Process Simulation for Temporary Cofferdams for Construction
LT Edward S. McGinley, III & LT Stanley W. Wiles

A cofferdam is a temporary construction method used in order to do construction in wet excavations. The cofferdam is constructed in order to keep water out of an excavation or work area with the aid of pumps. The cofferdam is a box like structure fabricated through the use of a steel bracing system and steel sheet piles.

Heavy equipment is used in the construction of a cofferdam. Pile drivers, cranes with clamshell buckets and concrete pumps trucks as well as pumps for dewatering are used in the construction process. The effective management of equipment on site as well as personnel is an important step in cost control and maintaining efficient productivity.

Project Description

The project used for this study was the U.S. 231 concrete bridge over the Wabash River on the southwest side of West Lafayette. The owner of the bridge is the Indiana Department of Transportation. The contractor for the project was Rieth Riley Construction Company. Construction began September 1993 and finished in August 1995. The final cost was $20.5 million. The bridge consisted of a 12-span , 2100 foot, twin-roadway bridge from abutment to abutment, with approaches to the abutments being part of a future contract. The bridge is a reinforced concrete sub-structure and super-structure, with pipe-pile group foundations. The super-structures consists of twenty-two piers; four of which were built in the river via cofferdam construction.

The first phase in building the river piers is the construction of a temporary bridge. The temporary bridge is needed for material handling and access to river construction activities. The next procedure is to construct a template for the cofferdam. The template serves two important measures. First, it is a guide for installing the sheet piles for the cofferdam. The sheet piles must be driven perfectly plumb to ensure the installation of the last pile. Second, the template provides all of the lateral support needed during the dewatering process. The template is installed on falsework to approximately the location of the desired cofferdam. The setting of the template takes about a half a day. The falsework is usually installed in one day. The spuds (H-piles) are driven through the cofferdam frame to provide an exact location and plumbness to the frame. The frame is then slid up the spuds to the required elevation of the frame and the falsework is then removed.

The next phase is to install sheet piles. The piles were started on the up-river side of the cofferdam. This eliminated the problem of working against the current during installation of the other three sides. The sheet piles should be driven enough (not to grade) to keep the water current and the wind from affecting them. To ensure a plumb cofferdam, every other sheet in the cofferdam is driven clockwise around the cofferdam and then the remainder should be driven counterclockwise. This process helps keep the cofferdam plumb during driving because the cofferdam slowly leans in the direction of the driving.

After the sheet piles are in placed, excavation was begun inside the cofferdam with a 1-1/2 cubic yard clam bucket. This process simply achieves the desired elevation of the pier being constructed in the cofferdam. The excavation process takes about a day and a half and another half day to clean it out with an "air-lift". The air-lift, designed by Reith Riley Construction Company, works as a vacuum and extracts the soil. It consist of installing an air hose from a compressor inside a small diameter pipe to create this vacuum.

Pipe piles were used for the foundation. The pipe piles were driven with a hydro-hammer S70 (70 Kip-feet), which is submersible if necessary. The crane then positioned the pipe pile close to its pre-determined destination. The impact hammer clamp on to it and drove it to its required depth. The piles are filled with concrete prior to the placing of the tremie seal.

When placing the tremie seal, the first couple of yards (7 sack, Class C mix, 2500 psi) were used to establish a cone on the bottom of the river. The pump truck’s discharge hose was submerged as far as possible without coming in contact with the river bottom. The tremie seal is usually poured slightly low to insure that the foundation may be formed correctly. Once the tremie seal has been placed and cured, the contractor dewatered the cofferdam and cut the pipe piles to grade. The tremie seal was cured for at least 3 days before dewatering.

The foundation for the pier was formed up. During this forming, the cofferdam was continued to be dewatered. Once forming was complete, the foundation was poured. The pier was then placed, the cofferdam was backfilled and filled back with water, and then the sheet piles were pulled.

Description of Simulation Models

Several versions of the sheet pile, box cofferdam were developed for simulation on Microcyclone. All the versions were done with construction of 100 cofferdams. No more than two cofferdams were simulated in construction at one time. The following resources were considered in the models:

1. Cranes: assisted in the falsework and template setting, setting and driving sheet piles, excavation (w/bucket), and driving bearing piles (w/pile driving rig).
2. Crews: three main crews; falsework/template, piles, and concrete crews.
3. Concrete Pump: used for tremie seal and concrete for footer and piles.

