£4m to design white elephant

Thomas Heatherwick's proposal for a "garden bridge" across London's Thames has been in the news again. BDOnline reports that Transport for London (TfL) will spend £4m to develop the bridge design through to planning consent stage.

I hope you were sitting down when you read that. Four. Million. Pounds. For the preliminary design of a concrete flowerpot, to be plonked in the middle of London with as little sensitivity as is imaginable.

It's the sort of sum that may well be applicable for a major bridge, but to me seems well out of proportion for a footbridge, particularly one where there is no compelling need for a crossing (it will be barely 250 metres away from the nearest existing Thames bridge, Waterloo Bridge), and there's nobody yet willing to pay for its construction. The hope is that private donors will pay for the entire project, with TfL committing nothing beyond the initial development costs, and carrying the risk that if the scheme goes nowhere, that money is simply wasted.

TfL's £4m figure comes from a committee paper available online, which reveals April 2014 as the target date for the planning application, meaning they must burn through roughly half a million pounds a month in the feasibility stage. The £4m isn't just for design development, it also covers consents fees, public consultation, and establishing a charity whose aim is to raise private funding for construction and maintenance. Nonetheless, it's a staggering expenditure, and even more eyebrow-raising when you note that this is a proposal which was never in TfL's business plan, but for which money has had to be found largely to satisfy a passing mayoral whim.

Maintenance alone is expected to cost anywhere from £3m to £5m, on top of an initial construction cost of from £60m to £100m. That would make this, I am pretty sure, the most expensive pedestrian bridge every built, by some way, three, four or even five times as expensive as London's Millennium Bridge, and trumping even the costly Gateshead Millennium Bridge on a pound per metre basis (flowerpot: up to £400k per metre; blinking eye: about £190k per metre, allowing for inflation).

These are only initial figures, with more accurate estimates anticipated at the end of September.

I find the entire idea beggars belief. This is a bridge with no real purpose (the additional green space created is peanuts relative to Green Park, 20 minutes walk away); which nobody (yet) is willing to pay for (other than the £4m of public funds to be washed down the drain); for which the likely cost must outweigh the likely
value on any conceivable scale; which would almost certainly become a public liability at some point in the future; and which judging from the images created so far looks like a massive blot on the landscape (probably more so when the scale and impact of the approaches at either end become apparent). If this doesn't meet the definition of a white elephant, what does?

Even in a time of austerity, there should be room for a little frivolity, a larger amount of fun, and for things that offset gloom with positive vision. However, if private donors really do have £60m or £100m to spare right now, it's hard not to think of more useful ways of spending their money.

Updated 30th July: Gateshead bridge costs amended (see comments).

"Pedestrian Bridges" by Andreas Keil

This new book on the design of footbridges is part of Detail’s Practice series, a line of guidance notes aimed mainly at architectural designers. Pedestrian Bridges (Detail, 112pp, 2013, also available in a German edition) [amazon.co.uk] is aimed at bridge designers generally, with plenty of material aimed at engineers as well as architects.

The lead author, Andreas Keil, is a prominent bridge designer with Schlaich Bergermann und Partner, responsible for several highly innovative designs, such as the Passerelle La Défense in Paris. Although this book includes many SBP designs as illustration, it is a general text rather than one which examines in detail the author’s own experiences.

The book has two obvious predecessors. The first is FIP’s Guidelines for the Design of Footbridges, edited by Keil's colleague Mike Schlaich; and the second is Footbridges by Schlaich with Ursula Baus. Schlaich and Baus’s book is not a design guide, but an extended review of examples of interesting structures from around the world, although in its attempts to explain and understand the work of widely-varied designers, it does offer considerable assistance and inspiration to designers of footbridges. The FIP Guidelines, on the other hand, cover very similar territory to Pedestrian Bridges, so I'll return to them in a bit.

Keil’s book attempts to be comprehensive, with sections on functional requirements, statics and dynamics, materials, structural types, finishings, economics, special bridges, and a short gallery of project examples. This is the engineer’s approach to bridge design, placing the options for response to functional demands at the centre stage. Architectural issues such as aesthetics, context, the pedestrian’s experience of the bridge, form a persistent subtext throughout.

