Thursday, 10 September 2026
Thursday, 10 September 2026 was a poor day on the railway: of the 19,770 passenger trains whose arrival was recorded, 87% were within five minutes, and a further 835 were cancelled outright.
An ordinary weekday timetable, with the two commuter peaks that give the network most of its work and least of its slack.
2,362 trains arrived more than five minutes late and 124 more than half an hour, against 17,284 that ran to time. The average delay across every train whose arrival was recorded was 1.6 minutes.
The reason attached most often to a late train was “This service has been delayed by a speed restriction” — 622 services carried it. Darwin publishes these as complete sentences written for passengers, so they appear here unedited.
The busiest hour was 07:00, with 1,669 booked departures and an average delay of 1.3 minutes.
The day in numbers
| Passenger trains recorded | 25,624 | |
|---|---|---|
| Cancelled | 835 | 3.3% |
| Arrival time recorded | 19,770 | 79.8% of those that ran |
| …arrived within 5 minutes | 17,284 | 87.4% |
| …arrived 5 to 30 minutes late | 2,362 | 11.9% |
| …arrived more than 30 minutes late | 124 | 0.6% |
| Average delay at destination | 1.6 min |
Every percentage among the bands is of the 19,770 trains whose arrival reached the archive — not of all 25,624, and the bands are exclusive. 5,019 trains ran without an arrival time being recorded, and nothing is claimed about how punctual those were. Delay is judged at the destination, a stricter test than the industry's own measure and not comparable with it.
By operator
38 operators ran passenger services on this date, from ScotRail on 3,035 runs down to TY on 4. Ordered by volume, not performance — any heading re-sorts it.
| Operator | ||||||
|---|---|---|---|---|---|---|
ScotRail SR | 3,035 | 0% | 2,195 | 97% | 97% | 0.2 min |
Northern NT | 2,792 | 2% | 2,579 | 89% | 88% | 1.4 min |
Great Western Railway GW | 2,134 | 7% | 1,431 | 82% | 74% | 2.7 min |
South Western Railway SW | 1,836 | 2% | 1,429 | 89% | 87% | 1.3 min |
Southeastern SE | 1,776 | 2% | 1,626 | 89% | 87% | 1.6 min |
Southern SN | 1,683 | 5% | 1,374 | 81% | 76% | 2.8 min |
London Overground LO | 1,602 | 3% | 1,508 | 98% | 95% | -0.2 min |
West Midlands Railway LM | 1,289 | 3% | 996 | 85% | 82% | 2 min |
Greater Anglia LE | 1,288 | 1% | 1,233 | 93% | 92% | 0.3 min |
Transport for Wales AW | 1,222 | 3% | 958 | 95% | 91% | -0.8 min |
Thameslink TL | 949 | 5% | 562 | 81% | 75% | 2.4 min |
London Underground LT | 778 | 0% | 190 | 95% | 95% | 1.5 min |
Elizabeth line XR | 698 | 11% | 494 | 80% | 69% | 1.9 min |
East Midlands Railway EM | 612 | 1% | 478 | 83% | 82% | 2.2 min |
Merseyrail ME | 591 | 6% | 330 | 97% | 87% | -0.3 min |
Great Northern GN | 440 | 4% | 328 | 80% | 76% | 3.2 min |
Avanti West Coast VT | 406 | 1% | 308 | 51% | 50% | 7.5 min |
c2c CC | 379 | 9% | 273 | 61% | 54% | 5.8 min |
Chiltern Railways CH | 350 | 7% | 324 | 98% | 91% | 0.2 min |
CrossCountry XC | 342 | 2% | 266 | 63% | 61% | 7.2 min |
TransPennine Express TP | 328 | 2% | 301 | 73% | 72% | 3.7 min |
London North Eastern Railway GR | 310 | 1% | 194 | 72% | 71% | 5.9 min |
Tyne and Wear Metro TW | 174 | 0% | 101 | 98% | 98% | -0.3 min |
Heathrow Express HX | 152 | 49% | 71 | 32% | 16% | 9.9 min |
Sheffield Supertram SJ | 98 | 0% | 41 | 100% | 100% | -0.9 min |
Gatwick Express GX | 72 | 22% | 22 | 77% | 45% | 4 min |
Charter/Other ZZ | 61 | 0% | 3 | 67% | 67% | 4.7 min |
Island Line IL | 52 | 0% | 52 | 100% | 100% | -0.8 min |
Eurostar ES | 52 | 0% | 26 | 81% | 81% | 4.7 min |
West Coast Railway Co WR | 30 | 0% | 8 | 88% | 88% | -15 min |
Grand Central GC | 21 | 0% | 19 | 47% | 47% | 13.2 min |
LS LS | 14 | 0% | 5 | 100% | 100% | -2 min |
Hull Trains HT | 14 | 0% | 14 | 64% | 64% | 7.6 min |
Lumo LD | 12 | 0% | 12 | 67% | 67% | 3.6 min |
Lumo Stirling LF | 10 | 0% | 10 | 80% | 80% | 4.7 min |
Caledonian Sleeper CS | 10 | 10% | 9 | 89% | 80% | -8.8 min |
North Yorkshire Moors Railway NY | 8 | 0% | 0 | — | — | — |
TY TY | 4 | 0% | 0 | — | — | — |
“On time” is of each operator's recorded arrivals; “incl. cancelled” counts a cancellation as a failure, which is what the industry's own measure does. Across the whole day they read 87% and 84% — why both are published, what each hides, and how the sort treats a small operator.
Why trains were late
8 different reasons were given for delay on this date, each reproduced in the operator's own words.
