In 1991, the Missouri and Illinois departments of transportation began voicing concern about the flow of interstate traffic between downtown St. Louis and St. Clair County, Illinois. Part of the solution was to build a new interstate bridge over the Mississippi River. By 2001 the two departments had determined the ideal location for the new bridge, and they had also invested money and resources securing a record of decision regarding the project from the Federal Highway Administration that would enable them to proceed. The project had already endured its share of starts and stops, and just as it seemed to have cleared many of the hurdles common in today’s large, complex transportation projects, the bridge’s design and construction price tag—$451 million—arrived.
The engineers wanted to give the region growing room with a grand design that would have eight lanes and a 2,000 ft span. (See “Long Cable-Stayed Span Planned for St. Louis,” Civil Engineering, November 2001, page 18.) Even with both states contributing, however, the funding gap amounted to $145 million. Without a life-saving injection of federal funding, which was becoming increasingly unlikely, it would be impossible to close the gap. It was evident that the two owners couldn’t have everything on their wish lists, but that was not reason enough to scrap the project.
The region needed help. Congestion had grown like a slow, suffocating cancer over the past 50 years. The Congressman William L. Clay Sr. Bridge, locally known as the Poplar Street Bridge, was burdened with the responsibility of carrying traffic from three interstates—55, 64, and 70—across the river, the only urban bridge to carry three interstates. As the only urban interstate bridge between the two states, the 1960s-era structure was fighting to keep pace with demand. Traffic volumes had increased substantially even during the previous 15 years, and the bridge’s level of service, according to criteria given in the American Association of State Highway and Transportation Officials’ A Policy on Geometric Design of Highways and Streets, or “Green Book,” had fallen to the second-lowest level, E, meaning that traffic flow had become irregular, that speeds varied rapidly, and that there were virtually no usable gaps in which cars could maneuver within the traffic stream.
Refusing to give up, the owners agreed to rescale the project by focusing on what was essential and forgoing the rest. As they dissected their original vision, they scaled back their goals to three while preserving the overarching objective of increasing safety and mobility. The retained goals were to relieve congestion on the Poplar Street Bridge, to realign portions of I-70 and numerous local roads on both sides of the state line, and to deliver the project within the available bistate funding of $306 million. HNTB, of Kansas City, Missouri, the owner’s representative on the original design and a guide in the reenvisioning process, became the principal designer and a member of the design/bid/build team responsible for making the seemingly impossible possible.
In the final, formal agreement between the two states, signed in 2008, the states agreed to jointly fund the main span and two approaches. The Missouri Department of Transportation (MoDOT) would contribute a lesser amount but would guarantee the top-end price, pledging to pay 100 percent of any overage. The MoDOT also assumed primary responsibility for the project and became the agency in charge. To help HNTB develop a design within the budget, the owners included the following bridge specifications in the agreement:
- A cable-stayed design, the most cost-efficient bridge type for the area;
- A main span of 1,500 ft, determined by the U.S. Coast Guard for purposes of ship passage, making it the third- longest cable-stayed span in the United States;
- The same alignment and pier locations as indicated in the record of decision for the original design;
- Four lanes of traffic, two in each direction, with the ability to restripe the pavement to create three lanes in each direction if needed.
The Mississippi River bridge would be the first built in the downtown St. Louis area in 50 years. To deliver it, the MoDOT wasted no time, putting the cable-stayed unit and the two approaches out for bids as three separate projects. The goal was to have the main span under construction as soon as possible, since this portion would take the longest to complete. Key project staff members, as well as the designers of a new interchange and the bridge approaches, were located in the same office to speed communication and facilitate decision making.
HNTB was well aware of the wind effects, river currents, and soil conditions that could be expected along the Mississippi. In recent years the firm had completed three bridge projects on the waterway: the widening of New Orleans’s Huey P. Long Bridge, for which the firm provided erection engineering (see “Broader Appeal,” Civil Engineering, October 2013, pages 58–63); the cable-stayed bridge at Cape Girardeau, Missouri; and the cable-stayed crossing at Greenville, Mississippi (see “Cable-Stayed Span to Cross Mississippi River,” Civil Engineering, June 2001, page 30). The MoDOT took advantage of that expertise to shorten the design phase.
