What happens when an inspector calls

Rail accident investigation was rather losing its way in the early 2000s. The scramble to apportion blame and demonstrate to the public that something was being done quickly meant that uncoordinated responses to accidents hindered a thorough railway investigation of an incident. It was not a situation that could be allowed to continue, if safety on the network was ever to improve.

In the wake of the Ladbroke Grove rail disaster on October 5 1999 (when 31 people lost their lives in a head-on collision between two passenger trains), Lord Cullen launched a public inquiry into not only the accident itself, but also the management and regulation of rail safety as a whole. 

Cullen concluded that the UK should set up an independent accident investigation body - a need that was reinforced by an EU Directive passed in 2004 requiring each member state to have its own accident investigation organisation, independent of regulator and industry. 

And so the Rail Accident Investigation Branch (RAIB) was born. The UK Government passed the Railways & Transport Safety Act of 2003, creating powers to form the new organisation. Career railway engineer Carolyn Griffiths, who had joined British Rail in 1979, was appointed to lead the new organisation. Griffiths stepped down from the role earlier this year, after 12 years as the RAIB’s leader. Simon French is currently the acting chief inspector, pending a permanent appointment.

Setting up the RAIB meant negotiations into what its relationship should be with the various bodies that already had an interest in railway safety - the industry itself, the police, and the Health and Safety Executive (HSE, into which Her Majesty’s Railway Inspectorate had been transferred in 1990).

Investigators and support staff had to be recruited and trained, offices found and equipment acquired, all at the same time as the detailed legal framework under which the whole thing would operate was being developed. It was a once-in-a-lifetime opportunity to start with a clean sheet and set up an organisation that would investigate thoroughly, efficiently and without fear or favour.

RAIL readers with long memories may recall that back in July 2004, while the RAIB was still in gestation, Nigel Harris interviewed Griffiths. He asked her how her new organisation was going to function, and the difference it might make to the way in which accidents and incidents on the UK’s railways and tramways were investigated. It’s now been ten years since the RAIB began its operational life on October 17 2005… so what has it achieved since then?

RAIB now employs 24 investigators (known as inspectors by the law). They are split into four teams, each headed by a principal inspector, and divided equally between two bases - at Derby and Aldershot. Each inspector was recruited for their expertise in railways, accident investigation and related topics, and all have extensive experience in their fields. 

Of the 22 inspectors and principals in post when the Branch (as the RAIB is also affectionately known) began operations in 2005, 11 are still serving. Four have since retired and seven have moved on to pastures new. This continuity has enabled the RAIB to build up a considerable (and respected) bank of skills and knowledge.

The RAIB’s inspectors provide an on-call team of at least four people (two from each office) who are available continuously, prepared to go anywhere in the country at 30 minutes’ notice, in response vehicles containing all the equipment needed for an on-site investigation. 

Their response is driven by a duty co-ordinator, drawn from a pool of trained inspectors, and who is on duty to take phone calls 24 hours a day, seven days a week. Notification of an incident usually reaches the Branch soon after it happens. If it becomes apparent that an event may warrant urgent on-site investigation, the duty co-ordinator will despatch a team to investigate. 

Frequent training exercises to test the readiness of the team and refresh their skills are an important part of day-to-day RAIB operations. There are also 18 support and administrative staff split between the two offices, and some of these are also available to attend investigation sites to provide technical and logistical assistance.

Over the past decade, RAIB investigators have been deployed to sites more than 450 times. 

Not all of those call-outs have led to published reports, as it is sometimes clear from a preliminary examination of a site that there are no safety lessons to be learned. Even so, the Branch has published more than 300 reports and bulletins in the public domain (all of which can be read on the RAIB website: www.gov.uk/raib). The full investigation reports, which include recommendations for changes that need to be made to improve safety, are the main output of the Branch. 

Sometimes an investigation may reveal that although an incident has identified shortcomings in the way things were being done, it’s not actually necessary to recommend changes to equipment or systems. In such cases, RAIB publishes a bulletin to inform the industry and the public about the event, and to highlight learning points that can be used in briefings to staff  (such as the importance of properly carrying out the requirements of the railway’s rule book).

Some have questioned the length of time it takes from an incident happening to the publication of the investigation report (it can be several months, or even longer at times). The Branch says it is very conscious of this, and takes criticism of the length of time taken very seriously. So why can it take so long?

The RAIB says the most important elements of the work are being thorough and being right. In respect of being thorough, it aims to add value and improve on what it sees coming from the industry’s own reports. 

Industry’s understanding of the techniques of investigation has improved significantly in recent years, but there is still a tendency for investigations to stop at a point where an immediate cause has been found, and not go ‘the extra mile’ to really question why things happened. This is where RAIB’s investigations differ: it keeps asking ‘why?’ until it gets to a point where everyone agrees that there are no more useful answers to find.

As for being right, the RAIB is adamant that its reports must be without error - what gets published has to be as accurate as it possibly can be.

The first stage of an investigation is the site work. This is when the most valuable evidence is usually gathered. As well as making sure that everything of relevance is captured, the RAIB team will also be trying to ensure that the site of the incident is handed back to the rail industry as quickly as possible, so that trains can start running again. Often, this can be before inspectors even get to the site, if it is clear from phone conversations that there is no reason for all or part of the scene to be ‘frozen’ any longer. 

