967   Statistical calculation and "M" attributes

Created: 03 Jan 2013

Status: In Force (green)

Part: Part 7-4 (2010; Edition 2)

Links:

Page: 167

Clause: Annex F

Paragraph: 5th bullet

Category: Issue for edition 2 of this part

Issue

Annex F says: "The content of a statistical LN can't include object/attribute which were not present in the original one."

This means that when I am interested only on an average of the original data and want to have an alarm from this averaged value (only) I need to model the limit supervision also for the original momentary data (?)

I discussed shortly about this with Wolfgang and his comment was: " formally your interpretation is right, i.e. you need to have the limit attributes already in the source. However, the original intention was only that process related attributes (with FC=MX) are not allowed to be added in the derived DO, like no 'new' DOs are allowed to be added. Please write a Tissue on this..."

Proposal

Please clarify this so that "dummy" data could be avoided. Note also that it should be possible to avoid range attribute in the original data (even it has FC=MX) while having it in the statistical data because that would then require the rangeC also.