With the news just in that our excellent NRM volunteer team have finished transcribing the final batch of state accident reports, covering 1900-1910 and amounting to over 8000 cases, it’s good to be able to share a guest blog post by one of our longest-serving volunteers, Philip James. In it he draws upon his experience of transcribing large numbers of accident records, noting some similar cases he’s seen over the years. This is just one advantage of the ‘long view’ which our project enables – it’s therefore great to see it being brought to the fore like this. Our thanks to Philip for all of his hard work, as well as this post!
This is my third post on my experience as a volunteer (see post 1 and post 2 for the others). My motivation for becoming a volunteer was partly an interest in railways and a desire to do something useful in this field. I also wanted to link my interest in railways with professional and technical skills acquired during my working life and earlier studies. This project has given me some ideas for doing this; a subject for a future post perhaps.
At present, like most volunteers on the project, most of my effort goes into the primary task of data entry and this will inform my observations for this piece. As work on phase three of the Railway Work, Life & Death project progresses I have had more thoughts on how it is progressing and my experience with it.
I can say little that is new from a technology perspective except to confirm that two large monitors are a considerable asset to aiding analysis and data entry. For preparing these guest posts, the snipping tool and optical character recognition tools have been helpful to extract data and images from source material and get them into a form suitable for use here. Even so, the print quality of some PDF documents is such that a significant level of manual editing is needed to produce readable text. The mark one eyeball is not redundant yet.
Project experience and issues
Accident Types and Interesting Questions
I have previously focused on the mechanics of digitising data for the project and some of the problems associated with this. Now that I have a reasonable pool of data on which to draw, it is possible to reflect on patterns and trends that are emerging and questions this may pose to researchers.
It occurs to me that other volunteers may have spotted trends in the data but will they be similar to my observations? If they are markedly different, does this imply a problem in the way individuals are working or do differences in the datasets explain it?
Many accidents could be categorised in more than one way so is there consistency in the way volunteers are thinking? Would it be desirable to have more than one way of categorising them? It may cause problems if we were to change the categorisation approach now.
The project co-leads will have access to a much larger data pool and it will be interesting to see how their analysis develops.
Injuries and Types of Accident
In the accidents I have seen about a quarter result in a fatality. About forty percent have to be categorised as ‘multiple injuries’ or ‘other’, mostly the latter, as the information about the injury does not allow a more specific attribution.
In terms of the location/ activity being undertaken at the time of the accident ‘about the track’ accounts for about a quarter and ‘shunting’ for well over a third. Sometimes there is more than one potential type for example, shunting may result in a person about the track being caught between vehicles. Few accidents appear to be primarily attributable to ‘workshops’ or ‘goods handling’ and ‘on railway property’ is a convenient catch all for a small number that don’t fit elsewhere.
Having completed fifteen batches of various sizes yielding 1,207 row entries I have noticed variations, sometimes marked, in the proportions of injuries and accident types for each batch. None of my batches are for Quarter 3 so I can only speculate about seasonal factors. A small batch could have uncharacteristic proportions that would not be evident in a larger one. There are no obvious trends. Without doing more analysis and using a larger data set, it is not possible to arrive at more meaningful conclusions.
Causes of Accidents
Accidents seem to have a limited set of headline causes, one or more of which be applicable to each case. In many cases, the victim is found to have a degree of responsibility for their own mishap. In some cases, it is the fault of another person. Some cases are attributed to misadventure or accident and in some a faulty system of working has been identified. The latter may imply a failure of policy or regulation by the Railway Companies or other authorities.
In many cases, the injured or deceased person has been deemed to be responsible in some way for their mishap. Mike Esbester informs me that around 95% of cases were found to be employee error. This suggests a greater willingness to accept accidents and less inclination to protect people from their mistakes and misjudgements. It is evident that there is a range of behaviours on the part of the accident victims ranging from those who have clearly caused their mishap to those who could have been more careful but the primary cause was elsewhere.
