From GridLAB-D Wiki
Jump to: navigation, search

--Jcfuller 23:19, 1 April 2014 (UTC) - In the example, what is the note about "wye spec"? I also can't find a reference for "voltage_wye". Bit confused with the clarifying example.

This should be on the NEVNode specification page and will be added if it is not. Added some clarifying text too.
Ftuffner 01:32, 18 April 2014 (UTC) Hopefully much clearer now. Adjusted voltage specification method to remove "delta" and "wye" fields.

--Jcfuller 23:19, 1 April 2014 (UTC) - It's unclear from this document whether this method will work with FBS or not. While I think after reading through, it should be obvious...it would be good to call it out specifically.

Ftuffner 21:44, 3 April 2014 (UTC) Well, not to be overly facetious, but it does mention Newton-Raphson at the beginning of every section. Added an explicit note to the top indicating this is NR-only.

--Jcfuller 23:19, 1 April 2014 (UTC) - Typo in Link/branch structure, 'from' variable description should be "from" not "to".

Ftuffner 21:44, 3 April 2014 (UTC) Copy-paste-itis strikes again. Fixed.

--Jcfuller 23:19, 1 April 2014 (UTC) - Is "link_type" only used for reliability, or is there a parent class of "link" that pulls in other classes?

Ftuffner 21:44, 3 April 2014 (UTC) link_type is only used for reliability, even in the current implementation. Special_lnk inside link objects serves a similar purpose for handling posting the admittance/impedance values and interpreting them (e.g., transformers). This portion of the functionality will no longer be necessary since link objects will be handling their admittance postings/updates directly. No "higher parent" class for link objects is expected beyond link, aside from the relevant powerflow_object class, which will need some minor changes for phase handling (but effectively does nothing).

Emailed comments

-- Under GLM inputs, how do you specify a WYE connected load? “1 2 3” gives you delta, would “1 2 3 4” give you WYE grounded? Would “1 4” be a single phase to ground load?

Ftuffner 21:44, 3 April 2014 (UTC) - Clarified this is just an example, but will make sure examples of both are on the relevant node page.

-- Will there be a volages_delta? We need to make sure that we can properly represent line-to-line, line-to-neutral, and single phase.

Ftuffner 21:44, 3 April 2014 (UTC) - Similar to the above comment, will insure this is addressed on the node specification page.
Ftuffner 01:32, 18 April 2014 (UTC) voltages have been redefined to a syntax similar to all other terminal-based connections. voltages_wye and voltages_delta are now just voltage

-- Node/bus Structure “voltage” is this still the line to neutral? We may need to adjust this since we are now allowing new combinations.

Ftuffner 21:44, 3 April 2014 (UTC) - Another item for the node specification page. The current plan is to support "L-G" and "L-L" voltage inputs. L-N will need to be discussed since it requires mapping a neutral and treating it differently than just another terminal.

-- Node/bus Structure “terminal” does every terminal have a unique number? Similar to how nodes have a unique name? So instead of node 32 w have terminal “45 46 47 48”?

Ftuffner 21:44, 3 April 2014 (UTC) - Better for the node specification page. It's semi-implied here (and is now fixed on the overall specification page) - terminals are bit masks. The numbers 1 - 64 will represent 64 unique phases that the user needs to keep track of. Phase 1 could be "A", as could Phase 2, 3, and 5. Or they could be "ABC". A 5000-node GLM may only need to use the numbers 1, 2, and 3 (for A, B, and C), or the user may elect to assign unique numbers to different "versions" of A, B, and C. The numbers just represent a unique "reference potential". Specific assignments are left to the user to keep track of.

-- Link/brank structure “terminals_from” is this an object or a single terminal within an object?

Ftuffner 21:44, 3 April 2014 (UTC) This is a single terminal designation within an object. This is used in conjunction with the from field to determine connection points on the corresponding node object.

-- Do you need to properly specific the from and to values for node and terminal? Is there a way to combine these in a single designation?

Ftuffner 21:44, 3 April 2014 (UTC) We are open to suggestions. They are kept separate for ease of implementation and flexibility. This effectively becomes the same as the phases property in the existing implementation, there are just phases for both sides now (since terminal 1 of the "from" side may not be what terminal 1 of the "to" side represents). This is also an internal programming structure. The link-object-level implementation should be discussed on the link specification page.