The phase at BUS9 (solution vector: node00007) is not initialized properly from the CIM powerflow data. Therefore, the steady-state initialization implemented in the MNA solver is executed first so that the initial phase at t=0 is already very close to the correct one.
Using PI-lines with parallel impedances to ground not only during the initialization but also the simulation or more accurate CIM load flow results would probably further improve the results because the values around BUS9 have still not converged completely.
%% Cell type:code id: tags:
``` python
pt.plot_timeseries(1,phasors['v9']['phase'])
plt.ylim([2.10405,2.1042])
```
%% Cell type:markdown id: tags:
The next plot is showing the same value, BUS9 phase during the initialization. The phase oscillation is larger than in the actual simulation.
We use Graphviz to render the network model into a layouted SVG figure
%% Cell type:code id: tags:
``` python
system
```
%% Cell type:markdown id: tags:
### Extending Network with Switch and Load
%% Cell type:markdown id: tags:
Here we add a new switch to Bus 9 that is triggered at 0.05s. This component could be added permantely to the CIM model as well using tools like Pintura.
%% Cell type:code id: tags:
``` python
## Switch
sw=dpsim.dp.ph1.Switch("Switch")
sw.R_open=1e9
sw.R_closed=0.1
sw.is_closed=False
## Load
load=dpsim.dp.ph1.PQLoadCS("Switched Load")
load.V_nom=230950
load.P=30000000
load.Q=0
## Topology
bus9=system.nodes["BUS6"]
gnd=dpsim.dp.Node.GND()
sw.connect([bus9,gnd])
system.add_component(sw)
```
%% Cell type:code id: tags:
``` python
system
```
%% Cell type:markdown id: tags:
### Running Simulation
The actual simulation is done by the C++ DPsim solver. Python is just used for configuration, scripting and analysis