Web links, documents and free of charge tools
Use the documents and tools listed below for the qualitative and
quantitative, occupational exposure risk assessment.
Not free of charge, but recognized standards are:
They are grouped
in the following Click-and-go headers:
1. Exposure assessment and compliance testing
2. Sampling and analytical methods
3. Exposure databases
4. Occupational Exposure Limit Values (OELV)
5. Compliance testing statistical tools
6. Lognormality and exposure variability
7. Compliance for mixtures
8. Exposure modeling
9. Chemical identification, physchem properties and hazard information.
For Specific Dutch tools click here (in het Nederlands).
1. Exposure assessment and compliance testing
Current days, workplace exposure assessment is described in different public
available, free of charge documents:
2. Sampling and analytical methods
689 for a reasonable price. The EU Standard (2018) to establish employers odds to be in compliance with
a limit values.
- A Strategy for Assessing and Managing Occupational Exposures. (Fourth edition
2015) of the AIHA
Free of charge sources with measurement methods are:
3. Exposure databases
For prior knowledge on substance workplace air exposures (raw and aggregated data), consult:
Public and private occupational Exposure Databases may be based on the US
guidelines on core information for workplace exposure measurements on chemical
agents, which made it possible to construct exposure distributions per SEG.
4. Occupational Exposure Limits Values (OELV)
As a license to operate and to test workplace air compliance, an employer needs
to have an OELV for all chemicals to which workers are exposed at his premises
or under his responsability. There are many (inter)national organisations that
OELVs differ in goal and quality and should therefore be chosen carefully.
For more information on the hierarchy of OELV's see Presentation
on H3##-statement based kick-off levels. Bilthoven NVT-AT 140925
Below OELV sources are presented in groups of different hierarchy.
4.1 Unbiased health based only OELV's
"Health based only" means that no political, socioeconomic and technical feasibility
factors have influenced the value (level & duration) of the OELV. Unbiased
means not established or financed by interest groups. Documents published before
1997 may suffer from limited quality control and transparancy. The web sites below
contain several hundreds of OELV documents often, however, on the same chemical
substances. The OELV's proposed for the same chemical substances may differ in
value, reflecting differences in scientific sources used, different hazard assessment
techniques (i.e. safety factors used) and differences in hazard perception. Compliance
with these limits may safeguard the employer from liability for health claims.
4.2 Health/Toxicology based
For the OELVs below the independency from interest groups and/or unbiased science is not guaranteed.
- Workplace Environmental Exposure
Levels (OARS WEEL's), originally developed by AIHA.
In 2013 transferred to the Occupational Alliance for Risk Science (OARS) which
is managed by Toxicology for Risk Assessment (TERA)
Begründungen zu Arbeitsplatzgrenzwerten stakeholders documents in German
underlying the legal limits
developed Criteria Documents including the Recommended Exposure Limits (REL). The older RELs may be based on analytical feasibility.
- REACH DNEL and DMEL. Based on product liability, suppliers needs to provide
users of their chemicals guidance on safe work (eSDS). The European
CHemical Agency (ECHA) has disseminated the REACH dossiers of twenty
thousand substances prepared by suppliers of chemicals in Europe. Of some
DNEL or DMEL are derived to safeguard the occupational and general public
use for inhalation, oral and skin exposure. The workplace air DNELs differ
from OELV's in that REACH is based on the specific use of a suppliers product
and not on the whole occupational load during the workers shift (TWA) or task(STEL).
Further DN/MEL are derived using fixed administrative assessment used to extrapolate from
route/dose/duration/species/response to human/worker inhalation or skin exposure.
Epidemiology and human findings are ignored at large in the DN/MEL. However DN/MEL can be
used as a starting point for developing a workplace OELV, if higher
hierarchy OELV's do not exist. The reference periods "long-term" and "short
term" should carefully be considered if converted to TWA or STEL/Ceiling.
The long-term inhalation, local and systemic DN/MEL's for the worker can be
found in the GESTIS
- DNEL comparible methods to develop limit values for pharmaceutical active ingridients exists from the EU Medicine Agency and is a work item within ASTM WK59975
4.3 Some regulatory OELV's in the EU
Most countries have regulatory limits. Non-compliancy may lead to a penalty.
