Corrections

When a number we published turns out to be wrong, we correct it and record it here: what it was, what it is now, and why. Newest first.

Follow corrections by RSS

  1. Substation capacity for new load, Telangana

    Firm N-1 MW a new load can draw at each EHV connection bus, with its range, limiting element and accuracy

    Corrected on

    Was
    v1: 118 buses, each with a figure; shared state-wide limit 396 MW; held-out flow error 49-56 MW
    Now
    v2: 135 buses (17 added), 100 with a figure and 35 marked not assessed with the reason; shared state-wide limit 377 MW; held-out flow error 45-51 MW; 86 of 118 buses change band
    Period
    Study window 14-27 Sep 2026; dataset v1, published 14 Oct 2026
    Why
    The v1 network carried a 213 km line that is really three lines in series, a duplicate of a metered line, yard trunks with four circuits mapped as two and lighter conductors than the transmission register gives; it left out the Nagarjunasagar pumped-storage plant's pumping, which forced its midday import onto one line; and default 220 kV ratings were computed at 230 kV. All are fixed from the SRLDC line register and the meters, and the study recalculated. Where the model cannot rate a station's supply, the station is now marked not assessed instead of given a figure.
    Reference
    DQ-55

    Open the corrected page

  2. Substation capacity for new load, Tamil Nadu

    Firm N-1 MW a new load can draw at each of 158 EHV connection buses, with its range, limiting element and accuracy

    Corrected on

    Was
    v1: shared state-wide limit 408 MW; median firm N-1 50 MW; held-out flow error 100-133 MW; the flow-correlation figure was the share of flow directions right (0.90-0.92), not a correlation
    Now
    v2: shared state-wide limit 342 MW; median firm N-1 53 MW; held-out flow error 79-122 MW; flow correlation 0.90-0.93; 32 of 158 buses change band, the limiting element changes at 115
    Period
    Study window 14-27 Sep 2026; dataset v1, published 10 Oct 2026
    Why
    The study's supply model counted the Mettur thermal station a second time as hydro generation (0.7-1.0 GW), and let some wind, solar and hydro enter the grid through power-plant switchyards instead of the substations they connect to. Its result also depended on the order a program read its inputs. All three are fixed and the study recalculated; the new figures are more accurate on data the model never saw.
    Reference
    DQ-54

    Open the corrected page

  3. IEX forecast scorecard

    RTM scorecard: the forecast graded and its naive benchmark

    Corrected on

    Was
    The RTM day-ahead model (rtm-dayahead-gbdt-longhist-cqr-v1), graded at 00:00 IST against the same RTM block the day before: 30-day MAE ₹2,124/MWh, skill −161%, band coverage 56% (to 6 Oct)
    Now
    The RTM next-day pre-auction forecast (rtm-next-day-v1.1), graded at the 11:00 IST DAM gate against the issue day's DAM price: MAE ₹1,077/MWh, skill −18%, band coverage 94% (27 Sep–5 Oct, 9 days)
    Period
    Delivery days from 27 Sep 2026
    Why
    The RTM day-ahead model was retired: it was trained on realised DAM prices but run on the DAM forecast, and it lost to a naive guess. The scorecard and the RTM forecast API now use the next-day pre-auction forecast. At its 06:25 IST issue a desk knows the issue day's DAM price but little of its RTM, so that DAM price is the naive guess. The retired model's grades stay in the archive and are no longer shown.
    Reference
    SP-21

    Open the corrected page

  4. Data-centre map

    Special Economic Zone count

    Corrected on

    Was
    276 "Official" SEZs
    Now
    275 operational SEZs (those drawn on the map)
    Why
    The page counted 276 official SEZs while drawing 275; it now shows the 275 operational SEZs on the map.
    Reference
    DQ-51

    Open the corrected page

  5. Data-centre map

    Substation count

    Corrected on

    Was
    16,922 substations (a stale count)
    Now
    28,251 substations (the map snapshot, as on /map)
    Why
    The figure was a hard-coded legacy count; it now reads the same map-snapshot count /map shows, from the metric registry.
    Reference
    DQ-51

    Open the corrected page

  6. IEX DAM price

    DAM blocks at or above ₹8,000/MWh in a delivery day

    Corrected on

    Was
    159 of 265 blocks
    Now
    64 of 96 blocks
    Period
    Delivery day 30 Sep 2026
    Why
    The count ran over 265 blocks from 28-30 Sep instead of the 96 blocks of the labelled day.
    Reference
    DQ-51

    Open the corrected page

  7. IEX DAM price

    DAM blocks at or above ₹5,000/MWh in a delivery day

    Corrected on

    Was
    165 of 265 blocks
    Now
    64 of 96 blocks
    Period
    Delivery day 30 Sep 2026
    Why
    The count ran over 265 blocks from 28-30 Sep instead of the 96 blocks of the labelled day.
    Reference
    DQ-51

    Open the corrected page

  8. IEX DAM volume

    DAM volume for a delivery day

    Corrected on

    Was
    529.1k MWh (about three days)
    Now
    184.6k MWh
    Period
    Delivery day 30 Sep 2026
    Why
    The volume added up about three days of DAM rows but was labelled as one delivery day; it now covers only the 96 blocks of that day.
    Reference
    DQ-51

    Open the corrected page

  9. IEX DAM price

    DAM minimum block price (the day's floor)

    Corrected on

    Was
    ₹1,332/MWh (a 28 Sep block)
    Now
    ₹2,000/MWh
    Period
    Delivery day 30 Sep 2026
    Why
    The floor was taken over the newest 500 IEX rows, which covered 28-30 Sep, so the 30 Sep floor was a 28 Sep block; it now uses only the 96 blocks of the named day.
    Reference
    DQ-51

