Packing Groups I, II, III: What a Fibreboard Box Must Pass

Your safety data sheet says UN packing group II. The quotation in front of you says "UN certified carton". Between those two statements sits a design-type test file that decides whether the box was dropped from 1.2 metres or from 0.8 metres, whether it was conditioned for a day at controlled humidity first or pulled straight off a Shenzhen production floor in August, and whether the liner it is made from absorbs more water in half an hour than the regulations allow. Forty centimetres of drop height is the entire practical difference between a packing group II box and a packing group III box, and no amount of looking at the two cartons will tell you which one you are holding.

The UN packing group is not a grade of packaging. It is a property of the substance, and it propagates into the packaging through a chain that is worth following end to end, because each link is a place where a supplier can answer imprecisely and a buyer can nod.

Where the packing group comes from

Packing group is the degree of danger within a hazard class: I is high danger, II medium, III low. It is assigned by applying the classification criteria for the relevant class to the substance itself — oral, dermal and inhalation toxicity data for Division 6.1, flash point and initial boiling point for Class 3 flammable liquids, corrosion rates and exposure times for Class 8, and so on. The result is published against the entry in the Dangerous Goods List, and mirrored in the hazardous materials table used in the United States.

Two consequences matter for packaging buyers.

First, the packing group belongs to the substance, so it is settled before any packaging conversation begins. If your supplier of the chemical has not given you a classification, no amount of carton specification will rescue the shipment. Get the classification, the UN number and the proper shipping name in writing first.

Second, not every class has packing groups at all. Class 1 explosives use compatibility groups, Class 2 gases have none, Class 7 radioactive material runs its own regime, and Division 6.2 infectious substances are handled through a separate packaging system. Asking "what packing group is it" about a gas cylinder shipment signals to a forwarder that the classification work has not been done.

From packing group to the letter on the box

The design type is tested at a performance level, and the level is recorded as a single letter in the UN packaging mark:

• X — tested to packing group I severity. Authorised for packing groups I, II and III.
• Y — tested to packing group II severity. Authorised for packing groups II and III.
• Z — tested to packing group III severity. Authorised for packing group III only.

The compatibility runs one way. An X box carries anything; a Z box carries only the mildest tier. Reading the whole mark string — code, letter, rated gross mass and the rest — is a separate skill worth having, and the field-by-field decode of a UN packaging mark covers it.

One detail that saves arguments later: a design type tested at a given gross mass may be marked at a lower gross mass than it achieved, but not a higher one. So a supplier can legitimately offer you the same physical carton marked Y20 rather than Y25 if that suits your fill. The reverse — marking above what was tested — is not a commercial concession, it is a false statement.

The test battery behind a 4G design type

For a fibreboard box, design-type qualification is a short list of tests plus a set of material requirements. It is short enough that you can ask for each element by name.

Conditioning: the 24 hours that decide the result

Fibreboard is hygroscopic, and its compression and drop performance move with moisture content. Before testing, fibreboard packagings are conditioned in a controlled atmosphere — commonly cited as 23 °C ± 2 °C at 50% ± 2% relative humidity for at least 24 hours, with an alternative atmosphere permitted. This is the single most skippable step in the whole battery and the one that changes results most.

The practical question for a buyer is not whether conditioning is required; it is whether the laboratory has a conditioning chamber and whether the report names the atmosphere and the duration. A drop test report on fibreboard that does not state the conditioning is not a complete report. It is a number obtained at whatever the ambient humidity happened to be that day, and in the Pearl River Delta that number varies by season. The same physics drives ordinary carton stacking failures, which is why compression strength has to be derated for humidity even outside the dangerous-goods world.

The drop test, and the heights that define each group

The drop test is where the packing group becomes a physical number:

Heights for the higher-density liquids are rounded up to the first decimal place. Confirm all of these against the current regulatory text before writing them into a specification — they are stable figures, but a specification that cites a number should cite the paragraph too.

Three points that reports either contain or quietly omit:

• Orientation matters as much as height. For boxes, the sequence covers a flat drop on the base, a flat drop on the top, a flat drop on the longest side, a flat drop on the shortest side, and a drop on a corner. A report showing a single flat-bottom drop has tested one of five conditions.
• Fill level is prescribed. Solids are filled to not less than 95% of capacity and liquids to not less than 98% before the drop, because a part-filled package behaves differently. A test run on an empty or lightly loaded box is not a design-type test.
• The pass criterion is about safety, not appearance. Crushed corners and split flaps are acceptable outcomes; loss of contents from an inner packaging, or damage to the outer that would compromise the inners in onward transport, is not. Expect a report to describe the condition of the packages after the drop, not simply to say "pass".

Stacking, vibration, and the Cobb number nobody quotes

The stacking test applies a load equivalent to the total mass of identical packages stacked to three metres, including the test sample, held for a minimum period commonly cited as 24 hours for fibreboard. The criterion is no deterioration that would affect transport safety and no distortion that would reduce strength or destabilise a stack.

