PVsyst Simulation Outsourcing: How to Brief a Vendor

You have decided to outsource the yield simulation. Good. That is the right call for most developers and EPCs. A PVsyst seat, a trained modeler, and a review engineer cost more than a full year of outsourced reports. The failure point is rarely the decision. It is the brief, and the same story repeats.

The vendor delivered a technically valid report. The lender rejected it anyway. The meteo file was wrong for the site, the P90 had no stated uncertainty basis, or the layout version in the report was two revisions old. Every one of those failures traces back to a missing line in the brief, not a missing skill in the vendor.

Quick answer. PVsyst simulation outsourcing works when the vendor gets six things in writing: a frozen, versioned layout, a named meteo dataset with fallback, a P50/P90 specification, loss parameters, the lender or IE checklist, and revision terms. A brief covering all six usually returns a bankable report in one revision cycle. Turnaround is agreed at scoping.

This guide is for Indian utility-scale developers lining up debt for a 25 MW project. It also serves US C&I pipeline managers and any EPC owner who has already decided to outsource and now wants the report to come back bankable the first time.

If you are still deciding whether to outsource at all, our founder’s playbook on outsourcing solar design covers that decision. If you want to know what a finished report costs and contains, read our companion piece on the PVsyst report service. This post covers the workflow in between: how you brief the vendor, and what happens after you do.

What PVsyst Simulation Outsourcing Actually Includes

Outsourcing a PVsyst simulation means handing a specialist team your project inputs and receiving an energy yield assessment (EYA) built in PVsyst, the industry-standard PV simulation software. The deliverable is not a software printout. It is an engineering document with a methodology section, a meteo data justification, a full loss tree, and exceedance-probability yield estimates that a lender’s independent engineer can audit line by line.

A complete outsourced scope usually covers five work packages:

1. Model setup. System definition, module and inverter selection from the PVsyst database, orientation, and electrical configuration.
2. Meteo data integration. Importing the site climate file, checking it against satellite sources, and documenting the choice.
3. Shading scene. A 3D near-shading model with terrain, rows, and obstructions. On flat utility sites this is mostly inter-row shading. On rooftops it includes parapets, HVAC units, and trees.
4. Loss tree and simulation. Soiling, availability, DC and AC ohmic losses, mismatch, degradation for year 1, and grid limits.
5. Uncertainty and reporting. P50, P75, P90, and sometimes P99 yields with an uncertainty budget, plus the report document itself.

What is usually excluded: site survey, geotechnical work, grid studies, and structural design. Those feed the simulation but are separate scopes. For the definition of bankability itself, see our guide to bankable PVsyst reports.

The Inputs a Vendor Needs From You Before Simulation Starts

The single biggest cause of rework is the vendor starting with half the inputs and guessing the rest. A professional vendor will send you an input checklist. Send everything on it in one package, before the clock starts. Here is what that package contains for a utility-scale project:

• Frozen layout (DWG plus PDF), version-numbered. Array boundary, row pitch, tilt or tracker type, inverter stations, and access roads. This is the master input. Everything in the report traces to it.
• Module and inverter datasheets. Exact models with PAN/OND files if you have them. If procurement is undecided, name the two candidates and ask the vendor to run both.
• Site coordinates and elevation. A KML boundary or survey file. Coordinates drive the meteo extraction, so accuracy matters to within the project boundary, not the district.
• Soiling expectations. Measured data if you have an operating reference plant nearby. Otherwise, regional studies. Our post on soiling loss modeling for India and Africa shows how much this one input moves the yield.
• Availability and grid constraints. Planned outage days, curtailment terms in the PPA, and any grid export limit.
• The lender’s checklist, if one exists. Many lenders and independent engineers publish their EYA requirements. Hand it over on day one.

Two of these deserve emphasis. First, the frozen layout. If your civil team changes the boundary after simulation starts, the shading scene and DC capacity both change, and the report restarts. Freeze it, version it, and treat every later change as a formal revision with a change note.