The chapter on functional requirements addresses issues which are frequently set out in national standards, such as minimum width, clearances, ramp gradients, staircase geometry, and parapet height and strength. In the UK, the relevant standard is BD 29, which is freely available online. Most of Keil’s commentary is, however, specific to Germany (with some reference to other international requirements), and I think it’s mildly unfortunate that in an English translation the opportunity has been lost to better internationalise the text. The wide variation in specifications around the world is a worthwhile topic, as it illuminates the arbitrariness of many individual requirements, and helps arm the creative designer if they should ever dare to test the boundaries of regulation.

The same issue is apparent in the section on static and dynamic performance (including aerodynamics). Although the book briefly compares the different loading standards, the key reference for pedestrian-induced vibration is the German HIVOSS approach, which is far from the only available guidance, and which has been reported as being more conservative and less versatile than the British approach (embodied in the National Annex to the Eurocodes). Despite industrial quantities of technical paperwork on the subject, dynamic analysis of pedestrian bridges has been a complex and specialist area, riddled by incompatible and sometimes incomprehensible research. The designer seeking guidance in this area will undoubtedly need to seek more in-depth advice.

The book’s chapter on design and construction is primarily a discussion of different structural forms, but opens by “reviewing and clarifying the tasks involved”, a consideration of the constraints to be documented and analysed before the opportunities for design can begin to be explored. Constraints include matters such as the site topography, geotechnical conditions, and the required bridge useage. This section is quite brief, which is a shame as this is a key area which is covered inadequately in most books on bridge design. Constraints arise not only from the site itself but from the client’s remit (including budget), from the demands of communities and stakeholders, from local contractor capability and resource, and from environmental limitations. For me, a thorough analysis of design constraints is a vital first step in creative design, and there is much more to be said than Pedestrian Bridges provides space for.

The discussion of different structural forms is thorough and clear, and as with the rest of the book, well-illustrated. As you might expect, Keil’s understanding of structural behaviour is acute, and the many diagrams and photographs both help explain the text and expand it by offering examples of inspirational design. As well as the more obvious forms, the book gives generous space to stress ribbon bridges, and to a discussion of the particular issues for curved bridges (although this is one case where the generally fine translation falls, with a confusing discussion of “carding” moments, which it took me some time to deduce meant torsion moments).

One question which readers may ask is whether this is a book aimed at the novice or at the initiate. I think its audience is clearly the former, but it contains much that will be of interest to the latter. I found a great deal that was new to me in some parts of the book (I won’t embarrass myself by saying which parts!) Other parts also helped to encapsulate and put structure to things that I have learned as a designer, but never allowed to coalesce from their varied and fragmentary project landscape.

Although Pedestrian Bridges covers very similar territory to the FIP Guidelines, there are some key differences. Keil's extensive discussion of materials and structural forms is entirely absent from the earlier document, which focuses primarily on performance requirements rather than design opportunities. Keil's book also has a useful section on economic aspects, with a survey of the costs of 22 footbridge examples given for comparison (not the first such effort). Both books offer a selection of case studies, with more in the FIP book, but no overlap between them.

In summary, for anyone inexperienced in pedestrian bridge design, this is a very helpful publication. Even experienced designers should find it a useful addition to their bookshelf. It is very well illustrated throughout, and clearly written and translated.

Sunderland's River Wear Crossing: A Post-Mortem

Okay, while the dust slowly settles on this week's announcement that Sunderland's hubristic River Wear Crossing scheme has been cancelled (in its current form, at least), I guess it's time to reflect.

This was a project conceived a few short years after the glitz and the glamour of millennial bridge fever, which saw Sunderland council looking with particular envy at their long-time rivals in Newcastle / Gateshead. The city of Glasgow were another local authority infected by the fever, although in September 2005, when Sunderland announced that their bridge design competition had a winner, it had not yet become apparent quite what a fiasco Glasgow's scheme would become.

The parallels with Glasgow's project are highly pertinent, with both competitions won by designs which were immediately and obviously structurally "challenging"; which were pursued with vigour by their promoters regardless of criticism; which received only two willing tenders, both well above the available budget; and which were binned without a hint of apology for the millions wasted in project development, to be replaced by simpler design-and-build projects where risk was eliminated by letting contractors take the lead.