Facing an aggressive (one-year) deadline to complete the new design, HNTB’s engineers began work in July 2008 and in 12 months developed plans that typically would have taken two years. Not only did the scaled-back design at $244 million meet the owners’ goals and budget, but also the speed at which the engineers delivered it saved the project $1.4 million per month in inflation costs. Knowing the MoDOT’s reputation for embracing innovation, HNTB incorporated two new approaches in its design. The first gave the MoDOT more flexibility in how it assesses a bridge’s response to seismic activity. The second was the decision to use steel anchor boxes rather than concrete corbels. The seismic innovation was based on a new approach: the conditional mean spectrum method. This method of evaluating the effects of seismic activity on buildings has long been favored by California architectural engineers. The approach considers the most expected (or mean) response spectrum of a structure under specified ground motions under the assumption that the occurrence of a target spectral acceleration value at the period of interest is the most likely scenario of earthquakes in the region.
Confident that this method had evolved to the point that it could be a valuable tool in bridge design, researchers at the University of Illinois at Urbana-Champaign and the University of California at Berkeley recommended it for this project. With approval from both owners and from the Federal Highway Administration, HNTB moved forward using this method, with great success. This inaugural application of the process for a bridge helped the designers produce a more economical design. Since the engineers had a better idea of the realistic demands on the structural system, they could avoid the unnecessary costs often incurred on highway bridges through the aggregated design spectrum.
Using time history records for the site in question, HNTB carried out a time history analysis of the proposed bridge. Because the new method did not call for the risks from the Wabash Valley and New Madrid seismic zones to be combined as an aggregate event, as would have been the case with an aggregated approach, HNTB was able to reduce the seismic loading that had to be considered in the design while maintaining the required level of safety. It was also able to reduce the size of the structural components and, consequently, the amount of rebar required and to avoid an expensive ground improvement program. The second innovation, the use of steel anchor boxes instead of formed concrete corbels inside the bridge’s pylon legs, was a first for HNTB in its cable-stayed designs.
The engineers knew that the tedious job of creating concrete corbels would slow the contractor and force crews to do more work high above the river. Steel anchor boxes would be both safer and more precise because they could be fabricated in a shop. Each steel anchor box was designed as an individual piece with vertical dimensions of 6 to 9.5 ft, which allowed a continuous shaft of boxes from the shortest cable to the longest, at the top of the pylon. To improve constructability, the anchor beams were bolted, not welded, to the boxes. Holes were drilled in the box walls with digitally controlled equipment, increasing the level of accuracy. As a result, the contractor was able to pretie the rebar and anchor box assemblies in the units while they were on a barge.
The work was thus done at a height of 20 ft rather than at several hundred feet above the water. Ultimately, the steel anchor boxes saved time, reduced costs, increased accuracy to accommodate the tight tolerances of the cable geometry, and reduced the amount of posttensioning needed along the perimeter of the pylon legs. This was the first time that a bridge in the United States was completed on the basis of the project delivery method known as design/bid/build alternative technical concept (ATC), and it was the first time that the MoDOT had used the method on a large-scale project. Like a preapproved value engineering proposal, the ATC invites innovation when design/build or other alternative procurement methods are prohibited.
Under the ATC process, each of the four prequalified contractors was handed HNTB’s baseline design and asked to confidentially propose changes that would meet or surpass the requirements in the request for proposals. HNTB’s design specified the following:
- Footings: The nominal depth of the water was 30 ft. There was approximately 40 ft of overburden, mostly loose to moderately dense sand and bedrock and high-quality limestone. The rock had a quality designation ranging from 90 to 100 percent; a recovery ratio—the percentage of the rock core retrieved during testing—close to 100 percent; and an unconfined compressive strength higher than that of the concrete specified for the rock socket. On the basis of these geotechnical conditions, HNTB designed the foundation to feature not dredged caissons but an array of 10 ft diameter drilled shafts roughly 85 ft deep that would be socketed into rock. This decision considered both the proximity of the bedrock to the surface and the cost.