If socio-economical feasebilty has lead to a level that exceeds the health based
level, then compliance with the regulatory limit will not safeguard the employer
from liability if exposure exceeds the health
EU Binding occupational exposure Limit Values.
Indicative Occupational Limits Values. adopted as regulatory limits in most
Germany TRGS 900.
UK EH40/2005 Workplace exposure limits. Table 1 WEL (air). Table 2: Biological monitoring guidance values (BMGVs)
valeurs limites d'exposition professionnelle (VLEP) indicative (VL), binding indicative (VRI) or binding restricted (=réglementaires contraignantes) (VRC).
list of legal limits (in Dutch)
Belgium KB Chemical agents.
Italy legislation for OELV.
Norway Limit values.
USA legal OELV's
The federal OELV's, different states like Texas and California have there own
4.4 Data-poor substances
While the sources under 4.1 through 4.3 mainly contain OELVs for the limited number
of data-rich substances, most substances do not have an OELV, due to the lack
of sufficient and relevant high quality human health & toxicology data.
However if GHS or CLP health H3##-statements are known, Kick-off
levels can be established.
H3##-statements to derive a Kick-off can be found in the ECHA database with hazard
and labeling information for all harmonized (6000) and notified (>100,000) dangerous substances in the EU.
Kick-off levels are used as a conservative starting point
for compliance testing.
A guidance for setting occupational exposure limits for data-poor substances has
been developed by ECETOC, see TR-101
5. Free of charge compliance testing statistical
tools (downloads and online)
6. Lognormality and exposure variability
- Since 2018 the online version of
BWStat developed by Tom Geens (Liantis) is available at the site of the
Belgium Society on Occupational Hygiene. It replaces the Excel tool BW_Stat_v2.1
released at the IOHA-BOHS
conference 2015 in London. BWStat performs several quantitative quality
checks, the EN
689 preliminairy (N=3->5) & the statistical test (N=6->50), and the
BOHS-NVvA between & within worker analysis.
(first version released at the AIHCe
1990 in Anaheim) is the EN689 (Windows) compliance testing tool with
See the YouTube demo
on how to enter sample size N and the descriptive statistics GM, GSD or how
to open a prepared text/CSV file with a .HYG extension , containing only the
numerical raw sample results. HYGINIST is useful if you want:
- (1) graphical and statistical Goodness-of-fit assessment for the sample
distribution (2 < N < 2000)
- (2) OELV compliance testing & exceedance probability using the 95% Upper
Tolerance Limit (UTL)
- with 70% confidence (EN
BOHS-NVvA guidance & French legal) or 95% confidence (IHSTAT,
- the maximum likelihood estimate (IHSTAT, employer's favorite)
- the unbiased estimate (useful if you want to optimize the sample
- (3) the arithmetic mean and its unbiased upper confidence limit (UCL),
useful to establish the long-term dose for chronic toxic substance based
on shift exposures (in Occupational Epidemiology) or the shift exposure
based on consecutive short-term sampling
- (4) testing location GM and dispersion GSD against priors or a second
distribution, and combining two exposure distributions
- to estimate GM and GSD within a detection range: Regression through
the log(outcome) and their corresponding Normal Order Statistics (=NOS
or Rankits) (see the demo on YouTube),
- to calculate the unbiased estimate of the descriptive statistics,
- to calculate the exact confidence that C95% is below the OELV,
- to use other tolerance percentages than the French, BOS-NVvA (2011)
and EN 689 (2018) 95% with 70% confidence
- to estimate the optimal sample size for an initial survey or periodic
measurements, based on prior knowledge of level (GM) and dispersion (GSD),
- to optimise the range of outcome from which GM and GSD are calculated
using censoring in the lognormal probability paper plot and the Shapiro
and Wilk goodness-of-fit test.
- to calculate Land's Upper Confidence Limit (UCL) for the arithmetic
mean (for long-term dose).
