Market simulation results are published every day before 18:00 on the FTP server: ftp.cwe-sf2.com (port: 21).
The login information was sent to you via the project partners in a specific communication just before first data publication.
If you would like to receive the login details, please revert to your respective power exchange account managers who will provide you with the access.
Market messages concerning simulation results
Errors in the fixed labeling of the CBCOs since 11 April 2015
As part of post-go live quality checks, the CWE partners discovered errors for some of the fixed labelling of CBCOs. From the 11th of April 2015 onwards, the labelling of CBCOs – in single cases – appears to be inconsistent with the labelling applied during the external parallel run until then. Indeed, for some labels there are two different IDs available for identical CBCOs.
The CWE TSOs corrected the erroneous labeling, which equals to 3% of the data, thereby aligning the labeling of CBCOs with their labeling during the previous parallel run phase. In order to correct the erroneous labeling, also other labels had to be renamed, which in total equals to 10% of changed labels. The correction has been made on 6th of August 2015 in the Utility tool on the CASC website.
We apologize for the inconvenience resulting from these errors. The CWE partners are making best efforts to avoid any other occurrences in the future and provide the market with reliable data to base their analysis on.
Ex-post publication of historical CBCOs with fixed labelling (until 11/04/2015)
Please note that the following files were recently updated on the Ftp server where parallel run results can be retrieved: PTDF_Fixed_CBCO_ID_all and PTDF_Fixed_CBCO_ID_Presolved_only (folder 2015).
The data is now available until 11/04/2015. Please note that the data for the period until Go-live on 20 May will be published as soon as possible before Go-live or right after Go-live while respecting the D+2 publication principle.
Please also note that in the Utility Tool the final “PTDFs” are updated with the fixed ID ex-post since 11 April; this will be done also from Go-live on in D+2 as well as for the “All CBCO fixed Label”.
Publication of simulation results until Flow-Based Go-live (12 May 2015)
Dear Market Participants,
The CWE FB MC partners are pleased to inform you about the extension of the external parallel run period.
It was previously communicated that the parallel run will end on the 15th of May 2015. As market parties expressed their wish to get the simulation results as near as possible to Go-live, the Project in the meantime could investigate possibilities to meet market expectations and we are now pleased to be able to best support the Market for a successful launch of Flow-Based MC.
Thus, FB parameters as well as simulated market results will be published on the Ftp server following the same process and with the usual timings until 19th of May (trading day) for delivery day 20th May.
Publication of aggregated D2CF data
We would like to inform you that the aggregated D2CF assumptions for the parallel run period have been published on the Ftp server in the “2015” folder. The historical aggregated D2CF data (vertical load, generation and bilateral exchanges) from 01/01/2013 to 19/02/15 were published.
Besides, please note that aggregated D2CF data since 20/02/2015 is directly published on the CASC website in D+2 as part of the general publication framework: http://www.casc.eu/en/Resource-center/CWE-Flow-Based-MC/Publication-CWE-Flow-based-External-parallel-run
Publication of LTnom & SA ATC for the 25th of January
Dear Market parties,
Please be informed that the indicators (LTnom & SA ATC) for the 25th of January, a day for which Default Flow based Parameters were applied, are now available on the CASC website in the dedicated publication section for Flow based DX//run data publications.
Publication of recomputed days based on default FB parameters
The CWE project partners inform you that four missed days of the parallel run for 2014 (28/02/2014 – 11/03/2014 – 06/04/2014 – 22/05/2014) could be recomputed based on default PTDFs which would have been used in a live environment. The default PTDFs for those days have been published on the Ftp server under the folder 2014 and “Recomputation based on default PTDFs”. In this folder, you will also find two other sets of default PTDFs corresponding to successful parallel run days (20/06/2014 + 04/08/2014) which have been computed for comparison.
