India has eight LNG import terminals. PPAC’s latest Ready Reckoner lists 53.5 million tonnes a year (MMTPA) of regasification capacity; counting Petronet LNG’s expansion of Dahej from 17.5 to 22.5 MMTPA, commissioned on 31 March 2026, the nameplate is about 58.5 MMTPA. One of them, Dahej, has run at or near full capacity for a decade. The other seven have mostly run at a fifth to a half of theirs. Two more are being built or planned at existing sites, and several new ones are approved or proposed, including Petronet’s ₹6,355 crore terminal at Gopalpur and Shell’s ₹21,600 crore expansion of Hazira.

This piece compares the two ways of adding regas capacity: expanding an existing terminal (more tanks, more vaporisers, sometimes a jetty) or building a new one. It uses the utilisation tables PNGRB publishes from PPAC data, the 2025 PNGRB expert committee report, and the cost and schedule of every terminal project for which figures could be found. The short answer: expansion has cost about a third as much per tonne of capacity and slipped about a third as long, but the bigger difference is that capacity at Dahej gets used and capacity elsewhere mostly has not, because the pipelines to carry the gas away were not ready.

One full terminal and seven part-empty ones

PNGRB’s April 2025 demand paper prints PPAC’s terminal utilisation for 2023-24 and the eight-year average from 2015-16 to 2022-23. Dahej averaged 98.2% over the eight years and 95.1% in 2023-24. Kochi, commissioned in 2013, averaged 13.1%; Ennore, commissioned in 2019, 12.0%; Mundra 24.9%. Across all eight terminals utilisation averaged 54.6% over the eight years and 50.5% in 2023-24. The committee report’s figures for April–December 2024 tell the same story: Dahej at 100.3%, the other six in service between 22.0% (Kochi) and 49.5% (Dhamra).

PNGRB, “NG Demand projections for 2030 and 2040” (April 2025), p.21, source line “PPAC” (first two utilisation columns and the totals); PNGRB High Level Expert Committee, “Vision 2040 – Natural Gas Infrastructure in India” (November 2025), Table 6.4(A), p.176 (April–December 2024); PPAC, Snapshot of India’s Oil & Gas data (Monthly Ready Reckoner), May 2026, Table 21 (capacity as on 1 June 2026, utilisation April–May 2026). PPAC lists Hazira at 6 MMTPA (5.2 in PNGRB’s February 2025 table) and still lists Dahej at 17.5 MMTPA, although its expansion to 22.5 MMTPA was commissioned on 31 March 2026; with it, nameplate capacity is about 58.5 MMTPA on PPAC’s figures. PPAC notes that only Dabhol’s 2.9 MMTPA high-pressure stream is commissioned. *Capacity-weighted average of PPAC’s terminal figures, this blog’s arithmetic.

PPAC’s newest terminal table that this piece could read — Table 21 of the Ready Reckoner for May 2026, published in June — shows the same split in April–May 2026. Dahej ran at 91.9%. Of the four new onshore terminals, Ennore ran at 26.4%, Mundra at 24.1%, Dhamra at 36.9% and Chhara, in its second year, at 10.5%; Kochi was at 26.3%, Hazira at 27.0% and Dabhol at 40.1%. Weighted by capacity, the eight terminals ran at about 48%.

Two details in that table bear directly on the expansion question. PPAC still lists Dahej at 17.5 MMTPA on 1 June 2026, two months after its 22.5 MMTPA expansion was commissioned. Petronet’s own results put Dahej at 65.6% of the enlarged capacity in April–June 2026, against about 92% a year earlier; on the old base the same volume would have been about 85%, according to a brokerage review of the results. The expansion arrived just as Dahej’s main long-term supplier, QatarEnergy, declared force majeure after the March 2026 attack on Ras Laffan, so for now the new 5 MMTPA is mostly headroom. PPAC also notes that only Dabhol’s 2.9 MMTPA high-pressure stream is commissioned, so its 40.1% understates how full the working part of that terminal is.