- to combine prior knowledge on GM and GSD with new sampling data (better
from Jérôme Lavoué of the University of Montreal. In English and French. His
Bayesian approach requires hardly existing knowledge on priors of the exposure
profile analysed. NDExpo
uses (like HYGINIST) the regression of detectables and corresponding Normal
Order Statistics/Rankits for point estimates of the non-deductible.
from INRS (Nancy,France). In French. With adjustments for non-detects
- Paul Hewett's IHDataAnalyst-Student
(IHDA-S) for <=25 samples including LoD adjustment. IHDA-S adds a deviant
approach to the general ML 95%-tile and Tuggle's non-central Student UTL95%,95%
by calculating likelihood's that the 95%-tile falls within 4 to 5 predefined
exposure rating categories. The Monte Carlo procedure calculating the likelihood's
uses a default GSD range of 1,05 thru 4, which is fine for task based Exposure
Scenarios testing against DNELs but too low for shift based multi-task compliance
testing against TWA8hr or STEL OELV in industrial or professional use. Further
the program includes a provision for Bayesian Decision Analysis which is displayed
(with an unknown how) if in tab "BDA Initial rating" the custom Prior probabilities
are changed and after clicking "Calculate All" in the green headed tab Statistics
- The EXCEL application IHStat,
developed by the AIHA Exposure Assessment Strategies Committee. In 11 languages
and for series up to 200 measurements without non-detects. IHStat uses two
archaic noncompliance statistics:
- the 'employers friendly' maximum likelihood test. This test is asymptotic
unbiased, meaning that the underestimation of non-compliance increases
if sample size decreases and GSD increases.
- the 'unrealistic' UTL95%,95%. Its tolerance limits are for sample size
6 or below and GSD>3 so wide that theý often exceed the range of natural
(vapour pressure, dustiness) limits.
- SPEED Not free
of charge. For the analysis of repeated, shift-long measurements on at least
some members of a random sample of workers from a job group
Based on observations in different countries and many situations (Oldham UK 1953,
Coenen GE 1966, Leidel US 1977) it is widely accepted that the body of a workplace
air exposure distribution is well represented by the lognormal distribution. In
the tails of the exposure distribution (<5% or >95%) the lognormal shape may not
fit as it can predict probabilities for non-existing or non-realistic concentration.
For example concentrations that exceed physicochemical limitations like maximum
dustiness for airborn solids, the atmospheric pressure for a gas or the saturation
concentration for vapours.
The sample Geometric Standard Deviation (GSD) is the descriptive statistic for the variability
(=dispersion) of the exposure data and responsible for the extrapolations in
the tails in lognormal distribution.
The GSD is:
the antilog (=EXP) of the standard deviation (s) of the logarithms of the results EXP(s)
the unbiased estimator of the population GSD [EXP(σ)] with minimum variance.
This means the most efficient estimator with an expectation equal to the population
EXP(σ). And this means in simple wording that the GSD equals EXP(σ)
on the average for every sample size.
by definition lager than one plus the analytical coefficient of variation (1+CVt)
(NIOSH 77-173 p124) .
considered to be an intrisic property for every substance/Similar Exposure
Group(SEG) combination, or its REACH equivalent the workers task/activity
Contributing Scenario (t-CS).
The numerical value of population GSD [EXP(σ)] of a substance air exposure distribution in a SEG/t-CS depends on workplace factors / operational
conditions like :
the the number of different tasks/handlings included and sequentially performed
during the observation period (task, shift and/or the working week),
in- or outdoor,
industrial or professional setting
technical and organisational risk management measures (ventilation, instruction etc),
But also on:
the way workers do their job (between worker differences),
Space-time factors like global positioning and altitude, local climat and
its variability througout the year.
Reliable EXP(σ) per SEG/t-CS may be retrieved from the institutional databases
[see 3. Exposure databases], as they are based
on core information for workplace exposure measurements on chemical agents
GSD's of well defined SEG/t-CS's are also published in peer review, scientific
journals. Two leading articles are:
and Rappaport (1993 table 6). The median (50%-tile) GSD's found for different
production factors within single workplace SEGs range between 1.7 and 3.6 .