That's what is to happen now to the New Wear Crossing - a new design-and-build tender stipulating the same alignment, same general principle as the original design (i.e. a cable-stayed bridge), and even the same foundation positions (to avoid having to re-assess any impact on river hydraulics, and land purchase issues, presumably).

Sunderland announced Spence Associates and Techniker as the winners of their RIBA-run contest to design a £43m highway bridge as part of the wider Sunderland Strategic Transport Crossing in 2005, but it would be three years before the winning design would actually be made public.

Quite why it was kept secret has never really been made clear, although I know that the design was subject to independent review from beyond the jury panel and the client's ordinary technical advisers. There was clearly some recognition of the bridge's highly radical departure from the structural engineering norm at the outset (the reviewer concluded that "construction of the bridge is challenging but achievable"). Even when, in November 2008, Sunderland took a positive decision to move ahead with the iconic design, they were still making clear their concern over its feasibility and talking about the possible need to "tone it down".

When the bridge design was finally unveiled, I commented immediately that "The iconic bridge design is simply amazing. Amazingly beautiful. And amazingly daft." Some of what was stated by its designers at the time was hugely misleading e.g. Techniker's patently absurd claim that "the bridge was not a technically difficult or radically new structure", or that a bridge "of similar size and span and of exactly the same construction and arrangement has been constructed a decade ago without any difficulty".

These were not the last statements made which seemed at odds with reality, with Sunderland council on several occasions reporting that the bridge was well-supported by the public, on the basis of a tiny survey which was, in fact, contradicted by a larger survey showing very nearly half of those questioned would have preferred a "tried and tested" design alternative.

In September 2008, I offered this summary of the issues with the design, and I do take some satisfaction from being able to say now that I told you so:

"Amongst all the consideration of cost versus quality that will probably dominate discussions, I think it's important not to forget about risk. Whether it will stand up or not, for a given budget or otherwise, it's undoubtedly a very high-risk proposal. And Sunderland might be wise to look at recent examples from Glasgow and Stratford-upon-Avon for cases where high-risk competition-winning designs were dropped only after first wasting considerable sums of public money."

Following the council's 2008 decision to invest funds in developing the design further, considerable effort was spent first working through the design problems and then preparing a full detailed design, with a number of technically experienced consultants brought alongside Techniker to provide assistance. The cost estimate soon rose well above the original £43m, and it became apparent that central government would only provide Sunderland with enough funding for a conventional bridge alternative, not for an iconic structure.

Nonetheless, the council eventually succeeded in putting together funding to match the cost estimate, and hence to put the highway project (featuring the bridge) out to tender. Indeed, they succeeded in getting the overall project budget down from £133m in 2009 to £118m in 2011.

The £118m estimate was based on a "mean probability" risk allowance, i.e. the council estimated what risks might push their costs up, and allowed for precisely 50% of them, in line with standard government guidance (the P50 risk level). They calculated that if all risks were to materialise in full, the project cost could rise to£130m, and that they would need to make financial provision for this eventuality. Sunderland have now removed the project website, so all their financial reports and other data are no longer shared with the public, although you can still find them, for now at least, via Google's cache.

Signs that the contractors' views of risk differed from the promoter's emerged in February this year, when two of the four firms bidding to build the scheme decided not to return a tender. We are told now that neither of the remaining two tenders returned "were priced within the current funding availability and approved budget for the scheme", and that to pursue them would require "significant additional funding".

What we have not yet been told is precisely what the tender figures are, i.e. to judge whether the original budget estimate or its risk allowance were ever, in fact, realistic. The bidders knew the available budget, and presumably gambled that Sunderland would not wish to lose face and abandon the scheme, finding additional money somehow.

So we are left with a number of unanswered questions:

• What were the tender prices from the two bidders? By how much did they exceed the budget estimate?
• How much of the cost over-run relates to the bridge rather than other aspects of the scheme?
• What was wrong with the budget estimate - what did the bidders price that was not in the estimate?
• How much money has been spent so far on developing the current scheme?
• Looking back to the original competition, why was this design ever chosen?
• Who or what is to blame for the decision to repeatedly press ahead with a project where the risk of failure was ever-present, and never sufficiently mitigated?

