Underwriter-Grade Risk Decomposition for Utility-Scale Solar & Battery Storage
Reduce underwriter friction on your next renewable placement. Carrier-defensible loss curves, parametric trigger pricing, and binding subjectivities you can drop into a submission today. No account data required from you.
Pilot placements in progress with Tier-1 broker partners · methodology auditable to federal & scientific datasets cited below.
——
Composite Risk Grade
—
Avg Annual Loss
—
PML 1-in-100 Yr
—
PML 1-in-250 Yr
Site Profile Footprint outlined · auto-zoomed satellite
Satellite
—
Illustrative submission · West-TX hyperscaler-offtake solar + BESS
Asset—Status—Location—Coords—TIV—Footprint—
Dominant Perils Score 0-100 · higher = more risk
Underwriting Verdict
—
Rate Guidance
—
Loss Decomposition by Peril % of Total Tail-Value-at-Risk
What the Broker Receives
1. Defensible Risk File
12-peril site-resolution decomposition · exceedance-probability loss curve · tail-value-at-risk attribution. Every line item traces to a published equation and an authoritative dataset — the underwriter scores against it, not around it.
2. Parametric Tail Cover
Pre-priced hail kinetic-energy, sustained wind, and solar-shortfall triggers. Worked example for this site: hail KE >10 J/cm² triggers $5M payout · ~92 bps on $5M ≈ $46k/yr premium · settles in 5 days via NEXRAD MESH shapefile vs ~9 months for indemnity hail.
3. Engineered-Solutions Credit
5 binding subjectivities mapped one-to-one to carrier credits — deductible buy-down, BESS sublimit lift, capacity expansion on hardening accounts. The artifact that lets a broker re-quote a placement.
Methodology + peril scoring deliverable today via Earthflow Core · loss-aggregation engine + carrier-validated credits in pilot with Tier-1 broker partners.
Methodology & Audit Trail
Authoritative Data Sources
Every number is auditable to a published, federal-grade source: hail kinetic-energy radar climatology · cloud-to-ground lightning strike density · national earthquake hazard maps and site soil profile · federal flood-zone mapping and drainage models · national wildfire perimeters, fuel models, and ignition probability · national soil survey · national wind and solar resource atlases · downscaled climate change projections through 2055.
Federal LOMA validating Zone X assignment (current FEMA confidence low; map amendment required pre-NTP).
05
Battery recertification
18-month cell-level battery management system recert post-COD.
↗ Live Interactive Demo
Re-price the layer in real time with parametric trigger sliders · Ask Cirra AI underwriter-style questions · One-click PDF download of this risk file.
earthflow.orbyfy.com/earthflowRE?site=meta
Orbyfy™EarthflowReinsurance Risk File
Site Profile & Composite VerdictPage 2 of 6
Tactical · Site Footprint
Regional · West-TX Peril Context
Footprint—
Transmission—
Coordinates—
Grid / ISO—
—
—
—
—
Capacity
—
Status / COD
—
TIV
—
Cedent
—
Layer
—
Period
—
Orbyfy™EarthflowReinsurance Risk File
Twelve-Peril DecompositionPage 3 of 6
Per-Peril Score · AAL Contribution · Trend · Evidence Sourced from federal hazard, soil, weather, and fire data
↗ Re-price any peril live
Toggle parametric trigger sliders on the live demo to reprice individual perils in real time.
earthflow.orbyfy.com/earthflowRE?site=meta
Orbyfy™EarthflowReinsurance Risk File
Stochastic Loss EconomicsPage 4 of 6
Loss KPIs Industry-standard 10,000-year Monte Carlo simulation — required to stably estimate the 1-in-250-year tail · engine output is demo synthesis pending pilot engagement
—
Avg Annual Loss
—
—
PML 1-in-100
—
PML 1-in-250
—
TVaR 99%
—
Reinst-adj PML
—
Tech Rate
Exceedance-Probability Curve As-built · with engineered mitigation overlay
As-built loss curveWith recommended mitigationVertical refs: 100-yr · 250-yr return periods
TVaR Contribution by Peril
Recommended Binding Subjectivities Mapped one-to-one to engineered-solutions credits
01
Class-4 hail-rated modules
ASTM E1038 Class-4 (highest impact rating), contractually fixed at construction award for West-TX SCS exposure.
