ECONOMICS · MARKETS · Capital
Infrastructure Economics
When Is a Spare Freight Path Worth Nothing?
The Barking derailment raises a question that spare track mileage cannot answer: can a container actually complete its journey when the normal corridor is constrained? Using London Gateway as a case study, Rokin Research models how an inland terminal can make nominal rail capacity worthless, why backlogs persist after reopening, and when reserve transport capacity might justify its price. The freight volumes and monetary values are illustrative scenarios, not measured losses from the derailment.
Rokin Research
Key takeaways
The transition’s investable bottlenecks are increasingly physical: connection queues, grid equipment, permitting capacity, and skilled delivery.
An alternative rail path has value only when compatible crews, port handling, inland receiving capacity and operating windows are available together. In an illustrative ten-day restriction, 1,350 TEU accumulate; even after normal service returns, 11 calendar days of extra capacity are needed for clearance. A fourth usable path saves £107,000 per modelled comparable event, but actual path availability, prices and incident frequency remain unverified.
On 23 October 2026, a GB Railfreight container train travelling from London Gateway partly derailed at Barking. Shipping line Hapag-Lloyd subsequently suspended its own rail service to and from the port and advised some customers to arrange road transport. DP World, meanwhile, said its terminal was operating a reduced rail freight service. Those statements concern different services, and neither quantifies delayed containers. [1] [2] [3]
That raises a larger economic question: what is a spare freight route worth if the container cannot complete its journey? An additional railway path is useful only when a train and crew can use it, the port can load it, and an inland terminal can receive it at the right time. Rokin Research's London Gateway study tests the consequences of that distinction through explicit scenarios, rather than claiming to reconstruct the disruption's measured cost. [6]
A spare track is not spare transport capacity
DP World lists eight on-dock rail tracks and 22 daily rail services at London Gateway. Those are useful descriptions of the terminal, not a count of affected trains on 23 October, and certainly not proof that 22 alternative services were available. Transport East's Thameside assessment also identifies competing passenger and freight demands along the wider corridor. [4] [5]
The correct unit of redundancy is an end-to-end movement. Imagine that the railway network can accept five additional train-equivalents a day, but compatible locomotives and crews can handle four, the port can load six, contractual windows allow four, and the inland destination can receive only three. Effective capacity is at most three. A sixth network path adds nothing if the receiving terminal still stops at three.
These hypothetical limits show how expansion at a non-binding stage can leave overall capacity unchanged. Road haulage has the same constraint: more lorries do not help if port gate appointments or delivery windows are already full. [6]
Reopening a route does not clear an inherited queue
Consider the study's central illustrative case. It represents 12 affected train-equivalents per day at 65 twenty-foot equivalent units (TEU) each: 780 TEU of daily freight requirement. A TEU measures container capacity, not individual containers. The model assumes 25% of normal rail service remains available (195 TEU/day), three alternative train-equivalents move another 195, and 170 additional loaded road lifts at 1.5 TEU each move 255. [6] [7]
Together those alternatives process 645 TEU per day, leaving 135 TEU of new unmet work per day. Over ten hypothetical restricted days, 1,350 TEU accumulate. Once ordinary service is restored, it can handle new arrivals, but an additional two train-equivalents, or 130 TEU/day of net headroom, are needed to reduce the old queue. Clearing 1,350 TEU then takes just over ten days, or 11 calendar days. Without extra headroom, reopening would stop the backlog from growing without making it disappear. [6] [7]
Waiting also accumulates while freight sits unresolved. Across the ten restricted days and subsequent recovery, the model produces approximately 13,760 queued TEU-days. This means the sum of outstanding TEU over time, not 13,760 distinct containers. The calculation includes the partial final recovery day. [6] [7]
The valuable path is the one that changes the bottleneck
Under the same ten-day illustrative assumptions, a fourth usable alternative rail path cuts the peak backlog from 1,350 to 700 TEU. But an extra network slot that remains blocked by a three-train inland receiving limit leaves the backlog unchanged. The comparison separates nominal from usable capacity. [6] [7]
