The fleet procurement manager reads “4,597 ultra-fast stations confirmed operational” and calculates accordingly. Then the first batch of vehicles deploys. The real charging time is not 9 minutes. The TCO model is broken before the fleet turns a wheel. Four specific miscalculations cause this. All four are avoidable with audit-grade data handling.
Mistake 1: Using Peak Power as the Operational Figure for Non-OEM Vehicles
The 9-minute claim is real. It is independently verified {VER — electrive.com on-site, March 5, 2026}: Ti3 measured 8 min 45 sec; Da Tang measured 9 min 24 sec. Both on BYD proprietary 1,500 kW Flash Chargers. Both BYD vehicles.
The audit finding that breaks the TCO model: no operator has published what power a non-OEM vehicle actually receives at a cross-brand charger. BYD confirmed physical access is open for all GB-standard EVs {CLM — Dongchedi, March 9, 2026}. BYD did not publish a power tier policy stating that non-OEM vehicles receive identical power allocation. Those are two different statements.
Until T2 data (Q2 2026 — first independent real-world delivery test) resolves this gap, the audit-grade planning estimate for non-OEM vehicles on a BYD Flash Charger is: 12 minutes (moderate scenario: 75% power) to 18 minutes (conservative: 60% power) {INF — scenario model}. A TCO model that uses 9 minutes for non-OEM vehicles is using a number that does not exist in the evidence base.
Mistake 2: Counting Geely’s 2,103 Stations as 2,103 Ultra-Fast Stations
Geely announced 2,103 stations covering 215 cities as of February 28, 2026 {CLM — Dongchedi, March 9, 2026}. Of those, 1,216 are ultra-fast. The remaining 887 are standard charging stations. The number fleet planners should use when calculating Geely ultra-fast coverage is 1,216 — not 2,103.
The per-city average of 9.8 stations across 215 cities {INF — calculated from CLM data} almost certainly masks concentration: tier-1 cities with 30–50 stations and tier-4 cities with 1–2. The city-level distribution has not been published {GAP}. A fleet operating in a tier-3 or tier-4 city should verify actual station availability at the specific base location using the network app before including Geely coverage in any procurement decision.
Mistake 3: Treating May Day as a Confirmed Milestone Rather Than a Verifiable Commitment
Wang Chuanfu committed to 1,000 highway Flash Charging stations operational before May 1, 2026 {CLM — Jingji Ribao, March 6, 2026}. This is the most specific, time-bound, independently verifiable infrastructure commitment in the market. Fleet managers are right to monitor it. The mistake is treating it as delivered before the date arrives.
Current build pace: approximately 1,344 stations/month {INF}. Required pace to hit annual 20,000 target: 1,621/month {INF}. Shortfall: 17% {INF}. The city/highway station split within the current 4,597 total is not disclosed {GAP}. Progress toward the May Day milestone cannot be independently tracked before May 1. Fleet procurement decisions for highway routes that rely on May Day coverage confirmation should not lock in before May 1.
BYD has not separately disclosed how many highway stations are included in the 4,597 total. All public reporting presents only the aggregate figure. There is no partial verification path before May 1. The only data source is BYD’s official announcement on that date.
Mistake 4: Ignoring the Highest-ROI Variable in the TCO Model
The cross-brand power degradation is real and material. For a 60-vehicle intercity fleet at 40% off-OEM charging frequency, the estimated annual fragmentation premium is RMB 295,200–346,000 {INF — scenario model based on assumed 75% power and 40% off-OEM frequency}.
Fleet managers focus on the power figure because it feels like the technical variable. The problem is that the power figure is outside their control. The off-OEM charging frequency is not. Routing planning and dispatch policy control which charger vehicles use. Reducing off-OEM frequency from 40% to 10% cuts the annual fragmentation premium by approximately 75%. This is a dispatch software configuration, not an infrastructure investment.
Procurement checklist before any ultra-fast charging fleet deployment:
- Gate 1: Is there an OEM-matched Flash Charging station within 5 km of the primary operating base, with at least 2 operational guns? Verify via app, not OEM press release.
- Gate 2 (highway): Is the route covered by OEM-matched ultra-fast stations at no more than 150 km intervals? Evaluate after May Day 2026 data for BYD highway.
- Gate 3 (cold climate): Does the route include regular below -20°C operation? Directional data: Song Ultra 10 min 21 sec at -20°C in Harbin {CLM — Dongchedi on-site test, March 17, 2026}. Standardized protocol test: Q4 2026.
- Flag all TCO models: cross-brand charging speed estimate is subject to T2 revision (Q2 2026). Do not lock in the figure as a confirmed cost input.
- Measure off-OEM charging frequency in the existing fleet before modelling new vehicles. The actual number may differ significantly from the 40% baseline assumption.
China Ultra-Fast Charging Infrastructure Fragmentation Audit
Full audit: CP-01 through CP-04 with data tables, build rate calculations, fleet TCO model, and 8 monitoring triggers. 30+ pages, data cutoff March 28, 2026.