Sub-Slab Characterization Of Gavins Point Spillway With Ground Penetrating Radar Mapping

Historic Missouri River flooding in June 2011, prompted the USACE1 to conduct physical inspection of exposed spillway slabs, supplemented with concrete coring and Ground Penetrating Radar (GPR) testing, as part of an overall risk evaluation of the six Missouri River Main Stem Dams in Montana, North Dakota, South Dakota, and Nebraska (USACE, 1998). In addition, the USACE mimicked operational conditions during the flood event to evaluate effects on the vertical/lateral subdrain systems beneath the spillway slab at Gavins Point Dam, near Yankton, South Dakota. The spillway gates were operated at 12,000 cfs flow with various configurations of gates discharging. Anomalous conditions observed during the half-day test suggested pressurization beneath the spillway slab. Further evaluation was performed with GPR on normally inundated portions of the spillway slab to evaluate the effects of this condition. During an 8-hour period, the Missouri River was lowered, exposing the upper spillway slab. Three teams simultaneously operated three 250-MHz GPR systems to gather bidirectional data on 2-foot centers in an area of approximately 54,590 square feet. Concurrently, concrete coring was performed near damaged vertical drains found during the half-day test. The GPR data were subjected to thorough data processing including: 1) amplitude mapping; 2) digitization and mapping of the top of bedrock; and 3) line-by-line inspection for trains of multiple reflections, shadows, and pull-up/push-down in travel time to bedrock. Based on knowledge regarding construction details, a conceptual model was developed from which prediction of the behavior of the GPR wavelet was made under various scenarios where void thickness and fill (water or air) were considered. Based on this conceptual model, the aerial extents and thicknesses of voids beneath the spillway slab were interpreted. Post-interpretation sonic drilling and borehole geophysical logging were performed to help constrain the conceptual site model and GPR anomalies. Interpretation of the void extents and a hydraulic explanation of void formation by considering multiple lines of geophysical, hydraulic, and other observational evidence provided an enhanced understanding of subslab conditions. This improved understanding was used to inform decisions on post-flood repair strategies as well as the implementation of risk reduction measures to improve the function and reliability of the slab subgrade.