26MAY

BP Rotterdam dark, ARA stocks slide

4 min read

08:52UTC

Both crude units at BP's 400,000 barrels per day Rotterdam refinery ran offline through May 2026 while ARA gasoil stocks fell to 13.56 million barrels, the lowest since July 2025.

← Back to GL 134C reverses the cliff, Brent -$14 Jump to analysis ↓

EconomicDeveloping

Key takeaway

BP Rotterdam at zero plus ARA gasoil at multi-year lows binds NWE refining margins into June.

BP's Rotterdam refinery (400,000 barrels per day, 19 million tonnes per year) had both crude units offline through May 2026: the first 200kbd unit on planned maintenance from early May, the second 200kbd unit shut concurrently ¹. That is one of Europe's largest sites running at zero crude throughput. Global composite refining margins printed the highest level in more than a year during the same period.

The stocks behind the refinery tell the second half of the story. ARA gasoil stocks sat at 13.56 million barrels in early May, down from April's 14.53 million barrels and the lowest level since July 2025 ². ARA jet hit a six-year low of roughly 600,000 tonnes on the 15 April baseline, naphtha was off 13.9 per cent on the week, and total ARA products fell 4 per cent. Mediterranean middle-distillate imports ran at a dataset-high 1.9 million barrels per day in May, the third consecutive month above the 10-year seasonal peak ³. ARA fuel oil was off 4 per cent month-on-month at 4.19 million barrels.

EIA's weekly to 8 May put US distillate stocks 6 per cent below the 5-year average at a 10.37 million barrels deficit; the gasoline draw came in at 4.1 million barrels against a 2.9 million barrel expectation ⁴. Both sides of the Atlantic basin ran simultaneously tight in the same fortnight. The IEA's May OMR recorded OECD on-land stocks falling 146 million barrels in April alone, with 2Q26 crude throughputs projected down 4.5 million barrels per day to 78.7 million barrels per day.

Composite refining margins at year-plus highs with stocks falling is the textbook signal of capacity binding rather than demand surging. The marginal-barrel refiner setting the European product price is running at maximum throughput on full crude inputs; everyone behind that refiner on the cost curve is making outsize money while everyone ahead is on planned or unplanned downtime. The constraint is rebuild capacity, not feedstock availability: until ARA stocks climb back above 15 million barrels or a Russian product export surge appears, the NWE gasoil crack has a structural anchor.

Deep Analysis

In plain English

Rotterdam is Europe's biggest oil refining hub. BP operates a 400,000-barrel-per-day refinery there, roughly the daily oil needs of a mid-sized European country. In May 2026, both of its crude-processing units were shut down for maintenance at the same time. This matters because refineries turn crude oil into petrol, diesel, and jet fuel. With BP Rotterdam not processing any crude oil, and ARA storage tanks for diesel and jet fuel running at their lowest levels in years, Europe's fuel buffer is thinner than usual. Other sources, mainly imports from the Mediterranean, are filling the gap, but the margin for error is small.

Deep Analysis

Root Causes

BP Rotterdam's dual-unit shutdown is not a supply shock in the force-majeure sense, as both outages are scheduled maintenance. The distillate market impact comes from the convergence of three independent factors in the same two-week window.

First, BP Rotterdam planned maintenance timed before summer turnaround season. Second, Shell Pernis on a separate turnaround running concurrently. Third, ARA inventory already depleted by the Iran-conflict-driven supply shock to Mediterranean sour crude flows since late February 2026.

The Hormuz closure compounded NWE refinery economics in a second way. Higher crude acquisition costs for Brent-linked spot cargoes versus alternatives squeezed run economics for refiners buying spot, creating incentives to bring forward turnarounds while margins were compressed. BP's timing may have been deliberately front-loaded to capture the post-Iran-conflict margin rebound on restart.

What could happen next?

Consequence
ICE Gasoil crack spreads held at year-plus highs as BP Rotterdam's 400kbd runs zero crude throughput; the crack provides a structural anchor for any NWE refining-margin long position.
Immediate · 0.85
Risk
If BP Rotterdam's second crude unit pushes into Q3 rather than Q2 restart, ARA gasoil stocks cannot rebuild ahead of the July summer driving peak, increasing the risk of a physical spot-market squeeze.
Short term · 0.7
Opportunity
Med distillate imports at dataset-high 1.9mb/d confirm arbitrage is working; TC14 (USGC to UKC) economics support transatlantic distillate flows that could begin replenishing ARA if US distillate exports increase.
Short term · 0.65
Consequence
European airlines and logistics companies face elevated jet fuel and diesel procurement costs through at minimum Q2 2026, with hedging desks exposed to above-trend crack-spread levels.
Medium term · 0.75

Source Landscape

This story draws on neutral-leaning sources from United Kingdom

United Kingdom×2

LeftRight

Primary parallel: The Shell Pernis catalytic-cracker fire of 2008 is the closest structural precedent. Shell's Pernis site lost 270kbd for approximately six weeks in 2008; ARA gasoil drew from 16 million barrels to under 12 million barrels in the same period, and ICE Gasoil M1 spiked roughly $80 per tonne relative to the forward curve.

The BP Rotterdam dual-unit shutdown removes a comparable volume (400kbd) for a comparable duration, but lands on an ARA inventory baseline (13.56mb gasoil) already lower than the 2008 pre-fire level of approximately 14.5mb.

