Comparison of MgO-GGBS, CaO-GGBS, and cement for construction of two-phase TRD cut-off walls in sand: workability, strength, and permeability

Cement–bentonite–soil cut-off walls installed using the trench remixing and deep (TRD) wall method have been widely used to control seepage. The TRD construction is usually executed in two-phase cut-off walls; however, previous studies mainly focus on the properties of the one-phase cut-off walls. Besides, chemical interactions between bentonite and cement will induce a reduction in the swelling potential and water retention capacity of bentonite, leading to a weakness in its seepage barrier performance. Thus, it is of great significance to develop alternative cementitious materials to replace cement for the construction of cement–bentonite–soil cut-off walls. Ground granulated blastfurnace slag (GGBS) has been increasingly used for partial or full replacement of cement in geotechnical engineering. In this study, the performance of cut-off walls constructed using the two-phase TRD method was fully evaluated based on the use of GGBS with the activators of magnesia (MgO) and quicklime (CaO) to replace cement. A series of laboratory tests were first conducted to evaluate the workability (i.e., flowability and bleeding) of fresh bentonite–sand mixtures with four types of binder slurry (i.e., cement, GGBS, MgO-GGBS, and CaO-GGBS). Subsequently, binder–bentonite–sand mixtures that achieved the required workability were selected for unconfined compressive strength (UCS) and permeability tests. The results indicated that both the flowability and the bleeding water of fresh mixtures were remarkably increased by the addition of cement or CaO-GGBS compared to that of GGBS or MgO-GGBS, easily exceeding the acceptable range. It may be attributed to the cation exchange of Ca 2+ released from cement or CaO hydration with Na + in bentonite; however, this interaction might be concealed for the one-phase cut-off walls. More importantly, MgO-GGBS showed much better performance in the strength and permeability than cement; for example, the average 28-day UCS could reach up to 1.12 MPa for the binder type of MgO/GGBS = 1:9 at the binder content of 10% and bentonite content of 3%, four times that of cement. On the other hand, it is revealed that the addition of a large amount of bentonite does not always make a significant improvement on the permeability of two-phase cut-off walls. Overall, the results promote the application of MgO-GGBS as a potential substitute for cement in cut-off walls using the two-phase TRD method.