Running Time series simulation of the European Low Voltage Test Feeder with 3 phase asymmetric loads

[nikofou]

Hello, i am new to pandapower and i wanted to run a timeseries simulation for the European Low Voltage Test Feeder Distribution Network (https://cmte.ieee.org/pes-testfeeders/resources/) (https://pandapower.readthedocs.io/en/v2.9.0/networks/3phase_grids.html). As i was reading through the documentation and the tutorials, i noticed that when running the time series simulation a datasource file must be provided describing the changed in load or generation for each timestep. But as this network is a three phase asymmetric network what form does the datasource dataframe take? Is there a predefined template? Also for this particular network i tried using the data contained in the .zip file in the ieee site for the load profiles. Thank you in advance.

[elodin]

Hi @nikofou, welcome to pandapower.

To keep things uncomplicated, i'd recommend you use one datasource for each phase. Your DataFrames will look exactly the same as the ones in the tutorial - you only use three of them instead of one. That could look like this:

import numpy as np
import pandas as pd

import pandapower.control as control
import pandapower.networks as nw
import pandapower.timeseries as timeseries
from pandapower.timeseries.data_sources.frame_data import DFData

net = nw.ieee_european_lv_asymmetric(scenario='on_peak_566')

n_ts = 12

df_a = pd.DataFrame(np.random.normal(1., 0.1, size=(n_ts, len(net.asymmetric_load.index))),
                    index=list(range(n_ts)), columns=net.asymmetric_load.index) * net.asymmetric_load.p_a_mw.values

df_b = pd.DataFrame(np.random.normal(1., 0.1, size=(n_ts, len(net.asymmetric_load.index))),
                    index=list(range(n_ts)), columns=net.asymmetric_load.index) * net.asymmetric_load.p_b_mw.values

df_c = pd.DataFrame(np.random.normal(1., 0.1, size=(n_ts, len(net.asymmetric_load.index))),
                    index=list(range(n_ts)), columns=net.asymmetric_load.index) * net.asymmetric_load.p_c_mw.values

ds_a = DFData(df_a)
ds_b = DFData(df_b)
ds_c = DFData(df_c)

const_load_a = control.ConstControl(net, element='asymmetric_load', element_index=net.asymmetric_load.index,
                                    variable='p_a_mw', data_source=ds_a, profile_name=net.asymmetric_load.index)

const_load_b = control.ConstControl(net, element='asymmetric_load', element_index=net.asymmetric_load.index,
                                    variable='p_b_mw', data_source=ds_b, profile_name=net.asymmetric_load.index)

const_load_c = control.ConstControl(net, element='asymmetric_load', element_index=net.asymmetric_load.index,
                                    variable='p_c_mw', data_source=ds_c, profile_name=net.asymmetric_load.index)

ow = timeseries.OutputWriter(net, output_path="./", output_file_type=".xlsx")

ow.log_variable('res_line_3ph', 'loading_percent')

timeseries.run_timeseries(net, run=pp.runpp_3ph)

Edit: calling runpp_3ph() instead of runpp()

[elodin]

The load profiles in the zip file are made for the balanced version of the grid. In order to use them, you'd first need to distribute them over the three phases. Also keep in mind, that they don't contain information about the loads power, but instead information about the scaling of the load.

[nikofou]

Hello @elodin and thank you for your answer! As i am trying out things in pandapower, i wanted to change the topology of the given example (of the EUROPEAN LV test feeder) and add 3ph loads or 1ph loads and then run the ts simulation again. Is there any advice/tutorial on that?

[zuno1999]

Hi @nikofou, I am also trying to add 3-ph loads to the model. Do you have any suggestions?