The material requirement buyers rarely ask about is water absorption. Fibreboard used for a 4G box has to meet a limit on the Cobb water absorption test — commonly cited as no more than 155 g/m² over 30 minutes on the outer surface, measured to ISO 535. That figure does not appear on the mark and seldom appears on a quotation, but it is the property that determines whether your box holds together on a wet quay or in the bottom of a rain-soaked container. It is also directly controlled by the liner specification and the sizing chemistry, which is why it belongs in the board conversation rather than the compliance conversation.

The United States adds a test the UN Model Regulations do not contain: a vibration test on the design type. A packaging qualified only against the UN battery may have no vibration data on file, which is a specific gap to check if the boxes are heading for domestic US movements rather than pure export traffic.

Two further construction requirements are worth quoting into a purchase order because they are cheap to verify on a sample: manufacturer's joints must be taped, lapped and glued, or lapped and stitched, with an adequate overlap; and where closure is achieved by gluing or taping, a water-resistant adhesive is required. Both are visible on a sample box in about a minute.

What "the design type" actually covers

This is where most non-compliance is created by people acting in good faith.

A 4G design type is not a box. It is a tested system: this outer carton, at these dimensions, in this board construction, containing these inner packagings, with this cushioning, closed this way. The test report describes the whole system. Change any element and you are outside it.

That means the following are all design-type changes, not procurement details: switching a 500 ml HDPE inner bottle for a glass one of the same volume; reducing the cushioning from a moulded EPE insert to loose fill; changing the number of inners per carton; increasing the outer dimensions; changing tape width or the number of strips at closure.

The lever that exists for buyers with a SKU range is the variation packaging — a design type qualified with cushioning and a defined mass of inner packagings, marked with a trailing V, which permits inner packagings to be varied within the tested envelope. If you ship eleven bottle formats in one outer, qualifying one variation design type instead of eleven separate ones is the difference between a manageable compliance file and an unmanageable one. It has to be specified at enquiry stage, because it changes how the test is set up.

Putting this to a supplier

The CMH listing for Jindong Packaging — a Shenzhen paper-packaging manufacturer operating since 2007 from a 12,000 m² plant in Xinqiao, Bao'an, with around 50 staff — names dangerous-goods (UN) cartons among its lines, and separately names EPE pearl-cotton and EVA foam cushioning and in-house board production on automatic corrugated lines. Cushioning made in the same plant as the outer is operationally relevant to design-type work, because the cushioning is part of what gets tested.

The listing does not state which packing groups any design type covers, which drop heights were used, which laboratory ran the tests, what the rated gross mass is, or whether the underlying reports can be released to a buyer. None of that should be inferred. Ask for it, in the same way you would ask any supplier, and read what comes back the way you would read any other supplier document — the general habits behind verifying a China supplier's claims and understanding what a management-system certificate does and does not prove both apply directly here.

Common questions

Can I use a Y-rated box for a packing group I substance if I under-fill it?

No. The performance level is a property of the design type, not of how full you make it. Packing group I requires a design type tested at packing group I severity, which means an X mark.

Does a UN test report cover my exact carton size?

Only if your size is within the scope the report states, or within a permitted variation. Dimensional changes generally require either a fresh test or an explicit variation provision. Ask the supplier to point to the clause that covers your size rather than accepting a verbal assurance that it is "the same box".

Who holds the test report, and does the testing have to be third-party?

The factory holds the qualification records, and you want a copy in your own compliance file — in an incident or an inspection you will be asked what basis you had for believing the packaging was compliant, and a supplier's assurance is a weaker answer than a report with a laboratory name and a date on it. Whether the testing must be third-party depends on the authorising state: some allow manufacturer testing under competent-authority oversight, others require designated laboratories. Ask which laboratory ran each test and whether it is designated by the authority named in the mark. Where you want independent assurance on the physical goods rather than on the design type, that is a separate job for pre-shipment inspection against your own agreed criteria.

What to ask your supplier next

Send this list with the enquiry, and keep the answers in the compliance file alongside the purchase order:

1. State the packing group the design type is qualified for, and the corresponding performance letter on the mark.
2. Send the drop test report, showing the drop height, the number of samples, the drop orientations and the post-drop condition of each.
3. Confirm the conditioning atmosphere and duration recorded in that report.
4. Send the stacking test report, with the applied load and its duration.
5. Send the Cobb water absorption result for the outer liner, with the method and the limit applied.
6. List every element of the design type: outer dimensions, board construction, inner packagings, cushioning, closure method.
7. Confirm whether it is a variation (V) design type, and if not, what qualifying one for my SKU range would cost and take.
8. Confirm the vibration test position if any part of the movement will be a domestic US shipment.
9. Name the testing laboratory and the authorising competent authority, with the reference and the date of the most recent requalification.

Jindong Packaging is a reasonable place to put that list — a CMH-listed Shenzhen carton maker building to drawing or sample on an ODM full-material basis, quoting sampling at about seven days and bulk production at about fifteen working days from 1,000 pieces on T/T terms. Questions 2, 3 and 6 are the discriminating ones. A supplier that can hand over a drop report naming the conditioning atmosphere and a design-type description listing the inner packagings is running a compliance file; a supplier that answers with a certificate number is showing you a document, and a document is not a test.