Monte Carlo simulation
SunSolve P90 uses Monte Carlo simulation to handle uncertainty:
- Samples randomly from probability distributions representing input uncertainties
- Runs thousands of yield calculations with different input combinations
- Generates statistical distributions of possible energy yields
- Calculates percentile values (P50, P90, P95) for risk assessment
Mathematical framework
Section link: Mathematical frameworkSunSolve P90 expresses each uncertain input as:
where is the base value and is a stochastic multiplier sampled from a probability distribution.
This approach enables flexible uncertainty modeling while maintaining realistic parameter ranges.
Types of uncertainty
Section link: Types of uncertaintySunSolve P90 handles uncertainties at different temporal scales:
Sim-to-sim: Systematic biases affecting all time periods equally for that simulation
- Measurement calibration uncertainties
- Site- or system-specific biases
Year-to-year: Long-term variability adjusting all values for that year
- Inter-annual weather variation
- Changes in soiling, availability, degradation
Step-to-step: Short-term fluctuations
- Weather measurement errors
- System availability changes
Distribution design
Section link: Distribution designThe tool includes an interactive probability distribution plotter to help visualize and design probability density functions (PDFs):
Understanding distributions:
- If a distribution has a mean of 1: equal probability of increasing or decreasing the input
- If skewed towards 0: more likely to decrease the input
- If skewed towards higher values: more likely to increase the input