PVPMC Workshop thoughts

[Anton Driesse]

Well, unfortunately I was not able to attend the workshop myself, but
wouldn't this discussion group be a perfect place for some of those who
did attend to share their  thoughts, reactions, highlights, etc?  
Honestly, I'd love to hear about it!

Anton

--
Photovoltaic Performance Labs Germany
Emmy-Noether-Str. 2
79110 Freiburg
Germany

+49-761-8973-5603 (Office)
+49-174-532-7677 (Mobile)

www.pvperformancelabs.com

[Mark Mikofski]

Here are my notes from the user group meeting:

1. near shade (from objects? row-to-row?)
2. no spatial coordinate system (use northings & eastings from some reference point? how to define elevation? sea level?)
3. no electrical coordinate system (for stringing modules?)
4. electrical mismatch due to shading
5. impacts of row-to-row, currently diffuse shade (Passias) and direct shade and isotropic diffuse shade (infinite sheds)
6. horizon profile
7. far shading given HOR file
8. tracker into bay coordinates
9. detailed balance limit
10. tandem cells (2 terminal, 4 terminal?)
11. more spectral examples
12. ground irradiance
13. validation
14. fraction of an hour at sunrise / sunset
15. electrical mismatch - distribution of irradiance - loss model correlation?
16. Logo page with users (power by pvlib) - CPR said okay?
17. output data into various formats, EG: to_sam(), to_pvsyst_csv()
18. cloud calc of entire NSRDB, KA suggest leave this to outside?
19. weather? operational data ingestion?, check/validate? use pvanalytics?
20. pysam, pvlib, pvlib-matlab - How to get started guides
21. documentation, "I'm a beginner, where do I start?"
22. ERA5, MERRA2, pull in near realtime data, commercial data
23. spectral correction factors, open source, data classes, spectral response, be smart about functions? PR currently from Anton, collaborate with Mark
24. PAN -> SAPM
25. parsing PAN / OND parser from Frederic Rivolier
26. PAN file generator
27. PVC collada file reader
28. cloudy sky detection (opposite of clear sky?)
29. PAN file as a standard?
30. solar variability index (already in pvanalytics?)
31. examples, if don't see create one
32. YouTube videos on how to github / PR
33. validated model? what qualifies? community acceptance, demand, how useful to group, peer review or conference?
34. volunteers? create an issue, new beginners
35. quarterly meetings?
36. quarterly releases?

Sorry, my notes are not too specific, my shorthand is a mystery

[Silvana Ayala]

Silvana dumplist 😊

-Detailed balance

-Tandem

-Mismatch correction factor

-Cabling and racking calculations (mass? Cost? Etc) for systems based on system size. Some ballasting?

-bifacial sheds -- > output ground irradiance for agripv

-cloud and world-maping color simulations for all the datasets

-Smart/options to load data and time shift 30+- minutes for solar position based on TMY, Solargis, etc etc. (sample code on bifacial_radiance)

-Early morning and late evenings sun position correction (sample code on bifacial_radiance)

-Electrical mismatch function from Deline paper 2021 (based on irradiance-nonuniformity)

-A 'powered by pvlib' image of logos on the welcome page (like the fancy companies do)

-Outputting/translating weather data into other formats

-Formal moodle training or online training so people can work through and earn ‘badges’ (.edx, etc)

-Simplified LCOE by Tim Silverman

-Weatherfile quality checkers

-SAM vs PVSyst vs comprisons

-pySMARTS pull in (Silvana owes this for 5 years now 😊 )

-AM1.5 modeling on smarts2 Example gallery

-Automatic comparison routines, i.e. applying the 7 clear-sky models at the same time?

-PVlib contributors research papers using PVlib made as more gallery examples (i.e. Kevin’s PVSCs)

-Hackaton as part of the next PVPMC ?

 

General discussion:

Mismatch : 1 voter

-Near object shading / near field shading models (pvlib no coordinate system)

-More row-to-row diffuse shading  (W. Hobbs)

-Rows with different tracking angles, like Nextracker smarttrack (can already be done)

-Add to example gallery the one-liner to model pvlib functions on full NSDRB databases.

-pySAM and pvlib: for someone to use python to run simulations, where to start.

-I'm a beginner where do I start. Section

-Era5 and MERRA2

-pvfit person -- code uses data classes to represent non negative functions like spectral esposnaivity, spectral irradiance, take QE.

-pvfit person -- panfile to sandia array performance model

-panfile generation, or parsing panfiles?

-PVsyst import files... for coordinate , shading scene files? .shd ?

-PVcase coordinates with collada.

-Bruno - suggestion to keep to pvc format. Pan files might change, they reserve the right to do so (and that's why not so documented)

-Beyond clear-sky, labeling sky for semi cloudy, mostly cloudy, all cloudy. Overcast day, for overcast models. related to Josh's paper variability. (Some in pv analytics).

-New user, faced with 7 clearsky model, which one to choose?

-Josh: for examples, if you don't see an example you want, but you don't feel reayd to provide it, put a comment? Could someone do this example?

