The periodic retest has a reputation as paperwork. I keep the test bay records, and the records say otherwise. Here is three years of them.
What the regulation requires
A composite IBC in hazardous materials service has a five-year service life from the moulding date on the specification plate. At the 2.5-year point it needs a visual inspection, internal and external, plus a leakproofness test at a minimum of 3 psi held for the specified dwell. At five years it needs that again plus a hydrostatic test to the plate rating.
The five-year life is absolute. No test extends it. That is the single most expensive misunderstanding in this field and I will come back to it.
The failure distribution
| Failure mode | Share of failures | Typical remedy | Part cost |
|---|---|---|---|
| Lid gasket degradation | 31% | Replace gasket | $9 – $22 |
| Valve seat gasket degradation | 23% | Replace gasket | $9 – $22 |
| Bottle stress cracking, bottom radius | 21% | Rebottle or retire | $135 – $210 |
| Valve body crack or distortion | 14% | Replace valve | $28 – $46 |
| Closure not seating (thread damage) | 7% | Replace lid, sometimes retire | $14 – $27 |
| Other / indeterminate | 4% | Investigate | — |
Periodic test failures by cause, 2020–2023
Fifty-four per cent of failures are a gasket. That is the headline and it is slightly embarrassing for the industry, because a gasket is a consumable that costs less than a sandwich and should be replaced on schedule rather than discovered at test.
Why gaskets dominate
Three reasons, in order of how often I see them.
- Wrong compound for the product. EPDM in aromatic service swells and loses its sealing geometry within months. Viton in hot water above 130 °C does the same. The gasket was never going to survive, and nobody checked compatibility at order time.
- Compression set. Any elastomer held under compression for 30 months takes a permanent deformation. Even a correctly specified gasket is a wear item on this timescale.
- Over-torquing. Somebody put a bar on the valve nut. The gasket extrudes, thins at one point and leaks there. We see this constantly and it is entirely avoidable — hand tight plus a quarter turn is correct.
More than half of all periodic failures are a nine-dollar part that nobody put on a replacement schedule.
| Compound | Right for | Fails with | Typical life in service |
|---|---|---|---|
| EPDM | Water, aqueous chemistry, caustics, most food | Aromatics, solvents, fuels, oils | 24 – 36 months |
| Viton (FKM) | Aromatics, chlorinated solvents, fuels, esters | Hot water >130 °C, amines | 36 – 48 months |
| Silicone | Hot-fill food, steam exposure | Solvents, oils, abrasion | 18 – 30 months |
| Nitrile (NBR) | Petroleum products, vegetable oils | Ozone, UV, strong oxidisers | 24 – 30 months |
Gasket compound against service, and what goes wrong
The 21% that matters most
Bottle stress cracking at the bottom radius is the minority failure and the one that justifies the whole test regime.
That radius is where the blow-moulded wall transitions from side to base. It is a geometric stress concentration even at rest, it carries the hydrostatic load of a full tote, and it takes the impact every time a pallet jack sets a unit down hard. Cracks initiate there, propagate slowly, and are invisible from outside because the cage is in the way.
A 3 psi air test finds them. A visual inspection from the outside does not. When people describe the periodic as a formality, this is the category they are gambling against, and the downside is a bottom-corner failure on a full unit of regulated liquid.
Worth noting what the test does not find: slow permeation, wall thinning that has not yet cracked, and fatigue in the cage welds. Those need the visual inspection, which is why the regulation requires both and why a test-only programme is incomplete.
The thing nobody checks until it is too late
We audit customer fleets and the same finding comes up every time: units that quietly left hazmat service without anybody noticing.
The service life runs five years from the two-digit moulding year in the marking string. A 2019 unit left service in 2024. Retesting it does not bring it back — there is no mechanism in the regulation for that. It is perfectly legal for non-regulated liquid indefinitely, and completely unusable for the regulated product it has been carrying.
In the last four fleet audits we ran, between 6% and 19% of units presented as "in service" were past date. Nobody was being careless. The plate is small, the year is two digits, and nobody owned the schedule.
| Fleet size | Past service life | Periodic overdue | Illegible plate |
|---|---|---|---|
| 180 units | 6% | 14% | 2% |
| 420 units | 11% | 22% | 5% |
| 95 units | 19% | 31% | 8% |
| 640 units | 8% | 9% | 3% |
Fleet audit findings, four customer fleets
The illegible plate column is the one people find most alarming. A corroded or painted-over plate means the container is not certified for regulated transport, and there is no way to re-establish an illegible marking short of full re-testing and re-marking.
What to do with this
- Put gaskets on a replacement schedule rather than a failure schedule. At every wash, not at every test.
- Specify gasket compound against the product at order time. The compatibility table is on our valves and fittings page.
- Photograph every specification plate in your fleet once, and build the periodic schedule from the moulding years. We will do this for you as a fleet audit.
- Treat a bottom-radius crack as terminal for that bottle. The cage is almost certainly fine — that is a rebottling candidate.
Reading the marking string character by character is covered on UN markings decoded.
Written by
Priya Raghunathan
Head of reconditioning
Runs the three wash lines and the test bay. Has rejected more totes than anyone else in the building, which is the point of the job.