Core Data Sources
SunScore™ projections are constructed exclusively from publicly available datasets. No proprietary utility data, real-time pricing feeds, or personally identifiable energy consumption records are used in the modeling process. All data is labeled by reference year. Rate and incentive inputs reflect publicly available 2024 filings unless otherwise noted.
NREL — National Renewable Energy Laboratory
Provides location-specific solar irradiation benchmarks, peak sun hour data by ZIP code, and regional production reference values sourced from the National Solar Radiation Database (NSRDB).
EIA — U.S. Energy Information Administration
Provides residential electricity rate averages at the state and regional level (Form 861), as well as historical rate variability data used in escalation sensitivity modeling.
IRS Publications
Provides current Federal Investment Tax Credit (ITC) schedules and applicable phase-down timelines used in incentive modeling based on federal tax law.
DSIRE — NC Clean Energy Center
Provides state-level incentive program reference data, including applicable property tax exemptions and net metering policy summaries across Texas service territories.
Texas Utility Rate Filings — PUCT
Publicly available rate filings submitted to the Public Utility Commission of Texas. This data is used for TDU-specific delivery charge modeling across Oncor, CenterPoint, AEP Texas, and TNMP service territories.
Solar Production Modeling Framework
The foundation of any SunScore™ projection is an estimate of how much electricity a solar system of a given size could reasonably be expected to produce in a given location over the course of a year. This production estimate is then applied against local electricity rate data to model potential energy cost offsets.
Baseline Production Equation
Annual Production = Size(kW) × PSH × 365 × (1 - Loss Factor)Peak Sun Hours (PSH)
Peak sun hours represent the number of hours per day during which solar irradiance is sufficient to generate electricity at a panel's rated capacity (1,000 W/m²). In Texas, modeled values Generally range from 4.5 to 6.0 hours per day, with Western and Central regions reflecting higher irradiation benchmarks.
System Loss Factor
SunScore™ applies a reference system loss factor to account for energy losses attributable to temperature, inverter efficiency, wiring, and module soiling. The factor is consistent with NREL's PVWatts default assumptions and is applied uniformly at the modeling level.
Utility Rate Escalation Assumptions
A 20-year solar savings projection is significantly influenced by how residential electricity rates change across the projection horizon. SunScore™ applies a rate escalation sensitivity framework to model this dynamic without making direct price predictions.
Rate Sensitivity Equation
Future Rate = Current Rate × (1 + Escalation Rate)nEscalation as Sensitivity Band
SunScore™ applies a modeled sensitivity band of 2–3% annual escalation. This range reflects historical residential rate variability in Texas as documented in EIA data. It is a reference scenario, not a forecast.
Reference Year Calibration
Base rate inputs reflect publicly available 2024 utility rate filings. Future rate modeling extends from this base year. Higher escalation assumptions produce higher modeled savings; lower assumptions produce lower estimated ranges.
Long-Term Performance Degradation Modeling
Solar panels experience a gradual reduction in energy production capacity over time. SunScore™ incorporates a degradation variable into its 20-year production modeling to ensure long-term savings projections reflect real-world performance declines.
Performance Degradation Equation
Remaining Output = Initial Output × (1 - Degradation Rate)nReference Assumption
SunScore™ applies a baseline of 0.5% annual degradation. This compounding effect means a system would produce approximately 90.5% of its initial output by year 20. This benchmark is consistent with NREL-published research.
Compounding Curve
This curve is applied to all multi-year production estimates. Actual degradation varies by technology and climate. SunScore™ uses sensitivity modeling between 0.3% and 0.8% to produce the resulting projection range.
SunScore™ Integration Layers
SunScore™ produces a modeled savings range by integrating five variable layers into a composite estimate.
Output: Modeled Savings Range
The integration of these layers produces a modeled 20-year savings range. This range is expressed as an estimated interval rather than a precise figure, reflecting cumulative uncertainty across production, rate, escalation, degradation, and incentive variables. No single output figure should be interpreted as a guaranteed or engineered outcome.
Data Governance and Update Policy
Quarterly Rate Review
Utility rate inputs are reviewed quarterly against PUCT filings and EIA updates. When material changes are identified, affected benchmarks are updated and re-labeled with the applicable reference period.
Incentive Verification
Federal and state incentive parameters are reviewed in conjunction with IRS publication updates and DSIRE changes. ITC percentage schedules are verified at minimum annually.
Version Control
The SunScore™ Projection Engine is versioned when material changes to modeling assumptions occur, maintaining an auditable record of assumption changes over time.
Plan Variability Disclosure
Modeling applies regional benchmark rates. It does not capture the full variability of individual retail electricity plans in the deregulated Texas market. Actual costs may vary materially.
Modeling Limitations
Transparency regarding the boundaries of SunScore™ projections is a foundational principle of this platform. The following limitations apply to all outputs.
Roof Shading & Obstructions
SunScore™ does not incorporate site-specific shading analysis. Trees or structures can materially reduce output below ZIP-level benchmarks.
Orientation and Tilt
Projections assume generally favorable orientation. suboptimal tilt or north-facing planes may produce significantly less electricity.
ERCOT Retail Plan Complexity
We do not model specific retail pricing, time-of-use structures, or demand charges. Homeowners should consult specific contract terms.
Buyback Plan Volatility
Texas does not mandate net metering. Buyback rates vary by provider and are subject to change. They are not modeled as a primary savings guaranteed variable.
Consumption Shift
Model assumes stable consumption. Adding EVs or energy-intensive appliances can materially alter the energy offset value of a system.
No Professional Site Survey
Projections do not reflect physical site assessment or structural evaluation. Always obtain a site-specific assessent from a licensed professional.
Methodology FAQ
SunScore™ draws exclusively from publicly available datasets including NREL solar irradiation benchmarks, EIA electricity rate data, IRS tax credit schedules, DSIRE incentive references, and Texas utility rate filings submitted to the PUCT. No proprietary or real-time data sources are used.
No. All SunScore™ outputs are non-binding modeled estimates based on publicly available reference data and defined analytical assumptions. Actual savings will vary depending on individual site conditions, consumption patterns, system specifications, utility rate changes, and future policy developments.
Utility rate inputs are reviewed quarterly. Incentive parameters are reviewed at minimum annually, with interim reviews when material policy changes occur. All published projections include a reference year label indicating the dataset vintage.
A single point estimate implies a precision that population-level modeling cannot support. SunScore™ outputs reflect the cumulative uncertainty across production, rate, escalation, degradation, and incentive variables. A modeled range is a more accurate representation of that uncertainty than a single projected figure.
No. SunScore™ operates at the ZIP code and regional benchmark level. It does not perform site-specific shading analysis, assess individual roof characteristics, or model individual retail electricity plan structures. These limitations are documented in the Modeling Limitations section below.