Streets
How much rooftop solar can one street take?
Blue dot = P5. White diamond = P50 (median). Pale bar = P5 to P95 over the trials (generic model only; P95 not held for the Indian model). Amber or light bar above it = 90% bootstrap range of P5 (Bootstrap 90% range of P5 from the Indian results themselves (results/indian_summary.csv p5_lo / p5_hi, 2,000 resamples of the 200 trials), held at every 11 kV voltage.) Dashed red line = 50% cap, unit assumed: kW on the DTR kVA rating; TGSPDCL to confirm.
Over limit before solar: Over the voltage limit before any solar: fix the tap/voltage first. The street is already above +6% with no solar (for example a 433 V DTR with the 11 kV side at 1.03 pu), so the result is not a solar limit and is never shown as 0 kWp safe.
P5 (screening value): the 5th percentile of the random trials (200 Indian model, 30 generic) of the largest rooftop solar that passes the screening tests at the sunniest hour. Not an approved capacity.
kWac is derived as kWp / 1.2 (the DC to AC ratio of reading R3) and is an estimate. Solar columns are panel kWp; the cap column is kW with the unit assumed; where the unit is not known it says so.
- Phase-managed (fix): each new roof connected to the next phase in turn (a proposed fix, not a rule).
- Phase unknown (random phase per home: assumed split, not measured): each home on a random phase, redrawn in each of the 200 trials (hosting/run_indian.py); no real phase data is used.
Phase-managed: a recommended phase per new roof, subject to discom feasibility. Not a rule: no Telangana clause for phase management or an export cap was found in the 2025 texts (a negative search, not proof).
Limits of these numbers. P5 is a screening value and can be optimistic. Checks run on the generic 400 V model (not repeated on the Indian model): in 91 of 540 roof-by-roof trials "more roofs is never better" fails, and OpenDSS gives 69.0 vs 73.35 kWp (urban) and 16.35 vs 18.0 kWp (rural), both unbalanced. Treat it as a range, not a guarantee. Tested: voltage rise to +6% at the sunniest hour, unbalance 3%, transformer and cable loading 100%, midday use 0.25 or 0.5 kW per home. Not tested: lower voltage limit (-6%) and evening peak.
Counts come from the stored 200 trials per case: the test that broke one roof beyond each trial's largest passing step. Pass or fail is not stored per test; nothing is invented for tests marked not tested. A transformer or cable can overload even if voltage passes. A 1.03 pu tap already uses part of the +6% margin; the Telangana LV limit is +6%/-6% (TSERC SoP Reg 5 of 2016, checked-correct).
- Reverse flow: not tested as a pass/fail limit (no limit is applied; the feeder view only shows the flow).
- Lower voltage limit (-6%) and the evening peak: not tested on the street model.
- Midday home use is assumed (0.25 or 0.5 kW per home), not measured. The standard screen uses minimum daytime load with maximum PV at the sunniest hour (checked-correct, literature).
- Telangana LV limit is +6%/-6% (TSERC SoP Reg 5 of 2016, checked-correct). Only +6% is tested; a 1.03 pu tap already uses part of that margin.
| Phase model | P5 kWp (panel) | P50 kWp (panel) | P95 kWp (panel) | P5 as kWac (derived, estimate) | Deciding limit | 50% cap, kW (unit assumed) | P5 / cap | P5 range (bootstrap) | Against the cap |
|---|---|---|---|---|---|---|---|---|---|
| Phase unknown (random phase per home: assumed split, not measured) | 8.9 | 18.0 | not held | 7.4 | voltage | 50.0 | 0.18 | 6.0 to 9.0 | below the cap |
| Phase-managed (fix) | 18.0 | 30.0 | not held | 15.0 | voltage | 50.0 | 0.36 | 18.0 to 18.0 | below the cap |
| Street type | 50% cap, kW | Phase unknown (random phase per home: assumed split, not measured) P5 | P50 | Phase-managed (fix) P5 | P50 | Three-phase solar (balanced connection) P5 | P50 | Deciding limit (phase unknown, else the case held) |
|---|---|---|---|---|---|
| Village street, 63 kVA, 700 m, ACSR Weasel | 31.5 | 0.0 | 6.0 | 0.0 | 9.0 | not modelled | voltage |
| Town street, 100 kVA, 400 m, ACSR Rabbit | 50.0 | 8.9 | 18.0 | 18.0 | 30.0 | not modelled | voltage |
| City street, 160 kVA, 300 m, ABC cable | 80.0 | 24.0 | 48.0 | 47.9 | 63.0 | not modelled | voltage |
| Apartment block, 315 kVA, 60 m, ABC cable | 157.5 | not modelled | not modelled | 288.0 | 291.0 | voltage |
The 50% cap can be tight or loose depending on the street: on the long overhead streets P5 can sit below the cap, while on a short street with a large transformer the cap is below what the network can carry. Read the rows above for the model shown.
The four cap readings with an engineering P5 are held for the generic 400 V sensitivity only (results/hosting_cap_unit_sensitivity.csv); the 50% cap and the 0.9 case are shown for every model.
50% cap: open question unit assumed: kW on the DTR kVA rating; TGSPDCL to confirm. The 0.9 case is read from the same stored P5 values (arithmetic on the cap, no new power flow); the precomputed rows are in the cap-unit table below.
- Peak solar output used in the screening test: 773 W per kWp of panel (the sunniest observed hour, 6 Mar 2025, physics model on archive weather, not metered) at a midday home use of 0.25 kW (hosting/run_indian.py).
- Every network value is an assumed typical value; no transformer, feeder or house data is held.
- Indian street model (hosting/run_indian.py): 11 kV/433 V DTR, 415/240 V declared supply; conductors ACSR Rabbit and Weasel (IS 398 resistance raised 15% for midday temperature) and aerial bunched cable; DTR ratings 63 / 100 / 160 / 315 kVA from common Indian stock. A 433 V DTR at 1.03 pu on the 11 kV side can be over +6% before any solar: that case is shown as over the limit before solar, not as 0 kWp safe.
- The generic 400 V model (generic transformer and cable) is kept as a sensitivity case; the OpenDSS and roof-by-roof checks were run on the generic model only.
- P5 is the 5th percentile of 200 random trials of which homes adopt solar and on which phase: a screening value, not a guarantee.
- kWac figures are derived from the kWp results with each reading's own factor and are estimates (the transformer rating stays in kVA).
- TGSPDCL guideline and agreement texts, CSTEP feeder report, MNRE table (PDFs) REAL not a dated file. Source: data/telangana_reports/*.pdf.
- Roof size: 5.91 kW average billed load per domestic net-meter service (REAL) against an assumed 3 kWp roof (EXAMPLE). Billed load is not confirmed to be installed solar kW, so this is a comparison for context, not a replacement of the assumed roof.
- State context: Telangana rooftop solar 695.9 MW as on 31.01.2026 (MNRE, REAL, a state total, not per street).
- Not held: Real transformer data (kVA, consumers, phase, 15-minute DTR or feeder load). Not public; the DTR list has to come from the department. Every network value here is assumed.
- Not held: Metered rooftop solar output. Not public and not downloadable (PVGIS stops at 2023; no metered source was found). Forecast accuracy is versus a physics model, not versus measurement.