The WHAT — grid hosting-capacity visibility

Where can the grid actually absorb rooftop? Red = opaque (no published data). That red is the bet.

Metric

––

No seed data yet

—

solar output firmed (storage-shifted)

The data story: hosting capacity is checked by every DISCOM for every application, but published by ~one state (Telangana) — DT/feeder capacity, quarterly, +50% DT cap. ~19 states only auto-approve ≤10 kW; the rest are case-by-case & opaque. The data exists; the visibility doesn't. Rooftop-installed GW are real (Sept 2025); CUF/firming illustrative.

What we're building on the ground

Real things, in real places, each with a number we update monthly. No abstractions: a plot, a dataset, a partnership. Each card shows what we measure, where it stands, and what it moves.

Don't fund what's already crowded: rooftop potential mapping (a national explorer already covers 120 cities) · residential rooftop + subsidy (PM Surya Ghar — 20.8 GW, +122% YoY) · MSME rooftop finance (well-served by large climate-finance players) · utility storage/firming (VGF live; FDRE ₹4.82–5.06/kWh). Discipline first: spend only on the white space.

1 · AgroPV demonstration plots — BET B (FUND)

Active

Where: 3 sites — marginal land · Partners: agrivoltaics research institutes + a developer + a farmer collective · Instrument: Grant (demonstration + MRV)

Plots commissioned1 / 3

Farmer income measured₹4k / ₹40k /acre/yr

Crop yield vs control plotmeasuring / ≥85%

Firm energy meteredlive MWh/yr

✓Site MoUs (3/3) ✓Plot 1 panels up ○First harvest measured ○Evidence pack published

↳ Moves: turns "nobody believes dual-use economics" into audited ₹ on a farmer's bank statement.

2 · MSME rooftop finance — lower priority

Crowded

Status: already well-served by large climate-finance players (blended finance, first-loss, credit lines) · Our role: monitor; engage only on a specific unserved sub-gap

Why lower priority: MSME rooftop financing is filling fast with larger, better-suited capital — blended finance, first-loss guarantees, dedicated credit lines, and bank–NBFC tie-ups. Adding money here would not be additional.

↳ Action: track; re-enter only if a specific sub-gap (e.g. micro-MSME <10 kW, informal credit) is left unserved.

3 · Feeder hosting-capacity layer — BET A (FUND)

Priority

Where: extend the existing national rooftop explorer + 1–2 pilot DISCOMs · Instrument: Grant + convening

The existing national explorer maps rooftop potential for 120 cities — it does NOT map feeder/DT hosting capacity. That missing layer is the bet.

Feeders mapped0 / 500

Hosting headroom on this mapno → live

Rooftop approval time90d → 30d

○Data-share MoU (the blocker) ○Feeder data ingested ○Headroom layer live

↳ Moves: gives the discom a concrete reason to say "yes" to rooftop — the #1 bottleneck.

4 · Convening — the energy table — FUND

Priority

Role: a neutral energy convener — the one canvas job with no owner · Brings together: research partners + DISCOMs + developers + funders around the firming-cost trade · Instrument: Convening

Convening table established0 / 1

Actors at the table0 / research + 2 DISCOMs + 2 funders

Joint hosting-data MoU signedno → yes

○First convening held ○Data-share agreed ○Firming-cost allocation drafted

↳ Moves: the neutral table where the firming-cost trade gets struck — unblocks Bets A & B.

Tracker values are illustrative placeholders for the format — to be updated monthly against real field data. The point is that every line is a number that moves, not a position.

Are we moving? — markers we track

Movement, not intent. One north-star outcome, then three horizons — each a shift with markers that are numbers carrying a baseline and a target. If a bar isn't moving, the strategy isn't.

◎ North star — what we're trying to achieve

Distributed solar (RTPV + AgroPV) delivering firm, dispatchable energy — measured in CUF-adjusted TWh and ₹/kWh-to-firm — with the DISCOM made whole and the farmer better off. Capacity stops being the scoreboard; firmed energy and a fair split of the firming cost become it.

