The reliability and timekeeping of trains are jointly assessed by the ‘Public Performance Measure (‘PPM’).
This is measured by the percentage of planned trains that are neither cancelled nor late. A train is ‘cancelled’ if it does not run or fails to complete half of its planned journey. For London and South East area train operators (‘TOCs’), a train is classified as ‘late’ if it arrives at its advertised destination five or more minutes after its scheduled arrival time.
As an example, in the three months ending 31 March 2006, South Eastern Trains – just prior to the handover to GoVia – had the following operating results overall.
| Number of trains planned | 131,249 |
| Cancelled | 0.8% |
| Late | 9.1% |
| Ran and on time (PPM) | 90.1% |
Source: London Travel Watch July 2006
For train operators serving London and the south east, PPM is calculated both for morning and evening peak services combined, and overall.
A criticism that could be made of the measure is the understatement of the importance of cancellations. A cancelled train and a train 5 minutes late score the same, although the disruption to passengers created by a cancellation – particularly of a heavily-used service in the rush hour – is likely to be greater.
The graph below shows the performance of the Kent franchisees since the measure was introduced in 1997-98.
Source: Office of Rail Regulation. Data are for Connex South Eastern until November 2003, South Eastern Trains until March 2006 and Southeastern thereafter.
In the chart, quarterly data for performance of peak services and all trains are shown as dotted lines, with the trends (four-quarter moving averages) as a solid lines. The quarterly data show a distinct seasonal pattern with performance falling off sharply in the October to December quarter (‘leaves on the line’ and bad weather). Around 40-45% of delays each year occur in the last quarter.
The sharp drop in performance in late 2000 reflects reaction to the Hatfield derailment. However, even before that accident it was evident that overall performance by Connex was deteriorating. Since Hatfield, performance has recovered to roughly its level of eight years ago, although there was a downward blip in the December quarter last year.
The recent improvement in performance has been experienced by all the London and South East operators reflecting a recovery in performance by Network Rail. The chart below shows the PPM of the Kent franchise operators relative to all TOCs in the region. Except for a period in 2001-2002, the performance of other London and SE TOCs has been better than that in Kent.
Source: Office of Rail Regulation.
Responsibility for delays is attributed either to a train operator or to Network Rail according to the cause. In addition, delays arising from poor traction during the autumn leaf fall are allocated to a ‘Neutral Zone’. In the chart below the recovery from Hatfield is apparent in the falling share of delays attributable to Network Rail.
Source: London Travel Watch. Data are for years ending March.
The most common causes for delay are malfunctions in trains and signalling problems, along with an undefined ‘other’ item attributed to Network Rail. ‘External’ causes are largely attributable to other TOCs using the network.
Source: London Travel Watch