Duration Input

Durations were determined from video. Both measurements from the construction process and from an interview with the project manager were used in determining approximate duration for the activities. Two activities were simulated with a small variance using a normal distribution: setting and driving sheet piles, and driving the bearing piles. These activities contained the most unknowns due to soil conditions and work process and had some variation in their durations due to these facts.

The duration of the dewatering activity was set at zero. The cofferdams were assumed to be dewatered overnight, so that work could begin in the morning on cutting the piles. It was also assumed that the work would be done in an eight hour workday for ease of calculation.

Resource Costs

The costs of labor and equipment were considered for the simulation of the sheet pile, box cofferdams. Indirect costs were not used in the simulation process. For a complete project analysis, indirect costs should be portioned to the respective activities of work and used in computation of savings for early completion. Because indirect costs vary from company to company, they were not used. Material costs were also not considered. The material costs do not change from simulation to simulation, so they would not factor into cost decisions as to the method based on a simulation.

In Microcyclone, costs can either be fixed or variable costs. Variable costs are only incurred when a resource is operating (such as, oil, fuel, etc.). Fixed costs occur whether or not a resource is operating (such as, rental costs). The equipment costs were split into variable and fixed costs. The fixed costs were based on approximate rental costs for comparison study. The variable costs were assumed costs for operation of the equipment. The crew costs were assumed to be variable costs. It was assumed that when the crews were not working on the cofferdam, that they would be reassigned to other parts of the project. The costs of labor were also assumed to be $80/day/man; while this is low, it is easy to convert to local area wages to apply the model. It should be noted that the costs were assumed on a daily basis (again, an eight hour working day was assumed for analysis purposes).

Simulation Versions

For our simulation, six different versions were run on the MicroCyclone program. Versions 1 through 3 used a crane to drive sheet piles, excavate and drive bearing piles. Versions 4 through 6 used one crane to drive sheet piles and excavate in addition to a separate crane used just for driving bearing piles (the activity for attaching the pile rig was removed because it is not necessary for these three versions). Due to site limitations, only one crane could excavate at a time (therefore only one excavation bucket was used to alternate between cofferdams), and no more than two cranes could be working at the same time. In all the versions only one concrete pump was used because it is not a constraining resource for this simulation (a simulation was done varying the concrete pump and productivity did not increase in any substantial amount).

Version 1

Version 1 used only one crane, one pile hammer, and one crew for the setting and driving of all piles and any excavation. Refer to the table below for a network summary for resources. Appendix A contains the network input, activity information and process report.

The total time to complete one cofferdam was 26.64 days at a cost of $26,387 per cofferdam. The crane was idle 9.4% of the time. The next restrictive resource was Crew B (25.6% idle).

Version 2

The total time to complete one cofferdam was 16.92 days at a cost of $32,133 per cofferdam. The crane was idle 36.6% of the time. The next restrictive resource was Crew B, again (57.9% idle).

Version 3

The total time to complete one cofferdam was 16.91 days at a cost of $32,109 per cofferdam. The crane was idle 36.4% and Crew B was idle 57.8% of the time.

Version 4

Total time to complete one cofferdam was 23.43 days at a cost of $21,835. Crew B was the most restrictive resource with an idle time of only 15.4%. The crane was the next most critical resource with an idle time of 39.3%.

Version 5

Total time to complete one cofferdam was 19.30 days at a cost of $33,964. The pile crane was the most restrictive resource with an idle time of 49.9%. Crew B was the next most critical resource with an idle time of 69.4%.

Version 6

The following table shows a time-cost tradeoff between the different versions.

To determine which version was a better alternative, we used the following rules. If both ‘CC’ and ‘CD’ were positive, then the new alternative is rejected. If both ‘CC’ and ‘CD’ are negative, then the new alternative is accepted. If ‘CC’ is positive and ‘CD’ is negative, accept the new alternative if the absolute value of ‘CC/CD’ is less than $1,000. If ‘CC’ is negative and ‘CD’ is positive, accept the new alternative if the absolute value of ‘CC/CD’ is greater than $1,000. The $1,000 is an assumed liquidated damage assessment on the project.

From the table above, version 6 is the best alternative for construction. Version 6 consists of two cranes used for setting falsework, driving sheet piles, and excavating; two crews (B); and two bearing pile driving cranes. While the crew idle percentages are high for the cofferdam simulation, our assumption from the beginning was that during idle periods the crews could be employed elsewhere on the project. For example, the concrete crew (Crew C) could work in the prefabrication yard during idle periods.

References