Subsea tunnels, unlike a typical land tunnel, can be vulnerable to seawater intrusion caused by unpredictable high water pressure during construction, which demands a high level of ground reinforcement. A novel technique of artificial ground freezing will be a promising alternative to conventional reinforcement or water-tightening technologies because the frozen ground formation withstands such high water pressure. The artificial ground freezing is carried out by circulating refrigerant such as brine or liquid nitrogen through freeze pipes pre-installed in the target water-bearing ground formation in which pore water becomes ice leading to an impervious frozen area with sufficiently high strength. In case of freezing with brine, there is a disadvantage that the longer freezing time is needed due to the high boiling point. On the other hand, liquid nitrogen, which freezes more quickly than brine, inevitably imposes a limitation that if vaporizes into nitrogen gas and may suffocate the tunnel workers in the isolated and enclosed circumstances such as long-span subsea tunnels.
In this research, considering the characteristics of these refrigerants, liquid air that is safe, capable of rapid freezing and suitable for long-span subsea tunnels is chosen as an alternative refrigerant in the artificial ground freezing. In order to evaluate the feasibility of artificial ground freezing using liquid air, an optimum mixture ratio of liquid oxygen and liquid nitrogen was suggested and the stability of liquid air was verified in accordance with the pressure and flow rate measured from laboratory experiments.
A lab-scale freezing chamber was devised to investigate changes in the thermal and mechanical properties of sandy soil corresponding to the variation of the salinity and water pressure. The freezing time was measured with different conditions during the chamber freezing tests. Further, the thermal conductivity of the frozen soil samples was measured and used as input values for the numerical analysis.
A theoretical model of heat flow for a single freeze pipe and freeze wall was developed to compute the energy and time required for freezing and maintaining frozen status.
The freezing energy and time requirement was calculated by the theoretical model of the heat flow to estimate the total amount of refrigerant required for applying the artificial ground freezing. Then, its validity was evaluated by comparing the results between the freezing chamber experiment and the numerical analysis. In particular, the freezing time showed no significant difference between the theoretical model and the numerical analysis. The amount of refrigerant for artificial ground freezing was estimated from the numerical analysis and the freezing efficiency obtained from the chamber test. In addition, the energy ratio for maintaining frozen status was calculated by the proposed formula.
The required time to freeze a cross passage tunnel and the amount of refrigerant were estimated to provide a preliminary economic analysis through simulating the process of heat transfer during freezing the cross passage tunnel, which is an essential structure in the subsea twin tunneling.
Analyzing the amount of refrigerant associated with the application of artificial ground freezing for constructing a cross passage tunnel in a virtual design, the applicability of rapid artificial ground freezing utilizing liquid air was evaluated. When utilizing liquid air as a refrigerant, the quantity of liquid air would increase compared to the liquid nitrogen. But considering the risks for accidents related to gas leaks of liquid nitrogen and ventilation cost for a long-span tunnel, it is fairly acceptable to use liquid air as a refrigerant for artificial ground freezing.