pattern and was only bounded by two piles. Some reduction in tip resistance was observed in the soft layers located above Layer 2, which can most likely be attributed to disturbance effects caused by the use of a vibratory hammer to create the pilot holes during driving. The post-densification analysis indicated that the sandy soil in Layer 3 was densified such that they would not likely liquefy during the SSE event. Axial Load Test Results The testing program consisted of dynamic compression testing using the pile driving analyzer (PDA), 1,000 kip (4.45 MN) Statnamic testing, static compression and reciprocal lateral load testing. The tests were performed at different times to evaluate the potential set up effects after installation. The axial compression capacity from the static load testing was determined using the Davisson method (Davisson 1972). Test pile TP-4, 18 in (457 mm) square PSC, was not tested. The final design of the production piles was based on the unit side friction f and end bearing q derived from the strain gage data s t obtained during the Statnamic and static load tests. In the preliminary design, the required pile tip elevation was about El. - 55.0 ft (El. -16.8 m) MLW. The downdrag load on the piles increased when incorporating the 4 ft (1.2 m) thickness of the structural fill for the working pad; therefore, the pile embedment was increased by an additional 2 ft (0.6 m). The pile tip elevation was confirmed by evaluating the design using the parameters derived from the pile load tests, which indicated that the compression capacity of the pile was similar between preliminary design and final design although the distribution of side resistance and end bearing were different. For the final design, the lower portion of Layer 2 (Layer 2b) was Layer 2a ( = 2 to 4 tsf [0.2 to 0.4 MPa]), and (2) the observed considerable increase of CPT tip resistance in Layer 2b (by about 22%) but no change in Layer 2a due to the pile driving. The pile length was computed using a design compressive load Static axial compression load test setup of 100 tons (890 kN) and the revised design parameters. A pile group efficiency of 1.0 was conservatively chosen due to the (negative) downdrag effect and to the (positive) densification effect in the sand layer. The center-to-center spacing of the piles was four times the width of the pile. The pile settlement was estimated to be insignificant when the allowable compression capacity was satisfied, and the piles had been embedded a minimum of 5 ft (1.5 m) into the very stiff-to-hard silt and/or clay (Marl Formation). The required quantity of piles was based on the imposed lateral loading and the SSE event. Using the results of the pile load tests, the final length of the 1,600 piles was shortened by about 7 ft (2.1 m), which resulted in a substantial savings in material and construction costs. Pile No. Pile Type End of Driving Results from Axial Load Tests kips (kN) 4 day 6 day Restrike Statnamic 12 day static load test TP 2 18 in (457 mm) sq PSC 376 (1,670) 629 (2,800) 743 (3,305) 847 (3,770) TP 1 18 in (457 mm) sq PSC 370 (1,645) 640 (2,875) 890 (3,960) --- TP 3 17.7 in (450 mm) diam IPC 246 (1,095) 642 (2,855) 857 (3,810) TP 5 18 in (457 mm) sq PSC 398 (1,770) 458 (2,040) --- Axial compressive capacities determined at different testing times The load-settlement curve for test pile TP2 indicated a plugging type failure with an increase in capacity over time (i.e., setup effect) of about 225% when comparing the results from the initial driving to the 12-day static load test. Examination of the test results indicated that the test piles did not fail during the 4-day restrike and Statnamic testing on day 6, as the pile displacements were relatively small. Therefore, the 4-day restrike and Statnamic tests served to verify the static load test results; however, the ultimate pile capacities were not determined by these two methods as the pile capacities were not fully mobilized. --- --- Pile Foundation Construction The sequencing of installation was to commence pile driving around the perimeter of the tank. The confinement by the perimeter piles resulted in additional densification of the sand layer. However, as the pile driving progressed from the edge towards the center of the tank, the pile driving became increasingly more difficult because the sands became too dense. About half way through installation, the piles could not be driven even with the use of a larger hammer. In addition, the casting of tension bar sleeves in the pile top would not allow the piles to be cut off at elevations higher than designed. Finally, predrilling had to be performed on every third or fourth pile to remove the column of soil to the sand layer to facilitate the pile driving. In hindsight, the problem should have been avoided by starting the pile driving from the center of the tank. considered a positive resistance; therefore, the negative skin friction layers were limited to the upper portion of Layer 2 (Layer 2a) and above. This delineation was based on (1) a relatively high CPT tip resistance in Layer 2b ( qc qc > 30 tsf [2.9 MPa]) compared with DEEP FOUNDATIONS • JAN/FEB 2019 • 17
