32 replies to this topic

[bergman]

Here's some background regarding my predicament:

We have HACCP plans that have been assessed by a third-party auditing firm (Silliker) and have passed their evaluation without fault. We are a dry food blending and packaging plant, with our major products being instant nonfat dry milk and bakery mixes. Through our hazard analyses, the only CCP we determined for these products is metal detection (our ingredients and processes are sufficiently "low risk"). Metal detection is achieved with in-line metal detectors after the final packaging, before the product is loaded into shipping cases. We have all the standard controls in place (documented checks every 2 hours with certified test pieces, corrective action procedures in place, training, etc etc etc.). In addition to these major products, we co-package a range of sugar subsitutes for a very small company and a range of dietary supplements for a local small company. These both can be considered dry food products in the same vain as our INDM and bakery mixes.

My problem is this - we don't have metal detection equipped on the lines the co-packaged products are run on, nor has it ever been a requirement or even a request for these 2 companies to implement such measures. Due to this fact, we have never invested the money into purchasing metal detectors for these lines. As such, I'm having trouble designing HACCP plans for these products without having metal detection involved. I can't reasonably justify that metal will not be a hazard in these products as there is just as much chance that metal fragments could be a problem with these as it would be for our 2 core products. In addition, the dietary supplements have an oxygen scavenger placed at the bottom of each bottle (per customer request) - these oxygen scavengers contain iron (they're also contained in a metallized film), which would obviously not get past a metal detector. The solution to that is a gravity fed or "throat" metal detector obviously, but that's out of the question at this time since we're not looking to invest into metal detection (the revenue generated by these products doesn't justify the investment).

So, my question is this - how do I justify NOT having metal detection for these 2 products? I can't say that metal won't be a hazard likely to occur.. And I doubt I could just say that the customer "didn't specifically request metal detection, so we don't have it".. I have everything in order from a HACCP standpoint for these products (hazard analyses, flow charts, etc.), but no metal detection to combat the hazard.. Am I hopeless unless we have a metal detector for these products? Please help!

[Larry007]

Bergman,

This is an interesting question. You stated that you have passed Silliker audits. Have they evaluated the production lines for these products? What was their conclusion?

One of the ways you might consider justifying not having a CCP would be based on historical data, although by itself it's pretty flimsy. But if you could document that
a) the product is metal detected at your supplier, prior to the problematic addition of the oxygen scavangers.
b) during transport/storage packaging integrity is maintained to prevent contamination. (this would be done by inbound checks and self audits)
c) the filling equipment is PM'd regularly and has NEVER had an incident of depositing metal into the product/container.
d) to date, there have been no customer complaints, regarding metal.

e) Optional - but recomended would be screens or magnets in place.

HACCP is suposed to be an objective risk analysis... in theory. In practice, it is a bit more subejective.

If it were my company, and I could document the above, I would be comfortable defending no CCP. At least, until the status of the points above changed. i.e. supplier audit showed their metal detection inadequate, my equipment threw off shavings, bolts, nuts, etc.. You get the idea.

If that were to happen, you would have to decide to metal detect or discontinue production, since there would be no way to "significantly reduce the hazard".

Hope this helps,

Larry

[Charles.C]

Dear Larry007,

Very nice comments IMO.

@Bergman - regret not seeing yr post first time around.

Might add that for point (a) a suitably relevant (eg HACCP-based) copy of the supplier's own certificated status might add a bit more strength if existent.
I recently posted an official US viewpoint on activities which might/might not be reasonably expected to act as potential metal contamination sources in seafood processing. You might be lucky enough to find something similar for yr product, ie supportive of "not likely". (you could perhaps justify yr own production as a customer driven scenario although the risk analysis will obviously still require some creative efforts, eg a 5x5 matrix with the risks on a borderline significant level ). Depends how picky the auditors are.

Rgds / Charles.C

[Larry007]

@Charles C. Thank you.

Your comment regarding supplier certification against a HACCP based planis excellent. I guess i assumed that.

@ Bergman

One of the key i have found for managing our system is to focus as much resonsibility as i can to the point of origin i.e the suppliers. The more they are in line the less i have to do internally. This of course requires i monitor them regularly but it is a system that works for us.

[bergman]

That is a great point - put the pressure on your suppliers. This makes total sense, and I'm glad you brought it to my attention. It's funny, b/c one of our suppliers was able to get away with no metal detection during a HACCP audit b/c they said that the next step in the supply chain (i.e. US) was equipped with it. If it worked for them, it should hopefully work for us (auditor dependent).

