Members of Challenge 2 teams may submit algorithms to the GO Competition Platform’s Sandbox using the datasets described below. If you do not see the blue Submit button, you are not part of a Team.

To participate in the Trial or Final Events, the team leader must also submit Team and Individual Registration documentation from the leader's Account (login) page.

Challenge 1 teams who haven't done so already need to reconstitute themselves for Challenge 2, i.e., the team leader needs to select the new members of the team.


Download (36.5 MB) the 11 synthetic networks composed of 63 scenarios used in Challenge 2 Trial  Event 1. The 4 industry datasets, composed of 20 scenarios, will not be made publicly available.

Any of the Trial 1 scenarios, including industry ones, may be executed individually in the Sandbox by putting the following in submission.conf




model scenarios Type
C2T1N00403 218,219,222,256,300  Industry
C2T1N00500 45,47,52,56,103 Synthetic
C2T1N00617 6,33,36,42,47 Synthetic
C2T1N00768 9,29,44,80,177 Synthetic
C2T1N02020 8,34,3856,70 Synthetic
C2T1N02055 3,8,21,32,34 Synthetic
C2T1N02312 1,142,160,254,278,370,416  Synthetic
C2T1N03411 11,134,140,145,149  Industry
C2T1N04441 1,12,221,224,292 Industry
C2T1N05746 6,11,15,25,29 Synthetic
C2T1N06469 3,7,16,41,42 Synthetic
C2T1N08028 1,2,13,16,26,27,30 Synthetic
C2T1N14204 2,12,13,27,31,31 Synthetic
C2T1N15810 2,13,16,34,41,46,56,59 Synthetic
C2T1N16955 3,4,5,6,8 Industry


Download (56.5 MB) the Version 7 Sandbox (C2S7N*) datasets; 16 models, 80 scenarios posted March 5, 2021. These synthetic datasets represent the kinds of problems that will be used in Trial 2, with the exception of the industry datasets. A spreadsheet listing the characteristics of each of the 80 scenarios is also available.  The differences between Divisions 1 & 3 and between Division 2 & 4 will represent the impact of line switching. None of the Version 7 datasets are exactly the same as the corresponding Version 6 datasets. Be sure to download the latest Evaluation code, described at the bottom of the Evaluation page, to score your own results.

Note on scenarios with negative values of zinf: Scenarios with a negative zinf(defined in section D.2 of the Problem Formulation) are possible and generally reflect high penalties imposed on the solution by a poor prior operating point. The GO Competition does not intend to use such scenarios in any Trial or Final Events. There were, however, 3 such scenarios in C2S7N*, specifically, C2S7N12209 scenario 4 (-16,673,022) and scenario 8 (-99,958) and C2S7N14212 scenario 3 (-13,205,325,795). The first two are the result of legitimate, but poor, prior operating points. The later, however, was the result of an error in constructing the data that resulted in a bad power balance. This scenario has been replaced with one with a zinf value of 7,610,313. The only difference between the new scenario and the old is better values in the raw file for the VM and VA parameters in the Bus Data block and PG and QG parameters in the Generator Data block. The new scenario replaces C2S7N14212 scenario 3 when making Sandbox submissions after March 12, 2021.

Please specify C2S7Nxxxxx (or C2S6Nxxxxx for Version 6) as the model along with dataset=C2_Sandbox and the scenario number  in submission.conf when making Sandbox submissions.

Download (665 kB) the Sandbox network model (C2S6N00617) with 10 scenarios. Added February 1, 2021. This new sandbox dataset has gone through more rigorous testing and validation to ensure that the new features included in Challenge 2 will have an impact on the best solutions.

Download (13.1 MB) the Sandbox set of 12 network models and 38 scenarios. Caution: The zip file from 11/20 padded the end C2S6N11152/scenario_001/case.json with many null bytes for unknown reasons. Do not use it.

This version (C2S6N*, posted Nov. 22, 2020) fixes some cost function and transformer tap issues and an error introduced in C2S5N*. It includes a csv file with details about each of the 38 scenarios (refererences to C2S5 should be C2S6). Fixing the transformer taps changed the number of buses in some network models so the naming is slightly different from the C2S4N* version as are the zinf  values given in the csv file. 


Use the submission.conf file described in Languages to specify necessary runtime parameters including software modues, dataset, network model, and scenario number. Copy and paste into the submission text box.

Example submission.conf file using python 3.7 with Intel Math Kernel Library (mkl) and compilers (replace MyRepo with your value):





source /share/apps/intel/2019u4/mkl/bin/ intel64

source /share/apps/intel/2019u4/compilers_and_libraries/linux/bin/ -arch intel64




DO NOT include solution files in your GitHub repository.