Modelling of subsidence due to water or hydrocarbon withdrawal from the subsoil

Abstract

Subsidence induced by human activity is experienced in many parts of the world. It is due either to withdrawal of underground fluids such as water, oil or gas, infiltration of water or due to mining activity. Its effects are sometimes dramatic both in the extension of the affected area and in the amount of surface lowering. But also relatively small amounts of surface lowering may have serious consequences if the affected area is located in coastal regions. Surface subsidence due to water withdrawal, for industrial or agricultural purposes has received most attention. A case book by Poland (1984) contains a collection of many documented cases and the most spectacular ones are listed here. For each case the amount of measured subsidence is given, the depth of the exploited layers, the areal extent of the involved region and the period when the phenomenon was observed; if the symbol + follows a date this means that the phenomenon was still active at the indicated date. The chosen cases are that of the Cheshire district in England (15m, 100–300m, 1500km², 1533–1977), Haranomachi City in Japan (2m, 100–200m, 25km², 1965–1978+), Tokyo (4.59m, 0400m and 800–2000m, 3240km², 1918–1975+), Wairakei, New Zealand (6–7m, 250–800m, 30km², 1952–1975+), Santa Clara valley in California (4.1m, 50–330m, 650km², 1918–1970), San Joaquin valley, also in California (9m, 60–900m, 6200km², 1930–1975), Tulare-Wasco, USA (4.3m, 60–700m, 3680km², 1930–1970), Huston-Galveston, USA (2.75m, 60–900m, 12000km², 1943–1973+) Mexico City (9m, 0–50m, about 225km², 1891–1978), Po — delta (3.2m, 100–600m, 2600km², 1951–1966), Ravenna (1.2m, 80–500m, about 600km², 1955–1977+) and Venice (0.15m, 70–350m, about 400km², 1952–1970) in Italy. Wairakei is a geothermal area; the subsidence of Ravenna is due to the combined effects of water withdrawal from the upper layers and the exploitation of deep gas reservoirs (at about 1800 m depth). In the Po delta water was pumped because it contained dissolved gas, which was separated at the surface due to the change of its solubility with pressure. In all these cases pumping of water disturbs the hydraulic equilibrium and acts as a forcing function on the coupled system soils plus water. Since the overburden remains constant, a decrease of water pressure due to pumping means an increase in effective stress, which produces deformation in the exploited layers. This deformation is transmitted to the surface and manifests itself as subsidence. The phenomenon is time dependent and continues until a new equilibrium state is reached. If water is pumped from superficial layers, desaturation and capillary effects may complicate the picture.

I wish to thank my co-workers G. Bolzon, R. Matteazzi, V. Salomoni, L. Simoni, H.W. Zhang, for their contributions to subsidence modelling over the last years and particularly D. Boso for her help in the preparation of the manuscript.