Methodology overview

InfraSure uses a chain of models. A hazard model gives physical-damage losses. A revenue model gives a merchant revenue estimate. The InfraRisk Index combines these results with site constraints into one score. This page shows how the parts connect. Each model has its own page.

The model chain

Model	Main outputs	Page
Physical-damage model	EAL, Hazard Loss, PML, Climate VaR	Physical-damage model
Revenue model	Risked Revenue, Revenue Risk	Revenue model
InfraRisk Index	Index score, benchmark percentile, red flags	InfraRisk Index and red flags
Future-climate adjustment	Future-climate Hazard Loss	Future-climate views

What the models cover

Area	Included
Geography	13,085 quarter-degree cells across the continental US.
Benchmark assets	100 MW solar plant; 100 MW wind farm with twenty 5 MW turbines.
Solar hazards	Hail, convective wind, hurricane, wildfire and flood.
Wind hazards	Convective wind and hurricane.
Revenue	P75 generation, historical local prices and adjustments.
Future climate	Hazard Loss under two emissions scenarios and three future horizons.
Portfolio VaR	Sum of individual asset P95 losses.

All results start from a standard benchmark asset. See Benchmark and asset-specific results. For what the models do not cover, see Limitations.

Example asset

The methodology pages use one example asset where they can: a 240 MW solar plant in Texas.

Result	Value
Asset value	$355.9M
Hazard Loss	1.71% (about $6.1M EAL per year)
Climate VaR (P95)	$34.9M, of which hail causes $34.8M

How each model page is organized

Each model page uses the same headings, so you can compare models:

Heading	Answers
Purpose	What question does the model answer?
Inputs	Which data does it use?
Method	How does it calculate the result?
Outputs	Which metrics does it produce?
Assumptions	What does it take as given?
Limitations	What does it not do?
Validation	How was it checked?
Version	Which methodology version does the page describe?

Methodology version

These pages describe methodology version 1.0, September 2026. Changes to the methodology are listed in What’s new.