We have HACCP certificates/audit reports for all of our suppliers (even if the supplies are "customer owned"), most of which have flow charts of their process (which invariably include metal detection). I will look through them and make sure that they have metal detection - I'm fairly certain the ones in question do. I believe the possibility for metal inclusion in transit, storage, or from our equipment I can justify we control through the receipt inspection process, storage SOPs, and our PMs, all of which are pretty sound.

Thanks very much guys for the responses - I think you may have saved me a lot of trouble!

Cheers,

Brad

[GMO]

I think if I'm honest you're thinking about this the wrong way. You're thinking about passing audits and making the HACCP plan but you're not thinking about 'is my food safe?'

Presumably as you process this in your site you would be the company responsible if something was found. You've also stated a ferrous hazard on your line which obviously cannot be controlled by your supplier because it's on your site.

I would start again and I would suggest that the only way you could make this work would be a combination of supplier and internal controls as you would perhaps for glass. The only problem is that as metal detectors are reasonably cost effective nowadays, you would be in hot water if you went to court after a metal find IMO.

Have you considered a second hand ferrous in foil detector? Sure, it's not perfect as it won't detect stainless steel or non ferrous but it would work on your specific iron hazard in metalised film and it shouldn't be a big outlay. You don't always have to buy new.

[Dr Ajay Shah]

I totally agree with the valid points made by GMO. You know the saying "You are guilty until proved innocent"


Cheers

[bergman]

I think if I'm honest you're thinking about this the wrong way. You're thinking about passing audits and making the HACCP plan but you're not thinking about 'is my food safe?'

Have you considered a second hand ferrous in foil detector? Sure, it's not perfect as it won't detect stainless steel or non ferrous but it would work on your specific iron hazard in metalised film and it shouldn't be a big outlay.

Yes, I realize the thought process is not tailored toward making the food safe, but there are several issues that contribute to this. I would much rather prefer to install a gravity fed MD, but we are limited by our processing equipment layout. A gravity fed MD (even a so called "throat" MD, which has the smallest footprint and necessary metal free zone) will not fit in the space we have to work with. Creating the space we need will take a considerable capital investment. These 2 particular products contribute ~ 2% of our total revenue (seriously), so no one wants to spend the kind of money necessary just to be able to put in a gravity MD for these 2 products. I have considered ferrous in foil detection, but considering all of our equipment is S/S (and all of our suppliers use metal detection), looking at only ferrous doesn't help reduce the metal hazard any appreciable amount, IMHO. X-ray is also an option, and I have been quoted around $40,000 USD for a brand new one - again, upper management does not want to shell out that kind of money for these products. With all of these things considered, that leaves our best option (to my knowledge) as installing magnet traps. This is probably the worst of all options from a food safety standpoint, but it is about all I'll be able to justify (and all upper mgmt will approve..)..

Thanks for the responses. I hear exactly what your saying and totally agree.. I am just stuck between a rock and hard place. I know what should be done, but I can't achieve that. SO, I have to go with the next best thing..

[GMO]

...and the next best thing is delisting the product IMO.

[mgourley]

A hazard analysis based upon likelihood of occurrence, severity and plant history is all you need to determine whether or not a metal detector is in fact, a CCP.
If you have determined that the likelihood is negligible, the severity is low and you have not had any customer complaints of metal in the final product, you don't need a metal detector.
Look up the FDA 7/25 mm rule regarding metal inclusion.

I may get a lot of flack for that position, but if you want to do HACCP scientifically and systematically, (which by definition it's supposed to be) rather than regulatory or political, the argument is valid.

[KTD]

A concern I have also addressed in the past is that metal detectors are often only as accurate as the environment and products involved. Mixed frozen and non-frozen components, raw and cooked, operation in a cold environment, etc. Under these variable conditions - often in the same finished product (muticomponent meal) - proper metal detection is difficult at best. Since CCPs are supposed to have near fair-safe monitoring systems, it is hard to justify MD as a CCP under these conditions.
Obviously, the process in this discussion would appear to hold good promise for acceptable MD operation...

[Tony-C]

I have everything in order from a HACCP standpoint for these products (hazard analyses, flow charts, etc.), but no metal detection to combat the hazard.. Am I hopeless unless we have a metal detector for these products? Please help!

Is sieving an option?

Regards,

Tony

[Foodworker]

Tony is right.

If you have a final, point of fill, sieve with a mesh size of maybe 1 - 2mm, in most cases a metal detector becomes superfluous. This would probably be an acceptable position for a BRC audit, I don't know about the other standards.

If your product has particles much bigger than this, it is difficult to use this as a justification for not having a metal detector.

The concept of metal detectors and HACCP plans has become blurred over the years.

Sometimes a company calls it a CCP because they already have the instrument and almost feel obliged to justify it in the risk analysis, even if the likelihood factor in their circumstances is in reality very low.

