A novel approach to the study of saline upconing beneath a partially penetrating well in a model confined aquifer has been undertaken using a geotechnical centrifuge. The model aquifer contains freshwater underlain by denser saline water, which is common in most coastal aquifers and inland areas where aquifers contain remnants of connate brines. A fine ground silica soil was adopted for the aquifer solid phase. A gravitational acceleration of 50g was imposed on the aquifer. The aquifer was subjected to controlled intermittent abstraction through a partially penetrating well at time dependent decaying flowrates. The distribution of saline water throughout the model aquifer was monitored in real-time using a novel electrical resistance based saline detector system developed as part of this study. Both the saline water upconing and reinstatement behaviour beneath the abstraction well were analysed in order to derive a better understanding of the non-equilibrium response of economic aquifers subjected to attempted high flow rate abstraction. A total of 7 centrifuge tests were executed to simulate the distribution of saline upconing due to anthropogenic action, whilst varying the duration of abstraction from the well, the period of reinstatement of the aquifer and the initial position of the saltwater/freshwater interface relative to the base of the well. Mariotte bottles were used to provide a source of freshwater and saltwater to the aquifer model during the experimentation. The research revealed a relationship between abstraction rate and equilibrium condition of the interface during reinstatement. The nature of the rise of the upconing was a complex function of initial pumping rate, initial location of the freshwater/saline interface relative to the well, rate of reduction in abstraction and duration of abstraction. The ratio between the normalised height of the cone and the cross-sectional area of the cone at its midpoint remained constant during a reinstatement cycle. From observations, the cone did not appear to decline immediately on cessation of abstraction. The mechanism of this delay in response could not be related to any of the observational metrics available. However, the angle of inclination of the interface reduced to a critical angle where the apex began to decline. This angle was found to be between 25° and 55°. The rate of decline reduced when the angle was less than 25°.