Imports have fallen rather than grown. PPAC puts LNG imports at 34,216 million standard cubic metres (MMSCM) in 2025-26, 4.2% below 2024-25, and at 4,532 MMSCM in April–May 2026, about 20% lower year on year (provisional). Press reports of PPAC’s monthly data give 2,604 MMSCM for June (5.4% lower), 2,915 for August (about flat) and 2,957 for September (8.5% lower). At the August–September rate India is importing about 26.6 MMTPA, roughly 45% of nameplate capacity. Supply shocks like the one at Hormuz move all terminals together; they do not explain why the gap between Dahej and the rest has lasted ten years.

Why the empty ones are empty: the pipe out

The PNGRB committee gives three reasons why “most LNG regasification terminals in India, except for PLL Dahej, operate at less than 50% capacity utilisation” (p.180): inadequate evacuation pipelines, destination clauses in long-term LNG contracts that tie cargoes to one terminal, and buyers who switch to other fuels when LNG is expensive. It names Kochi and Ennore as terminals “where connectivity faced prolonged delays despite their readiness”. Its own pipeline tables show how far behind the pipes still are.

PNGRB HLEC (November 2025): Tables 12.7 and 12.8 (pp.271–273) for length, utilisation and status; p.276 for the Gujarat evacuation figures; p.42 for the Bengaluru gap. GAIL’s integrated-network pipelines all carry the same 71.69% in Table 12.7, a network figure rather than one per line, so it is not shown. Dabhol breakwater dates from press reports.

Kochi’s terminal was finished in 2013; GAIL’s line from Kochi, whose cost rose from ₹2,915 crore to more than ₹5,750 crore over land acquisition and compensation, is still 644 km of an authorised 1,104 km and is not joined to the national grid. Ennore’s line is 1,080 km of 1,431 and ran at 5.9% of its capacity. Dabhol had no breakwater until April 2025 and closed for about four monsoon months a year. Mundra was ready in October 2018 but took its first cargo in January 2020 after a lease dispute.

The committee also flags a constraint that applies to expansion, not just to new terminals. Dahej, Mundra and Hazira have about 95–100 MMSCMD of regas capacity between them, and Dahej’s expansion and the Jafrabad floating terminal add about 37 MMSCMD more, yet the GSPL pipelines that evacuate gas from them carry “only around 40 MMSCMD” (p.276), and GSPL’s high-pressure network ran at 100.5% in 2024-25. Dahej also feeds GAIL’s pipelines, which is why it can run full; Hazira’s planned expansion to 26.2 MMTPA and Jafrabad would land in the same constrained corner of Gujarat.

What a tonne of capacity costs

Petronet’s latest Dahej expansion added 5 MMTPA for about ₹1,850 crore: ₹600 crore for send-out and ₹1,250 crore for two 180,000 m³ storage tanks, as the company described it in October 2023 (CRISIL put the total near ₹1,800 crore). That is about ₹370 crore per MMTPA. The send-out equipment alone came to about ₹120 crore per MMTPA, because the jetties, land, permits and pipeline connections were already there. The previous expansion, 10 to 15 MMTPA, cost about ₹3,000 crore for tanks and send-out, or ₹600 crore per MMTPA, before a second jetty.

The four new onshore terminals finished since 2019 cost between ₹950 and ₹1,290 crore per MMTPA: Chhara about ₹4,750 crore, Mundra about ₹5,040 crore, Ennore about ₹5,150 crore and Dhamra about ₹6,450 crore, each for 5 MMTPA. Petronet’s approved Gopalpur terminal is budgeted at ₹6,354.80 crore for 5 MMTPA, about ₹1,271 crore per MMTPA. Each of these also went over its first estimate: Chhara by about 12% (₹4,293 crore to ₹4,792 crore, per ICRA), Dhamra by about 15%, Ennore by about 19% against a 2012 figure.

Not every expansion is cheap. Shell’s plan to take Hazira from 6.28 to 26.2 MMTPA needs a new breakwater, a jetty for conventional carriers and 2.8 million m³ of dredging, and is costed at ₹21,600 crore, about ₹1,084 crore per MMTPA — new-terminal money at an old address. The cost advantage of expansion comes from reusing marine works and pipeline connections, not from the word “brownfield”.

Nominal rupees of the year each figure was reported; not adjusted for inflation. Costs come from company filings, rating rationales and press reports, mostly read through search summaries; Gopalpur’s figure is from Petronet’s exchange filing and Chhara’s from the PNGRB committee report. Kochi’s cost rose from ₹2,000 crore (2003 plan) to ₹4,200 crore (2012); later press figures go up to ₹4,700 crore.

