Enabling Power Network Studies with Geo-referenced Distributed Energy Resources: A Swiss Database

General Overview

This dataset provides the allocation of distributed energy resources (DERs) to the medium-voltage (MV) and low-voltage (LV) distribution grids, synthetically generated for Switzerland by Alfredo Oneto et al. in [https://doi.org/10.1016/j.segan.2025.101678]. The primary objective of this dataset is to complement the existing information on distribution grids, which includes network topologies, line attributes, and nodal peak powers, by incorporating a detailed node-level allocation of DERs. The dataset specifically includes photovoltaic (PV) systems, electric vehicles (EVs), heat pumps (HPs), and battery energy storage systems (BESS), along with yearly hourly-resolved profiles for conventional loads and DER attributes.

Data is provided for three reference years: 2030, 2040, and 2050. This dataset, in combination with the synthetic distribution networks, enables large-scale, high-resolution studies on power distribution network operations and expansion planning. Moreover, the DER data can also be used independently for various applications related to energy systems modeling and analysis.

The dataset was produced using publicly available open data and follows the methodology described in the article "[not yet published]." It provides DER allocations and hourly resolved profiles for all LV and MV nodes of the Swiss synthetic power distribution grids (PDGs) dataset.

Data Description

The dataset is structured into five main folders, corresponding to four DER categories (PV, HP, EV, BESS), conventional load profiles (Demand), and the power distribution networks (PDGs). This PDG data is available in folder 06_Grids. Each folder contains three subdirectories named 2030, 2040, and 2050, which house data relevant to the respective reference year. Additionally, the dataset includes a Python script that serves as an example of how to load grid and DER data and utilize them for power distribution grid operation optimization.

01_PV (Rooftop Photovoltaic Systems)

This folder contains data on rooftop photovoltaic (PV) systems connected to the MV and LV grids for each reference year. Each subfolder (2030, 2040, 2050) includes six files:

LV_generation.csv: Contains the average PV generation profiles for each LV node in the PDG dataset. Specifically, it provides the yearly PV generation profile (kW) in the form of 12 representative days (one per month) with hourly resolution, for each low-voltage grid and low-voltage node. Nodes absent from this file do not have PV installations. The fields include:

LV_grid: Unique identifier of the low-voltage grid.

LV_osmid: Unique identifier of the low-voltage node in the grid.

Hourly time-series of PV average generation for the 12 representative days, expressed in kilowatts (kW).

LV_P_installed.csv: Reports the nominal power of the PV systems (expressed in kWp) installed at each LV node. The fields include:

LV_grid: Unique identifier of the low-voltage grid.

LV_osmid: Unique identifier of the low-voltage node in the grid.

P_installed_kW: Nominal installed PV capacity in kWp

LV_std.csv: Provides the standard deviation of the representative PV generation profiles for each LV node, reporting the hourly standard deviation over the 12 representative days. Nodes absent from this file do not have PV installations. The fields include:

LV_grid: Unique identifier of the low-voltage grid.

LV_osmid: Unique identifier of the low-voltage node in the grid.

Hourly time-series of PV generation standard deviation for the 12 representative days, expressed in kilowatts (kW).

MV_generation.csv: Contains the average PV generation profiles for each MV node in the PDG dataset. Like LV_generation.csv, this file provides the yearly PV generation profile (kW) in the form of 12 representative days (one per month) with hourly resolution, for each medium-voltage grid and medium-voltage node. Nodes absent from this file do not have PV installations. The fields include:

MV_grid: Unique identifier of the medium-voltage grid.

MV_osmid: Unique identifier of the medium-voltage node in the grid.

Hourly time-series of PV average generation for the 12 representative days, expressed in kilowatts (kW).

MV_P_installed.csv: Reports the nominal power of the PV systems (expressed in kWp) installed at each MV node. Nodes absent from this file do not have PV installations. The fields include:

MV_grid: Unique identifier of the medium-voltage grid.

MV_osmid: Unique identifier of the medium-voltage node in the grid.

P_installed_kW: Nominal installed PV capacity in kWp

MV_std.csv: Provides the standard deviation of the representative PV generation profiles for each MV node, reporting the hourly standard deviation over the 12 representative days. Nodes absent from this file do not have PV installations. The fields include:

MV_grid: Unique identifier of the medium-voltage grid.

MV_osmid: Unique identifier of the medium-voltage node in the grid.

Hourly time-series of PV generation standard deviation for the 12 representative days, expressed in kilowatts (kW).

02_HP (Heat Pumps)

This folder contains data on heat pumps (HPs) connected to the MV and LV distribution grids for each reference year. The heat pump data are provided in terms of the parameters of a first-order thermal circuit model, describing the thermal conductance and thermal capacitance of the buildings, along with temperature profiles. The consumption profiles should be derived by applying an operational strategy defined by the user (e.g., constant temperature, minimum cost). The explicit thermal modeling of buildings enables the quantification of HP flexibility from a thermal standpoint, allowing its full integration into power system flexibility studies.

Each reference year includes the following three files:

LV_heat_pump_allocation.csv: Reports, for each low-voltage network and each low-voltage node, the characteristics of the heat pumps connected to that node, as well as the thermal properties of the served buildings aggregated at the nodal level. The fields include:

LV_grid: Unique identifier of the low-voltage grid.

LV_osmid: Unique identifier of the low-voltage node in the grid.

Nominal_power_kW: Nominal electrical power of the heat pumps connected to the node.

Thermal_capacitance_KWh/K: Thermal capacitance of the served buildings.

Thermal_conductivity_kW/K: Thermal conductivity of the buildings connected to the node.

COP: Coefficient of performance of the heat pump in full load conditions.

Temperature_profile_name: Identifier pointing to the corresponding temperature profile in the Temperature_profiles.csv file.

LV nodes absent from this file do not have heat pumps connected.

MV_heat_pump_allocation.csv: Similar to LV_heat_pump_allocation.csv, this file reports the characteristics of heat pumps connected at medium-voltage nodes. The fields include:

MV_grid: Unique identifier of the medium-voltage grid.

MV_osmid: Unique identifier of the medium-voltage node in the grid.

Nominal_power_kW: Nominal electrical power of the heat pumps connected to the node.

Thermal_capacitance_KWh/K: Thermal capacitance of the served buildings.

Thermal_conductivity_kW/K: Thermal conductivity of the buildings connected to the node.

COP: Coefficient of performance of the heat pump in full load conditions.

Temperature_profile_name: Identifier pointing to the corresponding temperature profile in the Temperature_profiles.csv file.

MV nodes absent from this file do not have heat pumps connected.

Temperature_profiles.csv: Provides the temperature profiles referenced in the allocation files. The fields include:

Temperature_profile_name: Unique temperature profile identifier.

Yearly, hourly resolved, ambient temperature profile (in degrees Celsius) obtained from real data recorded at weather stations in Switzerland in 2024.

03_BESS (Battery Energy Storage Systems)

This folder contains data on battery energy storage systems (BESS) co-installed with photovoltaic (PV) plants and connected to the medium-voltage (MV) and low-voltage (LV) distribution grids for each reference year (2030, 2040, and 2050). The data describe the storage capacity, power rating, and operational efficiencies of BESS units at a nodal level. For each reference year, two CSV files are provided:

BESS_allocation_LV.csv:This file provides information on the BESS units installed at the low-voltage (LV) nodes of the distribution grids. Specifically, it includes:

LV_grid: Unique identifier of the low-voltage grid.

LV_osmid: Unique identifier of the low-voltage node in the grid.

Battery_capacity_kWh: Total storage capacity of the BESS installed at the node, expressed in kilowatt-hours (kWh).

Nominal_power_kW: Rated power of the BESS unit, expressed in kilowatts (kW).

Charging_efficiency: Efficiency of the BESS during the charging process (unitless, between 0 and 1).

Discharging_efficiency: Efficiency of the BESS during the discharging process (unitless, between 0 and 1).

LV nodes not listed in this file do not have any BESS installed.