(2008) established reasonable worst case (RWC) total variances (σ2)
in well defined, but multiple workplace SEGs using a relative small but high
quality database, corresponding with a GSD=5.4 for solids and GSD=8.2 for liquids
These GSDs may even be too low as 0.5*LOD was used for the non-detectables up
TWA8 hours PAS measurements of SEG's in modern chemical industry
using well performed sampling plans and regression statistical analysis for non-detects
show GSD's up to 14. See BOHS
Conference 2013, GSDs
in the real world and The
distribution of long-term GSD's in chemical industry SEG's (in Dutch , English
states that GSD>3 indicates "process out of control or poorly defined groups".
Low GSDs are to be expected in:
highly controlled indoor workplaces, clean rooms etc.
REACH t-CS, assessing a single combination of Operational
Conditions (OC) and Risk Management Measures (RMM) in industrial use
SEG's of workers performing only a single-task at a fixed workplace
workplaces with a relatively high constant background that camouflage all
spatial and temporal variability of workers activity
So in a clean room and other situations where one would expect
a GSD at or just above 1+CVt, higher GSDs may indicate suboptimal controls or bad
technical sampling and/or laboratory handling.
GSD<1.5 in other situations then mentioned above may indicate "poor sampling strategy"
or "poor data handling" in example by:
small sample size underestimating the GSD
short sampling campaigns with measurements only on one or a few consecutive
sloppy handling of non-detectables
2-decades analytical detection methods (like gravimetric dust and inorganic acid)
using EM in stead of PAS
The resistance against high GSDs in IHStat may also be caused by the bad performance
of the two noncompliance statistics with GSD>3 and sample size <6.
Low GSDs occured in the ancient industrial hygiene era:
when workers in process industry and assembly lines performed a single task
during a whole shift,
with high background levels camouflaging the variability of workers activity,
when sampling techniques were insensitive, and
when officers in charge forbid their industrial hygiene sergeants to report
EN 689 (2018) supports a preliminary test in 5.5.2 for sample size 3,4 and 5.
Its use is justified by the INRS document
ND 2231. The document shows that the use of the preliminay test is only
justified for GSDs below 3-4.
Industrial Hygienist are advised, before testing compliance, to compare
the GSD found with an expected value, using:
measurement series performed before
GSDs reported in literature or if possible in large exposure databases
Read across with comparable substances and workplaces
Modeling, if possible
Tools to calculate and compare GSD's, and to test the Lognormal goodness-of-fit
can be found in 5. Compliance testing statistical
7. Compliance for mixtures
There are several methods for compliance testing of mixtures in the workplace
The three methods above imply, that the appraiser samples all components in
the mixture. Another approach for vapors with comparable effects is to select
the most critical component (based on the saturation concentration and the OELV)
and to adjust its OELV to the composition of the mixture in the air. XLUNIFAC
is the method of preference.
The methods above do not include synergy or antagonism, the phenomena that the
toxicokinetics and/or -dynamics of the health effects change due to the interaction
of the components in the mixed exposures
8. Exposure modeling
(TRanslation of EXposure MOdels) integrates six commonly used occupational
exposure models: ART v.1.5, STOFFENMANAGER® v.5.1, ECETOC TRA v.3, MEASE v.1.02.01,
EMKG-EXPO-TOOL and EASE v.2.0.
AIHA Tools and Links for Exposure Assessment Strategies
Advanced REACH Tool
Estimates the flux and the quantity penetrating the skin from the air or from
a liquid/solution in contact with the skin
UPERCUT A hazard identification tool for toxic risk following dermal penetration of chemicals
EPA exposure modeling
- the Raoult and XLUNIFAC
method for liquid mixtures. Adjusting the OELV of the lead/critical component,
becoming representative for the whole vapor mixture. Even if the fugacity
of the components behaves non-ideal.
- the Critical
Component or DPD+ approach.
9. Chemical identification, physchem properties
and hazard information
Information on Chemicals with the REACH and CLP information on >100,000
substances used in the EU. Check every EU Safety Data Sheet (SDS) or the harmonized
classification information using this site on recent changes
with nearly 10,000 substances, including information on occupational chemical
loads without an EU SDS obligation, like welding fumes or wood dust
Monographs Evaluation of Carcinogenic Risk to Humans
Health Council of the Netherlands Evaluation of the carcinogenity, genotoxicity
and reprotoxicity, including non-REACH substances
Dohsbase contains substance
identity information of >260,000 chemicals, health hazard information of
110,000 and physchem information of 40,000 substances
NIOSH Pocket Guide to Chemical Hazards
Safe work Australia
Useful free of charge physchem online databases are:
eChemPortal. A public web site with information on chemical properties. ECHA is a key collaborator
in the development of the OECD physchem software and hosting
A TOXNET® (TOXicology Data NETwork) database from the US National Library of Medicine covering chemicals and drugs
NIST Chemistry WebBook.