Following a request from FB User Group members, the FB parameters for winter situations have been published in the folder 2015 and “Winter day scenarios”: you will find there PTDFs for an hour corresponding to a typical winter situation (case A) and PTDFs for a stressed winter situation (case B).
The Project is confident that this additional information should further support MPs in the successful adaptation to the FB MC method in the CWE region.
Fallback Flow-Based parameters used for delivery day 25.01.15
Please note that the simulations performed for delivery day 25.01.15 as part of the daily parallel run were based on Fallback Flow-Based Parameters. These default PTDFs are used in case the parameters calculation could not be performed properly due to any issue occurring in the calculation process. Simulated data published for this day are therefore the default parameters as well as the market results based on those parameters. The root cause of the issue which occurred are currently investigated and further explanation will be given when possible.
Ex-post publication of all historical CBCOs with fixed labeling (until 28/01/2015)
The CWE Flow-Based MC Project would like to inform you that additional data is now published as part of the external parallel run in order to further support market needs in terms of data publication and transparency.
Following additional data is now available:
- Historical files with fixed labeling of presolved CBCOs since the beginning of the parallel run (the CBCOs remain anonymous but their ID is fixed) – available in excel on the FTP server in the new folder “PTDF (Fixed CBCO ID)”
- Historical files with fixed labeling of all (presolved and not presolved) CBCOs since the beginning of the parallel run (ID is still fixed, same than for the presolved publication and available even for non-presolved CBCOs) – available in excel on the FTP server in the new folder “PTDF (All Fixed CBCO ID)”
Please note, that due to algorithm changes needed to provide the full CBCOs fixed labeling files, already published data have been re-produced with new labels, and old data has been replaced and are no more valid for comparison with new data.
Starting from now, as it is still a manual process, data will published with fixed ID in the dedicated folder of the FTP server mentioned on a monthly basis. This process will be industrialized as of Flow based Go Live.
Ex-post publication of results for 09/04/2014
The simulation results for 09/04/2014 could be published and included in the regular file on the FTP server following successful recomputation. As this specific missed day was due to an issue at the very end of the FB parameters calculation process, it could be recomputed ex-post, after an adjustment of the algorithm, based on the valid input from TSOs already available. This technical recomputation enabled to calculate valid results.
Please note that the risk of having missed days due to technical issues should significantly reduce thanks to the implementation of appropriate operational safeguards and mitigation measures as well as the use of the new common TSO system.
Ex-post publication of parallel run Shadow Auction ATCs
Please be informed that the Shadow Auction ATCs have been recomputed as of the start of the daily parallel run for the following period (2014/02/25 to 2014/07/22). Data as of 2014/07/23 until now has already been published on a daily basis.
Please note that for the missing parallel run days, SA ATC values cannot be published: 2014/02/28, 2014/03/11, 2014/04/06, 2014/04/09 and 2014/05/22.
Message regarding hub-PTDFs
Project partners would like to draw your attention on the fact that the published FB parameters display so-called « hub-PTDFs » (i.e. giving the influence of the variation of the balance of a given hub on the critical branch). Hub-PTDFs, as explained during the market forum in March 2013 (http://www.casc.eu/media/CWE%20FB%20Publications/CWE%20MC%20FB%20Market%20Forum_2013_03_07.pdf; on p. 41-42 ), are computed using a sink-node as a reference, which location is changing over time as the rule to determine it depends on the topology of the grid. So in fact, PTDF(hub) actually means PTDF(hub) – PTDF(sink node).
As a consequence, analyzing and comparing hub-PTDFs over large periods of time could be misleading since there is an important probability that the sink node, therefore your reference, has changed within your sample. On the contrary, so-called « exchange-PTDFs » (that give the influence of an exchange between two hubs by substracting two hub-PTDFs) are consistent over time because they do not refer any more to the sink node (indeed: PTDF(HUB1-HUB2) = PTDF(HUB1) – PTDF(sink node) – (PTDF(HUB2) – PTDF(sink node)) = PTDF(HUB1) – PTDF(HUB2).
