Don’t scapegoat standards

With big changes afoot for the way railway investment is delivered, alongside constant discussion of “gold plating” and over-specification, is our adherence to standards the problem?

A version of this article also appeared in Issue 852 (9 May 2018) of RAIL magazine.

The way that infrastructure is delivered is changing.

As we draw to the end of Network Rail’s fifth and most turbulent Control Period, lessons are being learned and everyone has new ideas. Third party funding of major schemes is being pressed by the Department for Transport. The Transport Select Committee look like they are going to suggest that infrastructure investment is best delivered as part of a long-term pipeline. Even Network Rail are hinting that they want other groups to get involved in the earlier stages of infrastructure projects.

With the wider role of regulation, standards and control being ever-relevant and perhaps over-discussed (particularly in light of Britain’s possible future outside of the EU), it is worth remembering what standardisation means in the context of an evolving rail industry…

Seeing as the railways were responsible for some of the earliest standardisation (“Standard” track gauge was committed to vellum back in 1846) and some of the first British Standards (our rails still conform to BS11), railway engineers can talk about this subject with some authority.

Over the years, the industry has gained knowledge from research, from real-world experience and (often tragically) from our mistakes. My own discipline, permanent way engineering, has learned several lessons the hard way, from the consequences of cyclical loading on bolted rail joints (Hither Green, 1967) to understanding and predicting the occurrence of rolling contact fatigue (Hatfield, 2000).

In each case we undertake investigations into the causes of these incidents, trial new or improved ways of avoiding them and then record this learning in the form of a standard.

These are defined fairly succinctly by the International Organisation for Standardisation:

“A standard is a document established by consensus and approved by a recognised body that provides for common and repeated rules, guidelines or characteristics for activities or their results aimed at the achievement of the optimum degree of order in a given context.”

Put simply, standards allow us to maintain and improve four key facets of the railway: interfaces (making sure different railways fit together, for example through using the same track gauge), compatibility (making sure the various bits of one railway fit together, such as train footsteps and the platform edge), cost-effectiveness (such as avoiding repeating development work) and most importantly safety (through use of best practice and avoiding known errors).

Standards also conform to a hierarchy. The standards that have the highest legal importance are generally less detailed. Conversely, the most detailed standards are usually non-mandatory.

Mandatory standards include the European Technical Specifications for Interoperability (TSIs), the Railway Group Standards and British Standards/Euronorms.

Before anyone mentions it, I’ll point out that the TSIs are bound into UK law so, even if it wasn’t a clear economic benefit to maintain an interoperable railway, they are going nowhere if we leave the European Union.

The TSIs are also forming the basis of the design of new railways across the world so, if we want to export UK engineering skills, it makes sense to retain this skillset domestically.

“Voluntary” standards are those defined by the infrastructure operators like Network Rail. They still must be complied with to fulfil the operator’s safety case, which usually forms part of the legal arrangement to operate the infrastructure.

No matter where they fall in this hierarchy, standards never stand still and are in constant development. Industry is regularly consulted on them, updates are made and then are continually reviewed. Changes can come from legislation, research, user feedback and where a deviation/derogation is frequently submitted against them.

Engineers commonly input into the ongoing development of these standards; if a standard is out of date or is resulting in unnecessary work, it will likely be updated. The idea that the railway is designed to some static, out-dated rules (with costs soaring as a result) is clearly false, but so too is the idea that we don’t challenge the standards where we see their application as being disproportionate.

However, to suggest alternative ways to manage risk (which is ultimately what a deviation from the standard is) requires engineering experience, which is something that the UK is still sorely short of.

The UK-wide lack of engineers restricts the number who end up in engineering assurance; this lack of experience means that there is little or no flexibility from standards as there isn’t the raw knowledge to understand where risk is being safely managed in a new or different way. Particularly within Network Rail, the engineering skills shortage means that there isn’t the experience to sign off non-standard solutions.

This is a problem across all the technical disciplines, but both electrification and systems engineering have been a particularly obvious challenge during the Great Western route upgrade.

So, given the STEM skills shortage continues to put pressure on the availability of engineering experience, what alternatives do we have?

Enter the Common Safety Method for risk evaluation and assessment.

Whilst it has been enshrined in law since 2013, the railway industry has only recently started sinking its teeth into the useful applications of this process for quantifying risk.

When undertaking design, particularly outside of previously established practice, engineers must account not only for the risk of death but also non-fatal injuries and shock/trauma. It is useful to weight injuries and shock/trauma so that they can be included in the overall risk represented by a hazard or activity.

The resulting measure of harm is known as FWI, or “fatalities and weighted injuries”, and the RSSB’s Safety Risk Model lists the recorded or estimated FWI values against almost all hazards on the railway.

If you combine this dataset with a record of the number of people interacting with a given asset (such as the number of passengers who are on a given platform or who use a footbridge) as well as the government’s recommended value of life (just less than £2m per FWI), then you can start to associate a monetary value against a given risk that allows us to determine the proportionality of an intervention.

Engineers have only recently been getting to grips with this new approach to cost/benefit analysis (in the past, we usually left this to the economists), but it is already bearing fruit in the form of cost efficiencies, with examples including the modification of platforms and safe reduction of electrical clearances.

This isn’t the whole picture of course, and the high cost of infrastructure is a challenge requiring time and strategic thinking across organisational boundaries.

Nevertheless, in the great rush to find new and more efficient ways of building and modifying our railway, standards should not be scapegoated as an alleged barrier against good value. They are simply a way of recording our experiences for use by the engineers of the future.

Without them, there wouldn’t be a functioning railway.