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

I forecast 21% of Russia's H2 2025 seaborne crude export loading capacity will resolve as offline during 1-28 June 2026.

Context

The March outage was large, but it was a peak shock rather than a stable new level. Reuters reported on 25 March 2026 that Ukrainian strikes, pipeline damage, and tanker seizures had halted about 2.0 mb/d, or 40%, of Russian oil export capacity, with Primorsk, Ust-Luga, and Novorossiysk all affected (Reuters via The Moscow Times, 25 Mar 2026). By April, Russia had restored enough loading and rerouting capacity that Reuters estimated western-port crude loadings at about 2.2 mb/d for April, broadly in line with revised March, while Ust-Luga resumed on 7 April and Novorossiysk partially resumed on 9 April after a four-day halt (Reuters via Business Recorder, 30 Apr 2026).

The campaign did not stop. Ukraine hit Primorsk again on 3 May 2026 and said it also struck shadow-fleet tankers near Novorossiysk; Russia acknowledged a fire at Primorsk and said there was no oil spill (Reuters via StreetInsider, 3 May 2026, AP, 3 May 2026). Ukraine then said it struck the Tamanneftegas terminal, Astrakhan gas plant, and Yaroslavl refinery on 13 May 2026, causing fires at reservoirs and hitting primary refinery units (Reuters via Internazionale, 13 May 2026). So the central case is partial repair plus repeated harassment, not full restoration and not a permanent 40% outage.

Evidence

The base rate from this episode is that crude-loading stoppages have lasted days to two weeks unless renewed strikes arrive. Bloomberg tanker tracking put Russian seaborne crude flows at 2.32 mb/d in the week to 29 March 2026, down from 4.07 mb/d the prior week, showing the acute hit was real and operationally visible (Bloomberg via The Moscow Times, 31 Mar 2026). But the April IEA report, using Kpler and Argus data, said Primorsk resumed loadings two days after the attacks and averaged 1.4 mb/d in the week of 6 April, while Ust-Luga was still about 300 kb/d below 2025 levels; the same report said the March-April strikes reduced exports by 450 kb/d in the first week of April versus the March average, not by the full 2.0 mb/d peak (IEA Oil Market Report, 14 Apr 2026, pp. 20-21).

The full comparable export history matters because the fallback rule uses exports divided by 5.0 mb/d. These are crude exports in mb/d, current-vintage IEA/Kpler/Argus estimates from the April 2026 Oil Market Report; the sample is 8 annual/monthly observations covering 2022 through March 2026 (IEA/Kpler/Argus table, Apr 2026 vintage).

That table is the reason I do not put the forecast near 8%, even though physical capacity damage may be near that level by late June. If no Reuters or Bloomberg capacity stocktake appears by 14 July, the fallback formula is $100 \times (1 - \text{June seaborne crude exports}/5.0)$, floored at zero. A normal-looking 3.6 mb/d seaborne month resolves near 28%. Bloomberg's latest public flow datapoint I found was a 3.64 mb/d four-week average to 10 May 2026, down only slightly from 3.68 mb/d to 3 May, with Novorossiysk temporarily weak because of drone threats and bad weather and no visible satellite damage to the terminal (Bloomberg via UATV, 12 May 2026).

Terminal-level data point to impairment, not paralysis. KSE's April 2026 assessment, using Kpler and Argus data accessed on 7 April, put the 2025 seaborne oil exposure at Primorsk 22%, Ust-Luga 20%, and Novorossiysk 17% of Russian seaborne exports; those three ports account for about 59% of the target set (KSE Institute, 9 Apr 2026). In KSE's weekly Baltic series, Ust-Luga crude fell from 725 kb/d in the week of 16-22 March to 101 kb/d and 155 kb/d in the next two weeks, while Primorsk crude held around 780-848 kb/d; I read this as a severe Ust-Luga bottleneck plus storage and scheduling damage, not a full Baltic crude-loading collapse (KSE Drone Strike Impact Assessment, 9 Apr 2026). Reuters then reported that April loadings from Primorsk, Ust-Luga, and Novorossiysk were about 2.2 mb/d and that May could increase if external disruptions did not recur (Reuters via Business Recorder, 30 Apr 2026).

The latest macro oil data also cuts against a late-March-repeat baseline. S&P Global reported Russian crude exports of 4.1 mb/d in April 2026, up from about 3.4 mb/d in March in its Commodities at Sea and Maritime Intelligence Risk Suite data, with G7-linked tankers lifting 29.4% of April volumes (S&P Global, 11 May 2026). The IEA said Russian crude exports rose 250 kb/d month on month to 4.9 mb/d in April, driven partly by higher Baltic seaborne shipments and partial Druzhba resumption, while Russian crude output fell 460 kb/d year on year to about 8.8 mb/d (Reuters/IEA via MarketScreener, 13 May 2026, IEA Oil Market Report, 13 May 2026). High crude exports do not prove the ports are healthy, because damaged refineries can push crude into export channels, but they do show that March's system-wide loading shock had not persisted through April.

My distribution uses a 5.0 mb/d denominator and mixes five states: 24% on near-repair centered at 6.5%, 31% on moderate physical impairment centered at 15.5%, 27% on a fallback/current-flow case centered at 27%, 14% on renewed successful June strikes centered at 38%, and 4% on severe multi-node disruption centered at 54%. The mean is 21%, the median is 20%, and the 10th-90th percentile range is about 5.6%-39%. It puts 24% below 10%, 40% in 10%-25%, 27% in 25%-40%, and 9% above 40%.

What's non-obvious

The prompt's 40% anchor is a poor June point estimate. It was a peak stocktake during active fires, suspensions, queues, and pipeline-feed disruption. The best later evidence shows fast partial recovery: Primorsk resumed quickly, Ust-Luga came back after roughly two weeks, Novorossiysk restarted after a short halt, and total seaborne flows were back near 3.6 mb/d by early May (IEA Oil Market Report, 14 Apr 2026, Reuters via Business Recorder, 30 Apr 2026, Bloomberg via UATV, 12 May 2026).

The opposite trap is to forecast only true engineering capacity. The fallback can convert normal underuse of a 5.0 mb/d capacity base into a 25%-30% resolution even without much physical outage. I therefore forecast the resolution value above my best estimate of durable terminal hardware still offline at the end of June. This is a measurement-risk question as much as an oil-logistics question.

Limitations

Public reporting still does not give a clean berth-by-berth crude-loading capacity ledger for Primorsk, Ust-Luga, Novorossiysk, Kozmino, and smaller Russian crude terminals as of 17 May 2026. The best sources mix capacity, crude flows, petroleum products, Kazakh transit, damaged storage tanks, and tanker scheduling. That makes attribution hard.

The second gap is the resolver path. A Reuters or Bloomberg stocktake may count only strike-disabled capacity, which would pull the result lower. If no such stocktake appears, the CREA/Kpler fallback may count weak exports caused by weather, sanctions, OPEC+ behavior, maintenance, tanker availability, or buyer timing as if they were strike-driven capacity outages. That pushes the distribution upward.

The third gap is path dependence. A successful strike on Primorsk, Ust-Luga, or Novorossiysk during 1-28 June could dominate the monthly average. A quiet June, a ceasefire around energy infrastructure, or better air defenses around the export ports could push the result into the single digits. The current evidence supports a middle outcome: repair keeps the system working, but Ukraine keeps enough pressure on the western ports to prevent a clean return to normal.