Test Cases

The systems that ship with the package live in hermess/systems. Each one is a folder holding a sim_param.txt (network, devices, initialization) and a sim_dist.txt (disturbances), described in System files.

The list is available at runtime, so it never goes out of date:

import hermess

hermess.list_systems()
dae = hermess.simulate("ieee39", T_end=10.0)

Small systems

3bus

Three buses and three lines: a Sauer-Pai synchronous machine at bus 1, a grid-forming converter at bus 3 and a ZIP load at bus 2. The smallest system that still contains both a machine and a converter, which makes it the usual starting point and the one used by most of the test suite. Disturbance: a line opening.

3bus_loadstep

The same three-bus network with a single small load step instead of the line opening: a minimal setpoint-style event that stays in the linear regime while still exciting the grid-forming converter’s power and frequency response.

3bus_genrou

A three-bus triangle with two round-rotor GENROU machines (buses 1 and 3) and a constant-impedance load (bus 2), with excitation and turbine dynamics deliberately excluded (AVRCONST / GOVCONST). This is the cross-tool validation case for the GENROU machine model, compared against ANDES in CI; see Validation. Disturbance: a line opening.

3bus_gensal

The same network with two salient-pole GENSAL machines: the cross-tool validation case for GENSAL (see Validation). Disturbance: a line opening.

3bus_tgov1

The GENROU pair with a TGOV1 turbine governor closed around each machine and no other speed feedback (D = 0): the cross-tool validation case for TGOV1 (see Validation). Disturbance: a line opening, which leaves a power imbalance for the governors to pick up.

3bus_sexst

The GENROU pair with the simplified static exciter SEXST closed around each machine: the cross-tool validation case for SEXST (see Validation). Disturbance: a line opening, whose voltage step exercises the exciters.

3bus_avrst1a

The GENROU pair with the ST1A static exciter (transducer, transient-gain-reduction lead-lag, regulator lag): the cross-tool validation case for AVRST1A (see Validation). Disturbance: a line opening.

3bus_ieeedc1a

The GENROU pair with the IEEE DC1A rotating exciter (Sauer-Pai parameter set): the cross-tool validation case for IEEEDC1A (see Validation). Disturbance: a line opening.

3bus_avrac1a

The GENROU pair with the AC1A exciter in its small-signal form, at a high-gain fast-regulator parameter set: the cross-tool validation case for AVRAC1A (see Validation). Disturbance: a line opening.

3bus_psskundur

The GENROU pair with the SEXST exciter and the Kundur speed-input stabilizer: the cross-tool validation case for PSSKundur (see Validation). Disturbance: a line opening, whose rotor swings exercise the stabilizer path.

3bus_genrou_psse

The three-bus PSS/E benchmark system (60 Hz): an infinite bus, one saturation-free GENROU, and a constant-impedance load, transcribed from the ThreeBusMulti case of the PowerSimulationsDynamics.jl benchmark set. The cross-tool validation case against a PSS/E-produced trajectory (see Validation); run it with fn=60 and T_end=20. Disturbance: a line opening.

3bus_gfm_psid

Ideal source, droop grid-forming converter (D’Arco parameter set) and constant-impedance load: the cross-tool validation case for GridForming against PowerSimulationsDynamics.jl (see Validation). Disturbance: a load step.

3bus_sauerpai_psid

Ideal source, the six-state Sauer-Pai machine with constant excitation and mechanical power, and a constant-impedance load: the cross-tool validation case for SynchronousSubtransientSP against PSID (see Validation). Disturbance: a load step.

3bus_shaft5mass_psid

The Sauer-Pai machine on the five-mass torsional shaft HP-IP-LP-GEN-EXC (Sauer-Pai torsional data): the cross-tool validation case for Shaft5Mass against PSID (see Validation). Disturbance: a load step, whose electrical-torque step rings the torsional modes.

3bus_dynlines_psid

The Sauer-Pai system with the network itself dynamic (run it with line_dyn=True): every line current and bus voltage is a differential state. The cross-tool validation case for the dynamic network against PSID (see Validation). Disturbance: a load step.

omib_gfm_pscad

The one-machine-infinite-bus system of the PSCAD Test23 benchmark: the full D’Arco droop grid-forming converter (active damping, split power-filter corners) against an ideal source. Validated against both PSID and the PSCAD electromagnetic trajectory (see Validation). Disturbance: a reference-power step (the SETPOINT event).

omib_vsm_pscad

The OMIB of the PSCAD Test08 benchmark: the virtual-synchronous-machine converter (angle = "VSM" with the Kaura PLL and damped inner control). Validated against both PSID and PSCAD (see Validation). Disturbance: a reference-power step.

omib_gfl_pscad

The OMIB of the PSCAD Test24 benchmark: the current-injecting GridFollowing converter (PI power outers, current-mode inner, reduced-order PLL). Validated against PSID exactly and compared against the PSCAD trace (see Validation). Disturbance: a reference-power step.