Second, the lender checklist. A report built to a named checklist passes review. A report built to the vendor’s default template gets compared against the checklist later, and every gap becomes a query.

Meteo Data: The One Input That Decides Bankability

Meteo data is the largest single source of yield uncertainty, and the first thing a lender’s technical advisor checks. The choice is not academic.

On Indian utility projects we have seen a 2 to 4 percent swing in first-year energy between satellite sources alone. On a 25 MW plant yielding about 1,650 MWh/MW at a ₹3.00/kWh tariff, a 2 percent yield swing is roughly ₹25 lakh per year of revenue, up or down, for 25 years.

There are three families of sources. Solargis is a satellite-derived dataset with global coverage and a strong validation record; it is the most commonly accepted source for lender work in India, Africa, and the Gulf, per Solargis product documentation (2025). Meteonorm combines ground stations, satellite data, and interpolation, per Meteonorm (2025). NSRDB, from the US National Renewable Energy Laboratory, covers North America and is free, per NREL’s NSRDB (2025). Our detailed comparison of these sources sits in PVsyst meteo data: Meteonorm, Solargis, and NSRDB.

Dimension	Solargis	Meteonorm	NSRDB
Data type	Satellite-derived	Hybrid station + satellite	Satellite-derived
Coverage	Global	Global	North America focus
Lender acceptance (India/Africa)	Widely accepted	Accepted, sometimes questioned on interpolations	Rare outside US
Cost per site file	Paid, per site	License-based	Free
Best for	Bankable EYA in India, Africa, Gulf	Pre-feasibility, rooftop C&I	US projects, pre-feasibility

Your brief should name the primary dataset, the comparison dataset, and the rule for resolving a gap. A good line: “Primary: Solargis TMY for site coordinates. Cross-check: Meteonorm. If GHI differs by more than 3 percent, flag it in the report and use the more conservative source for the P50.” That single sentence eliminates the most common lender query we see.

How to Specify P50, P90, and Uncertainty in Your Brief

P50 and P90 are exceedance probabilities. The P50 is the yield the plant will exceed in 50 percent of years. The P90 is the yield it will exceed in 90 percent of years. Lenders size debt on the P90, because it represents a conservative, one-in-ten downside case. Our explainer on P50, P90, and P99 yield reports covers the math in detail.

Here is what most developers get wrong. They ask for “P50 and P90” and stop. The gap between those two numbers is entirely a function of the uncertainty assumptions, and those assumptions are negotiable unless you pin them down.

A vendor using optimistic uncertainties produces a tight P90 close to the P50. The yield looks great. The lender’s advisor then applies their own uncertainty budget, the P90 drops, and your debt sizing shrinks mid-negotiation.

Specify five items in writing:

1. Which exceedance levels. P50, P75, P90 at minimum. Some DFIs want P99. Ask the lender first.
2. The uncertainty budget. Irradiance uncertainty, model uncertainty, and inter-annual variability, each stated as a percentage. PVsyst’s own methodology for this is documented in the PVsyst uncertainty documentation (2025).
3. The time basis. Year 1 yield, and a stated degradation rate for later years.
4. Which probability the report headline uses. Lender reports should headline the P50 for energy and state the P90 prominently for debt sizing.
5. Long-term vs satellite-period basis. State the number of years of data behind the statistics.

A two-line example that works: “Report P50, P75, and P90 for year 1. Use irradiance uncertainty of 4 percent, model uncertainty of 3 percent, and inter-annual variability of 5 percent, or justify alternatives in the methodology section.” Now the vendor’s assumptions are auditable, and so is the final number.

The One-Page Simulation Brief

We compress everything above into a one-page document we call the One-Page Simulation Brief. It has six blocks. Every block is two or three lines. A vendor who receives all six can start immediately, and the report comes back with almost no structural queries.

