Summary:
- More than 96,000 route diversions have been recorded across 16 weeks. Week 16 shows the sharpest single-week decline in the dataset at 29%, suggesting the network is entering a new phase of stabilization, though diversion volumes still run well above pre-closure baseline.
- Jebel Ali lost the highest shipments in the dataset, with 15,944 shipments originally bound for Jebel Ali diverted to Khawr Fakkan, Fujairah, Salalah, Navi Mumbai, and Jeddah. Khawr Fakkan has absorbed the largest share, receiving 14,732 diverted shipments across all origin ports while simultaneously losing 4,548 of its own to other destinations.
- Navi Mumbai import dwell has risen every single week without exception, from 4.88 days in Week 1 to 19.85 days in Week 16. It is absorbing second-order congestion from being the landing point for Jebel Ali overflow.
- Port Elizabeth is the only Cape of Good Hope port with genuinely sustained elevation, averaging nearly 2 times its pre-conflict weekly vessel capacity across 16 weeks. Durban is moderately elevated but dipped below baseline in 6 of 16 weeks. Cape Town and Walvis Bay show period averages inflated by single-week spikes and are at or below pre-conflict levels by Week 16.
Overview
Sixteen weeks after the closure of the Strait of Hormuz, the container shipping network shows the first credible signs of stabilization at the routing level while congestion at key ports continues to deepen. The two trends are related: as fewer new diversions are issued each week, the accumulated backlog at ports absorbing rerouted cargo has nowhere to clear to.
Total diversions across the period reached 96,200 shipments. The pattern was not a simple decay from an initial spike. There were three distinct waves: an acute disruption phase in Weeks 1 through 5 averaging above 8,700 shipments per week, a mid-period of volatile but lower volumes from Weeks 6 through 13, and a secondary elevation in Weeks 14 and 15 before the sharp Week 16 decline to 3,836 shipments.
The diversion data reveals a clear hierarchy of impact. UAE ports, particularly Jebel Ali and Abu Dhabi, lost the most cargo. Khawr Fakkan and Fujairah, both on the Gulf of Oman side of the UAE coastline, became the primary absorbers. Saudi Arabia rerouted Dammam traffic through Jeddah in the acute phase, then Jeddah itself began cascading overflow to King Abdullah Port as a second-order consequence. India, specifically Navi Mumbai, absorbed a significant share of Jebel Ali overflow early on and is still paying the congestion price for it. The Cape of Good Hope route, meanwhile, has matured into a reliable structural alternative, with South African ports sustaining above-baseline throughput across the entire period.
Diversions across 16 weeks
The weekly diversion chart exposes a structure that a simple total count obscures. The highest single-day volumes occurred in the first two weeks, but the network did not then recover steadily. Week 4 rebounded 28% week-on-week after three weeks of decline. Week 8 surged 25% after two weeks of easing. Weeks 14 and 15 returned nearly identical elevated volumes of around 5,400 shipments each before the Week 16 decline.
Each re-elevation corresponds to a different carrier response: the initial emergency rerouting, a second wave as early capacity constraints at alternative ports forced further adjustments, and the late wave likely reflecting end-of-quarter cargo repositioning and further service string redesigns. Week 16’s 29% decline is the sharpest single-week drop in the dataset and could mark a genuine inflection, though it follows the pattern of previous false stabilizations.
The cumulative total of 96,200 diversions across 16 weeks has created a structural overhang of cargo sitting at ports not designed to be the final destination for these volumes. That overhang is what the dwell time data reflects.
The ports that absorbed diverted cargo
The geography of diversion follows a clear logic. Carriers needed ports outside the Strait, accessible quickly, with enough capacity to absorb rerouted volume. That pulled traffic toward two types of destinations: UAE ports on the Gulf of Oman side of the country, which sit just outside the Strait and require no additional transit distance, and ports further afield in India, Sri Lanka, and Southeast Asia that became viable as the disruption extended from weeks into months. The result is a network of substitutions that mirrors the original trade structure of the Gulf, with Khawr Fakkan emerging as the primary near-term fallback, and South Asian ports absorbing the longer-term structural overhang.
Jebel Ali accounted for 32% of all destination changes during the disruption, meaning shipments originally bound for Jebel Ali rerouted elsewhere. Of those shipments, 46% went to Khawr Fakkan, 17% to Fujairah, 5% to Navi Mumbai, 4% to Jeddah, and 3% to Salalah. The remaining quarter spread across destinations as far as Singapore (2%), Shanghai (1%), and Trivandrum (1%). Nearly half of Jebel Ali’s lost traffic stayed within the UAE, redirected to its Gulf of Oman coast. The other half dispersed across a wide arc, confirming that no single port was able to absorb the full volume.
Khawr Fakkan received 31% of all newly assigned final destinations across the period, the highest share of any port. Its inbound mix reflects the breadth of the disruption: 47% of its new arrivals came from Jebel Ali, but the remaining 53% came from other ports around the Strait — Fujairah (15%), Abu Dhabi (9%), Dammam (4%), Hamad/Doha (4%), Ajman (4%), and Jeddah (3%) among others. Khawr Fakkan is not absorbing overflow from one port. It is the collective alternative destination for the entire Gulf. At the same time, 15% of all destination changes involved shipments being redirected away from Khawr Fakkan to other ports, as congestion mounted and carriers sought alternatives further afield.
