Factors such as darkness, driving winds, changing tides, mechanical problems, difficult site conditions, exacting tolerances and ambitious project schedules all had an enormous effect on this seismic retrofit.

Even though bridges are common in our environment, they still inspire a certain awe in all of us. Whether beautiful or plain, the simplicity of their appearance often hides the complexity of the work involved with them. Mountain Pacific Surveys (MPS), Fairfield, Calif., had the opportunity to experience firsthand just how interesting and challenging bridgework can be.

Over the course of two years, MPS was on-site of the San Mateo-Hayward Bridge Seismic Retrofit project virtually every day. The project is part of the California Department of Transportation's (CALTRANS) aggressive program of repairing those structures most in need of seismic retrofit. The upgrade project on the San Mateo-Hayward Bridge, which spans the San Francisco Bay, carried a price tag of $125 million.

In February 1998, MPS was contracted by the joint venture partnership of Morrison-Knudsen/Traylor Brothers/Weeks Marine (MKTW) of San Francisco, Calif., to perform surveying services on the San Mateo-Hayward Bridge Seismic Retrofit project. Our work on the project was wide-ranging, from GPS control establishment to structural monitoring and from construction layout to as-built surveys. Factors such as darkness, driving winds, changing tides, mechanical problems, difficult site conditions, exacting tolerances and ambitious project schedules all had an enormous effect on how we carried out our work each day. Due to these varying situations, the same survey task from day to day often required different solutions. MPS party chief Mike Wilson and chainman Scott DeGracia teamed with MKTW project engineer Glen Smith and pile superintendent Norm Mitchell to solve each of these demanding and unusual survey situations in stride.

Many elements of construction could only be performed during the night, therefore our crew schedules often spanned day and night, occasionally seven days a week. For instance, a significant portion of the work was very close to a Pacific Gas and Electric Company (PG&E) twin tower high voltage power line. Pile driving work in this area could only be performed in the middle of the night when diminished power requirements afforded PG&E the ability to turn off the power lines closest to the construction activities. As this was a major undertaking for PG&E, the contract allowed a limited number of days to complete work in this area. As such, whenever a delay in this area occurred for any reason, we could expect to be on-site Saturday and Sunday trying to make up for lost time.

In the Beginning

Our first order of work was to establish an on-site horizontal control network based upon existing NAD 83 California HPGN stations in the vicinity. Four primary control points were established using Topcon dual-frequency GPS receivers (Topcon America Corp., Paramus, N.J.). This control served as the basis for all of our as-built and layout surveys. Two one-hour sessions were observed on the primary control points, as well as the three California HPGN points to which our final network adjustment was constrained. From this network, we as-built eight of the existing 38 piers along the length of the high rise section. Using this data and record drawing information, we computed a record position for the center of each pier along the original construction centerline. From this information, MKTW calculated the approximate locations of all existing footing corners, which in turn were used to guide the RTK-aided excavation operation around each footing.

Next, vertical control was established along the upper spandrel of each pier. Originally we had planned to run reciprocal trig levels from one pier to the next. However, our Geodimeter 650S Bergstrand total station (Spectra Precision Survey, Itasca, Ill.) could not complete the automatic fine level/compensation process due to the slight vibration caused by vehicle traffic. Not inclined to perform the trig levels without the compensator, we next tried a Wild T1600 total station (Leica Geosystems, Norcross, Ga.). Although the compensator was working properly, the vibration was sufficient to prevent accurate sighting to a target. Finally, we opted to use a Wild N10 spirit level (Leica Geosystems) that we generally use on building surveys for similar reasons. Since leveling this instrument is accomplished via a coincidence bubble rather than a compensator, we were able to overcome the vibration in the structure, although leveling across the structure became a much more tedious task than originally envisioned. Once established, the vertical control was used for everything from setting up tide boards to as-built surveys to layout for the various elements of construction.

