What 67% actually measures, and why it is not 73%

On 7 August 2026, India’s Petroleum and Natural Gas Minister told reporters that “the United States accounts for something like 67 per cent of our LPG imports.” In the narrow sense of supply security it is a genuine achievement: India has 332 million active domestic gas connections and has not run a national shortage through five months of the worst chokepoint crisis in modern energy trade. But the figure deserves more precision than a podium allows.

Two numbers are circulating and they measure different things. The ministry’s 67% is a broad characterisation of the current mix, with no published period or basis. Kpler, which builds import shares from satellite and AIS vessel tracking rather than customs filings, put the US above 73% in July specifically, on a record 0.91 million tonnes. Both are defensible. A monthly share taken at the peak always runs ahead of a multi-month average, and July was the peak. Use 67% for the sustained picture and 73% for July alone.

The direction is not disputed by anyone. In January the US supplied about 12%. In February, before the war, it was roughly 14% against a Gulf bloc holding 81%. By May the US was at 55% and the Gulf had fallen to 16%. By July the US was above 73%, the UAE had dropped to about 102,000 tonnes from more than 626,000 in February, Qatari volumes had fallen sharply, and Saudi Arabia had shipped India no LPG whatsoever. A trade relationship built over decades unwound in five months.

How the Gulf fell out of India’s import basket

The mechanics matter more than the geopolitics, because they explain why the shift was so total. India consumes roughly 31 million tonnes of LPG a year and imports about 65% of it, making it the world’s second-largest LPG importer. In 2024, about 90% of the 20.4 million tonnes imported came from four countries. More importantly, as the Minister of State told the Rajya Sabha in March, close to 90% of that imported LPG physically transited the Strait of Hormuz. India was not simply buying from the Gulf. It was buying through a single waterway 21 miles wide at its narrowest.

That waterway closed to normal commercial traffic on 28 February 2026. A ceasefire framework in April and a memorandum of understanding in mid-June briefly restored toll-free transits from around 17 June, but volumes never approached normal and the arrangement collapsed in early July after attacks on commercial vessels. Sourcing 90% of a staple household fuel through a corridor running at a fraction of capacity is not a diversification problem. It is a single-point-of-failure problem.

The deal that was only ever supposed to be 10%

The most misread part of this story is the timing. India did not turn to American LPG because of the Strait of Hormuz. The structured buying relationship predates the war by three months, and it existed for an entirely different reason.

In November 2025, Indian Oil, Bharat Petroleum and Hindustan Petroleum concluded a one-year contract for about 2.2 million tonnes of LPG from the US Gulf Coast for calendar 2026, priced off the Mount Belvieu benchmark and supplied by Chevron, Phillips and TotalEnergies Trading. It was the first structured US LPG contract in the Indian market, worth roughly 48 very large gas carrier cargoes, and it was widely read as a trade-balance gesture towards Washington during a tariff dispute. The volume equalled about 10% of annual imports. Analysts questioned the logic of hauling molecules 9,000 miles when Ras Tanura was four sailing days away.

Four months later that contract was the only fully functioning supply line India had. The minister has since been explicit that the decision was taken independently and not driven by Hormuz risk, which is worth taking at face value, since the negotiating team travelled to the US through the second half of 2025. India bought insurance for a commercial reason and then needed it for a strategic one.

The bill nobody budgeted for

This is where substitution stops being free. Gulf to India is one of the shortest major LPG routes in the world. US Gulf Coast to India is one of the longest, and in 2026 it got longer still.

The standard route from Houston to Asia runs through the Panama Canal, about 9,370 miles. As Gulf loadings collapsed and the global fleet repositioned westward, Panama’s Neopanamax auction slots became a bidding war: transit costs reached $1.076 million on 29 April, close to four times pre-conflict levels, with waiting times doubling. Operators did the arithmetic and sailed around Africa instead. The Cape of Good Hope route covers about 15,900 miles, adds 20 or more days each way, and by April accounted for roughly 43% of US-to-Asia LPG voyages, a share unmatched since 2016.

