8 replies to this topic

[Weebus90]

HI Everyone, 

I am preparing for an upcoming audit and am questioning the calibration of metal detectors. 

We currently have three metal detectors and perform verification checks on them hourly.

I am new in this company and no one seems to know when the last time they were calibrated was. 

 

Is it required to have an outside company calibrate these machines or can in-house personnel do this and have it be accepted?

 

Thanks for the help!

[Douglas NIcoll]

We used a manufacture's tech to calibrate the machine. The report would show the settings to get the most sensitive reading that the machine is capable of. - this is more of a validation step and Preventative Maintenance step. Using the metal bead wands or cards at the sizes your (or your customer's) specifications require is verification that the unit is capable of detecting that size. We put the machines on a yearly check with the service reps. It is when you experience degradation of performance and need to increase the sample bead size for a successful detection that you need to get the machines serviced and recalibrated.

[Snookie]

Most auditors and customers expect that your machines are re-certified annually.

[Setanta]

I second Snookie's comment. You will want an outside source (unbiased in the results) to certify your MD are working.

[Charles.C]

Dear All,

 

Based on my experiences, I would question that auditor's expect Preventive Maintenance for metal detectors ?.

 

Calibration of the standards used to "check" the MD is another equally interesting aspect?

 

Rgds / Charles.C

[George @ Safefood 360°]

It is generally accepted that you cannot 'calibrate' the metal detection head. You can simply adjust and confirm the sensitivity of the head under standard or known conditions. The check wands Charles refers to should be certified at purchase and assuming their integrity has not been compromised (i.e. they are not damaged) this should be fine for an audit. The safest route is simply have the supplier of the MD unit to conduct the service and check the sensitivity and give you the relevant report.

 

George

[mgourley]

BRC requires the following: (edited to show relevance to the current thread)

 

The company shall establish and implement documented procedures for the 

operation and testing of the metal or X-ray equipment. This shall include as a 

minimum: 

 

• the operating effectiveness and sensitivity of the equipment and any 

variation to this for particular products 

 

I would think that only an outside agency would be able to provide the "operating effectiveness and sensitivity" part of that requirement.

 

That information may have come with the documentation for the metal detector. But, if you are like many of us, what hole that documentation fell into is not known. It's probably best to have someone come in on an annual basis and provide a report that states "these are the capabilities of this particular metal detector. The metal detectors were tested and are operating within their capabilities."

 

Marshall

[Charles.C]

Dear mgourley,

 

I agree with the quantitative aspects of yr comments, and the reality check. The intrinsic problem may be additionally with BRC semantics ?.

 

Personally I have yet to find any “guaranteed” machine sensitivity (in the sense of a meter reading) for purchased MDs, or for the "calibration" check wands for that matter (equally [negligently?] I hv never asked for such specifics either ). My only agreed purchase "sensitivity" specification was for the unit to "visibly" detect/reject a certain "Xmm" wand, information as to which is typically given in brochures. People with better (non-anecdotal) knowledge than myself regarding what may be available are only too welcome to add it to this thread.

 

Previous threads hv discussed these issues and seem to mainly indicate absence of “hard” quantitative data ( choosing representative / reproducible  test conditions is presumably one tricky factor). Typical magnetic strengths (gauss readings)  have been mentioned as one possible criterion. For most people this would probably would require external assistance.

 

IMEX, external service technicians are also reticent over the validation of their own numerical MD “calibration” procedures also (eg “supplied” by the manufacturer). The recipient (beleaguered QA Mngr) of the conforming piece of paper is usually unconcerned about such niceties unless they are considering to take-over the job themselves.

 

I did note that no specific mention of “external”  or "calibration" appears in the BRC, MD-specific text 4.10.3 (although the latter word is an implied requirement as per sec.6.3).

 

Sadly, I would also add that for some, less-enlightened companies, money questions for external calibrations can also surface unless specific QA justifications are supplied. This is the “senior management commitment” issue again.

 

One obvious query is as to what BRC themselves (ie auditors) expect by the standard’s use of “sensitivity”. ?

My own auditorial experience  has been that “sensitivity”  is interpreted as a visible  +/- rejection action for the specified/certified "Xmm" check wand as per my original documenteded procedure ( any failure would clearly have routine CCP implications). Others may hv different experiences of course. Specific BRC auditorial non-acceptance of a visual criterion would certainly be interesting to hear about ?

 

Rgds / Charles.C

 

PS - again on semantics, i find the slight wording variations between bullets 2,3,4  of para 4.10.3.4 (and also when combined with para.4.10.3.5) rather suggestive/suspicious but only BRC (and the auditors?) know the whole story. So far.

 

PPS - as an example only, here is the Tesco (2009) viewpoint on "sensitivity" (eg para 3.2.1.1) -

 COP for Metal Detection&X ray Systems.doc   630KB   417 downloads

 

P3S - just noticed the OP did not refer to any particular standard but i guess  the posts are fairly generic, at least as far as sensitivity is concerned. SQF's specific requirements for a MD, if any, no idea. But may be relevant, and equally obscure .

[mgourley]

Charles,

 

On further reflection, I may rescind my comment about an outside agency.

 

Prior to my arrival at my current employer, they had set their Critical Limits on the metal detector at 3mm Fe and non-Fe and 4mm S/S. No one that I asked could accurately say how those limits were chosen. 

 

I ordered a set of test wands in .5mm sphere size increments from 2 to 4 mm for all three metals.

I then grouped all products on all three lines by size, density and formulation similarities. I then created a grid that listed each product variety, metal and size.

 

I then passed each product through the metal detector starting with the largest sphere test wand, stopping when a partcular size/type wand was not detected and rejected. 

 

What I found was that all products would detect and reject 3mm Fe, 3mm non-Fe and 4mm S/S. 

Some but not all products would detect and reject smaller wands.

 

From this, the validation of the Critical Limits was achieved, and I suppose one could say that I proved "the operating effectiveness and sensitivity of the equipment and any variation to this for particular products."

 

So, as long as my experiment were codified in a "procedure", I would think that it would meet the intent of the Standard.

 

Marshall

Edited by mgourley, 24 November 2013 - 03:18 PM.