Conversely, because metal detectors are almost universal in food factories and customers expect to see them, you have to be able to justify not using them, which is reverse logic.

[bergman]

Sieiving is an option, but I'm dubious that we can get it down to 1 - 2 mm. I'd be concerned with the product bridging at a sieve size like that. I am positive that at least one of the products in question (a brown sugar substitute) would not be sievable.

When initially setting up our HACCP plans, I was of the opinion that metal detection was not a CCP.. there are numerous threads here on the subject as it seems to be somewhat debatable. Some people have had success in audits, etc. not maintaining it as a CCP, others are adamant that it needs to be a CCP in any HACCP plan. I can see it either way. Others on our HACCP team agreed that it does not necessarily need to be a CCP, but one individual was adamant that we must have at least one CCP to have a HACCP plan. We took him at his word and went with it. Now that I've seen several other suppliers that have metal detection, but don't consider it a CCP, I'm leaning back toward my initial opinion.

Metal detection is a great tool to have in place and should be employed by just about everyone if practicable (I wish it weren't such an ordeal to implement it on this processing line..), but as a CCP? I guess, as Foodworker mentioned, it all depends on the likliehood of the risk. In 4 years of packaging these products, we have never had a complaint regarding metal in the product (or any other issues for that matter), so I am fairly confident that the hazard is low risk. Couple that with the fact that our suppliers for these products all have metal detection, we have other controls in place (receipt inspections, proper storage practices, self-audits, etc.), and the production process is simple with not a lot of chance for metal shavings or parts to fall into product, and I think we have proper justification to process without it. Essentially we dump straight to a ribbon blender, blend into a super sack, hook up super sack into bulk unloader, and on to packaging on a small-format filler.. There are moving parts in blending obviously, and also in the unloading and filling; however, there really are no small parts that could fall off into the product unless there was a catastrophe (e.g. auger broke off), in which case, we would dump all the product in contact with the equipment at the time. Likewise, the moving parts do not contact any surfaces that would generate metal shavings. This is all 3A equipment.. not saying that makes it immune from damage, malfunction, and the unknown, but the food contact surfaces are designed to be in contact with as little removeable/breakable parts as possible.

One more thing - the processing lines where we do have metal detection and that see metal contaminants from time to time can be linked to one supplier. They do not employ metal detection, and they process with a lot of aluminum equipment/containers (very old plant). Every time we see metal from the MDs, it has always been aluminum and always has the same appearance. The supplier has confirmed that this is a result of their processing (I know, I know.. why are we using them still??? Limited options in the industry and upper management supports use of this supplier - price and "friendship" - that's why). They have gotten by on AIB audits by saying that we have metal detection, so the hazard is reduced downstream. When we run a different product that does not use the ingredient provided by this supplier, we don't see metal (or at least have not in at least 2 years.. keep in mind, we are fairly small).

In any case, I greatly appreciate all of the responses. I think that I have enough information/documentation to support my approach for these products, and I also love to hear people's opinions on metal detection (or anything food safety/quality related!). Our audit is coming up soon, so I will post the response of the auditor regarding this matter (Silliker audit). Thanks again!

Cheers,

Brad

[Foodworker]

An additional control may be in line magnets but I would guess that your blenders are all stainless steel which would not be taken out by them. It all helps however.

I am surprised that AIB accepted the supplier's justification that you had metal detectors. You cannot expect your customers to do your quality control.

[GMO]

Personally I would be very wary at using "we've had no complaints" as a justification. Think about it; justification is basically validation which is the scientific evidence that your process works and that's not what customer complaints are. Complaints are like auditing, they are a verification activity. They are a snapshot, a sample, not proof. Whether someone finds something and then complains is dependent not only on whether they find the issue (it might not be there, it might not be visible or they might not notice it) and what mood they're in (they've had a bad day, they're disorganised and lose the FB etc.) The problem with using this as your validation as well is as soon as you have 1 piece of metal returned to you; it's blown out of the water.

I agree the situation with metal detection is a bit about face; it's almost a CCP because the technology exists, however, even in all of the arguments about whether or not it's a CCP; most people would still have them on their lines (perhaps as an oPRP) unless they have something like sieving or X-rays.

To make having no foreign body detection equipment a possibility I would think a lot of auditing would have to be done. Think about glass. There is no (reliable) detection equipment for it so it gets excluded as far as possible from the production environment (we can't do that with metal), we have glass breakage procedures and audit what glass there is to death. Imagine having a metal breakage procedure every time a piece of metal goes missing (do you really never lose any nuts or bolts???)

I work for a tiny company where cost is a premium and a second hand detector does not cost a lot. Personally I wouldn't be happy without some FB detection on the line whether or not it's a CCP and for all my colleagues occasional cowboy ways, neither would they.

