Giuseppe Gambolati

Land subsidence and uplift, ground ruptures, and induced seismicity are the principal geomechanic effects of groundwater withdrawal and injection. The major environmental consequence of groundwater pumping is anthropogenic land subsidence. The first observation concerning land settlement linked to subsurface processes was made in 1926 by the American geologists Pratt and Johnson, who wrote that “the cause of subsidence is to be found in the extensive extraction of fluid from beneath the affected area.” Since then, impressive progress has been made in terms of: (a) recognizing the basic hydrologic and geomechanic principles underlying the occurrence; (b) measuring aquifer compaction and ground displacements, both vertical and horizontal; (c) modeling and predicting the past and future event; and (d) mitigating environmental impact through aquifer recharge and/or surface water injection. The first milestone in the theory of pumped aquifer consolidation was reached in 1923 by Terzaghi, who introduced the principle of “effective intergranular stress.” In the early 1970s, the emerging computer technology facilitated development of the first mathematical model of the subsidence of Venice, made by Gambolati and Freeze. Since then, the comprehension, measuring, and simulation of the occurrence have improved dramatically. More challenging today are the issues of ground ruptures and induced/triggered seismicity, which call for a shift from the classical continuum approach to discontinuous mechanics. Although well known for decades, anthropogenic land subsidence is still threatening large urban centers and deltaic areas worldwide, such as Bangkok, Jakarta, and Mexico City, at rates in the order of 10 cm/yr.

Recent numerical studies based on a simplified lithostratigraphy of the Venice subsurface suggest that the city may be raised by pumping seawater into deep aquifers through 12 wells located on a 10 km diameter circle. Using an updated 3‐D reconstruction of the Quaternary deposits, developed very recently from about 1050 km of multichannel seismic profiles and eight exploration wells, along with a more accurate representation of the injection boreholes, novel finite‐element predictions are performed. The new model simulates the lithostratigraphy of the lagoon subsurface and allows for a reliable assessment of the water volumes injected into the geologic formations based on the actual bottom hole overpressure that can vary both in space and time. Pumping occurs into two Pleistocene sequences that are originated from the Alps and Apennine sedimentation and terminate just south and north of Venice, respectively, and the shelf portion of a Pliocene sequence that is rather continuous below the central lagoon with arenite layers to depths as much as 1000 m below mean sea level. With a proper tuning of the injection pressure the new hydrogeologic model allows for a prediction of a quite uniform 25–30 cm uplift over 10 years after the inception of injection. The gradient of the vertical displacement ξ_z does not exceed 5 × 10⁻⁵ and 1 × 10⁻⁵ in the whole lagoon and Venice, respectively, i.e., well below the most conservative bound recommended for the safety of the structures. If ad hoc calibrated injection overpressures are implemented in each single well, ξ_z may be reduced to as much as 0.1 × 10⁻⁵ throughout the city.

To prevent erosion of polluted land and discharge of contaminated groundwater and surface water into the Venice Lagoon, a 57 km long cutoff wall has been designed by the Venice Water Authority and is currently under construction along the canal banks of the Venice, Italy industrial harbor. To predict the impact on the hydrologic regime and mitigate the related inundation hazard in the nearby urban areas, a complex three-dimensional finite-element model has been developed and implemented over the multiaquifer system down to a depth significantly larger than the wall bottom. The model is initially calibrated against the regional piezometry and then refined locally to reproduce the groundwater volume drained along a 5-km long bank of a harbor canal already bounded. Major results from the simulations show that after the wall completion the subsurface discharge into the lagoon is successfully abated by as much as 85% relative to the preexisting rate with, however, an expected 1-m increase of the water table in...

A programme of Venice uplift because of seawater injection into a 600–800 m deep brackish aquifer underlying the lagoon has recently been advanced. As the actual spatial variability of the hydro‐geological parameters is to a large extent unknown, a controversial issue concerns the 25–30 cm heave prediction with the possible generation of differential displacements that might jeopardize the integrity of the architectural patrimony. It is shown that significant differential displacements have occurred in Venice in the past and are still presently occurring as evidenced by SAR interferometry. The results of a stochastic analysis addressing the variability of the hydraulic conductivity of the injected formation indicate that even a highly uneven aquifer expansion does not migrate to ground surface because of the smoothing effect exerted by the overburden. The predicted differential displacements are well below the values Venice is experiencing. Any a priori alarmism appears to be unfounded and unjustified.

Recent field evidence suggests that injecting fluids below the ground surface can induce an anthropogenic land uplift of a few tens of centimeters over a time interval that may range from a few months to a few years. At the same time, new modeling studies using a lot of realistic hydrogeological and geomechanical information from the northern Adriatic basin indicate that pumping seawater into a 600–800 m deep brackish aquifer below the Venice Lagoon might help raise the city uniformly by 25–30 cm over 10 years (a). This could provide Venice with an important innovative defence from and a substantial mitigation to the so‐called “acqua alta,” i.e., the increasingly frequent floods that plague the city. To test the feasibility of an actual program of anthropogenic Venice uplift, a pilot project is designed with the aim of investigating the occurrence over a limited area selected on purpose within or in the margin of the lagoon where three boreholes down to 800 m are drilled and seawater properly treated for geochemical compatibility is pumped into the selected aquifer during 3 a. Using an improved reconstruction of the geology and lithostratigraphy from a new seismic survey to be carried out in the lagoon subsurface, the pilot project plans the instrumentation of the injection wells and other boreholes for the continuous monitoring and accurate measurement of (1) pore water overpressure; (2) expansion of the injected unit by the radioactive marker technique; (3) compaction, if any, of the upper fresh water aquifer system with the aid of an extensometer; and (4) vertical and horizontal motions of land surface via spirit leveling, GPS and interferometric synthetic aperture radar. Preliminary numerical simulations show that a constant saltwater injection rate of 12 × 10³ m³ s⁻¹ from each well might provide a maximum 7 cm uplift at the center of the selected site over a 3‐a time, namely, a limited amount that is nevertheless accurately measurable and should not raise concerns for the stability of the buildings and the preservation of the infrastructures in the area. A continuous control of the experiment is envisaged based on much refined hydrologic and geomechanical models properly updated and calibrated to the detailed lithostratigraphy resulting from the new seismic campaign, the ad hoc field analyses, and the current field observations of the event. The completion of the pilot project is expected to require 4 a including an initial year needed for the necessary authorizations and the operative implementation of the injection program. The planned cost is in the range of 5 MEuro/a. The present paper addresses the major issues concerned with the design of the pilot project and discusses the results from the experiment simulations with a glance at their prospective application to an actual project of anthropogenic uplift of Venice.

Subsurface hydrocarbon removal induces a reservoir compaction that migrates to the ground surface and causes the land to subside. While compaction depends primarily on the fluid pore pressure drawdown, the thickness and the mechanical compressibility of the depleted formation, the amount of compaction reaching the surface depends on the depth and size of the producing field and the stiffness of the overburden. This paper presents a list of the most famed sites worldwide that have experienced anthropogenic land subsidence because of underground fluid withdrawal and reviews the basic mechanisms controlling the reservoir compaction and the related ground settlement. Finally, some mathematical models, that can be effectively used to simulate and predict the occurrence, and the most recent techniques for accurately and timely measuring and monitoring the deep compaction and the land surface displacements are described.