Frank Tuffner

Hello MA, Apologies for the delay getting back to you. Regarding P_Out, that is a set point. From the GridLAB-D perspective, anything that you can use a player on, you can also use HELICS on (properly configured, of course). Since you mentioned testing the same code with a triplex_load worked, that should at least rule out the Python. Looking at your inverter, I suspect it might just be over-writing and/or ignoring that value because I don't think you have the inverter set up to do exactly what you're...

Hello Jett, That is very strange. You are correct that it doesn't seem to be implemented anywhere. I'm trying to track down if that is a missed commit or something (amusing the the wiki entry is almost three years old and no one else has said anything). For now, you are correct that the inverter_dyn does not have a VSM mode enabled. I have notes out to some of the developers to see what happened here -- we clearly wrote up a specification and example, so I'm not sure where the code went. -Frank

Hello Jett, Apologies for the delay in responding. You're pretty close, but there's a couple minor items/nuances to include still. The first item is to either add flags DELTAMODE; to all of the powerflow objects, or just add to your module powerflow the all_powerflow_delta flag: module powerflow { enable_subsecond_models true; deltamode_timestep 10000000; //10 ms solver_method NR; all_powerflow_delta true; } You have most of them in it, but the transformer missing does appear to cause issues with...

Hello Jett, It looks like your file might have a couple issues, so that's probably most of the problem. If you look at your bus_X_output_recorder.csv files, you notice they turn to NAN values pretty quick, which is not a good sign. If you put a recorder on the status of your switch, you'll see that it starts OPEN and never closes. You'll need to fix the timesteps (use one of the recorder outputs as a guideline). The inverter_dyn object appears to be overloaded since the active power setpoint is the...

Hello again Shekiell, Based on your installation.txt and looking at how the python libraries are set up on my side, I think I see the issue -- it's related to how cmake and the Python libraries are interacting (it basically hates the "lib64" instead of "lib" folder). So you have a couple options: 1. Download the HELICS source code and compile it directly (then link the cmake for GridLAB-D to wherever you install it) 2. Copy the Python HELICS lib64 folder to lib (linking it might work too, I didn't...

Hello Shakiell, The "../install64" folder that is part of the -DCMAKE_INSTALL_PREFIX=../install64 command is just dictacting where the compiled GridLAB-D gets installed. The ".." just means "up one folder", so it would be in "gridlab-d/install64" folder (assuming you run this from "gridlab-d/cmake-build" or similar). If you leave the -DCMAKE_INSTALL_PREFIX item off, I believe it just will installed into the /usr/ path on the Linux machine, but I'm not 100% sure there. I run multiple instances/versions...

Hello Shakiell, You are correct that the "HELICS_FOCAL_361" is pointing to the HELICS installation - I should have been clear that was just an example from my computer. That folder is specifically looking for a four-folder structure with "bin", "include", "lib", and "share". It's mostly looking for the helics.h and similar files that are in the include/helics folder, as well as the libhelics.so file that is under the lib folder. Depending on how you installed HELICS, finding those libraries may be...

Hello Shakiell, That cmake command is likely it. I'll also note there's a semi-half-hidden step at the beginning of the GridLAB-D compile. Right before the git submodule command, you should ideally check out the release branch (master should be aligned, but that ensures you are getting the latest release): cd gridlab-d git checkout v5.3.0 git submodule update --init However, I suspect the lack of -DHELICS=ON and -DGLD_HELICS_DIR setting are the primary reason you're having issues. -Frank