Help interpreting rear irradiance profiles from bifacial_radiance simulations

[Rafael Campos]

Hi everyone — I know this group is focused on pvlib, not bifacial_radiance, but since many of you are experienced with solar modeling and irradiance simulations, I thought someone here might have some valuable insight into an issue I’m seeing with bifacial_radiance. Hope that’s okay!

I’m using the bifacial_radiance Python library to simulate rear-side irradiance in a single-axis tracker bifacial PV system, but I’m getting some odd results and would appreciate your thoughts. Specifically, it seems like the east and west sensors are somehow flipped: the east sensor measures higher rear irradiance during the morning, while the west sensor measures higher irradiance during the afternoon — which is counterintuitive.

Below is a snippet of the Python code I’m running, along with the CSV output for a specific hour (10 AM):

....

# Scene 5 variables

moduletype = 'CdTe'

originx5 = 14

originy5 = 14.95

sceneDict5 = {'gcr': GCR,'hub_height':hub_height, 'nMods':nMods, 'nRows': nRows, 'originx' : originx5, 'originy' : originy5}  

trackerdict = radObj.makeScene1axis(trackerdict=trackerdict, module=moduletype, sceneDict=sceneDict5, append=True)

.....

radObj.analysis1axis(

    sceneNum=4,

    modWanted=22,

    sensorsy=5,

    backsurfaceoffset=0.15,

    customname='5sens_back',

    debug=True

)

CSV output and hourly plot for a clearday:

x,y,z,rearZ,mattype,rearMat,Wm2Front,Wm2Back

14.570,26.300,1.112,0.994,a21.0.a0.CdTe.6457,a21.0.a0.CdTe.2310,673.742,75.359 (East-most sensor with lower z value but with highest rear-irrad)

14.336,26.300,1.355,1.236,a21.0.a0.CdTe.6457,a21.0.a0.CdTe.2310,674.487,64.706

14.100,26.300,1.598,1.479,a21.0.a0.CdTe.6457,a21.0.tube1.6457,673.430,1.222

13.870,26.300,1.840,1.722,a21.0.a0.CdTe.6457,a21.0.a0.CdTe.2310,673.966,59.457

13.640,26.300,2.082,1.965,a21.0.a0.CdTe.6457,a21.0.a0.CdTe.2310,673.161,63.769

I would really appreciate any insights, suggestions, or experiences you might have regarding this behavior. Has anyone encountered a similar issue with sensor orientation or irradiance inversion in bifacial_radiance? Any tips on verifying sensor placement or debugging coordinate systems would also be very helpful.

Thanks in advance for your time and support — looking forward to your feedback!

[cwh...@sandia.gov]

Hi Rafael,

Most of the irradiance reaching the rear surfaces is from ground reflections. In the morning, the shadows from the array are projected towards the west, so the ground under the east side of the array receives more direct sunlight and hence ground-reflected irradiance reaching the back of the east arrays is greater. 

Cliff