Metodiche standardizzate misurazione degli agenti D.Lgs.81/08

La EN 689 è una delle metodiche standardizzate per la misurazione degli agenti elencate nell’allegato ALLEGATO XLI del D.Lgs.81/08-Titolo IX art.225 c.2.
....
Art. 225. Misure specifiche di protezione e di prevenzione
...
2. Salvo che possa dimostrare con altri mezzi il conseguimento di un adeguato livello di prevenzione e di protezione, il datore di lavoro, periodicamente ed ogni qualvolta sono modificate le condizioni che possono influire sull'esposizione, provvede ad effettuare la misurazione degli agenti che possono presentare un rischio per la salute, con metodiche standardizzate di cui è riportato un elenco meramente indicativo nell'allegato XLI o in loro assenza, con metodiche appropriate e con particolare riferimento ai valori limite di esposizione professionale e per periodi rappresentativi dell'esposizione in termini spazio temporali.
...

Allegato XLI

UNI EN 481:1994	Atmosfera nell’ambiente di lavoro. Definizione delle frazioni granulometriche per la misurazione delle particelle aerodisperse.
UNI EN 482:1998	Atmosfera nell’ambiente di lavoro. Requisiti generali per le prestazioni dei procedimenti di misurazione degli agenti chimici
UNI EN 689:1997 UNI EN 689:2019	Atmosfera nell’ambiente di lavoro. Guida alla valutazione dell’esposizione per inalazione a composti chimici ai fini del confronto con i valori limite e strategia di misurazione.
UNI EN 838: 1998	Atmosfera nell’ambiente di lavoro. Campionatori diffusivi per la determinazione di gas e vapori. Requisiti e metodi di prova.
UNI EN 1076:1999	Atmosfera nell’ambiente di lavoro. Tubi di assorbimento mediante pompaggio per la determinazione di gas e vapori. Requisiti e metodi di prova.
UNI EN 1231: 1999	Atmosfera nell’ambiente di lavoro. Sistemi di misurazione di breve durata con tubo di rivelazione. Requisiti e metodi di prova.
UNI EN 1232: 1999	Atmosfera nell’ambiente di lavoro. Pompe per il campionamento personale di agenti chimici. Requisiti e metodi di prova.
UNI EN 1540:2001	Atmosfera nell’ambiente di lavoro. Terminologia.
UNI EN 12919:2001	Atmosfera nell’ambiente di lavoro. Pompe per il campionamento di agenti chimici con portate maggiori di 5 l/min. Requisiti e metodi di prova.

...

Annex A
(informative)

Assessment of exposure

A.1 General

This annex gives guidance on when measurements are advisable or if other approaches of assessment maybe used.

In practice there are different ways of assessing workers' exposure to chemical agents. This annex is intended as a support to the identification and assessment of inhalation exposure on workplaces.

In assessment, it may be necessary to consider together all activities in a working area, which is the spatially or organisationally delimited part of a site in which activities involving chemical agents are carried out by one or more workers and can be brought together in a risk assessment. It may encompass one or more workplaces or working processes. The main factors to be used to choose the appropriate method for assessing exposure are extensive, including the type and duration of exposure by inhalation at workplaces.

In addition to exposure measurements, the methods used to assess exposure generally include: reasonable

- worst-case measurements (under control);
- measurement of technical parameters, e.g. air velocity, air change rate);
- calculation of exposure (using appropriate models or algorithms);
- comparison with other workplaces, in the same enterprise or in other enterprises;
- control banding approaches, and
- good practice guidance for defined branches or tasks.

Reasonable worst-case measurements are performed to estimate the maximum exposures likely to occur with the controls in place.

When applying one of these methods and there is still uncertainty concerning the identification result and its evaluation, additional workplace measurements are recommended. In A.2 to A.9 different workplace situations are described together with information about possible methods to assess exposure and to derive suitable protective measures. Table A.1 gives an overview of workplace situations addressed in this Annex and the recommended way of exposure assessment.

...

Annex E
(informative)

Check of exposure measurements distribution, and identification of exceptional exposure within the SEG

E.1 General

Annex E gives graphical and statistical methods for the validation of the results and the SEGs, as required in 5.4, for the identification of individuals or groups whose exposure is not properly represented by the SEG, and for the decision to use the normal or the lognormal distribution in the compliances test (see Annex F). The graphical method supports the identification of a LOQ (see Annex H) and is applied using at least six exposure measurements in the SEG.

The measurement results may be examined to see if a normal or lognormal distribution fits the results better. In many circumstances, exposure variability is influenced by the multiplicative interactions of workplace factors (see 5.1.3, Annex A and Annex D). Measurements would be expected to approximate to a lognormal distribution, and this has often been observed. However, a normal distribution may appear with high control of workplace factors in constant conditions (see A.2) or if the influence of workplace factors is outshined by the random variation in the measuring procedure (see 5.2.2⁾.

The shape of the measurement results may be examined by a statistical test, such as the Shapiro and Wilk test, to see whether their distribution is closer to lognormal than normal distributions. For example, the results may first be tested as to whether their distribution fits a normal distribution. This will give a probability P1 of the hypothesis that the normal distribution fits the results. The test may then be repeated using the logarithms of the results, giving a probability p₂. If p₁ > p₂ it may be concluded that a normal distribution fits the results better than a lognormal distribution. The result of comparing p₁ and p₂ determines the distribution to be used for the compliance test (see Annex F).

The appraiser should take into account that the exposure distribution may deviate in the tails from the (log)normal model (see E.2.3) and that the normal distribution does not exclude negative concentrations. Truncation at higher concentrations (vapour saturation, explosion limit or maximum dustiness) can occur, like censoring in the lower tail due to the increasing number of results below the LOQ. Therefore, the check on the distribution and on pooling results are important before LOQ adjustment (see Annex H) and applying the models in the compliance test (see 5.5.3 and Annex F).

E.2 describes a graphical method for a lognormal distribution. To decide if a normal distribution describes the results better, the procedure should be repeated using normal probability paper.

E.3 refers to software that automates the graphical method of E.2, test goodness-of fit (see E.2.2) and test the consistency of the SEG (see E.2.4.3).

...

E.2.4.2 SEG consisting if two groups

Figure E.2 illustrates a log-probability plot which appears to be two intersecting straight lines rather than a single one. These measurements were divided into two different groups (with the central point in each, as this result could be in either group), and the two sets plotted separately on the same chart in Figure E.3.

Figures E.2 and E.3 suggest that the SEG includes two different groups with different exposure patterns, and the SEG should be divided in two, and more exposure measurements made so that the number of measurements in each SEG reaches the minimum required by the statistical test chosen in 5.5.3. If these extra measurements are added to the original plot, modifying Figure E.2, they may clarify whether one or two distributions are present, or if there is some other explanation.

Key

X exposure x_k (mg/m³)
Y probability P_k x 100

Figure E.2 - Example of exposure measurements which appear to come from two different populations within the SEG, each with a lognormal distribution of exposure

Key

X exposure x_k (mg/m³)
Y probability P_k x 100

Figure E.3 - Exposure measurements replotted as two different sets (with the central point included in both), confirming that these appear to be two different lognormally-distributed populations