Vertical Bifacial Photovoltaic System Model Validation: Study With Field Data, Various Orientations, and Latitudes
Bibliographic record
Abstract
Accurate modeling of photovoltaic (PV) systems is critical for the design, financial analysis, and monitoring of solar PV plants. For bifacial PV applications, models must additionally offer robust rear-side irradiance algorithms. However, bifacial PV irradiance models have yet to be sufficiently validated for east–west vertically oriented systems, where direct beam solar irradiation swaps at solar noon. Here, we validate five bifacial irradiance models with field data collected in Golden, CO, USA (40°N) and Fairbanks, AK, USA (65°N) for east–west vertical, north–south vertical, and south-tilted arrays. There is no clear best performer among subhourly models; Bifacial_radiance, bifacialVF, the System Advisor Model, and dual-sided energy tracer (DUET) comparably predict seasonal and daily changes in PV production, with root-mean-squared error (RMSE) falling in the range of 11–28% depending on the location and system orientation. PVSyst (v7.4.8), limited by hourly resolution, demonstrates RMSE in the range of 33–45%. The primary causes of high RMSE are similar for all models; using an irradiance cutoff of >100 W/m2, using clear-sky filtering, and removing time stamps with snow, lowers model RMSE to 4–13% for subhourly models and 12–25% for PVSyst. Regular meteorological station servicing is found to further decrease model RMSE by up to 3% abs. in Alaska. Finally, we model bifacial PV systems in over 250 locations between 15 and 85°N, finding that deviations between model-predicted annual insolation tend to be 2–3× higher for vertical PV systems than south-facing fixed-tilt systems. We discuss potential methods for improving vertical PV modeling and provide recommendations for high-quality field data collection in northern environments.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".