Firm distributed-RE energy tracked~0 → calibrate TWh/yr

Firm RE tariff (FDRE/RTC, sourced)₹4.82–5.06/kWh → falling

Low-hanging · 0–2 yrs · prove the unitActive

FROM counting nameplate GW → TO firm-energy + cost-to-firm accounting · PM Surya Ghar · PM-KUSUM

Firmed units live (audited CUF + ₹/kWh)1 / 3

Replicable templates published0 / 2

₹/kWh-to-firm proven on a real unitno → yes

Middle · 2–5 yrs · share the firming costPlanned

FROM DISCOM eats firming → TO cost shared via VGF / firm-PPA / green premium · VGF-BESS · Green Open Access · RDSS

Storage co-located (pilot)0 / 2 GWh

First-loss crowding ratio0 → 1:4

DISCOMs with hosting data live0 / 3

Long · 5–15 yrs · firm RE becomes defaultPlanned

FROM GW-target planning → TO firm-energy (TWh) + round-the-clock planning · 500 GW non-fossil 2030 · Net-Zero 2070

Cost-to-firm reduction0% → >40%

AgroPV dual-use on marginal landpilots → GW-scale

Firm distributed-RE share of new RE— → tracked

Measurement logic: outputs (units, GWh, ₹) are leading indicators we control; the outcome (firm TWh, ₹/kWh-to-firm, discom & farmer made whole) is the lagging indicator we're actually after. Targets are illustrative until calibrated to field baselines.

Actors on the canvas → where catalytic money goes

Map every job in the value chain to the actor who owns it. Where a willing commercial actor already does the job, the foundation stays out. Where the job is unowned, or an actor exists but the risk/return doesn't yet work, that white space is exactly where catalytic funding belongs — and the gap type tells you which instrument to use. Partner names are starting points to validate.

The actor canvas

Job in the chain	Actors / partners	Responsibility	Covered?	Catalytic entry
Policy · targets · subsidy	MNRE, SECI, BEE, state agencies	Set targets, run schemes	Govt	None — align, don't fund
Rooftop potential mapping	CSTEP (RTSE, 120 cities)	Map roof potential + savings	Crowded	None — extend to hosting (Bet A)
RTPV — premium / C&I	Tata Power Solar, Fourth Partner, Amplus, SolarSquare	Install & operate rooftop	Market	None — market works
RTPV — MSME / low-income	large climate-finance players + RESCOs	Serve thin-credit roofs	Filling	Lower priority
AgroPV — dual-use	Khethworks, Oorja, NISE/ICAR, FPOs	Prove dual-use farming + energy	No proof	Grant — demo + evidence
Evidence / MRV / modelling	CSTEP, CEEW, Prayas, WRI, TERI	Measure CUF, yield, income; model the system	Partial	Grant — independent MRV
Grid hosting data	DISCOMs, SLDCs, CEA + CSTEP (modelling)	Publish / model feeder headroom	Unowned	Data layer + convening
Storage / firming	IESA, developers, CEA, CSTEP (modelling)	Build & dispatch BESS; size firming	Partial	First-loss / VGF design
Finance / debt	IREDA, PFC/REC, SIDBI, cKers, Ecofy, NABARD	Lend to projects	Risk-averse	First-loss / guarantee
Aggregation	FPOs, RWAs, cluster bodies	Bundle small demand	Weak	Grant + convening
Offtake / DISCOM	DISCOMs	Buy / net-meter, bear firming	Stressed	De-risk via data + first-loss
Community / just transition	SELCO, FPOs, CSOs	Protect farmers & land	Partial	Grant + safeguards
Coordination	— (no neutral actor)	Convene the firming-cost trade	Unowned	Convening (foundation IS this)

How the canvas decides the funding instrument

Coverage on the canvas	What it means	Catalytic instrument
Market / Govt	A willing actor already does it	Nothing — stay out, just align
Unowned / no proof	Public good or unproven economics	Grant — evidence, demo, data, convening
Partial / risk-averse	Actor exists but won't move on the risk	First-loss / guarantee / concessional debt
Stressed bearer	One actor silently carries the cost	De-risk + redistribute (data + first-loss + VGF)

This is the rule: catalytic funding flows to the coloured gaps, never the green ones — and the gap's colour picks the instrument. Fund the white space, exit when the market arrives.

Baseline → goal (what we move)

Rooftop installed (national)20.8 → 30 GW resid (FY27)

PM Surya Ghar households~3M → 10M (1 cr)

AT&C losses16.3% → 12–15% (RDSS)

Firm RE tariff (FDRE/RTC)₹4.82–5.06/kWh → falling

Agrivoltaics pilots<25 → bankable model

DISCOMs with hosting data live0 → 2

Baselines sourced mid-2026 (pv-magazine, Mercom, IEEFA, PFC, SolarQuarter, IISD). AT&C trend contested (RDSS: falling; PFC: rising). Solar CUF ~19% is a domain estimate. ₹ asks indicative.