Figure 5. Excavator-on-a-barge system used the PDA combined with the CAPWAP method. The tests also indicated an increase in the capacity due to the setup effect, between end of drive and a restrike performed ten days later. The results from the experimental test ® site led to choosing the PDA test results as the pile acceptance criteria. The specifications called for testing at least one pile per tower, chosen mainly based on the pile position in relation to the barge, minimizing the time spent on the positioning of the barge. Some end-of-drive PDA test results in guides for navigating the barges through the less favorable locations. The buffalo trails. After the creation of the environmental reserve in 2003, the only major economic activity permitted in the region is raising buffalos. Those animals always follow the same path when returning to their stalls. Since the soil has very low resistance, the weight of the buffalos throughout the time deepened the terrain, allowing the use of the path for navigation. The buffalos also destroyed the native vegetation along that path, making it even easier to navigate, with no further environmental damage. The contractors could take advantage of those buffalo trails, through information given by locals, for navigating the barges through the Alagados. Driving Criteria and Dynamic vs. Static Load Tests The foundation designers sized the diameters and lengths of the piles utilizing semi-empirical methods widely used in Brazil. Those methods propose a correlation between the NSPT and the lateral adhesion factor and the end bearing of the piles. The designers established a minimum total number of hammer blows and a minimum penetration depth as field verification criteria. Early in the pile driving process, however, it became clear that the criteria defined in the project were not adequate. In several cases, the piles 56 • DEEP FOUNDATIONS • JAN/FEB 2014 reached the required depth, but not the total number of blows; in other cases, the driving reached the total number of blows, but the driven length was not sufficient according to the project. Even taking into account possible inac- curacies in the soil borings, the discrepancies between the predicted and actual behaviors were too great. This was attributed to the inadequacy of the semi-empirical methods used, which were developed based on correlations between NSPT and Static Load Tests (SLT) results carried out in the southern part of Brazil, where the geo- logical formations are much different. The contractor’s option was to perform load tests to verify the real capacity of the piles. The conditions in the Alagados region made executing SLT difficult and Pile Driving Analyzer (PDA ) are faster and less costly, allowing the testing of a larger number of piles. To validate the DLT, the parties to the project deemed it necessary to carry out comparisons of SLT and PDA results on test piles. They decided to create an experimental test site at the city of Porto de Moz, Pará, in the southern edge of the job site. They drove five test piles in the site and submitted them to SLT and PDA tests. The tests performed in this experi- expensive. Dynamic Load Tests using the ® mental test site indicated good agreement between the results of static load tests carried to failure with those obtained with the Alagados region indicated that the uplift capacities (calculated as the skin friction reduced by a factor of 0.8), and in some cases even the compressive capacities were not being reached, so the piles would have to be driven further. This would represent an increase in the cost of the job, in terms of labor and additional piles. Since the results obtained in the experimental test site showed that the soils in the region exhibited a substantial gain in skin friction with time, the parties involved determined that a new test would be carried out a few days after the first one, if the initial results were not satisfactory. Retesting the piles with the PDA at different times after initial driving showed that the setup effect increased the capacity of the piles to the levels required by the project, allowing great savings in time and cost. Conclusions Adequately sizing the equipment navigation in flooded regions similar to the Amazon is very important for economic and environmental reasons. The project’s large and heavy driving equipment was not adequate for navigating the floodplains of the rivers in the Amazon region, but light driving rigs on low draft barges seemed to be the ideal solution. This project shows that traditional foundation design methods may not be adequate for this region. New geotechnical studies are necessary, aimed at establishing more reliable design methods. Intensive use of Dynamic Load Tests, confirmed by Static Load Tests performed in the same region, should be mandatory. This is the only methodology with which one can currently rely.
