Comparing Solar Production Estimates: A Guide to Navigating Different Tools and Data Sources

Understanding Production Discrepancies When Rebuilding Remote Designs

When moving a project from one remote design software to another (such as rebuilding an Aurora design within Scanifly Remote or Scanifly On-site), it is common to see variations in production estimates in kWh, capacity, shading values (ie TSRF) and offset. Naturally, you may ask, “which software is more accurate”?

This guide will help you understand why these differences happen, how to establish an acceptable baseline, and when a discrepancy warrants a deeper look.

The Reality of Remote Data: Manual Modeling & Layers

A remote design uses non-real-time, offsite data (such as satellite or aerial imagery) to design a solar array. It’s a strong starting point, but still an interpretation of the property. Because no one has visited the physical site yet to collect the most updated, accurate information, remote tools require a designer to manually construct the scene.

When rebuilding a design, variations happen because:

Manual Scene Recreation: A designer must manually place shading obstructions, trees, and roof sizes based on their individual interpretation of the site. For example, two different designers might model the exact same tree with slightly different heights or canopy densities. Another common example might be how designers model subtle nuances in roof mounting planes, such as dormer sizing, crickets behind chimneys or nearby homes and structures, which ultimately affects the size of roofs and space for panels to fit.
Conflicting Data Layers: Remote platforms pull from various 2D imagery and 3D LiDAR layers. Depending on when those layers were captured or how a software engine processes them, minor structural differences can occur. This can be a different process for each organization, including which date or season Nearmap file is pulled, which LiDAR files are available, and if you refer to a LiDAR layer from NOAA, another US government agency, or Google3D.
Simulation Settings: Less obvious, but equally if not more essential, is double checking parity of simulation settings between two softwares. This includes ensuring consistency in production engine (eg PVWatts version 6 versus 8), system loss factors, and treatment of electrical hardware. Each variable could swing a project’s output by percentage points.

Note: Ultimately, a remote design is an estimate of a simulated environment. True precision is unlocked once high-resolution, on-site drone data or manual measurements are captured. The Solmetric SunEye and Pathfinder, frequent tools used in solar’s early days, also provided that onsite precision.

The +/-5% Rule: When Should You Investigate?

Within +/- 5%: If your Scanifly Remote estimate is within 5% of your original remote design, no action is needed. This is considered a comparable and acceptable variance between two remote modeling engines.
Beyond +/-5%: If the discrepancy is greater than 5% (higher or lower), it is worth diving into the data and requesting an audit from our team.

The "Spot the Difference" Audit Checklist

If you have a discrepancy larger than 5%, use this checklist to compare the specific inputs of both designs. Often, a major production swing comes down to a few mismatched settings.

If you would like to use this template, please use this link: Google Sheet - Solar Production Comparison

Tip: If you are comparing a simple, south-facing roof with zero shade, almost any discrepancy is going to be driven by the Simulation Engine or System Losses. However, if you are working with a highly cut-up roof, dormers, or heavy tree coverage, variations in Manual Scene Recreation (how the user drew the 3D environment) are usually the primary cause of the difference.

1. Project Basics

Scanifly Project Name & Project Link: Confirm exact Project Addresses match & the property is the same.

2. Geometry & Environmental Values

Roof Planes, Pitch, & Azimuth: Do the roof slopes and orientations match exactly? A few degrees of difference will alter production.
Shading Coefficients (ASA / TSRF): Check the specific shade values applied to the modules.

3. Equipment & Hardware Profiles

PV Module Model & Quantity: Ensure the exact same module profile and count are used.
Inverter Model: Verify the Scanifly simulation settings match the inverter/array ratio as the Aurora inverter selection, as different engines calculate clipping and efficiency curves uniquely.

4. System Losses & Derates

The "Hidden" Variables: Look at the fine print for soiling, snow, wiring, and age degradation.
Shade Loss Factors: Isolate the exact monthly vs. annual loss factors used in the original simulation.
Simulation Engine Version: Verify which calculation engine the other platform uses.
- Scanifly utilizes NLR/NREL’s PVWatts v8—the most current standard. Some legacy platforms still run on deprecated versions (like PVWatts v6), which can cause unexpected difference in production.

💡 Support Tip: Some design platforms make specific system loss and derate data difficult to find in their standard proposal views. You may need to export a full shade or simulation report to uncover the exact percentages being applied.

How to Handle Next Steps

If you have gone through the checklist and the designs still don't align, don't waste time going down a rabbit hole trying to force the math to match.

Option 1: Request a Scanifly Audit. Reach out directly to our support team with your Scanifly project link and a copy of the alternative design or shade report. We will gladly run a side-by-side audit with you to identify the exact technical driver of the variance.
Option 2: Transition to On-Site Reality. Remember that the ultimate source of truth is an on-site drone scan. If a remote design feels uncertain, moving the project to an on-site workflow will eliminate the guesswork and provide an indisputable 3D model of the property.