Simulation of desalination process in dredger fill soil under natural and engineering measures

Abstract: Land reclamation has become an important way to take land from the sea or wetlands, in order to create much more living space in over-crowded coastal areas. However, there is serious soil salinization of hydraulic fill, where most offshore sediments are added into the surrounding seawater with a generally high salt content. The hydraulic fill project also causes the groundwater discharge path to be extended. These impacts have posed a great challenge on the efficiency of investment and the speed of ecological environmental recovery. Therefore, it is very necessary to accurately predict the desalination process, rate and duration for land use planning and ecological environment restoration in reclaiming coastal areas. Taking the coating finishing project in the eastern new area of Taizhou in Eastern China as a research area, a DRAINMOD-S model was extended to the simulation and prediction of natural desalination process in hydraulic reclamation. The hydrological and desalination process in the model were calibrated using the hydrological and salt data from the regional monitoring points after one year of reclamation. Two modes were selected to implement the field test, including the natural desalination and the specific desalination process under comprehensive control measures. In natural desalination, the salt of dredger fill under the rainfall was discharged from the soil along the drainage into the nearby river. In specific desalination, the comprehensive treatments were mainly composed of the water conservancy, supplemented by biological control measures, thereby to speed up the output of salt. The results show that the DRAINMOD-S model better simulated the hydrological process and the salt leaching of hydraulic fill. The variation of groundwater level was basically consistent with the measured value. Specifically, the average deviation was -3.40 cm, while the average absolute deviation was 6.90 cm, and the root mean square error was 8.60 cm and the correlation coefficient was 0.79, as well as, the average deviation of soil salt concentration was -0.08 g/L, while the average absolute deviation was 0.20 g/L, and the root mean square error was 0.24 g/L and the correlation coefficient was 0.91. The soil salt of hydraulic fill was slowly decreasing under the natural rainfall and river drainage, where the annual average desalination rate was 0.45 g/L, while, the duration for the soil to reach the desalination standard was about 27 a. In contrast, the comprehensive treatments greatly shortened the desalination of dredger fill, where the average annual desalination rate increased to 0.73-1.80 g/L, while the desalination time reduced to 3-13 a. As such, the comprehensive treatments can be technically feasible to rapidly leach and drain the salt. In theoretical case, the better leaching can be achieved in the deeper buried depth and the smaller distance between underground pipelines, but the investment cost would be relatively high. Therefore, an optimal strategy can be further selected to balance multiple factors, such as the actual needs of project and investment budgets. The proposed model can be expected to serve the preliminary prediction for the land-building projects, land use management, and ecological environment reconstruction.