GridPACK v3.3 Release

As of GridPACK release 3.3, the minimal allowable version of CMake that is necessary to build GridPACK is CMake 3.3.0. Please update to a version of CMake greater than 3.3.0 on any platform that you are using to build GridPACK. On some platforms, newer versions of CMake are accessed using the command “cmake3” instead of “cmake”. This is discussed further in the build documentation.
A summary of updates to GridPACK is given below:
• The build documentation has been completely reviewed and revamped to make it simpler to build GridPACK. On most platforms, explicitly building libraries such as boost and MPI has been replaced by downloading modules using tools such as yum or get apt-get. We have made every effort to clarify the build process and reduce the number of steps required to get a working installation of GridPACK up and running. If you encounter any problems, feel free to contact us directly, either by emailing us at gridpack.account@pnnl.gov or by creating an issue at https://github.com/GridOPTICS/GridPACK/issues. Some of the improvements in the build system have required us to move to CMake 3.3.0 or greater so users should plan on upgrading their version of CMake. This should only require minimal changes to existing build scripts.
• We have added a new StatBlock module to GridPACK that is designed to gather the results of contingency analysis calculations into a single large distributed arrays and to perform elementary analyses on the results. The power flow contingency analysis application has been extended to use this functionality and now provides statistics that have been gathered over all contingencies in the system. For large power grid networks, this can represent 10’s of thousands of separate calculations. The European Open Model, which is included as a test case for the current contingency analysis application, has over 20,000 contingencies for a network with over 10,000 buses. Users can get a complete description of the available output for the contingency analysis application by looking at the README.md file in the contingency analysis application directory. Data that is collected from contingency analysis calculations includes
o A performance index calculation of the severity of each contingency
o RMS fluctuations of the voltage magnitude and phase angle for each bus across all contingencies
o The maximum and minimum voltage magnitude and phase angle for each bus across all contingencies
o RMS fluctuations in the real and reactive power for each generator, as well as the maximum and minimum values, across all contingencies
o The total number of faults found on each line across all contingencies
• New instructions for building GridPACK on MacOS High Sierra have been added to the web page.
• Static libraries are no longer required for GridPACK builds. This may be useful for interfacing with other languages or libraries, such as Python.
Bug Fixes:
• Isolated buses in contingency calculation were not getting properly reset for the next calculation (if the isolation was caused by removing a line from the network). This has been fixed so that all lines are reset to the same status as in the base case after each contingency.
• Matrix operations on complex matrices using an underlying build of PETSc based on real matrices were failing because the partitioning of the rows and columns of the real and imaginary parts was not necessarily the same. This has been fixed.
• Several problems with the dense matrix multiply were fixed, including unnecessary replication of communicators and a significant memory leak. This lead to improved performance and more robustness in the Kalman filter application.