Live loss-aggregation engine (real AAL, PML, TVaR per site) · parametric trigger pricing calibrated to your cedent portfolio · carrier-validated bps credits on subjectivities
Demo in This Artifact
TIV · cedent name · layer attachment · premium rate · all flagged DEMO chips on cover
Orbyfy™EarthflowReinsurance Risk File
Parametric Triggers & Climate StressPage 5 of 6
Parametric Trigger Pricing 5-day settlement · radically faster than 9-month indemnity average
↗ Settlement velocity advantage
Parametric: 5-day settle · Indemnity claim average: 9 months · 270-day cash-flow advantage to the cedent post-event.
Radar hail energy · Federal wind & solar atlas
Climate Stress · Long-Dated Risk Trajectory Through 2055 Site-specific composite + climate-sensitive perils projected under federal climate scenarios · demo synthesis
Risk Trajectory Heatmap · 2025 → 2055
Trajectory Analysis · RCP 8.5 Scenario
—
Orbyfy™EarthflowReinsurance Risk File
Methodology & Audit TrailPage 6 of 6
Authoritative Data Sources Every number traces back to a published federal or scientific dataset
Earthflow Physics AI™ engine · 70+ authoritative data sources · 1,180+ fields per site · sub-minute turnaround.
Pilot Status
Layer 2 loss-aggregation engine in active pilot with Tier-1 broker partners. Demo synthesis values on this artifact will be replaced by real engine output at pilot engagement.
Methodology Reference
Earthflow Reinsurance Methodology
Peer-reviewed equations · every data source cited · vulnerability-curve calibrations · peril-correlation matrix construction · climate downscaling approach. The underwriter-defensible reference companion to this risk file.
Attach this 6-page risk file to your next solar / BESS submission packet.
2
Quote the 5 binding subjectivities verbatim during carrier negotiation.
3
Use parametric trigger pricing as your tail-risk alternative when carriers won't quote indemnity hail or BESS thermal.
Drop-in ready for your submission
Renewal-Ready
Orbyfy Earthflow Recognition
Awards
Edison Awards 2026 Gold · IDC Leader (Utility Solutions) · Boston Business Journal Top Innovator
Innovation Cohorts
NVIDIA Inception · Greentown Labs Boston · Black & Veatch IgniteX · Siemens Next47 Start-up School
Contact
www.orbyfy.com
Insured Asset
—
—
Capacity
—
—
COD
—
—
Cedent
—
—
TIV
—
—
Layer
—
—
Period
—
—
🚧
Construction All Risk · CAR / EAR + DSU view active
—
Live Satellite View · Esri World Imagery
—
Maxar · Earthstar Geographics · z=16
Site Footprint
—
—
Distance to Coast
—
—
FEMA Flood Zone
—
—
Seismic SDC
—
—
—
—
—
Composite Risk GradeiComposite Risk GradeA 0–100 score and letter grade synthesizing all 12 perils, weighted by each peril's AAL contribution. The verdict (Bind · Bind w/ Subjectivities · Refer · Decline) follows a deterministic decision tree applied after the stochastic loss aggregation.14 risk modules + kpi_analysis.py composite + verdict rulesFirst-look bind / refer / decline triage in one number
—
—/100
—
—
Peril Decomposition · 12 Perilsi12-Peril DecompositionEvery peril scored 0–100 with AAL contribution, trend arrow, and 24-month sparkline. Click any tile to expand the Evidence Drawer — see the underlying data source, refresh timestamp, and short methodology line per peril.Physics AI™ engine · NEXRAD · NREL Wind Toolkit · USGS ASCE · FEMA NFHL · NIFC · SCADA · BMSIdentifies the dominant peril(s) · targets subjectivities · auditable to sourceEarthflow modules · 14 risk · 8 detection
Loss Economics · StochasticiStochastic Loss AggregationMonte Carlo simulation across 10⁴–10⁶ synthetic years producing AAL ± 1σ, PML 1-in-100/250, TVaR 99%, the exceedance-probability curve, and TVaR-by-peril attribution. Toggle the mitigation pills below the curve to see net-of-mitigation deltas.loss_curves.py · Stage 3 aggregation preserving peril correlationsLayer attachment · treaty capacity · tail-driver attribution
