The fact of sharing the same attributes "q" and "rangeC" for the attribute cVal (Vector type) generates some problems.
It was already highlighted in the tissue 65 regarding "db" attribute, but in case of "q" and "rangeC" is more difficult to solve.
In the case of "rangeC" attribute, which one of the two attributes within cVal (mag/ang) does it applies to? It doesn't seem to be possible to set a value suitable for both.
And with regard to "quality" attribute, there are real situations when it isn't possible to calculate the value of the attribute "ang" but however the "mag" of most of the currents and voltages are perfectly known. For instance, if the IED is using one phase voltage as angular reference for the rest of currents and voltages, and this voltage becomes 0, IED is obliged to indicate on all "ang" attributes an unknown value.
As a consequence of this, the "q" attribute of all currents and voltages becomes QUESTIONABLE and INCONSISTENT.
Proposal
Extend CMV common data class.
Discussion
Created
Status
will be discussed at the next meeting. During the Editor meeting, we will make a proposal
29 May 07
Future Improvement
The problem with the quality is not that the magnitude is known but the ang is unknown. The problem is that there can be transitory situations during which it's impossible to calculate the angs, for instance, when phase voltage measurements disappear due to a fuse failure while currents continue flowing.
So, during normal operation the IED is calculating vectors, but with abnormal conditions this IED is not capable of measuring the angles. Due to the fact that mag and ang share quality, one client would receive the mag of all currents and voltages as QUESTIONABLE during the abnormal conditions, when really the value of the currents can be perfectly valid.
In my opinion, we should have an additional quality for the ang attribute.
02 May 07
Ballot Period
With regard to the quality: if the magnitude is known, but the ang can not be calculated, then we are not calculating a vector - in that case we would calculate something with two data objects.
With regard rangeC: we support to have an additional range configuration and range attribute for the ang attribute.
19 Apr 07
Ballot Period
If we introduce a seperate quality for the mag and the ang attribute, we really need to consider, why we then would at all put these two values into the same common data class?
So far, there was at least one modeling principle followed as a design choice when to create a data class and when a data (except for "data of data", like the WYE, etc): They share quality and time stamp.
If elements have different quality, they should in my opinion be different data!
So, if the requirement as explained by the TISSUE is true, we should delete the CDCs based on vectors and create new data or new CDCs as "data of data"