There are three initialization options available for power-flow calculation in the drop-down menu.

• From input data: The values specified in the network input data file are used.

• Flat Start: The power-flow routine assumes the initial voltage magnitude of all PQ buses to be 1 p.u. and initial voltage angle of all buses to be 0 degrees.

• Smart Start: A linear approximation of the power-flow equations is solved beforehand to provide a good initial guess to the Newton-Raphson method.

Termination condition based on the tolerance of the mismatch equations (default 1E-8).

The algorithm terminates if the norm of the mismatch equation becomes smaller than the specified tolerance.

Maximum number of iterations allowed for Newton-Raphson to converge (default 100).

If the algorithm does not converge once this limit is reached, ePHASORSIM stops and the error message “Power-flow didn’t converge” is displayed.

Runs the power flow calculation.

Runs the power flow and also writes the results into following output files in a folder with name follows the format: '[XLS file name].PowerFlow'.

• 0_NetworkComposition.csv: Number of different components in the network (buses, machines, loads, etc).

• 1_Bus.csv:  Bus types, base voltage, voltage magnitude, and angle.

• 2_Machine.csv: Machine PQ and Q limits.

• 3_Statistics.csv: Number of NR iterations done and largest power mismatch.

• If checked, runs power-flow before running a dynamic simulation and the Solver initializes dynamic components based on voltage and power values calculated by power-flow.

• If unchecked, the voltages and power values in the input data are used for the initialization of the dynamic components.

The solver can automatically partition the admittance matrix into a block bordered diagonal format to perform parallel simulation. This uses a graph partitioning method where the number of desired blocks is defined by this parameter (a positive integral number).

Depending on the power system’s topology the value for this parameter varies. An ideal value for real-time performance can be sought using trial and error analysis. As a rule of thumb, the number of partitions should be 5-7 times the number of available CPU cores.

For example, with 10 cores available, the ideal value is likely between 50 and 70.

Refer to Parallel Simulation for more details.

Writes the network’s admittance matrix as a CSV file with a name that follows the format: '[XLS file name]_Y.csv'.

Displays instruction to get more logs and verbose output.

Finds the optimal number of threads (in terms of run-time performance) to be used for the parallelization of both network solver and integration of genunits. Refer to Parallel Simulation for more details.