This service has been delayed by a speed restriction
622 services
This service has been delayed by a late running train being in front of this one
566 services
This service has been delayed by a safety inspection of the track
417 services
This service has been delayed by a speed restriction over defective track
260 services
This service has been delayed by earlier engineering works not being finished on time
245 services
This service has been delayed by vandalism of railway equipment
151 services
This service has been delayed by a fault on a train earlier
100 services
This service has been delayed by a fault with the signalling system
88 services
Why trains were cancelled
A reason reached the archive for 493 of the 835 cancellations (59.0%), across 8 distinct reasons. A run cancelled part-way along often carries no reason at all.
This service has been cancelled because of vandalism of railway equipment
211 services
This service has been cancelled because of a safety inspection of the track
97 services
This service has been cancelled because of a fault on a train
44 services
This service has been cancelled because of more trains than usual needing repairs at the same time
33 services
This service has been cancelled because of a shortage of train drivers
28 services
This service has been cancelled because of a shortage of train crew
28 services
This service has been cancelled because of an operational incident
26 services
This service has been cancelled because of a fault on this train
26 services
The longest delays
The worst of them arrived 1h 52m behind schedule — London Waterloo to Woking, booked away at 23:20, run by South Western Railway. The ten longest range from 1h 52m down to 1h 10m, and they belong to 8 operators, Great Western Railway accounting for 2. What a delay on that scale usually means.
| Booked | Service | Late by |
|---|---|---|
23:20 | London Waterloo to Woking 2F73 · SW | 1h 52m |
16:15 | Tring to London Euston 2T48 · LM | 1h 43m |
04:35 | Reading to London Paddington 2P11 · GW | 1h 39m |
05:12 | Heathrow Airport T5 to London Paddington 1Y22 · HX | 1h 29m |
05:42 | Reading to Oxford 1Z10 · GW | 1h 25m |
21:39 | London Euston to Manchester Piccadilly 1H45 · VT | 1h 17m |
13:00 | London Kings Cross to Sunderland 1N93 · GC | 1h 14m |
09:52 | Aberdeen to London Kings Cross 1E15 · GR | 1h 14m |
06:49 | Liverpool Lime Street to Norwich 1L06 · EM | 1h 10m |
12:00 | Edinburgh to London Kings Cross 1E14 · GR | 1h 10m |
Through the day
Booked departures by hour, with the share that arrived within five minutes. Punctuality was weakest at 00:00, when 55% of 158 arrived within five minutes.
The filled part of each bar is the share of that hour's booked departures that arrived within five minutes. The quietest was 02:00 with 15.
What ran
Formations were recorded for 26,784 trains on this date, made up of 2,909 individual units and 12,724 separate vehicles. These figures cover the services an allocation was published for, not every train that ran.
25 different classes of train appeared, which is the signature of a full weekday — the long-distance fleets, the regional diesel units and the suburban electrics all working at once.
The most-used class was the 377, allocated to 3,182 of the day's recorded trains from a pool of 219 units.
The average train was 5.6 vehicles long. The longest formation recorded was 20 vehicles.
| Class | Trains | Units used |
|---|---|---|
| Class 377 | 3,182 | 219 |
| Class 465 | 1,627 | 100 |
| Class 720 | 1,525 | 129 |
| Class 450 | 1,509 | 118 |
| Class 150 | 1,385 | 95 |
| Class 158 | 1,350 | 133 |
| Class 345 | 1,281 | 67 |
| Class 165 | 1,170 | 66 |
| Class 156 | 1,169 | 94 |
| Class 387 | 1,109 | 88 |
| Class 375 | 1,035 | 101 |
| Class 385 | 989 | 66 |
Of 25,624 runs, 26,784 had a formation published (105%) — these are allocations rather than observations.
Buses
17,247 buses from 187 operators were seen on this date, across 3,086,155 position reports.
129,766 bus journeys were followed to their destination on this date. 92,229 arrived on time, 7,449 late, and 30,088 — 23% — arrived ahead of schedule.
Early arrivals are normal rather than exceptional. Bus running times carry enough slack to survive traffic, and it is concentrated in the last leg, so a clear run simply finishes sooner than booked.
The average arrival was 0 minutes late, and 780 journeys finished more than twenty minutes down.
| Journeys followed to destination | 129,766 | |
|---|---|---|
| On time | 92,229 | 71.1% |
| Early | 30,088 | 23.2% |
| Late | 7,449 | 5.7% |
| …more than 20 minutes late | 780 | 0.6% |
| Average arrival | 0 min late |
By operator
| Operator | Journeys | On time | Early |
|---|---|---|---|
Transport for London TFLO | 102,582 | 73% | 24% |
Nottingham City Transport NCTR | 2,580 | 79% | 17% |
Go North East GNEL | 1,731 | 71% | 10% |
First South Yorkshire FSYO | 1,655 | 58% | 26% |
First Leeds FLDS | 1,157 | 59% | 29% |
Bee Network BNML | 861 | 72% | 20% |
National Express West Midlands TNXB | 830 | 79% | 15% |
Bee Network BNSM | 811 | 79% | 15% |
National Express NATX | 783 | 30% | 32% |
Arriva Yorkshire WRAY | 725 | 58% | 24% |
First Potteries FPOT | 669 | 52% | 32% |
Preston Bus PBLT | 604 | 64% | 27% |
Of the operators with enough journeys to compare, Reading Buses ran closest to time (85%), and National Express furthest from it (30%).
Our figures, not the regulator's — why buses run early, the current league.
Recorded live from the Darwin and TRUST feeds on 10 Sept 2026 itself. How these figures are measured · Sources and licences