China's Fuel Oil Inventories in 2026: What the Trade Data Actually Shows

The number does not exist, and that is the finding

Searches for a China fuel oil inventory level return estimates, not data, because China publishes no fuel oil stock series. This is consistent with its treatment of crude, where the state releases no routine inventory reporting either, a pattern we examined in Pourquoi la Chine ne publie-t-elle pas ses réserves de pétrole ?. For fuel oil the opacity is less strategic and more structural: fuel oil is a commercial product held by refiners, traders and bonded terminal operators, and no single agency aggregates it.

What exists instead is a monthly trade record from the General Administration of Customs, and it is unusually informative because of how it is split. Imports are reported separately for ordinary trade, which is subject to import duty and consumption tax, and for bonded storage trade, which sits outside the tax net because the product is destined for international vessels. Exports are reported largely as bonded storage sales, representing low sulphur bunker fuel lifted along the Chinese coast.

That split matters. Ordinary trade imports are a proxy for refinery feedstock appetite among independent refiners in Shandong. Bonded trade is a proxy for bunkering inventory being positioned ahead of demand. A month in which ordinary trade falls and bonded trade rises is not a weak month, it is a different month, and conflating the two produces the misleading headline numbers that circulate whenever a monthly print moves sharply.

Projet 54China's fuel oil story is a bunkering story: bonded stock at Zhoushan, not strategic inventory.

A record year, then a 10 percent contraction

China's fuel oil imports totalled 21.6 million tonnes in 2025, about 376,000 barrels per day at the conventional conversion of 6.35 barrels to the tonne, down 10.4 percent from 2024's record high of more than 24 million tonnes, according to customs data published in January 2026.

The monthly shape was not a steady decline. Imports troughed at 1.30 million tonnes in May and again near 1.38 million tonnes in March, then built through the second half to 2.02 million tonnes in September, 2.15 million in November and 2.40 million in December. Within that, bonded storage trade did most of the work at the end of the year: December's 2.40 million tonne total contained 1.75 million tonnes of bonded trade against 653,000 tonnes of ordinary trade, a ratio of nearly three to one. In other words, the year ended with bunkering stock being built, not refinery feedstock.

Exports tell the complementary story. Marine fuel exports, consisting mainly of very low sulphur fuel oil, climbed 11.6 percent from a year earlier to 20.47 million tonnes in 2025. Monthly export volumes swung violently, from 1.23 million tonnes in October to 2.32 million in June and 2.24 million in December, which is characteristic of bonded bunker sales responding to quota releases and price spreads rather than to underlying consumption.

A tax decision, not a demand collapse

The driver was fiscal. An increase in the fuel oil import tax in early 2025, combined with lower tax rebates, softened demand from the independent refiners who buy imported fuel oil as an alternative feedstock when they run short of crude import quota. Emma Li, China senior market analyst at Vortexa, put it directly: China's 2025 decline in fuel oil imports was largely driven by weaker feedstock demand, amid a thinner fuel oil consumption tax rebate rate.

This is the single most important thing to understand about the series. Fuel oil import volumes in China are a tax arbitrage, not a consumption signal. When the rebate is generous and crude quotas are tight, independents switch to fuel oil and imports rise. When the rebate thins or quota is plentiful, they switch back to crude and imports fall, with no change whatsoever in how much product the country actually burns.

Li also set the 2026 expectation, and it has largely held: looking into 2026, refiners are expected to continue prioritizing crude oil over fuel oil as feedstock until quotas run short, or if the supply of certain crude grades get disrupted, such as heavy Venezuelan crude. That last clause is the conditional worth tracking. Independents run heavy, discounted barrels; if sanctioned or heavy grades become harder to source, fuel oil re-enters the feedstock slate quickly and import volumes move with it. The relationship between quota allocation and buying behaviour is the same mechanism we traced in China's crude stockpiling rate.

Volatility without a trend

The 2026 record through August is a series of sharp reversals rather than a direction. Imports fell to a record low in May, then recovered 76 percent month on month to 982,783 tonnes in June. By August imports had reached 1.3 million tonnes, up 15 percent on July but still 11 percent below August 2025, which tells you the recovery is off a very low base rather than a return to 2024 levels. Chinese refineries have continued to show limited appetite for imported fuel oil as feedstock, exactly as the tax mechanics predict.

Exports moved the other way and then reversed. June exports jumped 55 percent month on month to the highest level of 2026, helped by a price advantage: delivered 0.5 percent low sulphur marine fuel at Zhoushan and Shanghai was running around 50 US dollars per tonne below Singapore, Asia's largest bunker trading hub, which pulled bunkering demand north. Then August exports fell to 1.21 million tonnes, down 35 percent on the month and 27 percent year on year, the lowest since October 2024, with weather disruption at Zhoushan cited as a contributing factor.

Quota policy underwrote both moves. The first batch of 2026 low sulphur marine fuel export quotas totalled 8 million tonnes, with nearly 85 percent allocated to Sinopec and CNPC, and on 10 June 2026 the government added a further 5 million tonnes. Quota is the ceiling on bonded exports, so a quota release is a necessary condition for an export surge, and a quota exhaustion is sufficient to cause a collapse regardless of demand.

For anyone reading these prints as inventory signals, the discipline is to ask three questions in order. Was the move in ordinary trade or bonded trade? Was there a quota release or expiry in the window? And was there a price spread against Singapore large enough to redirect bunkering volume? Only after those three are ruled out does a monthly move say anything about stock levels.

Reading the series without being misled by it

The first implication is analytical. Anyone quoting a China fuel oil inventory number should be asked what it is derived from, because there is no primary series to cite. The defensible statements are about flows, about the bonded and ordinary split, and about Zhoushan throughput. Anything presented as a stock level is a model output, and should be labelled as an estimate, a discipline we apply across the China data set and explain in why China oil reserve estimates vary between sources.

The second is commercial. Zhoushan's rise to third place globally, ahead of Fujairah, is a structural shift in where marine fuel is sold, and it is driven by price and quota policy rather than by geography. For bunker suppliers, storage operators and trading desks, that concentrates counterparty and infrastructure exposure in a hub whose volumes can swing 35 percent in a month on quota and weather. Resilience planning should assume that volatility is the normal state, not an exception.

The third is directional. As long as crude import quotas for independent refiners remain adequate and the consumption tax rebate stays thin, fuel oil import volumes should stay structurally below the 2024 peak. The reopening condition is specific and worth monitoring: disruption to heavy or sanctioned crude supply would push independents back toward fuel oil feedstock, and the import series would respond within a quarter. That is the trigger to watch, rather than any broad measure of Chinese demand.