Methodology & data sources
Every rule, rate, and reference the engine applies. Each line cites the source and the effective date.
PG&E BEV2-S commercial DCFC tariff
Pacific Gas & Electric, BEV2-S schedule — CPUC Advice Letter filing
Effective: Filed 2024-12, current revision applies through 2026 review
Energy charge by TOU period + monthly subscription block. Demand charges waived under BEV2-S.
California LCFS credit price
CARB LCFS Data Dashboard — public posting
Effective: Snapshot 2025-Q4 ($59.34/credit average)
Used as default Y1 LCFS price. User can override.
IRS bonus depreciation
IRC §168(k); IRS Pub 946; TCJA phase-down schedule
Effective: Y1 = 60% in 2024, 40% in 2025, 20% in 2026, 0% from 2027
User-adjustable in wizard Step 4.
Federal 30C ITC (alternative fuel vehicle refueling property)
IRC §30C; Inflation Reduction Act §13404
Effective: Up to 30% with prevailing-wage uplift; sunset 2026-06-30 unless extended
Santa Clara County Business Personal Property (BPP) tax
Santa Clara County Assessor's Office BPP rate schedule
Effective: FY 2024–25, 1.0% of assessed value; 7-yr declining schedule
Income tax on operating profit — not modeled
Model design decision; see report Section 4 tax note
Effective: All cash flows are pre-income-tax. The engine values depreciation + the 30C ITC as tax-shield inflows at the user's marginal rate; it computes no federal or CA income tax on the project's profit. LLC franchise tax + gross-receipts fee are modeled as operating costs.
Address geocoding
Mapbox Geocoding API
Effective: Live lookup, cached per session
Used only to resolve county / utility territory / nearest supercharger distance.
Updates to underlying tariffs, LCFS prices, or IRS rules are reflected in the engine within two weeks of publication.
Location-derived utilization estimate
The wizard's “Location-derived” site-type option computes a starting utilization ramp from public location data: the distance, stall count, and site type of the nearest existing Superchargers (DOE Alternative Fuels Data Center); DC-fast charging stalls of every network within 10 miles, weighted by distance (same source); and county battery-electric vehicle registrations per 1,000 residents (CA DMV fuel-type dataset joined to state population estimates). Each factor maps to a score through a published breakpoint table, and the combined score selects a weighted blend of the three preset demand curves — so the estimate always falls inside the range the presets already span.
Every factor, score, weight, and data vintage behind an estimate is shown with it in the wizard. The estimate is directional: it is computed from public data about the area, not measured at the site, and it has not been validated against observed station throughput. It pre-fills the same editable fields the presets fill; every monthly value and growth rate remains yours to change. The P10–P90 range shown with an estimate is carried onto the report's Year-1 utilization sweep as shaded rows — the sweep itself is computed by the engine at every level regardless.
Coverage ratios
Alongside the headline metrics (payback period, 15-yr IRR, NPV, cumulative cash), the engine computes lender-grade coverage ratios for each year: FCCR (Fixed Charge Coverage Ratio — covers rent, insurance, taxes, AND loan service) and DSCR (Debt Service Coverage Ratio — loan service only, undefined for all-cash operators). Engine math is identical for every user.
For reference, SBA-504 / small commercial real estate conventions treat FCCR ≥ 1.25 and DSCR ≥ 1.20 as comfortable, 1.10–1.25 / 1.10–1.20 as marginal, and below 1.10 as distressed. These are widely cited banking reference bands, not thresholds the model imposes.
The coverage metrics (DSCR, FCCR) and financial figures (NPV, IRR, payback, cumulative cash) are descriptive model outputs. How they weigh against your own risk tolerance is your call.