Categorisation of new chemicals and the risk matrices

Categorisation of chemicals

What chemicals will be excluded from the new chemicals obligations?

There will be no change to the types of chemicals that will be excluded from the new chemicals obligations. Under the reforms:

  • the definition of a ‘chemical’ will not change—hence articles, radioactive chemicals and mixtures (including hydrates and alloys, but not their individual components) will continue to not be subject to NICNAS obligations
  • the definition of an ‘industrial chemical’ will not change—hence chemicals used solely as an agricultural or veterinary chemical, a therapeutic good, a food or food additive, will also continue to not be subject to NICNAS obligations, and
  • the definition of a ‘new industrial chemical’ will not change—hence chemicals listed on the AICS, as well as reaction intermediates, incidentally produced chemicals and naturally occurring chemicals will not be subject to new chemicals obligations.

Can industry use international information/data to categorise chemicals in accordance with the matrix?

Yes. However, if an introducer is relying on information that is subject to intellectual property protections (for example, REACH dossiers), then the introducer would be responsible for ensuring they have any necessary permissions to use this information for this purpose.

Will there be tools to assist with the categorisation of my chemical?

Yes. Stakeholders noted the importance of having access to simple tools to guide people intending to introduce new chemicals using the risk matrices and the relevant information requirements. NICNAS will:

  • develop simple online tools to guide people in the categorisation of chemicals
  • provide a service to those intending to introduce chemicals to assist them with categorisation, if required
  • post guidance material and tools on the NICNAS website, and
  • hold regular industry training workshops.

Why do I need to fully characterise the hazard of my chemical in order to categorise it?

Under the reforms, it will not be necessary for an introducer to evaluate every hazard of a chemical in order to categorise the chemical.

Introducers will only need to characterise the hazard of the chemical to the extent necessary to establish the category as either Exempted, Reported or Assessed.

The level of hazard characterisation required, and therefore the information requirements to establish the category for the chemical, will be commensurate with the predicted exposure.

For example, for Exposure Band 2 in the human health risk matrix, it would only be necessary to evaluate whether the chemical is known to have any of the hazards in Hazard Band D (which would determine whether the chemical will be categorised as Assessed or Reported). In order to establish whether categorisation as an Exempted chemical is appropriate, the introducer would need to evaluate the chemical against the criteria in all the hazard bands in order to establish that it was not hazardous.

While an introducer may not be required to fully characterise the chemical’s hazards in order to fulfil NICNAS requirements (for example, if a company decided to accept the categorisation of a Reported chemical instead of sourcing additional information that may indicate the chemical could be categorised as Exempted), knowing the hazards of the chemical being introduced may still be required under other legislation, such as work health and safety or product safety requirements.

Risk matrices

Why are risk matrices being used?

A matrix is simply a visual representation to describe risk as a function of hazard and exposure. It is being used to assist people to understand how the regulatory framework for new chemicals proposed in the NICNAS reforms is risk-based and proportionate, and creates greater certainty for industry in the categorisation of chemicals. By stratifying the regulatory treatment of new chemicals in this manner, the matrix shows how NICNAS’s assessment effort would be focused on higher risk chemicals, reducing the impost on those chemicals that are of lower risk.

The proposed framework described by the matrix replaces the need for a large number of certificate, permit and exemption categories that currently exist. While the current system is also somewhat stratified with exposure criteria (largely volume based) and some hazard criteria across the numerous new chemicals categories, it is a much less refined approach than the proposed framework (as depicted in the matrix). As identified in the Regulation Impact Statement, the current regulatory framework does not sufficiently accommodate risk in applying regulatory effort.

Why are there separate matrices for health and environment?

Having two separate risk matrices enables a simplified representation of human health and environmental exposure and hazard criteria. It allows for different exposure criteria to be set such that the indicative risk prediction is much more refined. For example, by having a separate environment matrix, release volume (which is more aligned to the actual risk posed to the environment), rather than introduction volume can be used to estimate risk.

Why are the risk matrices so complex?

As the risk matrices are a means to visualise a risk-based categorisation drawing on both hazard and exposure parameters, some people may find it more complex than the list of largely volume-based assessment categories that are currently in place for the notification and assessment of new chemicals. However, when compared more closely to the current system of multiple permit, exemption and certificate categories, each with their own set of applicability criteria (including some with both hazard and exposure criteria), the matrix actually represents a considerably simpler and more refined approach to determining the regulatory treatment required for a new chemical.

