Project

General

Profile

Issues #666

Missing GOOSE and SV subscriber test procedures: verfiy that the subscription of test signals follows IEC 61850-7-4 - Annex A.

Added by Thierry Dufaure almost 3 years ago. Updated over 2 years ago.

Status:
Closed
Priority:
Normal
Due date:
Discuss in Upcoming Meeting:
No
Clause Reference:
Annex A
61850 Standard:
7-4
Triggering Tissue:
Final Decision:
See solution 20210727
Initial Test Document:
Updated Test Document:
Test Case ID:
sGos23 (new) and sSvs16 (new)
Closed Reason:
Test Procedure Update
Triggering Tissue 2:
Triggering Tissue 3:

Description

New tests are required to verify that
- GOOSE subscriber
- SV Subscriber
interpret the q.test flag from the subscribed signals and accept the signal value depending on the current subscriber behavior (IEC 61850-7-4 Annex A).

Depending on te supported Beh value - the subscriber shall at least "process as invalid" signals that are flagged as q.test=true when the Beh = On.

This was discussed during the uca testing procedures meeting hold on May, 4th.


Files

subscriberTests_final.docx subscriberTests_final.docx 24.5 KB Michael Haecker, 05/07/2021 10:01 AM
Solution to redmine_666_20210602.docx Solution to redmine_666_20210602.docx 31.1 KB Richard Schimmel, 06/02/2021 06:25 AM
Solution to redmine_666_20210713.docx Solution to redmine_666_20210713.docx 31.5 KB proposal Thierry Dufaure, 07/13/2021 10:36 AM
Solution to redmine_666_20210727.docx Solution to redmine_666_20210727.docx 32.3 KB Richard Schimmel, 07/27/2021 10:04 AM
#1

Updated by Michael Haecker almost 3 years ago

Proposal for the update of the test procedures attached

#2

Updated by Michael Haecker almost 3 years ago

  • Status changed from New to In Progress
  • Discuss in Upcoming Meeting changed from No to Yes
#3

Updated by Richard Schimmel almost 3 years ago

Updated as discussed during TPWG. Remove LGOS/LSVS out-of-scope. Harmonize with other SV and GOOSE test procedures. Updated the step numbering. Refer to existing GOOSE ping-pong. Added abstract descriptions and specify testcases are mandatory.

#4

Updated by Richard Schimmel almost 3 years ago

A comment: The output data quality is determined by LN.Beh, so for steps 7 to 12, when LN.Beh is “test” or “test/blocked”, the DUT will/should/shall(?) publish (the "pong" message) with quality.test=true.

#5

Updated by Michael Haecker almost 3 years ago

The output data quality is determined by LN.Beh, so for steps 7 to 12, when LN.Beh is “test” or “test/blocked”, the DUT shall publish (the "pong" message) with quality.test=true.

This behaviour is described in Part 7-4, Annex A, Table A.1.

#6

Updated by Thierry Dufaure almost 3 years ago

Maybe we need to change the text "DUT ignores the data value changes flagged with quality test true" to something more explicit. The DUT may do something like "blocknig protection function" for sample value case, or be configured in case of GOOSE subscription to replace the incoming value. Ignore is not the correct word for exposing that the DUT acklowedged the presence of the test mismatch and react accordingly to the test mismatch expected behavior.

#7

Updated by Michael Haecker almost 3 years ago

sSvs16: agree. If the "DUT ignores the test samplesd flagged valueswith quality test true", then its SV subscribing function is not using the samples any more, processing analogues is discontinued. Propose to change into "DUT does not process the test samplesd flagged values with quality test true. An alarm message is created (when supported)." .

sGos23: agree. Propose to change into "DUT does not process the data value changes flagged with quality test true. DUT continues repeating GOOSE with the last values and does not send GOOSE with the changed values." .

#8

Updated by Thierry Dufaure almost 3 years ago

Propose to change expected results of sSvs16 into "DUT does not process the test samples flagged values with quality test true. Verify according to PIXIT Svs1a that the samples are NOT consumed.".
Propose to change expected results of sGos23 into "DUT processes the data value flagged with quality test true as described in the PIXIT (for instance: keep last non test value, substitute to a configured value, ...)".

#10

Updated by Richard Schimmel almost 3 years ago

TPWG agreed. Added PIXIT entry Gs12:

How does the subscriber handle incoming data flagged as test when the destination LN.Beh is On or Blocked?
For example: keep last non test value, substitute to a configured value, etc.

Is this behavior: fixed/configurable

#11

Updated by Richard Schimmel over 2 years ago

  • Status changed from Resolved to Closed
  • Discuss in Upcoming Meeting changed from Yes to No
  • Initial Test Document set to Ed2Amd1 TP1.0
  • Updated Test Document set to Ed2Amd1 TP1.1
  • Closed Reason Test Procedure Update added
  • Closed Reason deleted (--Not Set---)

Copy sGos23 also to Ed2

Also available in: Atom PDF