The Weekly Reflektion Week 39 / 2019

This week’s Reflektion is inspired by the Tay Bridge disaster in December 1879, which was the subject of a previous Reflektion in week 52/2017

The Tay Railway Bridge under construction looking south from Dundee. Dundee District Libraries, Photographic Collection

Are you aware of the criticality of the assumptions you make? Which performance indicators are you monitoring?

The Tay Bridge disaster occurred during a violent storm on Sunday 28th December 1879. The train from Wormit to Dundee was passing over the bridge when the bridge collapsed, killing the 75 passengers and crew onboard. The Tay Bridge had been opened on the 1st June 1878 and was passed as safe by the UK Board of Trade.

Faults in the design, construction, commissioning, maintenance and operations were found in the inquiry following the disaster with clear parallels with well construction in today’s oil industry. The bridge designer asked advice of other prominent engineers with regard to current best practice. In the UK, a sufficient allowance for wind pressure was considered to be 10 lbs/ft2. In the USA, a figure of 50 lbs/ft2 was used for the Brooklyn Bridge, and in France, 55 lbs/ft2 was used for the Eiffel Tower. Checking what others are doing, sometimes camouflaged as benchmarking is not unknown in the well construction industry. It’s important who you choose to benchmark yourself against.

Soil borings were made to determine the parameters for the foundations. The conclusion was that the foundations were to be set on ‘solid rock’. This ‘solid rock’ turned out to be a thin layer of conglomerate overlying more mud. The contractor had not worked in similar estuarine environments and lacked the necessary competence. Wells have been lost due to poor foundation of the wellhead. Do you automatically assume contractors have the necessary competence to gather the necessary data?

The change in foundation conditions led to a forced radical change in design.  The base of the foundation caissons was increased to spread the load and the load was reduced by moving from stone to iron columns for the piers. To compensate for the increased cost, the span of the girders was increased in order to eliminate one pier. Change management is also a significant challenge for the oil industry. A well design could take several man-months, but a new design may have to be prepared at very short notice.

The iron that was cast for the bridge columns was of poor quality. Flaws were regularly disguised by filling with a mix of beeswax, and iron filings for the sake of appearances. The wrought iron cross bracings were described as ‘best’ iron. ‘Best iron’ was actually the lowest of 3 grades, the other 2 being ‘best, best’ and ‘best, best, best’. Do our procurement requirements lead us to accept technically inferior quality?

The lugs to which the cross bracings were to be bolted had conical holes. Instead of drilling out the holes, undersized bolts were used. In addition to this, the construction gangs had no instructions on what torque should be used to make-up the bolts. Play in these bolts probably contributed to fatigue degradation in these lugs. How do we ensure changes are managed properly?

During the commissioning phase, a statement was made that the inspector “should wish, if possible, to have the opportunity of observing the effects of a high wind when a train of carriages is running over the bridge”. Yes, if only. Do we check that our equipment has been tested under relevant conditions?

The inquiry found that the maintenance inspector knew little of iron construction, and there were no instructions for the inspection of the iron components. During operation, the trains exceeded the speed limits that were set, and this contributed to weakening the bridge. The inquiry found that the North British Railway had no effective control of the construction or the operation. Passengers began

complaining and trading their season tickets in to return to the ferry boats due to ‘disturbing signs of weakness’. Some travellers on the Tay Bridge felt that the trains had a ‘prancing’ motion, yet no procedure existed to draw management attention to the problem. It would be interesting to know what performance indicators the railway company were using to monitor performance, in particular technical safety. Perhaps it was only the financial bottom line. Which performance indicators do you use? What are you actually trying to achieve with monitoring these performance indicators?

According to the inquiry, the project was sold on an underestimate of the cost, the contractor underbid the work to make sure he landed it, and the railway directors believed it all, because they wanted the bridge built. Time and cost pressure during the execution phase led to a set of incentives to get the bridge finished on time. Focus was on the financial bottom line. Is this something you have seen in any of your projects? Perhaps, some things never change.

It is interesting to note that causes of major accidents in one industry are just as relevant for the prevention of major accidents in others. All the more to learn folks.

Ref; Reflektion week 52, 2017 SPE 56924 Well Construction, Competency and the Tay Bridge Disaster

Reflekt AS