME. Thus the concentrations in the two compartments over time are plotted. Numerical results are also given: maximum concentrations, steady-state concentrations, mean concentrations over a predefined time period.

The model can be downloaded free of charge. The data should be entered in the right-hand section (yellow) and the results will appear in the left-hand section (blue). If choosing variable emission, in order to input variations in the emission profile, select the points then click on the point to be changed. The results can be visualised once REFRESH has been pressed.

Examples of model application

A few examples are given here to illustrate the possibilities of this model.

Example 1

A worker is painting in a 300m3 workplace without mechanical ventilation and without LEV. The solvent is white spirit with a VME of 525 mg/m3. The estimated emission of white spirit is 500 g/h. The model predicts exposure 3.2 times VME for the FF zone and 16.8 times for the NF zone.

If the workplace is equipped with good LEV (rLEV 90%), the results obtained for the same situation are 0.3 and 1.7 times VME respectively.

Example 2

A welder is working in a large 10,000m3 workshop without ventilation. The metal is soft steel, for which the applicable VME for iron oxide is 6 mg/m3 of fine dust. Fume formation is 1 mg/s (10 g/h). In these conditions the model predicts 0.06 times VME for the fume concentration in the workshop and 8.6 times VME for the welder himself.

If the welding workplace is equipped with poor LEV (rLEV 30%), local ventilation of 1500m3/h which is poorly positioned as usual, this gives 0.04 and 6.0 times VME respectively, which is a 30% reduction in exposure. With well positioned LEV (rLEV 90%) the result is 0.01 and 0.9 times VME.

Example 3

An aerosol of a product containing 50% isopropyl alcohol (VME 490 mg/m3) and 5% formaldehyde (VME 0.6 mg/m3) is sprayed to disinfect a biotechnology workplace. The workplace volume is 500 m3. The work takes 1 hour and uses 300 ml solution. In these conditions, the model predicts isopropyl alcohol exposure of 0.6 (FF) and 4.4 (NF) times VME, and for formaldehyde 50 (FF) and 360 (NF) times VME! The presence of mechanical ventilation (GEV) of 10,000 m3/h (20 exchanges/hour) during the disinfecting work (simultaneous disinfection of the ventilation unit ?) would lower the formaldehyde concentrations to 2.5 and 316 times VME respectively.

Comments

The results obtained are obviously approximate. A decision cannot be taken if figures close to VME are obtained, and in this instance it is probably necessary to implement measures or to reduce exposure somehow. However, if the model outcome is markedly lower or markedly higher than the decision-making criterion (standard), a decision can then be made without having to implement any exposure measures.

Estimating emission (E) is probably the most difficult aspect. Estimation methods as a function of procedures will have to be developed. It is foreseeable that this will be possible in the near future. IST is interested in this aspect. In this context it would be useful to share information and interests, successes and failures with such a method.

The model was developped by Raffaella Bruzzi, David Vernez and Pierre-Olivier Droz.