- Piers: The concrete anchor piers would be supported on four 10 ft columns extending below the surface and socketed into the bedrock.
- Approaches: The Missouri approach would comprise concrete girder spans varying in length from 110 ft to 144 ft that would abut the cable-stayed back span on the Missouri side. The Illinois approach, which had to cross above 21 railroad tracks, would comprise spans of steel plate girders ranging in length from 172 ft to 336 ft. Both of the approaches and the cable-stayed unit would have expansion bearings.
- Deck: The bridge deck, which was to be 94 ft wide, would be posttensioned with high-strength bars (as specified in ASTM International’s standard A722 [Standard Specification for Uncoated High-Strength Steel Bars for Prestressing Concrete]) 13/ 8 in. in diameter in the panels to reduce the potential for cracking. The deck would carry four vehicular lanes with full shoulders. (No pedestrian or bicycle lanes were included.)
- Superstructure: Steel plate girders were to be made composite with the concrete deck with cast-in-place closure strips. The edge girders would be made out of high-performance 70W steel (as defined in ASTM International’s standard A709 [Standard Specification for Structural Steel for Bridges]). The bridge’s floor beams were to be predominantly high-performance 50W steel, but some of the beams would be made of high-performance 70W steel. Weathering steel with a 100-year life was specified to lower the cost of maintenance and materials and to reduce the need to disrupt vehicular and boat traffic for painting. All floor beams were to be made composite with the concrete deck. The 87.4 ft long floor beams were to be spaced 14.1 ft apart on center in the back spans and 14.5 ft apart on center in the main span. The cables would be spaced 42.3 ft apart on the back spans and 43.5 ft apart on the main span.
- Pylons: Twin delta-shaped pylons 405 ft tall were to rise from the river, their triangular structure providing better aerodynamic behavior than offered by H-shaped pylons. The pylons were designed for seismic demands and for a vessel impact with a force of 6,900 kips. The tows considered in the design were a hopper barge three units wide by five long and a jumbo tanker barge two units wide by four long.
- Overall length: The bridge stretches for 1.2 mi with back spans of 635 ft and a required main span of 1,500 ft.
Two of the four contractors accepted the MoDOT’s challenge to improve the design. Each of the alternative designs they submitted was approved by the MoDOT and then refined by the contractor with HNTB’s assistance. In November 2009 the contract was awarded to MTA, a joint venture of Massman Construction Co., of Kansas City, Missouri; Traylor Bros., Inc., of Evansville, Indiana; and Alberici Constructors, Inc., of St. Louis.
After considering several areas for ATCs, MTA zeroed in on the drilled-shaft foundations for the river piers. The contractor owned large-capacity floating equipment and drill tools, including an 11 ft diameter core barrel with carbide-tooth roller bits. By using this specialized equipment, MTA could reduce the total number of drilled shafts in the river piers by increasing the diameter of the shafts from 10 ft to 11.5 ft. The rock sockets would be 11 ft in diameter and just 16.5 to 23 ft in length. This plan relied on the combined resistance of end bearing and side friction in the rock sockets. To confirm the design of the larger, shorter rock sockets, the MoDOT required MTA to perform an Osterberg Cell load test, which at the time set a world record of a combined end-bearing and side-friction capacity of 36,000 tons. MTA’s ATC was considered a minor plan alteration, but it had a major effect on the schedule and the budget. Reducing the number of drilled shafts cut the project cost by millions of dollars and shortened the schedule by several months.
Prior to the start of the project, HNTB and the MoDOT conducted a thorough risk analysis, determining and ranking every risk factor and developing game plans for each high-probability event. For example, the potential risk that scour might undermine the Illinois levee led to the design and installation of a rock blanket there. The ramifications of water surface elevations were addressed by a Federal Emergency Management Agency map revision. Since the U.S. Army Corps of Engineers was in the process of altering its maps to reflect other possible consequences, the bridge’s minor effects were added to the others. Moreover, when the ever-changing levels of the Mississippi threatened to delay the project, the team was ready for that, too.
Despite the team’s efforts to anticipate and prepare for contingencies and the MoDOT’s willingness to grant schedule extensions, dangerous high water cost the contractor 127 workdays over the life of the project. Flooding in the spring and summer of 2010 delayed the start of the foundation work on the river piers. To make up for lost time and to keep pace with the aggressive schedule in general, crews worked double shifts six days a week from the summer of 2010 to late November 2011, and for the remainder of the project they worked a 10-hour single shift Monday through Friday and an 8-hour shift on Saturday.
High water during the summer of 2013 threatened to delay work that was being performed to complete the deck of the main span. Crews therefore assembled the main structural steel piece by piece using crane barges fitted with ships’ anchors. (The steel pipes, called spuds, that are typically used for this purpose weren’t long enough to reach the river bottom and sink into it.) During those long days MTA relied on practices it has developed through its experience on other bridge projects to carry it to the finish line.
For example, the team employed balanced-cantilever construction for the superstructure. One cycle encompassed approximately 10 steps, including erecting steel sections, setting concrete panels, erecting stay cables, making adjustments, conducting extensive surveys of the deck’s geometry, and placing and curing concrete. A concrete mix with 70 percent blast furnace slag and a low heat of hydration was used in the mass concrete elements of the piers, namely, the footings, the subshafts, and the first two solid lifts of the tower. Here MTA implemented a thermal control plan that used embedded cooling pipes and insulation to control the temperature differential in the placements, and this plan shortened the thermal control period from months to a matter of days. Maturity meters embedded in the concrete were linked to a wireless system that enabled crews to remotely monitor the concrete’s temperature and strength during the curing process. Furthermore, self-climbing forms were used on the tower legs to improve the cycle time and reduce the need for cranes. A 350-ton capacity floating crane with 400 ft of boom and a 60 ft jib enabled workers to set rebar and anchor box gangs and to reach the tops of the towers.
Because of the design’s constructability, the high degree of teamwork on the part of the owner, the designer, and the contractor, and the universal commitment to meet the schedule, the contract had less than 0.2 percent overrun throughout the life of the project. After six years of cooperation and compromise by the two states’ departments of transportation, the state legislatures could not, at first, agree on what to name the bridge. Missouri wanted to name it after St. Louis Cardinal great Stan Musial, who had died just a year earlier, in January 2013. Illinois wanted to pay tribute to its armed forces veterans. In the end, the states found common ground in Musial’s background, which included service in the U.S. Navy during World War II.
On February 9, 2014, the Stan Musial Veterans Memorial Bridge opened on time, on budget, and with minimal change orders. Construction was fully completed the following month. Today, the bridge that almost wasn’t carries 40,000 vehicles per day, and volume is increasing steadily. It has reduced traffic on the Poplar Street Bridge by 20 percent and on the nearby Martin Luther King Bridge by half. Since the beginning of the year the MoDOT has seen an increase of 12,150 vehicle crossings (about 7.4 percent) on downtown St. Louis’s Mississippi River bridges. Had the new bridge not yet opened, most of that increase would have burdened the Poplar Street Bridge. The new bridge is siphoning truck traffic from other, older routes and has significantly reduced commute times for many. It has also reduced overall fuel consumption and accidents on both sides of the state line. What is more, the new bridge is a welcome alternative to long-distance drivers who want to avoid downtown St. Louis during rush hour.
Over the next 45 years, the bridge is expected to generate $25 billion in increased economic activity in the region. In keeping with this projected growth, the structure has been designed to accommodate an adjacent four-lane bridge in the future if demand warrants. The Stan Musial Veterans Memorial Bridge represents one of the greatest comebacks in recent transportation history, proving what can be accomplished when everyone focuses on a common goal and does the “impossible.” In addition to placing fourth on Roads & Bridges’ 2012 list of top bridges, the crossing was honored with the George S. Richardson Medal at the 2014 International Bridge Conference and with the 2014 Project of the Year Award from ASCE’s St. Louis Section. By Hans Hutton, S.E., P.E., Randy Hitt, P.E., and Thomas Tavernaro, P.E.