TOX, TIX and RAS score: the methodology behind DOHSBase Compare

This article assumes you are familiar with the basics: exposure assessment, occupational exposure limits and GHS/CLP hazard classification via H-statements. For the basics, start with Limit values for hazardous substances at the workplace. This article answers the question: how does DOHSBase Compare rank substances against each other on health risk, and what is the maths behind it?

Summary: DOHSBase Compare ranks substances by health risk using three connected indices, originally published by Geert Wieling and Theo Scheffers in the NVvA Newsletter of April 2006. The TOX index is a uniform classification on the basis of H-statements, with a choice between COSHH Essentials, TRGS 440 (default), ECETOC and SOMS. The TIX index (“Threshold exceedance Index”) measures the “ability to become airborne”: the logarithm of the ratio between the saturation vapour concentration (C_max) and the chosen occupational exposure limit (MAC), normalised to a 0–4 scale. The RAS score (Risk Assessment Score) is the product TIX × TOX on a 0–16 scale and produces a person-independent ranking. The methodology is intended as a desk-screening tool to make the combination of toxicity and exposure potential reproducibly visible — not as a replacement for workplace measurements or substance-specific risk assessment.

What DOHSBase Compare is solving

An occupational hygienist working in a process where multiple chemical substances are present at the same time regularly faces the question “which of these substances has the lowest risk from a health perspective, and which should we tackle first?”. The answer requires weighing factors that are each individually relevant: toxicity, the quantities used, contact duration, physical state, vapour pressure, the relevant limit values and the presence of specific endpoints such as carcinogenicity or reproductive toxicity.

Before 2005, no standardised way existed to make that weighing reproducible. Two occupational hygienists could rank the same list of substances differently, with the consequence that the outcome of the risk assessment depended heavily on who carried it out. Wieling and Scheffers labelled this in 2006 as “occupational hygiene alchemy” — positively framed as professional judgement, negatively framed as a black box. The lack of reproducibility made the outcome vulnerable to challenge by clients and regulators.

DOHSBase Compare was designed to objectify the ranking. The methodology hooks into two independently known quantities:

The intrinsic hazard properties of the substance — toxicity, classification systems and limit values.
The volatility and therefore the maximum airborne concentration the substance can reach.

Combining the two produces a standardised ranking that gives every user the same outcome for the same input.

The three indices

TOX index — uniform toxicity classes

The TOX index is not a new classification scheme but a uniform name for the existing hazard-banding systems. Several organisations developed classification schemes during the 1990s and 2000s that group substances into 3 to 5 classes for health damage on the basis of their H-statements (originally: R-phrases):

COSHH Essentials (UK Health & Safety Executive) — 5 classes (A through E)
TRGS 440 (German Technische Regeln für Gefahrstoffe) — 4 classes (1 low through 4 very high); the default in DOHSBase because the classification is based on a more recent adaptation of Annex I
ECETOC (European Centre for Ecotoxicology and Toxicology of Chemicals)
SOMS (Strategie Omgaan Met Stoffen, Dutch government)
ILO (International Labour Organization)

All schemes group substances into 3 to 5 classes on the basis of H- or R-phrases for health damage. The DOHSBase Compare user can select the scheme to work with; the TOX index then picks the appropriate scheme and translates the H-statements of the substance into the corresponding class number. The mapping below shows the original 2006 publication (R-phrase numbering at the time; the principle is identical for current H-statements):

Class	COSHH	TRGS	ECETOC	SOMS
4	Carc123, Mut123, ASen	Carc12, Mut12, T+, T48	ASen, T+, T48	Carc12, Mut123, Rep60,61, T+, T48
3	T+, T48, DSen, Rep	T+, T48, Sen, Mut3, Carc3, Rep60,61	Rep62+, DSen, T, Xn48, C, Xi41	Rep62+, T, C, Sen, Carc3, Narc67
2	T, Xn48, C, Xi37,41	Xn, Xi41, Rep62+	—	Xn, Xi41, Dry66
1	Xn	Xi36-38, Other	Xn, Xi36-38, Other	Xi36-38
0	Xi36, 38, Other	—	—	—

Source: Wieling & Scheffers, NVvA Newsletter 2006-01, table 1.

