For more than a decade, PlantPredict's physics-based energy models have been independently reviewed by the industry's leading engineering firms and used in the development of hundreds of gigawatts of utility-scale solar. This page is the evidence.
It's the quiet tax on every project financing: teams model in the tool that fits how they work, then re-run everything in PVsyst because that's the report the bank has always seen. Two models. Two sets of assumptions to reconcile. And when the numbers drift apart during IE review, the difference tends to come out of your energy estimate — and your debt sizing.
Lenders don't actually require PVsyst. They require confidence: a physics-based model, validated by independent engineers, with a methodology their IE can scrutinize. That's a standard PlantPredict was built to meet — and has been meeting for over a decade.
PlantPredict uses the same industry-standard, physics-based modeling approach lenders already trust — transposition, thermal, spectral, and loss models grounded in published research. In ICF's 2026 independent benchmark across 15 U.S. projects, the mean energy-yield difference between PlantPredict and PVsyst was −0.16% — statistically indistinguishable from zero. The model isn't the risk. Unfamiliarity is — and that's what this page exists to remove.
DNV GL, Black & Veatch, ICF, and Leidos each independently reviewed PlantPredict's algorithms and outputs in 2016–2017. In 2026, new independent engineering reviews from ICF and Solphi re-validate the platform — including its new 3D scene engine — against current hardware, current physics, and current lender expectations.
Every algorithm in PlantPredict is publicly documented. Your IE doesn't have to take our word for anything — they can audit the methodology line by line. That level of transparency isn't available from closed desktop tools.
Read the algorithm documentation →Bankability is earned one engineering review at a time. PlantPredict's validation record spans two eras: the foundational reviews that established the platform, and a new wave of 2026 studies validating its latest physics.
Benchmarked PlantPredict against PVsyst and operational data; concluded accuracy equivalent to the industry's bankable modeling tools.
Confirmed PlantPredict's suitability as a bankable energy assessment tool for utility-scale project finance.
Benchmarked against PVsyst across four diverse projects; concluded uncertainty comparable to PVsyst and other bankable platforms.
Full technology review; concluded PlantPredict is suitable for bankable, utility-scale PV energy assessments.
These reviews underpinned the financing of GW-scale portfolios — including projects where PlantPredict was the energy model of record.
First IE review of PlantPredict V12's 3D scene engine. All 18 algorithm categories rated “meeting expectations”; benchmark across 15 U.S. projects (1.1–235 MW) showed a mean yield difference vs. PVsyst of −0.16% — no systematic bias.
Head-to-head vs. PVsyst v8.1 across 20 simulation variants on three continents: every simulation within 1.0%, average difference under 0.5%, and no material deficiencies found.
Prompted by PlantPredict's next-generation 3D scene model (launched October 2025), these reviews benchmark the platform against today's plant designs — bifacial modules, terrain-aware backtracking, sub-hourly simulation — and the loss factors lenders scrutinize in 2026. ICF rated the 3D scene approach as “reflecting higher-fidelity approaches than is typical in solar industry-accepted PV performance modeling.”
Beyond IE reports, PlantPredict's accuracy is documented in the peer-reviewed literature — including the first-ever onymous blind assessment of PV modeling software, led by Sandia National Laboratories (Solar Energy, 2026), where PlantPredict stood alongside the industry's most established tools.
See the full validation library →Don't take the summary — read the studies your IE will want to see.
PlantPredict wasn't born as a startup's pitch deck. It was built inside First Solar — one of the world's largest solar manufacturers — to model the company's own utility-scale plants and underpin performance guarantees across its project fleet. When your own capital rides on the model's accuracy, the physics gets rigorous fast. Terabase Energy acquired PlantPredict from First Solar in 2021 and has been advancing it ever since.
Read the full PlantPredict story →A bankable model has to keep pace with the plants it models. PlantPredict's physics advances through open research, industry collaboration, and partnerships with the specialists lenders already respect.
PlantPredict engineers publish in the peer-reviewed literature and participate in the PVPMC community — including co-authoring the Sandia National Laboratories–led blind assessment of PV modeling software (Solar Energy, 2026), the first study of its kind. The people behind the model show their work in the forums where IEs and national labs stress-test modeling science.
Every model — transposition, thermal, shading, soiling, degradation — is publicly documented. IEs audit the methodology directly instead of reverse-engineering a black box.
Browse the Models & Algorithms docs →Launched October 2025: full 3D scene-based shade simulation for the plants actually being built in 2026 — terrain-aware backtracking, bifacial modules, complex topography — running on cloud GPUs. In Solphi's 2026 review, PlantPredict efficiently simulated a 1 GW+ tracker system with a full 3D scene — something not practical to replicate in PVsyst on the same hardware.
Lender-grade uncertainty quantification with PowerUQ, site-specific soiling with PVRADAR, and ray-traced bifacial research with SunSolve by PV Lighthouse. PlantPredict is the hub of the industry's most capable bankable modeling stack.
And because PlantPredict is built by Terabase — a software company, not an engineering consultancy — every IE firm can adopt it without funding a competitor.
Run PlantPredict alongside your existing tool on a live project. Teams consistently see agreement within the range documented in the independent studies — and get their results in a fraction of the time.
Send your independent engineer the validation library above. It's built to answer exactly the questions an IE review raises: methodology, benchmarking, and third-party results.
Our engineering team joins IE and lender diligence calls, walks through the methodology, and supports side-by-side reconciliation. You're not introducing a new tool alone — you're introducing one with a decade of review behind it and a team that shows up.
PlantPredict models have been used in the development and financing of utility-scale projects for years, supported by six independent engineering reports — DNV GL, Black & Veatch, ICF, and Leidos (2016–2017), and ICF and Solphi (2026). Multiple reports explicitly conclude PlantPredict is suitable for bankable, utility-scale energy assessments. Acceptance is ultimately each lender's call, which is why we provide the full validation library and join IE diligence directly.
In ICF's 2026 benchmark across 15 U.S. projects, the mean energy-yield difference was −0.16% — statistically indistinguishable from zero, with no systematic bias. In Solphi's 2026 study, every one of 20 benchmarked simulations landed within 1.0% of PVsyst, averaging under 0.5%. Same physics, same loss factors, same bankable numbers — and the studies document every comparison in detail.
Two generations of independent engineering firms: Leidos (2016), Black & Veatch, ICF, and DNV GL (2017), and new 2026 reviews from ICF and Solphi — plus peer-reviewed studies including the Sandia National Laboratories–led blind assessment (Solar Energy, 2026).
Yes — all algorithm documentation is public. Your IE can review every model without an NDA or a license, and the User Guide documents the full workflow.
docs.plantpredict.com/models →PlantPredict was created inside First Solar to model its own utility-scale fleet and is now developed by Terabase Energy, a software and automation company dedicated to utility-scale solar. Terabase is not an engineering consultancy — IEs and developers can adopt PlantPredict without a conflict of interest.
Join hundreds of organizations that trust PlantPredict for bankable energy predictions. Start with a free account, get a free trial of Pro features, or book a demo with our team.