Beyond Khawr Fakkan, the other net gainers are geographically instructive. Navi Mumbai received 3% of all destination changes, Singapore 2%, Hambantota 2%, and Mundra 1%. These are not Gulf-adjacent ports. Cargo that was bound for the Gulf is now being discharged in South Asia and Southeast Asia, at ports that serve entirely different inland markets. This is not rerouting within the same trade lane. It is a structural reallocation.
The transshipment picture adds a further dimension. Khawr Fakkan and Navi Mumbai are simultaneously absorbing more final-destination cargo and more relay traffic redirected through the TS fields. The two pressures are compounding, which is why dwell times at both ports are rising faster than at ports that absorbed only one type of demand shift.
Port congestion remains elevated in key ports
Diversion volumes are falling. Port congestion is not. The two metrics are on diverging trajectories because the dwell time data reflects accumulated overhang, not the current daily flow. Every week of elevated diversions adds to the backlog at receiving ports, and that backlog cannot clear until ship calls normalize and handling capacity catches up.
Navi Mumbai stands apart from every other port in this dataset. It has recorded a higher import dwell figure every single week for 16 consecutive weeks. The trajectory from 4.9 days in Week 1 to 19.9 days in Week 16 shows no plateau, no stabilization, and no reversal. This is not a port managing its way through a surge. It is a port whose congestion is still accelerating more than four months after the disruption began.
When Jebel Ali overflow began arriving in Weeks 1 through 8, it absorbed 1,475 shipments that the port was not configured to process at that velocity. Import dwell reflects the time from arrival to gate-out; rising dwell means cargo is sitting longer before it can be collected or onforwarded, which is a direct indicator of yard saturation and operational strain.
Salalah shows the opposite trajectory. It spiked to 30.1 days in Week 1 and continued climbing to a peak of 36.6 days in Week 5 as the closest large-capacity alternative to the Strait absorbed the initial surge. But it has declined steadily since, reaching 17.5 days in Week 16 and continuing to fall. Salalah had the infrastructure and experience to absorb the surge and work through it. Navi Mumbai did not.
Singapore import dwell has risen from 2.9 days pre-conflict to 9.2 days in Week 16, a 218% increase. Unlike Navi Mumbai, Singapore’s trajectory shows more variation week-to-week, reflecting the port’s greater operational flexibility and capacity depth. But the sustained above-baseline position confirms that transshipment cargo displaced from the Gulf is landing in Singapore and creating persistent demand above the port’s pre-disruption equilibrium.
Jeddah, elevated throughout at around 12 days versus a pre-conflict baseline of 4.6 days, reflects its role as the primary Saudi alternative gateway. Its dwell has not deteriorated to Navi Mumbai levels because Jeddah has deeper infrastructure for large-volume import handling, but it is running at 163% above baseline with no clear path to normalization as long as Gulf alternatives remain unavailable.
Cape of Good Hope vessel activity since the closure
Activity at Cape of Good Hope ports has increased since the closure, but the degree of elevation varies significantly by port. Port Elizabeth is the only port showing genuinely sustained above-baseline vessel capacity across most of the 16-week period. Durban is moderately elevated on average but inconsistent week to week. Cape Town and Walvis Bay show period averages that are inflated by isolated spikes and are at or below pre-conflict levels by the final weeks of the monitoring period.
TEU figures in this section represent the total capacity of vessels that called at each port, not the number of containers actually loaded or discharged. A port recording 100,000 TEUs in a given week means vessels with a combined capacity of 100,000 containers called there — not that 100,000 containers moved in or out.
With that context, Durban averaged similar to its pre-conflict weekly vessel capacity across the period but the trajectory is uneven. It exceeded its baseline in 10 of 16 weeks, with its highest volume in Week 5 at 110,770 TEUs against a pre-conflict average of 57,510. However it fell below baseline in Weeks 2, 3, 7, 8, 12, and 14, indicating that elevated activity has not been consistent.
Port Elizabeth is the standout in this dataset, averaging 2 times its pre-conflict weekly capacity and exceeding its baseline in 13 of 16 weeks. Its Week 6 reading of 46,865 TEUs against a pre-conflict average of 8,475 is the single largest relative spike of any port in the Cape dataset. Cape Town’s period average of 1.3 times is heavily influenced by its own Week 6 spike of 82,212 TEUs. Outside of that week, Cape Town’s performance is more modest, and by Week 16 it sits at 31,553 TEUs, below its pre-conflict average .
Walvis Bay averaged barely above baseline, and was below its pre-conflict level in 7 of 16 weeks. Its elevated readings in Weeks 4 and 13 appear to be isolated surges rather than a sustained trend. By Week 16 it sits essentially at baseline. Maputo tells a similar story: a period average only marginally above pre-conflict, and a Week 16 reading of 4,546 TEUs against a pre-conflict average of 8,090, well below baseline and declining.
The Cape of Good Hope data does not support a single narrative. Port Elizabeth’s consistent elevation is a genuine signal that at least some carriers have structurally integrated the southern route into their service designs. But the normalization at Cape Town, Walvis Bay, and Maputo by Week 16 suggests that much of the early surge was tactical rather than structural. The ceasefire reported in the final week of the monitoring period may also be a factor in the recent decline at these ports.
