The principles of classification

In this feature article Desmond Waight discusses the basic principles of classification of chemicals for transport.


A short while ago on one of the LinkedIn questions and answers, a person was seeking assistance in classifying certain mixtures intended for air transport.

The question posed was as follows

“Second Opinion: Product Shipment Classification & Identification.

I would appreciate a second opinion relating to the classification and identification of this shipment for air transport under IATA regulations.

The shipment details are as follows:

Product 1 consists of solution containing:


I wish to send 1L of the Product 1 in the one package as a fully declared DG shipment. I would appreciate members’ opinion on what the classification and identification name for both these products would be and state on the PSN label to see if it matches with my classification.”

My response was “Sorry — you cannot work out a classification based on the information above” and concluded with “Very regrettably, but based on the fact that you asked such a question, I have to say that you really do appear to need [re]training on classification of dangerous goods for transport”.

Other examples of similarly worded questions have been seen, including one in a dangerous goods safety adviser (DGSA) exam question, which spoke of a mixture containing constituents which were individually flammable and toxic, but without specifying the relative percentages so one was unable to work out to what extent the effect of the toxic component may have been diluted by the presence of the flammable component or vice versa.

So what are the principles of classification?

Determine the hazards involved

For transport, the first principle is to assess whether the goods to be carried, as will be presented for carriage, intrinsically (ie without consideration of the actual quantity to be carried) possess one or more of the hazard characteristics specified in the applicable legislated text.

In the case of the LinkedIn question, in relation to Product 1, the classifier would need to know the following.

  • The actual flashpoint, if any, of the mixture. This would normally need to be ascertained by testing. From this whether the Product 1 has Class 3 hazards and at which Packing Group (PG) level.

  • If testing shows a flashpoint below 60°C, but in excess of 35°C, then whether or not the product sustains combustion when tested according to the UN’s “Manual of Tests and Criteria (MTC)”. If it does not, then the product may be deemed not to have Class 3 hazards.

  • The acute toxicity of the mixture. If one knows the percentage of component A and the applicable LD50 value (or the percentage concentration of the active pesticide substance within component A and the LD50 value of the active pesticide substance), then it is possible to calculate the LD50 value for Product 1 and, thus whether or not, Product 1 has hazards of Class 6.1. Though the transport rules do not yet include the Globally Harmonised System (GHS) mixtures classification criteria, it would generally be acceptable to use the GHS classification and if acute toxicity category 1, 2 or 3 to read across as Class 6.1 PG I, II, or III.

  • The aquatic toxicity (acute and chronic) of the mixture resultant on the environmentally hazardous component B. Again, one would need the percentage concentration of component B in the Product 1 and the percentage breakdown of component B in order to calculate whether or not Product 1 meets the criteria for classification as environmentally hazardous.


While for air shipment, this classification would not need to be carried out if the earlier determination is not sure that the product has Class 3 and/or Class 6.1 hazards. It would need to be carried out because of the road transport normally needed before/after the air journey.

Having determined which hazard class(es) is (are) applicable and the PG level (if applicable), the next stage is to determine the precedent hazard.

Hazard precedence

If the hazard determination indicates only one or no hazard classes are applicable, that concludes the classification process (the next stage being identification, which is not being dealt with in this article).

However, if hazard of two classes are identified, then rules of precedence have to be applied. But, if one of the hazard types identified is Environmentally Hazardous Substances (EHS) (or Marine Pollutant (MP) as the sea rules call it), then this is not considered in this determination, but is treated as a supplementary classification under the International Maritime Dangerous Goods (IMDG) code and Alternative Dispute Resolution (ADR) provisions (but not for air as air assumes that if it has any other hazards then it is also automatically EHS; without the need for this to be mentioned).

Presuming that the Product 1 has, on investigation, been found to have both Class 3 PG III and Class 6.1 III hazards then application of the precedence rules, eg as in ADR, determines that the Class 6.1 hazard takes precedence, since the Product 1 is presumably a pesticide, with the Class 3 being a subsidiary hazard.


In ADR, as shown in the precedence table, there is a “Class 9” column. In the view of the author, this is misleading since Class 9 can never be a subsidiary hazard class. This is not included in the UN Orange Book and IMDG code.

Special issues

Multiple hazards

What, however, if an investigation shows that there are, neglecting any Class 9 or EHS/MP hazards, more than two hazards involved?

Here, one example included in the ADR text needs to be noted. In the ADR provisions where it deals with the allocation to primary and subsidiary classes of chemicals having multiple hazards, following the precedence table at (under the note 1 an example for the classification of a mixture) it includes the following note.

So, it would appear from this example that the third hazard is simply forgotten about.

The various regulatory texts do not specifically include instruction on what to do in the case where there are more than two hazards (apart from the hint from ADR’s example above); though there are some entries which show one primary and two subsidiary hazards (eg UN 2988 CHLOROSILANES, WATER-REACTIVE, FLAMMABLE, CORROSIVE, NOS).

The author of this article has in the past raised this with the UK CA who in 2012 advised:

“While we note that there are some entries in the Model Regulations that have two sub-risks, looking at this in general terms we do not think this is a good idea — periodically the issue of entries reflecting three hazards is raised (and all the various permutations there can be!). Hazard precedence can be used to determine the two main hazards and the overall classification. Specific entries, eg for various chlorosilanes are best dealt with on a case by case basis.”

So, unless the product fits on one of these listed entries, the general rule is, somewhat surprisingly, to ignore the hazard that comes last in the precedence routine.

Variable constituents

What, however, if the constituents may be present at different times in different percentages, such as can happen with certain waste streams?

Here, ADR uniquely in again suggests a solution:

“If this [classification based on actual known properties] is not possible without disproportionate cost or effort (as for some kinds of wastes), the substance, solution or mixture shall be classified in the class of the component presenting the major hazard.”

That is you treat a waste as if it has 100% of each hazardous component and apply the precedence rules as shown in the classification of mixtures example as shown above. Hence, if it happens that one particular drum of waste actually ends up with 90% of the flammable component and 10% in total of the toxic and corrosive components, it would still be classified as UN 2911 WASTE CORROSIVE LIQUID, TOXIC, NOS of Class 8(6.1).

Unknown hazards

What, however, if the goods are new chemical substances being sent out for testing — eg to determine their hazardous properties? The principle applied here is a precautionary principle and set out in a section on Classification of Samples (eg ADR 2.1.4, IMDG 2.0.4). This requires the consignor to make a best educated guess as to the classification and choosing the most severe PG (where applicable for that hazard). The section then applies restriction of quantity per package and type of package to minimise the risk in transport.