The risk matrix has been designed (and now refined further based on stakeholder feedback) to strike a balance between realising the maximum gains from applying a risk-based categorisation system, with the need to have a system which is practicable and facilitates compliance.

Why is the matrix read top down and not bottom up?

A bottom up approach to applying the matrix would not result in an assessment of the highest (most serious) levels of hazard. For example, if a bottom up approach is used and a chemical is identified as meeting the criteria for Hazard Band B for both human health and the environment, then the introducer may stop assessing the hazards at this point. However, the chemical may still meet criteria for Hazard Bands C, D or E. This bottom-up approach may result in incorrect categorisation of the chemical and inadequate protection of human health and/or the environment. Using a top down approach results in the highest hazards being identified first, resulting more quickly in the correct categorisation of a chemical.

The top down approach determines that the chemical does not have the most serious hazards, and allows categorisation of the chemical to terminate once the first positive result is obtained. While the matrix ensures that chemicals known to have these higher hazards are assessed pre-market, the information requirements have been developed so as not to require the generation of data for these endpoints, unless screening tests indicate a concern (for example, if in vitromutagenicity tests are positive, in vivodata may be needed to fully characterise the hazard). See Attachment B for more detail on how these higher hazards would be evaluated.


A good example of a chemical that would not be appropriately categorised if a ‘bottom-up’ approach was used is the phthalate, di(2-ethylhexyl) phthalate (DEHP). This chemical has low acute toxicity by all routes. It is not irritating to skin or eyes. It is a non-sensitiser. It is not classifiable for repeat dose toxicity. Genotoxicity results are negative. It is not considered to be carcinogenic. However, it is known to be dangerous with respect to male fertility. Therefore, unless the higher order hazard of known reproductive toxicity was considered first, this chemical would have been incorrectly placed into Hazard Band A, and categorised as an Exempted chemical.

Why should the absence of data be considered positive for hazard?

The amount and quality of hazard information required varies according to the exposure band. A data gap occurs where one or more of these requirements are not met. A data gap does not confirm the existence of a specific hazard, but does not allow an assumption that there is no hazard. The absence of data to characterise a hazard may require expert assessment to determine the need for risk management, particularly where the absent data may be critical in determining whether a chemical should be assessed by NICNAS to determine the need for (and appropriate type of) risk management.

How has the human health risk matrix been simplified?

The following key changes have been made to the health matrix.

Stakeholder feedback

Change to reflect stakeholder feedback

Too many hazard bands

The hazard bands for human health have been rationalised from 5 down to 4.

Reduce stratification to reduce complexity

Stratification of certain hazard endpoints across hazard bands have been removed, where no significant benefit in risk outcome was obtained (for example, the GHS categories for Single Target Organ Toxicity-repeated (STOT) have been combined into one hazard band).

Simplify consumer categories for use

The human health exposure categories for consumer use have been simplified. This was based on stakeholder feedback indicating that separating out rinse-off and leave-on cosmetics would be difficult to implement in practice and because, with the rationalisation of the hazard bands, there is less benefit in separating these two product categories.

Modify volume thresholds to better align internationally

Volume thresholds for the human health Exposure Bands 3 and 4 have been modified to create better international alignment—e.g. the introduction of the volume threshold of >10 T for non-cosmetic uses allows for alignment of the repeated dose information requirement with Canada and the EU.

Revise presentation of the matrix for the highest hazard band of Exposure Band 2

Presentation of the human health matrix for the highest hazard band of Exposure Band 2 has been changed to make it clear that a chemical known to have these hazards (carcinogenicity, mutagenicity and/or reproductive or developmental toxicity (CMR), or endocrine disrupting chemicals (EDC)) would require pre-market assessment.

The proposal remains that the introducer would not need to generate new information to characterise these hazards. Instead, for Exposure Band 2, an introducer would just need to determine whether their chemical was a known CMR or EDC by consulting specified international lists and considering any already available classification information.

This is consistent with current practice for chemicals introduced under exemption categories (which NICNAS advises in guidance cannot be CMR), and is also aligned with how CMRs are treated in the EU for cosmetics (in that they are prohibited for use in cosmetics if they are already classified as a CMR, unless they meet specific criteria which include (but are not limited to) that they have been evaluated and found safe by the SCCS).

With the proposed changes to the health risk matrix, is the reduction in pre-market assessment by NICNAS still expected to be more than 70%?

Yes. NICNAS has analysed the new chemical assessments completed in a three-year period (2012–2015) against the revised health risk matrix as proposed in this paper, and has confirmed that the reduction in pre-market assessment by NICNAS under the reforms is still expected to be more than 70%.

