Southeastern Detraining: Investigation over Assumption


On 2 March 2018, during a period of rail disruption caused by extreme weather conditions, an incident occurred just outside Lewisham station in London. That incident saw a number of passengers de-train from a Southeastern service which had been delayed for an extended period of time.

A number of official reports into the incident are currently being commissioned or discussed within the industry and government. As is our practice, we will avoid commenting extensively on the incident (and its causes) until those have been released. Given the extended debate online, however, it seemed best to provide some background on detraining incidents in general, and points worth considering in the context of any current discussion (such as at our own upcoming meet up later this week).

The risk of imminence

Whatever the individual causes of incidents such as this – and they vary – there are three things that historically have generally played a critical part.

Firstly, that correctly evaluating the risk of imminence during railway incidents is a problem. Imminence, in this context, is how far away (in time) the solution currently being pursued is to being successfully carried out. The oft-overlooked issue here is that in many individual instances the timescales involved in implementing the currently preferred solution are considered, but not whether that was the first solution attempted.

As a result, there can be a tendency to think of a solution as being ‘imminent’ for passengers, because that solution itself will only take a short time to implement, whilst ignoring the total time spent on previously failed solutions. Ultimately, there is a point where a ‘best’ solution for detraining in the mid-term must give way to the opportunity do something quicker, with a slightly elevated risk.

Previous incidents have repeatedly shown that the failure here can happen when the operator (or National Rail) procedures in place undervalue the opinion of the staff on the train in relation to those managing the ‘bigger picture’ of the wider incident.

This can be a dangerous oversight, as it is often the driver (or other train-based staff) who knows better than anyone the emotional and physical state of the people on their individual train. They know when repeated (and genuine) intentions of dealing with the problem ‘shortly’ need to give way to dealing with the problem now.

Other incidents – such as at Kentish Town in 2011 – have shown that this is a particularly critical thing to get right in (or around) London, where there is often a station IN SIGHT of the passengers. This means that what seems imminent to them can be very different from what seems imminent in the control centre.

The importance of decisions

Undervaluing driver input also relates to a second common failing: decision-making responsibility isn’t delegated appropriately to local operational managers or staff on the scene. Or, if it is, then staff are insufficiently trained or supported in making those decisions.

This leads, subconsciously or consciously, to the benefits of possible solutions not being properly evaluated, or re-evaluated even when they’re technically permitted. This is because the ‘safer’ option for those involved in the decision-making process is to delay making one until someone senior ‘signs it off’ – entirely understandable in such an environment.

The importance of trust

Thirdly, in recent years there has seemed to be a correlation between Operators with a bad reputation for service and unplanned detraining. How much this is correlation rather than causation remains to be seen, but it may be because it is sometimes forgotten that the railways operate on a currency of trust.

That trust – which is paid to an operator by their passengers – can be thought of as operating on a Keynesian model. It is earned, and banked, by an operator through the delivery of good services and a positive passenger experience during the ‘good’ times. It is then spent to maintain control and understanding during the ‘bad’.

An appreciation of the role that trust plays in the operator / passenger relationship is more critical now than it ever was before. In previous eras, the information provided by train staff was often the only such information available to passengers during an incident, no matter how much faith the passengers had in its accuracy. Regardless of trust, the only competing source was the passengers’ own instincts or knowledge.

Today, however, a variety of other, ever-present information sources exist, competing to provide information or advice about the situation at hand. Indeed, by preference, an increasing number of travellers consider social media – in whatever form – to be one of, if not their preferred, method of communication during an evolving situation. Regardless of the merits of that approach, it is the reality. Worse, for an operator, it is a reality that dramatically impacts their ability to maintain control of the narrative when trust is found wanting.

The importance of evaluating the evidence

How much of a role these factors, and more, played in the recent Southeastern incident remains to be seen. A number of accounts from passengers claiming to be involved have begun circulating. No doubt many of these will turn out to contain critical information about what went wrong and what the industry needs to change.

They are, however, by definition subjective accounts which can only ever present a single perspective on events. It is the combination of these accounts, along with those from the railway professionals involved, and the data and recording systems in play that day, that will yield the wider picture of what happened.

Once that full picture has been established, whether through the guise of an ORR investigation, an RAIB investigation (should it be deemed appropriate) or other means, we will look at it further here on LR.

Those reports will highlight just how much the railway has learned – or failed to learn – from previous, similar incidents, most recently at Kentish Town, East Croydon and Denmark Hill. They will also provide some guidance on what the industry needs to do to avoid such incidents in future.

One thing that should be made clear, however, is that those conclusions will likely highlight that, as in previous incidents, the main causes were related to management and to risk, not politics.

It is perhaps not altogether surprising that much of the debate swirling online since the incident have looked to frame it through a window of ‘nationalisation vs privatisation’ or of ‘driver only operation’.

Those debates perhaps have some indirect relationship to this incident, but they are almost certainly far from its cause. There is no intrinsic reason why a national operator would be more ‘trustworthy’ than a privatised one if the underlying service (or at least the perception of it) has been consistently poor. This is something that LR readers with long memories who experienced bad service during the British Rail era would no doubt attest to. Similarly, there is no reason why a two-person train crew would be better placed to overcome the risk management issue of imminency than a single operator, if both lack the power or backing from their employer to do so.

As human beings we quest for simple, clear solutions. It is wired into the fibre of our being. In the overwhelming majority of railway incidents such of this, however, the causes are both more complex and more mundane. Politics doesn’t make the railways work or fail, people and procedures do.

Establishing and evaluating the causes of this incident will take time and the conclusions almost certainly won’t be simple. We look forward to seeing what the results, however, and to sharing them with you here.

Should people wish to discuss the incident or share links, ahead of the release of reports, then this article is the place to do it. Be warned, however, that unsupported assertions about what happened or general comments on the politics of nationalisation or DOO deemed outside the scope of the discussion will be deleted without comment.

Written by John Bull