The Railway Inspectorate had an ongoing safety campaign that could be summarised as Lock, Block and Brake. This refers to the interlocking of points and signals, the division of the railway into blocks for signalling purposes and the fitting of brakes to all railway vehicles. Failures in Lock and Block are likely to result in serious accidents reported separately and few reports I have seen touch on them.
Many reports highlight the frailty of braking systems used on goods wagons. Some wagons had no brakes at all and frequently a wagon had only a hand brake operated by a lever on one side only and sometimes even this would be faulty. It was common practise for men to ride on wagons to apply the brakes or run alongside them for same. Sometimes sprags were inserted into wagon wheels to slow them and this was frequently reported as a dangerous practise, the danger usually being in the process of removing the sprag, a point made explicitly in one report. Sometimes men would scotch the wagon wheels using improvised wedges to stop them moving, also a dangerous practise.
Craig Shaw (the NRM volunteer coordinator) has made similar observations and this has prompted us to research the subject a little more. A Board of Trade ruling in 1911 stated that new wagons must have brakes operable from either side and existing wagons must be made compliant by the end of a transition period. With extensions this turned out to be by 1938 and for some wagons later still, no doubt due to the pressing needs of war and subsequent reconstruction.
This article by Jim Champ on Great Western Railway wagon brakes provides a detailed explanation. It also notes the displeasure of the GWR as they had their own design of brake operable from either side of the wagon but not compatible with the details of the Board of Trade ruling.
The Board of Trade also wanted to enforce the right-hand rule whereby the brake lever would be to the right when viewing a wagon from the side. This has the additional consequence that there will be a brake lever at each end of a wagon, important in some locations and one accident report refers to it.
The Regulation of Railways Act 1889 includes provision for Lock, Block and Brake giving the Board of Trade powers to impose changes.
Mike Esbester has also been following the ‘either-side brake’ question which he describes as a ‘long and complex story’ initially linked to the 1902 Prevention of Accidents rules and has found evidence of further extensions for some wagons into the late 1940s. He concludes that “a key problem was the private owner wagon fleet, collieries being a large part of this, and they were unwilling to foot the bill for something they regarded as ‘unnecessary’”.
Railway companies and collieries had large numbers of typically four-wheeled coal wagons that were akin to warehouses on wheels and were reluctant to accept the cost of fitting hand brakes to all of these, let alone continuous brakes operated from the engine. With regard to continuous brakes, these took rather longer to become universal and resulted in a new breed of railway vehicle, the diesel brake tender. At the time of transition from steam to diesel, there were still many unfitted wagons that relied on the locomotive, tender and brake van to bring them to a stand. A diesel locomotive often had less brake power than the steam engine and tender it replaced so brake tenders were created to supplement the braking force. They were in use until the late 1970s. This article from the Great Central Railway, Loughborough, describes their work to recreate such a vehicle. Typically, they were made by converting redundant rail vehicles and fitting weights to aid the braking force.
Issues with coupling also feature in many reports. Men often went between vehicles to couple manually or better still used poles to do this from a safe distance. When the couplings failed to work properly, they might resort to manual coupling and in an environment where shunting is taking place, the risk of being crushed is significant.
In many cases, they exposed themselves to avoidable risk without need. Typically, they would couple by hand instead of using a coupling pole or attempt to couple before the vehicles in question were stationary. Sometimes they would go between moving vehicles to couple them by hand. The inspectors’ reports include many such cases but seldom comment on the reasons for men breaking the rules although sometimes they highlight an absence of adequate rules or instructions.
Mike Esbester identifies wagon couplings and the need for automatic couplings as a similar issue to that with brakes: “It was the one that the unions picked up on, largely because there were possible technical solutions, and automatic couplings of freight stock in the USA was mandated in 1893”. The companies and private owners had large fleets of wagons and the cost of conversion would have been high ‘the companies put it at several million pounds in the 1890s’ and they successfully opposed change even to the present day.
