From GridLAB-D Wiki
Jump to: navigation, search

B Van Kirk 2/6/2014 13:08


For underground lines with paired (phase-neutral) conductors (e.g. from_terminal "1 4; 2 4; 3 4";) I think it should be made clear that while '4' is being used with each pair because the neutrals are bonded at either node (right?) they will be modeled as separate conductors.

Ftuffner 01:41, 21 March 2014 (UTC) This seems like a reasonable clarification to me. We need to be clear that only the endpoints are commonly-bonded, but the individual neutrals are handled separately in the line equations.

Overall comments

Ftuffner 15:17, 24 March 2014 (UTC) - Any "newly created" fields should be typed (i.e., size, if appropriate, and casting -- complex [3])

Ftuffner 21:53, 3 April 2014 (UTC) It has been suggested that maybe there's a better way to "shorten" the specification for from/to and their corresponding terminals (rather than 4 fields). Suggestions are welcome at this point since this method was selected for ease of implementation/flexibility.

Overhead Lines

Ftuffner 18:10, 21 March 2014 (UTC) - I think we should clarify the 5x5 example to be a "5x5 with neutral". It seems a little odd when you say it is n-by-n with n-phases and the number of phases is 5, but then your matrix is 6x6.

vank348 05:30, 11 April 2014 (UTC) - I believe this issue has been resolved.

Ftuffner 18:10, 21 March 2014 (UTC) - Pointers to the individual matrices are not planned to be maintained in the current NEVBRANCHDATA structure, since they aren't needed now. Is there a compelling reason to continue to include them?

vank348 05:30, 11 April 2014 (UTC) Agreed. This bit was only included for consistency with the underground line section.

Ftuffner 18:10, 21 March 2014 (UTC) - The description of how the matrices are put into the Newton-Raphson powerflow are a little confusing. I'm not sure we really need that level of detail in the specifications. At the very least, I think this could use a slight rephrasing.

vank348 05:30, 11 April 2014 (UTC) We should meet to discuss this. I'm not sure how to make it clearer or how much detail to include or whether to cut this section entirely.

Underground Lines

Ftuffner 18:10, 21 March 2014 (UTC) - Class:underground_line - The example mentions a node 3. I think this is just a typo.

Ftuffner 18:10, 21 March 2014 (UTC) - Class:line_configuration - First paragraph has the last sentence effectively repeated

Ftuffner 18:10, 21 March 2014 (UTC) - It's not entirely clear what the "resistance array" is for. A link to the equations page here might help (or quick explanation).

andyfisher 19:13, 24 March 2014 (UTC) - The resistance array is there to hold the resistance properties for each conductor for calculating the Y array.

Ftuffner 18:10, 21 March 2014 (UTC) - It's also "pseudo-code", not "sudo-code". Too much Linux for you!

andyfisher 19:13, 24 March 2014 (UTC) - Wow! That is sad. sudo get me a sandwich!

Ftuffner 18:10, 21 March 2014 (UTC) - Does anything special/particular need to be added for the "ordinary line" neutral case (separate conductor)? I don't see that really covered in here.

andyfisher 19:13, 24 March 2014 (UTC) - Nope! Solid conductors are treated as well insulated lines and are treated just like overhead lines such that the resistance and gmr listed for the conductor are used verbatim rather than any intermediary calculation for tape-shields and concentric neutrals.

Transformers

Ftuffner 15:17, 24 March 2014 (UTC) - For transformers with tertiary windings, how is the rating/impedance handled? I found an example of a "harmonic stabilization" tertiary winding (http://www.mgmelectricalsurplus.com/documentation/mcgraw3ph1nameplate.jpg) that has a different rating than the overall transformer. I'm not sure how that affects things. I'm also not sure if this differs with transformers connected with an "active" tertiary winding. Do we need to extend the transformer_configuration object to cover this better?

vank348 05:30, 11 April 2014 (UTC) Using the Dugan modeling approach, I'm not 100% on how this would be modeled, but I don't imagine it being too difficult. More importantly, if the transformer exists it makes sense that we should support it. I think it would make sense then to allow the ratings to be specified in the array format.

Ftuffner 15:17, 24 March 2014 (UTC) - For custom transformer types, I'm not sure I follow the "winding connection field should contain exactly as many elements as there are terminals listed in the transformer object".

vank348 05:30, 11 April 2014 (UTC) I think this is clearer now.

Ftuffner 15:17, 24 March 2014 (UTC) - The custom transformer example is a bit confusing. How difficult would it be to make a diagram with this example?

vank348 05:30, 11 April 2014 (UTC) I'll put together a diagram.

Ftuffner 15:17, 24 March 2014 (UTC) - The paragraph after the transformer_configuration examples repeats several elements of the paragraph before the examples.

vank348 05:30, 11 April 2014 (UTC) I believe this issue is resolved.

Ftuffner 15:17, 24 March 2014 (UTC) - The "NEVTransformer" link isn't properly formatted.

vank348 05:30, 11 April 2014 (UTC) I believe this issue is resolved.

General Comments

--Jcfuller 23:18, 1 April 2014 (UTC) - Can units be included in the array formats? Should they be?

Ftuffner 23:34, 3 April 2014 (UTC) I'm not sure they have any relevance in this particular section (link), but this is a good question. I would assume "yes", but the logistics may make things difficult. Posing as question over in the data format page.

