High error

[Suraj Bhanu Jat]

I have been comparing the actual power output with the power output from pvlib model. But still getting huge errors. 

I found that the error may be because the GHI is less than DNI and DHI but I am not sure. 

Is there a problem with the weather data, or actual power output data or am I doing something wrong while calculating the power output using pvlib. 

GHI,DNI, DHI, Temp and Wind are actual measurements. Is there possibility that the actual measurements might be wrong? 

why is the percentage error so high when I compare the actual power output and power output from physical model.

The power output in the code is just for single string but actual physical model power is from multiple string with different panel specifications, but the method for calculation is same.

Attaching the code for reference.

Please advise.

Thanks in advance,

 

[kevina...@gmail.com]

Hello Suraj,

The three irradiance components are related to each other by this equation: GHI = DHI + DNI * cos(solar_zenith)

cos(zenith) < 1, so GHI must always be less than (or equal to) DHI + DNI.  But that's not the case in the plot in your notebook.  So there must be something wrong with the irradiance data.

The DNI curve in particular looks very unrealistic to me.  When GHI is unstable like it is in your plot, it is usually due to fluctuations in DNI.  I suspect there is something wrong with the DNI sensor, or perhaps the data are mislabeled and what is called "DNI" in this dataset is actually something else.

Kevin