BESS_allocation_MV.csv:Similarly to LV_heat_pump_allocation.csv, this file reports the characteristics of heat pumps connected at medium-voltage nodes. The fields include:

MV_grid: Unique identifier of the medium-voltage grid.

MV_osmid: Unique identifier of the MV node in the grid.

Battery_capacity_kWh: Total storage capacity of the BESS installed at the node, expressed in kilowatt-hours (kWh).

Nominal_power_kW: Rated power of the BESS unit, expressed in kilowatts (kW).

Charging_efficiency: Efficiency of the BESS during the charging process (unitless, between 0 and 1).

Discharging_efficiency: Efficiency of the BESS during the discharging process (unitless, between 0 and 1).

MV nodes not listed in this file do not have any BESS installed.

04_EV (Electric Vehicles)

This folder contains data on the charging demand and flexibility potential of electric vehicles (EVs) in the low-voltage (LV) distribution grids for each reference year (2030, 2040, and 2050). The data are derived from an open mobility database for Switzerland, obtained by Parajeles et al. [https://arxiv.org/abs/2504.03633].

Due to the high granularity of the original mobility dataset, direct integration at the individual vehicle level would be computationally intractable. To address this, the charging events have been aggregated and processed at the LV node level. The dataset provides both uncontrolled charging consumption profiles and flexibility bounds (considering V1G smart charging), along with the maximum flexible energy that can be activated per day. This structure enables studies on EV demand both as an inflexible and a flexible load, facilitating large-scale assessments of EV flexibility at high temporal and spatial resolutions.

The EV profiles are first defined at the municipality level and then disaggregated to LV nodes using a share factor. Each reference year contains four CSV files:

EV_power_profiles_LV.csv:This file reports the non-controlled charging profile and flexibility bounds for each municipality, represented by its BFS code. The columns contain:

BFS_municipality_code: Unique identifier for the municipality.

Profile_type: Indicates the type of profile—Upper, Base, or Lower.

Time-series data for 8,760 hours of the year, expressed in kilowatts (kW).

The Base profile represents the uncontrolled charging demand, while the Upper and Lower bounds define the maximum and minimum power levels, respectively, in the case of flexible charging (V1G). EV power consumption can be shifted between these bounds, with daily energy displacement constraints provided in EV_flexible_energy_profiles_LV.csv.

EV_flexible_energy_profiles_LV.csv:This file provides the maximum flexible energy that can be shifted per day, ensuring that EV battery capacity constraints are not violated. The columns include:

BFS_municipality_code: Unique identifier for the municipality.

Time-series data for 365 days of the year, expressed in kilowatt-hours (kWh).

EV_energy_profiles_LV.csv:This file ensures overall energy balance by reporting weekly EV energy consumption at the municipality level. The columns include:

BFS_municipality_code: Unique identifier for the municipality.

Time-series data for 52 weeks of the year, expressed in kilowatt-hours (kWh).

EV_allocation_LV.csv:This file disaggregates the municipality-level EV profiles to individual LV nodes. Each LV node is assigned a share of the corresponding municipality’s profile. The columns include:

LV_grid: Unique identifier of the low-voltage grid, where the number before the dash (-) corresponds to the BFS code of the municipality (e.g., 1-1_0_7 belongs to municipality 1, and 852-2_1_2 belongs to municipality 852).

LV_osmid: Unique identifier of the low-voltage node in the grid.

EV_share: Share factor indicating the fraction of the municipality-level profile assigned to the node.

05_Demand (Conventional Load)

This folder contains data on the conventional (non-dispatchable) electricity demand for low-voltage (LV) and medium-voltage (MV) distribution grids at a nodal level for each reference year (2030, 2040, and 2050). The demand profiles are expressed in per unit (p.u.) relative to the nodal peak power reported in the grid dataset. All profiles have an hourly resolution and span one full year.

Low-Voltage Demand

For LV loads, representative commercial and residential profiles were identified based on the municipality’s energy mix. Each LV node’s demand profile is computed as a weighted combination of these municipality-level profiles, considering the specific commercial and residential consumption mix at the node level.

