What percentage of Russian seaborne crude oil export loading capacity will be offline at end of June 2026 due to Ukrainian strikes (vs the H2 2025 baseline)?

TL;DR

My forecast is 32%: a Reuters/Bloomberg capacity-stocktake mode is in the low teens, and the fallback export-volume mode is near 42%.

Context

The March shock was real. On 25 March 2026, Reuters put about 2.0 mb/d, or 40%, of Russian oil export capacity offline after Ukrainian strikes suspended Primorsk and Ust-Luga loadings and left Novorossiysk behind schedule; the same report put Novorossiysk capacity near 700 kb/d (Reuters via The Moscow Times, 25 Mar 2026) . That was a point-in-time peak, not the June base case.

By late May, the visible system had mostly repaired. Bloomberg tanker tracking put Russian seaborne crude exports at 3.64 mb/d in the four weeks to 31 May and said Baltic and Black Sea export terminals that had faced Ukrainian disruption had largely resumed normal operations (Bloomberg via The Moscow Times, 2 Jun 2026) . The current June risk is different: Reuters says western-port crude loadings could fall because Russia is raising refinery runs while output is lower, while fresh Ukrainian strikes around Novorossiysk, Volgograd, and Volna keep the right tail alive (Reuters via Kitco, 8 Jun 2026) (AP, 13 Jun 2026) .

Evidence

The historical backbone is that actual Russian seaborne crude exports normally run far below the 5.0 mb/d capacity denominator. The table below is the full seaborne crude series in the IEA May 2026 Oil Market Report: units are mb/d; annual rows are calendar-year averages; monthly rows are monthly averages; the vintage is current-series IEA/Kpler/Argus data published on 13 May 2026; N = 8 comparable observations (IEA Oil Market Report, 13 May 2026) .

This matters because the fallback rule is not a clean capacity measure. A normal 2025 seaborne crude month of 3.47 mb/d would resolve to $(1 - 3.47/5.0) \times 100\% = 30.6\%$ under the fallback, before adding any June 2026 strike effect (IEA Oil Market Report, 13 May 2026) .

The port concentration is high enough that single-node outages matter. KSE, using Kpler vessel data accessed on 7 April 2026 and published on 9 April 2026, put 2025 oil-loading activity at Primorsk 1.26 mb/d, Ust-Luga 1.15 mb/d, Novorossiysk 0.97 mb/d, Nakhodka/Kozmino-area 1.00 mb/d, and Murmansk 0.30 mb/d; KSE also said the targeted Primorsk, Ust-Luga, and Novorossiysk ports accounted for about 59% of Russian seaborne oil exports (KSE Institute, 9 Apr 2026) .

The repair base rate points against a persistent March-scale physical outage. Bloomberg said weekly seaborne crude flows fell to 2.32 mb/d in the week to 29 March from 4.07 mb/d the prior week, with 22 tankers loading 16.23 million barrels versus 37 tankers loading 28.5 million barrels the week before (Bloomberg via The Moscow Times, 31 Mar 2026) . KSE then found Ust-Luga crude volumes fell from 725 kb/d in the week of 16-22 March to 101 and 155 kb/d in the next two weeks, while Primorsk crude stayed near 780-848 kb/d (KSE Institute, 9 Apr 2026) . But Carnegie estimated that Ust-Luga had resumed loading by 7 April and shipped about half of pre-strike volumes over the following two weeks, after total oil and product shipments fell to 3.5 mb/d from 25 March to 11 April versus 5.2 mb/d from 1 January to 24 March (Carnegie, 20 Apr 2026) . The IEA May report also has April seaborne crude exports at 3.76 mb/d and says Russian total exports recovered in the second half of April, though not to the three-week pre-Baltic-attack peak (IEA Oil Market Report, 13 May 2026) .

The strongest current evidence for a low physical-capacity reading is May. Reuters said western-port crude exports through Primorsk, Ust-Luga, and Novorossiysk rose 15% month on month in May to 2.5 mb/d from 2.2 mb/d in April, the highest since September 2025, despite ongoing Novorossiysk attacks that briefly suspended loadings (Reuters via MarketScreener, 2 Jun 2026) . CREA's May bulletin, published 11 June 2026, said crude export volumes rose 8% month on month and Ust-Luga crude loadings rose 49% month on month after the March-April strike-related reduction (CREA, 11 Jun 2026) . Bloomberg also reported that all crude berths at Novorossiysk were back in operation by mid-May for the first time since early April (Bloomberg via BusinessMirror, 20 May 2026) .

