Evaluation of Compaction Methods for Pipe Trench Backfill in Areas of High Water Table

Abstract

Determining the density of backfill material and attaining the desired degree of compaction in the backfill under a high water table, typical operations in South Florida, present a difficult problem to the Florida Department of Transportation (FDOT) engineers and contractors. As a first step in seeking the solution to this problem, variations in soil conditions, determined by standard penetration test (SPT) N-values, that corresponded with varying methods of compaction were evaluated. In the experimental phase of the project, five field conditions were tested: Condition A, control dry, backfilling in the dry condition above water; Condition B, dump soil, no compaction, backfilling under the water table, no tamping; Condition C, dump soil, compact per FDOT specifications, no dewatering, backfilling under the water table, tamping with a 2 by 4 as specified; Condition D, control wet, compact per FDOT specifications, dewatering during pipe installation and backfilling with periodic compaction according to FDOT specifications; and Condition E, dump soil, no compaction, dewatering during pipe installation and backfilling by dumping. For each of the five field conditions, SPT N-values were obtained at seven locations. As expected, the highest N-values were obtained for the control dry condition (Field Condition A). The next best set of N-values was obtained for the control wet condition (Field Condition D). This result was also expected. In Field Condition B, backfill was dumped without tamping. In Field Condition C, backfill was compacted by hand tamping in accordance with FDOT specifications (Section 125-8.3.3, 1996). It was evident from the SPT values that the result did not noticeably improve from Condition B to Condition C. In Condition E, dewatering was continued after the pipe segments had been placed and bedded. Lifts of backfill material were dumped without compacting or tamping. The N-values obtained from Field Condition E were not significantly different from the values obtained for either Condition B or C. The provision of hand tamping in the FDOT specifications is questioned as a result of these findings, and the benefit of dewatering, although obvious, is substantiated.