Update psse parser for 3winding transformers #917

hei06j · 2024-06-16T11:47:26Z

This PR is to address the issue #912, which also partly addresses the issue #888. The issue #888 still stands, but with this PR, it solves by changing the transformer sbase to >70 (it is 60 now and not solving).

PSSE three-winding transformer supports

zero NOMVx values, which in this case, the transformer x-th winding nominal voltage is the same as the bus basekV it is connected to,
cw = 1 (for off-nominal turns ratio in pu of winding bus base voltage)
cw = 2 (for winding voltage in kv)
cw = 3 (for off-nominal turns ratio in pu of nominal winding voltage, NOMV1, NOMV2 and NOMV3.)

which are now supported in the PSSE parser as part of this PR.

ccoffrin · 2024-06-23T17:30:11Z

src/io/psse.jl

                    if transformer["CW"] != 1  # NOT "for off-nominal turns ratio in pu of winding bus base voltage"
-                        sub_data["tap"] /= _get_bus_value(bus_id, "base_kv", pm_data)
-                        if transformer["CW"] == 3  # "for off-nominal turns ratio in pu of nominal winding voltage, NOMV1, NOMV2 and NOMV3."
-                            sub_data["tap"] *= transformer["NOMV$m"]
+                        if transformer["CW"] == 2
+                            sub_data["tap"] /= _get_bus_value(bus_id, "base_kv", pm_data)
+                        elseif transformer["CW"] == 3  # "for off-nominal turns ratio in pu of nominal winding voltage, NOMV1, NOMV2 and NOMV3."
+                            if !iszero(transformer["NOMV$m"])
+                                sub_data["tap"] *= transformer["NOMV$m"] / _get_bus_value(bus_id, "base_kv", pm_data)
+                            end
                        end
                    end


Could you elaborate about what the correct semantics of this should be?

The way the code is organized seems risky to me. For example, what happens if transformer["CW"] == 3 and iszero(transformer["NOMV$m"]) == true

iszero(transformer["NOMV$m"]) == true means that the transformer winding m nominal voltage is the same as the bus base voltage it is connected to, i.e. _get_bus_value(bus_id, "base_kv", pm_data). So in the case of transformer["CW"] == 3, the tap value won't change. For when transformer["CW"] == 2, the winding voltage is in kv and does not need to account for transformer["NOMV$m"].

Given the complexity of this sepecific section I suggest we re-write the code int he form,

if transformer["CW"] == 1 ... elseif transformer["CW"] == 2 ... elseif transformer["CW"] == 3 ... else error - unsupported value for "CW" on transformer .... end

Recommendation implemented. :)

ccoffrin

Ok, I think this is very close to ready. I do have one last question about float equality testing to resolve.

ccoffrin · 2024-06-26T20:59:02Z

src/io/psse.jl

@@ -434,8 +434,13 @@ function _psse2pm_transformer!(pm_data::Dict, pti_data::Dict, import_all::Bool)
                    else
                        br_r, br_x = transformer["R1-2"], transformer["X1-2"]
                    end
-                    br_r *= (transformer["NOMV1"]^2 / _get_bus_value(transformer["I"], "base_kv", pm_data)^2) * (pm_data["baseMVA"] / transformer["SBASE1-2"])
-                    br_x *= (transformer["NOMV1"]^2 / _get_bus_value(transformer["I"], "base_kv", pm_data)^2) * (pm_data["baseMVA"] / transformer["SBASE1-2"])
+                    if transformer["NOMV1"] == 0.0


Instead of x == 0.0 I am wondering if isapprox(x, 0.0) would be more robust.

@odow, can you comment the the best practice in Julia for float equality tests?

ccoffrin · 2024-06-26T20:59:10Z

src/io/psse.jl

@@ -448,8 +453,13 @@ function _psse2pm_transformer!(pm_data::Dict, pti_data::Dict, import_all::Bool)
                        br_r *= (transformer["WINDV2"]/_get_bus_value(transformer["J"], "base_kv", pm_data))^2
                        br_x *= (transformer["WINDV2"]/_get_bus_value(transformer["J"], "base_kv", pm_data))^2
                    else  # "for off-nominal turns ratio in pu of nominal winding voltage, NOMV1, NOMV2 and NOMV3."
-                        br_r *= (transformer["WINDV2"]*(transformer["NOMV2"]/_get_bus_value(transformer["J"], "base_kv", pm_data)))^2
-                        br_x *= (transformer["WINDV2"]*(transformer["NOMV2"]/_get_bus_value(transformer["J"], "base_kv", pm_data)))^2
+                        if transformer["NOMV2"] == 0.0