By my reckoning, the bridge amounts to some £77.5m of the £118m estimate, once you share out project management costs, land purchase, risk and indexation in proportion to actual construction costs. That's some way beyond the original £43m assumed at competition stage, and we must assume that even that figure has been blown past by the actual tender prices.

The financial reports for the bridge stated a figure of £13.3m for "fees" for the project, including items like consultants, designers, public consultation, licenses, consents, surveys, project management etc. Some of this will, I guess, relate to future fees e.g. for project supervision during the construction phase. But most of those headings relate to activities which should already have been completed, and hence I'd guess perhaps some £10m has been spent on developing the scheme to its current state, and the majority of that completely wasted on a design which will never be built.

Looking back to the start of the whole process, I recall emailing Sunderland in 2008 to ask for a copy of the competition jury's minutes, and never received a reply. I think there is much to be said for transparency, and to the opportunity it offers to receive criticism gracefully, and to benefit from it. Perhaps future design competition organisers could consider that.

It looks to me like Sunderland Council will wish to rapidly move on: to put this fiasco behind them and focus on a new, less challenging, possibly far more context-appropriate replacement project. But if others are to learn from this unhappy experience, I hope that answers to the above questions do emerge.

Ding dong

I will hopefully have more to post about this at the weekend, but for now:

Sunderland City Council scraps plan to build £118million bridge across the Wear

Bridge collapse! The rise and fall of Sunderland’s ill-fated Iconic River Wear crossing

Scared off? Sunderland’s landmark bridge architect Stephen Spence reveals his biggest regret over project axe

Sunderland scraps costly Wear bridge plans

£118m Iconic Sunderland bridge was a “ridiculous waste of money” blasts councillor

The official council report can be found online, although it offers few answers to the key questions such as: just how much over budget were the two tenders submitted? How much money has been spent on the abortive design?

Bridges news roundup

Alfenz Bridge / Marte Marte Architects
Intriguing concrete truss bridge with truss diagonals which vary geometry according to the intensity of the shear force.

Schanerloch Bridge by Marte Marte Architects
A bridge with a twist, simple and sculptural.

£6.6m footbridge swings into action
It has been a long time coming, but Hull's "iconic" swing bridge opened to the public in mid-June, with a formal opening ceremony on June 27th. It has already been dubbed a "bridge to nowhere" by one critic. Elsewhere, reasons reported for the lengthy delay in its completion include having to raise the bridge by a mere 6 inches; the bridge's "tracking system" being too small; and concern over vandalism of the windows in the (currently empty) retail unit that sits on the bridge's pivot. One press officer grumbles "personally, I hate the word iconic".

Bay Bridge troubles began with design
Lengthy story of woe blames acceptance of inappropriate bolt specification for failure of Bay Bridge bolts.

Analysis: River Torrens footbridge in Adelaide cost higher compared with Helix footbridge in Singapore
Explain please. The Torrens bridge is not architecturally spectacular, nor does it seem to be structurally anything special (indeed it looks quite dull in most the images I've seen). So how can it possibly cost more than the structurally highly complex Singapore Helix bridge? Is it all down to its "waterfall feature"?

Contemporary Bridge: Berlin
This is an "ideas" competition for a 90m long "iconic" or "landmark" footbridge in Berlin. There are financial prizes, and some glory to be garnered, but essentially it's just a bit of fluff to keep workless architects occupied for a week or two, nobody has any plans to build any such bridge.

Palo Alto to consider design contest for 101 bike bridge
An invited competition is planned to "vet many designs simultaneously in the least amount of time and funding".

Designs unveiled for Lincoln High Street level crossing bridge
It took me a little while to work out where the bridge is in this story's visualisations, but it is there. It strikes me as massive and disproportionate to the issue it tries to address.

Scotland's Oldest Bridges
This is what the internet was invented for!