Interestingly, a 3,000' portion of the original 1929 structure had been left intact to provide public pedestrian access to the bay. Known as the "fish pier," this portion of the original structure parallels the present day bridge and provided an excellent vantage point for the westerly 20 percent of the project. In fact, some of the construction control points for the 1967 structure were found on the fish pier and used in our control network. MKTW also used the fish pier as a lay down and staging area, and as access for both workers and heavy equipment.

Because the fish pier is in a seemingly severe state of decay, we had to perform a series of load tests along the length of the structure to evaluate its worthiness. These tests were performed in three stages using concentrated loads of 20,000, 40,000 and 60,000 lbs. The procedure included setting four rows of nails at approximately 30' intervals for the length of the structure and measuring elevations at each location with no load at each of the test loads, and again at no load upon completion. The load was carried on a trailer and would allow the full test weight to be concentrated directly over the survey monitor points. The complete sequence took four days to complete and required in excess of 1,000 measurements to the monitor points. The fish pier held up well, although steel trench plates were required in a few areas for reinforcement.

Survey operations, as well as all construction activities, were hampered by limited access to work areas. Access to the upper spandrel of each pier and to structural elements along the underside of the bridge deck is provided via a 2'-wide catwalk that runs along the top of each pier for the length of the high-rise section. At a healthy pace, the crossing requires about a 30-minute walk to complete, and this crossing was made hundreds of times. Since many elements of construction were at or below the water level, access to these areas was accomplished only by crew boat or skiff. Often a crew member could be stranded an hour on a lower spandrel waiting for a boat to become available to provide passage back to land.


The driving operation for the 88 steel piles, which ranged from 5' to 12' in diameter and 160' in length, required extensive layout. These piles had a specific relationship to the existing pier foundations, and as such, the existing foundations had to be as-built surveyed prior to pile placement. The tops of the existing foundations were in the -10 to -22 elevation range. Therefore, the as-built surveys were completed in cofferdams using control brought down from the upper spandrels of adjacent piers. After locating the foundation, coordinates were calculated for each new pile, of which there were four per pier.

The actual layout of the piers was handled a few different ways, depending upon various combinations of clearance under the bridge, water depth and the ability to position survey instruments. MPS explored the possibility of RTK positioning; however, since the majority of our work was directly under the existing bridge, it was determined our window of opportunity would be too small. Pile driving operations occurred throughout the day and night; therefore, we needed to develop systems that could be employed anytime at a moment's notice.

Prior to pile driving, a template barge was spudded in and the piles were lifted vertically and placed in a jig on the template barge. We were able to adjust the template jig along both an x and y axis. Thus, piles were able to be fine-tuned into position by using two eccentric reference points on the template jig with a known relationship to the center of the pile. A preliminary measurement was taken to the reference points, an adjustment was calculated and implemented, and then a final check was made to the reference points.

After the correct horizontal position was achieved, the piles were monitored for plumb during the driving operation with two instruments at positions as near to 90 degrees to one another as possible. Alternately, some of the piles were located using the same principles as outlined above; however, the total stations were replaced by vertical plumb lasers placed in the box girder of the superstructure. In this situation, the pre-determined position of the targets was transferred by conventional survey methods up to the inside of the superstructure box girder where the lasers were set up. The template jig below would be positioned such that when the visible laser hit the target locations on the jig, the correct layout was achieved. This method proved beneficial in that the position of the template jig could be monitored throughout the driving process by simply checking the position of the target relative to the laser.

In a few situations there was not enough vertical clearance to position the template barge under the existing bridge. In these instances it was not possible to use the template jig or attach a prism to the pile to obtain distance measurements. Therefore, these piles were positioned by calculating the azimuth to the outside diameter of the pile from two existing control points and guiding the pile into position by providing corrections to the crane operator.