For example, in the case of a train striking a pedestrian or cyclist at a level crossing, the train involved can be examined once it has been taken to a depot, and the level crossing can be photographed and surveyed without impeding the passage of trains. What might be crucial in this situation is recording exactly where the train stopped, and it’s often possible to arrange for that information to be recorded by people on site, before the train is moved.

Once the site team has returned to base, the first decision to be made is what (if any) level of investigation is required. This discussion takes place in a meeting of all the inspectors (usually on the Monday following the site work). The inspectors who went to the site present their findings - the course of events, the results of any interviews, reconstructions or tests already carried out, a review of previous relevant investigations (which may go back long before the founding of the RAIB), and a recommendation on what to do next. 

Aside from more traditional investigation methods, RAIB staff also keep a close eye on social media. What is said on forums, message boards, Facebook and Twitter may all find its way into the mix when it comes to deciding what action to take.

There have been more than 450 incidents on the railway over the past decade. Thankfully serious railway accidents are rare in the UK, but how is it decided which incidents the RAIB should investigate?

In the Regulations under which RAIB operates (see panel, page 60), there is a definition of ‘serious accident’. If an event falls into this category, it must be investigated. The rest of the investigations undertaken by the RAIB are at its own discretion, based on the decision taken at the Monday meeting. The main criterion is the amount of safety learning that is likely to come out of the investigation.

While it is important to inform interested parties about what has happened, the main value of an RAIB investigation (which justifies the resources put into it) lies in its potential to make the future railway safer. This safety learning can even come from events that at first sight may appear trivial, or from the actions of people not connected to the railway. 

For instance, in 2013, a collision between a train and a car at Jetty Avenue level crossing (Woodbridge) did not result in any personal injury. However, the RAIB investigation found that at this level crossing, safety relied on users looking out for approaching trains. Many motorists did not understand how they were supposed to use the crossing, while Network Rail did not understand how motorists were actually behaving and how limited the view of approaching trains actually was. 

The resulting recommendations seek to achieve fundamental changes in the way that safety is managed at user-worked level crossings, which currently represent one of the biggest contributions to risk on the railways. 

After the decision to investigate has been made, the next step is to let the wider world know of RAIB’s intended actions. This involves drafting a website entry (which often starts the ball rolling for what will later appear in the press and on internet news sites), and letting interested parties know what information is going to be released. The RAIB wants, in particular, to make sure that injured and bereaved people know that their accident is going to be the subject of wider attention, before the details are made public.

When things need putting right in a hurry, the RAIB needs to act fast. If the work on site has uncovered a serious problem with a piece of equipment or an operating practice, the Branch can issue urgent safety advice, either to the industry or just to the companies directly affected. 

Such advice can also be used to summarise the RAIB’s emerging findings, at a later stage in the investigation. And when this advice is sent out, it appears in the final report for reference. So far, this process has been invoked 31 times, for issues ranging from the design of the switches involved in the derailment at Grayrigg (see case study, page 57) to the details of cotter pins on steam locomotives.

As an investigation progresses, evidence is gathered and analysed. Regular reviews are held to confirm that the right evidence is being collected, and that the analysis is correct. Where there are technical issues, it’s likely that testing will take place. Additional interviews with people who have information about what happened also help fill in the gaps in the investigators’ knowledge. 

Peer review is a vital and hugely valuable part of the process, and takes place at several levels, up to the final analysis review with the chief inspector. At each stage the investigator is challenged to justify the conclusions they have drawn and the recommendations they are proposing.

During and after the analysis review process, the report is being written. It too has to survive peer review at several levels, and only once the chief inspector is happy with it can it be released for consultation. This process happens in the weeks running up to publication of the report, and is a stage required by law. It is seen by the Branch as an important opportunity for the people and organisations that have an interest in the investigation to have their say. 

The investigation team will contact each of the organisations that are likely to be affected by any recommendations, as well as the injured or bereaved, to share with them what the RAIB has found and what it intends to say in the report. Just as importantly at this stage, it provides some much-needed answers for injured parties and bereaved families. The formal consultation then gives each party sight of the draft report, and two weeks to comment on it.

All this should mean that any errors are picked up before publication, and any doubts about whether recommendations are desirable or practicable can be dealt with. Once all the comments from the consultation process have been considered (in some cases, changes to the report can mean that a re-consultation is required), the final text is submitted to the Secretary of State for Transport, and published on RAIB’s website three days later.

Since its inception, the RAIB has made a total of 1,330 recommendations, covering aspects across the entire industry. Almost all of these have been addressed to the Office of Rail and Road, which confirms that around 95% of recommendations have been implemented, showing the very real difference the investigative work of the Branch is making.

 “The RAIB has come a long way in its first ten years,” says Acting Chief Inspector Simon French. “I am really proud of what we have achieved to date, the professionalism of the team, and our important contribution to making the UK’s railway among the safest in Europe.”

The Branch is also striving to further improve the service it provides to the industry. While it says it will never compromise on quality, it is determined to find more effective ways to communicate its findings to the public and the industry. 

Each year it publishes an Annual Report (this year’s was published in August), which lists the recommendations made to the industry and the action that has been reported to implement them. 

Over the course of its first ten years, the RAIB has shone a light onto the minutiae of railway operations, to uncover the weaknesses that hinder safe working. While it is sometimes uncomfortable for the individuals and companies involved to come under such close scrutiny, it is done without judgement and becomes the catalyst for significant safety improvements. 

• This feature was published in RAIL 786 on October 28 2015