--Jcfuller 23:18, 1 April 2014 (UTC) - In the line spacing, how does GLD know which coordinate system you are trying to use (rather than a mistake of leaving out or adding some data)? Will there be logical checks for the geometrical configuration?

Ftuffner 23:34, 3 April 2014 (UTC) Logical checks of sorts are a requirement for this. We should rephrase this to indicate as much. Also, the logical progression of "how do I imply which coordinate system I'm in" should be addressed.

--Jcfuller 23:18, 1 April 2014 (UTC) - "Unlike the transfer admittance, this will be a complex-valued matrix... " - implies that the transfer matrix is NOT complex valued...is this true? Doesn't seem to be the case when looking at the next set of eqs.

Ftuffner 23:34, 3 April 2014 (UTC) Changed the phrasing to be a little more clear, hopefully.

--Jcfuller 23:18, 1 April 2014 (UTC) - I don't follow how to specify the parallel (non-electric) connections. In the example, it shows two connections (UGL_Space and UGL_Space2), but its unclear what the relationship (spatially) is between them.

--Jcfuller 23:18, 1 April 2014 (UTC) - The notation is a bit confusing in "The Init, Presync..." section. Its unclear why you need a [2][N][N] array -- is the [2] indicating a break down of real and imaginary?

User:andyfisher:andyfisher 9:54, 17 April 2014 (UTC) - Yes the [2] is to split real and imaginary.

--Jcfuller 23:18, 1 April 2014 (UTC) - Is transformer really a link anymore, or some combo of link and load? Treated as a link?

--Jcfuller 23:18, 1 April 2014 (UTC) - Where do regulators end up in this? As a separate class, or a special class of transformers (like OpenDSS)?

--Jcfuller 23:18, 1 April 2014 (UTC) - " As with the previous implementation, impedance values can be specified in conventional per unit (%) or absolute values. " - we can? How do you do that?

--Jcfuller 23:18, 1 April 2014 (UTC) - Unclear how you are going to do error checking in the transformer configuration. Is "winding_coupling "1 2;1 3";" okay for both

from_terminal "1";
to_terminal "2; 3";

and

primary_terminal "1";
secondary_terminal "2";
tertiary_terminal "3";

Do they both configure the same way?

--Jcfuller 23:18, 1 April 2014 (UTC) - In the 3rd transformer_configuration example, how does the 500 kVA rating get split amongst the different winding connections (this might be a 300 / 200 or 400 / 100 connection)?

--Jcfuller 23:18, 1 April 2014 (UTC) - Typo in "model implementation" to link to NEVTransformer. Does this page exist yet?

Ftuffner 23:34, 3 April 2014 (UTC) Syntax error - it wasn't double bracketed.


Email comments

-- It looks like this answers part of "specify the from and to values for node and terminal" in the Data format spec. For backward compatibility, how will this be handled?

-- For line spacing: we should show an example for 6 or 7 conductors to better highlight the structure.

-- Is there a reason we use semicolons to separate integers in the from_terminals?

Ftuffner 23:34, 3 April 2014 (UTC) This implies different conductors (to differentiate neutral-containing versions), but does need to be clarified.

-- The name of the third underground cable should be changed. “Insulated” makes it sound like the other two types are not insulated.

-- For transformers: the documentation will need to include clear drawings of the terminal connections or else there will be problems. (not a specification issue)

Ftuffner 23:34, 3 April 2014 (UTC) Semi a specification issue. This will need to be documented and may be useful for the specifications so the "specific instances" are defined for implementation.


More comments

--Melizondo 07:14, 20 April 2014 (UTC) Table 2, definition of conductor, “from_phases” and “to_phases:” shouldn’t they be “from_terminal” and “to_terminal?”

Ftuffner 16:09, 22 April 2014 (UTC) Indeed it should. Fixed.

--Melizondo 07:14, 20 April 2014 (UTC) GLM inputs section, line spacing examples: There are errors in the text, the first 2 specifications are full matrix and condensed upper diagonal of distances between conductors, and the last specification is Cartesian coordinates. In the text, the first 2 are referred as Cartesian coordinates and the last one as absolute distances. (the name field in the arrays are correct, the error is just in the text)

Ftuffner 16:09, 22 April 2014 (UTC) Fixed while fixing other minor text items in that section.

--Melizondo 07:14, 20 April 2014 (UTC) Underground Lines section, Class:Line_configuration: why is the special note special? It might be worth explaining. Shouldn’t also this special note appear in the overhead lines section?

--Melizondo 07:14, 20 April 2014 (UTC) Underground Lines section, The Init, Presync, Sync, and Postsync Functions: why is this section needed? Shouldn’t be more appropriate to have this section in the implementation page? Or are these necessary because of some particular aspect of NEV that is different from previous implementation? If so, it might be good to write this up front

--Melizondo 07:14, 20 April 2014 (UTC) Underground Lines section, The Init, Presync, Sync, and Postsync Functions, General, fifth paragraph: should this link be the same “Underground Line Equations” that appear in the other paragraphs of this same section?

--Melizondo 07:14, 20 April 2014 (UTC) Underground Lines section, The Init, Presync, Sync, and Postsync Functions, Needed Equations: these are the same as the equations in the link “Underground Line Equations,” or are these necessary because of some particular aspect of NEV that is different from previous implementation? If so, it might be good to write this up front. Or you may want to eliminate the page referred with the link “Underground Line Equations” and just use the equations in this page.