Question: Modelling of monofacial/Bifacial on fixed & single axis tracking system

[Charlie Linck]

Hi!

I used the pvfactors_timeseries function to model the monofacial and bifacial irradiance on a single-axis tracking PV system using pvsystem.SingleAxisTrackerMount and sat_mount.get_orientation. I believe these results are correct.

Now, I want to model again monofacial and bifacial PV panels but on a fixed system so I decided to use pvfactors_timeseries again, but instead of using surface_tilt data from the orientation function, I used my own fixed calculated optimal fixed azimuth and tilt (named optimal_azimuth and optimal_tilt). But this actually gives me better results than my tracking system, which seems odd to me?

Has anyone tried this or have an idea where I might have gone wrong? Or a better function solution in which I can model a fixed monofacial and bifacial system?

In the attachment, I added pictures of my code.

Looking forward to hearing from you.

Many thanks in advance!

Best regards,

Charlie

[kevina...@gmail.com]

Hi Charlie,

One thing that jumps out is that `axis_azimuth=180` for the fixed-tilt system implies it is an east- or west-facing array.  I wonder if that's what you actually intended?  For a normal equator-facing array you'd want axis_azimuth=90 or 270.

In general, bifacial modeling results depend on more than just tracking vs fixed racking.  For example the different heights above ground will change how the shaded ground patches affect rear-side irradiance.  If you want to double-check the pvfactors results, you can compare them with the output of the infinite sheds model (added in pvlib 0.9.1) -- of course it's a different model and won't give exactly the same results as pvfactors, but the two should be broadly consistent if they're modeling the same system.

Kevin

[Charlie Linck]

Hi Kevin,

I have my axis-azimuth orientation that North =0, so that why I implemented axis-azimuth =180 (facing South).

Good point, I will look into the sheds model. Myabe I this will show better results.

Thanks!

Charlie

[kevina...@gmail.com]

Charlie, I'm not sure I understand your last message.  pvlib has a fixed azimuth coordinate system that the user doesn't have any influence over, i.e. pvlib always has North=0.  The "axis_azimuth" input to the pvfactors function represents the azimuthal alignment of the array's axis of rotation.  Note that pvfactors was intended for single-axis tracking PV arrays, but you can model fixed-tilt systems by pretending a fixed-tilt array is a tracker array that never rotates.  But to treat a South-facing fixed-tilt array like a tracker array, its hypothetical axis of rotation would be aligned East-West, prompting my suggestion of an axis_azimuth of 90 or 270.

Kevin