TIX index — ability to become airborne

The TIX index (Threshold exceedance Index) expresses how likely it is that a substance will exceed its exposure limit by simply evaporating. The starting question is: if the substance could evaporate freely up to its saturation vapour pressure, how far above the limit value would that maximum air concentration be?

The calculation has two steps.

Step 1: calculate the maximum air concentration C_max. From the vapour pressure at room temperature:

C_max = M / 22.4 × 273 / T × P_T / 760 × 10⁶ [mg/m³]

where:

M = molecular mass of the substance
M / 22.4 = conversion factor from ppm to mg/m³ at 273 K
T = temperature in Kelvin
P_T = vapour pressure in mm Hg
P_T / 760 = relative vapour pressure; for an ideal gas equal to the molar fraction of the substance in the surrounding air, and therefore equal to the concentration in ppm or vol %

The formula simplifies to:

C_max = 16036 × M × P_T / T [mg/m³]

The calculation does not account for additional exposure from misting, agitation of the liquid surface or dustiness of solids. For liquids, the formula is a reasonable upper bound. For solids, TIX has limited meaning: dustiness is the dominant factor and is not modelled in DOHSBase. There is also no generally accepted measure for dustiness, let alone a database parameter for it.

Step 2: scale the ratio to a 0–4 index. The ratio C_max / MAC, where MAC is the chosen exposure limit, produces a value below 10⁷ for most substances. To bring that ratio into a workable scale, the logarithm is taken and scaled:

TIX = ¹⁰log(C_max / MAC) × 4 / 7

Values above 4 are capped at 4. TIX is set to 0 once C_max / MAC drops below 1 — a substance whose vapour cannot reach the limit value receives no volatility points.

The C_max / MAC ratio itself is identical to the historical RIR index (Risk Index Ratio) proposed by Mutgeert in 1979 (De Veiligheid no. 55, pp. 355–361). DOHSBase adds the logarithmic scaling and the 0–4 cap to that historical ratio so that TIX can be multiplied directly with TOX.

RAS score — Risk Assessment Score

The actual ranking is produced by multiplying TIX and TOX:

RAS = TIX × TOX

RAS values lie between 0 and 16. Observations on the DOHSBase database have shown that all three indices have good discriminatory power — the distribution of scores spreads sufficiently across the scale to give a meaningful ranking in typical occupational hygiene practice.

DOHSBase Compare presents the result as a table sorted by descending RAS score. Substances for which the calculation is not possible — for example because the limit value is missing — appear in a separate section. For those substances DOHSBase supplies a kick-off value when at least one H-statement for health damage is present, so that the RAS calculation can still be performed against a conservative reference.

A worked example: phenol, methanol and toluene

A small chemical plant prepares several hundred tonnes of a product based on a mixture of roughly equal quantities of phenol, methanol and toluene. The work takes place at normal pressure and temperature. Mists and skin contact are excluded by the prevailing control regime. Vapour emissions cannot be ruled out during drum filling, a 15-minute task carried out every two hours. The plant manager asks: which substance should we tackle first?

A professional-judgement approach offers several plausible answers:

Toluene, because it is the only one with a reproductive-toxicity classification?
Phenol, because it is the only one with an immediate effect (corrosive), so the impact of controls would be visible quickly?
Methanol and toluene, because those are the highly flammable ones?

Three defensible answers, three different work priorities — exactly the “alchemy” the Compare methodology is meant to prevent.

Input data for the three substances:

	Phenol	Methanol	Toluene
CAS#	108-95-2	67-56-1	108-88-3
Physical state	Solid	Liquid	Liquid
C_max (mg/m³)	401	21 × 10⁴	14 × 10⁴
EU classification & symbol	T (toxic), C (corrosive), Mut.Cat.3	F (highly flammable), T (toxic)	F (highly flammable), Xn (harmful), Repr.Cat.3
R-phrases	23/24/25; 34; 48/20/21/22; 68	11; 24/2325; 39/23/24/25	11; 38; 48/20; 63; 65; 67
15-min limit (mg/m³)	16 (SCOEL)	520 (MAC)	9 (Sweden); national MAC withdrawn
Skin notation	H	H	H

Applied to the TRGS scheme, using the 15-minute limit (which fits the 15-minute drum filling task in the example), the RAS ranking comes out as:

	TOX	TIX	RAS	Rank
Phenol	3 (R23/24/25, 68, 48/20/21/22)	0.8	2.4	3
Methanol	3 (R23/24/25)	1.5	4.5	1
Toluene	2 (R63)	1.5	3.0	2

Source: Wieling & Scheffers, NVvA Newsletter 2006-01, example + table 2.

The ranking shows that methanol has the highest RAS score and therefore — from a health perspective, given the control regime in this scenario — deserves first priority. Toluene comes second, phenol third. The plausible-looking alternative of phenol-first (since the corrosive effect of controls would be quickly visible) does not hold up after the calculation: phenol does have a higher TOX, but its solid state and low vapour pressure make TIX so low that the resulting risk in this exposure scenario is smaller than methanol’s.

The same ranking holds — with different absolute numbers — when the 8-hour limits are used instead. The result is robust to the choice of averaging window.

Limitations of the RAS ranking

The RAS score is explicitly a screening index, not an absolute risk value. Three caveats stated explicitly in the original 2006 publication, and still fully applicable in 2026:

RAS has no meaning as an absolute value. A RAS score of 6 does not mean “twice as risky as 3”; it means the substance occupies a higher position in the ranking. The methodology is intended for comparing alternatives and creating a first triage, not for quantifying the probability of limit exceedance.
Physical state and dustiness are modelled incompletely. C_max describes only the vapour route. For powders and granular materials the actual exposure potential depends strongly on dustiness, a property for which no generally accepted measure exists and which DOHSBase does not model. For these substances RAS is a lower-bound estimate.
Actual risk must be confirmed at the workplace. A Compare ranking gives a desk-first impression; the actual assessment requires air measurements and an EN 689 conformity test or an equivalent analysis.

The advantage of RAS, and the principal reason to prefer it over loose professional judgement, is that the ranking is person-independent. Same input, same outcome, regardless of who carries out the assessment. For clients and regulators that is a substantially stronger starting point than a non-reproducible black-or-white judgement formulated by a single expert.

Source publication and historical lineage

The TOX/TIX/RAS methodology was first published by Geert Wieling and Theo Scheffers in the NVvA Newsletter of April 2006 (NR1 2006, pp. 7–11), under the title “Ranking chemical substances with DOHSBase”. The full article is available as a local copy.

The TIX formula has an older ancestor: the RIR index (Risk Index Ratio) proposed by B.J. Mutgeert in De Veiligheid no. 55, 1979, pp. 355–361. Mutgeert introduced the C_max / MAC ratio as a dimensionless measure of the possibility that a vapour exceeds its limit value. Wieling and Scheffers retained this core, added the logarithmic scaling (to bring the ratio into a workable 0–4 band) and coupled it to the TOX index to produce a multipliable risk score.

Since 2009 the Compare methodology has been externally validated by the RIVM (the Dutch National Institute for Public Health and the Environment) in the peer-review report of the AWARE project (RIVM 320023001, 2009). The report uses the DOHSBase Compare methodology explicitly as a reference framework for describing the TIX index — one of the earliest documented external evaluations of the DOHSBase calculation core. The Arbokennisnet knowledge dossiers and the sectoral Arbocatalogus Carrosserie Schadeherstel (2021, authored by Geert Wieling) use the RAS score as a standardised ranking method in concrete work processes. See DOHSBase in the literature for the full overview.