Pont Jean-Jacques Bosc bridge design competition

The lengthy design contest for a new 110m euro highway bridge over the River Garonne at Bordeaux is heading for its concluding stage. Launched in June 2011, five design teams were shortlisted in November of that year. The local municipality has recently agreed to enter negotiations with the two top-placed teams, featuring Rem Koolhaas and Dietmar Feichtinger. A final winner is expected to be announced in December this year, with bridge construction taking place from mid-2016 to late 2018. Since the scheme has been consulted on since 2009, this makes even the snail-like pace of project development in the UK seem less bad.

This is a substantial construction project, as the river must be about 550m wide at the location of the bridge, and all the designs exhibit a very generous deck width, particularly given that the actual highway itself is not that large and most of the width is given over to footway and cycleway provision.

Here are the five shortlisted designs:

Setec TPI / Marc Barani

From the images, it's hard to see why the bridge needs triple masts at each of its four support piers - the walkways are exceptionally wide, but it looks like a more conventional twin masts or portal pylon solution could have been chosen. That said, it's an attractive looking design at first sight.

One thing that is unclear is how the spans are stabilised, as none of the usual solutions adopted for multiple span cable-stayed bridges are clearly visible (very stiff towers - Rio Antirrio; intersecting cables - Forth Replacement Crossing; tower ties - Ting Kau; very stiff deck - Mersey Gateway). Without one of these measures in place, the bending in the deck and towers becomes excessive, as the load pulling down one span (and pulling its towers together) is not well resisted by the adjacent spans being pulled upwards.

Marc Mimram Engineering

I like much of Mimram's proposal, especially the first image above with its elegant approach span arches and gently sloping footways. However, the main-span arches feel confusing to me, particularly the way they penetrate the deck and spring from below - the relationship between arch and deck seems to lack any structural clarity.

The choice of slightly fanned quadruple arch ribs also seems to offer little in either architectural or engineering terms. The ribs will do little to provide lateral stability to the arch, which will be governed by lateral bending at the deck level or at the arch supports. As with so many competition designs, there seems to be a pursuit of originality purely to make the bridge unique, not for any other justification.

Dietmar Feichtinger / Schlaich Bergermann und Partner

This design bears more than a passing resemblance to Feichtinger's Passerelle Simone de Beauvoir, which also features undulating walkways separated in level from the road deck, held apart with a series of steel struts. Unlike the lenticular Paris bridge, the Bordeaux proposal does not derive so directly from the structural form, allowing the spatial requirements to take precedence. At the middle of the bridge, the depth between walkway and roadway becomes sufficient that a partly-sheltered pedestrian plaza is created by joining the two walkways together below the roadway.

While the desire to give pedestrians a variety of different experiences is laudable, the effect overall is to make the bridge significantly deeper than it really needs to be, and I wonder whether the benefit makes this worthwhile. Road users are obliged to pass over a mid-river hill, obscuring views of their destination.

OMA / WSP

Dezeen has more to say on this design. From the images provided, it's difficult to say much about the structure, which seems to be self-consciously minimal, a beam bridge on a series of pier columns. I am told that the deck is a simple concrete slab on steel plate box girders, designed to be easily launched across the river.

As the images make clear, the emphasis is not on the bridge but on its flexibility of use, a huge tabula rasa which can be reconfigured for different traffic modes, events, fairs etc. This seems quite admirable, but it still seems a surprise for  a bridge where the architecture is so quiescent to be considered a potential competition winner.

RFR / BET

The final shortlisted entry is the most structurally expressive, a multi-span suspension bridge which seems to me to be poised carefully between restraint and flamboyance. The challenge of providing stiffness in such an arrangement is addressed by using "A-frame" towers to carry the cables (traditional vertical towers work well for a three-span suspension bridge, but less well for more spans).

The towers are inclined laterally which creates interesting cable profiles but seems to me to add an unnecessarily jaunty air to what is otherwise a straight, non-skewed crossing of the river. The deck widens out at the tower positions, creating large balcony-type areas. I'm reminded a little of the unhappy Golden Jubilee Bridge in London, although this is certainly a more elegant design. My main concern is simply that this is quite an odd choice of solution for such a wide bridge. The suspension system will normally permit a very slender bridge deck, but here its width will govern its depth. It is also likely to be many times more expensive than some of the alternatives, as this is just not an economic construction technology for a highway bridge of these spans.