mosaik-pvgis 1.2.3

PVGIS Simulator for mosaik

This component simulates PV power output based on PVGIS (Photovoltaic Geographical Information System) photovoltaic performance data (the data is free to use). PVGIS provides information for any location in Europe and Africa, as well as a large part of Asia and America.

The PV simulator does not require any input, the only need to configure the PV system, geographical location and select an available reference year.

Since PVGIS only provides hourly data, any other step size causes the data to be aggregated or splitted accordingly:

• For energy, the estimate is summed within a time step that is greater than one hour, or devided by time intervals otherwise.

• For power, the estimate is repeated with the last hourly value for a time step that is less than or equal to one hour, or is averaged otherwise.

Note that a time step greater than one hour may not make sense for the available estimates since they are based on hourly data.

PV output data:

• P[MW] - estimation of power produced

• E[MWh] - estimation of enrgy produced

PV system input data:

• scale_factor - multiplies output [coef], might be used as a contol signal

Configuration:

• scale_factor [coef] - multiplies production once model is created, 1 is equal to 1 kW peak power installed

• latitude [grad]

• longitude [grad]

• slope [grad] - inclination angle for the fixed plane

• azimuth [grad] - orientation angle for the fixed plane

• optimal_angle - if True, calculates and uses an optimal slope

• optimal_both - if True, calculates and uses an optimal slope and azimuth

• system_loss [%] - system inefficiency

• pvtech - PV technology: “CIS” (defaul), “crystSi”, “CdTe”, “Unknown”

• database - solar radiation database, PVGIS-SARAH2, PVGIS-NSRDB, PVGIS-ERA5, PVGIS-SARAH (default)

• datayear - reference year from database

An example scenario is located in the ´demo´ folder.

Other options

Please pay attention to the input data. If you want to use only Direct Normal Irradiance input data as part of the co-simulation, then mosaik-pv is suitable, if you want to use full weather information (global irradiance, wind speed, air temperature and pressure) then mosaik-pvlib is suitable. If you are satisfied with historical performance estimates for a particular location or have no other input data, then mosaik-pvgis is the best solution which is based on PVGIS performance data.

Installation

pip install mosaik-pvgis

If you don’t want to install this through PyPI, you can use pip to install the requirements.txt file:

pip install -r requirements.txt

• To use this, you have to install at least version 3.2.0 of mosaik.

• It is recommended, to use the mosaik-csv library to export the results.

How to Use

Specify simulators configurations within your scenario script:

SIM_CONFIG = {
    'PVSim': {
        'python': 'mosaik_components.pv.pvgis_simulator:PVGISSimulator'
    },
    'CSV_writer': {
        'python': 'mosaik_csv_writer:CSVWriter',
    },
    ...
}

Initialize the PV-system:

# Create PV system with certain configuration
PVSIM_PARAMS = {
    'start_date' : START,
    'cache_dir' : './', # it caches PVGIS API requests
    'verbose' : True, # print PVGIS parameters and requests
}
pv_sim = world.start(
                "PVSim",
                step_size=STEP_SIZE,
                sim_params=PVSIM_PARAMS,
            )

Instantiate model entities:

PVMODEL_PARAMS = {
    'scale_factor' : 1000, # multiplies power production, 1 is equal to 1 kW peak power installed
    'lat' : 52.373,
    'lon' : 9.738,
    'slope' : 0, # default value,
    'azimuth' : 0, # default value,
    'optimal_angle' : True, # calculate and use an optimal slope
    'optimal_both' : False, # calculate and use an optimal slope and azimuth
    'pvtech' : 'CIS', # default value,
    'system_loss' : 14, # default value,
    'database' : 'PVGIS-SARAH3', # default value,
    'datayear' : 2016, # default value,
}
pv_model = pv_sim.PVSim.create(1, **PVMODEL_PARAMS)

Connect with PV-simulator:

world.connect(
                    pv_model[0],
                    csv_writer,
                    'P[MW]',
                )

world.run(until=END)

Notes

The simulator configuration can be specified with the sim_params parameter in world.start:

• start_date : preferable format is “2020-07-17 12:00:00”

• end_date : slightly optimizes memory by slicing historical time series in the range [start_date : end_date] inclusive, if provided

• cache_dir : ‘./’ - if not False, a local dir and file are used to store PVGIS API requests (may cause some problems in case of lack of space or restrictions on file discriptors)

• verbose : True - output PVGIS API requests to stdout

• gen_neg : False - if True, multiplies output by -1

Changelog

1.2.3 - 2025-05-07

• Leap year error is fixed

• end_date parameter is added

• arrow removed from dependencies

1.2.2 - 2024-11-05

• Units correction: 1) Renamed output attribute P[MW] -> E[MWh]. 2) New P[MW] attribute <- repeat the last hourly P[MW] for a time interval less or equal than one hour, take the average value of the hourly P[MW] for the time interval exceeding one hour

• New caching component used

• Tests were added

1.2.1 - 2024-10-04

• PVGIS API update

• Minor changes

1.2 - 2024-02-13

• Control signal and gen_neg parameter are added

• Minor changes

1.0 - 2024-01-11

• Restructured as a mosaik component

• Readme and examples added

• Initial release of version implemented by Danila Valko

Authors

The adapter was initially developed by Danila Valko.