2.1 PVLib: ModelChain with Solcast weather data

Running PVLib's ModelChain example using Solcast API's data for the weather data. We will use one of the "unmetered locations" available from the Solcast API and the SDK:

In [1]:

#!pip install pvlib

#!pip install pvlib

In [2]:

from solcast.unmetered_locations import UNMETERED_LOCATIONS

solcast_location = UNMETERED_LOCATIONS['The Colosseum']
lat, lon = solcast_location["latitude"], solcast_location["longitude"]
lat, lon

from solcast.unmetered_locations import UNMETERED_LOCATIONS solcast_location = UNMETERED_LOCATIONS['The Colosseum'] lat, lon = solcast_location["latitude"], solcast_location["longitude"] lat, lon

Out[2]:

(41.89021, 12.492231)

In [3]:

import pandas as pd
import pvlib
from pvlib.pvsystem import PVSystem
from pvlib.location import Location
from pvlib.modelchain import ModelChain
from pvlib.temperature import TEMPERATURE_MODEL_PARAMETERS

temperature_model_parameters = TEMPERATURE_MODEL_PARAMETERS['sapm']['open_rack_glass_glass']

sandia_modules = pvlib.pvsystem.retrieve_sam('SandiaMod')
cec_inverters = pvlib.pvsystem.retrieve_sam('cecinverter')

sandia_module = sandia_modules['Canadian_Solar_CS5P_220M___2009_']
cec_inverter = cec_inverters['ABB__MICRO_0_25_I_OUTD_US_208__208V_']

import pandas as pd import pvlib from pvlib.pvsystem import PVSystem from pvlib.location import Location from pvlib.modelchain import ModelChain from pvlib.temperature import TEMPERATURE_MODEL_PARAMETERS temperature_model_parameters = TEMPERATURE_MODEL_PARAMETERS['sapm']['open_rack_glass_glass'] sandia_modules = pvlib.pvsystem.retrieve_sam('SandiaMod') cec_inverters = pvlib.pvsystem.retrieve_sam('cecinverter') sandia_module = sandia_modules['Canadian_Solar_CS5P_220M___2009_'] cec_inverter = cec_inverters['ABB__MICRO_0_25_I_OUTD_US_208__208V_']

In [4]:

location = Location(latitude=lat, longitude=lon)

system = PVSystem(
    surface_tilt=20, 
    surface_azimuth=200,
    module_parameters=sandia_module,
    inverter_parameters=cec_inverter,
    temperature_model_parameters=temperature_model_parameters
)


mc = ModelChain(system, location)

location = Location(latitude=lat, longitude=lon) system = PVSystem( surface_tilt=20, surface_azimuth=200, module_parameters=sandia_module, inverter_parameters=cec_inverter, temperature_model_parameters=temperature_model_parameters ) mc = ModelChain(system, location)

We will use the Solcast SDK to retrieve accurate real-world irradiance and and weather data:

In [5]:

from solcast import live

help(live.radiation_and_weather)

from solcast import live help(live.radiation_and_weather)

Help on function radiation_and_weather in module solcast.live:

radiation_and_weather(latitude: float, longitude: float, output_parameters: List[str], **kwargs) -> solcast.api.Response
    Get irradiance and weather estimated actuals for near real-time and past 7 days
    for the requested location, derived from satellite (clouds and irradiance
    over non-polar continental areas) and numerical weather models (other data).
    
    Args:
        latitude: in decimal degrees, between -90 and 90, north is positive
        longitude: in decimal degrees, between -180 and 180, east is positive
        output_parameters: list of strings with the parameters to return
    
    See https://docs.solcast.com.au/ for full list of parameters.

In [6]:

solcast_resp = live.radiation_and_weather(
    latitude=lat, 
    longitude=lon,
    output_parameters=['ghi', 'dni', 'dhi', 'air_temp', 'wind_speed_10m'],
    period='PT5M',
    hours=78
)

solcast_weather = solcast_resp.to_pandas()
# Solcast API data is "period end" while PVLib expects instantaneous
# To account for this, relabel the Solcast data to "period middle"
solcast_weather.index = solcast_weather.index - pd.Timedelta(minutes=2.5)
# timeseries in local time for readability
solcast_weather.index = solcast_weather.index.tz_convert('Europe/Rome')

solcast_resp = live.radiation_and_weather( latitude=lat, longitude=lon, output_parameters=['ghi', 'dni', 'dhi', 'air_temp', 'wind_speed_10m'], period='PT5M', hours=78 ) solcast_weather = solcast_resp.to_pandas() # Solcast API data is "period end" while PVLib expects instantaneous # To account for this, relabel the Solcast data to "period middle" solcast_weather.index = solcast_weather.index - pd.Timedelta(minutes=2.5) # timeseries in local time for readability solcast_weather.index = solcast_weather.index.tz_convert('Europe/Rome')

Now let's run the ModelChain with this weather data

In [7]:

mc.run_model(solcast_weather)

mc.run_model(solcast_weather)

Out[7]:

ModelChain: 
  name: None
  clearsky_model: ineichen
  transposition_model: haydavies
  solar_position_method: nrel_numpy
  airmass_model: kastenyoung1989
  dc_model: sapm
  ac_model: sandia_inverter
  aoi_model: sapm_aoi_loss
  spectral_model: sapm_spectral_loss
  temperature_model: sapm_temp
  losses_model: no_extra_losses

Plot the DNI fetched from the Solcast API against the AC power of the site:

In [8]:

import matplotlib.pyplot as plt

fig, ax1 = plt.subplots()
ax2 = ax1.twinx()
mc.results.weather.dni.plot(ax=ax1, color='g')
mc.results.ac.plot(ax=ax2, color='b')

ax1.set_xlabel('time')
ax1.set_ylabel('DNI', color='g')
ax2.set_ylabel('AC Power', color='b')

import matplotlib.pyplot as plt fig, ax1 = plt.subplots() ax2 = ax1.twinx() mc.results.weather.dni.plot(ax=ax1, color='g') mc.results.ac.plot(ax=ax2, color='b') ax1.set_xlabel('time') ax1.set_ylabel('DNI', color='g') ax2.set_ylabel('AC Power', color='b')

Out[8]:

Text(0, 0.5, 'AC Power')

In [ ]: