Watertightness and leakage prevention is crucial in every tunnel joint design. Immersed tunnels have been utilizing the Gina-Omega gasket solution since the 1960s to prevent leakage from occurring. However, after almost 50 years of service life, leakages were detected in the immersion joints, which leads to an investigation into the source of the problem. It was uncovered that the bolts of the Gina gaskets had failed due to the immense load exerted by the sand above them. It was apparent that the load on the gaskets builds up due to an increase in soil stresses. A hypothesis was formed regarding this phenomenon; the sand inside the joint gap was exposed to almost 50 years of loading and unloading cycles from the expansion and contraction of the tunnel elements due to seasonal changes in temperature. This, in turn, densifies the sand inside the joint gap, which results in rising soil stresses. A 1:3 scaled physical model of the joint gap was designed and constructed to test the validity of this hypothesis. The model joint gap is equipped with a static lining on one end and an actuated lining on the other, hence the device is able to imitate the annual joint contraction and expansion cycle. A barrel containing sand is fixed onto the top of the model joint gap and acts as a reservoir of sand, allowing for more sand to enter the joint gap. The device is also equipped with 2 load cells and an LVDT, which allows for the measurement of horizontal soil stresses and vertical gasket displacement respectively. Two holes, with a flap covering each of them, are installed on the side of the model joint gap, which allows for a penetrometer test to be conducted on the joint gap sand. Multiple experiments with varying configurations and test conditions were performed. The results show that although the multiple loading and unloading cycles apply the same displacement for every cycle, the soil stresses increase with time. The Gina gasket show an apparent “walking effect,” where the gasket moves continually inwards. While a similar test conducted without the presence of sand fails to produce any “walking effect.” Penetrometer measurements show that the soil increases in density over time. The investigation is continued further with finite element analysis using the geotechnical modeling software PLAXIS. The joint gap part of the device is modeled in the program and is subjected to loading conditions and configuration similar to the physical model. Results of the simulation show indications of gasket “walking effect,” as well as stress-strain behavior similar to the experiment results. A comparison analysis between the results of the physical and finite element model is subsequently conducted. The finite element analysis allows for the calculation of the sand vertical stresses, which is previously unable to be measured during the experiment. It is observed that the vertical stresses rapidly escalates due to the high friction between the sand and the lining wall. A further analysis is conducted to estimate the force required to push back the displaced Gina gasket. Due to the good agreement between the results of the physical and finite element models, a modified version of the finite element model is used to predict the resulting pushing-back forces. Finally, the research is concluded with the validity of the previous-mentioned hypothesis. It was also confirmed that the soil undergoes densification, proven by the penetrometer readings as well as further validated by the stress-strain behavior of the sand. It is also proven that at higher horizontal strain, the increase in density, stresses, and the “waking effect” is more pronounced. The force analysis produces values of force needed to push back the Gina gasket. It was concluded that pushing back the entire Gina gasket upwards would require a high amount of force. However, affecting only 1/3 of the gasket bottom area would result in the gasket merely being pushed aside while failing to push the soil upwards.