Driving an actual pile was a rather disjointed process. During the first step, a pile would be driven to what was termed "excavation depth," which was approximately 15' above water level. At this point, the hammer would be removed and MPS would take three shots on the pile to determine a coordinate for the center of pile and a depth for the excavation. Next, MKTW would excavate the inside of the pile down to the depth of the slurry that would later be pumped into the pile. Upon completion of the excavation, the pile would be driven down to the water level, where one more set of three shots was recorded. Finally, the hammer would be replaced once again and the piles would be driven home to their design elevation. Since the piles were all driven well below the water surface, final tops of pile elevations were determined by the use of a level rod that was fabricated from a 35' tide gauge graduated to 0.01' attached to a 6" square steel tubing welded to the outside of the hammer. The rod was monitored during the final driving sequence, which would halt once we obtained the pre-determined rod reading, indicating the design elevation had been reached. Run out for the last portion of pile driving was essentially non-existent; so, the final horizontal coordinates for each pile were extrapolated from the two data sets, one obtained at water level and the other obtained 15' above water level. Final pile locations were routinely falling in the +/-0.1' range horizontally and within +/- 0.02' vertically of the design position. During the entire pile operation vehicular traffic was uninterrupted, probably providing an interesting show for many motorists.

'Dem Bones are Heavy!

After placing the piles and a 3' concrete cap over the existing foundation at each pier, a structure, known as the "dogbone" was set over the piles to structurally tie them together. The dogbones were fabricated off-site in two pieces and shipped by barge to the project site, then set in place and joined together by a closure pour. Each dogbone section weighed 550 tons; manipulating them was no easy task. The crane used to load them onto the dogbone barge was originally used to load military cargo at the Hunters Point Naval Shipyard. Seeing only limited use over the past years, there was a question as to whether the crane structure was safe to be used for the task at hand. A structural analysis of the crane was performed and a series of load tests was prescribed to evaluate the capabilities of the structure. MPS was called upon to devise a monitoring strategy in which incremental elevation changes of 0.003' could be detected during the execution of the various load tests. Twelve mini prisms were mounted to the structure at key locations as directed by the structural engineers, from which six each were visible from two control points on the ground away from the crane. With two Geodimeter 650S Bergstrand instruments, the 12 targets were monitored during a daylong sequence of increasing load tests, culminating in the crane structure performing as predicted by the engineer's model, all the way up to 125 percent of the anticipated load.

The primary horizontal control points were monitored with secondary control points outside of the test area. Vertical control was established and monitored throughout the test with the aid of a Zeiss Ni2 auto level (Carl Zeiss, Thornwood, N.Y.) mounted on an elevator tripod. The elevator head tripod allowed us to measure and monitor our instrument height very accurately, and by utilizing frequent checks to secondary control points outside the work area, we were able to accurately monitor each instrument's horizontal position throughout the test sequence. This effort assured us that positional displacements observed were indeed in the structure, and not introduced as a result of the failure of our monitoring methodology to detect systematic or procedural errors.

The San Mateo-Hayward Bridge Retrofit project presented many challenges not encountered in our everyday work. Difficult site conditions and constantly changing variables kept us hopping. With a little ingenuity, however, we successfully overcame each challenge. Our company made a commitment to be the right hand of MKTW from a surveying perspective. As such, we worked all hours of the day and night as the project schedule dictated and were available at a moment's notice--all in an effort to provide an unparalleled level of service to the project and our client.

We came away from this project with a greater appreciation for the complexities of a project of this magnitude, coupled with the affirmation that the application of modern day technologies and time-tested survey techniques can work together to solve even the most unusual and difficult survey situations.

Bridge History

Constructed in 1929, the San Mateo-Hayward Bridge was the longest bridge in the world at that time. Consisting entirely of a low-level structure at elevation 35, a lift section was soon added to accommodate ship traffic through the bay. By the late 1950s, vehicle traffic would be brought to a standstill a half-dozen times a day due to ship traffic. This functional inadequacy, coupled with the extensive maintenance required as a result of the corroding superstructure, prompted its replacement. The structure consists of two traffic lanes in each direction for the trestle portion and three traffic lanes in each direction for the high-rise section. California recently received bids to construct a parallel trestle that will increase the capacity to three lanes in each direction for the entire length of the span in an effort to relieve congestion of more than 77,000 vehicles per day.

The work performed under the MKTW contract consisted primarily of retrofitting the superstructure, substructure, piers and pier foundations.