P50 / P90

P50 and P90 are probabilistic energy-yield metrics used in solar project forecasting to estimate expected annual generation under different confidence levels. These metrics help solar designers, developers, EPCs, investors, and lenders evaluate both expected system performance and financial risk exposure.

P50 represents the median expected energy output—there is a 50% probability the system will generate more and a 50% probability it will generate less.
P90 represents a conservative forecast—there is a 90% probability the system will meet or exceed this energy value.

In professional solar designing workflows, accurate P50 and P90 estimates directly influence system sizing, financial modeling, loan underwriting, PPAs, and overall project bankability—especially for commercial and utility-scale projects.

‍

Key Takeaways

P50 represents expected energy output; P90 represents conservative, lender-grade output
P90 is always lower due to forecast uncertainty
Narrow P50–P90 gaps indicate high-quality modeling and stable climate
Essential for bankable solar project documentation

‍

What It Is

P50 and P90 are statistical benchmarks derived from long-term solar irradiance data, system design assumptions, and uncertainty modeling.

P50 = Expected annual energy generation under typical weather
P90 = Conservative energy estimate accounting for downside risk

Design teams calculate these values during Solar Layout Optimization, Shadow Analysis, and annual yield simulations to ensure projects are both technically accurate and financially dependable.

These metrics form the backbone of:

Investment-grade energy reports
EPC contracts and performance guarantees
PPA negotiations
Lender due diligence and credit approval

How It Works

P50 and P90 values are generated using long-term energy simulations combined with probability distributions that reflect real-world uncertainty.

1. Collect Long-Term Solar Data

10–20+ years of historical irradiance data
Satellite and ground-based weather datasets
Typical Meteorological Year (TMY) files
Seasonal and year-to-year climate variability

This data is a core input for accurate annual energy yield modeling.

2. Model System Performance

Energy modeling incorporates technical parameters such as:

PV module characteristics and PV module efficiency
Inverter efficiency and inverter loading ratio
Temperature losses using cell temperature coefficient
Shading losses evaluated through Shadow Analysis
Soiling, mismatch, wiring, and availability losses
Long-term degradation rate

3. Apply Uncertainty Factors

Uncertainty modeling accounts for variables such as:

Measurement and sensor error
Modeling assumptions
Weather variability
Degradation uncertainty
Climate-change impacts
O&M and availability risks

Reducing these uncertainties narrows the P50–P90 gap and improves project bankability.

4. Generate Probability Curve

All modeled scenarios are combined to create a probability distribution of annual energy outcomes.

5. Extract Probabilistic Outputs

P50 → 50th percentile (expected output)
P90 → 10th percentile (conservative output)

Lenders typically size debt using P90, while developers and EPCs design systems around P50 to reflect realistic operating performance.

Types / Variants

P50 (Expected Value)

Median annual energy yield
Used for feasibility studies and solar design validation
More optimistic than lender-grade values

P75

Moderately conservative scenario
Common in commercial solar projects
Balances optimism and financial caution

P90 (Bankable Value)

Industry-standard for project financing
Used in loan structuring and PPAs
High confidence for repayment capacity

P95 / P99

Extremely conservative projections
Used in utility-scale and grid-risk assessments
Reflect worst-case climate and performance scenarios

How It’s Measured

P50 and P90 are expressed as annual energy generation, typically in:

kWh/year (residential)
MWh/year (commercial)
GWh/year (utility-scale)

Conceptual Formula

P50 = Median(Expected Generation Distribution) P90 = 10th Percentile of Generation Distribution

Key Inputs That Influence Results

Irradiance variability
Performance Ratio
Temperature and shading losses
System losses and degradation
Climate uncertainty assumptions

Practical Guidance

For Solar Designers

Use P50 for layout optimization, system sizing, and early-stage proposals.
Reduce uncertainty by validating assumptions with Shadow Analysis and accurate site data.

For EPCs

Incorporate P90 into risk assessments and performance guarantees.
Align availability and O&M assumptions conservatively.

For Developers

Present both P50 and P90 in feasibility studies to clearly show expected vs. risk-adjusted yield.
Use P90 during PPA negotiations with conservative off-takers.

For Sales Teams

Explain P50 as expected performance and P90 as a financial safety benchmark.
Communicate results clearly using Solar Proposals and ROI visuals from the Solar ROI Calculator.

For Investors & Lenders

Base underwriting and debt sizing on P90.
Request independent yield assessments with detailed uncertainty breakdowns.

Real-World Examples

Residential Rooftop System (8 kW)

P50: 11,200 kWh/year
P90: 10,300 kWh/year

The homeowner uses P50 for savings estimates and P90 for conservative ROI projections.

Commercial Carport System (500 kW)

P50: 760 MWh/year
P90: 705 MWh/year

The bank approves financing based on P90 to ensure reliable debt coverage.

Utility-Scale Solar Farm (100 MW)

P50: 210 GWh/year
P90: 192 GWh/year

The utility relies on P90 for grid planning and long-term supply commitments.

‍

Releated Terms

‍