Mainline - train operating companies

Management maturity

Overview: our RM3 assessments found varied performance (between RM3 level 2 and level 4) in employee engagement and consultation, internal communication arrangements, safety culture, change management and proactive indicators. 

There were two empty passenger train derailments, seven train collisions with vehicles at level crossings (resulting in two fatalities), and two low-speed collisions between passenger trains and empty passenger trains at Glasgow station.

We continue to encourage more progress on the use of activity-based safety performance indicators and outcomes because evidence shows that these can change behaviours and reduce unsafe acts.

Evidence: our RM3 approach is now embedded in most operators' safety management systems and is being used as part of their own auditing arrangements. Train operator staff continued to attend our training course on using RM3 in a manner consistent with our approach.

Working with the Association of Train Operating Companies (ATOC), we produced a suite of high-level RM3 evidence matrices on key risk topics such as train crew management. We intend to extend this to include passenger safety (but not at the platform-train interface), station infrastructure management and workforce safety. 

We also updated our RM3 assessment audit protocols for our inspectors and mainline train operators on evaluating the key risk topics during inspection and audit, to aid consistency of assessment criteria. 

In December 2014, we held our annual RM3 review meeting with 75 industry invitees. It was run jointly with ATOC, hosted by RSSB and with guests from Network Rail and the freight industry. This successful event reviewed TOC performance and looked at integrating and developing their use of RM3. We highlighted improvements made by the train operating community over the last four years. Some are making progress towards level 4 'predictable' risk management in several areas, but improvements to the implementation, culture and monitoring and review lag behind the leadership, policy and planning criteria.

Train protection and warning system

Overview: We continue to push operators to enhance their Train Protection and Warning Systems (TPWS) where reasonably practicable, so that it has in-service monitoring functionality. In 2014-15 SPADs highlighted some shortcomings with Mark 1 equipment. Progress has been improving. Chiltern Railways now have a progressive TPWS upgrade plan and c2c plans to upgrade its entire fleet to at least mark 3 TPWS as standard. We anticipate further train operator upgrade commitments in 2015. 

We continue to monitor industry plans to roll out the European Train Control System, which includes the European Rail Traffic Management System (ERTMS). However, until ERTMS is fitted network-wide, which is a long way off, the residual risk of and from SPADs remain the highest potential catastrophic hazard facing train operators.

Evidence: there were 299 mainline SPADs, a 4% increase, but overall SPAD risk declined 7% as measured against the September 2006 baseline. SPADs now represent more than 15% of the overall Precursor Indicator Model (PIM) risk, 10% of all train accident risk and 0.6% of total accident system risk. The numbers of multi-SPAD signals, where trains have passed them twice or more in the last five years, increased gradually over recent years. This trend is likely to be associated with a more congested mainline network and drivers facing more red signals. 

We are monitoring the development of the RSSB-led Strategy Project Group's 10 year SPAD mitigation strategy. Work to examine whether the higher risk SPAD categorisation process - in which we were involved - has amplified the actual risks posed by SPADs continues. There is some anecdotal evidence that SPAD causation categorisation has in the past been too risk averse given the actual circumstances. 

We prosecuted a now former First Capital Connect train driver who failed to set up his cab secure radio, passed a signal at danger, reset the train control system and then drove off without alerting the signaller at Hitchin in October 2013 – a so called 'reset and continue' SPAD. We take SPAD and 'reset and continue' SPAD events very seriously because they undermine basic system protection control arrangements. 

Mainline charter train SPAD

Our investigation into the potentially very serious SPAD at Wootton Basset Junction on the mainline network in March 2015 involving West Coast Railway (WCR) continues. Our work highlighted significant weaknesses in WCR's board governance arrangements and led to our enforcement action in May 2015. We continue to monitor the implementation of WCR's post-incident improvement plan, which included significant organisational change.

Driver management 

Overview: our short-term focus remains on improving driver management and maintaining pressure on operators for train protection enhancements. Specific issues include drivers losing concentration and becoming fatigued or distracted. We encourage the use of Network Rail's 'VariSPAD' workshops to rectify the causes and impact of high risk SPADs and of on-train data recorder downloads to provide more reliable driver performance assessment data.