AAL—
AAL ±1σ band—
PML 1-in-100—
PML 1-in-250—
TVaR (99%)—
Reinst-adj PML—
TVaR by peril:
Premium GuidanceiPremium GuidanceTechnical rate = (AAL / TIV) + risk load + expense load + capital charge against PML 1-in-250. The per-peril deductible matrix reflects the recommended subjectivities for this site.Stage 4 financial layer · market-calibrated load factorsIndicated rate benchmarked vs the cedent's portfolio book
Technical Rate
—%
—
Indicated Premium
—
—
Construction Phase Dashboard · CAR/EAR + DSUiCAR/EAR + DSU Construction ViewActive when Phase = Construction and the site has a construction block. COD countdown, DSU revenue-at-risk curve, EPC performance scorecard, supply-chain status, 90-day weather-delay forecast, and 6-phase Sentinel-2 imagery timeline.Earthflow Detect phase_analyzer.py · 3-source weather (NOAA + NREL + Open-Meteo) · EPC / CMMS feedCAR/EAR + DSU pricing · schedule-risk transparency · supply-chain alertsphase_analyzer.py · Sentinel-2 + 3-source weather
—
Days to COD
—
—
DSU Revenue at Risk · per day of delay
Daily revenue loss (P50)—
90-day cumulative—
EPC Performance Scorecard
Supply Chain Statusmodule + BOS inventory
90-day Weather Delay ForecastOpen-Meteo + NOAA
Parametric Triggers · Live Distance-to-ThresholdiParametric Trigger PricingPre-agreed sensor thresholds with live distance-to-trigger gauges. Drag any slider to re-price the technical rate in real time. Common triggers: hail kinetic energy, sustained wind, GHI underperformance, BESS cell ΔT, storm-surge depth.NEXRAD MESH · NWS gauges · on-site weather + BMS · parametric_pricer.js5-day settlement velocity vs 9-month indemnity claims averageNOAA · NEXRAD · NREL POA
⚡ Settlement velocity: Parametric triggers settle in 5 business days vs. 9-month average for indemnity claims · radically faster liquidity to the cedent post-event.
IoT / Condition Monitoring · Predictive MaintenanceiIoT Condition MonitoringLive inverter health grid, string-level fault tree, BESS cell ΔT, on-site weather station POA + hail detector, EL scan schedule, and predictive-failure alerts. All telemetry is read-only by architectural design.SCADA Modbus TCP · BMS REST/MQTT · weather station — see Data & Cyber methodologyCatches breakdown signals before they become claims · feeds energy-shortfall triggeriot_stream_aggregator · SCADA + EL scans
Inverter Health Gridhealthy · watch · fault · predicted-failure
BESS Thermal Healthcell-level BMS
On-site Weather StationPOA + hail detector
—
Predictive-Maintenance Alertslast 7 days
Equipment Pedigree · Reinsurance Audit TemplateiEquipment PedigreeModule + inverter + racking + BESS spec template covering Tier-1 status, ASTM E1038 hail class, glass thickness, frame, MTBF, certifications, stow capability, warranty backing, and BESS chemistry/BMS. Modeled on reinsurance factory-audit conventions (Munich Re, Swiss Re, HSB), not a specific published rubric.EPC bill-of-materials · manufacturer spec sheets · Earthflow Detect satellite as-built verificationCounterparty + warranty assessment · drives Stage 2 vulnerability mitigation factorsequipment_pedigree.py · Tier-1 + ASTM E1038
Climate Scenario Stress Test · 30-yr HorizoniClimate Scenario Stress TestRCP 4.5 and RCP 8.5 horizons across 7 time periods (today → 2055) for the composite and the most climate-sensitive perils (hail frequency, hurricane intensity, wildfire season-length, sea-level rise).CMIP6 ensemble · NEX-GDDP downscaling · LOCA2 (CONUS)Long-dated treaty pricing · climate-uplift reserve · TCFD / ISSB-alignedRCP 4.5 / 8.5 · TCFD · ISSB · CMIP6
.
