10 replies to this topic

[classic]

We have certificates of confomity for the test pieces used to verify the operation of our metal detectors.  At a past audit the auditor said we should have certificates of calibration rather than conformity.  On the certificates of conformity for the test pieces reference is made to a certificate. 

 

Please can anyone advise if we should be asking and retaining the actual certificates at the time of purchase or is the confomity sufficient and if we do need calibration certicates how is the best way of getting then for test pieces we have had in use for some time.  I wouldn't wish to get an NC at the next audit.  We are operating to BRC v6

 

Any help would be appreciated.  Thanks

[Slab]

I'm unsure about BRC requisites, but it is always the best practice to have "metrological traceability" for any standards, whether it be thermometers, scale weights, salinometers, etc.  The problem with "conformity" is that they are of usually dubious quality lacking traceability.  The main disadvantage to COT is the added cost, which usually marks up the price by 300%.  If it's non-regulatory, either by governmental bodies or customer requirements, then it's a benefit you will have to decide upon. From the wording by your auditor it sounds as if its a recommendation rather than a requirement. 

Maybe this section of BRC standards will be of some help;
 

6.3 Calibration and control of measuring and 

monitoring devices

The company shall be able to demonstrate that measuring and monitoring equipment 

is sufficiently accurate and reliable to provide confidence in measurement results.

 

Clause

Requirements

6.3.1

 

The company shall identify and control measuring equipment used to monitor CCPs, 

product safety and legality.

This shall include as a minimum:

..a documented list of equipment and its location

..an identification code and calibration due date

..prevention from adjustment by unauthorised staff

..protection from damage, deterioration or misuse.

 

6.3.2

All identified measuring devices, including new equipment, shall be checked and where 

necessary adjusted:

 

..at a predetermined frequency, based on risk assessment

..to a defined method traceable to a recognised national or international Standard 

where possible.

 

Results shall be documented. Equipment shall be readable and be of a suitable accuracy for 

the measurements it is required to perform.

 

6.3.3

Reference measuring equipment shall be calibrated and traceable to a recognised national or 

international Standard and records maintained

 

6.3.4

Procedures shall be in place to record actions to be taken when the prescribed measuring 

and monitoring devices are found not to be operating within specified limits. Where the 

safety or legality of products is based on equipment found to be inaccurate, action shall to 

be taken to ensure at-risk product is not offered for sale.

[Charles.C]

Dear Classic,

 

It might hv helped if you post a copy of yr COCo. ?

 

I guess it depends on what auditorial detail of information (eg as prev. post) is expected on the document stated to be a "certificate of conformity" or  "calibration certificate".

I note there is no reference to a mandatory  "certificate" in above text or involvement of external bodies ?

 

I have no idea regarding any defined statistical difference between COCa / COCo but some COCo's appear to have quite complete-looking data embodied within them, eg -

 

 sample - certificate of conformity.pdf   204.85KB   161 downloads

 

(i see the magic word "uncertainty" present plus NIST. However the "intermediary standards" comment might disturb a typical auditor although may in actuality be "best practice")

(or is the auditorial problem that the COCo is typically provided by an instrument's own manufacturer, ie not strictly independent ?)

 

It's also possible IMO that yr auditor simply does not know what is the minimum statistical data to "qualify" so is more "confident" to accept  a document stated to be a "certificate of calibration".

 

I suggest that the BRC words "risk assessment" in 6.3.2 are statistically meaningless and can be ignored provided that some kind of relevant reference is available as supporting evidence for actual practice.

 

I doubt that many holders of COCas can readily explain the full meaning of reported "uncertainty" percentages on their own certificate (including myself).

 

Regardless, if the financial difference is not extravagant, seems easiest to get a, titled, COCa.

 

Rgds / Charles.C

[classic]

I've posted 2 certificates of conformity which I have on file for metal detection test pieces.  We have some even older ones in use which we purchased some time ago and can't find anything other then the actual pieces have a ref number on them.

 

Regards

 

Classic

Attached Files

•  C2.pdf   251.39KB   144 downloads
•  C1.pdf   318.37KB   100 downloads

[cazyncymru]

Ask them for this kind of thing

 

Caz x

Attached Files

•  107128B.pdf   1.44MB   173 downloads

[chrisbird616]

Not sure if this is helpful or applicable, but our company's practice is to have the test pieces calibrated at the same time as the metal detector units. The external calibration contractor then just includes reference to the test pieces on the calibration certificate (i.e. we receive and keep on file a single calibration certificate for a metal detector units and the three associated test pieces).

