The impact of groundwater drawdown and vacuum pressure on sinkhole development. Physical laboratory models

A considerable proportion of the damaging sinkholes worldwide correspond to human-induced subsidence events related to groundwater withdrawal and the associated water-table decline (e.g. aquifer overexploitation, dewatering for mining). Buoyancy loss in pre-existing cavity roofs is generally claimed to be the main underlying physical mechanism. It has been also postulated that rapid water-table drawdowns may create a vacuum effect in the subsurface and contribute to enhance sinkhole activity in karstic terrains with a low effective porosity cover. Our laboratory physical model explores the role played by vacuum pressure induced water-table drops with different magnitudes and rates on sinkhole development, simulating an invariable mantled karst comprising cavernous bedrock and a low-permeability cover. The multiple tests performed include real-time monitoring of the water level drawdown (magnitude, duration, rate), the negative air pressures in the bedrock cavity and the cover, and several features of the subsidence phenomena (deformation style, size, magnitude, rate). The main findings derived from the test results include: (1) Vacuum pressure may trigger the development of cover collapse sinkholes in areas with low-permeability covers. (2) Different water-table decline patterns (magnitude, duration, rate) may result in different subsidence styles or rheological behaviours: sagging versus collapse. (3) Ground fissuring, frequently related to extension at the margin of sagging depressions, may cancel or significantly diminish the vacuum effect. (4) An overall direct relationship between the water-table decline rate and the subsidence rate. Some possible strategies are proposed to ameliorate the adverse effect of the negative air pressure on sinkhole hazard, which most probably has a local impact restricted by the concurrence of rapid water drawdowns and low-permeability covers.