Good, the case files are now uploaded to GitHub as pull requests (case533mt_lo.m and case533mt_hi.m).
Best, Gabriel Från: bounce-127945200-95792...@list.cornell.edu <bounce-127945200-95792...@list.cornell.edu> För Ray Daniel Zimmerman Skickat: den 15 december 2023 17:15 Till: MATPOWER-DEV-L <matpower-de...@list.cornell.edu> Kopia: lovisa.tho...@svk.se Ämne: Re: Proposed 533-bus system based on real data Sounds good … you might also something like case533mt_lo.m and case533mt_hi.m for the names. Not a big deal either way to me. Ray On Dec 15, 2023, at 7:33 AM, Gabriel Malmer <gabriel.mal...@iea.lth.se<mailto:gabriel.mal...@iea.lth.se>> wrote: Hi again, Ray, thank you for suggestion, I was thinking about that as well. I will then upload them as two separate case files (case533mt.mand case533mt2.m), with description of the loads in the help text. Mirko, I investigated the rated currents when merging parallel branches according to your suggestion. There are only 15 edges in the system with parallel branches out of 577 edges in total. In these 15 cases the parallel branches are almost perfectly symmetrical, the largest difference in ratedCurr being less than 2% (687 A with current division, 700 A w/o). Kind regards, Gabriel Från: bounce-127939752-95792...@list.cornell.edu<mailto:bounce-127939752-95792...@list.cornell.edu> <bounce-127939752-95792...@list.cornell.edu<mailto:bounce-127939752-95792...@list.cornell.edu>> För Ray Daniel Zimmerman Skickat: den 12 december 2023 19:10 Till: MATPOWER-DEV-L <matpower-de...@list.cornell.edu<mailto:matpower-de...@list.cornell.edu>> Ämne: Re: Proposed 533-bus system based on real data Hi Gabriel, I prefer to avoid anything that requires modifying a file in the official distribution, even if just to uncomment a single line. So, there are two options for including high load data without putting it in the same case file. 1. Provide the high load case as a separate case file that is identical except for the loads (and the description of the loads in the help text). 2. Provide the high load case as a change table, suitable for use with apply_changes()<https://github.com/MATPOWER/matpower/blob/master/lib/apply_changes.m> (see contab_ACTIVSg200.m<https://github.com/MATPOWER/matpower/blob/master/data/contab_ACTIVSg200.m>, for an example). I think I would prefer the first option. Thanks again, Ray On Dec 11, 2023, at 5:13 AM, Gabriel Malmer <gabriel.mal...@iea.lth.se<mailto:gabriel.mal...@iea.lth.se>> wrote: Hello, Ray, thank you for looking into our case file. We are very glad that you want to include this in the next MATPOWER release. The comments and copyright + license lines are perfectly fine. As an appendix to the file, I added the opportunity to also run the maximum net load hour of 2022. Have a look and see if you find this ok as well. Mirko, thank you for pointing this out regarding the rated currents. You are correct in that we made a simplification when assuming all parallel branches can be fully loaded. I think the grid is relatively balanced, but if someone wants to run an OPF using our ratedCurr they might want to have some additional security margin. Kind regards, Gabriel Från: bounce-127932157-95792...@list.cornell.edu<mailto:bounce-127932157-95792...@list.cornell.edu> <bounce-127932157-95792...@list.cornell.edu<mailto:bounce-127932157-95792...@list.cornell.edu>> För Mirko Todorovski Skickat: den 7 december 2023 22:27 Till: MATPOWER Developer List <matpower-de...@list.cornell.edu<mailto:matpower-de...@list.cornell.edu>> Ämne: Re: Proposed 533-bus system based on real data Hello, I'm glad there is new test system based on real data. While browsing through http://lup.lub.lu.se/student-papers/record/9124806, I noticed something worth discussing. Figure 3.3 (p. 20) illustrates merging parallel branches into a single equivalent branch by simply adding the individual branch rated currents. However, this method isn't entirely accurate. For instance, if we consider two 20 kV cables from https://nkt.widen.net/content/iveqn7cqgj/pdf/230529_AXAL-TT-Pro-3_24kV.pdf?u=gj0n1y, with cross-sections of 25 mm^2 and 50 mm^2 and rated currents of 100 A and 145 A respectively, the equivalent cable rated current wouldn't be 245 A. Generally, all cables exhibit similar reactance per unit length regardless of their cross-sections. However, the resistance ratio is approximately 2 due to the second cable's doubled cross-section compared to the first one. Consequently, applying the current divider principle becomes necessary to determine the proportion of total current flowing through each parallel cable. In the attached file, I've computed three scenarios: 1. Equal length for both cables, 2. The first cable is 20% longer, 3. The second cable is 20% longer. The equivalent rated current for the parallel combination is calculated so that one cable carries current equal to its rated current. The results for the three scenarios are: 1. 223 A, 2. 210 A, 3. 238 A. I believe it would be beneficial to perform similar calculations for the actual cables in the test system to validate if corrections to the rated current are necessary. Best regards, Mirko On 12/6/23 18:28, Ray Daniel Zimmerman wrote: Hello Gabriel, I’d be happy to include this in the next MATPOWER release. I’ve attached a version with a few tweaks to the comments at the beginning of the file (no data changes). Please check that they are ok with you and the providers of the original data. I added some of the description of the system that you included in your e-mail. And I added copyright and license lines. Whenever possible we like to include an explicit license for all MATPOWER files, BSD 3-clause license for code and the Creative Commons Attribution 4.0 International license for data. If these are ok with you, please go ahead and create a pull request on GitHub to add the file. Thanks again for this contribution! Ray On Nov 30, 2023, at 10:38 AM, Ray Daniel Zimmerman <r...@cornell.edu><mailto:r...@cornell.edu> wrote: Absolutely, we would be interested in including this. Let me have a look at it and get back to you about any suggested modifications. Ray On Nov 29, 2023, at 9:13 AM, Gabriel Malmer <gabriel.mal...@iea.lth.se><mailto:gabriel.mal...@iea.lth.se> wrote: Dear Matpower-developers, My name is Gabriel Malmer and I’m a PhD student in power systems at Lund University. For my master’s thesis, me and my colleague Lovisa Thorin developed a 533-bus distribution system in Matpower, based on real system data from the local DSO Kraftringen in southern Sweden. Kraftringen have agreed to have this Matpower-case publically available, and I was wondering if this would be of interest for the Matpower data-package? See attached file, case533mt.m. If this is of interest, please let me know and I can add it to the Matpower GitHub. Also feel free to suggest modifications to make the case-file more adequate. Some general information: The system covers about 20x30 km and servers about 30,000 inhabitants + an industrial area. Total net loads between -5 and 50 MW. Right now, the loads included are the net loads from the minimum load hour of 2022. For more information, see http://lup.lub.lu.se/student-papers/record/9124806. <image001.png> Kind regards, Gabriel Malmer Doktorand LTH, Lunds universitet Avd. för Industriell Elektroteknik och Automation (IEA) Postadress: Box 118, 221 00 Lund Besöksadress: Ole Römers väg 1, 223 63 Lund <case533mt.m> <case533mt.m>