 

This has always been acceptable to auditors, including this year's BRC Food v6.

[Charles.C]

Dear Classic,

 

Thks for uploads.

 

IMO there is auditor confusion occurring over the interpretation of words like conformity, calibration as applied in different situations. Or simply nitpicking going on.

 

A detailed look at the limitations of the possible titles and the semantic confusions over COCo/COCa (and others)  is here –

 

 Manufacturing Calibration Certificates.pdf   106.46KB   110 downloads

 

eg, A comment on COCo –

 

So, what is a ‘Certificate of Conformance’? Typically available when an instrument is purchased, it is now generally recognized that such a document has little value as an assurance of product performance. Of course, the manufacturer expects that the product conforms to its specification but, in this sense, the document simply affirms that the stipulations of the customer’s purchase order or contract has been duly fulfilled.

 

 

Additionally, as a  partial definition, - “calibration is simply the process of comparing with a known standard and reporting the results.”

 

So what data (and standard?) is required to be reported after a hypothetical  annual “calibration” for a MD test piece ?

 

The answer is apparently to assure compliance with  respect to BS EN ISO 9002, UKAS, AFBMA calibration requirements, eg  -

http://www.metaldete...est_pieces.html

http://www.completes...est-pieces-p789

http://www.detectame...test-pieces.php

 

No explicit details were visible in above links. ? I hv no idea what the UKAS requirement means, AFBMA probably refers to steel ball size tolerances, we all know what iso9002 is.

Perhaps, in view of its intended usage,  the test-piece certificate should contain something in respect to a standard MD’s response characteristics to the test-piece ?  But in practice, none seem to do so.

 

The detail on Caz’s example is certainly more visually impressive than yr "certificates" (which do hv some slightly odd statements / ambiguities) together with  references to iso 3290, iso9001. iso3290  appears to be another standard for dimensional  accuracies of steel balls. Hmmm.

 

In comparison to all the above, the calibration of a reference checkweight has an obvious practical expectation other than compositional data.

 

All very confusing (to me) but i think you should hv challenged the auditor as to why, for example, yr C2 item was inadequate.

 

Rgds / Charles.C

 

PS (added later) –  It's an interesting topic so i expanded a little.

 

Just for info., here is a summary of a generic list ( ISO17025) regarding info. to be given in a Calibration Certificate (example attached at end post) -

 

5.10.2 TEST AND  CALIBRATION  CERTIFICATES

• Test Report/ Calibration Certificate shall contain

1. Title (Test Report/ Calibration Certificate)
2. Name & address of the laboratory
3. Unique identification  (like serial number)
4. Name & address of the Customer
5. Identification of Test Method
6. Condition & identification of Test/ Calibration item
7. Date of receipt & date of test/ calibration
8. Reference to sampling plan, if any
9. Test/ calibration results
10. Name & designation of persons authorizing report
11. A statement to the effect that the results relate to only the items tested/ calibrated

 

5.10.4  CALIBRATION CERTIFICATE

5.10.4.1  In addition to the requirements listed in 5.10.2, calibration certificates shall include the following,
                 where necessary for the interpretation of calibration results:

a)     Environmental conditions, if applicable

b)    Uncertainty of measurement

c)     Evidence measurement traceability

5.10.4.2  Statement of compliance shall include  which parameters of specification are met /

               not met

5.10.4.3  Any adjustments/ repairs carried out shall be reported

5.10.4.4  No recommendation on calibration interval

 

 

As far as I can see, the BRC standard makes no explicit requirement regarding  onward certification / calibration of  the metal detector calibration standards (test-pieces) although it could be interpreted as expected within para. 4.10.3.4. However posted experiences here suggest that such a  possibility is not implemented, ie auditors appears to generally accept COCos. The standard does specify  shape / materials.

 

I have no particular expertise regarding MD test-pieces but based on previous posts, can suggest -

 

The minimum requirements for a Manufacturer’s Specification for the test-pieces  is data regarding the size, shape and material plus appropriate  assertions as to the control / traceability of the manufacturing history / process (eg involving ISO9001, traceable to National Standards).

No specific performance criteria / data seem to be required.

 

Regarding uploads, the C1/C2, COCos are basically asserting that the item (ie a ball) complies with the manufacturer’s specification by virtue of the data supplied  and/or cross-referred to. The level of assertion is defined by the specific text.