LV_basicload_shares.csv:This file provides the share of commercial and residential demand for each LV node. These shares are used to scale the standard commercial and residential profiles of the municipality where the LV grid is located. The fields include:

LV_grid: Unique identifier of the low-voltage grid, where the number before the dash (-) corresponds to the BFS code of the municipality (e.g., 1-1_0_7 belongs to municipality 1, and 852-2_1_2 belongs to municipality 852).

LV_osmid: Unique identifier of the low-voltage node.

Commercial_demand_share: Share of commercial demand at the node.

Residential_demand_share: Share of residential demand at the node.

Commercial_profiles.csvThis file contains the hourly per-unit load profile for commercial demand at the municipality level, identified by the BFS municipality code.. The fields include:

BFS_municipality_code: Unique identifier for the municipality.

Time-series data for 8,760 hours of the year, representing the commercial load profile in per unit.

Residential_profiles.csv:This file has the same structure as Commercial_profiles.csv but contains the residential demand profiles instead.

Medium-Voltage Demand

For MV loads, a single standard profile is assumed for all MV nodes not connected to a secondary substation.

MV_load_profile.csv:This file provides the hourly per-unit load profile for MV nodes, assuming a representative consumption pattern. 

Power_pu: Time-series data for 8,760 hours of the year, representing the MV demand in per unit.

06_Grids

This folder contains the power distribution grids. In this repository, we report only the grids for the integrated medium-low voltage system 459_0. Refer to the original grid repository by Alfredo Oneto et al. in [https://doi.org/10.1016/j.segan.2025.101678] to gather all the grids, replacing the LV and MV .zip files. The folder contains three files:

LV.zip: This .zip archive contains all the low-voltage PDGs data, like network topology, branch flow limits, line impedances, voltage levels, and nodal peak powers.

MV.zip: This .zip archive contains all the medium-voltage PDGs data, like network topology, branch flow limits, line impedances, voltage levels, and nodal peak powers.

dict_folder.json: This file contains a dictionary that maps the BSF municipality number, the number before the dash (-) of the LV_grid code, to the corresponding subfolder in the LV.zip archive. This file is used to access the LV_grid data. For more details about the structure of the LV.zip archive refer to the article [https://doi.org/10.1016/j.segan.2025.101678] and the original PDG dataset [https://doi.org/10.5281/zenodo.15167589].

For more details about the data structure refer to the original PDG dataset [https://doi.org/10.1016/j.segan.2025.101678].

07_Complementary_data

This folder contains complementary data, which are not needed to use the DERs dataset but may be of use for its interpretation and modification. It contains the following files:

Municipalities_2022_01_crs2056.geojson:This GeoJSON file contains the municipal boundaries of Switzerland as part of the swissBOUNDARIES 3D dataset. The dataset includes detailed geometric representations of administrative units in Switzerland, the Principality of Liechtenstein, and border exclaves of neighboring countries. The municipalities are represented as polygons, based on EPSG:2056 coordinate reference system.

08_Data_loader.py

This Python script provides an example of how to load distributed energy resources (DERs) data for Low Voltage (LV) and Medium Voltage (MV) grids for a specified simulation year and time interval. The script is designed to facilitate the loading and processing of the DER data present in the dataset.

How to Use:

Initialize the GridLoader: Create an instance of the GridLoader class by specifying the grid type (LV or MV), grid name, start date, end date, and data year.

Load Grid Data: Use the load_grid method to load the grid's nodes and edges data.

Load DER Data: Call the respective methods to load PV data, HP data, EV data, BESS data, and conventional load profiles.

How It Works:

GridLoader Class: The main class that handles the loading and processing of grid and DER data.

Data Loading Methods: Methods like load_pv_data, load_hp_data, load_ev_data, load_bess_data, and load_conventional_load_profiles are used to load specific types of DER data.

Time Interval Filtering: The script filters the data based on the specified start and end dates to ensure only relevant data is processed.

Data Processing: The script processes the loaded data to create profiles that can be used for simulations.

Disclaimer

The data presented here is compiled and processed to the best of the author's knowledge. The authors do not bear responsibility for the use of this data.

Please exercise caution when using this data for scientific research, and inform the authors if you uncover any inaccuracies.