Today
2030
2035
2040
2045
2050
2055
Treaty Fit · Portfolio AggregationiTreaty Fit MetricsFour-metric scorecard against the cedent's existing book — aggregate concentration, top-5 correlation, diversification benefit, and layer-attachment band fit. Computed via portfolio Monte Carlo with peril correlations preserved (not naive AAL summation).Cedent bordereau · correlation-preserving portfolio aggregationBordereau aggregation · concentration / capacity allocation decisionscedent book · 4-metric scorecard
Methodology · Data Sources · Audit Trail
Hazard Data
NEXRAD MESH · radar-derived hail kinetic energy (NOAA)
NLDN · National Lightning Detection Network (Vaisala)
A single coordinate goes in. A defensible per-peril decomposition comes out. Below: the 11 named public datasets we ingest, the 13 published engineering equations we run, and the 10 outputs we ship in the underwriter risk file.
🛰Fused Earth Data
70+ authoritative datasets ingested in parallel. A single upstream outage never blocks the other modules.
70+ datasets · CONUS coverage
⚛Physics AI
19 parallel modules implementing published engineering equations. AI extends physics but does not decide.
19 modules · 0% opaque weights in headlines
📄Underwriter File
1,180 fields per site in ~49 seconds. Per-peril loss curves at $/MW/yr granularity, every line item with data-source provenance.
Exceedance Probability curve10/100/250/1000 yr marks
Parametric trigger pricingHail KE · sustained wind · GHI
Binding subjectivities (5)Mapped to engineered credits
Climate stress 2030–2055RCP 4.5/8.5 letter grades
Treaty fit metricsConcentration · correlation · layer
~49 seconds end-to-end1,180 fields per site
Physics is system of record. AI extends physics (downscaling, satellite feature extraction) but does not decide. Every headline number traces to a published equation + named public dataset — auditable by HSB's underwriting team, defensible in treaty negotiation.
Why Earthflow Fuses Asset Data + Earth Data + Physics AI
A solar + BESS site doesn't price itself. It's priced by the perils that surround it. Underwriting risk only emerges when you fuse what the broker knows about the asset with what authoritative public datasets know about the ground, sky, and physics around it.
Looking at a solar + BESS site in isolation — its TIV, its module pedigree, its inverter MTBF, its BESS chemistry — tells you one part of the story. But the site doesn't fail because of itself. It fails because of its environment and its physics. A hail corridor 50 mi wide. A peat bog two parcels away. A 1-in-100 wind gust on an under-spec'd racking class. A seismic Vs30 the as-built design didn't account for. A wildfire WUI 2.6 mi south. None of that is in the broker's submission packet. It lives in NOAA SPC catalogs, USGS Vs30 grids, NASA lightning flash density, FEMA NFHL Layer 28, SSURGO Histosol classification, NREL Wind Toolkit, and published engineering standards (ASCE 7-22, FEMA HAZUS-MH, FM Global PRG 18-1, IEC 62305-2). Our thesis is simple: asset data alone is not enough to underwrite a renewable. You have to fuse it with Earth observation — at the asset level — and run it through physics-grounded models.
🔆SITE / ASSET DATA
TIV / capacity$M, MWdc, GWh storage
Module pedigreeMake/model, ASTM E1038 hail class, glass mm
Subjectivities5 binding, mapped to engineered credits
Concrete example. Without environmental fusion, Fighting Jays Solar (Fort Bend County TX, 350 MWdc) was “a solar site in Texas.” With it — fusing NOAA SPC's 764 hail events in 30 yr within a 50 mi catchment with Heymsfield-Wright terminal-velocity physics — it was “a 4.5″ hailstone, 802 J kinetic-energy site with 96.8% module-break probability and $104k/MW Expected Annual Loss.” Earthflow flagged it Very High pre-event. March 2024 took it down for ~$50M insured. The data was there. The fusion is what makes it underwritable.
70+ authoritative datasets · 19 Physics AI™ modules · 1,180 fields per site in ~49 seconds — fused at the asset level, for every renewable account on the broker's submission desk.
🛰EARTH DATA · 70+ AUTHORITATIVE SOURCES
Every Earthflow peril output traces back to a named public dataset. No opaque ML weights drive headline numbers. Below is a sampling of the production stack — ingestion is parallel, and a single upstream outage never blocks the other modules.
NOAA SPC
FEEDSHail · Tornado · SCS climatology
COVERAGE1955–2025 · 50-mi catchment
USGS NSHM 2023
FEEDSSeismic Sa(T) · Vs30 site class
STANDARDASCE 7-22 §11 NSHM 2023
NASA LIS/OTD
FEEDSLightning flash density
STANDARDIEC 62305-2 collection-area
NREL WTK-LED
FEEDSWind resource · V₇₀₀yr
COVERAGE2007–2023 · 2 km grid
FEMA NFHL L28
FEEDSFlood SFHA · floodplain depth
FALLBACKHAND · JRC Surface Water
USDA SSURGO
FEEDSHistosol · K-factor · OM%
RESOLUTION30 m · all CONUS counties
GridMET
FEEDSERC · KBDI · drought stress
CADENCEDaily · 4 km grid
Sentinel-1 SAR
FEEDSSubsurface moisture · peat
RESOLUTION10 m · 6–12 day revisit
PRISM + NARR
FEEDSPrecipitation · IDF design
STANDARDNOAA Atlas 14
NWCG · NIFC
FEEDSFire regime · perimeter history
STANDARDNWCG PMS 425-1 regions
MODIS / VIIRS
FEEDSFire hot-spot history · NDVI
CADENCEDaily · 250–375 m
USFWS IPaC
FEEDSPermitting · listed species
COVERAGECONUS + 20-state RE framework
⚛12 PERILS · DATASET → PHYSICS → OUTPUT
Each peril fuses one or more datasets through a published engineering standard or peer-reviewed equation. The output isn't a class label — it's a $/MW/yr expected annual loss your treaty desk can defend.
🧊 Hail / SCS
NOAA SPC · 30-yr · 50-mi catchment
Heymsfield-Wright · FM Global PRG 18-1
🌪 Tornado
NOAA SPC + ASCE 7-22 §32 · 3000-yr MRI
ASCE 7-22 §32 EF-scale
💨 Wind / Hurricane
NREL WTK-LED + ERA5 V₇₀₀yr
ASCE 7-22 §27 design pressure
🌍 Seismic
USGS Vs30 + NSHM 2023
HAZUS-MH 2.1 DS2 fragility
⚡ Lightning
NASA LIS/OTD flash density
IEC 62305-2 collection-area
🌊 Flood
FEMA NFHL L28 · JRC SW · HAND
FEMA SFHA + depth proxy
🔥 Surface Wildfire
LANDFIRE FBFM40 + GridMET ERC/KBDI
Rothermel spread + Keetch-Byram
🟫 Peat / Subsurface Fire
SSURGO Histosol + Sentinel-1 SAR + KBDI
Rein 2009 · Huang-Rein 2017 smolder
🌧 Erosion
SSURGO K-factor + PRISM + topographic slope
RUSLE (USDA NRCS)
🔧 Equipment Breakdown
SCADA fault log · MTBF actuarial · CMMS
Cox proportional-hazards survival
🔋 BESS Thermal Runaway
BMS · cell-level temp · NMC/LFP chemistry
NFPA 855 · UL 9540A
☀ Module / PV Degradation
EL imaging · I-V trending · drone IR
Kimber soiling · Faiman temperature
Worked example · Peat ignition score, end-to-end.
Peatville MI (43.620°N, -82.830°W). SSURGO returns Fibric Peat, 75% organic matter, ≥2.0 m depth. Sentinel-1 C-band SAR returns 32.1% subsurface moisture. GridMET ERC returns 65. Push those through the Rein 2009 smoldering equation (spread rate = 0.5 + 4.5 × deficit1.5 cm/hr) and the Huang-Rein 2017 burn-depth function: Peat Score 75.6, surface score 43.4, compound bonus ×1.10, Combined 71.8 (HIGH). Output to the underwriter file: 7-day smoldering burn at 0.77 m depth, +23% surcharge, detection difficulty Very Hard. All 11 numbers reproduce from peat_fire_risk.py. A reinsurer can defend a +23% surcharge in a treaty negotiation by saying: "we fused SSURGO, Sentinel-1 SAR, and Rein thermodynamics. Here's the waterfall."
EARTHFLOW UNDERWRITER · DEMO BUILD · Figures are illustrative · Data shapes mirror production Earthflow pipeline output · Not for production binding decisions