The strongest current evidence for a high fallback reading is the 8 June Reuters loading program. Reuters said June crude loadings from the three western ports could fall to 1.7 mb/d from 2.5 mb/d in May, with Russia seeking to raise crude runs by 250-400 kb/d while lower output limits feedstock (Reuters via Kitco, 8 Jun 2026) . Pairing that 1.7 mb/d western-port figure with roughly 1.1-1.2 mb/d of non-western seaborne crude implied by Bloomberg's 3.64 mb/d total and Reuters's 2.5 mb/d western figure gives a June seaborne crude center near 2.9 mb/d, so the fallback value centers near $(1 - 2.9/5.0) \times 100\% = 42\%$ (Bloomberg via The Moscow Times, 2 Jun 2026) .

Fresh strikes add upside, mostly through the Black Sea chain, but they are not yet quantified as multi-week mb/d losses. Ukraine's General Staff said the 7-8 June strike on the Grushovaya oil transshipment base near Novorossiysk caused fires in two oil storage tanks and quality-assessment infrastructure, and AP reported on 13 June that drone debris sparked a fire at a Black Sea export terminal in Volna that Russian outlets described as transporting crude oil, petroleum products, and LPG (Ukrainska Pravda, 10 Jun 2026) (AP, 13 Jun 2026) . Those reports raise the high tail, but they do not yet show a renewed shutdown of Primorsk, Ust-Luga, Novorossiysk, or Kozmino on the March scale.

My calculation uses a resolution-source mixture. I put 35% on a qualifying Reuters/Bloomberg capacity stocktake and center that path at 16%, because the main crude-loading system had recovered enough to support high May flows while recurring Black Sea and pipeline strikes leave moderate residual damage risk. I put 65% on the fallback export-volume path and center that path near 40-43%, using the 8 June Reuters June-loading program and a 5.0 mb/d denominator. The final distribution has mean 32.1%, median 36.2%, and a 10th-90th percentile interval of 8.2-47.9%. It puts 13% below 10%, 16% from 10-25%, 35% from 25-40%, 31% from 40-50%, and 6% above 50%.

What's non-obvious

The question can resolve high even if Russia's ports are mostly usable. The fallback formula converts any gap between June seaborne crude exports and 5.0 mb/d into “offline capacity,” even if the gap comes from lower output, refinery feedstock choices, OPEC+ behavior, tanker scheduling, sanctions friction, or buyer timing. The 2025 IEA seaborne crude average alone maps to 30.6% on that formula (IEA Oil Market Report, 13 May 2026) .

The second trap is that Ukrainian refinery strikes can increase crude exports before they reduce them. May's high crude exports came partly because refinery outages freed crude for shipment, while June's expected export drop is tied to refineries trying to raise throughput amid fuel shortages and lower crude output (Reuters via MarketScreener, 2 Jun 2026) (Reuters via Kitco, 8 Jun 2026) . That makes low June exports a weak proxy for damaged loading capacity.

Limitations

The largest gap is direct terminal-by-terminal crude-loading capacity during 1-28 June. Public sources give tanker-flow aggregates, scheduled loadings, strike reports, and some satellite interpretation, but not a clean daily table of usable mb/d at Primorsk, Ust-Luga, Novorossiysk/Sheskharis, Kozmino, Murmansk, and minor terminals. The 8 June Reuters article is a loading-program and feedstock story, not a capacity-offline stocktake (Reuters via Kitco, 8 Jun 2026) .

The resolution-source risk is large. A Reuters/Bloomberg capacity stocktake would probably exclude much of the export-volume weakness caused by refinery routing and output cuts. The fallback rule would include it mechanically. The same real-world June could therefore resolve near 10-20% under a physical-capacity stocktake and near 40-45% under the fallback formula.

The remaining half of the window can still move the result. A direct late-June hit on a Baltic terminal, Sheskharis, Kozmino, a feeder pipeline, or tankers at the port would matter more than the historical averages. I do not treat the absence of a public Kozmino outage report as proof that no Kozmino risk exists; it only keeps the current central case concentrated in western-port and Black Sea disruptions.