Maxwell v12 → Catalog Mapping¶
Phase 4, step 4.4 (draft for review). Maps the partner's 全球电池运营财务模型v12.xlsx onto our BU catalog and transaction map — line-by-line provenance, adopt / contrast / reject verdicts, per ADR 0002 posture (adopt terminology and structure; distrust incumbent cost bases).
Classification: competitive analysis (modeling purpose 3) — market intelligence recast into framework-aligned language, serving as input data and validation oracle. See Documentation Organization Proposal.
What v12 actually is¶
v12 models a vertically-integrated swap operator: one entity owns the battery fleet AND runs the swap service. Revenue arrives directly from riders; electricity and battery management are its costs. In catalog terms, v12 is BCS + SNS merged into one BU — the H1-minus-H2 internal transfer is invisible. This is the common Chinese operator structure (NIO-style). It is not wrong — but it cannot answer where profit is made and what ROI is sensitive to per participant, which is our stated purpose. Unbundling it onto our BU catalog is the mapping's core job.
Workbook structure: Input (3 business lines), Region_DB (7 regions), Battery_DB (6 battery types), Battery_Operation_Model (120 months × 3 lines, cohort batch ledger), Battery_Batch_Ledger, Dashboard/Sensitivity_*.
Mapping table¶
| v12 element | Location | Maps to (ours) | Verdict |
|---|---|---|---|
| 业务线 1–3 (3 lines, 3 configs) | Input | separate model configurations | adopt pattern |
| 目标区域 + Region_DB FX | Input / Region_DB | market profile (USD basis) | adopt |
| 电池型号 + Battery_DB (Ah, V, kWh, USD, EFC, SOH) | Battery_DB | structured battery classes in profile | adopt |
| 超配系数 auto-calc + bounds | Input B24–27 | BCS overprovisioning identity | adopt w/ rework |
| 收费模式 按次/按度电 | Input | H1 composite tariff (subset) | adopt w/ rework |
| 单次换电补电比例 0.8 | Input | kWh/swap driver | adopt |
| 单车日耗电量 12 kWh | Input | daily energy demand (demand chain) | adopt |
| 充电安全系数 / 峰值集中系数 | Input | demand/supply ratio modifiers | contrast |
| 起始/成熟不可用比例 0.1/0.05 | Input | BCS availability/loss assumptions | contrast |
| 电池管理费 7-component buildup | Input B29–44 | BCS opex + exogenous overheads | adopt w/ rework |
| 阶段划分 扩张/稳定/退出 24/24/12 | Input / col AN | expansion/stable/exit scenario phases | adopt |
| 项目转让电池折扣 0.5 | Input | exit residual treatment | contrast (eMe) |
| 批次台账 cohort SOH (矩阵_*, M_SOH) | cols AP–AZ, BA–KF | BCS batch retirement ledger | adopt (core asset) |
| EFC-based batch retirement | Battery_DB + AG | EFC retirement convention | adopt |
| 收入 Z | col Z | H1 revenue — belongs to SNS | unbundle |
| 电费 AA | col AA | H3 energy cost — belongs to SNS | unbundle |
| 管理费 AB | col AB | BCS opex line | unbundle |
| EBITDA AC = Z−AA−AB | col AC | merged P&L — split per catalog | rework |
| 期末残值 (flat % of cost) | Dashboard | eMe second-use function | contrast (deviation register) |
| IRR 月度→年化 | Dashboard | standard inquiry set (RoI) | adopt |
| Dashboard vs Sensitivity IRR mismatch | known defect | single IRR convention (Phase 5) | reject (defect) |
| Sensitivity block stale pasted values | Sensitivity_* | reproducible sensitivity harness (Phase 5) | reject (defect) |
| 所得税 toggle | Input | policy/fiscal parameter group | adopt |
What v12 structurally lacks (our catalog adds)¶
- SNS/BCS separation — no H2 availability fee, no price wedge, no per-participant profit pooling. The integrated operator hides exactly what we expose.
- Station economics — no station/slot/charger CAPEX, no rent, no station labor, no integer capacity, no utilization-vs-capacity constraint. Swap counts are demand-only, never capacity-limited (demand/supply ratio absent).
- Rider economics — no vehicle purchase, no TCO vs ICE, no affordability check. Riders are just a battery count (车上电池数).
- EPS as a counterparty — grid tariff is a bare cost input; no PPA, no anchor-load value.
- Composite tariff — 按次 or 按度电 only; no subscription component.
- Rider modality — no vehicle tenure dimension; VSP modality absent (ADR 0003).
Verdict summary¶
- Adopt: cohort batch ledger, overprovisioning auto-calc with bounds, phase segmentation, region/battery DB pattern, EFC retirement, USD basis (already aligned with our unitary ruling), monthly cadence.
- Contrast: cost bases (battery USD, grid tariff, duty), availability/peak/safety coefficients, flat residual, transfer discount 0.5.
- Reject: merged P&L as final output (keep as a consolidation view — sum of catalog BUs — so results remain comparable with Maxwell), IRR convention mismatch, stale sensitivity block.
The integrated view matters: our unbundled catalog must be able to re-consolidate to v12's merged P&L as a cross-check — consolidation = eliminating H2 transfers. If our separated model, consolidated, disagrees with v12's integrated math on identical inputs, the discrepancy localizes a coupling bug. That makes v12 a test oracle, which is its highest use.
Next step (4.5): assemble Phase 4 — consolidate methodology pages, run KRR, close the phase.