Azimuth and tilt

[ThomasGeffroyCatie]

Hi,

I try using azimuth and surface tilt parameters for better estimates, without success.
The value for those two parameters are not taken into account.

    system = pvsystem.PVSystem(
        module_parameters={
            'pdc0': 975,
            'gamma_pdc': -0.0045,
            "surface_tilt": 7,
            "surface_azimuth":  180
        },
        inverter_parameters={
            'pdc0': 1000,
        },
        temperature_model_parameters=temperature_model_parameters
    )
    mc = ModelChain(system, location, aoi_model='physical',spectral_model="no_loss")
    
    weather = df[["PeriodEnd", "Azimuth", "Dni", "Dhi", "AirTemp", "WindSpeed10m"]]
    weather = weather.set_index("PeriodEnd")
    weather.index = pd.to_datetime(weather.index)
    weather.columns = ['ghi', 'dni', 'dhi', 'temp_air', 'wind_speed']

    mc.run_model(weather)

Is it spectral_model that need to be set to something different ?

Thanks

[kandersolar]

Howdy @ThomasGeffroyCatie, surface_tilt and surface_azimuth should not be specified inside of module_parameters but instead as separate inputs to PVSystem (see docs). module_parameters lists intrinsic module properties needed to model output power from effective irradiance and cell temperature whereas tilt and azimuth are "external" properties of the broader system. So if you move those inputs out of module_parameters like this I think you'll get the intended results:

    system = pvsystem.PVSystem(
        surface_tilt=7,
        surface_azimuth=180,
        module_parameters={
            'pdc0': 975,
            'gamma_pdc': -0.0045,
        },
        inverter_parameters={
            'pdc0': 1000,
        },
        temperature_model_parameters=temperature_model_parameters
    )

[ThomasGeffroyCatie]

Thanks, it solved this issue.

But I have the same problem with the inverter_parameters. In the examples I see for example

system = pvsystem.PVSystem(arrays=[array], inverter_parameters={'pdc0': 1})
mc = modelchain.ModelChain(system, loc, spectral_model='no_loss')

But changing pdc0 to 0 or 1000 doesn't change anything in my prediction.
Any idea why ?

Thanks

[cwhanse]

@ThomasGeffroyCatie can you share code to reproduce that outcome? The code you have posted looks correct.

[ThomasGeffroyCatie]

def get_modelchain_dc(df,inverter):
    temperature_model_parameters = TEMPERATURE_MODEL_PARAMETERS['sapm']['open_rack_glass_polymer']
    system = pvsystem.PVSystem(
        inverter_parameters=dict(pdc0=inverter),
        surface_tilt=15,
        surface_azimuth=160,
        module_parameters={
            'pdc0': 975,
            'gamma_pdc': -0.0045,
        },
        temperature_model_parameters=temperature_model_parameters
    )
    mc = ModelChain(system, location , aoi_model='physical',spectral_model="no_loss")
    
    weather = df[["PeriodEnd", "Azimuth", "Dni", "Dhi", "AirTemp", "WindSpeed10m"]].copy()
    weather = weather.set_index("PeriodEnd")
    weather.index = pd.to_datetime(weather.index)
    weather.columns = ['ghi', 'dni', 'dhi', 'temp_air', 'wind_speed']

    mc.run_model(weather)
    dc_df = mc.dc.to_frame()
    dc_df.columns = ["pdc"]
    return dc_df,mc

Call with different values :

for inverter in [0,1,1000]:
    print("*"*15,"inverter",inverter, "*"*15)
    dc_df,mc = get_modelchain_dc(df,inverter)
    print(dc_df.describe())

And the results:

*************** inverter 0 ***************

/home/thomas/.local/lib/python3.8/site-packages/pvlib/modelchain.py:499: pvlibDeprecationWarning: ModelChain.dc is deprecated and will be removed in v0.10. Use ModelChain.results.dc instead
  warnings.warn(msg, pvlibDeprecationWarning)

                 pdc
count  261776.000000
mean      186.613741
std       250.600602
min        -0.755653
25%         0.029174
50%         1.606278
75%       388.336066
max       847.890800
*************** inverter 1 ***************

/home/thomas/.local/lib/python3.8/site-packages/pvlib/modelchain.py:499: pvlibDeprecationWarning: ModelChain.dc is deprecated and will be removed in v0.10. Use ModelChain.results.dc instead
  warnings.warn(msg, pvlibDeprecationWarning)

                 pdc
count  261776.000000
mean      186.613741
std       250.600602
min        -0.755653
25%         0.029174
50%         1.606278
75%       388.336066
max       847.890800
*************** inverter 1000 ***************
                 pdc
count  261776.000000
mean      186.613741
std       250.600602
min        -0.755653
25%         0.029174
50%         1.606278
75%       388.336066
max       847.890800

/home/thomas/.local/lib/python3.8/site-packages/pvlib/modelchain.py:499: pvlibDeprecationWarning: ModelChain.dc is deprecated and will be removed in v0.10. Use ModelChain.results.dc instead
  warnings.warn(msg, pvlibDeprecationWarning)

[cwhanse]

dc_df = mc.dc.to_frame()

You are extracting the DC output of the array, which is not affected by the inverter size. pvlib applies clipping to the AC output of the inverter.

[ThomasGeffroyCatie]

Yes of course.
Thanks @cwhanse