How will gaps in information (data) be managed in the matrix?

Evaluation of a hazard endpoint is made based on total weight of evidence, which means that all available information relevant to the toxicity of the chemical is considered together. This may include results of in vitro and in vivo tests, in silicomodelling, the chemical’s physical and chemical properties, epidemiological and clinical studies and documented case reports.

If the information to characterise a particular hazard is not sufficient or readily available, it is the introducer’s responsibility to either source or generate the information if it is a requirement for the exposure band into which the chemical falls. For the purposes of categorisation, if information cannot be obtained, the hazard of the chemical will fall in the highest band for the purposes of applying the risk matrix where there is a data gap. An absence of data does not necessarily indicate a specific hazard. However, in the absence of data, there may be a need for expert judgement, particularly where the data gap is critical in deciding whether a chemical could be categorised as an Assessed chemical.

If the data gap results in the chemical falling under the Assessed category, NICNAS will undertake the assessment and determine the implications of the gap. Depending on the likely risk of the chemical, NICNAS may require the introducer to obtain further information. Alternatively, given the inherent uncertainty arising from the lack of data, NICNAS may use a precautionary approach to the risk assessment. This may include recommendations for controls to mitigate a risk that cannot be ruled out based on the available data.

In some cases, waivers and exclusions may apply to particular information requirements, similar to the operation of the current NICNAS framework. Guidance on established waivers will be provided as part of the details for the hazard endpoints. Given the extensive range of possible waivers, introducers will be able to consult with NICNAS regarding additional information waivers on a case by case basis.

  1. In some cases, a chemical’s physical properties prevent certain types of tests from being undertaken. For example, if a chemical is corrosive it is not necessary to conduct an acute oral toxicity test.
  2. For the purposes of categorisation, if a chemical is only ever introduced in a formulated product at low concentrations, or the introducer holds data on the irritation potential of the product, the requirement for irritation information on the neat chemical may be waived.


What will happen to PLCs under the reforms?

PLC criteria will be maintained under the reforms, with some changes proposed to better align them internationally.

For human health, most polymers that meet the PLC criteria will be categorised as Exempted chemicals. Those with lung overloading potential that are proposed to be used in aerosol cosmetics, and those with known hazards (i.e. pre-existing GHS classifications) may be categorised as Reported or Assessed chemicals.

For the environment matrix, polymers that meet the PLC criteria will be categorised as Exempted chemicals.

What are the proposed changes to the PLC criteria?

NICNAS is aiming to:

  • achieve greater international alignment with the equivalent criteria of USA and Canada
  • clarify current criteria, and/or
  • change the criteria to ensure greater consistency with the proposed new legislation.

It is proposed that changes be made to the following aspects of the PLC criteria:

  • Polyesters: proposed changes include additions to the list of prescribed reactants for polyesters (outcome of work undertaken by the OECD Clearing House on New Chemicals (CHNC))
  • Reactive functional groups: alignment of the list of moderate concern functional groups to that of USA and Canada
  • Removal of the “not a hazardous chemical” PLC criterion
  • Clarification of the stability criterion
  • Addition of a criterion such that polymers containing certain perfluoroalkyl moieties cannot be PLCs (this criterion will be closely based on the equivalent criterion for polymer exemption in the USA), and
  • Other minor changes, for example, amendments to the number average molecular weight boundaries (i.e. >/≥ or </≤).

How will polymers be treated in the matrix?

For human health, it is proposed that the chemicals risk matrix would not apply in the same way to polymers with high molecular weight (i.e. NAMW >1,000 Da) as for other chemicals. Instead, the hazard characterisation for these polymers would be determined in accordance with a rule-based decision tree. This characterisation would mean that:

  • all polymers that fall into Exposure Band 1 will be categorised as Exempted
  • for the vast majority of polymers with NAMW >1,000 Da, it would not be necessary to hold hazard data on the polymer. However, any existing hazard classification of the polymer would need to be considered when characterising the polymer’s hazards (for example, if the polymer is already classified under the GHS as a corrosive).
  • for polymers that do not meet the PLC criteria (and have NAMW >1,000 Da), the characterisation of their hazard would be largely based on structural considerations, such as the amount of reactive functional groups on the polymer. The vast majority of such polymers are expected to be categorised as Reported

Polymers with high frequency of known sensitising groups (such as isocyanates) would be characterised such that introduction in uses with:

  • medium to high exposure (Exposure Band 3 or 4) would result in the polymer being categorised as Assessed
  • low exposure (Exposure Band 2) would result in the polymer being categorised as Reported.

For environmental risk, it is proposed that the risk matrix for new chemicals would also apply to polymers. For the purposes of that matrix, it will be assumed that:

  • High molecular weight (NAMW >1,000 Da) non-ionic and anionic polymers, assuming no hazardous degradation products or toxic counter-ions:
    • are not bioaccumulative, and
    • do not have acute ecotoxicity endpoints ≤100 mg/L, or chronic ecotoxicity endpoints ≤1 mg/L if their Functional Group Equivalent Weight (FGEW) is >5,000.

Therefore, all high molecular weight (NAMW >1,000 Da) non-ionic and anionic polymers with FGEW >5,000, assuming no hazardous degradation products or toxic counter-ions, would fall into Hazard Band A and would be categorised as Exempted. Hazard data for these polymers would not need to be held.

  • High molecular weight (NAMW >1,000 Da) cationic and amphoteric polymers are not bioaccumulative. However, characterisation of their potential ecotoxicity would be required to determine whether the polymer falls into Hazard Band A, B, C or D. This would be largely based on structural considerations and defined structural activity relationships (SARs).

How do I categorise chemicals if I am an importer of finished products who does not have access to information about the chemicals?

It would be the responsibility of introducers to ensure that they meet the NICNAS requirements relating to the chemical they propose to introduce. Introducers may rely on advice from others but they must be satisfied as to the veracity of this advice because it is the introducer who is ultimately responsible for meeting NICNAS requirements in relation to the chemical. This is the same legal principle that applies in other areas – for example, employers may rely on advice from others as to the safety of the products their workers use, but ultimately the employer must be satisfied (and will be responsible for ensuring) that they are meeting their obligations under the work health and safety legislation.

How can a chemical in a consumer product applied to the skin be considered to be 100% released to the environment given some absorption to skin?

Currently, the default assumption used for environmental release of a cosmetic chemical for the purposes of a regulatory risk assessment is 100% of the introduction volume. Details of the environmental risk assessment process are outlined in the Environmental Risk Assessment Guidance Manual for industrial chemicals. A release assumption of 100% is not always appropriate and, if required, the assumption is refined on a case-by-case basis during risk assessment. Industry will need to self-categorise under the reformed approach and there is currently no OECD Emission Scenario Document for cosmetic uses. NICNAS is prepared to work with the cosmetics industry to develop realistic release assumptions for different types of cosmetics, or accept submissions providing evidence justifying a different assumption.

Why has the “<1% introduced” scenario been included in Exposure Band 2?

The placement of this scenario in Exposure Band 2 is part of the risk-proportionate approach to the reforms and is reflective of the lower risk posed by such a scenario. If a chemical is imported at <1% concentration (regardless of the volumes at which it is imported), the risk to workers is lower than if the same chemical were to be imported at a higher concentration and then either reformulated or incorporated into an end use product in Australia at <1% concentration. It is also lower risk to workers than if the same chemical were to be manufactured (i.e. synthesised) in Australia and then reformulated to <1% concentration in end use products.

The concentration limit of <1% has been chosen as it aligns with many of the concentration cut-offs for classification of mixtures under the GHS.

I’m introducing 1,001 kg of a non-cosmetic chemical; why does this suddenly change my exposure band for human health? What is the difference in risk between 1,000kg and 1,001kg?

Having less clearly defined thresholds would result in ambiguity for anyone attempting to categorise their chemical under the matrix. For the purpose of removing complexity, it is necessary to define thresholds. This approach is consistent with other international regulators.

I am planning to import a new chemical into Australia that will be stored at my company’s warehouse and remain unopened. It will subsequently be exported in its entirety. What is the appropriate exposure band for such a chemical?

Exposure Band 2 for both human health and the environment.

The scenario described is not covered by the definition of ‘transhipment’ as the chemical does not remain under customs control. However, a ‘Contained import/export’ scenario has been added to Exposure Band 2 for human health. The requirements to meet the ‘Contained import/export’ scenario in Exposure Band 2 are as follows:

  • the new chemical is imported into Australia, and
  • the new chemical does not undergo any form of treatment or processing during its time in Australia (i.e. the packaging in which the new chemical is contained remains unopened), and
  • the new chemical remains the responsibility of the importer/exporter at all times whilst it is in Australia, and
  • the entire quantity of the chemical leaves Australia.

Last update 17 February 2016