The Mahape data centre L&T Vyoma broke ground on in January 2026 is the first phase of a 100 MW campus planned in Navi Mumbai. That is the entire public description of the campus: 100 MW, planned, basis not stated. Six months later the company described the campus as two facilities of about 40 MW each, which totals 80 MW and leaves 20 MW unaccounted for. No campus land area is published. This registry counts the 40 MW of phase one through its own record and counts zero here, so a plan cannot inflate a state total.
1At a glance
2Forty plus forty is eighty
| Source | Campus figure | Detail |
|---|---|---|
| L&T Vyoma release, 21 Jan 2026 | 100 MW | Part of a 100 MW campus planned in the city |
| The Hindu, 21 Jan 2026 | 100 MW | First phase of a planned 100 MW campus |
| Inc42, Aug 2026 | two facilities, ~40 MW each | Campus potentially scaling up to 100 MW |
| Operator specification page | 40 MW IT | Phase one only; no campus figure |
| This registry | 0 MW | Plan not counted; phase one counted separately |
The gap is 20 MW — one fifth of the plan — and it is not a rounding issue. Either there is a third building nobody has described, or the second building is larger than 40 MW, or the 100 MW is a site ceiling that includes capacity not yet designed.
Every one of those is a normal way for a developer to talk. None of them is a facility, and none can be counted. This is the fourth campus-scale plan in this registry where the headline exceeds the named components — the L&T 100 MW, the AirTrunk 601 MW against 600, the Iron Mountain 150 MW against 147.5, and the Raigad 3 GW against nothing.
Indian campus announcements are consistently stated as a round number, named in smaller pieces. 100 MW becomes two 40s. 250 MW becomes 148 plus a Chennai allocation. 600 MW becomes five cards.
In most cases the round number is a genuine site ceiling and the pieces are what will actually be built. Both are useful; only one is capacity.
A data product that publishes the round number alone is a news feed. One that publishes the round number, the named pieces and the difference between them is a diligence tool.
3How much land a 100 MW campus needs
No campus land area is published, so it has to be inferred from the operator’s own phase-one density of 16.3 MW per acre.
| Density assumption | Density | Land needed | Basis |
|---|---|---|---|
| L&T’s own phase-one density | 16.3 MW per acre | 6.1 acres | 40 MW on 2.45 acres |
| AirTrunk BOM2, same node | 9.0 MW per acre | 11.1 acres | 120 MW on 13.3 acres |
| CapitaLand Chennai, four acres gross | 13.3 MW per acre | 7.5 acres | 53 MW gross on 4 acres |
| Likely range | — | 6 to 11 acres | Density dependent |
Six to eleven acres in Mahape is not a small parcel. The MIDC industrial belt at Mahape and the neighbouring nodes are built up, and the 2.45 acres used for phase one is consistent with buying infill plots rather than assembling a greenfield campus.
Which is the most likely explanation for the missing 20 MW: a campus plan that depends on land not yet acquired. If the site exists, the plan is credible; if it does not, the 100 MW is a target.
4The Mahape node
| Record | MW | Status |
|---|---|---|
| AirTrunk BOM2 Mahape (PRJ-181) | 120 | Facility, at least; stage unverified |
| L&T Vyoma campus plan (this record) | 100 | Plan; 40 MW with a groundbreaking |
| Mahape published total | 220 | Mixed scopes and stages |
| Maharashtra counted total | 786.4 | IT load, additive rows only |
| Mahape as a share | 28% | Of the state’s counted capacity |
220 MW published in one node of Navi Mumbai — 28% of everything Maharashtra has counted — and none of the L&T figure exists yet.
This is the same finding the registry recorded on PRJ-181: the Mahape to Airoli belt publishes something close to 600 MW against roughly 30 MW live. Adding a 100 MW plan to a 120 MW claim does not change the ratio; it confirms it.
What does change is the competitive picture. AirTrunk at 120 MW and L&T at 100 MW in the same node means Mahape is no longer a secondary Navi Mumbai location. It is where the next phase of the market is being built, and both of them need the same substation capacity.
5Procurement
6Limits and open questions
- Plan stage. No land area, no phase schedule, no capex and no basis for the 100 MW figure.
- The named components total 80 MW against a stated 100 MW. Twenty megawatts unaccounted for.
- Six to eleven acres is inferred from three density assumptions and is not a disclosed figure.
- The campus figure may be a site ceiling rather than a build plan. Not stated.
- Zero megawatts counted at campus level. The 40 MW phase one is counted once, through PRJ-201.
- What would close it. A land area and a phase schedule. Either alone would resolve half the page.
7Sources
Sources: the Larsen & Toubro Vyoma press release of 21 January 2026; The Hindu and ETV Bharat coverage of the same date; Inc42 coverage of the Vyoma.AI restructuring and campus plan; and the L&T Vyoma Mumbai data centre specification page.