Parameter Taxonomy¶

How parameters are organized and governed. Schema lives in models/parameters/schema.py; market profiles in models/parameters/profiles/.

Five groups¶

Group	Contents	Role
Environment	FX, interest, fuel, electricity prices	market context, not controllable
Policy / fiscal	import duty, VAT, income tax	decision-relevant; changes by year
Operating	CAPEX, prices, labor fees, scale, horizon	what the operator controls
Assumptions	utilization, degradation, failure rates, overprovisioning	behavior hypotheses to stress-test
Intermediary	computed variables	checking + comparison basis; never inputs

Every parameter carries a provenance tag:
maxwell-v12 — extracted from Maxwell's v12 (reference artifact)
industry-verify — published baseline; contrast case per ADR 0002
oves-knowledge — OVES first-hand knowledge; base case when documented
pending-input — awaiting owner decision
One market profile = one market × modality (e.g. Kenya OVRB base case), a versioned YAML file
pending-input (value null) is legal; the profile reports what is missing instead of guessing
Intermediary variables are computed by the engine — they appear in profiles only as comparison targets
Deviation-register parameters (e.g. residual value) carry the deviation note inline; sensitivity reports both values

First profile: profiles/kenya_ovrb_base.yaml. Highlights:

Extracted from Maxwell's v12 Line 1 (Kenya Nairobi) with provenance tags
Known ambiguity recorded, not resolved silently — e.g. v12 unit battery 3.4 kWh vs 72V100Ah spec
Pending owner inputs: vehicle procurement price, normalized swap price ($/kWh), target WACC
Deviation register entry inline: residual value 32% (v12) vs 0% (industry) vs eMe function (pending)

Computed by the engine, compared across BUs and markets: