"I would say it's the most complicated moveable bridge I've ever designed, and I've designed a lot of these things," says Tampa, Fla.-based Jim Phillips, lead bridge designer with Hardesty & Hanover LLC (H&H). 

After securing approval and a budget for the crossing, Renfrewshire Council in 2015 recruited Sweco UK to help develop its design and procurement, says the owner's project director, Barbara Walker.

While various types of crossing including tunnels were reviewed, site constraints made a two-leaf swing bridge the best option, says Christopher Cardno, who was on the Sweco project team from the start.

The local port authority required a 90-m-wide navigation channel with 45 m of headroom, potentially to allow passage of offshore oil rigs, he explains. The nearby airport flight path set a ceiling of just over 40 m. And space limitations on both riverbanks ruled out a single-leaf swing bridge, he adds. 

The reference design by Sweco, with architect Kettle Collective, was "gold plated" and likely to break the budget, says Cardno, who is now head of design with Jacobs UK Ltd. Based on that plan, the 2019 bids were "quite significantly over budget," confirms Walker.

Three joint ventures had shortlisted to bid, but one quickly dropped out, to Walker's disappointment, she says. The remaining two were invited to bid for the reference design and also to negotiate a price after a period of value engineering.

In March 2021, a team led by Irish Republic-based John Graham Construction Ltd. won the design-build contract worth $103 million. It beat a joint venture including American Bridge Company and Denmark's Cowi A/S. 

Graham, which declined to discuss the project, subcontracted the steelwork to a joint venture of Netherlands-based Hollandia Infra BV and Iemants NV of Belgium. That team hired H&H for the detailed design, supported by Roughan & O’Donovan Ltd., Dublin.

The bridge halves each sailed for around a week to the River Clyde. Photo courtesy Smulders

With site work about to start, the owner exercised a break clause in Sweco's contract and rebid the site supervision work, awarding it to Jacobs. But "we did have some continuity," says Walker. By then Cardno had moved to Jacobs, taking six years of project experience with him.

Recognizing "a lot of things that made the bridge much more complicated and quite a bit more expensive,” H&H did “a kind off complete redesign," says Phillips. For example, the original 30° skew of the pylons relative to the deck axis was eliminated. And the 30° skew of the deck ends went down to 6°. 

Instead of having seven cable stays on the front and three at the back "we changed it to three and three, which was much more efficient and also balanced the loads in the pylons," says Phillips. 

The decks were narrowed, and the pivot positions moved 2 m forward to reduce the counterweights, making the whole bridge lighter. It also "contributed to reducing the maximum overturning load on the slewing bearing," he adds.

"In a lot of swing bridges, the bridge is balanced so the total dead load center of gravity is right at the pivot point," explains Phillips. But in this case, unbalancing the structure was found to be optimal. "It's part of a process we went though to balance all the loads on the slewing bearing in the various conditions."

In the final design, the deck comprises 2.2-m-wide edge beams which are generally around 2.4 m deep. Deck cross girders are 750 mm deep and double that at cable stay locations.

At each bearing, the deck structure comprises a full width steelwork box, which is roughly 3.7 m deep and 10 m long. Internal stiffening plates connect the box to the edge girders and pylons. And at the center, a 6.7-m-dia cylindrical diaphragm mirrors the bearing beneath. A 50-cm projection of this diaphragm below the box's flat bottom is bolted to a slewing bearing, which is anchored within a concrete "pivot pier” on the riverbank.

Diagram courtesy Hardesty & Hanover

Because of space limitations, Hollandia and Iemants each built one complete half the bridge in their respective home workshops and shipped the structures to the site. In week-long journeys, the south half was barged from Rotterdam this April and the north one followed from Flushing in May.

To move the roughly 1,600-metric-ton, 38-m-tall, 18-m-wide and 92-m-long structures, heavy lift specialist Sarens NV used a combination of multi-wheeled self-propelled modular trailers (SPMT) and floating sheerlegs. "Site conditions in Rotterdam and Flushing required different SPMT configurations due to space and ground pressure constraints," recalls Kenny Decoster, Saren's technical solutions leader. And "extensive ballasting was necessary to compensate for tidal and load variations during operations in Glasgow." 

The south bridge section was rolled to the Rotterdam quayside on six 10-axle SPMTs and loaded onto a 122-m-long barge by three floating sheerlegs. In Flushing, site conditions allowed the structure to be rolled directly onto the barge on 32-axle SPMTs. Sheerlegs and SPMTs were used to place the structures onto their bearings when they arrived on the Clyde.

A bridge half being rolled onto its bearings. Photo courtesy Roughan & O’Donovan

For the sea journey, the contractors stressed the cable stays to about 80% of the final values, says Phillips. "We left ourselves the ability to make the final adjustment ... on site," he adds. 

To accommodate temporary and different loads during transportation on SPMTs and lifting with sheerlegs, "the leaves included diaphragms and strengthened webs," Phillips adds. "Both of those conditions are quite a bit different from the final condition." 

The successful marrying of both bridge halves has been "pleasantly surprising," says Walker. And following intensive value engineering, the crossing remains close to budget, allowing for design changes, she adds. "We're expecting it to open [this] October, and as far as we're concerned that's pretty much on target."