Illustrative intervention | Peak backlog after ten days | Private planning-cost proxy | Saving versus base |
Three usable alternative rail paths (base) | 1,350 TEU | £684,115 | — |
Four usable alternative rail paths | 700 TEU | £577,115 | £107,000 |
Extra network path, inland capacity unchanged | 1,350 TEU | £684,115 | £0 |
Equivalent extra road collection (65 TEU/day) | 700 TEU | £638,865 | £45,250 |
Source: Rokin Research model v0.3, ten-day constructed scenario; assumed constant-2026-GBP incremental private logistics costs, not observed London Gateway expenditure. The rail and road comparison holds the additional freight handled constant. It does not establish that either option is physically available. [6] [7]
The fourth rail path's £107,000 per-event saving already includes its assumed extra operating premium. The road alternative handles the same additional 65 TEU/day but saves only £45,250 with the study's higher assumed road premium. Neither figure is a market quotation or a measure of social welfare. A real comparison would need actual gate access, inland capacity, tariffs, passenger impacts and environmental costs. [6]
Duration and timing can be as important as capacity
With demand and clearing capacity fixed, doubling the hypothetical restriction from ten to twenty days doubles the peak backlog from 1,350 to 2,700 TEU, but quadruples queue exposure from about 13,760 to 55,038 TEU-days. Earlier cargo waits through more of the interruption, and the larger inherited stock takes longer to clear. These are selected scenarios, not a probability interval or a forecast of the Barking incident. [6] [7]
The model also tests a hypothetical 2,500-TEU working storage buffer. A twenty-day case exceeds it by 200 TEU, but the study does not attach a speculative overflow penalty. Similarly, concentrating arrivals earlier in a ten-day period raises queue exposure to about 16,100 TEU-days even though the total cargo is unchanged. Averages conceal some operating constraints. [6]
An event saving is not an investment return
A reserve-path contract must be evaluated before the next interruption. In the model, one additional usable path saves £107,000 in a comparable ten-day event. Assuming, purely for illustration, 0.25 comparable events per year, that translates into £26,750 in expected annual private planning-cost avoidance before any standby fee or capital spending. [6] [7]
If securing the capacity also requires £50,000 upfront, the study's ten-year, 3% real-discount example reduces the maximum sustainable annual availability fee to about £20,900. That calculation is not a recommendation to buy a particular path: the frequency, contract terms, construction needs and real ability to activate it are all unverified. It also excludes impacts on other rail users and therefore cannot establish a public benefit-cost case. [6]
The investment decision starts at the actual constraint
For freight operators, the lesson is not to buy redundant assets indiscriminately. It is to identify which missing link prevents an entire delivery, compare feasible rail and road packages at the same service level, and check how much capacity remains after the main route reopens. A nominal alternative may be worthless; a compatible receiving slot or a coordinated operating agreement may be the critical addition.
The Barking derailment makes this a timely question, but the answer established here is narrower and useful: resilience is the ability to complete an additional movement, not the number of routes that appear on a network map. [6]
Research note. This article explains a constructed freight-capacity model using publicly documented incident facts and published infrastructure context. Its freight volumes, restriction length, diversion capacity, queue totals and monetary savings are not measured consequences of the October derailment. The technical report records the equations, alternative scenarios, assumptions and evidence gaps; the Excel model permits readers to test those inputs. [6] [7]
Sources
[1] c2c (23 October 2026), Barking freight train derailment: Advice for customers. Open source
[2] Hapag-Lloyd (27 October 2026), UK: Rail Services to and from London Gateway Suspended Due to Train Derailment. Open source
[3] The Loadstar / DP World statement (27 October 2026), No rail freight from London Gateway and Tilbury after storms. Open source
[4] DP World (Undated specification page, accessed 30 October 2026), London Gateway - infrastructure and facts. Open source
[5] Transport East (Undated strategy page, accessed 30 October 2026), Thameside Taskforce. Open source
[6] Rokin Research (30 October 2026), When a Freight Corridor Is Blocked, How Much Spare Capacity Prevents a Backlog?. Exact research asset supplied with this package.
[7] Rokin Research (30 October 2026), London Gateway Rail Redundancy Model, v0.3. Exact research asset supplied with this package.