2006 Philadelphia Annual Meeting (22–25 October 2006)

Paper No. 3
Presentation Time: 8:00 AM-12:00 PM

EFFECTS OF GROUNDWATER FLOW PATTERNS ON SLOPE STABILITY


KAUNDA, R.B., Geosciences, Western Michigan University, Kalamazoo, MI 49008, CHASE, Ronald B., Department of Geosciences, Western Michigan University, Kalamazoo, MI 49008 and KEHEW, Alan, Dept. of Geoscience, Western Michigan University, 1187 Rood Hall, Kalamazoo, MI 49001, ronald.chase@wmich.edu

Studies conducted by Chase and others (2001) during the last 10 years along the southeastern shoreline bluffs of Lake Michigan show that bluff recession continues during times when the lake is well below its mean level. Extensive stratigraphic, structural, and temporal studies of bluff failures along a 16 km coastal reach have shown that most bluff recession is likely to occur at locations underlain by interlayered sand and clay and during the winter-spring cycle. The clay layers act as aquitards and create perched water tables and bluff-face seeps in overlying sands at two or more levels above the main water table. Perched water levels are monitored hourly at three test sites by a series of vibrating wire piezometers. During winter months, surface seeps freeze resulting in a rise of the perched water tables that increases pore pressures and lowers effective stresses resulting in slope movement.

Occasional catastrophic failures and many slow slumps have been identified during the winter-spring cycle of 2004-05. For example, a significant March 7 event occurred that was identified in the majority of the 44 borehole inclinometers placed in 12 nearby monitoring wells. Flow net analyses were conducted before, during, and after the movement using hydraulic head data recorded hourly from 14 nearby piezometers. The results show that the flow patterns before the slope failure were different from those after the failure. It is feasible that ongoing minor catastrophic failures alter the local and regional ground water flow patterns, hence affecting the slope evolution and causing future failure events to be less predictable.