3bus_marconato_pscad

The three-bus system of the PSCAD Test25 benchmark (60 Hz): two Marconato machines (Type II governor and fixed torque, AVRSimple exciters) with every line dynamic (run it with line_dyn=True and fn=60). Validated against PSID and the PSCAD voltage trajectory (see Validation). Disturbance: a reference-power step.

kundur

The Kundur two-area system, twelve buses with four subtransient machines and three ZIP loads, exercising the automatic voltage regulators. Disturbance: a load step. The classic system for inter-area oscillation studies.

kundur_conv

The two-area system with two of the four machines replaced by grid-forming converters, for comparing converter-dominated against machine-dominated dynamics on an otherwise identical network. Disturbance: a load step.

IEEE 39-bus systems

ieee39

The standard IEEE 39-bus (New England) benchmark: 39 buses, 46 branches, ten subtransient synchronous machines and nineteen ZIP loads. It is simulated in the time domain to study the electromechanical response to a sequence of disturbances:

• short circuit on branch 5-8 (t = 7.00 s),

• fault cleared by opening the branch at both ends (t = 7.04 s),

• load change at bus 5 (t = 9.00 s).

Figure 1: The IEEE 39-bus network.

ieee39_conv

A converter-penetrated variant of the same network: five Sauer-Pai synchronous machines, two grid-forming and three grid-following converters, nineteen ZIP loads. Disturbance: a bus fault and its clearing.

Figure 2: The IEEE 39-bus network with grid-forming and grid-following converters.

ieee39_ideal

The 39-bus network with constant-power loads instead of ZIP loads and a longer disturbance sequence: a line fault followed by the line opening, a bus fault and its clearing, and a load step. Inherited from the parent project, where it served as the idealized-measurement scenario, and kept as a functional test that the simulator reproduces the expected response.

ieee39_conv_old

An earlier converter-penetrated variant, kept for comparison with results produced before the converter models were reworked. Prefer ieee39_conv for new work.

The 14-generator South East Australian system

A 59-bus, 76-branch, five-area, 50 Hz benchmark with fourteen aggregated power stations and five static var compensators, from M. Gibbard and D. Vowles, Simplified 14-Generator Model of the South East Australian Power System, University of Adelaide, revision 4, 2014. The parameter tables are transcribed into sea_data.py and sea_dynamics.py; the report itself is not redistributed.

The case folders are generated by build_sea_system.py and can be validated against the published load flow and rotor modes with validate_sea.py. Both are repo-only developer tools (run from a source checkout, not shipped in the wheel), because they write the case folders next to themselves:

python hermess/systems/sea14gen/build_sea_system.py 1
python hermess/systems/sea14gen/validate_sea.py 1
python hermess/systems/sea14gen/build_sea_system.py 1 --no-pss

sea14gen/case1

Operating case 1 of the benchmark: twelve round-rotor (GENROU) and two salient-pole (GENSAL) machines with ST1A and AC1A exciters and speed stabilizers, five SVCs and thirty-two ZIP loads. Disturbance: a bus fault and its clearing.

sea14gen/case1_nopss

Operating case 1 with the power system stabilizers removed, which is how the benchmark exposes its poorly damped inter-area modes.

sea14gen/case1_conv

Operating case 1 with three of the stations replaced by converters, two grid-forming and one grid-following, keeping the SVCs and the load pattern.

sea14gen/case2 … sea14gen/case6

The remaining five published operating conditions of the benchmark, spanning heavy to light system loading, each with the exciter and stabilizer settings of the report and a _nopss variant with the stabilizers removed. All folders are generated by the current build_sea_system.py and validated against the published load flow and rotor modes with validate_sea.py. Disturbance in every case: a 100 ms bus fault and its clearing.

Case	Total generation	Total load
1	23030 MW	22300 MW
2	21590 MW	21000 MW
3	25430 MW	24800 MW
4	15050 MW	14810 MW
5	19060 MW	18600 MW
6	14840 MW	14630 MW

sea14gen/smib

A single GENROU machine against an infinite bus, used to cross-check the machine implementation against the benchmark equations in isolation. No disturbance.