Ground condition

The soft ground in this area consists of peat soil (6–7 m thick) where natural water content, w_n = 300– 800% and unit weight, γ_t = 10–11 kN/m³ overlaid on a sand layer of 5–6 m thick and soft clay of 14–15 m thick. Natural water content, w_n = 70–100% and unit weight, γ_t = 14–16 kN/m³.

Construction condition

The size of the vacuum consolidation area was 45 m × 30 m. Vertical drains were installed up to 26 m deep in a square grid pattern at intervals of 0.8 m × 0.8 m. A highway embankment of about 6.1 m high was constructed at a rate of 14.1 cm/day.

Field monitoring data

Figure 10.2(a) shows the field monitoring data. The vacuum pressure measured at the pump was about 85–92 kPa during the vacuum operation. The vacuum pressures measured under the airtight sheet were low at the beginning of operation and gradually increased to 50–60 kPa prior to the start of embankment operation, as shown in Fig. 10.2(c). The consolidation settlement induced by vacuum pressure and embankment load was about 241 cm.

It was found that decreasing the vacuum pressure somehow correlates with the quantity of ground settlement. Figure 10.2(d) shows the vacuum pressure difference between at the pump and under the airtight sheet. The change of pressure in the elevation head calculated by settlement data was plotted as a comparison in Fig. 10.2(d). It can be seen that the amount of vacuum pressure difference between at the pump and under the airtight sheet was larger than the amount of pressure change due to elevation head loss. This is probably due to some leakage of vacuum pressure through the intermediate sand layer.

Ground condition

The soft ground in this area consists of peat soil (− 4 m thick) where the average natural water content, w_n = 600% and unit weight, γ_t = 10–11 kN/m³, followed by soft clay (3–6 m thick) where the average natural water content, w_n = 85% and unit weight, γ_t = 15 kN/m³.

Construction condition

The size of the vacuum consolidation area was 60 m × 60 m. Embankment thickness was about 13.0 m high and was constructed at a rate of 21 cm/day.

Field monitoring data

Figure 10.2(b) shows the field monitoring data of Noshiro. The vacuum pressure measured at the pump was about 68–88 kPa during the vacuum operation. The average vacuum pressure measured under the airtight sheets was about 60 kPa at the beginning of operation. The vacuum pressure at the pump slightly decreased during embankment backfill up to 6 m and then gradually returned to more than 80 kPa. The consolidation induced by the vacuum pressure and embankment load was about 237 cm. Figures 10.2 (b) and (d) show the vacuum pressure difference between at the pump and under the airtight sheet and elevation head loss by the consolidation settlement. It can be clearly seen that the vacuum pressure differences and elevation head loss were closer than in the Kushiro case. It could be that the loss of vacuum pressure at this site was lower than the previous one because there was no intermediated sand layer.

According to selected case histories where the conventional vacuum consolidation system was applied in Japan at different locations and under different ground conditions. It can be seen that the decreasing of vacuum pressure is caused by elevation head difference during the consolidation settlement under the vacuum consolidation systems shown in Fig. 10.1. The loss of vacuum pressure means the loss of applied load. This problem affects the quality of the ground improvement. The common solution is to put additional surcharge loading on the improved area, however, this results in an increase in construction cost and time.

Vacuum operation

Vacuum pressure measured under the airtight sheet was maintained constantly at about 80 kPa along to the consolidation settlement between October 2008 and December 2009. The water discharge at the beginning of vacuum pumping was about 400 m³/day and gradually decreased during consolidation settlement to about 200 m³/day.

Surface settlement

Settlement of about 1 m was induced by vacuum consolidation prior to embankment construction. The maximum surface settlement at the termination of the vacuum operation in December 2009 was 10.6 m. Ground upheaval in the nearby area was relatively small.

Differential settlement

Large settlements were observed in the Ac1 and Apt1 layers, about 5.0 m, followed by the Ap3 layer, about 3.0 m.

Construction rate

The average embankment construction rate at the start was 3 cm/day as per the original design criteria. However, it was increased to 8 cm/day after embankment thickness was about 9 m. The total construction period was significantly shorter by more than a year compared to the trial embankment.

Excess pore water pressure

Maximum excess pore water pressure induced by the embankment load was about 50 and − 25 kPa at GL − 7.0 m (Apt1) and GL − 17.0 (Apt3) m respectively. Additionally, the excess pore water pressure at GL − 30 m (a malfunction in the middle period) was also effectively reduced by vacuum operation. Note that only PVDs for the CVC installed up to 20 m deep were directly connected to the horizontal drain and vacuum system, while the existing PVDs, installed up to 34 m deep, were not directly connected to the horizontal drain. In this case, it can be assumed that the vacuum pressure was also distributed in the sand mat passing through the existing PVD.