 

Examining uploads –

C1

The COCo gives values for size and material for the “ball”.

The specification states that  ball is manufactured to AFBMA standard (AFAIK this may cover size and composition, [see sample attached below]) (evidence in referred certificate).

The COCo does not claim the item is traceable to National Standards.

The COCo does not  identify the material in “detail” (may be in referred certificate).

The COCo has a rather obscure introductory text, especially following "may" (probably due to a generic form)

The cross-referenced certificates are apparently not provided.

 

 AFBMA standard for metal balls.pdf   1.45MB   68 downloads

 

C2

The COCo gives values for size and  material for the “ball”.

The specification states that  ball is manufactured to AFBMA standard (AFAIK this may cover size and composition, [see sample attached below]) (evidence in referred certificate).

The COCo does claim that the items are traceable to National Standards. (evidence??)

The COCo does not  identify the material in “detail” (may be in referred certificate).

The cross-referenced certificates are apparently not provided.

 

"Con" Comments

 

Both COCos apparently lack the cross-referenced certificates.

Upload C1 has one significant omission (see underlined).

 

Whether auditor dissatisfaction was concerning lack of details, cross-certificates, age of COCo (?) only the auditor knows, so far.

Regardless, IMO / IMEX the COCo “method” should in principle be acceptable.

 

Note: if  “Calibration Certificate” is envisaged

 

I predict some test-piece manufacturers happily use this title although content is not dissimilar to typical COCo.  A question of interpretation. Whether will be more auditorially acceptable is another matter.

Strictly a Certificate of Calibration (eg as per ISO17025) requires inclusion of numerous items (see top list). Any COCa  surely requires some statistical data further to typical COCo content evidencing that a calibration procedure has been performed for appropriate parameters, eg some uncertainty results.

 

A detailed example here –  Sample calibration certificate for checkweights as per ISO17025.pdf   51.19KB   44 downloads

Edited by Charles.C, 11 November 2013 - 05:47 PM.
revised PS

[George @ Safefood 360°]

I think this is very simple. Metal detection check 'wands' or pieces usually state the thickness of the shim in the plastic moulded casing. If it states for example 2.2mm S.S. then you need a certificate which shows this was measured checked and confirm to be the thickness against a certified standard. The Purchase document refers to the certificate for each piece so I would simply request a copy of these certificates and hold them on file. It is not possible to 'calibrate' a shim of metal moulded into plastic unless you are Harry Houdini.

 

George 

Edited by George Howlett, 12 November 2013 - 01:35 PM.

[Charles.C]

I think this is very simple. Metal detection check 'wands' or pieces usually state the thickness of the shim in the plastic moulded casing. If it states for example 2.2mm S.S. then you need a certificate which shows this was measured checked and confirm to be the thickness against a certified standard. The Purchase document refers to the certificate for each piece so I would simply request a copy of these certificates and hold them on file. It is not possible to 'calibrate' a shim of metal moulded into plastic unless you are Harry Houdini.

 

George 

Dear George,

 

Not too sure about using the  word “simple” in any calibration context.

 

Admittedly it may be simple from most auditor’s POVs (ie an appropriate piece of paper) although the current OP suggests that maybe the paper’s desired contents  may also have some bearing (pun intended).

 

Also unsure about yr reference to the thickness of a  “shim”. All the MD  stick / card  products I could locate seem to be encased balls (unless some cunning semantics is in use) ? (see the links below)

Perhaps you have a link to some “shim” types, or hv used them ?

 

So far have only seen certificate references to official standards (material / dimensions) for balls though thickness measurement is obviously referenceable to some standard if desired. And geometry ?

 

I am sure that the MD response may (will?) be different in respect to shape (and encasement?)  but this (strangely IMO) seems to not be considered in relationship to MD detection test standards. From memory, I only saw one supplier / brand who mentioned that an “adequate” MD response was part of the design specification. No details given unfortunately.

 

http://www.loma.com/lo_ts_metal.shtml

http://www.teststandard.com/

http://www.testick.com/

http://www.test.ie/C..._detectors.html

 mettler-toledo - test_samples, certificate of conformity.pdf   228.79KB   80 downloads

 

Rgds / Charles

[Evelyn O Brien]

i

[Charles.C]

Hi Evelyn,

 

Welcome to the Forum !

 

Looks like you clicked the wrong button

 

No problem to try again.