These days, outside preservation, hand coupling and unbraked wagons are things we are little aware of but sometimes incidents involving these features arise. This link, courtesy of Underground News refers to a runaway train on the underground on 13 August 2010. It was reported in the October edition. Prompt and effective action by the signallers and line controllers successfully resolved the incident.
An automatic coupling would help facilitate continuous braking but is not a prerequisite as brake and other connections can be made separately by hand while using a non-automatic coupling. Clearly for traditional freight wagons this would be a big overhead although for modern vehicles in relatively fixed formations, less so. This could explain why you often see empty container flats in trains.
In conclusion we can probably say that the practise of hand coupling has been largely eliminated mainly by the adoption of new ways of handling freight. Much as it pains me to say it, the credit for this probably goes to the reforms introduced by Dr Beeching, sometimes described as Britain’s most hated civil servant. [Controversial Transport Minister, Ernest Marples has escaped with relatively little censure despite business links to road construction.] As an aside, those interested in Dr Beeching’s legacy might be interested in a mock trial that took place some years ago; details courtesy of Railwatch, December 2001 edition.
Accidents about the track regularly expose failures to appoint lookouts or a lookout having an additional job to do. In a lot of cases, men were expected to keep their own lookout although when doing work that required concentration on some feature of track or signalling, so the need for dedicated lookouts is recognised by the inspectors.
Observations and Questions
From over 1200 Excel line entries I have seen nine accidents reported on the London Underground and some of those occurred during conversion of the sub-surface lines from steam to electric traction. Does that mean the underground is safer or less prone to accidents?
If conversion to electrification is a potential cause of more accidents, then other schemes we might look at are the former Southern Region lines, the Mersey rail network and the Tyne and Wear electrification, since dismantled. Also, the Liverpool Overhead Railway, since dismantled which used electrification from the outset. I have not seen any reports to suggest a correlation but I may not have seen reports for the appropriate years.
Glasgow also has an Underground railway taking the form of twin tunnels and a circular route. I have not seen any accidents attributed to it. For many years it had cable-hauled trains and until the 1970s there were no switches or points. Trains were lifted out of the running tunnels into a depot above the line. These distinctive operating characteristics could lead to types of accident not seen elsewhere.
I have not seen any accidents attributed to tramways and light railways. These may be outside the scope of the accident reports and their operating characteristics might bring them into closer proximity to the public resulting in a different approach to safety.
The primary purpose of a railway may also be significant. A railway with an intensive service and frequent trains may present few opportunities for men to work on the track during operating hours and hence few accidents. A lightly-used railway may have few accidents to track workers as there are few trains to avoid. Perhaps railways with moderate service levels are the most dangerous to those working on the track but will the evidence support this?
Many of the accidents involve ‘shunting’, ‘goods handling’ or ‘caught between vehicles’, almost all activities associated with freight. A railway with little or no freight traffic might therefore have few accidents attributable to this cause. The London Underground might be such an example although in the era of these accidents we are considering, it did handle some freight.
Other companies handled a mixture of traffic so any analysis would have to separate the passenger and non-passenger elements. I suspect most of the accidents involved freight in some form but will the evidence confirm this?
Some accident reports include references to live animals. In all the cases I can recall we are dealing with horses used for shunting or haulage of carts. Some reports include horseboxes but none refer to horses or any other animals being loaded, unloaded or moved by rail. Occasionally a horse might be injured or killed in an incident or its behaviour might contribute to it.
I’m not sure what conclusions can be drawn from this. Perhaps the railways were aware that animals have minds of their own and need to be carefully managed when on railway property.
The Railway Inspectors
The Railway Inspectorate traditionally recruited its inspectors from the ranks of the Royal Engineers. In the past, the Army had a significant mileage of railway on its own lands and some of its Engineering Officers would have technical and operational skills relevant to the role. They would also have the man management skills and presence to interrogate witnesses, some of whom might be reluctant to cooperate.
