17 replies to this topic

[Simon]

My brains a bit whacked, this may have been asked before. How often should migration testing be carried out on a food contact material? Are there any legal requirements (any country)? What would seem a normal length of time that test results can still be considered credible (scientifically)?

For anyone not familiar with migration testing - it is carried out to demonstrate a materials suitability to be in contact with food; the testing itself is usually carried out using various stimulants such as olive oil and acetic acid, with I imagine time and temperature and to see what if anything migates from the material into the stimulant.

Thanks,
Simon

[okido]

Hi Simon,

For food contact material a validity period of six years for migration reports is common for me.
After three years an evaluation on relevant regulations and new scientific insight is obligatory.

Have a nice day, Okido

[Simon]

Hi Simon,

For food contact material a validity period of six years for migration reports is common for me.
After three years an evaluation on relevant regulations and new scientific insight is obligatory.

Have a nice day, Okido

Hello Okido.

I can understand that a continued eye should be on changes to legislation, science, or product formulation and would warrant new migration tests.

Outside of this why 6 years? Just for the heck of it??? This is the part I do not understand. Why the need to re-test at all and where does the frequency come from?

[Charles.C]

Dear Simon,

Outside of this why 6 years?

Why not Yearly ? (Laboratory / Auditor / Consumer Viewpoint.)

I'm surprised BRC would accept anything more . (No icon for vampires unfortunately)

Rgds / Charles.C

[Simon]

Dear Simon,
Why not Yearly ? (Laboratory / Auditor / Consumer Viewpoint.)

I'm surprised BRC would accept anything more . (No icon for vampires unfortunately)

Rgds / Charles.C

Why not monthly, weekly, daily???

[Charles.C]

Dear Simon,

OK, If one needs a specific reference point with international acceptance, how about 1.5 x World Cup.

(Still trying to work out who Jane Wagner is. No relation to Robert I suppose.)?

Rgds / Charles.C

added - JW - She is the author of The Search For Signs of Intelligent Life in the Universe, The Incredible Shrinking Woman etc.

I might have known.

[Charles.C]

Dear Simon,

Despite my attempt to infiltrate football into wider fields, I do agree yr question merits a proper response both for yr topic and in general.
I actually ran into a similar conundrum over documentation of reference weights for use in calibration of balances. Auditors demanded a yearly renewal of testing documentation – their justification was that a maximum of 1 yr is “standard practice” in the “trade”. Take it or leave it. Similar logic also applied to thermometers, metal detectors etc. Clearly on a technical level these items are not equivalent but how to quantitate such ??

Accordingly I was frankly astonished at Okido’s 6 yr comment. I would certainly like to know his auditor’s comment (perhaps he is very skilful at diversionary tactics ?). I’m not professionally in food contact packaging so I don’t know how high on the list of potential risks, migration ranks but I get the impression this is not a minor parameter ? However I have experienced it from the user’s side and some customers simply demand a yearly renewed certification of “their” packaging which includes this test. BRC etc food auditors IMEX tend to be remarkably trusting of packaging in general so do not probe too much into such things, enough problems already for their knowledge base perhaps + no specifics in the food "rules" from memory and no "trade" thumb rule to fall back on (yet).

Surely other posters have experienced similar situations ?

Rgds / Charles.C

[okido]

Hi Charles,

perhaps he is very skilful at diversionary tactics

Thanks for the compliment Charles, no pun intended.

A dozen of auditors from customers, and auditing agencies have gone trough the system of migration testing but the period of 3-6 years have never been an issue.
They are rather pleased that we have a system in place that is systematically used.
Regarding calibration standards I had the same discussion with an auditor.
We had a length standard that was used on monthly basis to calibrate rulers.
The standard was calibrated yearly (standard practice for the calibration organization), and no strange matters were reported, the lab that carried out the test even asked me once if we were actually using the standard.
This yearly calibration became a pointless exercise bringing no added value and we changed it.
The length standard spends its day in the box stored in a climatised room. After every calibration with no deviations the calibration period is increased with one year, it is now once in five years. We do the same all other standards we have in-house.
We can show documented process variation for standards and an occasional auditor raised his eyebrows about the applied statistics but that is all.

Have a nice day, Okido

[Charles.C]

Dear Okido,

We had a length standard that was used on monthly basis to calibrate rulers.

Made of tungsten carbide perhaps ??

What did they calibrate it against ? (airfreight to France ??)