Project partners therefore recommend market parties to use « exchange PTDFs » for analyses spanning over large period of time.
New FRM values will be used for Go Live and implemented in the parallel run as of July 26th. Compared to the FRM-values used before, in average the FRM values decrease by one percentage point.
Maximum increases for single lines are 10 percentage points, maximum decreases are 9 percentage points for single lines.
Please note that Shadow Auction ATC values are now again published as part of the daily parallel run process (http://www.casc.eu/en/Resource-center/CWE-Flow-Based-MC/Publication-CWE-Flow-based-External-parallel-run). The SA ATCs for the period missing will be recomputed and published as soon as possible.
A technical issue leading to potential errors in the publication of SA ATCs has recently been identified. In order to guarantee the integrity of data made available to market participants, CWE partners decided to stop the publication of SA ATCS until the issue is completely addressed, which is expected in the coming days. We apologize for the inconvenience and commit to restore the situation and provide dedicated explanations in the best possible timings.
The so-called “private system” (the part of the FB system that allows TSOs to carry out analyses on their own) contains a “filter button” that allows to check only the CBs overloaded by some LTA and/or NTC corners (to prevent unnecessary data loading).
This functionality in the system has been modified recently and consequently misunderstood by the operator as he concluded that there were no LTA overloads due to a lack of results on the screen.
As a result, the operator did not integrate all possible remedial actions, which limited the available capacity and resulted in a loss of social welfare when compared to the ATC MC solution. In order to avoid such an issue in the future, an additional check will be made.
In general, this is the kind of mistakes that can be expected in a parallel run where a new (FB) process is operated (tested) in parallel with the actual (ATC) process. When the operators are more familiarized with the new processes and tools, this kind of mistakes will not happen
On May, 18th, one TSO faced issues with its ECP server, which is responsible for providing input files as messages to the Flow Based Common System. As this TSO also provides input from the ATC Common System to the Flow Based Common System via ECP, these files, including LT nominations on some borders, were initially not available.
As a back-up solution, the files have been downloaded from the ATC Common System and adjusted for naming convention in order to manually upload the files to the Flow Based Common System.
After the upload of this file, the validation was successful. However, the file was empty and this resulted in errors in the computation to perform an ATC check, which prevented the results from being created. This was clearly an unexpected effect as the ATC check is not a mandatory step for the computation of FB parameters.
This issue will not occur after go-live as this file will be replaced by reference NTCs and therefore become obsolete. As a short-term mitigation measure, it will be ensured that in case of an unavailability of this input, the process will skip this check and continue with the provision of results. Parameters for this day have been re-computed as the correct file is available and will be published ex-post.
One TSO had issues to create a D2CF file as by accident a fix for its calculation module was installed on their production environment. This delivery failed and caused the inability to create and deliver a D2CF for the merging. As a back-up, an alternative reference day was used to perform the merge (used was the file of 21-05), but the merging resulted in high adjustment on load in order to compensate for the discrepancies linked to the changed date. Based on this, the quality of D2CF was disqualified by the TSO and he decided to not start final computation.
As of the usage of the next Flow Based Common System release in July, it can be chosen to remove the effected timestamps, which – depending on the number of timestamps removed – will either lead to spanning (if only 1 or 2 timestamps are missing) or to the provision of so-called “fall-back Flow Based Parameters” (fb-FBP).
More information on spanning or fb-FBP can be found in the section 4.6 of the Approval Document: http://www.casc.eu/en/Resource-center/CWE-Flow-Based-MC/Approval-Documents
In general, the decision not to go for normal business process due to bad D2CF quality is based on the violation of thresholds for adjustment values.
Please note that in this case, the re-computation is unfortunately not feasible due to the impact of D2CF on results of other TSOs, and therefore the complete process during this business day cannot be reproduced ex-post.
