Master branch:
Development branch:
bifacial_radiance contains a series of Python wrapper functions to make working with RADIANCE easier, particularly for the PV researcher interested in bifacial PV performance. Please see the instructions here, notebook examples in the \docs\ folder of the repo, and discussion on the Wiki for more details!
https://youtu.be/4A9GocfHKyM This video shows how to install the bifacial_radiance software and all associated softwares needed. More info on the Wiki. Instructions are also shown below.
A GUI has been added in version 3.0. The GUI reads/writes all input parameters necessary to run a simulation, and runs the specified simulation by calling the correct functions with the specified parameters. So no need to use a journal or a script! But you still need to install following the procedure below.
To run the gui, import bifacial_radiance and run bifacial_radiance.gui()
This software requires the previous installation of RADIANCE from https://github.com/NREL/Radiance/releases.
Make sure you add radiance to the system PATH so Python can interact with the radiance program
If you are on a PC you should also copy the Jaloxa radwinexe-5.0.a.8-win64.zip executables into program files/radiance/bin: http://www.jaloxa.eu/resources/radiance/radwinexe.shtml
**Note: bifacial_radiance is not endorsed by or officially connected with the Radiance software package or its development team.
- clone the bifacial_radiance repo to your local directory or download and unzip the .zip file
- navigate to the \bifacial_radiance directory using anaconda command line
- run
pip install -e .( the period . is required, the -e flag is optional and installs in development mode where changes to the bifacial_radiance.py files are immediately incorporated into the module if you re-start the python kernel) - for best compatibility, deploy in an Anaconda 2.7 environment, or run
pip install -r requirements.txt
Copy gencumulativesky.exe from the repo's /bifacial_radiance/data/ directory and copy into your Radiance install directory.
This is typically found in /program files/radiance/bin/.
**Note: GenCumulativeSky is detailed in the publication "Robinson, D., Stone, A., Irradiation modeling made simple – the cumulative sky approach and its applications, Proc. PLEA 2004, Eindhoven 2004." The source is available from the authors here.
Keep scene geometry files separate from the bifacial_radiance directory. Create a local directory somewhere to be used for storing scene files.
This makes sure the PATH is updated
from bifacial_radiance import RadianceObj # the main container for working with radiance
Now that the module is loaded, let's use it.
demo = RadianceObj(name = 'Testrun', path = 'myfolder') #create a new demo run. Files will have the Testrun prefix, and be saved to 'myfolder'
demo.setGround(0.3) # input albedo number or material name like 'concrete'. To see options, run this without any input.
# Now download an EPW climate file for any global lat/lon value :
epwfile = demo.getEPW(37.5,-77.6) # pull EPW data for any global lat/lon
# let's load this epw file into our MetObj container class.
metdata = demo.readEPW(epwfile) # read in the EPW weather data as metdata object. Run this with no input parameters to load a graphical picker
# if you'd rather use a TMY3 file, select one that you've already downloaded:
metdata = demo.readTMY() # select an existing TMY3 climate file. return metdata object.
Now that we have ground albedo and a climate file loaded, we need to start designing the PV system. Fixed tilt systems can have hourly simulations with gendaylit, or annual simulations with gencumulativesky
# create cumulativesky skyfiles and save it to the \skies\ directory, along with a .cal file in root
demo.genCumSky(demo.epwfile)
--- optionally ----
demo.gendaylit(metdata,4020) # pass in the metdata object, plus the integer number of the hour in the year you want to run (0 to 8759)
# note that for genCumSky, you pass the *name* of the EPW file. for gendaylit you pass the metdata object.
The nice thing about the RadianceObject is that it keeps track of where all of your skyfiles and calfiles are being saved. Next let's put a PV system together. The details are saved in a dictionary and passed into makeScene. Let's start with a PV module:
# Create a new moduletype: Prism Solar Bi60. width = .984m height = 1.695m.
demo.makeModule(name='Prism Solar Bi60',x=0.984,y=1.695) #x is assumed module width, y is height.
# Let's print the available module types
demo.printModules()
the module details are stored in a module.json file in the bifacial_radiance\data directory so you can re-use module parameters.
Each unit module generates a corresponding .RAD file in \objects\ which is referenced in our array scene.
Starting in version 0.2.3 there are some nifty module generation options including stacking them (e.g. 2-up or 3-up but any number) with a gap, and torque tube down the middle of the string.
Since version 0.2.4, orientation of the module has been deprecated as an input. Now, to define the orientation it has to be done in the makeModule step, assigning the correct values to the x and y of the module. x is the size of the module along the row, therefore for a landscape module x > y.
# make a 72-cell module 2m x 1m arranged 2-up in portrait with a 10cm torque tube behind.
# a 5cm offset between panels and the tube, along with a 5cm array gap between the modules:
demo.makeModule(name = '1axis_2up', x = 1.995, y = 0.995, torquetube = True, tubetype = 'round',
diameter = 0.1, zgap = 0.05, ygap = 0.05, numpanels = 2)
Now we make a sceneDict with details of our PV array. We'll make a rooftop array of Prism Solar modules in landscape at 10 degrees tilt.
module_name = 'Prism Solar Bi60'
sceneDict = {'tilt':10,'pitch':1.5,'clearance_height':0.2,'azimuth':180, 'nMods': 20, 'nRows': 7}
# this is passed into makeScene to generate the RADIANCE .rad file
scene = demo.makeScene(module_name,sceneDict) #makeScene creates a .rad file with 20 modules per row, 7 rows.