Two benchmarks help read these numbers. Storage drives much of the cost: Indian terminals carry about 55,000–95,000 m³ of tank capacity per MMTPA (two 160,000–200,000 m³ tanks per 5 MMTPA), and Dahej’s two new tanks cost about ₹625 crore each. Abroad, a 2017 Oxford Institute for Energy Studies paper put a generic 3 MMTPA onshore terminal at US$700–800 million and a floating storage and regasification unit (FSRU) terminal at about half that; new FSRU vessels have since been ordered at US$320–364 million each, and Bangladesh pays about US$250,000 a day to charter one.

How long it takes, and how late it runs

Measured against the first commissioning date each project announced, Dahej’s latest expansion ran about 12 months late (March 2025 to March 2026). Its two tanks took about 37 months from the September 2021 construction order to commissioning in late 2024. The new onshore terminals ran 22 to 35 months late, and from board approval to commissioning Ennore took about 52 months. The committee report’s own planning assumption is that “construction of an LNG terminal typically takes three years”.

The slowest projects are the ones that depend on new marine works. Dabhol’s breakwater, planned in 2018 to take about three years, was finished in April 2025. The Jaigarh FSRU, due in late 2018, did not start until 2022 at the earliest. Jafrabad, due by April 2020, is still not operating: its breakwater was damaged by monsoon and cyclones, and its FSRU was chartered to, and then sold to, Turkey’s Botaş. Elsewhere FSRUs have been the quick option — Germany’s first took about six months in 2022 — but only where a government cleared the jetty, the pipeline and the buyers at the same time.

The recurring causes are the same across projects: breakwaters and channel dredging (Dabhol, Chhara, Jafrabad, Ennore, Hazira’s expansion), evacuation pipelines not ready at commissioning (Kochi, Ennore, Chhara), lease and concession disputes (Mundra), environment and forest clearances (Ennore, Hazira, Chhara’s pipeline through the Gir eco-sensitive zone), and FSRU charters (Jafrabad, Jaigarh).

What it means for the regas charge

A terminal’s costs are mostly fixed, so its charge per unit of gas rises as utilisation falls. The table below spreads the capital cost of four kinds of 5 MMTPA project over the gas they handle at different utilisation levels. At 90% a Dahej-type expansion needs about ₹12/MMBtu to recover its capital; a Dhamra-type new terminal at 30% needs about ₹126.

This blog’s illustration, not a published tariff: capital recovered over 25 years at 12%, plus fixed operating cost of 2.5% of capex a year, spread over 5 MMTPA × 52 MMBtu per tonne × utilisation. It leaves out power, boil-off, return on equity above 12% and tax. For comparison, the charges in force for 2025 were about ₹66.05/MMBtu at Dahej, Hazira and Mundra and ₹79.80 at Kochi and Ennore (PNGRB HLEC, p.179), each rising 5% a year; ICRA estimates a 5 MMTPA terminal needs more than ₹81 at 30–55% utilisation to service its debt.

That arithmetic explains three observations. Petronet could offer GAIL capacity in the Dahej expansion at ₹52.05/MMBtu, more than 25% below its standard Dahej charge, as Argus reported in December 2025. The newer terminals charge more — ₹79.80 at Kochi and Ennore against ₹66.05 at Dahej for 2025 — and still struggle: ICRA estimates a 5 MMTPA terminal needs more than ₹81/MMBtu at 30–55% utilisation to service its debt. And the PNGRB committee’s complaint that regas charges escalate 5% a year regardless of how full a terminal is (p.181) is a complaint about who pays for empty capacity.

How much capacity demand actually needs

PNGRB’s base case (“Good-to-Go”) has gas demand rising from 187 MMSCMD in 2023-24 to 297 MMSCMD in 2030, with LNG demand doubling “to ~200 MMSCMD”. At PNGRB’s own conversion (190 MMSCMD = 52.7 MMTPA) that is about 55.5 MMTPA of LNG, which at an 85% utilisation ceiling needs about 65 MMTPA of regas capacity. India already has about 58.5 MMTPA; the under-construction Jafrabad and the approved Gopalpur would take it to about 68.5. The committee lists another 25.2 MMTPA as proposed, and Hazira’s cleared expansion would add about 20 MMTPA more.

