Start-up procedure and monitoring for metal detector

We are in the process of installing a metal detector. Can anyone share with me any imformation on start-up procedure and monitoring activities. we process poultry [raw chicken product to be adequetly cooked by the consumer].

Hi Joelchenda,

We do a check on start up and then every hour after that.

You need to get test pieces - ferrous, non ferrous and stainless. The size of the test pieces will be determined by how sensitive you are setting your metal detector - your supplier should be able to advise you on this.

A good start up test would see all pieces run through (with product) and all detected and the same every hour. Make sure you document, write up a SOP and have  a check sheet that gets signed by the operator etc. Your metal detector count should be a minimum of 3 per hour (all test pieces). You will need to set corrective actions for higher counts of course. And best practice would see rejected product be discarded into a lockable cabinet.

Dear joelchenda,

And make sure yr test pieces come with a certificate stating an appropriate traceability source.

The size of the test pieces may also relate to yr customer's (minimum) requirements, if any, or legislatory issues.

Rgds / Charles.C

 

 

The size of the test pieces may also relate to yr customer's (minimum) requirements, if any, or legislatory issues.

 

 

Aperture size and product type will also play a large role in determining this.

Here is a pretty good rundown on the basics for the OP;

 MD Basics.pdf   70.78KB   639 downloads

[edit]

I thought it might help to add this from the FDA fisheries guidance;

 Chapter 20 Metal Inclusion.pdf   1.95MB   407 downloads

Aperture size and product type will also play a large role in determining this.
 
Here is a pretty good rundown on the basics for the OP;
 
MD Basics.pdf

 
Hi Joelchenda,
 
In addition to what was already said, it is important to review your customers COP's - they may require you to carry out a Failsafe checks, memory test etc.
Basing on this as well as the standard business practice an your certification you have to prepare a MD procedure and clarify the frequency of the tests, responsibilities and corrective actions (what to do if the test fails). On the top of the general procedure it is a good practice to have a more specific document issued to the shop floor - work instruction or standard operating procedure. The more detailed (descriptions, pictures) this is the better for you. Make it simple, logical, clear-cut and easy to follow. Remember to verify it - once the document is ready, take it with you on the line/s and do what is says. Whilst this is an obvious point, very often it gets forgotten.
 
An important point is also to carry out a comprehensive training to CCP operators - include videos etc. Operators must exactly know what are the corrective actions in the event of the standard (consecutive) or memory test fail. You also have to clarify what actions should be taken in an event of a good pack being rejected on the machine.
This means you have to place those actions somewhere in your NCR/CA/Hold procedure.
 
You will need to get your engineer involved too.
 
The monitoring activity will include recording every test (i.e.: start of every product run, after break, machine breakdown, every hour and at the end of each product run) by the trained operator, signed records. It is advisable to include simple table to record the actions for any fails/rejected good pots.
Records must be kept for reference as part of your due diligence program.
 
Remember to adjust your HACCP plan accordingly :)
 
All of the above will obviously need adjusting to your process but I hope this will help you in some way - I also attach White Paper re: MD from Mettler Toledo.
 
Good luck!

How do I attach files?

(feeling silly..)

Before replying you need to open up the "more reply options" editor.

Should be self explanatory from there.

 Untitled.png   124.65KB   9 downloads

 

How do I attach files?

 

 

 

(feeling silly..)

Dear qatfs,

No need to feel silly, it's not entirely obvious as various other people have also found out.

If remains unclear, can see this post -

http://www.ifsqn.com...udy/#entry70723

And thks for yr preceding post which contains some excellent tips. In case of unfamiliarity in Zambia, I guess NCR = Non- Conformance Result or similar.

And as Slab implied, the implemented test pieces may well be a (forced) compromise between intrinsic machine capabilities (and specific product matrix/process) and external business requirements. Tables of typical MD design sensitivities/aperture/product matrices are available on this forum. Price is also a factor when one examines detailed brand data.

Rgds / Charles.C

First, check to see if any of your customers require any specific monitoring activities.

For start up you definitely want to test your product with each of your three test wands/cards at least once per card. I would recommend running each wand/card through three times to verify rejection. After your start up is complete, verifying each test wand/card once per hour and at the end of the run is advisable. 

Good luck!

Regards,

Greg

Hi there,

As promised (although a bit delayed) - White Paper attached - maybe it will help someone.

@Charles C. - thank you for help in taking my first wobbling steps at the forum :)

Hi there,

 

As promised (although a bit delayed) - White Paper attached - maybe it will help someone.

 

@Charles C. - thank you for help in taking my first wobbling steps at the forum :)

 Metal Detector Validation.pdf   1.14MB   532 downloads

If you purchased the metal detector new you should work with the manufacturer and incorporate the manufacturers procedure(s) into yours. 

I worked in a place where they tested the metal detector every hour when they had just started up then moved them out to every 4 hours later after performing a risk analysis.  Also you need to determine if it will be a CCP and if so what will the critical limits be.  It has to be added into your HACCP plan, if you have one, and on the block diagram for it.  Just a few things to check to make sure you remembered.

Also the way you put the test piece through should be dictated by the manufacturer in accordance with the way they have determined a piece of test equipment should run through it.  If you have a reject bar though you will need to test it with a typical product piece so that you can be sure that the product is being kicked off properly by the arm.

Just an add-on to previous post-

The in-process frequency  of implementing wand checks, possibly regardless of haccp etc, may well be determined by the subsequent corrective action in the event of a checking failure. The typical  CA involves a segregation/recheck of all product passed through MD since last "good" check. This can be a very large number if infrequent checking/substantial output.

Rgds / Charles.C

While some very good points have been made about test sizes, customer requirements, aperture etc., the manufacturer needs to validate the machine at startup and they can help you to determine what the smallest sizes you can test for with your product and have the fewest rejects or where the machine is most stable.