Evidence: we continue to see examples of interruptions in driver's concentration and/or distractions causing, or at least exacerbating, potential safety incidents. We look to the whole industry to identify common patterns, learn lessons and implement appropriate remedies, which may include infrastructure enhancements.

Low-adhesion

Overview: our assessment of the sector's low-adhesion prevention and rail-head enhancement work on the mainline network, found that vegetation management remained substandard in some areas, but overall had improved on previous seasons. 

Evidence: leaf fall contributes to rail-head low adhesion by creating a slippery leaf mulch on the rail-head which can lead to trains sliding past signals for a considerable distance even after brakes are applied, or of trains becoming 'invisible' to the signalling system as wheel-to-rail contact is lost. 

We saw significant evidence from our inspections of systematic vegetation clearance and good evidence that rail-head treatment work has improved. However, it was inconsistent on some routes, such as long stretches in Anglia, which led to a heightened potential safety risk and knock-on effects on service performance.

Station management, train dispatch and the platform-train interface

Overview: overall harm to passengers and the public at stations increased 2% compared to 2013-14 but reduced 2% when normalised by the 4% increase in passenger journeys (the best available approximation of increases in station usage by passengers and public). Our pressure and the industry's own concern means that most operators now have good platform-train interface (PTI) risk management arrangements in place. Our inspections generally found strong and consistent PTI management processes. Overall mainline PTI harm reduced by 21%, but when normalised by increases in passenger journeys it decreased 24%.

We found evidence of good performance in safety leadership, risk assessment, worker engagement and safety culture around the management of the PTI. Generally, standards of train dispatch are at satisfactory to good levels, but some station risk assessment issues remain, such as curved platforms, or where platform furniture obscures views during platform-train dispatch.

Evidence: managing crowd congestion or taking appropriate proactive steps to avoid it continues to be a big challenge right across the network. These pressures are due to increases in passenger numbers, service frequency and disruption, including the impact of one-off and everyday rush-hour events, and from station infrastructure enhancement works. We saw examples of poor management of station crowding at Birmingham New Street, Finsbury Park (following the engineering overruns at Euston station), London Bridge and Paddington stations in 2014-15.

Procedures for proactively identifying and dealing with station crowding include having a dedicated control room, automated congestion monitoring with critical levels triggering a proactive response, use of CCTV to better understand passenger behaviour and staff resourcing levels, and the effective control of passenger flows during train dispatch. The industry's people on trains and station risk working groups are seeking to improve crowd management.

A cross-sector PTI risk management strategy was launched in January 2015 – in which we were involved and endorsed – and provides an analytical tool for enhancing PTI controls, including influencing passengers' behaviour. It shows a good example of effective partnership working.

We continue to inspect station passenger management plans to monitor their PTI risk management processes because of the high level of fatality and major injury risk posed. This includes periodic visual inspections to ensure proactive management is effective and dynamic. The future passenger crowding challenge means we must keep pushing operators' day-to-day responsiveness to crowding. 

As mentioned previously, we are also focused on the use of engineering solutions to design out risk for new or enhanced infrastructure, stations and rolling stock. This approach will help deliver our longer term asset management work plan, which includes the alignment of relayed track and platforms to reduce PTI gaps. Looking ahead, residual PTI risk will remain a perennial challenge until engineering solutions, such as PTI gap fillers, or platform-edge doors, (which can prove prohibitively expensive when retrofitted), are installed.

Rolling stock risks

Overview: we found rolling stock maintenance was generally of a high standard, but very little evidence that fleet departments are using the data they collected about rolling stock management to enhance a balanced suite of performance and safety indicators. Although they expressed a willingness to do this with their contractors. We continue to monitor trends in vehicles returning to service with failures after overhaul or external maintenance, but saw some improvements after the worsening trend in 2013-14. 

Evidence: PIM-measured risk from train and rolling stock increased 4%. We found evidence of train operators failing to fully recognise the full consequences of introducing new rolling stock. For example, the emergency evacuation plan arrangements for disabled passenger assistance and stepping distances for Thameslink's new 700 class fleet arrangements were not considered at the design stage by the train operator.

However, more broadly, our relationships with train operators and train service providers are open and honest and provide us with a strong forum to challenge them when necessary.