The higher “Good-to-Best” case (365 MMSCMD in 2030) would need more. PNGRB gives no LNG split for it; if domestic production stayed near the ~100 MMSCMD the committee assumes in the base case, LNG would be about 265 MMSCMD or 74 MMTPA, roughly 86 MMTPA of capacity at 85%. By 2040 both cases need two to three times today’s capacity.

The tonnage is not the binding problem for 2030 in the base case; location and connection are. Raising Kochi, Ennore, Mundra, Dabhol and Dhamra from their April–May 2026 utilisation to 70% would add about 9.8 MMTPA of throughput from 25 MMTPA of terminals that already exist — more than any single expansion now planned except Hazira’s — at the cost of pipelines rather than tanks.

Expansion or new build: the metrics that decide it

The record suggests a short list of numbers that would tell, before a project is approved, whether its capacity will be used:

• Capital cost per MMTPA, with the marine works shown separately. Recent range: about ₹370–600 crore for expansions that reuse jetties; ₹950–1,300 crore for new onshore terminals or expansions that need a breakwater.
• Evacuation headroom at commissioning: pipeline capacity booked and built from the terminal gate, in MMSCMD, against the terminal’s send-out. Kochi and Ennore show what happens when this is zero; the Gujarat cluster shows it can be negative for expansions too.
• Months from board approval to commercial operation, and slippage against the first date: about 37 months for tanks at an existing site, 52 or more for a new terminal, open-ended where a breakwater or FSRU charter is involved.
• Utilisation in year three, the test every terminal since Dahej has failed: Kochi 13% average over its first decade, Ennore 12%, Mundra 25%.
• Regas charge needed at 50% utilisation, against the charges buyers already pay (₹66–80/MMBtu in 2025).
• Contract flexibility: whether the anchor LNG contracts allow cargoes to be landed at any Indian terminal, or tie them to one by destination clause.

What doesn’t follow from any of this

None of this says new terminals are never justified. East-coast demand cannot be served from Gujarat without pipelines that are themselves unfinished; terminals add storage, which the committee notes covers only about 12 days of national gas consumption across 23 tanks; and a second terminal in a region adds resilience that utilisation figures do not measure. Dahej’s low cost also rests on two decades of earlier spending on jetties and connections that a new site has to pay for once. The cost figures are nominal rupees from 2012 to 2026, several come from press reports or rating rationales rather than audited accounts, and some figures from the same project disagree (Kochi’s cost appears as ₹2,000, ₹3,600, ₹4,200 and ₹4,700 crore in different years). The 2026 terminal figures are press reports of PPAC data that this piece could not read directly.

What would fix this

• PNGRB should make registration of any new LNG terminal or expansion conditional on evacuation capacity — booked pipeline capacity from the terminal gate, scheduled to be in service within 12 months of the terminal — and publish that test with each registration under the 2025 regulations, so that Kochi’s decade at 13% is not repeated.
• PPAC should publish terminal-wise capacity, utilisation and LNG receipts each month as a machine-readable table, stating the capacity used as the denominator, rather than only as a table in the Ready Reckoner PDF. Its May 2026 edition still listed Dahej at 17.5 MMTPA two months after the expansion to 22.5 MMTPA, and its terminal percentages for April–May imply more gas than the import figure it gives for the same two months.
• PNGRB should publish, in one table, the regas and truck-loading charges and booked capacity that terminals must now file twice a year, as its 2025 committee recommended.
• The Ministry of Petroleum and Natural Gas should require state-owned buyers to seek destination flexibility across Indian terminals in new long-term LNG contracts, and report cargo swaps between terminals, so capacity in one port can absorb cargoes contracted for another.
• Before Hazira’s expansion and Jafrabad are commissioned, GSPL and PNGRB should publish a funded plan to raise Gujarat’s evacuation capacity from about 40 MMSCMD toward the 130 MMSCMD of regas capacity the cluster will have.

These are proposals from this blog, not recommendations any agency has adopted, except where the PNGRB committee is cited.

Sources and caveats

About this article: Researched, written and edited by Umashankar Triplicane Dwarakanathan, with AI research assistance; every figure is meant to trace to the primary source cited. See the Editorial Policy for how sourcing, AI use and corrections work.