    Open the corrected page

  10. IEX DAM price

    Daily average DAM price

    Corrected on

    Was
    ₹7,107/MWh on /iex-market; ₹6,973/MWh on /bess
    Now
    ₹7,652/MWh
    Period
    Delivery day 30 Sep 2026
    Why
    The page averaged every DAM row among the newest 500 IEX rows, 265 blocks over about three delivery days, under one date; it now uses only the 96 blocks of the named day.
    Reference
    DQ-51

    Open the corrected page

  11. Installed generation capacity

    Installed capacity: total, renewable, thermal and nuclear

    Corrected on

    Was
    1,109 GW total; renewable 104.1 GW (9%); thermal 500.4 GW; nuclear 17.6 GW
    Now
    555 GW total (CEA, 31 Aug 2026); renewable 295.6 GW (53%); thermal 250.2 GW; nuclear 8.8 GW
    Why
    The page added the CEA report's region rows to its state rows, counting every megawatt twice, and left the renewables bucket out of Renewable.
    Reference
    DQ-44

    Open the corrected page

  12. IEX DAM price

    Daily average DAM price

    Corrected on

    Was
    ₹5,346/MWh on /iex-market; ₹5,466/MWh on /bess
    Now
    ₹7,448.55/MWh (the simple mean of IEX's 96 blocks)
    Period
    Delivery day 29 Sep 2026
    Why
    Both pages averaged every DAM row among the newest 500 IEX rows, about 2.75 delivery days, under one date. Fixed under DQ-51.
    Reference
    DQ-43

    Open the corrected page

  13. National electricity demand

    Latest national demand

    Corrected on

    Was
    130.3 GW
    Now
    230.8 GW (28 Sep 2026, 10:13 IST)
    Why
    The page read a cached generation total, dropped its timestamp and never refreshed it.
    Reference
    DQ-40

    Open the corrected page

  14. State demand ticker

    Hour-on-hour change in state demand

    Corrected on

    Was
    0.0% for 24 of 34 states
    Now
    0 of 34 states at exactly 0.0%; no change shown without a comparable reading
    Why
    Readings stamped at half past the hour were copied into two hours, so the ticker compared a reading with its own copy.
    Reference
    DQ-32
  15. Home page

    Searchable regulatory documents

    Corrected on

    Was
    More than 5,100 documents and passages
    Now
    595 documents from 5 authorities
    Why
    The figure counted 4,581 passage chunks as documents.
    Reference
    DQ-31

    Open the corrected page

  16. State generation mix

    Day-ahead forecast line on the generation-mix card

    Corrected on

    Was
    Flat at about 8 GW after a 12.5 GW actual stack (Andhra Pradesh); gaps at the seam of -31% (Andhra Pradesh), -74% (Himachal Pradesh), -20% (Delhi)
    Now
    Drawn only within 15% of the last actual hour; otherwise withheld, with a note
    Why
    The line plotted a fuel split of the demand forecast on top of a generation stack.
    Reference
    DQ-30

    Open the corrected page

  17. State power flows

    State net flow and in-state generation

    Corrected on

    Was
    Maharashtra +39,676 MW net export on 26,290 MW demand (25 Sep); all-India net +203,664 MW
    Now
    Maharashtra -2,930 MW net import (demand 24.2 GW, generation 21.3 GW); all-India net -1,982 MW
    Period
    22 Jan-26 Sep 2026 (a full double count from 20 Aug)
    Why
    In-state generation added up three overlapping production models for the same state-hour.
    Reference
    DQ-29

    Open the corrected page

  18. State peak demand

    Jammu and Kashmir rolling peak demand

    Corrected on

    Was
    5,197 MW
    Now
    3,634 MW
    Period
    From 25 Jul 2026
    Why
    Jammu and Kashmir's data was split across two names, and one of them served a peak frozen since 25 Jul.
    Reference
    DQ-26
  19. State fuel mix

    Coal generation by state against CEA DGR-2

    Corrected on

    Was
    Assam +16 to +21%, Haryana -11 to -15%, Odisha -10 to -12%, Uttar Pradesh -5% against DGR-2 (12-14 Jul); Haryana 0 MW coal in 16 of 24 live hours
    Now
    Within 2% of DGR-2 on 407 of 408, 391 of 391 and 424 of 425 state-days; Haryana coal in 31 of 31 hours, mean 3,932 MW
    Period
    12 Jul-18 Aug 2026, and live hours from about 20 Sep 2026
    Why
    A deduplication step rebuilt each state-hour's fuel split from stale or partial model rows, which distorted anchored coal and in Haryana turned coal into hydro.
    Reference
    DQ-24
  20. State fuel mix

    Coal generation, Chhattisgarh, Maharashtra and Uttar Pradesh

    Corrected on

    Was
    0 MW
    Now
    Chhattisgarh about 19.0-19.6 GW, Maharashtra about 13.8 GW, Uttar Pradesh about 17.8-20.5 GW (equal to CEA DGR-2)
    Period
    4-7 Aug 2026
    Why
    The CEA daily report (DGR-2) feed behind the state coal anchor froze on 7 Jul 2026, and the days it missed were never re-anchored.
    Reference
    DQ-8
  21. State fuel mix

    National coal generation (sum of states)

    Corrected on

    Was
    50-53 GW
    Now
    145-158 GW
    Period
    1-8 Aug 2026
    Why
    The CEA daily report (DGR-2) feed behind the state coal anchor froze on 7 Jul 2026, and the days it missed were never re-anchored.
    Reference
    DQ-8