1. Conversion of Cali Cyme Feeder Model [20h]

[vank348]

Summary of CYME-to-glm Conversion Work

2/28/2014:

• Added database flexibility where equipment database tables (i.e. CYMEQSHUNTCAPACITOR) can be located in primary database, in addition to secondary database.

3/07/2014:

• Added support for shunt capacitors specified in single three-phase field as opposed to in three separate per-phase fields
• Added workaround where network voltage not specified with swing bus

3/14/2014:

• Added handling of conductor distances specified in geometric mean distance (GMD) phase-to-phase and GMD phase-to-neutral (CYMEQAVERAGEARRANGEMENT table in database)

3/21/2014:

• Added handling for CYM section devices with location specified as '0' where previously only values of '1' or '2' were allowed. This results in topological change wherein new nodes need not be created for these devices, since they are already explicitly represented in CYM model.
• Added support for three-winding autotransformers. Currently approximates three-winding transformer in GridLAB-D as three-phase regulator.

3/28/2014:

• Added handling for underground and overhead lines with length equal to '0'. Lines are removed in the glm implementation, with the connected nodes instead being linked via a parent-child notation.
• Added handling for nodes with no phase designation that are result of previously described automated parent-child linking.

4/04/2014:

• Began adding a check that must be done to eliminate 'grandchildren' occurences. These need to be replaced with 'sibling' reltationships wherein two or more nodes can share the same parent node, but that parent node cannot itself have a parent denoted.

[vank348]

4/10/2014: -Resolved grandchildren issue using sibling fix described above. Added check to remove nodes that are present in the CYMNODE table but not connected elsewhere in the database. Began addressing overhead line terminal node phase mismatch issue.

[dpinney]

Quick question: do we have written instructions for getting the required .mdb files out of Cyme? I recall seeing a help file that described some of it, but I can't find it online.

[jcfuller1]

I believe that MDB is the natural storage method for CYMEDIST models. Here's a link to a copy of the user's manual: http://www.scribd.com/doc/204048645/87/Import-Export-Library-Online-Update

[vank348]

4/14/2014: -Resolved collapsed node phase mismatch issue. Began addressing cases where newly created nodes for mid-section devices are mistakenly deleted by conversion script.

[dpinney]

Thanks I didn't have that manual. I'm trying to develop a tutorial on the level of "windows-r, type c:/program files/, email the file named X..."

[vank348]

Summary of CYME-to-glm Conversion Work

2/28/2014:

• Added database flexibility where equipment database tables (i.e. CYMEQSHUNTCAPACITOR) can be located in primary database, in addition to secondary database.

3/07/2014:

• Added support for shunt capacitors specified in single three-phase field as opposed to in three separate per-phase fields
• Added workaround where network voltage not specified with swing bus

3/14/2014:

• Added handling of conductor distances specified in geometric mean distance (GMD) phase-to-phase and GMD phase-to-neutral (CYMEQAVERAGEARRANGEMENT table in database)

3/21/2014:

• Added handling for CYM section devices with location specified as '0' where previously only values of '1' or '2' were allowed. This results in topological change wherein new nodes need not be created for these devices, since they are already explicitly represented in CYM model.
• Added support for three-winding autotransformers. Currently approximates three-winding transformer in GridLAB-D as three-phase regulator.

3/28/2014:

• Added handling for underground and overhead lines with length equal to '0'. Lines are removed in the glm implementation, with the connected nodes instead being linked via a parent-child notation.
• Added handling for nodes with no phase designation that are result of previously described automated parent-child linking.

4/04/2014:

• Began adding a check that must be done to eliminate 'grandchildren' occurences. These need to be replaced with 'sibling' reltationships wherein two or more nodes can share the same parent node, but that parent node cannot itself have a parent denoted.

[vank348]

4/20/2014: -Resolved terminal node phase mismatch issue. Correcting capacitor voltage set point translation error. Have had to put work here on hold temporarily due to a major issue that came up with another project. This will be top priority again next week.

[dpinney]

Hi vank348, could you send me an email? My address is dwp0 ampersand nreca.coop. I'd like to discuss this code. Thanks!

[vank348]

5/1/2014: Resolved capacitor voltage set point issue. Began fixing another phase mismatch problem. This one arose from sections with devices (fuses, etc) not getting caught with the existing fix.

[vank348]

5/6/2014: Resolved section device phase mismatch problem. Began tackling related phase mismatch issue where collapsed (parent-child node) nodes have conflicting phase information.

[vank348]

5/9/2014: Addressed new phase mismatch issue with wholistic fix that turns every node 'family' (any set of nodes connected by at least one parent-child relationship that are electrically common) into a single parent with the remaining nodes being children (as described previously), this time collecting all phase information and applying combined phase coding to all nodes. Working out bugs.

[vank348]

5/16/2014: Resolved phase issues with parent-child relationships. Began resolving issue where three-phase nodes are getting labled as split-phase nodes (ABCS).

[vank348]

5/16/2014: Resolved split-phase mislabelling issue. Began addressing issue where collapsed nodes result in parallel lines (duplicates in admittance matrix).

[vank348]

5/21/2014: Resolved parallel line issue. Currently trying to find source of issue where 'split connection' error occurs with line being connected to two phases simultaneously.

[vank348]

5/27/2014: Resolved split connection error and another parallel line issue. Exciting milestone: topology errors appear to be resolved. Currently investigating reason for failure of feeder power flow to converge.

[afisher1]

The previous developer working on this ticket has left the lab. I have taken over. My assessment is that the file produced was not able to converge due to malformed parent child relationships. Debugging of the code is currently underway.

[afisher1]

I've been able to determine that this script doesn't handle the unique situation where objects are listed as transformers but their equipment database shows them as regulators. So I’ll have to write some code to handle that. Also, I have modified the script to handle extracting more than two objects from a single link connection as I know this is a possibility for malformed parent child relationships. I probably won’t be able to address these issues till late this week at the earliest. The fixes shouldn't take me more than a day to code up once I do get to them.

[dpinney]

Per call on 25 July, we might be able to get something working by 8 August.

[dpinney]

I've received a glm. :smile_cat: