You need to do a risk assessment, eg -
(1) Is there a significant hazard at the optical sorter step ?
(2) If yes, can you validate that the sorter "eliminates" the hazard or reduces it to an acceptable level ?
If the answers are YY, the Codex decision tree will support yr designating it as a CCP. (the critical limits may be based on "best practice" data from a manufacturer since there must surely be some kind of specification as to the efficiency/capability of removal of certain particle sizes ?)
Presumably if the answer to (2) is No, it is not a CCP and if the answer to (1) is yes, you need to re-design the flow, eg use MD/X-Ray.
Some users may set it as a PRP or other entities. This choice depends on yr own particular flavour of haccp methodology.
PS - from memory, at yr location, metallic contaminants >2mm are considered a significant hazard
Thanks for the fast reply!
We have done a risk assessment:
1) Is there a significant hazard at the optical sorter step?
Yes there is: The product come from the field and can be contaminated with small pieces of glass, stone, metal (covered by the metal detector) plant material (no food safety risk)
2) can you validate that the sorter "eliminates" the hazard or reduces it to an acceptable level?
We now see that when we used the optical sorter (almost) nothing is found by the x-ray so we can conclude that the optical sorter reduces the hazard to an acceptable level (99.9 % purity)
So this indeed means that we need to make a CCP and depending on the product/quality it will go through the x-ray or optical sorter.
But then the critical limits will be a problem, because depending on the product the settings are different.
There is big difference between sesame seeds and pumpkin seeds in size and colour, so if we need to test if the object will be shot out, there need to be different objects for all programms
How would other users set it as a PRP? There is a risk for the consumer so you will always need to check if its a CCP or not.