Dear Sean Archer,
Thks yr reply. I appreciate yr thoughts.
You may understand my personal interest if you have a look at my recently uploaded (excel) yoghurt hazard analysis and refer my CCP1.
(added - see http://www.ifsqn.com...dpost__p__50651 )
As you noted, the situation partly revolves around the expanding “definition” of a PRP. However, in my own example, seems to me the control measure (CM ) is well-matched to para 7.2.3 (f) in ISO 22000.
1. The control measure described in PAS220 must provide measurable results not results subject to personal observations.
2. The control measure described in PAS220 must be applicable within a considerablly short time
I think "measurable" results does not exclude personal (eg visual) observations (see 7.6.3). The "timely manner ...immediate corrections" of 7.4.4 was from memory sort of compromised to "validatably feasible" after the standard(s) was launched. No argument that continuous monitoring is optimum.
However i agree that, on a case-by-case basis, lack of items 1,2 may create an absence of readily monitorable (or suggestable) critical limits which is a reasonable stipulation for a
CCP. The lack may also relate to the feasibility of validation, a necessity for both OPRP/CCP
As it happens both critical limits and validation should be possible in my example.
I am unsure if an OPRP is forbidden to
have ( pseudo-critical?) limits? . I think possible although the converse situation was a major motivation for the original introduction of OPRP terminology.
It is also certainly possible to find published examples of OPRPs which are seemingly not PRPs at all, eg specific, localised, process activities acting equivalently to CMs.
I am inclined to the conclusion that def.3.9 is simply inaccurate as written, ie an OPRP does not need to be directly "traceable" to a PRP as implied by 3.9, but can be totally deduced,
or not, from para. 7.4.4.
Above was sort of logic i used in yoghurt example but it would be nice to be able to justify it (or not).
Rgds / Charles.C
PS - I appreciate yr point re flexibility of ISO 22000 but this is what has (partly) contributed to the GFSI problem, hence FSSC. Not really ISO’s fault I suppose, just business driven.
Not sure if PAS220’s additional menu now de facto ISO 22000 "approved" by virtue of later issues in the 22000 series ? (unseen by myself, just speculating).(maybe it depends on 22003).
PPS - Apologies to FSSM, this is all totally OT of course.