Same note as above.

ccoffrin · 2024-06-26T21:00:35Z

src/io/psse.jl

-                    br_r12 *= (transformer["NOMV1"] / _get_bus_value(bus_id1, "base_kv", pm_data))^2 * (pm_data["baseMVA"] / transformer["SBASE1-2"])
-                    br_r23 *= (transformer["NOMV2"] / _get_bus_value(bus_id2, "base_kv", pm_data))^2 * (pm_data["baseMVA"] / transformer["SBASE2-3"])
-                    br_r31 *= (transformer["NOMV3"] / _get_bus_value(bus_id3, "base_kv", pm_data))^2 * (pm_data["baseMVA"] / transformer["SBASE3-1"])
+                    if transformer["NOMV1"] == 0.0


same note as above.

ccoffrin · 2024-06-26T21:00:43Z

src/io/psse.jl

-                    br_x12 *= (transformer["NOMV1"] / _get_bus_value(bus_id1, "base_kv", pm_data))^2 * (pm_data["baseMVA"] / transformer["SBASE1-2"])
-                    br_x23 *= (transformer["NOMV2"] / _get_bus_value(bus_id2, "base_kv", pm_data))^2 * (pm_data["baseMVA"] / transformer["SBASE2-3"])
-                    br_x31 *= (transformer["NOMV3"] / _get_bus_value(bus_id3, "base_kv", pm_data))^2 * (pm_data["baseMVA"] / transformer["SBASE3-1"])
+                    if transformer["NOMV2"] == 0.0


same note as above.

ccoffrin · 2024-06-26T21:00:50Z

src/io/psse.jl

+                        br_x23 *= (transformer["NOMV2"] / _get_bus_value(bus_id2, "base_kv", pm_data))^2 * (pm_data["baseMVA"] / transformer["SBASE2-3"])
+                    end
+
+                    if transformer["NOMV3"] == 0.0


same note as above

ccoffrin · 2024-07-04T17:10:23Z

Looks good. Thanks for the contribution!

* Improve performance of calc_connected_components (#914) --------- Co-authored-by: mtanneau3 <[email protected]> * Add in-place conversion to basic data format (#915) * Add in-place conversion to basic data format * Add unit test for in-place make_basic_network --------- Co-authored-by: mtanneau3 <[email protected]> * No deepcopy when renumbering components (#916) * No deepcopy when renumbering components * In-place _renumber_components and unit tests * Fix typo --------- Co-authored-by: mtanneau3 <[email protected]> Co-authored-by: Carleton Coffrin <[email protected]> * add notes to changelog and prep for release * Update psse parser for 3winding transformers (#917) * fix psse parser 3 winding tranformers * update readme * add note to changelog and readme * fix quote counting check in psse parser --------- Co-authored-by: Mathieu Tanneau <[email protected]> Co-authored-by: mtanneau3 <[email protected]> Co-authored-by: Rahmat Heidari <[email protected]>

fix psse parser 3winding tranformer

a723c8b

frederikgeth assigned frederikgeth and ccoffrin Jun 16, 2024

update readme

42bc06d

hei06j mentioned this pull request Jun 16, 2024

Not able to solve pti model #888

Open

ccoffrin reviewed Jun 23, 2024

View reviewed changes

ccoffrin mentioned this pull request Jun 23, 2024

Release v0.21.2 #919

Merged

transformer parser updated

a31253e

ccoffrin reviewed Jun 26, 2024

View reviewed changes

update psse.jl

893cfde

ccoffrin changed the base branch from master to rc-v0.20 July 4, 2024 16:49

ccoffrin changed the base branch from rc-v0.20 to release-prep July 4, 2024 17:02

ccoffrin merged commit fa79083 into lanl-ansi:release-prep Jul 4, 2024
6 of 7 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Update psse parser for 3winding transformers #917

Update psse parser for 3winding transformers #917

hei06j commented Jun 16, 2024

ccoffrin Jun 23, 2024

hei06j Jun 23, 2024

ccoffrin Jun 24, 2024

hei06j Jun 25, 2024

ccoffrin left a comment

ccoffrin Jun 26, 2024

ccoffrin Jun 26, 2024

ccoffrin Jun 26, 2024

ccoffrin Jun 26, 2024

ccoffrin Jun 26, 2024

ccoffrin commented Jul 4, 2024

Update psse parser for 3winding transformers #917

Update psse parser for 3winding transformers #917

Conversation

hei06j commented Jun 16, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ccoffrin left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ccoffrin commented Jul 4, 2024