-Numphocus, quarterly funding cycle. Type of person that can do stuff, 3-5 k. Google summer of code

-Monthly, or quaterly reviews?

-CH volunteers to set up quarterly group meetings.

-Submit email on pvpmc website for pvlib mailing lit.

-Video explaining how to github.

-

[Anton Driesse]

Thank you so much Mark and Silvana for sharing your notes!  It wasn't exactly what I was expecting, but then again I'm not sure what I was expecting.  I guess there is no substitute for having been there though! :(

Anton

--
You received this message because you are subscribed to the Google Groups "pvlib-python" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pvlib-python...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pvlib-python/4e9ce6b3-5a98-490b-8555-06e244f3fc7en%40googlegroups.com.

[Kenneth J. Sauer]

Hi Anton,

We missed you at PVPMC 2022! I guess the main (obvious) note you're
looking for is that it simply wasn't the same w/out you there 😂, & I
genuinely mean that! I'll toss my brief notes into the mix to continue
the group synopsis. Please read at your leisure (no response
required/no rush).

"""
KJS Executive Summary Notes - Sandia PVPMC 2022
- Reproducibility error component in simulation uncertainty from large
differences in PV modeling results across users (observed in modeling
round-robins)
- Significant subhourly clipping adjustments in PV modeling (ideally,
adjustment models directly use, or are empirically informed by, 1-min.
data sets & preferably are compatible w/ conventional, hourly PV
simulation modeling); results can be significant (up to 4% annually
depending on geographical location)
- Represent actual tracking in models: Model actual tracking
algorithms instead of idealized tracking (which itself is different
than so-called "TrueTracking", given non-ideal patterns & rates of
tracker rotation in the real world, etc.); also, somehow take into
account performance impacts of hail & wind stow algorithms used to
mitigate damage from severe (or even catastrophic) weather events
- Get better at (& standardize) soiling rate measurements as well as
snow/soiling modeling (incl. for bifacial); explore "true-ups" of
long-term predictions based on ground-based measured soiling data
(also, explore ways to measure-correlate-predict MCP soiling, albedo,
& spectrum for certain cell technologies, etc.)
- FEDIS model of diffuse irradiance IAM losses by Yu Xie
(https://github.com/xieyupku/FEDIS/blob/main/fedis.py/)
- Include multi-surface (multi-albedo) modeling option for PV systems
which span multiple albedos (replacing a global, project-level
definition)
- Split PV modules (i.e., half-modules, incl. w/ half-cut cells) have
self-shading advantages for fixed tilt & single-axis tracking (incl.
w/ backtracking on), given the relationship b/w electric power loss &
shade coverage factor (i.e., shade area fraction % = shade area below
shade line divided by total module area). Designing the optimal
tracking algorithm is (thus) a complex optimization problem involving
many considerations incl. terrain-dependent self-shading, electrical
mismatch, & AOI losses (as you deviate from idealized solar (sun)
position tracking), down to the design of the PV cells (CdTe vs. Si,
half-cut) & modules (split aka "twin" modules, bypass diode
configuration), namely, down to the effects of shade line or shade
area fraction % on PV module performance.
- On the horizon: Improvements in short-term forecasting accuracy to
support energy market trading
- Relationship b/w open-source, Python-wrapped NREL SAM aka pysam (a
PV simulation modeling software program) & open-source, Sandia PVLIB
aka pvlib-python (a library of functions used in PV simulation
modeling). Consider absorbing the latter into the former, namely,
aligning the latter's overlapping functions w/ the former's (ideally
via direct code use w/out translation), ensuring exact consistency of
algorithms, mathematics, & results. Also, incl. NREL's PVWatts in the
former category (PV simulation software program that could directly
import & use the pvlib-python function package).
"""

I'll follow up if I think of anything else of note.

--Thanks,
Ken

> To view this discussion on the web visit https://groups.google.com/d/msgid/pvlib-python/8d70f113-2a16-9a16-a328-c33ed8f9092a%40pvperformancelabs.com.

[Anton Driesse]

Hi Ken,

I missed you too!  By the way, your carefully worded comment leaves open
the possibility that it was actually better w/out me there! 😉

Your summary notes are definitely a good addition to the collective
memory and help to give some of the terser bullet points from Mark and
Silvana context and connections.  I find your last bullet particularly
interesting, especially in the overall evolving landscape of open and
closed source PV system simulation software.

Many thanks, and I hope that others find these notes useful as well!

Anton

[MENDYL ABDERRAHMANE]

Hi all,

I would like assistance in downloading the ERA5 clear sky data for different sites, from ECMWF  , but i think there is an issue with the website, any one can assist, thanks 

 

To view this discussion on the web visit https://groups.google.com/d/msgid/pvlib-python/152994d1-2728-0c2a-5fe4-1646e4f27f70%40pvperformancelabs.com.

--

            Mendyl Abderrahmane 

   Ph.D. - Department of Meteorology, 

   Eötvös Loránd University,

   Budapest,