Edited by GMO, 01 July 2011 - 03:45 PM.

[bergman]

Believe me - when they told me that, I was just as surprised. I had never heard of passing QA/QC on to your customers as acceptable.

You're correct - our blenders are all S/S, but we will be installing magnets in the near future. Better to have at least something, instead of nothing at all. In the long run, I'm hoping that we will be able to reconfigure that particular processing line so that we can fit in a throat MD. Safeline has a nice option that only needs about 8" of vertical space to install.

[GMO]

Believe me - when they told me that, I was just as surprised. I had never heard of passing QA/QC on to your customers as acceptable.

You're correct - our blenders are all S/S, but we will be installing magnets in the near future. Better to have at least something, instead of nothing at all. In the long run, I'm hoping that we will be able to reconfigure that particular processing line so that we can fit in a throat MD. Safeline has a nice option that only needs about 8" of vertical space to install.

Do you put this product into outer cases? What about using a case sized metal detector?

[bergman]

GMO - can't say that we never lose any bolts, but in my experience with this organization (2 years), we have had one bolt come off on the packaging line (vertifcal form-fill-seal) that was picked up by the metal detector equipped on that line. Other than that, the only metal we have seen has always been aluminum from the one supplier I referenced above.

I see your point on FB detection, and agree that it should be employed - trying to get my colleagues to see this has, disappointingly, been a challenge for these particular products. In their minds, if the customer doesn't specifically request it, then we don't have to do it. This feeling has been deeply entrenched here for some time, and since coming on with this organization, changing this sentiment has been one of my greater challenges. I have been successful in some ways, but not on all fronts..

[GMO]

GMO - can't say that we never lose any bolts, but in my experience with this organization (2 years), we have had one bolt come off on the packaging line (vertifcal form-fill-seal) that was picked up by the metal detector equipped on that line. Other than that, the only metal we have seen has always been aluminum from the one supplier I referenced above.

I would argue on your statement above that metal detection is justified. On my risk assessment matrix, I would still class that as medium risk due to the impact ont he consumer had you not had the detection equipment in place...

[bergman]

I see your point, but we're talking about 2 totally different processing lines here. The one where the bolt was observed (last year, incidentally) is no where near the same as the equipment on the production line in question (that doesn't have a MD). The line where the bolt came off is a fully automated vertical form-fill-seal - there are many more contact points with moving and removeable/breakable parts on this line. The line in question, with no MD, is essentially a small-format filler - bags are practically hand-filled. Product is gravity fed into its small hopper from a bulk unloader, product is then gravity fed again from the small hopper onto a vibratory pan, from the pan product is vibrated into a filling tube which dispenses product upon pressing a foot pedal controlled by an operator. So, the operator is literally holding the package under the filling tube as product flows into the package. The other line, where we have MDs and the bolt was observed, operates nothing like this, so drawing conclusions from its history to apply to the small-format filler is comparing apples and oranges.

[Charles.C]

Dear GMO,

I agree the situation with metal detection is a bit about face; it's almost a CCP because the technology exists, however, even in all of the arguments about whether or not it's a CCP; most people would still have them on their lines (perhaps as an oPRP) unless they have something like sieving or X-rays.

As an OPRP ?? Why so ?? Why not simply make it a CP (ie low risk control measure) and reduce the number of CCPs / OPRPs ?

Unfortunately the MD in a final position (and sometimes upstream separators also) is the classic textbook HACCP / CCP. Not to mention a proud illustration of the Codex Tree. Somewhat ironically, Codex are also a major justification for some of the other alternatives.

Rgds / Charles.C

[Tony-C]

Dear GMO,

As an OPRP ?? Why so ?? Why not simply make it a CP (ie low risk control measure) and reduce the number of CCPs / OPRPs ?

Unfortunately the MD in a final position (and sometimes upstream separators also) is the classic textbook HACCP / CCP. Not to mention a proud illustration of the Codex Tree. Somewhat ironically, Codex are also a major justification for some of the other alternatives.

Rgds / Charles.C

If you decide the hazard is not significant let's say due to the likelihood being low then it is neither a CCP or OPRP

Many people will say having metal detection is the ideal scenario but being realistic and from Brad's post it seems that the risk of metal contamination is small and this is backed by historical evidence. Therefore if it was my business I would consider I had taken 'reasonable precautions' and be investing my money elsewhere.

[GMO]

Hmm, that's the thing. I think the poster thinks it is justified even though the risks are low but the bosses aren't listening... That's what really worries me.

[mgourley]

I think CCP's are supposed to be determined based upon Likelihood, Severity and Plant History.
If the likelihood is low, the severity is low and the plant history is negative on that line, then it's not a CCP.

Marshall