Counter-parallel: Hurricane Harvey in August 2017 shut roughly 2.3mb/d of US Gulf Coast refining capacity for three weeks. Despite the scale of the US outage, ARA stocks built during the Harvey period as European refiners increased runs to capture the arbitrage. The key difference: Harvey displaced US demand to European imports; the BP Rotterdam shutdown displaces NWE supply with no equivalent offset.

ICE Gasoil M1 crack spreads versus Brent dated have printed at the highest level in over a year during May 2026. The specific per-tonne crack level was not retrieved in this window (flagged for U#2), but directionally the margin breakout from the multi-year refining margin composite high confirms that the throughput removal is flowing directly into cracking economics.

For a European airline buying jet fuel, the six-year low on ARA jet stocks (roughly 600kt at 15 April baseline) signals a structural input-cost squeeze: the jet-gas oil spread has widened, adding to airline fuel hedging costs at a moment when summer capacity is fully committed.

For a diesel-dependent logistics company, ARA gasoil at 13.56mb versus the July 2025 low provides a narrow inventory buffer, and any further demand upside or supply disruption in May-June hits the physical basis immediately.

Consensus view: Energy Aspects analyst Amrita Sen argues that simultaneous planned maintenance at BP Rotterdam and Shell Pernis represents the worst possible timing from an ARA inventory standpoint.

Both sites pull crude throughput in May when Med sour import cargoes are already diverted to Cape of Good Hope rerouting, lengthening voyage times and slowing ARA replenishment. The refining margin breakout to year-plus highs is a direct function of the throughput gap, not a structural profitability shift.

Counter-view: FuelsEurope Director General John Cooper argues the ARA stock draw is real but overstated as a supply-security concern. European distillate import logistics can flex on a six-to-eight-week lag, and the 1.9mb/d Med distillate import rate, already at a dataset high, shows that supply chains are responding to the draw signal. The risk is a two-to-three-week tightness window before the import response lands, not a structural deficit.

Key tension: Whether the BP Rotterdam second crude unit returns inside Q2 2026 or slips into Q3, which determines whether ARA gasoil stocks can begin rebuilding before the summer driving demand peak or must run through June on import cover alone.

Sources:Argus Media / Insights Global·Argus Media / Insights Global·Reuters / OilPrice.com

Mentions:APS →ETS →Argus Media →Cape of Good Hope →OECD →Mediterranean →Insights Global →EIA →Vortexa →Shell Pernis Refinery →Shell →ICE Gasoil →International Energy Agency →Rotterdam →

First Reported In

Update #1 · GL 134B out, Rotterdam dark, OPEC+ pending

Argus Media / Insights Global· 18 May 2026

Read original →

Causes and effects

This Event

BP Rotterdam dark, ARA stocks slide

One of Europe's largest sites went to zero throughput inside the same fortnight Atlantic-basin distillate stocks ran tight on both sides of the ocean.

Led to

EU gasoil imports crash to 695kbd

BP Rotterdam's both-units-dark position compounded the EU gasoil import collapse; domestic cracking capacity withdrew exactly as the 695kbd import shortfall opened.

Occurred 15 May 2026

Read story →

Different Perspectives

Indian / Asian refinery buyers

The Adani $275m OFAC settlement for 32 Iran-LPG violations, posted 18 May, recalibrated the compliance-cost calculus for every Indian buyer holding Russian cargoes loaded under the lapsed GL 134B; GL 134C restores cover but the Cuba carve-out and the Cuba-tainted cargo class force per-voyage due diligence on the full logistics chain.

Shell / TotalEnergies NWE refining

With BP Rotterdam's 400kbd dark on both crude units and the ICE Gasoil crack near $54/bbl as Brent fell $14, NWE refiners running full crude capture a crack-to-crude ratio of roughly 56%, well above the 30-35% historical norm; every barrel cracked into gasoil on non-Hormuz feedstock earns extraordinary margins.

VLCC owner / Baltic Exchange freight desk

The BDTI at 2,249 on 20 May is still pricing a war the market no longer fully believes; GL 134C removes the compliance bid from Baltic Aframax TD7 and TD19 ahead of any VLCC print, because owners reprice forced-rerouting premiums faster than they reprice an all-time-high composite index.

Goldman Sachs / Energy Aspects sell-side macro

The Brent-Dubai EFS narrowing from above $6/bbl confirms the light-sweet war premium is deflating, not dead; the 30-60 day MOU window means the $14 Brent decline has priced a scenario where Hormuz is functionally open by July, leaving the flat price exposed to a re-spike if mine clearance stalls.

EU Council sanctions directorate

The 20th package's maritime-services ban deferral, contingent on G7 coordination at Kananaskis, reflects Hungary, Slovakia and Austria wielding the unanimity veto to block a measure that would raise NWE seaborne costs for states whose Russian crude arrives by pipeline and faces no freight exposure.

Rosneft / Russian export ministry

Russian export revenue at $19.0bn in March on Urals FOB ~$76/bbl, $28 above the G7 $47.60 cap, confirms the cap has no effective bite at current flat price; the shadow fleet's Russian-flag share rising to 21% shows Moscow absorbed Western vessel-services constraints by re-flagging out of P&I reach.