Sometimes the only way to discover the truth is for witnesses to tell all, even if it may be detrimental to them personally to do so. I have not seen any cases in these reports or others unrelated to the project where the evidence of HMRI has been used in criminal proceedings either against companies or their servants and this is probably an essential feature if investigations are to discover the truth and prevent further accidents.
The absence of litigation has consequences. Mike Esbester makes the observation “the inspectors were in most cases polite in their criticism of company managements but I think I detect barely-concealed frustration at times, particularly when you get the same sorts of accidents occurring time and time again, sometimes in the same locations.” I add his words to this post after seeing details of four accidents in three months, all meriting the same recommendation. This grouping was not an isolated example and the frustration continues to be barely-concealed. I was frequently able to copy and paste details from one accident summary to another such was their similarity.
A feature of the inspectors’ reports is that some make reference to similar accidents while others do not. At first sight, this might seem like an inconsistency or double standard but there could be another explanation. Where a company has declined to introduce or enforce rules concerning a dangerous method of working, then the inspectors might feel obliged to express their views firmly. Where the company was seeking to comply and had been let down by its servants, then failures might be seen as isolated cases. The inspectors would know something of the attitude of individual companies towards safety but would not be able to express this explicitly in reports.
Another feature is that inspectors may refer to other accidents they have reported on but not usually to reports prepared by their colleagues. Perhaps they were unable to compare notes in a timeframe to make this possible. There are also detail differences in the style of reports prepared by different inspectors on similar accidents although not so different as to make one question their recommendations.
Mike continues, “the inspectors were stuck between a rock and a hard place, as their remit was to investigate and make recommendations for changes – but as they couldn’t insist upon changes or enforce them, they had to rely upon persuading the companies to take action. Keeping them onside was therefore vital, and no doubt explains some of the cautious or even deferential wording at times. This problem afflicted all of the inspectorates (e.g. the factory inspectors) set up in the 19th century, so it wasn’t confined to the railways.” Companies and their servants could be prosecuted by other means, typically following a police investigation so they did not have immunity.
At the risk of digressing from railway issues, there is a philosophy that any business should be able to take and manage its own risks and the state should in general not intervene. It might be argued that companies have to make a profit and poor safety will eventually impact their reputation and profit. That said, safety precautions are likely to impact profit sooner hence their reluctance to act.
What the Railway Inspectors did was to inform the debate regarding the number, type and magnitude of risks and their impact on real people, many of whom had limited influence over events. To this day, understanding risk is important for any business as some people are naturally risk tolerant, particularly when a risk they take might have bad consequences for somebody else rather than themselves.
A Philosophical Perspective
Digitising the accident records potentially makes them more usable for researchers, particularly those who need to look for trends and find evidence to support or refute ideas. This quotation from Lord Kelvin comes to mind.
‘In physical science a first essential step in the direction of learning any subject is to find principles of numerical reckoning and practicable methods for measuring some quality connected with it. I often say that when you can measure what you are speaking about and express it in numbers you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind: it may be the beginning of knowledge, but you have scarcely, in your thoughts, advanced to the stage of science, whatever the matter may be.’
This project is advancing our knowledge of this subject and aiding measurement. From the numbers will come lines of enquiry and evidence to support or refute ideas.
 1907 Q4 Appendix C, North British Railway, 31/12/1907, Gavel, Report by C Campell.
 1907 Q4 Appendix C, Northern Counties Committee, 31/12/1907, Londonderry, Report by C Campell.
 This merits further analysis and I hope to say more in a future post.
 Four in 1904 Q2 App B, one in 1907 Q4 App C, three in 1912 Q2 App B and one in 1937 Q4 App B.
 Sometimes referred to as the Dockers’ Umbrella.
 1907 Q4 Appendix C.