Correction of measured pore water pressure

The reading of pore water pressure directly from the piezometer is required to be corrected by the location of piezometer after consolidation settlement to obtain the correct value. The data from differential settlement gauges installed at the same depth as the piezometers were used for the correction.

A corrected pore water pressure profile is shown in Fig. 10.11. It can be clearly seen that that the excess pore water pressure at piezometer at GL − 7 m was lower than the hydrostatic pressure in the beginning and was a bit higher than hydrostatic pressure during the completion of the embankment construction. In contrast, the pore water pressure at GL − 17 m was always lower than the hydrostatic line after the correction.

Therefore, the stability of the embankment was evaluated to be more stable than the uncorrected case. Moreover, the pore water pressure induced by embankment load was more clearly identified. The maximum excess pore water pressure of GL − 7 m at the finished level of embankment was about 20 kPa and then decreased to − 54 kPa at the termination of the vacuum operation.

For the piezometer at GL − 17, the pore water pressure was reduced to − 60 kPa before subjected to embankment load, and became about − 48 kPa at the termination of the vacuum operation. After that, the pore water pressure returned to its hydrostatic condition.

Lateral displacement

In the initial period, the vacuum consolidation induced the inward lateral displacement by about 20 cm. After the ground was subjected to the embankment load, outward movement was induced by a maximum of about 1 m. Note that the maximum lateral displacement was about 10 m deep where the tip of the sheet piles were located.

Ground deformation shape

Figure 10.12 shows the soil profile of the improvement area after termination of the vacuum operation and borehole locations. A total of nine boreholes were carried out along the cross section of the embankment. Of concern was that the large differential settlement might be observed at the boundary of the improvement area and adjacent area. However, after the inspection of the airtight sheet elevations, it was found that the settlement surface along the cross section was a smooth curve shape, probably due to the restraint effect of the reinforced layer constructed underneath the embankment.

Moreover, the surface elevation of the counterweight berm at the left side of the embankment was close to the adjacent paddy field. It was found that the settlement under the counterweight berm was about 5 m. Also, the ground surface at the counterweight berm on the right side that was leveled at the beginning of construction was upheaved after the end of construction.

Standard penetration test

Figure 10.13 shows the SPT-N value profile both before and after improvement. Note that the design depth of improvement by vacuum consolidation is up to an elevation of − 20 m deep. It can be seen that SPT-N value was changed from 0–2 to 10–20 after improvement. Moreover, SPT-N value of the soil layer under an elevation of − 20 m to − 34 m was also changed to 10–20 and only slightly changed below 34 m. In addition, no significant difference of SPT-N value under the center of the embankment and side slope could be observed.

Natural water content

The natural water content was significantly decreased in the range of the vacuum consolation zone, as shown in Fig. 10.14. The maximum natural water content prior to the construction was about 400% and became about 200% after improvement. Unlike the increase of the SPT-N value from elevations of − 20 m to − 34 m, the reduction of natural water content in this zone is not clear. Also, natural water content both under the center of embankment and side slope are similar.

Undrained shear strength

Figure 10.15 illustrates the undrained shear strength from the Triaxial unconsolidated undrained test. Also, the theoretical lines are plotted based on the expression in Eq. (10.2). The solid line and dotted line represents the undrained shear strength under the center of embankment and side slope, respectively.

$S_u=\left(P_0+\mathrm{\Delta }{P}_{emb}+\mathrm{\Delta }{P}_v\right)\times m$

where

S_u = undrained shear strength

P₀ = initial effective stress

ΔP_emb = total stress induced by embankment load

ΔP_v = total stress induced by vacuum load

The total unit weight of embankment (γ_t) obtained from the field = 21.5 kN/m³ and the effective unit weight (γ′) calculated by the Boussinesq equation = 13.1 kN/m³. These values were used to calculate ΔP_emb. Moreover, vacuum loads were obtained from the reduction of pore water pressure. Here, the values of ΔP_v above and below the As3 layer were assumed to be 53.6 kPa and 48.2 kPa, respectively. It can be seen that the theoretical calculations are good representations of the undrained shear strength in the range of the vacuum consolidation zone, but higher than the field data lower than 20 m where the vacuum consolidation system was not applied. It can be considered the primary consolidation mostly completed in vacuum consolidation zone but quite slower outside the vacuum consolidation zone. Note that the degree of consolidation was assumed to be 100 in all calculation value.

Consolidation yield stress

The consolidation yield stress obtained from the oedometer tests are plotted, as shown earlier in Fig. 10.16, with the theoretical line calculated by Eq. (10.3).

$P=P_0+\mathrm{\Delta }{P}_{emb}+\mathrm{\Delta }{P}_v$

Unlike the undrained shear strength, consolidation yield stress by oedometer tests are mostly higher than the theoretical calculation, especially in the zone improved by vacuum consolidation. It can be seen that the difference in consolidation yield stress between the center of embankment and side slope is not clear.