If such a situation happens after Go-Live, that is if some mandatory inputs are of too bad quality to run a normal process, TSOs will rely on the so-called fb-FBP to guarantee the market coupling process. Such situation will NOT provoke a decoupling. It is possible, even though not deliberately triggered, that such situation will be illustrated during the parallel run when the new version of the FB system is deployed as of July.
The CWE Flow-Based Market Coupling Project would like to inform you about changes in the parallel run reports publication format as of 13th May 2014.
From the launch of SWE full price coupling the 13th of May on (delivery day 14th May), Spanish (ES) and Portuguese (PT) market clearing prices will be included in the reports published as part of CWE FB parallel run. This will impact the column format of the overview_MCP.csv file and the tab MCP of the daily files report_yyyymmdd.xls.
During the short clock change a (temporarily) local tool provided wrong values for NTC for two specific timestamps (hour 22&23). Due to this, exceptional high PTDFs were created for these specific timestamps. As a consequence of a validation check, these hours were excluded from the results. It has been decided to publish this day ex-post, as the results are available and the impact on the results is both limited and only applicable on these two specific timestamps.
An issue occurred during the creation of the merged D2CF. There was a bug in the merging tool, which creates necessary input for the FB system, that resulted in the selection of the wrong reference day. The reference day for the 6th was 30th of March, the short clock change with only 23 hours. Consequently a new reference needed to be used, therefore the Saturday before was chosen. Unfortunately, at that time the system rejected the manual change by the operator. Due to this, incorrect results have been created for the 6th. Efforts were done to correct manually, but this was not feasible in time.
A fix for this issue has been implemented to ensure this issue will not occur again.
The encountered issues in the process for publication of simulation data for delivery date 9th April have been investigated and the explanation can be found in the excel table below. Unfortunately, there will be no ex-post publication of the results as the they are considered not representative.
Please click here for more information
We would like to inform you that due to technical issues the simulation results for delivery day 11/03 will not be published.
Further explanation will be given in the table below which is regularly updated
We would like to inform you that due to technical issues the simulation results for delivery day 28/02 will not be published. Further explanation will be given in the table below which is regularly updated
Due to switches of systems in preparation of the launch of the daily parallel run publication, the availability of Flow Based parameters results in the utility tool will only be available within the coming weeks.We apologize for the inconvenience.
Please be informed that the parallel run report for week 6 has been published today.
Please be informed that there will be no updates of the CWE parallel run publications during the holiday season.
As a consequence, parallel run results of weeks 51 and 52 will be published on 9 January together with the normal publication of week 1 (2014).
Please note that due to an unfortunate manual error, the parallel run report for week 45 was discovered to erroneously present “intuitive” FB results as “plain” FB results. The results have since been corrected.
The CWE FB project would like to remind you to update your copy for week 45 in case you maintain offline copies of the weekly parallel run reports.
We apologize for the inconvenience.
After the 1st quarter of 2013, CWE partners would like to provide an overview on the different types of errors which led to missing data for some days.
Please find hereafter a synthetic table displaying the missing days together with explanatory factors.
Please be aware that only representative days (successful simulations) are available.
Results of May 20th in week 20.
Please note that the results of May 20th published on Thursday May 23rdpresented an error that was identified only after the publication. Because of a wrong manual action in the (not yet) industrialised part of the process, a critical branch with very low PTDFs was kept instead of being detected and removed by the local operator. This critical branch led to unrealistic market results, not representative of the flow-based allocation results under a fully operational process.
As the results of the whole qualification/coordination process remain valid, we were able in this specific case to republish the market results that would have been obtained without this critical branch. On June 12th, the corrected market results were published. If you’ve already downloaded week 20, please replace it by this new file. We are sorry for any inconvenience this error may have caused.
The Flow-Based Market Coupling parallel run data is public and provided by CWE project partners for market parties and stakeholders. Please note that it can however not be disseminated or displayed without the project partners’ approval. Moreover, any commercial use of this data is also subject to project partners’ approval as the data is owned by the CWE project.