I think you got few comments because the auditor had never met it before or after

I like yr calibration frequency adjustment procedure but I doubt you can validate it ?. (I tried this myself in an attempt to extend from 1 to 2 years and was immediately asked for a reference, I gave up gracefully and used the audit to justify the cost).

Rgds / Charles.C

[Simon]

A dozen of auditors from customers, and auditing agencies have gone trough the system of migration testing but the period of 3-6 years have never been an issue.

I know, I know, it's fine. But I need to know why??? PLEEEEEAAAASSSSSE!!!

[okido]

Hi Charles,

Validation that an increased calibration period for a length standard has no adverse effects on food safety or other wise sounds very reasonable.
However, the length standard is in the box when it returns from the calibration lab and it proves to be accurate.
It remains in the box ready to be used if we would have a new shipment of rulers or a new supplier. It is maybe used once every 4 years.
For me it is crystal clear that a bar of high grade steel with engraved with lines and numbers that we call a standard will not change its shape or size over time when it is stored in a box in a climatized room. I could consider oxidation but the steel is rather resistant for this.
An auditor who would challenge the increased calibration period based on a missing scientifically validation would clearly show to me that he is not adding much value to the audit on this point and I would not hesitate to let him know this.

I know, I know, it's fine. But I need to know why??? PLEEEEEAAAASSSSSE!!!

Regarding the 3-6 year period for migration testing I will delve into the archive to find the scientific justification for this Charles.

Have a nice day, Okido

[Charles.C]

Dear Okido,

I certainly agreed with the principle of yr comments so I tried to back them up a bit.
Actually in the case of standards (primary-secondary) / instruments / calibration intervals, I suppose these are metrological questions which can be a very complex area indeed. It seems the topic of selection of calibration intervals remains “open” (eg 3, 5.below). One “basic” (1.,para5) and one more detailed treatment (2.) which I looked at supported yr extending concept quite well (the latter also suggested an arbitrary criterion for the extension [see method 1] but was still silent on the amount of extension reasonable). On the other hand to illustrate how far you can go to reliably measure length try (4.).
A few of the above would probably rebut most auditors “standard practice” argument nonetheless I failed to find a numerically satisfactory answer / procedure. I hope you can do better.

1. http://www.emfdosime...son/Swanson.htm

2. http://www.dap.de/95doc/DAP-TM-08.pdf

3. http://store.ncsli.o..._Pract_P438.cfm

4. http://www.ktu.lt/do...s/Santrauka.pdf

5. http://focus.ti.com/...2b/szzk012b.pdf

@ Simon, In a similar way, I don’t know the answer to yr plea either. Good luck Okido (or anybody).

Rgds / Charles.C

[okido]

Hi Simon,

I am not able to find a numerically satisfactory answer / procedure yet in written form.
However after talking to some people I conclude that the procedure works like this:
Laboratories build up a database with the data of raw materials and finished products from different customers over time.
If you send your PET bottle for migration testing the labo verifies test results against the data already in their database and the regulations.
Over time they build up statistics about materials and usage in products and with some mathematical modelling you can do quit accurate predictions.

Have a nice day, Okido

[Simon]

Hi Simon,

I am not able to find a numerically satisfactory answer / procedure yet in written form.
However after talking to some people I conclude that the procedure works like this:
Laboratories build up a database with the data of raw materials and finished products from different customers over time.
If you send your PET bottle for migration testing the labo verifies test results against the data already in their database and the regulations.
Over time they build up statistics about materials and usage in products and with some mathematical modelling you can do quit accurate predictions.

Have a nice day, Okido

Thanks Okido.

[Aur0re]

Hi!

In France, DGCCRF accept a validity of the tests of 5 years if, since then, no modification of the composition of the packaging is made. (note n°2004-64)

In fact all my customers told me to realise one test of migration every day (at least) and it’s what I do.

Aur0re

[Simon]

In fact all my customers told me to realise one test of migration every day (at least) and it’s what I do.

Aur0re

Are you serious Aur0re, you do this every day? If it's true can you explain your test method please?

Regards,
Simon

[Aur0re]

Oh no! I'm sorry. It's a mistake, I carry out this test every year! It was one of my customer who would like to carry out this test on all production!

Sorry for this mistake

[Simon]

Oh no! I'm sorry. It's a mistake, I carry out this test every year! It was one of my customer who would like to carry out this test on all production!

Sorry for this mistake

I thought daily sounded a bit much, it would be a full time (and very dull) job for somebody. I still cannot get to the bottom of why a customer requests testing to be carried out once every year. It just sounds nice and neat.

Forgive them Father for they know not what they do.

Regards,
Simon