Format#

The ANDES config file uses the format provided by Python module configparser. The syntax is like the following:

[System]
freq = 60
mva = 100
...

[PFlow]
sparselib = klu
linsolve = 0
tol = 1e-6
...

[TGOV1]
allow_adjust = 1
adjust_lower = 0
adjust_upper = 1

In the above, System, PFlow and TGOV1 are two sections. freq = 60, for example, is a pair of option and value in the [System] section. Note the space before and after the equal sign.

The meaning of the fields in each section can be found in Config reference, which contains the default values and acceptable values for each option. The values for config fields can be a string or a number. Fields with acceptable values being (0, 1) can only accept 0 or 1 to indicate true or false. Non-binary values for such options will cause unexpected errors in the program.

Limits in models#

All models have three config options:

  • allow_adjust: allow limits of limiters in this model to be adjusted if the inputs, at steady state, is out of the limits. allow_adjust = 0 is the global off-switch for this model.

  • adjust_lower: allow reducing the lower limit to the input value, if the input at steady-state is below the lower limit. This is disabled by default.

  • adjust_upper: allow increasing the upper limit to the steady-state input. This is enabled by default.

Note that setting allow_adjust = 0 is equal to setting adjust_lower = 0 and adjust_upper = 0, but the former saves some time for function calls.

The limit adjustment feature is to alleviate issues caused by the model parameters. Commercial tools have a more sophisticated mechanism for autocorrection, and the limit adjustment is part of it. However, if you see an limit adjustment warning or even initialization error, it is important not to rely on autocorrection but fix the data by yourself. Autocorrected data can yield some results but issues can remain hidden.