Apply it like this on your next project: copy the six headings into a document, fill two or three lines under each, and attach it to the purchase order. The brief then becomes the acceptance test. When the report arrives, you check it against the brief, not against memory. That is the whole trick.

Project Phases and Revision Terms

Turnaround for a lender-grade yield assessment is agreed at scoping, once the vendor has seen a complete input package. Be wary of any promise that sounds too fast for a lender-grade report, because the shading scene or the uncertainty analysis is usually what gets skipped.

A realistic engagement runs through these phases:

Phase	What happens	Your job
Input review	Vendor checks the package, raises gaps	Answer queries promptly
Model setup	System definition, meteo import, shading scene	Approve the setup summary
First simulation	Loss tree, first P50/P90 run	Review the loss tree first
Draft report	Full document with methodology	One consolidated review pass
Final	Revisions incorporated, final issue	Sign off against the brief

Two rules protect this schedule. First, consolidate your review comments into one pass. Five dribbled comment rounds cost more calendar time than one thorough one. Second, agree the revision terms in the brief: the number of included rounds and how layout changes are handled. Our solar design SLA and turnaround guide has sample wording. That last clause is what keeps a “small boundary tweak” from becoming a free re-simulation.

The performance ratio and specific yield in the draft are your fast sanity checks. For a well-designed fixed-tilt plant in high-irradiance India, a first-year PR in the 78 to 84 percent range is normal. A draft showing 88 percent means someone set a loss to zero. Find it before the lender does. Our guide on common PVsyst errors that kill bankability lists the usual suspects.

Freelancer vs Specialist Firm vs In-House: Who Should Run Your Simulation

There are three realistic sourcing options, and each wins in a different situation (our freelancer vs firm comparison goes deeper). The wrong choice here costs more than any modeling error.

Dimension	Freelance PVsyst operator	Specialist engineering firm	In-house team
Cost basis	Lowest, self-reviewed	Quoted to scope, includes review	Loaded salary plus licenses
Turnaround	Variable	Agreed at scoping	Depends on workload
Review depth	Self-reviewed	Second-engineer review	Depends on staffing
Lender acceptance	Weak, no institutional record	Strong, documented track record	Strong if experienced
Continuity risk	High, single person	Low, bench depth	High, attrition
Best for	Pre-feasibility screening	Bid-stage and lender-grade work	Continuous multi-GW pipelines

Verdict. For a financed project, use a specialist firm with a documented lender track record, and verify it the way you would verify a contractor: named reference projects, a sample report, and the review process in writing. For pre-feasibility screening, a freelancer is fine. Build in-house only when simulation volume justifies a dedicated modeler plus a reviewer, because a bankable report needs both.

The continuity point deserves weight. India added record solar capacity through 2024 and 2025, per MNRE (2025), and global annual installations passed 600 GW in 2024, per the IEA-PVPS Snapshot (2025). Experienced PVsyst modelers are in short supply everywhere. An outsourcing relationship with a bench-backed firm is, in practice, a hedge against your own attrition.

Red Flags When Choosing a PVsyst Vendor

The vendor conversation itself tells you most of what you need to know. These are the patterns we see behind failed engagements:

1. No meteo question in the first call. A vendor who accepts any site without asking about climate data will use whatever file is on their hard drive.
2. No version control on inputs. If they do not ask for a layout version number, they cannot tell you later which layout the report models.
3. No methodology section in their sample report. Ask for a redacted sample. If the methodology is two paragraphs, the report will not survive an IE review.
4. A P90 with no uncertainty budget. Ask “what uncertainties sit between your P50 and P90?” A one-line answer like “standard values” is a fail.
5. Unlimited revisions at no charge. This sounds generous. It means the quote assumes the first version will be wrong, or that “revisions” exclude everything you will actually need.
6. No named reviewer. Bankable work needs a second engineer signing the review sheet. Ask who reviews, and what they check.
7. No lender references. A firm doing lender-grade work can name the institutions their reports have passed, at least by category. Our post on the lender acceptance register explains how that record works.

