Methodology & Limitations

Telangana EV Charging Opportunity Index · research beta · built 1 August 2026

Read this first. Every score is a modeled estimate from public data, and the index is a research beta that has not been validated against real charging performance. Nothing here is a measurement of charging demand. The model ranks relative opportunity so you know where to look — it does not tell you a site will work. That requires a survey, a DISCOM load sanction, and verified traffic counts.

1. What is computed

Telangana is divided into 21,813 hexagonal cells (H3 resolution 8, about 0.74 km², inside Greater Hyderabad; resolution 7, about 5.16 km², elsewhere), covering 112,291 km². For each cell and each segment:

OpportunityIndex = normalise( Demand − EffectiveSupply ), scaled 0–100.

Both terms are put on a common footing before subtraction: each raw layer, and the supply total, is converted to a z-score, clipped at ±3σ, and rescaled to 0–100. Subtraction therefore happens between two comparable 0–100 quantities, and the result is min-max rescaled to 0–100.

Demand is a weighted sum of normalised layers (z-score, clipped at ±3σ, rescaled 0–100). Effective supply is the distance-decayed sum of existing charger power within a segment-specific radius — 3 km urban and 25 km on corridors for cars, 30 km for trucks.

2. Weights

Car charging	%	Truck charging	%
Dwell points of interest	22	Freight corridor & truck traffic	30
Residential density	18	Industrial / logistics density	22
EV 4-wheeler registrations	18	Warehouse proximity	12
Arterial traffic	12	Highway junctions	10
Offices	10	Truck-halt ecosystem	10
Affluence (not available)	8	Mining / heavy industry	8
Parking & fuel retail	6	HT power feasibility	8
Power feasibility	6

A layer we could not obtain is never scored as zero — zero would mean "measured, none present". Its weight is removed and the remainder renormalised. The car model therefore runs on 92 of 100 points; trucks on 100 of 100.

3. Sources

Layer	Source	Accessed	Licence
Existing stations	Public CPO locators and official publications, aggregated	2026-08-01	public sources
Station power / connectors	Open Charge Map	2026-08-01	CC BY 4.0
Charging consumption & sanctioned load	TGSPDCL via Telangana Open Data Portal	2026-08-01	Public Domain
EV registrations	VAHAN, Ministry of Road Transport & Highways	2026-08-01	Government of India
Roads, land use, POIs, boundaries	OpenStreetMap via Overpass & Nominatim	2026-08-01	ODbL
State policy	Telangana EV & Energy Storage Policy 2020–2030; G.O. Ms. No. 41, 16 Nov 2024	2026-08-01	Govt of Telangana
Central guidelines	Ministry of Power / BEE EV charging infrastructure guidelines, 17 Sep 2024	2026-08-01	Government of India
Basemap	CARTO, OpenStreetMap	live	ODbL / CARTO

4. Station sizing

Gun counts come from an M/M/c queueing model: the smallest number of guns keeping the 90th percentile wait under 10 minutes at peak. Peak hour is taken as 18% of daily sessions.

Sizing is calibrated to measured throughput, not to industry rules of thumb. Hyderabad's metered data for August 2024 shows 449,414 kWh across 261 billed charging connections — about 57 kWh per connection per day, roughly 3 car sessions. The busiest locality observed (Gachibowli) runs about 144 kWh/day per connection, roughly 7 sessions. Session bands are anchored to that observed range, then grown to 2030. Common industry assumptions of 15–30% utilisation are roughly double what Hyderabad actually shows today, and sizing to them would over-build guns and overstate payback.

5. Validation — and what it did not show

Modeled demand was tested against measured electricity consumption for 85 Hyderabad localities.

Test	Spearman ρ	p
Modeled demand vs measured kWh	+0.044	0.69
Modeled hotness vs measured kWh	+0.071	0.52
Modeled demand vs sanctioned grid load	−0.298	0.006

The rank correlation is not statistically significant. This model is not validated against measured consumption. Mean consumption does rise across demand quartiles (1,066 → 1,495 → 1,560 → 2,479 kWh), which is directionally encouraging, but the medians are noisy and the effect is carried by a minority of high-throughput sites.

One result is significant, and it points the other way: cells with higher modeled demand tend to have lower sanctioned grid capacity. Read charitably, the model is finding genuinely under-served areas. Read cautiously, high-scoring sites may sit where grid capacity is scarce — meaning higher connection costs and longer utility lead times. Either way, treat the power position of any shortlisted site as an open question.

6. Gap register

Gap	Status	Effect
Charger power (kW)	Stated for about 2% of stations	Supply is imputed from operator priors — the single largest weakness. assumed
Gun / port counts	Almost never published	Imputed. assumed
Operator coverage	5 of 20+ active operators	Supply understated where a missing operator already builds.
Private & captive charging	Invisible in all public data	Cannot be closed. Real supply is higher than shown, most of all in dense urban areas.
Affluence layer	Not acquired	8 points removed from the car model.
Some station records	Published 31 Mar 2022	Counted at half weight; current status unverified.
Urban resolution	Fine detail for Greater Hyderabad only	Warangal, Nizamabad, Karimnagar and others are scored on ~5.2 km² cells.
Bus charging	Withheld	Depot locations can only be inferred and the electrification-allocation layer is unavailable, so bus scores are not published as siting guidance.

7. Stability — and one place it fails badly

Every weight was re-run at ±20%. The top-25 lists hold 97.0% (cars) and 96.3% (trucks) of their members on average. Rankings are robust to the weight choices — though that measures stability, not correctness.

Weights are not the real uncertainty here. Charger power is imputed for 97.9% of stations, so we re-ran the model with every imputed kW scaled ±50%, leaving measured values untouched:

Segment	Top-25 retained at kW −50%	at kW +50%
Cars	88%	100%
Trucks	24%	100%

The truck ranking is substantially an artefact of assumed charger power. The DC ≥ 50 kW filter that defines truck-relevant supply passes 54 stations at our assumed values but only 8 if those assumptions are 50% too high — so 76% of the truck shortlist changes. Truck results are therefore graded D and must be treated as a rough geographic prior, not a ranking. Car results are materially more robust (88–100% retained) but still rest on the same imputed supply data.

7b. Resolution normalisation

The grid mixes H3 resolution 7 (~5.3 km²) and resolution 8 (~0.76 km²). All demand layers except one are fixed-radius measures — "how much of X lies within N km of this point" — which are independent of cell size. The exception was EV registrations, an absolute per-cell count, where a 7× larger cell would carry 7× the count for no real reason. That layer is now scored as density per km². Without this correction, cell size leaked into the ranking.

7c. Regulatory documents tracked

Beyond the September 2024 guidelines above, the refresh pipeline tracks:

Document	Status
PM E-DRIVE public charging station operational guidelines	Verified live, 1 Aug 2026
TGERC tariff schedule FY 2026–27	Verified live, 1 Aug 2026
Ministry of Power amendment to the charging-infrastructure guidelines (Jan 2025)	Not independently retrieved — the host returned HTTP 403. Referenced, not yet read, and therefore not relied on here.

Regulatory coverage requirements answer a different question from commercial opportunity and are deliberately not merged into this index.

8. Reproducibility

The browser calculation and the Python reference implementation are tested against each other on 100 random points and agree exactly. Point queries run entirely in your browser against a precomputed index; no query data is transmitted.