OK, we're almost done. RADIANCE has to combine the skyfiles, groundfiles, material (*.mtl) files, and scene geometry (.rad) files into an OCT file using makeOct. Instead of having to remember where all these files are, the RadianceObj keeps track. Or call .getfilelist()
octfile = demo.makeOct(demo.getfilelist()) # the input parameter is optional - maybe you have a custom file list you want to use
The final step is to query the front and rear irradiance of our array. The default is a 9-point scan through the center module of the center row of the array. The actual scan values are set up by .makeScene and returned in your sceneObj (sceneObj.frontscan, sceneObj.backscan). To do this we use an AnalysisObj.
analysis = AnalysisObj(octfile, demo.name) # return an analysis object including the scan dimensions for back irradiance
analysis.analysis(octfile, demo.name, scene.frontscan, scene.backscan) # compare the back vs front irradiance
print('Annual bifacial ratio average: %0.3f' %( sum(analysis.Wm2Back) / sum(analysis.Wm2Front) ) )
Beginning in v0.2.5 we can query specific scans along the array.
# Do a 4-point scan along the 5th module in the 2nd row of the array.
scene = demo.makeScene(module_name,sceneDict)
octfile = demo.makeOct()
analysis = AnalysisObj(octfile, demo.name)
frontscan, backscan = analysis.moduleAnalysis(scene, sensorsy = 4, modWanted = 5, rowWanted = 2)
frontresults,backresults = analysis.analysis(octfile, demo.name, scene.frontscan, scene.backscan)
print('Annual bifacial ratio on 5th Module average: %0.3f' %( sum(analysis.Wm2Back) / sum(analysis.Wm2Front) ) )
# And you can run the scanning for another module.
frontscan, backscan = analysis.moduleAnalysis(scene, sensorsy = 4, modWanted = 1, rowWanted = 2)
frontresults,backresults = analysis.analysis(octfile, demo.name, scene.frontscan, scene.backscan)
print('Annual bifacial ratio average on 1st Module: %0.3f' %( sum(analysis.Wm2Back) / sum(analysis.Wm2Front) ) )
For more usage examples including 1-axis tracking examples, carport examples, and examples of scenes with multiple sceneObjects (different trackers/modules/etc) see the Jupyter notebooks in \docs\
RadianceObj(basename,path): This is the basic container for radiance projects.
Pass in a basename string to name your radiance scene and append to various
result and image files. path points to an existing or empty Radiance directory.
If the directory is empty it will be populated with appropriate ground.rad and view
files.
Default behavior: basename defaults to current date/time, and path defaults to current directory
RadianceObj.getfilelist() : return list of material, sky and rad files for the scene
RadianceObj.returnOctFiles() : return files in the root directory with .oct extension
RadianceObj.setGround(material_or_albedo, material_file): set the ground to either
a material type (e.g. 'litesoil') or albedo value e.g. 0.25. 'material_file' is a
filename for a specific material RAD file to load with your material description
RadianceObj.getEPW(lat,lon) : download the closest EnergyPlus EPW file for a give lat / lon value.
return: filename of downloaded file
RadianceObj.readWeatherFile(weatherFile) : call readEPW or readTMY functions to read in a epw or tmy file. Return: metdata
RadianceObj.readEPW(epwfilename) : use pyepw to read in a epw file. Return: metdata
RadianceObj.readTMY(tmyfilename) : use pvlib to read in a tmy3 file. Return: metdata
RadianceObj.gendaylit(metdata,timeindex) : pass in data read from a EPW file.
Select a single time slice of the annual timeseries to conduct gendaylit Perez model
for that given time
RadianceObj.gencumsky(epwfilename, startdt, enddt) : use gencumulativesky.exe to do an entire year simulation.
If no epwfilename is passed, the most recent EPW file read by readEPW will be used. startdt and enddt are optional
start and endtimes for the gencumulativesky. NOTE: if you don't have gencumulativesky.exe loaded,
look in bifacial_radiance/data/ for a copy
RadianceObj.makeOct(filelist, octname): create a .oct file from the scene .RAD files. By default
this will use RadianceObj.getfilelist() to build the .oct file, and use RadianceObj.basename as the filename.
RadianceObj.makeScene(moduletype, sceneDict) : create a PV array scene with nMods modules per row and nRows number of rows. moduletype specifies the type of module which be one of the options saved in module.JSON (makeModule adds a customModule to the Json file). Pre-loaded module options are 'simple_panel', which generates a simple 0.95m x 1.59m module, or 'monopanel' which looks for 'objects/monopanel_1.rad'. sceneDict is a dictionary containing the following keys: 'tilt','pitch','clearance_height','azimuth', 'nMods', 'nRows'.
Return: SceneObj which includes details about the PV scene including frontscan and backscan details
RadianceObj.getTrackingGeometryTimeIndex(metdata, timeindex, angledelta, roundTrackerAngleBool, backtrack, gcr, hubheight, sceney): returns tracker tilt and clearance height for a specific point in time.
Return: tracker_theta, tracker_height, tracker_azimuth_ang
AnalysisObj(octfile,basename) : Object for conducting analysis on a .OCT file.
AnalysisObj.makeImage(viewfile,octfile, basename) : create visual render of scene 'octfile' from view 'views/viewfile'
AnalysisObj.makeFalseColor(viewfile,octfile, basename) : create false color Wm-2
render of scene 'octfile' from view 'views/viewfile'
AnalysisObj.analysis(octfile, basename, frontscan, backscan) : conduct a general front / back ratio
analysis of a .oct file. frontscan, backscan: dictionary input for linePtsMakeDict that
is passed from AnalysisObj.makeScene.
MORE DOCS TO COME: