Skip to content

STEPSS v3.84

Choose a tag to compare

@apetros apetros released this 09 Nov 20:02

STEPSS v3.84

Rebuilt for RAMSES 3.84 (was 3.83), URAMSES 3.84 (was 3.83).

Bundled components

Component Version Upstream release Published
RAMSES 3.84 v3.84 2026-11-09
Helios 1.5.0 v1.5.0 —
DYNGRAPH 1.4.0 v1.4.0 —
CODEGEN 5.5 v5.5 —
URAMSES 3.84 v3.84 2026-11-09

Bundled examples

Example Version Upstream release Refreshed
Kundur two-area 1.4.0 v1.4.0 —
IEEE Nordic 1.3.0 v1.3.0 —
5-bus tutorial 1.2.0 v1.2.0 —
6-bus microgrid 1.1.0 v1.1.0 —

Upstream release notes

RAMSES v3.84 — v3.84 - network frequency update scales line and transformer reactances correctly

With $NET_FREQ_UPD T, line and transformer reactances now follow the network frequency in the right direction, and reported branch flows match the network solution.

This option rescales the network admittances at every time step with the per-unit centre-of-inertia speed ν of each island. The series reactance of lines and transformers was rescaled the wrong way: an inductive branch received X/ν instead of νX, and a series capacitor νX instead of X/ν. The network therefore became stiffer during an over-frequency and weaker during an under-frequency, the reverse of the physical effect. Shunt susceptances and impedance loads were already rescaled correctly.

Series reactances now scale as νX when inductive and X/ν when capacitive, with the resistance unchanged. On the Kundur two-area system, after the trip of one generator leaves the frequency 2 % below nominal, the final frequency moved by +0.046 Hz with the previous rule and moves by −0.020 Hz now, relative to a run without the option, and the inter-area angle now decreases, as a stiffer network should make it, instead of increasing.

Branch flows reported as observables, shown in the run-time display, or read by discrete controllers (phase-shifting transformer and line protection) did not take the option into account, and differed from the network solution by about 2|ν−1| of their value: 16 MW on a transformer carrying 813 MW in the case above. They are now computed with the same rescaled admittances as the network solution.

Simulations with $NET_FREQ_UPD F, the default, are unchanged and give the same trajectories as 3.83. The rescaling rules are described under Network Frequency Update.

WECC wind turbine models. WT3 and WT4 now start in steady state and run through a solid fault. Their REEC_A reactive-power and voltage control chain left several signals inconsistent at t = 0, so RAMSES reported the injectors as not in steady state and a limiter switched on and off from the first time step; a zero-impedance fault then produced a NaN in the network mismatch. All four QFlag/vFlag combinations and PFFlag = 1 now initialise without residual. Setting Tiq, Tpord or Tg below 0.005 s now passes the signal through, where it previously halved it, or with Tg below 0.001 s held the active current at zero. An out-of-range REEC_A limit now ends the run with a message naming the injector, instead of stopping the host process.

Plot step of zero. $PLOT_STEP 0, which records the observables at every time step, could crash a simulation on Linux and Windows, typically with a segmentation fault shortly after the first disturbance. The observables buffer was sized by dividing by the plot step, and on those platforms the result became a negative size. The buffer is now sized correctly for any plot step, and $PLOT_STEP 0 runs normally on all three platforms.

No data file changes are needed.

URAMSES v3.84 — v3.84 - network frequency update scales line and transformer reactances correctly

Toolchain and kit provenance

The modules_*/ directories in this release ship pre-compiled RAMSES
libraries and .mod files. They are committed to the repository, so the
Source code archive below (or a clone checked out at this tag) is the
whole release; there are no separate kit downloads to fetch and unpack.

A .mod file can only be read by the gfortran generation that wrote it, so
to add your own Fortran models you need a compiler emitting the same module
ABI version as your platform's kit. make -f <makefile> check-deps verifies
this before building.

Makefile Kit directory Built with .mod ABI Target Runtime floor
build/Makefile.linux modules_l GNU Fortran (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0 15 x86_64-linux-gnu ldd (Ubuntu GLIBC 2.39-0ubuntu8.9) 2.39
build/Makefile.macos modules_m GNU Fortran (Homebrew GCC 16.2.0) 16.2.0 16 aarch64-apple-darwin24 macOS 15.8.2 arm64; needs brew install gcc openblas
build/Makefile.windows modules_wg GNU Fortran (Rev4, Built by MSYS2 project) 16.2.0 16 x86_64-w64-mingw32 static - self-contained, no MSYS2 runtime required

Exact compile flags, BLAS build and RAMSES commit for each kit are recorded in
that kit's BUILDINFO.txt, for example modules_l/BUILDINFO.txt.

The Windows/Intel kit in modules_wi/ (used by build/msvs/URAMSES.sln) is
not refreshed by this release: no CI produces Intel-format modules. See
modules_wi/BUILDINFO.txt for what it was last built from.


With $NET_FREQ_UPD T, line and transformer reactances now follow the network frequency in the right direction, and reported branch flows match the network solution.

This option rescales the network admittances at every time step with the per-unit centre-of-inertia speed ν of each island. The series reactance of lines and transformers was rescaled the wrong way: an inductive branch received X/ν instead of νX, and a series capacitor νX instead of X/ν. The network therefore became stiffer during an over-frequency and weaker during an under-frequency, the reverse of the physical effect. Shunt susceptances and impedance loads were already rescaled correctly.

Series reactances now scale as νX when inductive and X/ν when capacitive, with the resistance unchanged. On the Kundur two-area system, after the trip of one generator leaves the frequency 2 % below nominal, the final frequency moved by +0.046 Hz with the previous rule and moves by −0.020 Hz now, relative to a run without the option, and the inter-area angle now decreases, as a stiffer network should make it, instead of increasing.

Branch flows reported as observables, shown in the run-time display, or read by discrete controllers (phase-shifting transformer and line protection) did not take the option into account, and differed from the network solution by about 2|ν−1| of their value: 16 MW on a transformer carrying 813 MW in the case above. They are now computed with the same rescaled admittances as the network solution.

Simulations with $NET_FREQ_UPD F, the default, are unchanged and give the same trajectories as 3.83. The rescaling rules are described under Network Frequency Update.

WECC wind turbine models. WT3 and WT4 now start in steady state and run through a solid fault. Their REEC_A reactive-power and voltage control chain left several signals inconsistent at t = 0, so RAMSES reported the injectors as not in steady state and a limiter switched on and off from the first time step; a zero-impedance fault then produced a NaN in the network mismatch. All four QFlag/vFlag combinations and PFFlag = 1 now initialise without residual. Setting Tiq, Tpord or Tg below 0.005 s now passes the signal through, where it previously halved it, or with Tg below 0.001 s held the active current at zero. An out-of-range REEC_A limit now ends the run with a message naming the injector, instead of stopping the host process.

Plot step of zero. $PLOT_STEP 0, which records the observables at every time step, could crash a simulation on Linux and Windows, typically with a segmentation fault shortly after the first disturbance. The observables buffer was sized by dividing by the plot step, and on those platforms the result became a negative size. The buffer is now sized correctly for any plot step, and $PLOT_STEP 0 runs normally on all three platforms.

No data file changes are needed.

Artifact

stepss.jar — SHA-256 a137e66caf44c180e883196f304675c8538f726c9786723c09f10358a04abb31