The freight market responded exactly as it should when tonne-miles explode and vessel availability shrinks. A record 150 very large gas carriers repositioned toward the US Gulf Coast by mid-March, and for weeks ballast vessels outnumbered laden ones, a dislocation larger than the one at the start of the pandemic. About a tenth of the global VLGC fleet spent part of the spring waiting outside the Middle East Gulf, sitting off western India, or trapped inside the Gulf entirely. US LPG exports hit a dataset record of 2.8 million barrels a day in April.

What it costs at the kitchen door, and who actually pays

Indian households have felt almost none of this directly, which is itself the story. A 14.2 kg domestic cylinder in Delhi costs Rs 942. The economic cost of filling it, at post-crisis contract prices and post-crisis freight, is above Rs 1,600. The gap does not disappear. It moves.

Most of it sits on the balance sheets of the three state oil marketing companies, which the petroleum ministry confirms are absorbing an under-recovery of about Rs 700 per cylinder. Cumulative domestic LPG under-recovery reached Rs 60,000 crore by the end of the last financial year, up from Rs 41,338 crore the year before, against which the Cabinet approved Rs 30,000 crore in compensation. On petrol and diesel the same companies were separately running losses of Rs 600 to 700 crore a day.

The rest has been passed to the poorest users by the quiet route. In June the government cut subsidised refills under the Pradhan Mantri Ujjwala Yojana from nine a year to four, keeping the Rs 300 per cylinder rate intact but shrinking the annual benefit from Rs 2,700 to Rs 1,200. For a household that took its four cylinders in April, May, June and July, the concession ended this month. Retail price rose Rs 29, which the ministry accurately described as about one rupee a day, though that framing sits alongside an entitlement that fell by more than half.

Supply itself held up better than the finances. Refinery LPG output was pushed from about 35,000 tonnes a day before the war to roughly 53,000, against demand near 72,000. Distributor backlogs stayed under four days and the average delivery cycle held near two and a half days even through periods of panic booking. The pressure was absorbed elsewhere: a Natural Gas Control Order was issued on 9 March under the Essential Commodities Act, prioritising piped domestic gas and CNG while moderating industrial supply, and commercial LPG for restaurants and small businesses was still running at only 70 to 75% of pre-crisis volumes in June. Households were protected. Commercial users were not.

Why 67% may not survive the autumn

Every part of this arrangement is contingent on the strait staying shut, and as of this week it may not be.

On 5 August, Iran said it had reached agreement with Oman on a proposed shipping route through the Strait of Hormuz, with a joint statement in final drafting and the Iranian parliament reviewing terms. The structure would route vessels into the Gulf via Iranian waters and out via Omani waters. US and regional officials have signalled a deal is close, and Brent fell below $80 on the news before settling near $82.

If that holds, the economics reverse quickly and unsentimentally. Gulf cargoes reach western India in days rather than weeks, and Middle East freight has run at roughly half the American rate throughout. Saudi and UAE producers have idle export capacity and every incentive to reclaim a 20 million tonne a year buyer. The 2.2 million tonne American contract will be honoured because it is contracted. The spot volumes stacked on top of it, which are the bulk of the current 67%, have no such protection.

The more durable questions are structural. India has been steadily raising the share of energy that does not transit Hormuz at all, and the minister used the same appearance to promote the Rs 84,000 crore Samudra Manthan deep-sea exploration mission. Import terminals, coastal storage and long-term offtake are decisions with ten-year horizons that have just been handed five months of extremely persuasive evidence. A reopened strait restores the cheap route. It does not restore the assumption that the cheap route will always be there.

Every figure is attributed to a named primary source and dated. Import shares and cargo volumes come from Kpler vessel-tracking data as reported by Business Standard, ThePrint and CNBC. Freight rates come from the Baltic Exchange, with routing and canal cost data from Argus, Vortexa and S&P Global Commodity Insights. Domestic price, subsidy and under-recovery figures come from the Ministry of Petroleum and Natural Gas, PIB releases and Rajya Sabha replies. Transit counts come from IMF PortWatch.

Where sources disagree, we say so rather than choosing the cleaner number. Two Kpler-derived figures for May US volumes have been published, 630 kt and 666 kt. The ministry’s 67% and Kpler’s 73% are different measurements, not a correction of one another. The April share in Figure 1 is our own calculation from supplier volumes and is flagged as such. This is a live conflict and figures move weekly. Last verified 8 August 2026.