The meteo and uncertainty checks above are PVsyst-specific and catch problems earliest. The broader partner-vetting checks, from insurance to bench depth, apply to simulation vendors exactly as they apply to design vendors.

How Heaven Designs Helps With PVsyst Simulation Outsourcing

Our simulation team runs PVsyst yield work for utility-scale developers, EPCs and C&I portfolios using the workflow in this post: an indexed input package, a named meteo source with a cross-check, a stated uncertainty budget, and a report checked against your lender’s checklist before it leaves our review desk.

For an example of our engineering scope on a complex industrial site, see our 11 MW Gulf tin-shed rooftop case study.

• Solar Ground Mount Design: utility-scale layouts, tracker yield, and the frozen layout your simulation brief starts from.
• MW-Scale Project Management Consultancy: owner’s engineer support, including EYA review and lender query handling.

If you already have the input package ready, send it through the contact page to get a quote for your scope. We reply within 1 business day, and turnaround is agreed at scoping. And if you already have a report from another vendor, our guide on how to validate a PVsyst report before your lender does walks through the review checklist we use.

FAQ

What drives the cost of PVsyst simulation outsourcing?

Cost depends on scope: report tier, project size, meteo data purchases, the 3D shading scene, the uncertainty analysis, and revision rounds. Site-specific Solargis files are usually passed through at cost, so ask whether meteo data is included.

Send your input package to get a quote for your scope.

What should I send a vendor before they start a PVsyst simulation?

Send a frozen, version-numbered layout in DWG and PDF, module and inverter datasheets, site coordinates with a KML boundary, soiling and availability expectations, grid or curtailment constraints, and the lender’s EYA checklist if one exists. Package it as one indexed folder with a README listing every file and version. A complete package up front is the difference between a clean delivery and a long email thread.

Which meteo dataset should I specify in the brief?

For lender-grade work in India, Africa, and the Gulf, specify Solargis as primary with Meteonorm as cross-check. For US projects, NSRDB is free and widely accepted. Write the fallback rule in the brief: if the two sources differ on GHI by more than 3 percent, the report flags it and uses the more conservative value for the P50. That rule preempts the most common lender query.

What P90 assumptions should a bankable report state?

A bankable report states the exceedance levels reported (P50, P75, P90 at minimum), the uncertainty budget behind them, and the data period. Typical starting points are 4 percent irradiance uncertainty, 3 percent model uncertainty, and 5 percent inter-annual variability, with justification for any deviation. The report should headline P50 for energy and state P90 prominently, because lenders size debt on the P90.

How long does an outsourced PVsyst report take?

Turnaround is agreed at scoping and depends on the report tier and a complete input package. The schedule usually breaks into input review, model setup, first simulation, draft, and final. Your response time on queries is the biggest variable you control.

Is a freelancer good enough for a bankable PVsyst report?

Usually not. A freelancer is a reasonable choice for pre-feasibility screening where no lender reads the output. For financed projects, the report needs a second-engineer review, a documented methodology, and an institutional track record that a single operator cannot show. The risk is not modeling skill; it is review depth, continuity, and lender acceptance.

What is the difference between a bid-stage and a lender-grade PVsyst report?

A bid-stage report supports your tariff calculation with a P50 estimate, standard meteo data, and a simplified loss tree. A lender-grade report adds a site-specific paid meteo dataset, a full 3D shading scene, P75 and P90 exceedance values, a stated uncertainty budget, and a methodology section an independent engineer can audit. Plan for a lender-grade rebuild after financial close shortlisting if you commissioned only a bid-stage study.

Can I reuse one PVsyst report across multiple lenders?

Yes, if the report was built to a lender-grade standard from the start. Most lenders and independent engineers accept the same core EYA, then ask institution-specific queries on top. The practical approach is to build the report against the strictest checklist you expect to face, then answer incremental queries per lender. A report built to the weakest checklist gets rebuilt, which costs more than building it right once.