Good afternoon.

Every godown below (S.No 1–52) is transcribed directly from the scanned TSWC register you sent — the full 52 of 52 listed godowns, real names, districts and rated capacities, 5,94,400 MT total.

• S.No 23–27 (Suryapet Dist and Khammam-I) were added from a second rescan of the page section that was originally cut off between the two scanned sheets — the register is now complete with no known gaps.
• One correction from that rescan: S.No 28–31 (Khammam-II, Raghunadhapalem, Kothagudem, Palvancha) were previously tagged "Khammam Region" — the source register has no such heading, and Khammam District and Badradhri (Kothagudem) Dist are both printed under "NALGONDA Region" (confirmed by the printed Region Total of 1,60,550 MT, which only reconciles when Nalgonda + Suryapet + Khammam + Badradhri Kothagudem districts are summed together). Corrected to Nalgonda Region.
• This register only covers TSWC-owned ("Own") godowns — it has no FCI, CWC or private-agency data, and no commodity or occupancy/utilization figures, so those columns from the earlier sample mockup have been removed rather than filled with guesses.
• One internal mismatch: the printed Siddipet district subtotal is 65,700 MT but its 6 line items add up to 65,000 MT — flagged in that row's notes, not silently corrected.

Combining the scanned TSWC register with the "TGSWC Warehouses Locations" Google My Maps export and OpenStreetMap geocoding is what makes the map on this page possible — but each of those sources has its own limits, and stacking them compounds a few risks worth naming plainly:

• Neither source has real GPS. The scanned register has no coordinates at all, and the Google My Maps KML export — despite showing pins visually on the map — turns out to carry no <coordinates> or <Point> geometry for any of its 288 markers either. Every "exact" location on this page is a reconstruction — matched by village name, then looked up on OpenStreetMap — not a coordinate pulled directly from either source.
• Village names collide. Telangana has many villages that share a name across different districts (multiple "Chityal"s, "Mailaram"s, "Jakkapur"s and so on). A name-only match can silently pick the wrong one if the district isn't also checked — we cross-checked district on every match here, but a future refresh of this data needs to keep doing that, not just match on name.
• Districts were redrawn twice (2016 and again in 2019/2021), so a godown's district label depends on which year its source was written down. "Warangal Urban", for instance, no longer exists — it's now split between Warangal and Hanumakonda districts. The scanned register and the Google Maps export don't always agree on which district a shared village belongs to.
• Not every godown has an exact GPS match. Of the 52 TSWC-owned godowns on this page's map, 45 now resolve to an exact coordinate — 42 straight from TSWC's own warehouse coordinate export, plus 3 more geocoded to their real village on OpenStreetMap; the remaining 7 (4 with no usable coordinate anywhere, plus 3 corrected against public sources below) fall back to a district- or mandal-headquarters approximation rather than risk plotting a wrong village.
• TSWC's own coordinate export had errors of its own. Cross-checking every depot against Telangana's state border and against every other depot in its district turned up 3 rows that were wrong by 50-230 km — S.No 5 (Suguru), 41 (Gundlapahad) and 45 (Choutuppal) — plus one more (S.No 12, Allapur/Toopran) that was ~110 km off with no public source agreeing with it. All four are marked with a warning in their map popup; the first three now use a verified public coordinate (mandal headquarters, sourced from Wikipedia and public village directories) in place of TSWC's figure, and S.No 12 uses its OpenStreetMap match instead, which independently agrees with Wikipedia to within half a kilometre.
• Both sources are static snapshots — the scan is dated before December 2023, and the Google My Maps export is a one-time download from August 2026. Neither is a live feed, so a godown added, closed, or relocated after those dates won't appear or update here automatically.

These come from a second, different source — TSWC's statewide, signed "Business Report JULY 2026" scan — not the pre-2023 Own register above. It lists every warehouse TSWC stored stock in that month whether TSWC owns it or not, and — unlike the Own register — already gives both rated capacity and that month's occupancy/agency breakdown in the same table, so no separate xlsx merge was needed for these rows.

• 499 of the report's 510 non-Own rows were transcribed (366 AMC/Hired, 132 Investor, 1 PEG) — 378 AMC/Hired, 131 Investor and 1 PEG per the report's own printed abstract. The AMC/Hired total is short by about 50,000 MT of capacity (roughly 11 rows' worth) that fell in a page-seam gap during transcription; the Investor total is short by about 18,000 MT despite one extra row, suggesting a misread figure somewhere rather than a missing row. Neither gap has been guessed at or backfilled.
• None of these 499 have a coordinate source of their own — the report lists a warehouse name and district, nothing else. As a follow-up pass, each row's name was stripped down to a likely village (dropping suffixes like "AMC", "Rice Mill", company names) and checked against English Wikipedia, requiring the article to explicitly confirm the right district before accepting a match — 172 of 499 (34%) resolved this way, all real village-level coordinates, not warehouse-exact. The other 327 either had no Wikipedia article, an article with an empty coordinates field, or were pure company/facility names with no village to look up — those still fall back to the district headquarters, spiraled outward so they don't stack. One tentative match (a "Nizamsagar" candidate 150 km from Kamareddy district's centre) was found and discarded rather than kept, on the same reasoning as the Own register's bad-coordinate corrections. Use the category filter to isolate just the Own register on the map if exact placement matters — most of it does have a real coordinate.
• Districts here don't fully match the Own register's district list — this report covers a few districts (e.g. Adilabad, Jayashankar Bhupalpally, Jogulamba Gadwal, Komaram Bheem Asifabad, Mancherial, Mulugu, Rajanna Sircilla, Sangareddy, Vikarabad, Hanamakonda, Jangaon) that have no TSWC-owned godown in the Own register at all — only hired/investor/PEG capacity. Their district-headquarters coordinates were newly added for this update and haven't been independently re-verified the way the Own register's were. One of these, Nagarkurnool, was missed entirely in the first pass — its 18 rows were silently plotted at the map's generic fallback point near Jangaon, ~120 km away, until a later re-check caught the gap and added its coordinate. Worth knowing in case a similar district-list gap is still hiding elsewhere in a future data refresh.
• The per-agency occupancy split (FCI/CSC/NAFED/Markfed/etc.) is lower-confidence than the capacity and total-occupancy figures for a number of rows, particularly on the more crowded pages — the transcription cross-validated row totals against printed district subtotals everywhere, but a few individual agency columns on dense pages were a best-effort read. Rated capacity and total occupancy are solid throughout; treat any single agency's number with a bit more caution.