The US National Institute of Standards and Technology (NIST)
. Information on chemical health and safety data received by EPA and EPA's assessments and regulatory actions for specific chemicals
. The US SCR Syrres subsstance properties databases
EPI Suite is not only a large database with measured physchem data but also
with physchem estimates based on the SMILES code.
Data Chemical details for 8,599 unique substances (GSIDs) and DSSTox standard
chemical fields (chemical name, CASRN, structure, etc) for EPA ToxCast chemicals
and the larger Tox21 chemical list
The CambridgeSoft chemistry and biology reference database
Useful free of charge toxicological databases are
Toxic Substances Portal
Hazardous Substances Data Bank (HSDB)
Zelfinspectie I-SZW. In vier stappen beoordelen of uw bedrijf gezond en veilig werkt met gevaarlijke stoffen volgens de Arbeidsinspectie
Grenswaarden Stoffen op de Werkplek van de Social Economische Raad (SER-GSW)
met Europese werkplek atmosfeer grenswaarden en REACH DNELs.
Veilig werken met chemische
stoffen.nl Door de sociale partners binnen de SER ontwikkelde leidraad
voor het veilig werken met chemische stoffen.
Hiermee is het gat gedicht
ontstaan na het afschaffen van de meeste wettelijke en bestuurlijke
MACs in 2007. Het bevat ook de goede praktijken die in de afgelopen decades zijn
ontwikkeld door bedrijven en branche verenigingen.
DOHSBase Database met het grootste en hierarchisch
opgebouwd bestand aan grenswaarden en meetmethoden
AI interne instructie toetsen van de meetresultaten aan luchtgrenswaarden (2007).
Gevaarlijke Stoffen Tot stand gekomen door multidisciplinaire samenwerking tussen de beroepsverenigingen van arbeids- en organisatiedeskundigen, bedrijfsartsen, arbeidshygiënisten en veiligheidskundigen
NVvA en Arbo-pagina
De website van de Belgische Arbeidshygiene vereniging BSOH ontsluit een groot aantal pratische tools
Wil je meer weten over de Arbeidshygiëne in Nederland dan zoek dan op
de webstee van de Nederlandse Vereniging
voor Arbeidshygiëne. Voor Arbo in het algemeen is er de Startpagina
Bij de beoordeling van de blootstelling is voor veel stoffen de bijdrage van
de opname via de intacte maar onbeschermde huid belangrijk. Van de homepage
van Wil ten Berge is het freeware programma SKINPERM
te downloaden dat op grond van een aantal fysisch chemische parameters de
potentiële huiddoorlatendheid van een gas, damp of een vloeistof schat.
En met aanvullende gegevens over kleding, besmet oppervlak e.d. wordt de mogelijke
opname in het lichaam geschat. Een goede aanvulling dus op NEN 689 die volledig
op de werkplekatmosfeer is gefocust.
U vindt dit programma door te surfen naar http://home.planet.nl/~wtberge/
SKINPERM schat de hoeveelheid van een chemische stof die via huid in het
lichaam wordt opgenomen indien het agens:
- Zich als damp of gas in de werkplek atmosfeer bevindt of
- Als vloeistof of opgelost in een vloeistof met de huid in contact komt.
De opname van vaste stoffen en aerolsolen is niet met SKINPERM te schatten.
Paul Schepers, wetenschappelijk onderzoeker bij de Universiteit Nijmegen, heeft
het zoeken naar toxische eigenschappen van chemische stoffen op het internet
wel zeer gebruiksvriendelijk gemaakt. Zijn afloopschema Patchwork
leid je naar de meest betrouwbare, publiek vrij toegankelijke, databestanden.